Download Purification and properties of virus particles, infectious subviral

Journal of General Virology (1994), 75, 1849 1857. Printed in Great Britain
1849
Purification and properties of virus particles, infectious subviral particles,
cores and VP7 crystals of African horsesickness virus serotype 9
J. N. Burroughs, R. S. O'Hara, C. J. S m a l e , t C. Hamblin, A. Walton, R. Armstrong
and P. P. C. Mertens*
A F R C Institute for Animal Health, Pirbright Laboratory, Pirbright, Woking, Surrey GU24 ONF, U.K.
Methods were developed for the purification, at high
yield, of four different particle types of African horsesickness virus serotype 9 (AHSV-9). These products
included virus particles purified on CsC1 gradients which
contain proteins apparently directly comparable to those
ofbluetongue virus (VP1 to VP7); virus particles purified
on sucrose gradients which also contain, as a variable
component, protein NS2; infectious subviral particles
(ISVPs), containing chymotrypsin cleavage products of
VP2; and cores, obtained by treating purified ISVPs
with 1 M-MgCIz to remove the components of the outer
capsid layer (VP5 and VP2 cleavage products). Additional protein bands migrating with apparent Mrs
lower than that of VP5 were detected during SDS-PAGE
analysis of virus particles. These appear to be conform-
ational variants of VP5 and are identified as VP5' and
VP5". BHK-21 cells infected with this strain of AHSV9 produce large quantities of flat, usually hexagonal
crystals ofVP7, a major group antigen and core protein;
these were also purified. Either 20 mg of virus particles,
20 mg of ISVPs or 10 mg of cores as well as 20 mg of
VP7 crystals could be purified from approximately
8× 109 infected cells. None of the preparations of
particles or crystals showed any detectable contamination with BHK-21 cell proteins or antigens, as
determined by SDS-PAGE or indirect ELISA. Virus
particle and ISVP preparations had similar specific
infectivities for BHK-21 cells (approximately 1 × 109
TCIDs0/A260 unit) but the infectivity of cores was
approximately 105-fold lower.
Introduction
fluorocarbon extraction of infected cell homogenates,
followed by sucrose or caesium chloride gradient
centrifugation. However, using these methods we were
unable to recover significant amounts of purified virus
particles of BTV serotype 1 from South Africa (BTVISA)
from either sucrose or CsC1 centrifugation gradients.
Similar difficulties were anticipated with AHSV. The
yield of AHSV particles was not reported in earlier
studies and the evaluation of particle purity was limited
to electrophoretic analyses of the virus proteins, followed
by staining with Coomassie blue or silver, or fluorography. Using these techniques, it is relatively easy to
identify the major AHSV outer capsid proteins (VP2 and
VP5) and core proteins (VP3 and VP7). The identification
of the minor core proteins and non-structural proteins
has been less straightforward, owing to the possibility of
low-level contamination with host cell or non-structural
virus proteins, together with variations in the expression
or labelling levels and the migration of some proteins
during electrophoresis. In consequence, some confusion
still exists over the identity of some AHSV proteins,
which has resulted in the use of a protein nomenclature
(Grubman & Lewis, 1992) that is inconsistent with that
previously used for BTV, the prototype orbivirus
(Gorman et al., 1985). An analysis of the genome
African horsesickness is caused by an arthropod-borne
virus which is recognized primarily as a pathogen of
horses and other Equidae. African horsesickness virus
(AHSV) is a member of the Orbivirus genus, within the
family Reoviridae (Holmes, 1991) and possesses a dsRNA
genome composed of 10 segments (Bremer, 1976)
packaged within the core of a double-layered icosahedral
protein capsid (Oellerman et al., 1970). AHSV is
relatively closely related to bluetongue virus (BTV) (the
prototype orbivirus) and epizootic haemorrhagic disease
virus, two other serogroups of orbiviruses with which it
shows some serological cross-reaction and RNA sequence homology (Huismans & Eaton, 1992; Gorman et
al., 1983; Van Staden et al., 1991; Nel et al., 1992).
Previous reports on the purification and structural
protein components of AHSV (Oellerman et al., 1970;
Bremer, 1976; Grubman & Lewis, 1992) are all based on
procedures published for BTV (Verwoerd, 1969;
Verwoerd et al., 1972; Huismans, 1979; Huismans et al.,
1987a, b; Mecham et al., 1986). These methods involve
I Presentaddress: The RobensInstitute,The Universityof Surrey,
Guildford, SurreyGU2 5XH, U.K.
0001-2278 © 1994SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
1850
J. N. Burroughs and others
segment coding assignments of AHSV (R. S. O'Hara,
J. N. Burroughs & P. P. C. Mertens, unpublished results)
has helped us to identify the primary gene products and
consequently to confirm the nature and identity of the
individual virus structural proteins. AHSV particles
purified by the methods described here have a protein
composition essentially similar to that of BTV particles
and we were therefore able to retain a protein nomenclature consistent with that for BTV and as
recommended by Gorman et al. (1985).
