J. gen. Virol. (I978), 39, 231-242
23I
Printed in Great Britain
Prevention of D e a t h in Semliki Forest Virus-infected Mice by
Administration o f Defective-interfering Semliki Forest Virus
By N. J. D I M M O C K AND S. I. T. K E N N E D Y *
Department of Biological Sciences, University of Warwick, Coventry CV4 7AL
(Accepted t7 November I977)
SUMMARY
Adult mice inoculated with Semliki Forest virus (SFV) were protected from a
lethal infection of the central nervous system by intranasal administration of
defective-interfering (DI) SFV. DI SFV was prepared by eight passages at high
m.o.i, in BHK zI cells. Mice were treated with unpurified, unconcentrated tissue
culture fluid which had been u.v.-irradiated to inactivate the infective virus present.
Prevention of death was maximal when the DI virus was administered simultaneously with the infecting inoculum, and under the same conditions multiplication of infective virus in the brains of treated mice was reduced by IC-fold. It was
shown that DI SFV was propagated in mouse brains following intranasal inoculation and it was concluded that protection was brought about through the intrinsic
interfering capacity of the DI virus.
INTRODUCTION
Defective-interfering (DI) virus is produced during the multiplication of many different
viruses (Huang, r973). DI viruses are non-infective and multiply only with the assistance of
'standard' infective virus, hereafter referred to as standard virus. This results in decreased
production of infective progeny or ' interference'. DI virus is dependent upon and interferes
only with the multiplication of the parental virus or a closely related strain.
In the majority of instances DI viruses have been generated in cell culture but DI forms
of influenza (von Magnus, I95 I), Rift Valley fever (Mims, t956), lymphocytic choriomeningitis (Lehmann-Grube, I97 0, vesicular stomatitis and rabies viruses (Holland & Villarreal,
I975) occur in vivo in embryonated chickens' eggs or in mice. There is speculation that DI
viruses may have a biological role in the modulation of virus diseases (Huang & Baltimore,
I97o), but the only evidence for this has been the protection of mice lethally infected with
vesicular stomatitis virus (VSV) by very high doses of DI VSV in adult and neonatal mice
(Holland & Doyle, ~973; Doyle & Holland, I973; Holland & Villarreal, I975). However,
DI VSV could not be produced or perpetuated in adult mice although the protection
afforded by DI VSV was greater in adults than in the neonates.
Recent studies have provided much information about the molecular biology of the
defective-interfering particles of Semliki Forest virus (DI SFV; Bruton & Kennedy, i976;
Bruton et al. I976). DI SFV produced in BHK 2I cells has RNA which lacks a considerable
portion of the central genomic region of standard SFV and interferes efficiently with the
* Present address: Department of Eiology, University of California, San Diego, La Jolla, California
92o93, U.S.A.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
232
N.J. DIMMOCK
A N D S. I. T. K E N N E D Y
multiplication of standard SFV in B H K 2I cells. This report describes experiments which
investigate the capacity of DI SFV to moderate the disease caused by intranasal infection
of adult mice by standard SFV.
METHODS
Virus and cells. Three times plaque-purified ts ÷ SFV (standard virus) and the extracellular
fluids harvested after eight passages in B H K 2i cells at a constant multiplicity of 5° p.f.u./
cell (passage 8 virus) were those described by Bruton & Kennedy (I976). These virus
preparations were clarified by centrifuging at IOOOOg for 3o rain at 4 °C and stored at
- 7 0 °C. Monolayer cultures of B H K 2I cells, clone I3, and of primary chick embryo
fibroblasts were grown in 5 cm plastic Petri dishes as described by Morser et al. (I973).
Infectivity assays. The infectivity of SFV-containing samples was determined by plaque
assay on chick embryo fibroblasts as described before (Kennedy & Burke, I972) except
that o.o2 ~ DEAE-dextran was included in the agar overlay and the incubation temperature
was 3° °C. These measures increased the size of plaques.
Mice. Random-bred mice of the Porton strain were purchased from Allington Farm,
Porton and were used about I week after delivery. Male mice weighing about 2I g were
used at approx. 5 weeks of age.
Infection. All mice were infected intranasally. Individual mice were anaesthetized with
ether until just unconscious. A measured volume (20/zl) was delivered with a micropipette
on to the nose which the mice then sniffed up into both nostrils. Mice were examined twice
daily for signs of sickness, paralysis or death up to Io days after infection. When tissue
samples were required mice were killed with ether. Blood was obtained from the heart
before removing other organs and allowed to clot overnight at 4 °C. Lungs were washed in
PBS. Olfactory bulbs were dissected out after removing the top of the skull. Lungs, olfactory
bulbs and brain (less olfactory bulbs) were dispersed in medium I99 containing 5 % calf
serum and stored at - 7 0 °C.
Ultraviolet (u.v.) irradiation. One ml portions of virus suspension in 5 cm plastic Petri
dishes were continuously agitated at room temperature in a position zo cm beneath an
ultraviolet lamp (Anderman and Co. Ltd, London) for the times specified in individual
experiments. The emission wavelength of the lamp was 256 nm and the dosage was 6o ergs/s/
cmL After irradiation samples were stored at - 7 o °C.
