Journal of General Virology (1992), 73, 3185-3194.
3185
Printed in Great Britain
Equilibrium and kinetic analysis of Autographa californica nuclear
polyhedrosis virus attachment to different insect cell lines
T. J. Wickham,lt M. L. Shuler, 1 D. A. Hammer, 1 R. R. Granados 2 and H. A. Wood 2.
1School of Chemical Engineering and 2Boyce Thompson Institute for Plant Research, Cornell University,
lthaca, New York 14853, U.S.A.
The kinetic and equilibrium attachment of Autographa
californica nuclear polyhedrosis virus (AcMNPV) to
seven insect cell lines was evaluated. Kinetic experiments revealed differences of up to 10-fold in the
infection rates among cell lines. Equilibrium binding
also varied between cell lines and was saturable. The
Tn 5B1-4 and Tn F cell lines had the highest virus
binding affinities and infection rates and exhibited
diffusion-limited attachment. The rate of infection
appears to be limited by the rate of attachment. For the
Tn 5B1-4 cells the physical to infective particle ratio
for AcMNPV was 5-3. From the Scatchard analyses,
the cell lines Tn 5BI-4 and Tn F displayed affinities of
2-35x 10 L° M-1 and 1-60x 101° M-1, respectively, with
6000 and 13 700 binding sites per cell. The insect cell
line Hz 1075, which is not susceptible to A c M N P V
infection, displayed a much lower, but saturable,
binding of A c M N P V with 900 sites/cell and an affinity
of 1-1xl01° M-1. Unlabelled AcMNPV, but not
Lymantria dispar M N P V could compete with labelled
A c M N P V for binding sites. There were 93 to 96%
reductions in virus cell binding following pretreatments
of cells with three proteases, suggesting the involvement of a cellular protein component in virus binding.
Tunicamycin, an inhibitor of N-linked glycosylation
and expression of some membrane proteins on the cell
surface, reduced virus binding in a dose-dependent
manner suggesting a role for glycoprotein(s) in binding.
However there was no evidence for the direct involvement of oligosaccharides in attachment. Metabolic
inhibitors of oligosaccharide trimming and competition binding assays using simple sugars caused no
measurable reductions in virus binding. These findings
suggest that A c M N P V attachment to insect cells is
receptor-mediated via a glycoprotein component(s);
the direct involvement of oligosaccharide moieties in
binding is unlikely.
Introduction
the infection process and is responsible for the secondary
infection of cells in vivo. In tissue culture, the nonoccluded virus (NOV) is 1500 times more infectious than
the OV (Volkman et al., 1976). The majority of the NOVs
have been observed to enter cells in vivo by fusion at the
cell surface (Wang & Kelly, 1985). However the
infectious entry route into tissue culture cells has been
shown to be primarily through receptor-mediated endocytosis; a minority appear to enter through some other
route, possibly by fusion at the cell surface (Volkman et
al., 1986). The 64K glycoprotein present in large
amounts in the NOV envelope is involved in fusion and
appears to be activated below a pH of approximately 5.5
(Volkman et al., 1986). The NOV attachment and entry
kinetics of a closely related virus, T. ni MNPV, have
been measured (Wang & Kelly, 1985); however attachment, the earliest step in the AcMNPV NOV infection
process, has not been studied in detail. The existence of
receptor-mediated attachment, the type of receptor, and
the physical parameters such as the number and affinity
The baculovirus Autographa californica nuclear polyhedrosis virus (AcMNPV) is used widely as a vector for the
expression of heterologous proteins in insect cells
(Luckow, 1990). It is a rod-shaped virus with a doublestranded, circular D N A genome of 125 kb. AcMNPV is
capable of infecting a number of cell lines, including ones
from Spodoptera frugiperda (IPLB-21AE) and Trichoplusia ni (Tn-368). There are two forms of the virus. The first
is occluded within crystallized protein polyhedra that are
formed in the nucleus late in infection. The occluded
virus (OV) is liberated from polyhedra following ingestion by insect larvae and is responsible for the primary
infection of insect midgut cells. The second form of the
virus is non-occluded. It buds from infected cells early in
t Present address: The Scripps Research Institute, 10666 N. Torrey
Pines Road, La Jolla, California 92037, U.S.A.
0001-0647 © 1992 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
3186
T. J. W i c k h a m
and others
of virus-binding sites involved in attachment, have not
been investigated for AcMNPV.
For most viruses, attachment studies have involved
the identification of the viral receptor because its
identity and expression are often key to determining
tissue tropism and host range. Although receptor
expression is not the sole determinant of susceptibility,
infection is initiated through receptor-mediated binding
which thus represents a necessary step in the infection
process. Specific cellular proteins have been identified as
viral receptors, such as the CD4 receptor for human
immunodeficiency virus (HIV) (Dalgleish et al., 1984;
Klatzmann et al., 1984). However others utilize more
ubiquitous cell surface molecules, such as the sialic acid
residues bound by influenza virus (Paulson et al., 1986).
Numerous studies on virus-cell interaction have used
Scatchard analysis to examine the virus binding to cells
to obtain parameters such as the virus-cell affinity and
the number of viral sites on the cell (Alcami et al., 1989;
Basak & Compans, 1989; Epstein et al., 1984; Fries &
Helenius, 1979; Roberts, 1989; Schlegel et al., 1982;
Schwarz & Datema, 1984; Taylor & Cooper, 1989;
Schlegel et al., 1982; Schwarz & Datema, 1984; Taylor &
Cooper, 1989; Trimble & Maley, 1984; Verdin et al.,
1989;
Volkman
e t al.,
1986;
Wunner
e t al.,
1984).
Additionally, a theoretical study of virus attachment has
investigated the different possible interpretations of
viral Scatchard analyses with respect to the viral site
number obtained and the relationship of the overall virus
affinity to the individual receptor affinity (Wickham et
al., 1990). Metabolic inhibitors, receptor analogues, and
receptor-cleaving enzymes have also been used to
identify and determine the receptor for a particular virus
( M e t t e n l e i t e r e t al., 1 9 9 0 ; P a u l s o n e t al., 1 9 8 6 ; W u D u n n
& S p e a r , 1989).
Here the physical parameters involved in the kinetic
and equilibrium attachment of the non-occluded form of
A c M N P V t o i n s e c t cell l i n e s a r e e x a m i n e d . E x p e r i m e n t s
designed to determine the existence and nature of an
A c M N P V r e c e p t o r , w h e t h e r it is a s p e c i f i c p r o t e i n o r a
more common oligosaccharide component, are reported.
