JOURNAL OF INTERFERON RESEARCH 7:17-27 (1987)

Mary Ann Liebert, Inc., Publishers

The Effect of Interferon Treatment

in Rabies Prophylaxis in Immunocompetent,
Immunosuppressed, and Immunodeficient Mice
RUGIMAR MARCOVISTZ, PEDRO MANUEL LEAL GERMANO,1
YVES RIVIÈRE, HENRI TSIANG, and ARA G. HOVANESSIAN

ABSTRACT
The development of rabies is modulated by many interacting factors, most of which are
dependent on the host immune response. For this reason, we studied the action of inter¬
feron (IFN) treatment on street rabies virus infection in mice, immunocompetent or im¬
munosuppressed with cyclophosphamide. In immunocompetent mice, paralysis of hind
limbs is the first symptom characteristic of rabies disease before weight loss and general
prostration leading to death. Paralysis does not occur in immunosuppressed mice, which
develop a shaggy hair and eventually lose weight and die. Administration of interferon
(10s units, intraperitoneally) 1 h after virus inoculation and every 24 h led to a delay in
the onset of first disease signs, but in general did not rescue immunocompetent or immu¬
nosuppressed mice from death. In both types of mice, rabies virus production in the
brain was reduced by 1 log in response to IFN treastment. In immunocompetent mice
treated with IFN, there was a significant increase of antibody synthesis against rabies
virus. As expected, antibody synthesis in immunosuppressed mice was almost negligible.
However, in mice treated with IFN and cyclophosphamide there was still significant anti¬
body synthesis specific for rabies virus. IFN administered intravenously, subcutaneously,
or intraperitoneally crosses the blood-brain barrier to cause enhanced levels of the two
double-stranded RNA-dependent enzymes, the protein kinase and 2',5'-oligoadenylate
(2-5A) synthetase in the brain. However in spite of this effect, IFN treatment seems to be
unable to prevent the evolution of rabies disease in immunocompetent and immunosup¬
pressed mice. Since the suppressing effect of cyclophosphamide is nonselective on both
the cellular and humoral immune responses of mice, we investigated the action of IFN in
rabies virus-infected athymic nude mice, which lack cells. Athymic nude mice infected
with street rabies virus become cachectic and die without any apparent symptom of paral¬
ysis of hind limbs. IFN treatment (administered as above) protected nude mice against
rabies infection. Three months after virus inoculation and 2 months after the end of IFN
treatment, 7 of 8 IFN-treated mice remained in perfect health. These results illustrate that
the efficacy of IFN treastment against the evolution of rabies disease in mice is dependent
on the suppression of the T-cell-mediated immune response of the host.

Département de Virologie, Institut Pasteur, 75724 Paris Cedex 15, France.

'P.M.L.G. had a fellowship of Brazilian Government (CNPg).

17
 MARCOVISTZ ET AL.

INTRODUCTION

is usually transmitted by animal bites; it migrates from this site through the periph¬
Rabies
virus
eral nerves to reach the central nervoussystem, where the virus starts to replicate.<u Some viral
replication might even take place in striated muscle cells at the site of virus inoculation, because it
has been demonstrated in hamsters infected with street rabies virus.(2) The production of interferon
(IFN) has been detected as early as 24 h after rabies virus inoculation and throughout the virus infec¬
tion.(3_6) The disease usually progresses toward paralysis and death. Thus, rabies virus infection
takes over in spite óf the production of IFN early after virus inoculation. This is perhaps due to the
fact that the amount of IFN produced during rabies virus replication at the site of inoculation is not
high enough to stop virus replication and its spread through the peripheral nerves. For this reason,
exogenous IFN has been suggested as a therapeutic agent to treat rabies virus infection. In vari¬
(7"13)