Measurement of virus-specified RNA and of the interference index of DIparticles. Preparations containing standard virus or DI particles, which on occasions were supplemented with
standard virus, were used to infect B H K monolayer cultures at a m.o.i, of about IO p.f.u./
cell. The monolayers were previously labelled for i6 h with 2 #Ci/culture of ~H-uridine.
After adsorption for I h at 37 °C the inocula were replaced with Glasgow modified MEM
containing one-tenth the normal phosphate concentration, I #g/ml actinomycin D, 2o mMHEPES (pH 7'4) and 2 ~o dialysed calf serum. After a further I h at 37 °C the fluids were
removed, the cultures washed twice with warmed phosphate-free Earles's solution containing I #g/ml actinomycin D, ao mM-HEPES (pH 7'4) and 2 ~o dialysed calf serum and
each culture incubated in 2 ml of this medium containing o'5 mCi 32P-orthophosphate/ml
for a further 6 h. The cultures were then washed three times with ice-cold PBS, once with
5° mM-tris (pH 7"4) containing Ioo mM-NaC1 and I mM-EDTA (TNE) and the cell cytoplasm solubilized by incubating the monolayers under z ml T N E containing 2 % (w/v)
Triton N I o I for 2 min at room temperature. Each cytoplasmic extract was carefully
removed (leaving the nuclei attached to the plastic dish), made to 2 % in SDS and a total
nucleic acid extract was prepared as described by Bruton & Kennedy (i975). The extract
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
Treatment oJ" infected mice with D I S F V
233
was alcohol precipitated at - 20 °C in the presence of IOO #g carrier t-RNA. After recovery
from alcohol by centrifuging and drying in a stream of N2, the extracts were counted and
approx, equal amounts of each extract (as measured by aH-uridine radioactivity) were
analysed on a 1.7 % (w/v) acrylamide + o. 13 % bisacrylamide + o.5 % agarose gel which
was prepared and run as described before (Clegg & Kennedy, 1974). After electrophoresis
the gels were dried under vacuum at 9o °C and autoradiographed on Kodirex X-ray film
for I to 3 days. The autoradiographs were then scanned using a Joyce-Loebl microdensitometer and the total areas under the RNA peaks corresponding to (A) standard virusspecified R N A species (42S, 38S, 33S and 26S RNA) and (B) DI particle species (DIssl
and Dlssz) were determined (Bruton et al. 1976). The ratio B/A will be referred to as the
interference index (but does not necessarily reflect any interference occurring in vivo).
Radioimmunoassay. A modification of the technique described by Dalrymple et al. (1972)
was used. Specifically, IO #1 portions of 35S-methionine labelled purified SFV [approx.
o'oo5 #g, 15oo ct/min and prepared as described by Kennedy (I974) but using methioninedeficient culture medium] was mixed with twofold dilutions in triplicate of test antigen
(5o #1) and 5o #1 of a iooo-fold dilution of SFV-specific, kaolin adsorbed antibody prepared
in rabbits as described before (Kennedy, 1974). This contained antibody to the envelope
but not the core antigens. All dilutions were done using PBS lacking Ca 2+ and Mg 2+ ions.
After incubation for I h at 37 °C, 5o #1 of goat anti-rabbit IgG was added to each sample
and incubation continued for I h at 37 °C and then for 16 h at 4 °C. This was centrifuged
at 7ooog for IO min at 4 °C and IOO #1 of each supernatant fluid was removed for counting,
as was the remaining mixture of supernatant fluid and precipitate. The extent of precipitation was calculated from the difference in radioactivity between these two samples based
upon the mean of the triplicate assays of each test antigen dilution. All samples were dissolved in o.I ml soluene-35o (Packard Instrument Co., Illinois, U.S.A.) and counted as
previously described (Atkins et al. I974b). The dilution of test antigen which gave 5o%
precipitation was determined by interpolation. Controls were included in all assays by
omission from the complete assay mixture of (i) test antigen, (ii) specific antibody and (iii)
goat anti-rabbit IgG. In another control the complete mixture was prepared with preimmune antibody in place of specific antibody. In no case was non-specific precipitation
greater than 5 %. The relationship between total antigenic mass and infectivity of SFV is
shown in Fig. I. From this graph the total antigenic mass of virus preparations either
containing or devoid of DI particles was converted into 'p.f.u. equivalents' (see below).
Preliminary experiments showed that ultraviolet irradiation of standard SFV for 200 s
did not measurably alter its ability to combine with antibody.
Since the infectivity titre of standard virus was known, the dilution which reduced precipitation of labelled SFV by 5o % could also be expressed as p.f.u. In order to equate the
antigen present in the passage 8 virus with standard virus, the dilution of passage 8 virus
which reduced precipitation by 5o% was determined and expressed in terms of 'p.f.u
equivalents'. Thus passage 8 virus reduced precipitation of the radio-labelled SFV by 5o %
at a dilution 35o-fold less than standard virus and therefore contained 35o-fold fewer
'p.f.u. equivalents' than standard virus.