We describe methods for the high yield purification of
four different particle types of AHSV serotype 9 (AHSV9) from infected BHK-21 cells; in addition, flat, usually
hexagonal crystals of the major core protein and group
antigen VP7 were purified. The purification of such
crystals from orbivirus-infected cells and their characterization has not previously been reported.
Methods
Virus. AHSV-9 [South African isolate, 2 May 1973, designated 90/61
(MB3, BHK4)] and AHSV-2 (MB3, BHK2) were supplied by Dr B. J.
Erasmus. They were plaque-purified three times in BHK-21 cells. The
growth and partial purification of AHSV was carried out exactly as
described for BTV by Mertens et al. (1987).
Virus particles. The partially purified virus was resuspended in
10 mM-DTT and l % sodium N-lauroyl sarcosine (NLS) until the
densely turbid suspension had clarified (usually within 15 min), and
was then centrifuged through 10 ml 40% (w/v) sucrose (containing
0.1% NLS) onto a 5 ml 50% (w/v) sucrose cushion (Beckman SW28
rotor, 2 h, 60000g, 4 °C). The virus particles, which formed a blue
opalescent band at the interface of the sucrose solutions, were recovered
and dialysed against 0.1 M-Tri>HCI pH 8.0 containing 0.1% NLS.
Virus particles could also be purified using CsC1 gradient centrifugation. Preformed gradients were made by carefully overlaying 2 ml
of 139 g/ml CsCI solution with 2 ml of 1.30 g/ml CsCI solution
followed by 1 ml of 40% (w/v) sucrose (all solutions in 0-1%
NLS/Tris-HCI buffer). Material clarified as above was layered onto
these gradients and centrifuged in an SW40 rotor at 130000 g for 1.5 h
at 4 °C, The virus particles (a sharp, white band at the interface of the
CsC1 solutions) were recovered and further purified in self-forming
CsCI gradients [1.33 g/ml CsCI in 0.1% NLS (SW40 rotor, 130000g,
> 17 h, 4 °C)I at a density of 1.33 g/ml and dialysed as above.
IS V P s / V P 7 crystals. The partially purified virus material was treated
with 20 mg/ml chymotrypsin (Sigma) and 1% NLS for 1 h at 37 °C;
this caused the highly turbid suspension to increase in clarity but it did
not become completely clear. This material was then centrifuged
through a sucrose gradient with steps of 4 ml of saturated (i.e. 66 %
w/v) sucrose solution, 8 ml of 50 % (w/w) sucrose and 8 ml of 40 %
(w/v) sucrose, all containing 0-1% NLS (SW28 rotor, 2 h, 60000g,
4 °C). An opalescent blue band of ISVPs was recovered from the upper
sucrose interface and dialysed against Tris-HC1 buffer. A dense,
greyish-white band was recovered at the lower sucrose interface and
this was further centrifuged on preformed discontinuous CsCI gradients
(density 1.24 to 1.36 g/ml, 40% sucrose, without NLS, SW40 rotor,
1.5 h, 130000 g, 4 °C). A white band was recovered from the interface
of the CsCI solutions and was centrifuged on a self-forming CsC1
gradient (1-31 g/ml CsCI without NLS, SW40 rotor, > 17 h, 130000 g,
4 °C). The material again formed a sharp white band, had a density of
1.31 g/ml and was composed of crystals of the major core protein, VP7.
It was recovered and dialysed against Tris-HC1 buffer.
Virus cores. When material which had been treated with chymotrypsin and NLS was centrifuged on CsC1 gradients using the same
conditions as for virus particles, a band of ISVPs was obtained with a
density of 1.34 g/ml. Following dialysis to remove NLS and treatment
with 1.0 M-MgC12, the resulting product was centrifuged on preformed
CsC1 gradients (density 1.30 to 1.44 g/ml CsCI, 40% (w/v) sucrose
without NLS, SW40 rotor, 1-5 h, 130000 g, 4 °C). A sharp, white band
of core particles was obtained with a density of 1-36 g/ml which was
harvested and dialysed against Tris HCI buffer.
Antisera. Guinea-pig antisera were raised against disrupted BHK-21
cells, virus purified in sucrose or CsCI, ISVPs, cores and VP7 crystals
as described by Hamblin et al. (1991).
lndireet ELISA. Diluted purified antigens were adsorbed to wells of
microtitre plates. The specificity and endpoint titre of each serum were
then determined by titrating a twofold dilution series of the serum
against antigen as described by Hamblin et al. (1990).
Specific infectivity determinations. The infectivity of different AHSV
particle preparations was determined by the production of c.p.e, using
a 10-fold dilution series added to microwell plates containing BHK-21
cells. The final infectivity, expressed as TCIDs0/ml, was calculated as
described by K/irber (1931). From the A~60/ml of each preparation,
specific infecfivities could then be expressed as TCIDs0/A260 unit.