ous animal species, it has been shown that rabies disease can be prevented by exogenous IFN or by
inducers of IFN.<14_16> For example, Postic and Fenje'81 reported that IFN treatment protected rab¬
bits against rabies if administered simultaneously with the virus. However, when IFN was given 24 h
after virus inoculation, then only a prolonged incubation period was noted. In another study in rab¬
bits, Ho et al. '*> emphasized the route of administration of IFN. Daily intramuscular or intravenous
injections of IFN up to 3 weeks gave less favorable results than when IFN was given both intramus¬
cularly and intravenously at the time of rabies virus challenge. Cynomolgus monkeys infected with
rabies virus were protected by repeated intramuscular administration of human leukocyte IFN be¬
ginning 24 h after virus inoculation.(10_12) Administration of IFN after appearance of clinical symp¬
toms did not influence the progress of the disease.( 10) Daily administration of human IFN intramus¬
cularly near the site of virus inoculation protected 7 out of 10 monkeys.( 12> All these results illustrate
that IFN might be effective in rabies prophylaxis. However, some data are controversial with re¬
ports of partial protection by IFN or no protection at all.(17) It has become evident now that IFN
treatment is useless once the replication of rabies virus has started in the central nervous system. On
the other hand, early administration of IFN leads to a significant delay in the appearance of first
clinical symptoms which accompany rabies infection.
Rabies virus infection in mice and in hamsters results in the production of IFN.'2"61 We have pre¬
viously shown that two peaks of circulating IFN are detectable during rabies virus infection: an early
peak 24-48 h after inoculation followed by another peak late in the infection. The production of the
first peak of IFN is due to the response of mice to rabies virus at the site of inoculation, whereas the
second peak of IFN is due to the production of high levels of IFN in the brain in response to rabies
virus replication.'5 6) IFN in the brain and in the circulation of rabies-virus infected mice is of /ß
type. The first peak of IFN may play a role in the onset of the rabies disease because its neutraliza¬
tion with anti-mouse IFN-a//3 globulin accelerates the development of the disease.<6) Rabies virus-
infected mice die in spite of IFN produced during this infection. Thus, it might be that IFN is pro¬
duced in insufficient amounts or is produced too late in the infection to exert an efficient antiviral
effect in the brain. It should be emphasized also that the development of rabies is modulated by
many interacting factors such as the host immune response. For example, the occurrence of paraly¬
sis is dependent on the efficiency of an immune response as it has been shown by experiments using
athymic nude mice"" or cyclophosphamide-treated mice."181 However, the lethality mechanisms
might be the consequence of neuronal function impairment.119 20)
To dissociate the specific role of IFN from that of the overall immune response, we investigated
the action of exogenous IFN in immunocompetent, immunosuppressed, and immunodeficient athy¬
mic mice. Before performing these experiments, we first demonstrated that IFN is capable of induc¬
ing an antiviral response in neuroblastoma cell cultures against infection with rabies virus, and sec¬
ond, we could show that IFN in the circulation crosses the blood-brain barrier.

MATERIALS AND METHODS

Cell Cultures: The mouse neuroblastoma cells (NB-41A3) were grown in FIO Ham medium with
15% horse serum and 2.5% fetal calf serum (FCS), C243 cells in MacCoy medium with 5% FCS,

18
 ACTION OF IFN IN RABIES VIRUS-INFECTED MICE

baby hamster kidney cells (BHK-21), and chick embryo-related cells (CER) in minimal essential me¬
dium (MEM) with tryptose and 5% of FCS.

Mice: Pathogen-free C3H/He mice were obtained from Institut Pasteur (Paris); Swiss, nu/nu
and Nu/+ mice (nude mice homozygote and hétérozygote for the nude gene) were from Iffa-Credo
(L'Arbresle, France). Nude mice were maintained in a pathogen-free environment.