RESULTS
Multiplication of virus after intranasal inoculation
Three mice were sacrificed at intervals after infection with standard SFV and the infectivity present in serum, lungs, brain (less olfactory bulbs) and olfactory bulbs was titrated.
Fig. 2 shows the amount of virus/g of each tissue sampled. Virus was not detected in the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
N . J . D I M M O C K AND S . I . T . K E N N E D Y
234
I
I
I
I
I
1
1
2
~
3
I
4
~9
I
I
I
~8~
0
" 7F
i 6l
>-,
O
Q
~d 41
¢
e 3
>-
O~O
.~ 2
"5
[,,.
e-
I
6
I
I
7
8
Infectivity (log 0 p.f.u./ml)
I
9
Fig. I
I
Days
Fig. 2
Fig. ~. Relationship between total antigenic mass and infectivity of SFV. A fixed amount of
asS-methionine labelled purified SFV was mixed, in triplicate, with dilutions of partially purified
unlabelled SFV (Kennedy, I974) and assayed for infectivity and for total antigenic mass as
described in Methods. The coordinates of each point are the mean values of the respective assay.
Fig. z. Multiplication of SFV in mice infected intranasally with 2 x io5 p.f.u. Three mice were killed
at the indicated times and various tissues excised and pooled. Infectivity titres are presented as
log10 p.f.u, per g tissue or per ml serum. 0 - - 0 , Brain; O - - O , lung; A - - A , olfactory bulbs;
A--/~, serum. Arrows show that no p.f.u, was detected.
lungs until I z h p o s t infection (p.i.) and then r e m a i n e d at lO2 to Io z p . f . u . / g tissue. T h e
level o f v i r u s / m l serum was similar b u t was n o t detected until 3 days p.i. I n c o n t r a s t there
was a r a p i d increase to 1o 7 p . f . u . / g in the olfactory bulbs between I2 a n d 24 h p.i. Infectivity in the b r a i n rose m o r e slowly a n d only exceeded t h a t o f the olfactory bulbs after
3 days. Similar results were also o b t a i n e d with i n t r a n a s a l a d m i n i s t r a t i o n o f Sindbis virus
which suggests t h a t in P o r t o n mice these viruses a r e t r a n s m i t t e d f r o m the nose to the olfact o r y b u l b a n d t h e n to the brain. A n alternative i n t e r p r e t a t i o n t h a t virus infects some other
o r g a n first, enters the b l o o d stream a n d then infects the C N S was n o t s u p p o r t e d b y experiments in which mice were infected via the tail vein, The resulting p a t t e r n o f infection differed
f r o m t h a t following intranasal a d m i n i s t r a t i o n since virus a p p e a r e d first in the serum a n d
then in the b r a i n a n d lastly in the olfactory bulbs, essentially as r e p o r t e d b y Bradish et al.
(1975).
I n the succeeding experiments we a t t e m p t e d to p r o t e c t SFV-infected mice with D I SFV.
Since D I S F V does n o t direct the synthesis o f virus p r o t e i n s (Bruton et al. 1976) and therefore consists o f virus c o m p o n e n t s m a d e entirely b y s t a n d a r d S F V , it is likely t h a t D I S F V
follows the same r o u t e o f infection as the s t a n d a r d virus. Hence we a d m i n i s t e r e d D I S F V
by the intranasal r o u t e as it seemed to offer a direct access to the m a i n site o f virus multiplication in the brain.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
Treatment o f infected mice with DI S F V
t
1
I
I
I // I
235
I
m
100
•
8O
u.d
d.
e-
60 E
,2
"6
40
1[
I
I
f
20
40
60
I
80
Time (s)
I ,,/ I
100
I
16 0
20
Fig. 3. Effect of u.v. irradiation on the infectivity of SFV and on the interference index of passage
8 SFV. Separate ~ ml portions of SFV and passage 8 SFV were irradiated for various times as
described in Methods. Each irradiated SFV sample (]oo #1) was then plaque assayed (O--O) and
from the results of the plaque assay standard virus was added to half of each irradiated sample of
passage 8 virus so that its infectivity might be restored to that of non-irradiated passage 8 virus.
The interference index of these samples ( A - - A ) together with that of the remaining irradiated
samples not supplemented with standard virus ( A - - A ) was determined.
Lethality of passage 8 SFV
Preliminary experiments indicated that for standard virus there were 6 x IO2 p.f.u, per
LDso when inoculated by the intranasal route. Passage 8 virus contains D I SFV together
with about lO6 p.f.u./ml infective SFV (Bruton & Kennedy, I976 ). Titration of this material
in mice by the intranasal (or intracerebral) route gave a pattern of infection and a p.f.u.
per LDs0 similar to that of standard virus alone. Clearly this combination of standard virus
and D I SFV was not moderating the infection and could not be used in protection
experiments.