Methods
Cells, media and virus. The following cell lines were used. T. ni BTITN-F (Tn F) and BTI-TN-MGI (Tn M) are novel cell lines (R. R.
Granados et al., unpublished results) established from T. ni fat body
and midgut tissues, respectively. Another novel cell line, BTI-TN-5B 14 (Tn 5B1-4; R. R. Granados, unpublished results) was established
from T. ni eggs. BTI-EA-88 (Ea 88; R. R. Granados et al., unpublished
results) is an attached cell line derived from Estigmene acrea BTI-EAA
(Granados & Naughton, 1976). Other cell lines used were TN-368
(Tn 368; Hink, 1970), IPLB-SF21AE (Sf-21 ; Vaughn et al., 1977), Sf-9
(Summers & Smith, 1987) and Heliothis zea 1075 (Hz 1075; Goodwin et
al., 1982). All cells were grown in TNM-FH medium [Grace's insect
cell medium (Gibco) plus lactalbumin and yeastolate] supplemented
with 10~ foetal bovine serum (FBS; Hyclone) (complete medium).
Fourth tissue culture passage AcMNPV clone 1A (Wood, 1980) was
used in all binding experiments. The G isolate of Lymantria dispar
multiple nuclear polyhedrosis virus (LdMNPV) was also used in
competition experiments (Smith et al., 1988).
Growth o f labelled and unlabelled virus particles. Unlabelled virus was
obtained by inoculating Sf-21 cells at an m.o.i, of 10 p.f.u./cell for 1 h.
The inoculum was then removed, replaced with fresh medium, and the
cells were incubated at 28 °C. AcMNPV NOV was harvested at 2 days
post-infection (p.i.) LdMNPV NOV was harvested at 8 days p.i.
For 3zp-labelled virus, ceils were infected as described above except
that the inoculum was replaced by Grace's medium minus phosphate
containing 10% dialysed FBS (Hyclone) and 100 p_Ci/ml of 32pi
(NEN). Cells were then incubated at 28 °C and at 6 h p.i. 0-I volume of
complete medium was added to replenish phosphate. Virus was
harvested from the culture supernatant at 2 days p.i., which is before
cell lysis normally occurs.
Both unlabelled and labelled virus was concentrated and purified by
centrifugation at 500 g for 15 min to remove ceils. The supernatant was
then centrifuged at 50 000 g through a 35 ~ (w/w) sucrose cushion. The
virus pellet was allowed to resuspend overnight in Grace's medium.
Negative staining of the resuspended pellet showed that over 9 0 ~ of
the virus particles were intact. The radioactive virus activity was
typically 10-5 c.p.m./virus particle.
Measurement o f virus particle and infectious particle concentration.
Unlabelled virus particle concentration was determined by measuring
DNA concentration. Purified virus was digested with 2mg/ml
proteinase K (Boehringer-Mannheim) overnight at 55 °C to release
viral DNA. Digested samples were added in 1 to 100 ktl volumes to 2.0
ml of TNE buffer pH 7.4 containing 100 ng/ml of Hoechst 33258 dye
(Polysciences). Background from free DNA or broken viral particles
measured before proteinase K treatment was always less than 10~ of
the total. Phage 2 DNA at a known concentration was used as a
standard. Using calf thymus or Escherichia coli DNA also gave the
same standard curve as the 2 DNA. The virus particle concentration
was calculated from the M r of the viral genome and the measured
AcMNPV DNA concentration using a TKO 100 fluorometer (Hoefer
Scientific) (Labarca & Paigen, 1980). The concentration of infectious
virus was determined by a centrifugation plaque assay (Wood, 1977)
using Tn 5B 1-4 cells (described below under measurement of infection
kinetics). The ratio of physical to infectious particles could then be
calculated.
Measurement o f attachment kinetics. Radioactive virus particles in
Grace's medium plus 10% FBS were added to confluent Tn 5B1-4 cells
(2 x 106/35 mm well) at 4 °C in a volume of 1.0 ml. The viruscontaining medium was then removed from each sample at various
times up to 10 h. The ceils were washed twice with Grace's medium and
then removed by scraping to measure the number of bound TCAprecipitable counts.
Measurement o f infection kinetics. Tn 5B1-4 cells were added to 35
mm wells at l-5 x 106 cells/well and allowed to become firmly attached.
This is an optimum density of Tn 5B1-4 cells, selected to be low enough
to support high amounts of polyhedra (Wood et al., 1982), but
sufficiently high for cells to reach confluence, preventing the formation
of overly large plaques. This density gave rise to small plaques with
large numbers of polyhedra which could be detected easily using phasecontrast microscopy. Virus stock was diluted to 5500 p.f.u./ml in
complete TNM-FH medium and 1.5 ml was added to duplicate wells
and incubated at 22 °C. The inoculum then was removed, the cells were
washed twice with Grace's medium, and overlaid with 1.5~ agarose
(FMC Bioproducts) in TNM-FH medium containing 5 ~ FBS.
Infected foci were counted under a phase-contrast microscope
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
A c M N P V attachment to insect cell lines
equipped with a grid reticle to determine the number of infected cells at
each time point. For the other cell lines, 106 cells were added per well in
six-multiwell tissue culture plates. Virus stock diluted to 5000 p.f.u./ml
in Grace's medium plus 10 ~ FBS was added (1.0 ml) to the wells for 15
min after which the medium was removed and the cells were washed
twice with Grace's medium. The cells were overlaid and the number of
plaques determined in the same way as for Tn 5B1-4 cells. The time
course of Tn 5B 1-4 infection showed that 15 min was a sufficient time
in which to determine the initial infection rate. Fifteen min was used
for the other cell lines without concern about large decreases in rate due
to depletion of infectious virus in the medium because their infection
kinetics were all slower than Tn 5B1-4 cells. In measuring the effect of
cell concentration on the number of infectious foci produced, 0-5, 0.75,
1.0 or 1.5 x 106 cells were added per 35 mm well, exposed to 500
p.f.u./ml in 1.0 ml for 15 min and the resulting infectious foci were then
counted under a microscope. Cell density was kept below 1.5 x 106 cells
to minimize cell density effects in the plaque assay, as already noted
(Wood et al., 1982).
Equilibrium binding studies. Cells were added to Falcon 24-multiwell
plates (Becton Dickinson). The cells were then allowed to attach and
were centrifuged at 2000 g for 15 min to strengthen their attachment.