Rabies Virus: The challenge virus strain (CVS) of rabies virus adapted on BHK-21 cells was used
to infect culture cells (10 pfu/cell).
The street rabies virus strain of salivaryglands from naturally rabid foxes was kindly supplied by
Dr. J. Blancou (Centre National d'Études sur la Rage et les Animaux Sauvages, Malzéville, France)
and was used to inoculate mice at 10 LD 50/mouse (into the footpad).
The infectivity of CVS in supernatants of NB41A3 cells was measured by the plaque-forming
units (pfu) method on CER cells.'2"
The infectivity of rabies virus in brain extracts was titrated in mice as described in World Health
Organisation Monograph number 23.'22)

Assay of Antibody against Rabies Virus: Blood was drawn from the retroorbital plexus and the
serum was separated and assayed by the fluorescent focus inhibition test.'23'

Preparation of IFN: IFN /ß was prepared by infection of mouse sarcoma C243 cells with New¬
castle disease virus. The IFN was purified on a column of CM-Sepharose. The partially purified
preparation of IFN had a specific activity of 5 IO8 U/mg of protein.'24'

Preparation of Cell and Tissue Extracts: Monolayer cultures of NB41 A3 cells in Falcon flasks (25
cm2) were washed with PBS before addition of 1 ml of lysis buffer containing 10 mM HEPES (pH
7.6), 10 mMKCl, 2 mMMg(OAc)2, 7 mM 2-mercaptoethanol, and 0.5% NP-40. After 5 min, the
cytoplasmic extracts were collected and centrifuged at 1,500 g for 15 min, and the supernatants
stored at -80°C.
Frozen tissues were homogenized by the use of a glass Dounce homogenizer in 2 ml of low-salt
buffer containing 10 mM HEPES (pH 7.6), 10 mM KC1, 2 mM Mg(OAc)2, 7 mM 2-mercaptoetha¬
nol, and aprotinin at 100 U/ml. This homogenate was left for 15 min at 4°C before addition of
NP-40 at a final concentration of 0.5%. Each suspension was then sonicated for 10 s and centri¬
fuged at 1,500 g for 20 min.'25' Tissue extracts were stored at -80°C.

Assay of 2-5A Synthetase andp67K Kinase: Both enzymes were assayed after partial purification
on poly(I) poly(C)-Sepharose. <25_27> The levels of 2-5A synthetase were estimated by the capacity of
·

the enzyme to synthesize 2-5A. Accordingly 2-5A synthetase levels are given as units corresponding
to 1 nmole of [3H]AMP incorporated into 2-5A per milligram of protein per hour. The activity of
the p67K kinase was assayed by its phosphorylation, i.e., the 67K protein. The 32P-labeled 67K pro¬
teins were localized in the polyacrylamide gels after electrophoresis and pieces of gel (0.5 cm) con¬
taining the 32P-labeled proteins were cut and the radioactivity measured by liquid scintillation. The
results are given as 32P04 incorporated into the 67K protein per milligram protein per hour.

Immunosuppression: Fresh cyclophosphamide (Sigma) was prepared for each experiment. Each
mouse was injected intraperitoneally with 150 mg of cyclophosphamide per kg in 100 µ\ 7 h after
virus inoculation. A second dose of 75 mg of cyclophosphamide per kg was given 7 days after virus
inoculation.

Statistical Analysis: The difference between experimental and control groups was analyzed by
Student's i-test.
-

19
 MARCOVISTZ ET AL.

RESULTS AND DISCUSSION

The response of mouse neuroblastoma cells to treatment with IFN
The central nervous system is the target organ in rabies virus infection and accordingly neurons
have been described to be the most susceptible cell type to accommodate replication of rabies virus.
Here we investigated the response of a mouse neuroblastoma cell line by monitoring the induction
of two double-stranded (ds) RNA-dependent enzymes, the 2-5A synthetase, and a specific protein
kinase (p67K kinase). Treatment of mouse neuroblastoma cells with mouse IFN-a/j8 resulted in a
dose-dependent increase in the level of both of these enzymes (Table 1). Rabies virus infectivity was
also assayed in control and IFN-treated neuroblastoma cells. Virus yield was then titrated on CER
cells to illustrate the resistance of IFN-treated cells to rabies virus infection. Treatment of cells with
103 and 104 U/ml of IFN led to a 102- and 104-fold reduction of rabies virus yield, respectively (Table
1). These data therefore indicate that neuroblastoma cells respond to IFN by the development of an
efficient antiviral response against rabies virus.