Effect of u.v.-irradiation on infectivity and
interfering capacity of DI SFV
Fig. 3 shows that infectivity of SFV falls by over 6 log10 units in 6o s. Since we proposed
to use an irradiated passage 8 preparation in protection experiments, we needed to show
that it still possessed interfering capacity. Accordingly, after irradiation the infectivity was
restored to its original value by the addition of an appropriate amount of standard virus.
The interference index for B H K cells was then measured as described in Methods. There
was no significant drop in interference index until after I8O s irradiation (Fig. 3) despite an
assertion to the contrary by Huang (x973)In the following experiments infective virus present in passage 8 SFV preparations was
inactivated by irradiation for Ioo s. This is then referred to as 'u.v. D I SFV'.
x6
VIR 39
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
236
N . J . D I M M O C K AND S. I. T. K E N N E D Y
T a b l e i. Treatment o f standard SFV-infected mice with
u.v. D I S F V or u.v. standard S F V
Days after infection
r - -
Treatment*
Symptoms
SFV
Dead
Paralysed
Sick
Well
SFV+
U.V. DI SFV
Dead
Paralysed
Sick
Well
SFV+u.v. SFV
Dead
Paralysed
Sick
Well
3"0
201"
3"9
4"2
5"0
5 .2
5'9
I
5
12
t
I
12
8
i
i
5
I
i
4
3
2o
I9
I
I
t
18
I
3
IO
I
zo
I I
20
4
4
Ii
I
4
6"9
8"9
10"2
Total
dead
20
2
1
18
I7
~6
I
I
I
16
4
16
I6
I
I
I
4
9"o
I7
I
4
4
4
3
3
* Mice were treated with u.v. DI SFV ( 2 × IO~ 'p.f.u. equivalents') or u.v. SFV at z h before infection,
at infection and 4 and 23 h afler infection. All inoculations were given intranasally under light ether anaesthesia. Control mice were inoculated with medium, with u.v. DI SFV or with u.v. standard SFV and remained
healthy.
? Number of mice inoculated.
Prevention o f death in SFV-#~fected mice by u.v. D I S F V
I n the following experiments mice were i n o c u l a t e d i n t r a n a s a l l y with Io LDs0 o f s t a n d a r d
S F V (6 x io 3 p.f.u.). G r o u p s were t r e a t e d with u.v. D I S F V (2 x lO5 'p.f.u. e q u i v a l e n t s ' ) or
an equivalent a m o u n t o f u.v. s t a n d a r d SFV, estimated by the r a d i o i m m u n o a s s a y described
in M e t h o d s . T r e a t m e n t s were given at 2 h b e f o r e infection, at infection a n d 4 a n d 23 h p.i.
T h e progress o f the disease is r e c o r d e d in T a b l e I in terms o f the n u m b e r o f sick, p a r a l y s e d
and d e a d mice. T o summarize, o f mice given s t a n d a r d S F V alone, I o o % died; o f those
given u.v. s t a n d a r d S F V in a d d i t i o n , 8 0 % died; o f those given u.v. D I S F V in a d d i t i o n ,
1 5 % died. This crucial experiment was r e p e a t e d on several occasions a n d the following
generalizations e m e r g e d : (I) a d m i n i s t r a t i o n o f u.v. D I S F V as described a b o v e always
p r o t e c t e d infected mice to the extent t h a t over 50 % survived. (2) I f infected mice t r e a t e d
with u.v. D I S F V b e c a m e ill they always died. On no occasion did a sick or p a r a l y s e d m o u s e
recover. (3) W h e n an infected m o u s e treated with u.v. D I S F V b e c a m e sick the progress o f
the disease a n d its s y m p t o m s were always within the range observed for mice given only
s t a n d a r d SFV. (4) T h e p a t t e r n o f disease in infected mice treated with u.v. s t a n d a r d S F V
differed in no way f r o m t h a t in mice given only S F V except that, on occasion, a few mice
w o u l d survive. This v a r i e d between o a n d 2o %.
W e c o n c l u d e d t h a t the D I S F V p r e p a r a t i o n was in some w a y m o d i f y i n g the lethal effect
o f S F V a n d p r o c e e d e d to investigate h o w the D I virus p r e p a r a t i o n was involved.
Effect o f u.v. D I S F V on multiplication o f standard S F V in mice
Infection a n d t r e a t m e n t s were carried out as described above. C o n t r o l mice which
received diluent, u.v. D I S F V (2 x IOn 'p.f.u. e q u i v a l e n t s ' p e r dose) o r u.v. s t a n d a r d S F V
r e m a i n e d healthy (Table 2). N o infective virus was isolated f r o m t h e m showing t h a t the
i n a c t i v a t i o n h a d been effective ( a n d that cage-to-cage t r a n s m i s s i o n d i d n o t t a k e place).