Electrophoretic analysis. The [aaSlmethionine-labelled proteins present in BHK-21 cell harvests and extracts and in purified AHSV particle
preparations were analysed using 11% polyacrylamide gels with a
Laemmli (1970) buffer system followed by fluorography (Bonner &
Laskey, 1974).
Electron microscopy. Purified AHSV preparations were stained with
either 2 % methylamine tungstate pH 8.2 or aqueous uranyl acetate and
examined in an electron microscope.
Results
Partial purification of AHSV-9 from infected tissue
culture material was achieved using a combination of
detergent treatments (Triton X-100 and sodium deoxycholate) and discontinuous sucrose gradient centrifugation using the procedures described for BTV
(Mertens et al., 1987). When this material was compared
to the original tissue culture harvest, or cytoplasmic
extracts (Fig. 1, lanes 1 and 2 respectively) some increase
in virus particle purity was detected but significant
amounts of non-structural A H S ¥ proteins were still
present (data not shown). Final purification of intact
virus particles was achieved by treatment with DTT
(10 mM) and NLS (1%), followed by centrifugation on
discontinuous sucrose gradients containing 0-1% NLS.
Virus particles purified in this way were free of detectable
cellular proteins and the majority of the non-structural
virus proteins, but did contain proteins designated VP1
to VP7, which are directly comparable in terms of
relative amounts and MrS to those present in BTV
particles (Mertens et al., 1987). However, these preparations also contained protein NS2 as a variable
component (Fig. 1, lane 8), sometimes in much larger
amounts than in BTV particles purified using either
sucrose (BTVISA) or CsC1 gradients (BTV4) (Mertens et
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
Afi'ican horsesickness virus purification
1
2
3 4
5
6
7
8
9
1
1851
2
:%
vm
-
VPI/2~
VP3VP5
VP5'
VP4 -
- VP5
- VP5'
VP5"
i VP5
VP5'
VP5Z
NS2,
VP7
VP6 VP7 -
Fig. 2. Fluorograph of an 11% polyacrylamide Laemmli slab gel. Lane
1, [3~S]methionine-labelled purified AHSV-2 virus, 1% 2-mercapto-
ethanol in sample buffer; lane 2, as lane 1, but with 0.1% 2mercaptoethanol in the sample buffer.
Fig. 1. Fluorograph of an 11% polyacrylamide Laemmli slab gel. Lane
1, [sSS]methionine-labelled AHSV-9-infected BHK-21 cell harvest; lane
2, cytoplasmic extract of AHSV-9-infected cells; lanes 3 and 9, virus
from CsC1 gradients; lane 4, ISVPs from sucrose gradients; lane 5,
ISVPs from CsCI gradients; lane 6, cores; lane 7, VP7 crystals; lane 8,
virus from sucrose gradients. Asterisks indicate the chymotrypsin
cleavage products of VP2.
al., 1987). In contrast to BTVISA or BTV3 (which are
relatively unstable in high concentrations of CsC1 and
lose their outer capsid proteins), it was also possible to
purify DTT/NLS-treated AHSV-9 virus particles by
CsC1 gradient centrifugation. The particles purified by
this method had a density of 1'33 g/ml, and did not
contain any detectable NS2 protein (Fig. 1, lane 3).
Incorporation of 0.1% NLS into these CsCI gradients
was found to be necessary to obtain sharp bands of
purified material. Maintenance of both types of virus
particle (with or without NS2) in 0.1% NLS was
essential to prevent irreversible aggregation. Storage of
purified virus particles at 4 °C in 0.1 M-Tris-HC1 pH 8.0,
1% NLS had no detectable effect on virus infectivity
over a 12 month period. However, some breakdown of
VP2 was observed even after short periods of storage,
leading to the detection of additional protein bands
migrating between VP3 and VP5 (Fig. 1, lane 3) that
were not present in freshly purified virus (Fig. 1, lane 8).
In comparable preparations of BTV particles these
minor bands were detectable by Western blotting using
antisera to synthetic peptides derived using sequence
data for VP2 (data not shown) thereby confirming their
identity as VP2 breakdown products.
If the partially purified material was treated with
chymotrypsin and NLS and centrifuged through a
sucrose gradient, modified particles (identified as ISVPs)
were recovered from the upper sucrose interface. In these
particles the outer capsid protein, VP2, had been cleaved
into five smaller products identified as VP2a to VP2e
(Fig. 1, lane 4) which have been shown to be derived
from virus structural proteins by Western blotting using
antisera to purified AHSV-9 virus particles (data not
shown). This contrasts with ISVPs of BTVISA in which
case only three chymotrypsin cleavage products were
consistently associated with the particle. Material
forming a dense greyish-white band was also recovered
from the lower sucrose interface and was further purified
on preformed and self-forming CsC1 gradients. This
material, which formed a sharp white band with a
density of 1-31 g/ml, had a crystalline appearance when
visualized by light microscopy and was composed
entirely of the major core protein and group antigen VP7
(Fig. 1, lane 7). These crystals, which were present during
the early stages of virus purification, are thought to form
within the infected cells but are not thought to be the
product of infection with defective interfering particles,
since they were also produced during replication of virus
which had been plaque-purified an additional three times
and during infections at both high and low multiplicities.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
1852
J. N. Burroughs and others
Table 1. Specific infectivity (TCID~o/A26 o unit) of
purified A H S V - 9 virus particles, ISVPs and cores
Viral material
smaller, their yield was lower and they had a more
rounded appearance.