The response of the brain and spleen in mice treated with IFN
In a preliminary step toward understanding the action of IFN in rabies virus prophylaxis in mice,
we investigated the response of normal mouse tissues toward treatment with IFN. This response was
estimated by assaying the level of 2-5A synthetase and p67K kinase. These enzymes provide conven¬
ient markers to assess the response of different tissues to IFN and thus they show indirectly the pres¬
ence of IFN in a particular tissue.'28"30'
Table 2 gives the levels of 2-5A synthetase and the p67K kinase in the spleen and brain of normal
mice injected with IFN (5 10" units) by four different routes: intravenously (i.V.), intraperi¬
toneally (i.p.), subcutaneously (s.c), and intracerebrally (i.e.). All the different treatments resulted
in enhanced levels of enzymes both in the brain and spleen. These enhanced enzyme levels were
comparable in the spleen of mice in response to IFN treatment by different routes. On the other
hand, in the brain there was a better response when IFN was administered i.e. These data illustrate
two major points: (i) IFN injected i.v., i.p., or s.c. crosses the blood-brain barrier to enhance the
level of 2-5 A synthetase and the protein kinase in the brain; (ii) IFN injected i.e. passes into the cir¬
culation to cause the induction of these enzymes in the spleen.

Table 1. The Response of Neuroblastoma Cells to IFN

IFN 2-5A synthetase p67 kinase Virus titer

(U/ml) (nmoles f3HJAMP/W cells) (32P cpm/W cells) (pfu/ml)
None 0.6 1,375 5.0 108
102 2.1 1,830 1.0 108
103 14.9 9,256 1.2 106
10" 22.9 18,040 3.0 x 10"
Mouse neuroblastoma cultures (NB-41A) in duplicate flasks were treated with different
doses of IFN (NIH U/ml) for 24 h. One flask was used for the preparation of cell extracts to
assay the activities of 2-5A synthetase and p67K kinase. The cells in the second flask were in¬
fected with CVS-fixed rabies virus (10 pfu/cell) and incubated for 48 h at 37°C. The superna¬
tants were harvested and assayed for rabies virus infectivity on CER cell cultures.

20
 ACTION OF IFN IN RABIES VIRUS-INFECTED MICE

Table 2. The Activities of 2-5A Synthetase and Protein Kinase in the Spleen
and Brain of Mice Treated with IFN

Route of injection 2-5Asynthetase p67K kinase

Organ of IFN (nmoles/mg/h) (32P cpm/mg/h)
Spleen No treatment 183 ± 56 5,220 ± 2,756
IFN-i.p. 345 ± 37 12,373 ± 1,492
IFN-i.v. 375 ± 22 11,695 ± 1,262
IFN-s.c. 396 ± 25 10,342 ± 2,216
IFN-i.c. 335 ± 41 12,025 ± 1,885
Brain No treatment 0.1 ± 0.1 3,266 ± 1,567
IFN-i.p. 1.9 ± 0.4 9,287 ± 1,513
IFN-i.v. 2.1 ± 0.3 9,555 ± 1,125
IFN-s.c. 1.9 ± 0.3 7,525 ± 1,045
IFN-i.c. 4.5 ± 0.7 15,860 ± 1,984
The activities of 2-5A synthetase and p67K kinase were assayed in the spleen and in the brain
of normal mice injected with 5x10" NIH units of mouse IFN- + ß by different routes: intra¬
peritoneally (i.p.), intravenously (¡.v.), subcutaneously (s.c), and intracerebrally (i.e.). Five
mice were studied in each experiment.