The medium in pairs of wells was then successively removed and
replaced with 0.2 ml of Grace's medium at 4 °C containing a known
virus concentration and c.p.m, per particle. (Inclusion of 10~ FBS or
1% albumin in the medium had no effect on virus binding and these
were not included in the binding medium.) The cells and virus were
then incubated at 4 °C for 18 h to reach equilibrium. Following the
incubation, the medium was removed and the cells were washed twice
with Grace's medium to remove unbound virus. The cells were then
removed and the TCA-precipitable radioactivity could be measured to
determine the number of bound viruses per cell. Non-specific binding
was determined by measuring the percentage of bound c.p.m, in the
presence of a 20-fold excess of unlabelled virus.
Competition studies
with unlabelled baculoviruses. Unlabelled
LdMNPV and AcMNPV preparations were added to 4 x 105 Tn 5B1-4
or Hz 1075 cells in 15 mm tissue culture wells at 4 °C. Ten rain later a
small amount of 32p-labelled AcMNPV was added to the wells. The
contents of the wells were mixed gently and incubated at 4 °C for 2 h.
The cells were then washed twice with PBS and the cell-bound TCAprecipitable c.p.m, was measured.
Pretreatment of cells with proteases. Three proteolytic enzymes were
used to pretreat cells, followed by extensive washing of the cells and
measurement of radiolabelled virus binding: proteinase K (BoehringerMannheim), 1.0 mg/ml, pH 7.0, 25 °C; trypsin (Boehringer-Mannhelm), 0.5 mg/ml, 25 °C; chymotrypsin (Boehringer-Mannheim) 1-0
mg/ml, 25 °C. Cells were pretreated with the proteases in PBS at pH 7.0
for 1 h and then washed five times with PBS containing protease
inhibitors [ 1 mM-PMSF (Sigma) and 5 mg/ml trypsin inhibitor (Sigma)]
to remove residual activity. None of the enzyme treatments affected
cell viability as measured using trypan blue and over 95 % of the cells
remained attached to the well following protease incubation and
washing. Radioactive virus in Grace's medium plus 10~ FBS was
added to each well and allowed to bind for 2 h at 4 °C. The cells were
then washed twice to remove unbound virus, and the cell-bound c.p.m.
was determined. Controls for all the enzymes were exposed to the same
conditions as the enzyme-containing samples.
Pretreatment of cells with glycosylation inhibitiors. The effect of
inhibitors of N-linked glycosylation was tested by growing cells in the
presence of 10 ~ttd-swainsonine (Boehringer-Mannheim), 4 mMdeoxynojirimycin (dNM) (Boehringer-Mannheim), or 0-2, 1.0 or 4-0
~tg/ml tunicamycin (TM) (Boehringer-Mannheim) for 3 days and then
comparing virus binding with cells grown in the absence of inhibitors.
SW, dNM, and 0-2 and 1.0 ktg/ml TM had no effect on the doubling
3187
time or viability of the cells compared to the control. Cells grown in 4.0
ktg/ml TM were slightly vacuolated and had a longer doubling time than
control cells although viability was not affected. Equivalent amounts of
TM, SW and dNM-treated cells along with control cells were added to
Falcon 24-multiwell dishes and allowed to attach. The medium was
then removed and replaced with radioactive virus-containing medium
(Grace's plus 10~ FBS) which was allowed to bind for 2 h at 4 °C. After
washing of the cells, the amount of cell-associated virus was measured.
Monosaccharide competition experiments. For sugar competition
experiments 8 to 9 ~ (w/w) solutions of the competing sugars were
prepared in 10 mi-Tris-HC1 pH 7.5, so that a final osmolarity of 380 to
410 mosM was obtained. Glucose, mannose, galactose, N-acetylglucosamine and fucose (all obtained from Sigma) were used since all are
incorporated into oligosaccharide structures. Xylose and fructose,
which are not found in animal oligosaccharide structures, were used as
controls. Radioactive virus was pelleted and resuspended in 10 mMTris-HC1 pH 7.5. Ten ktl of the virus solution was added to 200 ~tl of the
sugar solution which was then allowed to bind for 2 h at 4 °C. After
washing of the cells, the amount of cell-associated virus was measured.
Electron microscopy. A confluent monolayer of Tn 5B1-4 cells was
incubated with virus at a concentration of 1011 virus particles/ml at
4 °C overnight. The cells were then washed twice with Grace's medium
and fixed with 4 ~ glutaraldehyde at room temperature. The fixed cells
were stained with 1% OsO, in 0-1 M-sodium cacodylate buffer pH 7,
and washed with distilled water. The sample was then dehydrated in
ethanol and propylene oxide and embedded in epoxy plastic. Samples
were then thin-sectioned. Negative staining of virus samples was
accomplished by staining with 2% aqueous phosphotungstic acid, pH
7.2.
Results
Measurement of physical to infectious particle ratio
The physical to infectious particle ratio for AcMNPV
when measured with Tn 5BI-4 cells was 5.3 and
corresponded to 19~ of the particles being infectious.
This ratio is lower than in previous reports (Volkman et
al., 1986; Wang & Kelly, 1985); the differences are
probably due to differences in assay methodology (see
Discussion). Negative staining of AcMNPV NOV
following purification and concentration showed that
over 90 ~ of the virus particles were intact with complete
viral envelopes.
Infection rates of different cell lines
The infection rates of cell lines were determined by
exposing attached cells to a dilute concentration of virus
for various times, followed by washing the cells and
overlaying with agarose. The resulting number of
plaques then corresponded to the number of cells that
had been infected during a given exposure time to the
virus. The infection kinetics of Tn 5B1-4 cells were
studied in detail because these had the highest infection
rates. The time course of infection of these cells is shown
in Fig. 1. The rate remained nearly constant for the first
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
3188
T. J. Wickham and others
3000
Table 1. Infection rate constants and percentage of virus
bound at equilibrium measured for various cell lines
2000
E 1000
0
0
30
60
90
Time (min)
120
Fig. 1. Infection kinetics of Tn 5B1-4 cells by A c M N P V . A virus
inoculum was diluted in T N M - F H medium plus 10~ FBS to 5500
p.f.u./ml. Tn 5B1-4 cells (1.5 x 106) were inoculated with 1.5 ml of
diluted virus and incubated at room temperature. At various times the
inoculum was removed and the cells were washed twice with Grace's
medium. The cells were then overlaid with agarose and incubated for 3
to 4 days, at which time plaques were counted under a microscope.
hour of exposure and then started to decline as the
remaining virus in the inoculum became depleted. The
rate measured at 15 min was taken as the initial infection
rate and was found to depend linearly on cell concentration (r 2 =0-99) in a range between 0.5 × 106 and
1-5 x 106 cells/35 mm well. There was also a linear
dependence on virus concentration (r 2 = 0-97) under the
condition of dilute virus concentration ( < < I p.f.u./cell)
that was used for these measurements (data not shown).