When material which had been treated with chymotrypsin and NLS was centrifuged on CsC1 gradients
using the same conditions as for virus particles, the outer
capsid was not stripped off (in contrast to BTV3 or
BTVISA; Mertens et al., 1987) and a band of ISVPs was
obtained at a density of 1.34 g/ml. These particles had
the same protein composition as those purified using
sucrose gradients (Fig. 1, lane 5).
Following dialysis to remove NLS, ISVPs were treated
with 1-0 M-MgC12, resulting in their conversion to virus
cores, which were then purified by CsC1 gradient
centrifugation at a density of 1-36 g/ml, and contained
proteins VPI, VP3, VP4, VP6 and VP7 (Fig. l, lane 6).
AHSV cores appeared to be unstable in 1.5 N-MgC12,
conditions known to release cores of BTV4 from the
outer capsid components (Mertens et al., 1987), whereas
Specific infectivity
Virus particles from sucrose*
Virus particles from CsCI*
ISVPs from sucrose
ISVPs from CsC1
Cores
1.7 x
0.5 x
0.7 ×
0.9 ×
1.5 X
109
109
109
109
10 4
* Particles were purified and maintained in 1-0% or 0-1% NLS
respectively and are equivalent to the 'disaggregated' BTV virus
particles described by Mertens et al. (1987).
It was subsequently found that VP7 crystals were at least
partially soluble in 1.0% NLS and 10mM-DTT
(conditions used for virus particle purification) and
although the crystals could also be purified from the
gradients used for final isolation of virus particles, by
adding a 66% (w/v) sucrose step, they tended to be
l
I
I
I
[
I
I
I
1
1
I
(b)
F
I
[
I
I
I
~x Normal serum
Anti-BHK
• Anti-virus (sucrose)
1.2
O Anti-virus (CsC1)
0.8
t
'
0.4
~ O
• Anti-ISVP s
-
[] Anti-cores
_
• Anti-VP7 crystals
I
I
I
I
I
I
,r
_
|
J
i
J
i
2-0 -
(f)
(h)
g)
16
1.2
0-8
0-4
-2
-3
4
-5
6
7
2
3
~
-5
-6
7
-2
-3
~,
-5
-6
-7
Dilution (loglo)
Fig. 3. Indirect ELISA using twofold dilution series of normal guinea-pig sera (A) or guinea-pig antiserum raised against B H K cellular
antigens ('k'), virus particles purified on sucrose (O), virus particles purified on CsCI gradients (©), ISVPs purified on sucrose gradients
(ll), cores ([]) and VP7 crystals (A)- These sera were titrated against each antigen. (a) BHK cell antigen; (b) unpurified AHSV; (c)
virus from sucrose; (d) virus from CsC1; (e) ISVP; (f) cores; (g) VP7 crystals; (h) A. californica NPV polyhedrin.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
African horsesickness virus purification
1853
Fig. 4. Electron micrographs of purified particles. (a) Virus particles in sucrose, stained with 2 % methylaminetungstate. (b) Virus
particles in CsCl. (c) ISVPs in sucrose. (d) Cores. Samples shown in (b to d) were stained with aqueous uranyl acetate. Bar marker
represents 20 nm.
at 0.5 M-MgC12 only a low rate of conversion to cores was
obtained.
Although the amount of [~SS]methionine label incorporated into VP4 was low, VP4 was consistently
detected in all preparations of intact and modified
particles and comigrated with the primary translation
product of genome segment 4 (O'Hara et al., unpublished), thereby confirming its identity as a virus
structural protein. The VP5 band detected in purified
virus could sometimes be resolved into two distinct
bands by S D S - P A G E under standard conditions
(identified as VP5 and VP5'; Fig. 1, lane 9). Purified
AHSV-2 virus particles were also analysed by S D S P A G E with decreasing concentrations of 2-mercaptoethanol in the sample buffer. At a concentration of 0"1%
(v/v) another band (VP5") running slightly faster than
VP5' was also detected (Fig. 2, lane 2); but was not seen
under normal reducing conditions (Fig. 2, lane 1). As the
amount of reducing agent was decreased further, from
0-01% to zero, the relative amount of VP5" increased
and VP5' disappeared. When VP5 and VP5' were
extracted from gels and compared by Cleveland mapping
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
1854
J. N. Burroughs and others
(b)
$
fb
Fig. 5. Electron micrographs of VP7 crystals, stained with 2% methylamine tungstate. (a) Typical preparation, (b) crystal with
hexagonal profile, (c) lattice detail from hexagonal particles and (d) crystalline fragment showing edge detail. Bar marker in (d)
represents 4.3 gm in (a), 260 nm in (b), 104 nm in (c) and 73 nm in (d).
they produced indistinguishable polypeptide cleavage
products. In addition, when these three proteins (VP5,
VP5' and VP5") were reanalysed by SDS PAGE, under
fully reducing conditions, they all comigrated with
unpurified VP5 from intact virus particles (data not
shown).