Action of exogenous IFN in rabies virus-infected mice: immunocompetent

or immunosuppressed
It has been reported that suppression of the immune response by cyclophosphamide in rabies
virus-infected mice both increases the overall mortality rate and delays onset of disease signs and
death for 1-2 weeks.'18' Therefore, it was of interest to study the action of IFN on the outcome of
rabies virus infection in a host in which the immune response is suppressed and in which the onset of
disease signs are delayed. Cyclophosphamide has the capacity to suppress in a nonselective manner
both cellular and humoral immune responses.13" Cyclophosphamide does not seem to interfere with
the production of IFN during rabies virus infection in mice (data not shown). Similarly, action of
IFN in normal mice immunosuppressed by cyclophosphamide is comparable with that observed in
immunocompetent mice, at least for the induction of 2-5A synthetase and the protein kinase in the
lung and spleen (data not shown).
In immunocompetent mice, paralysis of the inoculated limb is the first symptom characteristic of
rabies infection before weight loss and general prostration leading to death. Paralysis of hind limbs
does not occur in immunosuppressed mice, where the first sign of onset of rabies disease is shaggy
hair accompanied with an incoordination of both hind limbs before weight loss and general prostra¬
tion.
Mice infected with street rabies virus were treated with IFN in the absence or presence of the im-
munosuppressor, cyclophosphamide. The results of this experiment are summarized in Table 3 and
Fig. 1. In immunocompetent mice, IFN treatment did not prevent the appearance of paralysis and
there was only a slight protection: 7 out of 14 survived compared with 10 out of 15 in the untreated
infected mice. All immunosuppressed mice treated or not with IFN died of rabies. The day-of-onset
disease signs and the day of death varied individually within the mice of each group (Fig. 1). The
most pronounced effect of IFN treatment was observed on the day of onset of paralysis, thus indi¬
cating a longer incubation period in IFN-treated mice (p < 0.001).

21
 MARCOVISTZ ET AL.

Table 3. Effects of IFN Treatment in Rabies Virus-Infected Mice, Immunocompetent

or Immunosuppressed

Incubation Day of Morbidity3 Mortality

Treatment period3 death3 (days) number of mice (°/o)
None 10.3 ± 1.5 15.2 ± 5.4 4.9 ± 4.0 10/15 (66.6)
IFNb 13.7 ± 6.0 19.7 ± 5.3 6.0 ± 4.0
(pi < 0.01)d (pi < 0.01) (pi > 0.3)
7/14 (50)
Cyclophosphamide 11.8 ± 2.8 15.8 ± 2.4 4.1 ± 2.5 14/14 (100)
IFN and cyclo¬ 13.8 ± 2.6 16.2 ± 1.6 2.3 ± 1.9
(p2 0.30)
15/15 (100)
phosphamide1: (p2 < 0.01)e > (p2 < 0.01)
¡"Arithmetic mean ± standard deviation of each group presented in Fig. 1.
b10" units of IFN were injected i.p. 1 h after virus inoculation and every day until paralysis in immunocompe¬
tent mice or until the first symptoms of shaggy hair and incoordination in immunosuppressed mice.
cOne 150 mg/kg of cyclophosphamide dose given 7 h after virus inoculation and once a week at 75 mg/kg
doses (i.p.).
dpl, The difference between immunocompetent mice treated or not with IFN.
ep2, The difference between immunosuppressed mice treated or not with IFN.

VIRUS VIRUS & IFN VIRUS & CY VIRUS, CY & IFN

days 3o
110/15) (7/141 »15/151
27
3

m
c

21
V
(0
S(0 18
.c
(0
IS mil , mil
TJ

F
»-
o

il
«- S 12
(O
O
c
9 III II
>
E

number of mice in experimental groups

FIG. 1. Time of onset of disease signs and death in immunocompetent (virus; virus and IFN) and
in immunosuppressed (virus and cyclophosphamide; virus, cyclophosphamide, and IFN) mice in¬
fected with rabies virus and treated or not with IFN. Mice (C3H/He) were inoculated with street
rabies virus (10 LD50) into the footpad (day 0). IFN treatments and cyclophosphamide (CY) injec¬
tions were as in the legend of Table 3. Vertical lines represent the morbidity period of individual
mouse starting with the day of onset of clinical signs until death. The numbers at the top of each col¬
umn indicate the number of mice dead in each group.