Thus the infection rate could be expressed in the form
r i = kiCV
where r i is the infection rate (infections/
ml. min), C is the cell concentration (cells/ml; varied by
changing the number of cells per well), V is the virus
concentration (p.f.u./ml; varied by changing the dilution), and k i is the infection rate constant (ml/cell. min),
which depends on the cell line. This expression for the
infection rate is a more general form than originally
given by de Gooijer et al. (1989), who assumed viral
infection was first order with cell concentration when
non-occluded infectious virus was in excess. When virus
is not in excess, the infection rate also depends first order
on viral concentration. Based on this dependence the
infection rate constant, ki, could be calculated from the
cell concentration, the initial infection rate and the virus
concentration. The initial infection rates and infection
rate constants of the other cell lines were similarly
measured to determine which cell lines were most
susceptible to A c M N P V (Table 1). Tn 5B1-4 and Tn F
cells showed the highest infection rates (ki = 4.4 x 10 -9
cell/ml, min) compared to the cell lines Sf-21, Sf-9 and Tn
368 which showed infection rates five- to 10-fold lower.
Detailed comparison between our measured infection
rates and those determined by de Gooijer et al. (1989) is
not possible without detailed knowledge of viral concentrations in their experiments.
Cell line
Infection rate constant*
(ml/cell.min X 10-9)
Binding at 4 °C:~
(%)
Sf-9
Sf-21
"In 368
Tn M
Ea 88
Tn F
Tn 5B1-4
0-44 (+ 0.9)
0.48 (+ 0-6)
1-65 (+ 0.1)
2.02 (+0.3)
2-27 (_+ 0.1)
4.35 (_+0-3)
4.40 (_+ 0.3)
14 (+ 1)
18 (+ 1)
22 (+ 5)
34 (_+2)
34 (+4)
50 (_+3)
51 (_+ 2)
* Measured using the number of plaques formed at 5 days after a
15 min incubation at 22 °C with 5000 p.f.u./ml and 106 cells/ml.
t Measured using TCA-precipitible c.p.m, following incubation o f
32p.labelled virus with 5 x 105 cells/15 m m tissue culture well (0-2 ml).
Attachment rates to Tn 5B1-4 cells
The attachment kinetics of A c M N P V to Tn 5BI-4 cells
were examined in detail. The attachment rate differs
from the infection rate; clearly attachment is a prerequisite for infection, and we expect the infection rate to be
slower than the attachment rate. The same linear
dependence on cell concentration and virus concentration were found for the initial attachment rate as was
found for the infection rate. This dependence held only at
initial times since the binding would approach equilibrium as time progressed. The initial attachment rate could
be expressed as ra = k, C V , where r, is the attachment
rate (c.p.m./ml.min), C is the cell concentration
(cells/ml), V is the radiolabelled virus concentration
(c.p.m./ml), and ka is the attachment rate constant
(ml/cell. min) which has the same units as the infection
rate constant, ki. From the initial slope of the line in Fig.
2 k, for Tn 5B1-4 cells was 5-2 x 10-9 ml/cell.min,
satisfying the requirement that since attachment precedes infection, k a > ki. This value was very similar to the
measured kl, which was 4-4 × 10 -9 ml/cell.min. Since k i
is only slightly greater than ka, this suggests attachment is
the rate-controlling step for infection, and that postattachment steps occur quickly on the time-scale of
attachment. Rates of attachment and infection for HIV
also appear to be related similarly, which has led to the
development of a mathematical model in which infection
kinetics are adequately treated by attachment kinetics
(Layne et al., 1989).
Calculation of maximum attachment rate constant based
on diffusion-limited attachment
The maximum possible attachment rate represents the
case where attachment is diffusion-limited and every
virus collision with the cell surface results in attachment.
The attachment rate constant ka(max) in this case can be
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
A c M N P V attachment to insect cell lines
60
i
I
0
o
i
i
3189
1400
i
50
o
~" 40
1200
30
0
1000
20
800
10
O
d~
I
100
I
I
I
600
I
200 300 400
Time (rain).
500
600
Fig. 2. Attachment kinetics of AcMNPV to Tn 5B1-4 cells at 4 °C.
Radiolabelled virus particles (18500 c.p.m.) were added to 2-0 x 106Tn
5B1-4 cells in a 35 mm tissue culture plate in a total volume of 1.0 ml.
Binding came to equilibrium by 10 h.
simply calculated from the virus diffusion coefficient Dv
and the radius of the cell a such that katmax~= 4aDv for an
attached cell (Smoluchowski, 1917; Shoup & Szabo,
1982). By calculating this constant it is possible to set a
m a x i m u m limit on the attachment rate of the virus and
also to determine whether the binding of the virus to the
cell is diffusion-limited. An attachment rate constant
that is close to the m a x i m u m calculated rate indicates
diffusion-limited attachment. Using the Stokes-Einstein
relation and accounting for the rod shape of the virus
(Atkins, 1982), the diffusion coefficient for A c M N P V
was calculated to be 2.0 x 10 -8 cm2/s at 22 °C. The radius
of an attached cell is 1 3 ( + 2)I.tm so that the m a x i m u m
possible attachment rate constant was calculated by the
above equation to be 6-2 x 10 -9 cm3/cell.min. Tn 5B1-4
cells had the highest attachment and infection rate
constants which were 5 . 2 x 10 -9 cm3/ceU.min and
4-4 × 10 -9 cm3/cell.min, respectively, indicating diffusion-limited attachment of the virus to these cells. Since
the rate constant for infection is very close to the rate
constant for attachment, this suggests the rate of
infection of these cells by A c M N P V is limited by the
diffusion of the virus to the cell surface. The infection
rate constant for the Tn F cell line shows that it is also
close to the diffusion-limited rate.
Equilibrium binding at 4 °C
The amount of radioactive virus that bound overnight to
cells at 4 °C was measured for each cell line. Large
differences in binding were observed between the cell
lines as had been observed for their respective infection
rates. Tn 5BI-4 cells displayed the highest binding
whereas Sf-9 cells displayed the lowest (Table 1). A
comparison of the equilibrium binding values with the
infection rates for each cell line showed a correlation
between affinity of the cell line for the virus and its
.o 400
,¢
200
0
I
0
I
i
I
30
60
90
120
Unlabelled particles/cell (x 10-3)
150
Fig. 3. Competition of unlabelled LdMNPV (I-1)and AcMNPV
with labelled AcMNPV for binding to Tn 5B1-4 cells. Cells
(4 × 10s/15mm well) were incubated at 4 °C with 2500 c.p.m, of 32p.
labelled AcMNPV and various concentrations ofunlabelled AcMNPV
or LdMNPV for 2 h. Points represent the average of triplicates.
infection rate (Table 1). Sf-9 and Sf-21 cell lines
displayed both the lowest binding and infection rates
whereas Tn F and Tn 5B1-4 cells both displayed the
highest binding and infection rates.