In a typical preparation, AHSV was purified from
forty 850 cm 3 roller bottles of BHK-21 cell monolayers
(about 8 x 109 cells). These procedures yielded approximately 50 A26o units of particles, which is equivalent to
either 20 mg of virus particles, 20 mg of ISVPs or 10 mg
of cores, (assuming the same relationship between A~6o
and particle concentrations as with BTV) and approximately 20 mg of VP7 crystals.
There was no significant difference in specific in-
fectivity of virus particles or ISVPs purified using sucrose
or CsC1 (Table 1). However, the value for cores was
approximately 10~-fold lower. These results are similar to
those for purified particles of BTVISA and BTV4
(Mertens et al., 1987, 1993).
To confirm that all the AHSV preparations were free
of contaminating BHK-21 cell antigens, antisera were
raised in guinea-pigs to the different purified particles
and VP7 crystals. The specificity and endpoint titres of
each serum were determined by an indirect ELISA. A
B H K cell antigen preparation gave a strong reaction
with its homologous antiserum (Fig. 3a). Unpurified
AHSV antigen gave a strong reaction with both
homologous anti-virus serum and anti-BHK cell serum
(Fig. 3b). However, although all of the purified AHSV
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
African horsesickness virus purification
antigens reacted strongly with their own homologous
antisera they failed to react significantly with either antiBHK cell or normal guinea-pig sera (Fig. 3c to g).
Nuclear polyhedrosis virus (NPV) polyhedrin from
Autographa californica was used in negative antigen
controls and failed to react with the sera under test (Fig.
3 h). The low reactions with BHK cell antigens, seen with
all of the antisera raised against the purified AHSV
preparations, at high antiserum concentration, were of a
similar order to those obtained with normal guinea-pig
serum (Fig. 3 a). These data demonstrate that none of the
AHSV antigens purified by the methods described
contained levels of BHK cell antigens that were detectable by these ELISAs.
Virus particles from sucrose gradients were examined
by electron microscopy following staining with 2%
methylamine tungstate pH 7'2 (Fig. 4a). They had a
poorly defined surface structure similar to that previously
observed in some particles by Oellermann et al. (1970)
who stained with phosphotungstic acid. However, the
appearance and diameter of the virus particles and ISVPs
purified using CsC1 gradients and then stained with 2 %
methylamine tungstate was similar to that of core
particles, demonstrating not only a clear difference
between the virus particles purified on CsC1 and sucrose
gradients but also suggesting that in some circumstances
this stain can remove the outer capsid layer of AHSV.
When virus particles purified using sucrose gradients
were stained with a saturated aqueous solution of uranyl
acetate (data not shown) some evidence of surface
structure in the outer capsid layer was apparent. Similar
results (with the more diffuse outer capsid appearance,
distinct from that of cores) were also obtained with virus
particles purified in CsC1 and stained with uranyl acetate
(Fig. 4b). However, slightly better evidence of surface
structure and axes of fivefold symmetry was observed on
AHSV-9 ISVPs stained with uranyl acetate (Fig. 4c),
which were not previously seen with whole virus or
ISVPs of BTV using similar stains (Mertens et al., 1987).
Core particles showed a clearly defined outer capsid
structure (Fig. 4d) similar to that which has been
described for cores of BTV, regardless of the stain used,
although the ring-shaped surface capsomers reported by
Oellerman et al. (1970) appeared to be divided, usually
into six wedge-shaped sections. Preliminary X-ray
diffraction data indicate that these are trimers of VP7
(D. Stuart, personal communication). Diameters of all
particles were not significantly different to those reported
for BTV (Mertens et al., 1987).
A typical preparation of VP7 crystals examined at low
magnification (Fig. 5a, b) showed large numbers of
intact and fragmented flat hexagonal structures with a
maximum diameter of 8000 nm. When examined at
higher magnification by electron microscopy following
1855
staining in methylamine tungstate it was found that these
structures were composed of a highly ordered twodimensional crystalline lattice (Fig. 5 c, d).
Discussion
Effective methods have been developed for the purification of ISVPs, cores and VP7 crystals of AHSV-9. The
VP7 crystal structure and the AHSV particles containing
NS2 protein have not been described before. It is
conceivable that NS2 is copurified with virus particles or
is bound to the outer virus surface in some way rather
than representing a true structural component and is
released by high CsC1 concentrations.
Grubman & Lewis (1992) were unable to detect either
of the [3~S]methionine-labelled minor proteins (VP4 or
VP6) described by Bremer (1976), or any proteins
analogous to VP4 and VP6 of BTV, in AHSV particles
purified using the method of Mecham et al. (1986).