22
 ACTION OF IFN IN RABIES VIRUS-INFECTED MICE

Effect of IFN treatment on antibody synthesis against rabies virus in

immunocompetent and immunosuppressed mice during
rabies virus infection
IFN treatment of immunocompetent mice infected with street rabies virus resulted in a 10-fold re¬
duction of rabies virus production in the brain. On the other hand, there was a significant increase
(p < 0.02) of antibody synthesis against rabies virus in response to treatment with IFN (Table 4).
Here, therefore, we have a good example in which IFN treatment triggers enhanced antibody syn¬
thesis. This is one of the properties of IFN that has been reported previously by several
workers.(32"35'
As expected, antibody synthesis in immunosuppressed mice was almost not detectable. However,
in mice treated with IFN and cyclophosphamide, there was still a significant antibody synthesis
against rabies virus (Table 4). These antibodies were of IgM class. The mechanism by which IFN
somehow by-passes the inhibitory action of cyclophosphamide on the antibody synthesis remains to
be clarified. It should be noted, however, that there was no apparent difference in the clinical symp¬
toms observed in immunosuppressed mice treated or not with IFN.

Action of IFN treatment in rabies virus infection of nude mice

Athymic nude mice provide a useful experimental animal model to investigate the role of cells in
the development of clinical signs and pathogenesis during a viral infection.'36' For example,
paralysis of hind limbs of mice during street rabies virus infection is considered to be a consequence
of events in which cells seem to play a major role. Therefore, athymic nude mice infected with
street rabies virus do not show paralysis, which is the routine clinical symptom observed in normal
mice.'17'
Homozygous nude (background Swiss; nu/nu) mice infected with street rabies virus became
cachexie Such infected mice started to lose weight significantly 15 days after virus inoculation. This
loss of weight increased progressively until the death of all animals (8/8) between 22th and 30th day
of infection (Fig. 2). IFN (105 units) administered 1 h after virus inoculation and every 24 h up to 30

Table 4. Virus and Antibody Titers in Immunocompetent and

Immunosuppressed Mice during Street Rabies Virus Infection

Virus titerb
Treatment3 (LD50/ml) Antibody titer (1U)^
None 1032 25.4 ± 8.6 (6)d
IFN 102 36.8 ± 8.5 (6)
(pi < 0.001)e
Cyclophosphamide 10"2 0.48 ± 0.04 (5)
IFN and cyclophosphamide 1042 11.7 ±8.6 (6)
(p2 < 0.001)f
aThe different treatments were as in Table 3 and Fig. 1.
bVirus in the brain of mice was titrated in Swiss mice; the average from 2 mice.
cAntibody production against rabies virus was titrated by the fluorescent focus reduction
test. ,u

dThe number of mice in each group.

epl, The difference between immunocompetent mice treated or not with IFN.
fp2, The difference between immunosuppressed mice treated or not with IFN.

23
 MARCOVISTZ ET AL.

30

O)
» 20

2 '- .
2
m

10

5 10 15 20 25 30
Time ; days
FIG. 2. Weights of rabies virus infected nude (nu/nu) mice treated or not with IFN. Two groups
of 8 nude athymic nude mice (males, 2 months old) were inoculated with street rabies virus (10
LDso/mouse) into the footpad. One group was treated with IFN (105 units) administered i.p. 1 h
after virus inoculation and every day up to 30 days. The arithmetic means of body weights (the ordi¬
nate) of each group were measured on different days as indicated (abscissa). A, Rabies virus-in¬
fected mice; the number below the different points indicates the number of mice dead. ·, Rabies
virus-infected mice treated with IFN.

days protected nude mice against rabies infection (Fig. 2). One month after the end of IFN treat¬
ment (i.e., 60 days after virus inoculation), one mouse developed rabies and died a few days later. In
this mouse, diagnosis of rabies virus was made by immunofluorescence study of brain sections. The
rest of the IFN-treated mice (7/8) remained healthy until the end of experiment, 90 days after virus
inoculation.
These results indicate that the mortality in infected homozygous nude (nu/nu) mice treated with
IFN was significantly reduced (12.5%) with respect to the untreated infected mice (100%). On the
other hand, however, treatment of heterozygous (nu/+ ) mice with IFN did not protect mice against
rabies virus infection (Table 5). Synthesis of antibody (IgM) specific for rabies virus was detectable
up to 60-80 days after virus inoculation in IFN-treated mice (Table 5).