Competition of unlabelled baculoviruses with labelled
A c M N P V for binding
To test the specificity of binding of A c M N P V , competition experiments were performed using progressively
higher concentrations of unlabelled A c M N P V and
another nuclear polyhedrosis virus, L d M N P V . Unlabelled A c M N P V was able to compete with labelled
A c M N P V for attachment to Tn 5B1-4 cells, whereas
L d M N P V at the same concentrations as A c M N P V
could not compete with labelled A c M N P V (Fig. 3).
Viral Scatchard analysis for different insect cell lines
The equilibrium binding of A c M N P V to different
confluent insect cell lines was measured using standard
Scatchard analysis to compare their relative viral
affinities and number of saturable sites. Total binding
was measured by adding progressively higher concentrations of unlabelled virus along with labelled virus. For
selected virus concentrations, 20-fold amounts of excess
unlabelled virus was added and the percentage of
radiolabelled virus that continued to bind in the presence
of this excess virus was measured. Specific binding was
assessed by subtracting the binding values obtained in
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
3190
T. J. Wickham and others
14
i
L
I
1.0
'
- .-'°°'t
12
~
°° ° ° ° °
, l0
j...-
i
0.8
°.
°-
. ....:
Pi
i
0.6
°
~0-4
.o°,'
' .° - - - •
0.2
m
2
,'"
0'
0.0
."~"
I
0
,
t
20
40
60
Particles added/cell (x 10 3)
,
80
Fig. 4. Saturable binding of AcMNPV to Tn 5B1-4 (A) and Hz 1075
ceils (O). Tn 5B1-4 or Hz 1075 cells (6-5 x 10s/15 mm well) were
incubated with virus particles at various concentrations (in 0.2 ml) for
18 h at 4 °C. The amount of TCA-precipitable bound c.p.m, was
determined to measure total binding (dotted lines). Non-specific
binding for both cell lines was 9 ~ of the total amount of virus added
and was determined by the amount of binding that occurred with the
addition of a 20-fold excess of particles to selected samples. Specific
binding (solid lines) was obtained by subtracting the non-specific
binding from the total binding. Points represent the average of
duplicates.
the presence of 20-fold excess from the binding values in
the absence of excess virus (Schlegel et al., 1982; Taylor
& Cooper, 1989). The percentage of radiolabelled
particles that continued to bind in the presence of excess
virus represents non-specific binding. Comparison of the
total and specific binding showed that under nonsaturating conditions specific binding of the virus
accounted for 75~ of the total binding for Tn 5BI-4 cells
whereas for Hz 1075 cells specific binding accounted for
only 25 ~ of the total binding (Fig. 4). Specific binding
data for Tn F, Tn 5B1-4 and Hz 1075 cells were
converted into Scatchard plots (Fig. 5). The Hz 1075
insect cell line was used as a non-permissive control. Tn
F and Tn 5B1-4 cells had affinities of 1-60 x 101° and
2-35 x 1010 M-1, respectively, and specific binding sites
per cell of 13700 and 6000, respectively. The Hz 1075
line, although not capable of supporting AcMNPV
infection, displayed a lower, but saturable, binding of
AcMNPV with approximately 900 sites/cell. The affinity of 1.1 x 101° M-1 was comparable to those of the Tn
cell lines.
An estimate of the exposed area per cell of each line
was calculated by dividing the total area of the tissue
culture plate by the number of cells added to the plate
that allowed a fully confluent monolayer to form. This
area was used to calculate the number of binding sites per
I.tm2 of cell surface. The number of specific binding sites
i
"
5
10
Bound particles/cell (x 10-3)
15
Fig. 5. ScatchardplotofspecificAcMNPV binding to three insect cell
lines. Total and non-specific binding data (as in Fig. 4) were converted
to specific binding and plotted by Scatchard analysis to determine the
binding parameters for Tn F (11} (4.8 x l0 s cells/well), Tn 5B1-4 (O)
(6-5 × 105 cells/welt) and Hz 1075 cells (Jk) (5.5 x l0 s cells/well). Lines
were obtained using linear regression.
for Tn F, Tn 5B1-4 and Hz 1075 cells were 33, 20 and 2
per ~tm2 of cell surface, respectively. The maximum
number of virus particles that could fit as a monolayer
over 1 btm2 could be calculated from their dimensions.
AcMNPV particles are 0.07 p.m in diameter and 0.33 btm
in length. Thus, approximately 200 particles/btm z could
fit, tightly packed, onto the cell if oriented end-on or 43
particles could fit if oriented on their side. A comparison
of these values with the measured number of sites per
~tm2 indicates that between 17~ and 76~ of the surface
of Tn F cells could be occupied by virus particles,
depending upon the orientation of the particles.
Electron micrographs of A c M N P V particles binding to
Tn 5B1-4 cells
Electron micrographs of virus particles binding near
saturation to Tn 5B1-4 cells show that virus particles
cover a large fraction of the cell surface (Fig. 6a). Virus
particles appear to bind over the entire surface and some
virions are bound to other cell-bound particles. A few
particles are clearly visible in coated pits (Fig. 6b). The
small degree of internalization seen here might be real, or
might be an artefact due to warming the cells to room
temperature to perform glutaraldehyde fixation. At 4 °C
some particles were observed to be internalized by the
cell as previously observed in Sf-21 cells (Volkman et al.,
1986); however, the electron micrographs (Fig. 6 a) show
that the internalized (endocytosed or fused) virus
represented less than 5 ~ of the total particles. Additionally, endocytosed virus particles, as shown in Fig. 6(b),
represented a small minority of the total cell-associated
particles.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
A c M N P V attachment to insect cell lines
3191
Fig. 6. Thin-section electron micrograph of AcMNPV particles (a) bound to Tn 5B1-4 ceils under saturating conditions, and (b) bound
and internalized by coated pits and vesicles, respectively.
Effect of pretreatment of cells with proteases on
A c M N P V binding
Protease pretreatments of cells were tested in an attempt
to identify whether the potential AcMNPV receptor had
a protein component. Table 2 summarizes the results of
pretreatments with the proteases trypsin, chymotrypsin
and proteinase K. All three pretreatments of Tn 5B1-4
cells caused a 93 to 96% decrease in binding.