A l l mice given S F V alone h a d high virus infectivity in olfactory bulbs a n d in the rest o f
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
Treatment o f infected mice with D I S F V
237
Table 2. Treatment o f standard SFV-infected mice with u.v. DI S F V or
u.v. standard S F V : effect on virus multiplication*
Olfactory
bulb
Brain
Diluent
< I'O
< x'o
Well
u.v. D I SFV
< I'O
< I'O
Well
u.v. SFV
Treatment
State o f
health
< I-O
< I'O
SFV
7"4
7"6
6.8
7"2
9"7
8"7
9"4
7'3
Paralysed
Well
Well
Paralysed
S F V + u.v. SFV
6.8
6'9
6"7
6.2
7"7
8.6
8"7
8"3
Well
Sick
Sick
Sick
3'z
< I'O
3"5
6.8
3"3
< I'O
3"4
9"4
S F V + u . v . D I SFV
Well
Well
Well
Well
Paralysed
* All mice were inoculated intranasally as before. Infected mice received 1o LD~0. Those treated with
u.v. SFV or u.v. D I SFV (2 × IO~ 'p.f.u. equivalents ') were inoculated at 2 h before, at infection and 4 a n d
23 h after infection. Mice were killed 4 days after infection. Results are expressed as Iogz0 p.f.u./mouse.
the brain. Titres of virus from infected mice treated with u.v. standard SFV were in the
same range. At the time the experiment was terminated some mice in each of these two
categories were still healthy, some sick, while others were paralysed; these differences were
not reflected by virus infectivity in the brains. Three out of four infected mice treated with
u.v. D I SFV remained well. Brain titres were reduced by about 5 log10 units and olfactory
bulb titres by about 3 log10 units. Infectious virus could not be isolated from one mouse.
Lastly, one u.v. D I SFV treated mouse was paralysed; titres of infective virus in its olfactory bulbs and brain were typical of mice having SFV alone. We concluded that the
treatment of infected mice with u.v. D I SFV but not u.v. standard SFV could reduce virus
multiplication to a high degree.
Effect of number of doses of u.v. D I S F V on standard SFV-infected mice
Initially mice were treated with an empirically derived schedule at 2 h before, at infection
and 4 and 24 h after infection which proved to be effective. Later experiments were designed
to test the effects of each of these inocula (Table 3). The most effective single administration
of u.v. D I virus was together with the infecting virus. However, prevention of death was
more efficient with an additional pre-infection inoculation with DI virus. This treatment
proved to be as effective as the full course. DI virus administered at 4 and 24 h after standard
virus was of no benefit to the mice.
Prevention of death in mice infected with increasing amounts of
standard S F V by decreasing amounts of u.v. D I S F V
Table 4 shows that undiluted u.v. DI SFV prevented death in 8o % of mice inoculated with
io LDs0 of standard virus. However this protection was lost when u.v. DI SFV was diluted
Io-fold. Protection was also reduced when mice were given IOO LD~0 and undiluted u.v.
16-2
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
238
N.J. DIMMOCK
AND
S. I. T. K E N N E D Y
Table 3. Effect of the number of inoculations of u.v. D1 SFV on
standard SFV-infected mice*
Incidence of death or paralysis
A
r
u.v. D I SFV at:
- 2h
.
+
+
oh
.
+ 4 h
.
Day 4
+ 24 h
Day 5
Day 6
No.
%
No.
%
No.
%
56
6
I9
13
15
4
5
6
94
25
32
38
38
15
5
5
9
94
32
32
56
+
+
.
+
--
+
--
9
,
3
--
+
--
--
2
--
+
+
+
I
6
6
8
50
* G r o u p s of I6 mice were injected intranasally at time o with IO LDs0. U.v. D t SFV ( 2 x , o s ' p . f . u .
equivalents ') was administered intranasally at the times indicated.
Table 4. Prevention of death in mice infected with increasing doses of
standard SFV by decreasing doses of u.v. DI SFV
Days after infectiont
u.v. D I S F V / m o u s e
A _ _
SFV/mouse
c- ~
,-~
LDso
p.f.u. (a)
Dilution
'p.f.u.
equivalents'*
(b)
b :a
No. ~
-
-
i
4'7
Dead
4"o
Dead
5.0
Dead
5.7
Dead
%
No.
%
No.
%
No.
%
Io
66
13
87
t5
too
2
0
6
o
13
O
6'7
3
[ o
-0'7
3
o
ro
io
I0
I0
61 io 3
-
6X I&
,o °
4 × ,o 5
66'7
o
6XIO3
6X 103
IO-a
IO-2
4×IO *
4 x IO3
6'7
0"7
I00
ioo
6 x io*
6xio 4
IO°
4xIO s
IOOO
IOOO
6 x IO 5
6 × 105
I0°
-
4 × 105
I 2
'3
3
zo
3
9
8
60
53
II
12
73
80
II
I4
2o ]
2o
o
I4
93
I4
93
I5
lI
73
II
73
I2
IOO
i
8o t
20
o
II
73
'5
IOO
'5
IO0
I4
93
I3
loo
15
IOO
73
93
* 'P.f.u. equivalents' are calculated f r o m the equivalent antigenic content of u.v. D I SFV and standard
SFV estimated by r a d i o i m m u n o a s s a y (see Methods).
t Boxes show protection of >I 2 o % (3/15) mice over a period of 3 or m o r e days.