CONCLUSIONS

In spite of the in vitro and in vivo evidence that treatment of mouse neuroblastoma cells with IFN
provides protection against eventual infection with rabies virus, and that IFN in the circulation
crosses the blood-brain barrier, IFN treatment of mice infected with rabies virus does not prevent
the development of rabies disease. IFN treatment stimulates a significant enhancement in the level
of antibodies against rabies and causes a delay in the appearance of the first signs of paralysis and

24
 ACTION OF IFN IN RABIES VIRUS-INFECTED MICE

Table 5. Mortality, Virus Infectivity, and Rabies-Neutralizing Antibody Synthesis in

Street Rabies Virus-Infected Nude Mice

Days after Mortality3 Virus titer Antibody Titer(IU)

Mice virus inoculation % (no.) (LDso/ml) total Ig IgM
Control
nu/nu 30 100 (8/8) 1061 2.6 NTb
nu/+ 25 85.7(6/7) IO60 NT

IFN-treated
nu/nu 22 2.6 NT
30 4.5 4.0
60 12.5(1/8) <103c 3.0 3.0
80 1.7 1.5
nu/+ 28 77.7(7/9) 1060 NT

Athymic nude mice (nu/nu) and their heterozygous littermates (nu/+ ) were inoculated with street rabies virus
(10 LDso) into the footpad. The titers of virus and antibody are the arithmetic means of the titers obtained from
three mice per group. The mice were bled from retroorbital plexus; the sera were assayed before and after incu¬
bation with Protein A-Sepharose (90 min at 4°C) to separate the anti-rabies antibodies IgM from those of IgG.
IFN (10s units) was administered i.p. 1 h after virus inoculation and every day up to 30 days.
aMortality until 90 days after virus inoculation.
bNT, Not tested.
The only dead mouse; the virus was detected by immunofluorescence of brain sections.

death. However, although IFN might be functional during the incubation period of rabies virus, the
evolution of rabies disease depends on other mechanisms related to the immune response of the
host. For this reason, we investigated the action of IFN in rabies virus-infected mice during immu-
nosuppression by cyclophosphamide. This substance is capable of suppressing in a nonselective way
both cellular asnd humoral immune responses of the host. Cyclophosphamide-treated mice died ir¬
respective of IFN treatment. However, an intriguing observation was the enhancement of anti-
rabies antibodies in immunosuppressed mice in response to treatment with IFN.
The efficacy of the IFN treatment was illustrated by using athymic nude mice infected with street
rabies virus. IFN treatment protected nude mice against rabies infection. Several months after the
IFN treatment, 7 of 8 nude mice remained in perfect health and among these survivors no virus was
detectable in the brain. These results indicate that the efficacy of IFN treatment against the evolu¬
tion of rabies disease in mice is dependent on the suppression of the T-cell-mediated immune re¬
sponse of the host. It is tempting therefore to suggest that combination of IFN treatment with a spe¬
cific T-cell suppressor might be used efficiently in postexposure prophylaxis of rabies.

REFERENCES

BAER, G.M., SHANTHA, T.R., and BOURNE, G.H. (1968). The pathogenesis of street rabies virus in
rats. Bull. WHO 38, 119-125.
2. MURPHY, F.A., and BAUER, S.P. (1974). Early street rabies virus infection in striated muscle and later
progression to the central nervous system. Intervirology 3, 256-268.
STEWART II, W.E., and SULKIN, S.E. (1966). Interferon production in hamsters with rabies virus.
Proc. Soc. Exp. Biol. Med. 123, 650-653.
4. WIKTOR, T.J., KOPROWSKI, H., and RORKE, L.B. (1972). Localized rabies interferon in mice. Proc.
Soc. Exp. Biol. Med. 140, 759-764.