Effect of inhibitors of N-linked oligosaccharide addition
and trimming
The inhibitors of N-linked glycosylation dNM, SW and
TM were used to investigate further the role of N-linked
glycoproteins in binding (Table 3). These agents are
routinely used to study the glycosylation of recombinant
glycoproteins expressed in insect cells by using the
AcMNPV vector (Jarvis & Summers, 1989; Oker-Blom
et al., 1989). SW and dNM inhibit glycolytic trimming
reactions of N-linked glycoproteins following the addition of the high-mannose structure. TM prevents the
addition of the entire high-mannose oligosaccharide onto
Table 2. Effect of protease pretreatments on AcMNP V
binding to Tn 5B1-4 cells
Protease
Control (no protease)
Proteinase K
Trypsin
Chymotrypsin
Radioactivity bound (c.p.m.)
1163 (+
77 (+
48 (+
57 (+
73)
28)
24)
16)
N-linked glycoproteins and can also inhibit the subsequent expression of glycoproteins on the cell surface
(Elbein, 1987). Growth of Tn 5BI-4 cells for 3 days in the
presence o f d N M and SW had no effect on the binding of
AcMNPV, compared to the control. Growth of the cells
was unaffected by dNM, SW, and 0-2 and 1-0 ~tg/ml TM.
TM at a concentration of 4 Ilg/ml caused slight
vacuolation of the cells; however, the cells remained
viable and growing. TM caused a dose-dependent
reduction in the binding of AcMNPV to Tn 5B1-4 cells.
At the highest concentration of TM, binding was
reduced to 21~ of the control. This residual binding
represents non-specific binding.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
3192
T. J. Wickham and others
Table 3. Effect of metabolic inhibitors on A c M N P V binding
to TN 5B1-4 cells*
Metabolic
inhibitor
Tunicamycin
Swainsonine
Deoxynojirimycin
Inhibitory action
Concentration
Binding
(percentage
of control)
N-linked glycosylation
and transport of
proteins to cell surface
Trimming to complex
forms at an
intermediate stept
Early trimming
steps producing
high-mannose
oligosaccharides t
0-2 ktg/ml
1-0 p,g/ml
4.0 ttg/ml
10 ttM
54 (+ 2)
33 (+ 1)
21 (__.3)
108 (+ 12)
4 mM
95 (+ 4)
* Cells grown for 3 days in the presence of inhibitors and incubated
with 32P-labelled virus for 2 h at 4 °C. Percentage of control bound is
the average of triplicates.
t The effects represent those in mammalian cells since the exact
mechanism of these inhibitors in insect cells is not known. They do,
however, cause an increase in Mr of N-linked glycoproteins in insect
cells, suggesting that they function similarly.
Effect of monosaccharides on A c M N P V binding
Further competition experiments were conducted using
the major sugar components that are incorporated into
glycosylaminoglycans, O-linked and N-linked oligosaccharides. The monosaccharides tested were glucose,
mannose, N-acetylglucosamine, fucose and galactose.
None of these was able to reduce binding significantly at
the highest concentrations that could be tested (6 to 7
w/w). Higher concentrations could not be used because
the osmolarity would affect cell viability and integrity.
Discussion
Measurements of the fraction of physical particles that
are infectious revealed a higher proportion of infectious
particles than previous reports (Volkman et al., 1976;
Wang & Kelly, 1985). In previous studies, less than 1 ~ of
the particles were found to be infectious, whereas the
present study found that at least 19~ of the particles
were infectious. This higher fraction is probably due to
the differences in methodologies used to measure
infectious particles. Typically, a diluted virus inoculum is
incubated with Sf-21, Sf-9 or Tn 368 cells for 1 h and then
removed. On the basis of the infection kinetics that we
measured for these cell lines, 1 h is not enough time for
even 30 ~o of the infectious particles to attach and infect
these cells under normal assay conditions (2 x 106 to
4 x 106 ceUs/ml). The method for measuring infectious
particles employed in the present study involved concentrating the virus and cells onto the bottom of the tissue
culture well, thereby increasing the probability of cell-
virus interactions (Wood, 1977). The titres of virus
measured using this method are typically five- to 10-fold
higher than titres measured by other methods (unpublished observations).
A comparison of the attachment rate constant ka and
infection rate constant k i of Tn 5B1-4 cells with the
calculated maximum possible attachment rate constant
ka,nax) shows that AcMNPV attachment to Tn 5B1-4
cells is very close to diffusion-limited, and that infection
is very close to attachment-limited. This finding implies
a strong binding interaction between the virus and cell
since at the diffusion limit every collision of an unbound
particle with the cell results in attachment. At 4 °C,
attachment reached its equilibrium value between 3 and
5 h, and remained unchanged from that time to 18 h. This
suggests that the small degree of internalization seen in
the electron micrographs (Fig. 6) was not persistent or
significant. Continued internalization would have led to
an increase of cell-associated virus, which was not seen.
Furthermore, the lack of change of cell-associated virus
levels from 3 to 18 h suggested the minor internalization
did not promote further attachment, or stimulate viral
release. We conclude internalization has a minor effect,
if any, on attachment at 4 °C.
The close agreement between ka and ki for Tn 5B1-4
cells also indicates that the viral attachment step may be a
controlling factor in the establishment of infection in
these cells. Thus, an unbound infectious virus particle
will nearly always bind after its first collision with the
cell, after which it will probably proceed to infect the cell
productively rather than detaching, for instance. Additionally, the correlation between the virus binding of the
cell lines at 4 °C and their respective infection rates
demonstrates that cells with a higher affinity for the virus
also become infected faster (Table 1).
Pretreatment of the cell surface with proteases decreases saturable binding by 93 to 96~ implicating a
protein in binding, rather than a non-specific interaction
with lipid components in the cell membrane. These
experiments were performed after 2 h of incubation, at
which time virus binding was very close to its equilibrium value (90~o), suggesting the virus was well bound
during these protease experiments. Additionally the
reductions in binding caused by tunicamycin also
implicate a protein, or more specifically a glycoprotein.
Tunicamycin inhibits the addition of the entire core
oligosaccharide onto the protein. However the failure to
glycosylate some proteins can also inhibit their expression on the cell surface due to altered folding, increased
proteolysis, and/or inhibition of intracellular trafficking
(Elbein, 1987). Because of these secondary effects, the
reduction in binding cannot be taken to mean that an Nlinked oligosaccharide is directly involved in binding.