:~ G r o u p s of I5 mice were treated with u.v. D I SFV at z h before infection and at infection.
D I SFV and no protection was obtained against infection with Iooo LDso. Thus, to prevent
death by Io LDs0 we require a ratio of at least 67 'p.f.u. equivalents' D I SFV to I p.f.u.
standard SFV, while with Ioo LDs0 a ratio of 6"7 gave some protection.
Effect of varying the time of administration o]"u.v. DI SFV on
standard SFV-infected mice
There was no doubt that treatment of mice with u.v. DI SFV (2 × Io 5 'p.f.u. equivalents')
at 8, 2 or even I h before infection had no effect on the course of disease or time of death of
infected mice. Similarly u.v. D I SFV administered at I h or later after infection did not
alter the course of disease. Onset of illness was delayed by about 24 h when u.v. D I SFV
was given 3o min following infection but by day 5 the normal extent of disease was evident.
B H K cells were also pre-treated with u.v. D I SFV at various times before standard virus
was added at time o. Propagation of DI virus was measured as described in Methods.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
Treatment o f infected mice with D I S F V
239
D I virus was detected in cells inoculated with u.v. D I SFV at 2 and 4 h but not at 6, 8 or
I2 h before addition of standard virus.
Mode of action of protection by u.v. DI SFV
Interference with attachment of infectious virus ?
Prevention of death was most efficient with 67-fold excess D I SFV over standard infective
virus. If the number of attachment sites was limited it could be suggested that D I virus was
physically preventing standard virus from adsorbing to and infecting cells. However the
failure of an amount of u.v. standard SFV equivalent to D I SFV to protect mice indicated
that this was not the explanation.
Induction of interferon ?
The induction of interferon by D I SFV in chick embryo cells in culture at 24 h after
infection was abolished progressively by u.v.-irradiation and after IOO s no interferon could
be detected by the method of Atkins et al. 0974a). Mice were tested for the induction of
interferon by four intranasal inoculations of u.v. D I SFV (containing no infective virus)
at o, 2, 6 and 24 h. No interferon could be detected in serum collected at 8 and 3 x h.
The absence of detectable serum interferon did not mean that interferon was not being
produced and exerting an inhibitory effect in some localized site in the mouse. Sindbis
virus was the most suitable probe for detecting local interferon since after intranasal
inoculation Sindbis could be found at high levels in the olfactory bulbs and then the brain
with kinetics of appearance similar to those of SFV.
Sindbis and u.v. D I SFV were inoculated together and the infectivity present in the brain
and olfactory bulbs at 4 days after infection was titrated. Neither u.v. D I SFV or u.v.
standard SFV inhibited the multiplication of Sindbis virus. Together these results indicate
that interferon played no detectable part in the protection by u.v. D I SFV described in this
report.
Interference by the production of defective-interfering virus
In order to postulate that mice survived a lethal SFV infection through the interfering
property of D I SFV it is necessary to demonstrate that u.v. D I SFV is propagated in mice
inoculated intranasally. Mice were given a single dose of u.v. D I SFV (z × Io 5 'p.f.u.
equivalents') in 3 x io 6 p.f.u. Controls consisted of mice inoculated with 3 x io 6 p.f.u.
alone. Brains ( + olfactory bulbs) were harvested after 3 days and were assayed for the
presence of D I SFV by the enhanced D I SFV R N A assay described in Methods. Results
in Table 5 show that D [ SFV R N A was present only in brains of mice inoculated with u.v.
D I SFV. A control was devised to see if the assay would detect the amount of u.v. D I SFV
present in the mouse inoculum if it had not been propagated. This consisted of a mouse
brain homogenate mixed (a) with the mouse inoculum (u.v. D I S F V + s t a n d a r d SFV as
above) or (b) with z × ~o9 p.f.u, standard SFV which is the level of virus in the brain after
it has multiplied for 3 days. These mixtures were then assayed by the enhanced D I SFV
R N A system and no D I SFV R N A was detected. Thus we conclude that D I SFV was
propagated in the brains of infected mice following intranasal inoculation.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
240
N.J. DIMMOCK
AND
S. I. T. K E N N E D Y
Table 5. Synthesis of virus R N A in B H K cells infected with olfactory lobe and
brain homogenates from mice inoculated intranasally with S F V or S F V + u.v. D I SFV*
Intracellular level of virus R N A speciest
f
Standard SFV R N A s
DI RNAs
r~
42S + 38S
33S + 26S
SFV
Inoculum
40
58
28
t5
t8
9
ND
ND
ND
SFV + u.v. D I SFV
57
43
64
18
14
23
~I
76
Io
* Mice were killed 3 days after infection.
t Arbitrary units determined as described in Methods.
$ ND, not detected.