However, they did report a series of minor bands
detected by silver staining, which they suggested included
VP4 and VP6. The virus purification procedure used did
not include detergent and reducing agent treatment
which we have found to be essential to release virus
particles from cellular materials. When Mecham et al.
(1986) originally used this method for BTV, most of the
preparations contained a protein of M~ close to that of
NS1, which may indicate that the method gives incomplete purification. Minor bands detected by silver
staining in whole virus preparations may therefore
represent contaminants rather than real structural
components. Cellular proteins have been identified which
migrate near VP4 during electrophoresis, are still
synthesized late in infection by AHSV and are more
highly labelled with [35S]methionine than AHSV VP4
(which consistently labelled less well than VP4 of BTV)
(Fig. 1, lane 1; O'Hara et al., unpublished). This may
explain why Bremer (1976) reported two minor protein
bands in preparations of virus particles of AHSV, VP4
and VP4a, the identity of which relative to VP4 of BTV
remains uncertain. The data reported here and to be
reported by O'Hara et al. indicate that by comparison
with BTV, VP6 and possibly VP4 of AHSV were
incorrectly identified in earlier reports.
In vitro translation of AHSV genome segment 6
(numbered according to its migration during electrophoresis on 1% agarose gels) yielded at least three
related proteins, the largest of which comigrated with the
major outer capsid protein VP5 during SDS-PAGE
(O'Hara et al., unpublished results). Grubman & Lewis
(1992) refer to the faster migrating translation product of
genome segment 6 of AHSV-4 as VP6, although no
evidence was presented to show that this product was
equivalent to VP6 of BTV (which is encoded by genome
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
1856
J. N. Burroughs and others
segment 9) and its Cleveland map analysis was similar to
that of VP5. We have observed in purified virus three
distinct forms of VP5 with different migration rates in
SDS-PAGE (VP5, VP5' and VP5"), which after recovery
and re-analysis by SDS-PAGE all comigrate and
therefore appear to be of similar M r. It is considered
likely that VP5 can adopt different conformational
structures which are at least partially dependent on the
reducing environment used and which alter its rate of
migration during SDS PAGE. Laviada et al. (1993) also
detected a second protein band in the region of VP5
when purified particles of AHSV-4 were analysed by
SDS-PAGE. They confirmed its relatedness to VP5 by
its reaction with a VP5-specific monoclonal antibody.
We have identified the different forms of VP5 as VP5'
and VP5", and have identified the different major bands
detected after in vitro translation of segment 6 of AHSV9 as VP5, VPSa and VPSb (O'Hara et al., unpublished)
as suggested by Gorman et al. (1985). The equivalence of
the VPSa or VP5b bands to either VP5' or VPS", if any,
remains to be determined.
The nomenclature for AHSV proteins previously used
by Grubman & Lewis (1992) is incompatible with that
for BTV (Gorman et al., 1985; Mertens et al., 1987) and
would cause confusion in Orbivirus protein nomenclature
as a whole. Grubman & Lewis (1992) did not identify any
other AHSV minor structural proteins which were
directly comparable to VP6/VP6A found in purified
particles ofBTV (Mertens et al., 1984, 1987; Verwoerd et
al., 1972). The minor proteins that we have consistently
identified in all of the AHSV-9 particles include VP1,
VP4 and VP6. No other minor protein bands were
detected in the preparation of cores. These minor
proteins which comigrate during SDS-PAGE with the
major in vitro translation products of AHSV genome
segments 1, 4 and 9 respectively (O'Hara et al.,
unpublished results) were also present only in infected
cell lysates. In each of these respects these proteins are
similar to the directly comparable minor core proteins
VP1, VP4 and VP6 of BTV (Mertens et al., 1984, 1987).
The purified AHSV particles and VP7 crystals have
provided a direct and indirect source of useful diagnostic
reagents (both antigens and antibodies) which are free
from host cell protein contamination. VP7 is a major
group antigen in both BTV and AHSV (Gumm &
Newman, 1982; Oldfield et al., 1990) and purified VP7
crystals have been used as a reagent for group-specific
tests. Antisera raised against virus particles, ISVPs, cores
or VP7 have also been used in serogroup-specific ELISAs
which are suitable for the detection of both AHSV and
AHSV-specific antibodies in field samples and infected
tissue culture (Hamblin et al., 1991, 1992).
Disc-shaped aggregates of expressed AHSV-4 VP7
have been observed in recombinant baculovirus-infected
insect cells (Chuma et al., 1992), although their fine
structure was not examined. The flat, usually hexagonal
crystals of VP7 produced by AHSV-9 infection of BHK21 cells have a highly ordered hexagonal lattice consistent
with a dimeric or trimeric subunit structure. X-ray
diffraction studies using crystals of recombinant baculovirus-expressed VP7 from BTV have indicated that the
protein is grouped in a trimeric structure (Basak et al.,
1992; Hewat et al., 1992a, b) and in vitro assembly
studies have demonstrated both dimer and trimer
formation (Wade-Evans & Mertens, 1993). The hexagonal arrangement observed in the VP7 crystal lattice
appears to have direct structural similarity to the
segmented, ring-shaped capsomers that are visible on the
outer surface of the AHSV core [shown here and
previously reported by Oellermann et al. (1970)] and
which are also composed ofVP7 (Hewat et al., 1992a, b).