Rather, the reduction indicates that some component of
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
A c M N P V attachment to insect cell lines
glycoprotein(s), and not necessarily the oligosaccharide
moiety itself, may be involved in binding. A lack of
direct participation of N-linked oligosaccharides in
binding is supported by competition experiments using
monosaccharides and inhibitors of N-linked glycosylation. Neither of the experiments was able to demonstrate
any reductions in attachment. Although the exact
mechanisms of the trimming inhibitors have not been
elucidated in insect cells, they have been shown to be
active in insect cells (Roberts, 1989; our unpublished
observations) by increasing the apparent Mr of N-linked
glycosylated proteins, thus accomplishing the general
alterations in the N-linked structures that were intended.
The competition of unlabelled AcMNPV, but not
LdMNPV, with labelled AcMNPV indicates that the
binding component has specificity. Electron micrographs of AcMNPV binding to cells also shows the virus
localized in coated pits which are responsible for the
internalization of ligands through receptor-mediated
endocytosis. Although electron micrographs of cells at
4 °C show internalization, it is small in extent compared
to levels of attachment (as determined by counting
bound and internalized virus in Fig. 6 and similar
electron micrographs). Therefore it appears valid to treat
measurements of cell-associated virus as representing
bound virus at 4 °C. Further evidence that internalization is minimal at 4 °C is the lack of increase in cellassociated virus from 4 to 18 h; active internalization
would lead to a progressive increase in cell-associated
virus. The significant variations in virus binding and
infection rates between cell lines is suggestive of
variations in the expression of a component or components mediating binding.
Calculations of the maximum number of AcMNPV
particles that physically fit onto the surface of a cell
reveal that a large fraction of the surface of some cell
lines is occupied by virus particles. If AcMNPV (0.07
gm x 0.33 gm) were tightly packed end-on to cover the
surface of Tn F cells completely (417 p.m2), a maximum
of only 85000 virus particles could fit as a monolayer. If
the virus particles were placed sideways onto the cell, a
maximum of 18100 could fit. Tn F cells were found from
Fig. 5 to possess 13 700 specific sites enabling binding of
17~ and 76~ of the maximum that could fit tightly
packed in end-on and sideways orientations, respectively. Electron micrographs of particles bound at near
saturating conditions also show that a large fraction of
the surface is occupied by virus particles. Theoretical
treatments of virus attachment have shown that space
can become limiting during virus binding even in the
presence of receptor-mediated attachment if the receptor
number is much greater than the number of spaces on the
cell to which virus can bind (Wickham et al., 1990).
When this occurs, the resultant Scatchard plot can have a
3193
constant slope with a site number approaching the
maximum number of virus particles that can fit onto the
cell surface, similar to what is seen for Tn F cells (Fig. 5).
This interpretation of spatial saturation does not rule out
the existence of a specific interaction of AcMNPV with
the cell surface. In the above context, a virus 'site' is
related to the area that a single particle occupies on the
surface of a cell. Based on the measured number of sites
and from electron micrographs of particles binding to
these cells under saturating conditions, it appears that
the amount of available binding space begins to become
a limiting factor during AcMNPV attachment to Tn F
cells. This interpretation of the binding data may be
applicable to other virus systems with large virus site and
receptor numbers.
AcMNPV is capable of infecting at least 34 insect
species; however, most of these species require large
doses of the virus to become infected (Evans, 1986). The
inability of LdMNPV to compete with AcMNPV for
binding, the differences in infection rates and binding
between cell lines, and the reductions in binding caused
by tunicamycin and protease pretreatment suggest
specificity of binding. The discovery of an AcMNPV
receptor will be important in discovering how its
function relates to the relatively broad host range of
AcMNPV and how differences in the sequences and
expression of the receptor influence insect susceptibility.
This work was supportedby the NationalScienceFoundationwho
sponsored this research under NSF EET-8807089 and by the Cornell
Biotechnology Program in the form of a predoctoral fellowship
awarded to T. J. Wickham. The Cornell BiotechnologyProgram is
sponsoredby the NewYorkStateScienceand TechnologyFoundation,
a consortiumof industries, the U.S. Army Research Officeand the
National Science Foundation.
References
ALCAMI, A., C,~a~SCOSA, A. L. & VINUELA, E. (1989). Saturable
binding sites mediate the entry of African swine fever virus in Veto
cells. Virology 168, 393-398.
A TKINS, P. W. (1982). PhyMcalChemistry, pp. 809-856. San Francisco:
W. H. Freeman.
BASAK, S. & COMPANS, R. W. (1989). Polarized entry of canine
parvovirus in an epithelial cell line. Journal of Virology 63, 31643167.
DALGLEISH, A. G., BEVERLEY, P. C. L., CLAPHAM, P. R., CRAWFORD,
D. H., GREAVES,M. F. & WEISS, R. A. (1984). The CD4 (T4) antigen
is an essential component of the receptor for the AIDS retrovirus.
Nature, London 312, 763-767.
DE GOOIJER, C. n., VANLIER, F. L. J., VANDEN END, E. J., VLAK, J. M.
& TRAMPER, J. (1989). A model for baculovirus production with
continuous insect cell cultures. Applied Microbiology and BiotechnoIogy 30, 497-501.
ELBEIN, A. D. (1987). Glycosylation inhibitors for N-linked glycoproteins. Methods in Enzymology 138, 661-709.
EPSTEIN, R. L., POWERS, M. L., ROGART, R. B. & WEINER, H. L. (1984).
Binding of t25I-labeled reovirus to cell surface receptors. Virology
133, 46-55.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12
3194
T. J. W i c k h a m and others
EVANS, H. F. (1986). Ecology and epizootiology of baculoviruses. In
The Biology of Baculoviruses, vol. 2, Practical Application for Insect
Control, pp. 89-132. Edited by R. R. Granados & B. A. Federici.
Boca Raton: CRC Press.
FRIES, E. & HELENIUS, A. (1979). Binding of Semliki Forest virus and
its spike glycoproteins to cells. European Journal of Biochemistry 97,
213-220.
GOODWlN, R. H., TOMPKINS, G. J., GET'FIG, R. R. & ADAMS, J. R.
(1982). The characterization and culture of virus replicating
continuous insect cell lines from the bollworm, Heliothis zea
(Boddie). In Vitro 18, 843-850.
GRANADOS, R. R. & NAUGHTON, M. (1976). Replication of Amsacta
moorei entomopoxvirus and Autographa californica nuclear polybedrosis virus in hemocyte cell lines from Estigmene acrea. In
Invertebrate Tissue Culture. Applications in Medicine, Biology, and
Agriculture, pp. 379-389. Edited by E. Kurstak & K. Maramorosch.
New York: Academic Press.