DISCUSSION
Our data suggest that standard SFV, inoculated in a drop on to the nose, infects the CNS
via the olfactory nerves. The precise cellular route by which viruses may scale the olfactory
nerve is discussed by Johnson & Mims (I968). The accounts of aerosol infection of hamsters
with SFV (Henderson et al. I967) and of mice with West Nile virus (Nit et al. I965) are also
particularly relevant here. It seems likely that intranasal infections follow specific routes
since, for example, herpes simplex virus type I appears preferentially in the trigeminal
nerve compared with the olfactory bulbs (Rabin et al. I968).
Using the intranasal route we have demonstrated that exogenously applied DI virus can
prevent the appearance of disease and eventual death which results from infection with
Semliki Forest virus. We could detect no production of interferon in response to the u.v.
DI virus administered. DI virus was propagated under the conditions used. The effect on
the immune system of administering a total of 8 × Io 5 'p.f.u. equivalents' of u.v. DI SFV
was not fully investigated. However, preliminary experiments in which mice were immunosuppressed with cyclophosphamide (2oo mg/kg) at the time of infection showed that
protection by u.v. DI SFV was as effective as in mice not given the drug (data not shown).
We conclude therefore that the prevention of disease and death was due to the intrinsic
interference of the defective-interfering virus. This was borne out by our finding that after
treatment with DI virus there was a Io 4 to ~oS-fold reduction of infectious virus in the brains
of those infected mice which showed no signs of illness. Although we were able to protect
the majority of infected mice with DI virus there were always some which succumbed. The
clinical signs, time of death and amount of infectious virus present showed that DI virus
did not ameliorate the disease to any noticeable extent. Since protected mice had low virus
titres in olfactory bulbs it seems that interference with virus multiplication took place at
this location or between the olfactory bulbs and a putative primary site of infection in the
nose.
Protection by other DI viruses in animals has been demonstrated in mice with DI (yon
Magnus) influenza virus inoculated intranasally (yon Magnus, 195I), and DI VSV inoculated intracerebrally (Doyle & Holland, I973; Holland & Doyle, ~973; Holland & Villarreal,
i975). It is not certain that protection was afforded by the interfering capacity of yon Magnus
virus as the existence and possible involvement of interferon was not known at that time.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
Treatment of infected mice with DI S F V
24I
Holland and co-workers failed to demonstrate protection via the intranasal route with
DI influenza virus, but the p.f.u. : H A ratio of infective virus found in the lungs of treated
mice was less than with untreated, showing that interference had taken place. DI VSV fails
to protect when given intranasally (Holland & Doyle, I973) whereas the same amount of
DI VSV does protect when administered intracerebrally; this probably reflects the larger
(about Io3-fold) intranasal inoculum of infective virus. It would be interesting to see if mice
could be protected with DI VSV from smaller doses of infective virus in the light of our
findings in this report.
Several points arise in a comparison between Doyle & Holland's 0973) successful protection of mice with D t VSV and also ours with DI SFV. Firstly, these findings establish
that protection by DI viruses in vivo can occur with both negative (VSV) and positive (SFV)
stranded RNA viruses. However, it is not completely clear that the protection afforded by
DI VSV is due to its intrinsic interfering capacity since the best protection in adult mice is
not accompanied by generation or perpetuation of DI virus (Holland & Villarreal, I975).
Secondly, while Io 1° purified DI particles were required per mouse for protection in VSV
infections (Holland & Doyle, 1973; Holland & Villarreal, I975), we were able to use unpurified DI SFV present in tissue culture fluids. Thirdly, protection has now been demonstrated by both the intracerebral route (Holland and co-workers) and the intranasal route.
Although the latter cannot be described as a natural route of infection, apriori it causes less
trauma and approximates more closely to natural infection than crushing a hole through the
skull!
A major problem in the therapeutic use of DI virus is introducing it into cells containing
multiplying virus. This is probably the reason why we failed to propagate DI SFV by injecting DI virus into the brains of mice when over IO~ p.f.u, were present (data not shown).
The problem is aggravated if the infecting virus departs from the initial superficial site of
injection. This was apparently the reason why we failed to obtain protection with DI SFV
administered after infection. Prophylaxis was also ineffective although this may reflect degradation of the DI virus since 'prophylaxis' of B H K cells was limited to 4hprior to infection.
We feel that the small amount of DI virus required for protection, the use of u.v.-irradiation to remove infective virus and the use of unpurified tissue culture fluids encourages the
investigation of the possible therapeutic use of DI virus in situations where the site of infection is known and there exists a way of delivering the DI virus to that site. The convenience
of intranasal/aerosol administration suggests that respiratory infections might be particularly suitable for study.
We thank W. J. Ball, A. S. Carver, Gail Duffy, Celia Ellis, W. H. Moore and Caroline
Stark for their various contributions to this work. Maggie Colby performed the interferon
assays. Actinomycin D was a gift from Merck, Sharp and Dohme, Hoddesdon, Herts.
REFERENCES
ATKINS, G. J., JOHNSTON, M. D., WESTMACOTT, L. M. & BURKE, D. C. (I974a). I n d u c t i o n o f interferon in chick
cells by temperature-sensitive m u t a n t s of Sindbis virus. Journal of General Virology 25, 381-39o.