The major inner capsid protein of rotavirus (VP6), which
forms the ring-shaped capsomers on single-shelled
particles, has also been shown to polymerize, under
appropriate conditions, to form a hexagonal lattice
(Ready & Sabara, 1987). The functional significance of
the VP7 crystals, if any, remains to be determined. It is
possible that these crystals represent a by-product, rather
than an essential component of the AHSV replication
process.
The purification methods described here represent
significant improvements over those previously published
for AHSV both in terms of yield and purity. They also
differ from those previously published for BTV (Mertens
et al., 1984) in the nature of the centrifugation conditions
used for the purification of two intact virus particle
forms (plus and minus NS2) and the isolation of VP7
crystals. The purified particles have provided a useful
source of pure AHSV RNA for molecular studies,
including gene assignment, cloning and sequencing
studies. They also provide material of sufficiently high
purity and in the relatively large quantities that are
essential for our current crystallization and X-ray
diffraction studies of the different particle types (Basak et
al., 1992).
We wish to thank Dr C. Bostock and Dr A. M. Q. King for critical
evaluation of the manuscript, Mr B. Clark for production of figures
and Mrs C. Chisholm for preparation of the manuscript. This study has
been jointly supported by the Ministry of Agriculture, Fisheries and
Food and by the Commission of the European Community contract,
African Horsesickness in Europe (contract number 8001-CT91-0211).
References
BASAK, A., GRIMES, J., BURROUGHS,J. N., MERTENS, P. P. C., ROY, P.
& STUART, D. I. (1992). Crystallographic studies on the structure of
the bluetongue virus inner capsid. In Bluetongue, African Horse
Sickness and Related Orbiviruses, Proceedings of the Second International Symposium, pp. 483490. Edited by T. E. Walton & B. I.
Osburn. Boca Raton: CRC Press.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09
African horsesickness virus purification
BONNER, W. M. & LASKEY, R. A. (1974). A film detection method for
tritium-labeled proteins and nucleic acids in polyacrylamide gels.
European Journal of Biochemistry 46, 83-88.
BREMER, C. W. (1976). A gel electrophoresis study of the protein and
nucleic acid components of African horse sickness virus.
Onderstepoort Journal of Veterinary Research 43, 193-200.
CHUMA, T., LE BLOIS, H., SANCHEZ-VIZCAiNO,J. M., DIAZ-LAVlADA,
M. & ROY, P. (1992). Expression of the major core antigen VP7 of
African horsesickness virus by a recombinant baculovirus and its use
as a group-specific diagnostic reagent. Journal of General Virology
73, 925 931.
GORMAN, B. M., TAYLOR, J. c~ WALKER, P. J. (•983). Orbiviruses. In
The Reoviridae, pp. 287-358. Edited by W. K. Joklik, H. FraenkelConrat & R. R. Wagner. New York: Plenum Press.
GORMAN, B.M., GHALEB, H, W., HUISMANS, H., JOKLIK, W.K.,
K~4UDSON,D. L., MCPHEE, D. A., MURPHY, F. A., RITCHIE, A., ROY,
P. & TAYLOR, J. (1985). W H O / F A O working team report: molecular
virology. Bluetongue and related orbiviruses. Progress in Clinical and
Biological Research 178, 689-695.
GRUBMAN, M.J. & LEWIS, S.A. (1992). Identification and characterisation of the structural and nonstructural proteins of African
horse sickness virus and determination of the genome coding
assignments. Virology 186, 444-451.
GUMM, I. D. & NEWMAN, J. F. E. (1982). The preparation of purified
bluetongue virus group antigen for use as a diagnostic reagent.
Archives of Virology 72, 83-93.
HAMBLIN, C., GRAHAM, S.D., ANDERSON, E.C., CROWTHER, J.R.
(1990). A competitive ELISA for the detection of group-specific
antibodies to African horse sickness virus. Epidemiology and Infection
104, 303-312.
HAMBLIN, C., MERTENS, P. P. C., MELLOR, P. S., BURROUGHS,J. N. &
CROWTHER, J.R. (1991). A serogroup specific enzyme-linked
immunosorbent assay for the detection and identification of African
horse sickness virus. Journal of Virological Methods 31,285 292.
HAMBLIN, C., ANDERSON, E.C., MELLOR, P.S., GRAHAM, S.D.,
MERTENS, P. P. C. & BURROUGHS,J. N. (1992). Detection of African
horse sickness virus antigens and antibodies in young equidae.
Epidemiology and Infection 108, 193-201.
HEWAT, E. A., BOOXH,T. F., LOUDO~, P. T. & RoY, P. (t992a). Three
dimensional reconstruction of baculovirus expressed bluetongue
virus core-like particles by cryoelectron microscopy. Virology 189,
10-20.