HINK, W. F. (1970). Established insect cell line from the cabbage
looper, Trichoplusia nL Nature, London 266, 466~,67.
JARVlS, D. L. & SUMMERS,M. D. (1989). Glycosylation and secretion of
human tissue plasminogen activator in recombinant baculovirusinfected insect cells. Molecular and Cellular Biology 9, 214-223.
KLATZMANN, D., CHAMPAGNE, E., CHAMARET, S., GRUEST, J.,
GUETARD, D., HERCEND, T., GLUCKMAN, J. C. & MONTAIGNIER, L.
(1984). T-Lymphocyte T4 molecule behaves as the receptor for
human retrovirus LAV. Nature, London 312, 767-768.
LABARCA,C. & PAIGEN, K. (1980). A simple, rapid, and sensitive DNA
assay procedure. Analytical Biochemistry 102, 344-352.
LAYNE, S. P., SPOUGE, J. L. & DEMBO, M. (1989). Quantifying the
infectivity of human immunodeficiency virus. Proceedings of the
National Academy of Sciences, U.S.A. 86, 4644-4648.
LUCKOW, V. A. (1990). Cloning and expression of heterologous genes in
insect cells with baculovirus vectors. In Recombinant DNA Technology and Applications, pp. 2-25. Edited by C. Ho, A. Prokop & R.
Bajpai. New York: McGraw-Hill.
METI'ENLEITER, T. C., ZSAK, L., ZUCKERMANN,F., SUGG, N., KERN, H.
& BEN-PORAT, T. (1990). Interaction of glycoprotein glII with a
cellular heparinlike substance mediates adsorption of pseudorabies
virus. Journal of Virology 64, 278-286.
OKER-BLOM, C., PETI'ERSSON, R. F. & SUMMERS, M. D. (1989).
Baculovirus polyhedrin promoter-directed expression of rubella
virus envelope proteins, El and E2, in Spodopterafrugiperda cells.
Virology 172, 82-91.
PAULSON, J. C., ROGERS, G. N. MURAYAMA,J.-I., SZE, G. & MARTIN,
E. (1986). Biological implications of influenza virus receptor
specificity. In Virus Attachment and Entry into Cells. Proceedings of
the American Society for Microbiology Conference, Philadelphia, pp.
144-151. Edited by R. L. Crowell & K. Lonberg-Holm Washington,
D.C. : American Society for Microbiology.
ROBERTS, T. E. (1989). Fatty acid acylated baculoviruspolypeptides and
aspects of intracellular protein processing. Ph.D. thesis, Queen's
University, Kingston, Ontario, Canada.
SCHLEGEL, R., WILLINGHAM, M. C. & PASTAN, I. H. (1982). Saturable
binding sites for vesicular stomatitis virus on the surface of Vero
cells. Journal of Virology 43, 871-875.
SCHWARZ, R. T. & DATEMA, R. (1984). Inhibitors of trimming: new
tools in glycoprotein research. Trends in Biological Sciences 9, 32-34.
Snouv, D. & SZABO, A. (1982). Role of diffusion in ligand binding to
macromolecules and cell-bound receptors. Biophysical Journal 40,
33-39.
SMITH, I. R. L., VANBEEK, N. A. M., PODGWAITE,J. D. & WOOD, H. A.
(1988). Physical map and polyhedrin gene sequence of Lymantria
dispar nuclear polyhedrosis virus. Gene 71, 97-105.
SMOLUCHOWSKI, i . V. (1917). Versuch einer mathematischen theorie
der koagulationskinetic kolloider losungen. Zeitsehrift J~r Physikalische Chemie 92, 129-168.
SUMMERS, M. D. & SMITH, G. A. (1987). A manual of methods for
baculovirus vectors and insect cell culture procedures. Texas
Agricultural Experimental Station Bulletin B1555, 1-56.
TAYLOR, H. P. & COOPER, N. R. (1989). Human cytomegalovirus
binding to fibroblasts is receptor-mediated. Journal of Virology 63,
3991-3998.
TRIMBLE, R. B. & MALEY, F. (1984). Optimizing hydrolysis of N-linked
high-mannose oligosaccharides by endo-fl-N-acetylglucosaminidase
H. Analytical Biochemistry 141, 515-522.
VAUGHN, J. L., GOODWIN, R. H., TOMPKINS, G. J. & MCCAWLEY, P.
(1977). The establishment of two cell lines from the insect Spodoptera
frugiperda (Lepidoptera :Noctuidae). In Vitro 13, 213-217.
VERDIN, E. M., KING, G. L. & MARATOS-FLIER, E. (1989). Characterization of a common high-affinity receptor for reovirus serotypes 1
and 3 on endothelial cells. Journal of Virology 63, 1318-1325.
VOLKMAN,L. E., SUMMERS,M. D. & HSlEH, C.-H. (1976). Occluded and
non-occluded nuclear polyhedrosis virus grown in Trichoplusia ni:
comparative neutralization, comparative infectivity, and in vitro
growth studies. Journal of Virology 19, 820-832.
VOLKMAN, L. E., GOLDSMITH, K. & HESS, R. T. (1986). Alternate
pathway of entry of budded Autographa californica nuclear polyhedrosis virus: fusion at the plasma membrane. Virology 148, 288-297.
WANG, X. & KELLY, n . C. (1985). Baculovirus replication: uptake of
Trichoplusia ni nuclear polyhedrosis virus particles by insect cells.
Journal of General Virology 66, 541-550.
WICKHAM, T. J., GRANADOS, R. R., WOOD, H. A., HAMMER, D. A. &
SHULER, M. L. (1990). General analysis of receptor-mediated viral
attachment to cell surfaces. Biophysical Journal 58, 1501-1516.
WOOD, H. A. (1977). An agar overlay plaque assay method for
Autographa californica nuclear polyhedrosis virus. Journal of lnvertebrate Pathology 29, 304-307.
WOOD, H. A. (1980). Autographa californica nuclear polyhedrosis virusinduced proteins in tissue culture. Virology 102, 21-27.
WOOD, H. A., JOHNSTON, L. B. & BURAND, J. B. (1982). Inhibition of
AcNPV replication in high-density Trichoplusia ni cell cultures.
Virology 119, 245-254.
WUDUNN, D. & SPEAR, P. G. (1989). Initial interaction of herpes
simplex virus with cells is binding to heparan sulfate. Journal of
Virology 63, 52-58.
WUNNER, W. H., REAGAN, K. J. & KOPROWSKI, H. (1984).
Characterization of saturable binding sites for rabies virus. Journal of
Virology 50, 691-697.
(Received 24 September 1991 ; Accepted 10 August 1992)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 18:59:12