ATKINS, G. J., SAMUELS,J. & KENNEDY, S. I. T. (1974 b). Isolation a n d preliminary characterization o f temperature-sensitive m u t a n t s of Sindbis virus strain AR339. Journal of General Virology 25, 371-38o.
BRADISH, C. J., ALLNER, K. & EITZGEORGE, R. 0975). I m m u n o m o d i f i c a t i o n a n d the expression of virulence in
mice by defined strains of Semliki Forest virus: the effects o f cyclophosphamide. Journal of General
Virology 28, 225-237.
BRUTON, C. J. & KENNEDY, S. I. "L (I975). Semliki F o r e s t virus intracellular R N A : properties of the multis t r a n d e d R N A species a n d kinetics o f positive a n d negative s t r a n d synthesis. Journal of General Virology
28, 111--I27.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58
242
N.J. DIMMOCK
AND
S. I. T. K E N N E D Y
RRUTON, C. J. & KENNEDY, S. ~. T. 0976). Defective-interfering particles o f Semliki Forest virus: structural
differences between s t a n d a r d virus a n d defective-interfering particles. Journal of General Virology 3x,
383-395.
BRUTON, C, J., PORTER, A. & KENNEDY, S. I. T. (I976). Defective-interfering particles o f Semliki Forest virus:
intracellular events during interference. Journal of General Virology 3 x, 397-4t6.
CLEGG, J. C. S. & KENNI~DY, S. I. T. (I974). Polyadenylic acid sequences in the virus R N A s]zecies o f ceils
infected with Semliki Forest virus. Journal of General Virology 22, 331-345.
DALRYMPLE, J. M., TERAMOTO, A. Y., CARDIFF, R. D. & RUSSELL, P. K. (1972). R a d i o i m m u n e precipitation o f
g r o u p A arboviruses. Journal of Immunology xo9, 426-433.
DOYLE, M. & HOLLAND, J. J. (I973). Prophylaxis a n d i m m u n i s a t i o n in mice by use of virus-free defective T
particles to protect against intracerebral infection by vesicular stomatitis virus. Proceedings of the
National Academy of Sciences of the United States of America 7o, 21 o5-2 Io8.
HENDERSON, D. W., PEACOCK, S. & RANDLES, W. J. (1967). O n the pathogenesis of Semliki Forest virus (SFV)
infection in the hamster. British Journal of Experimental Pathology 48, 228-234.
hOLLAND, J. J. & DOYt~E, M. (I973). A t t e m p t s to detect h o m o l o g o u s autointerference in vivo with influenza
virus a n d vesicular stomatitis virus. Injection and Immunity 7, 526-53 I.
HOLLAND, J. S. & WLLARREAL, L. P. (1975). Purification o f defective interfering T particles o f vesicular s t o m a titis a n d rabies viruses generated in rive in brains of n e w b o r n mice. Virology 67, 438-449.
hq3ANG, A. S. (I973). Defective interfering viruses. Annual Review qf Microbiology 27, I o I - t ~ 7 .
rIt~ANG, A. S. & BALTIMORE, 1) (1970). Defective viral particles a n d viral disease processes. Nature, London
2 2 6 , 325-327.
JOHNSON, R. T. & MIMS, ¢. A. C. (I968). Pathogenesis o f viral infections o f the n e r v o u s system. New England
Journal of Medicine 278, 23-3o.
KENNEDY, S. L T. (I974). T h e effect o f e n z y m e s on structural a n d biological properties o f Semliki F o r e s t
virus. Journal of General Virology 23, 129-I43.
KENNEDY, S. I. T. & BURKE, O. C. (I972). Studies on the structural proteins o f Semliki Forest virus. Journal of
General Virology x4, 87-98.
LEHMANN-GRUBE, F. (197I). L y m p h o c y t i c choriomeningitis virus. Virology Monographs xo, J-r73.
MIMS, C. A. C. (I956). Rift Valley fever in mice. IV. Incomplete virus: its p r o d u c t i o n a n d properties. British
Journal of'Experimental Pathology 37, I29-I43.
MORSER, M.J., KENNEDY, S. I. T. & run, RE, D. C. (I973). Virus-specified polypeptides in cells infected with
Semliki Forest virus. Journal of General Virology 2x, I9-29.
NIR, v., BEEmER, A. & GOLDWASSER, R. A. 0965). W e s t Nile virus infection in mice following exposure to a
viral aerosol. British Journal of Experimental Pathology 46 , 443-449.
RAB~N, e. R., JENSON, A. ~. & MELNICK, J. ~. (~968). Herpes Simplex in mice: electron m i c r o s c o p y of neural
spread. Science x6~, I26-127.
y o n MA6NtrS, P. (~95D" P r o p a g a t i o n o f the P R 8 strain o f influenza virus in chick embryos. III. Properties
o f the incomplete virus p r o d u c e d in serial passages o f undiluted virus. Aeta Patholegica et Microbiologica Scandinavica ~9, ~56-~ 8~.
(Received 28 June I 9 7 7 )
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 20:41:58