HEWAT, E. A., BOOTH,T. F. & RoY, P. ( 1992 b). Structure of bluetongue
virus particles by cryoelectron microscopy. Journal of Structural
Biology 109, 61 69.
HOLMES, I.H. (1991). Family Reoviridae. In Classification and
Nomenclature of Viruses, Fifth Report of the International Committee
on Taxonomy qf Viruses, pp. 186-199. Edited by R. I. B. Francki,
C. M. Fauquet, D. L. Knudson & F. Brown. New York: SpringerVerlag.
HUISMANS, H. (1979). Protein synthesis in bluetongue virus infected
cells. Virology 92, 385 396.
HUISMANS, H. & EATON, B. (1992). Working team report on molecular
virology. Identify nucleic acid sequences and antigens/epitopes for
use in diagnostic procedures and molecular epidemiology. In
Bluetongue, African Horse Sickness and Related Orbiviruses,
Proceedings of the Second International Symposium, pp. 982-989.
Edited by T. E. Walton & B. I. Osburn. Boca Raton: CRC Press.
HUISMANS, H., VAN DER WALT, N. T., CLOETE, M. & ERASMUS, B. J.
1857
(1987a). Isolation of a capsid protein of bluetongue virus that
induces a protective immune response in sheep. Virology 157,
172 179.
HUISMANS, H., VAN DHK, A. A. & BAUSKIN, A. R. (1987b). In vitro
phosphorylation and purification of a nonstructural protein of
bluetongue virus with affinity for single-stranded RNA. Journal of
Virology 61, 3589 3595.
K~RBER, G. (1931). Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche. Archly fiir Experimentelle Pathologie
und Pharmakologie 162, 480-483.
LAEMMLI, U.K. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, London 227,
680 685.
LAVIADA, M.D., ARIAS, M. & SANCHEZ-VIZCAiNO, J.M. (1993).
Characterization of African horsesickness virus serotype 4-induced
polypeptides in Vero cells and their reactivity in Western blotting.
Journal of General Virology 74, 81-87.
MECUAM, J. O., DEAN, V. C. & JocmM, M. M. (1986). Correlation of
serotype specificity and protein structure of the five U.S. serotypes of
bluetongue virus. Journal of General Virology 67, 2617 2624.
MERTENS, P. P. C., BROWN, F. & SANGAR, D. V. (1984). Assignment of
the genome segments of bluetongue virus type 1 to the proteins which
they encode. Virology 135, 207-217.
MERTENS, P.P.C., BURROUGHS, J.N. & ANDERSON, J. (1987).
Purification and properties of virus particles, infectious sub-viral
particles and cores of bluetongue virus serotypes 1 and 4. Virology
157, 375-386.
MERTENS, P. P. C., BURROUGHS,J. N. & MELLOR, P. S. (1993). Infection
of Culicoides vector species by different orbivirus particle types. In
IXth International Congress of Virology, Glasgow, Abstracts (W21-7).
NEL, L. H., PICARD, L. A., RoY, P. & E1UISMANS,H. (1992). Studies on
the nucleic acid sequence and expression of African horse sickness
virus and epizootic hemorrhagic disease virus NSl-encoding genes.
In Bluetongue, African Horse Sickness and Related Orbiviruses,
Proceedings" of the Second International Symposium, pp. 369-375.
Edited by T. E. Walton & B. I. Osburn. Boca Raton: CRC Press.
OELLERMANN, R.A., ELM, H.J. & ERASMUS, B.J. (1970). Characterisation of African horse sickness virus. Archiv fiir die gesamte
Virufforschung 29, 163 174.
OLDFIELD, S., ADACHI, A., URAKAWA, T., H1RASAWA, T. ~; ROY, P.
(1990). Purification and characterization of the major group-specific
core antigen VP7 of bluetongue virus synthesized by a recombinant
baculovirus. Journal of'General Virology 71, 2649 2656.
READY, K. F. M. & SABARA, M. (1987). In vitro assembly of bovine
rotavirus nucleocapsid protein. Virology 157, 189-198.
VAN STADEN, V., THERON, J., GREYLING, B. J., HUISMANS, H. & NEL,
L. H. (1991). A comparison of the nucleotide sequences of cognate
NS2 genes of three different orbiviruses. Virology 185, 500 504.
VERWOERD, D.W. (1969). Purification and characterisation of bluetongue virus. Virology 38, 203 212.
VERWOERD, D. W., ELS, H.J., DE VILLiERS, E.-M. & HUISMANS, H.
(1972). Structure of the bluetongue virus capsid. Journal of Virology
10, 783-794.
WADE-EVANS, A. M. & MERTENS, P. P. C. (1993). Analysis of interactions between structural and non-structural bluetongue virus
proteins translated in vitro. In IXth International Congress of
Virology, Glasgow, Abstracts (P21-19).
(Received 29 November 1993; Accepted 25 February 1994)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 01:15:09