25
 MARCOVISTZ ET AL.

5. MARCOVISTZ, R., TSIANG, H., and HOVANESSIAN, A.G. (1984). Production and action of inter¬
feron in mice infected with rabies virus. Ann. Virol. (Inst. Pasteur) 135E, 19-33.
6. MARCOVISTZ, R., GALABRU, J., TSIANG, H., and HOVANESSIAN, A.G. (1986). Neutralization of
interferon produced during rabies virus infection in mice. J. GEn. Virol. 67, 387-390.
7. BAER, G.M. (1981-1982). The effect of interferon on rabies infection of animals. Tex. Rep. Biol. Med.
41, 526-531.
8. POSTIC, B., and FENJE, P. (1971). Effect of administered interferon on rabies in rabbits. Appi. Micro-
biol. 22,428-431.
9. HO, M., NASH, C, MORGAN, C.W., ARMSTRONG, J.A., CARROLL, R.G., and POSTIC, B.
(1974). Interferon administered in the cerebrospinal space and its effect on rabies in rabbits. Infect; Im¬
mun. 9, 268-293.
10. HILFENHAUS, J., KARGES, H.E., WEINMANN, E., and BARTH, R. (1975). Effect of administered
leukocyte interferon on experimental rabies in monkeys. Infect. Immun. 11, 1156-1162.
11. HILFENHAUS, J., WEINMANN, E., MAJER, M., BARTH, R., and JAEGER, J. (1977). Post-exposure
administration of human interferon to rabies infected monkeys. J. Infect. Dis. 135, 846-851.
12. WEINMANN, E., MAJER, M., and HILFENHAUS, J. (1979). Intramuscular and/or intralumbar post-
exposure treatment of rabies virus infected cynomolgus monkeys with human interferon. Infect. Immun.
24, 24-31.
13. ATANAZIU, P., BARROETA, M., TSIANG, H., and FAVRE, S. (1970). Inhibition in vivo de la multi¬
plication du virus rabique par un interferon endogène. Ann. Virol. (Inst. Pasteur) 119, 767-777.
14. FENJE, P., and POSTIC, B. (1970). Protection of rabbits against experimental rabies by poly(I) -polyfC).
Nature 226, 171-172.
15. LEVY, H.B., BAER, G., BARON, S., BUCKLER, CE., GIBBS, C.J., IABAROLA, M.J., LONDON,
W.T., and RICE, J.A. (1975). A modified polyriboinosinic-polyribocytidylic acid complex that induces in¬
terferon in primates. J. Infect. Dis. 132, 434-439.
16. JANIS, B., and HABEL, K. (1972). Rabies in rabbits and mice: Protective effect of polyriboinosinic-poly-
ribocytidylic acid. J. Infect. Dis. 125, 345-353.
17. GUILLON, J.C., and TSIANG, H. (1980). Rôle de l'interféron et du thymus dans la pathogenèse de l'in¬
fection rabique chez la souris. Ann. Virol. (Inst. Pasteur) 131E, 229-245.
18. SMITH, J.S., McCLELLAND, CL., REÍD, F.L., and BAER, G.M. (1982). Dual role of the immune re¬
sponse in street rabies virus infection in mice. Infect. Immun. 35, 213-221.
19. TSIANG, H. (1982). Neuronal function impairment in rabies-infected rat brain. J. Gen. Virol. 61, 277-
281.
20. GOURMELON, P., BRIET, D., COURT, L., and TSIANG, H. (1987). Electro-physiological and sleep al¬
terations in experimental mouse rabies. Brain Research (in press).
21. BUSSEREAU, F., FLAMAND, ., and PESE-PART, D. (1982). Reproducible plaquing system for rabies
virus in CER cells. J. Virol. Meth. 4, 277-282.
22. WORLD HEALTH ORGANIZATION. (1973). Laboratory Techniques in Rabies. M.M. Kaplan and H.
Koprowski (eds.). WHO Monograph Series no. 23.
23. SUREAU, P., ROLLIN, P.E., and ZELLER, H. (1982). Correlation entre l'épreuve immunoenzymatique,
la séroneutralisation et la réduction de foyers fluorescents pour le titrage des anticorps rabiques. Comp.
Immunol. Microbiol. Infect. Dis. 5, 143-150.
24. KRUST, B., RIVIÈRE, Y., and HOVANESSIAN, A.G. (1982). p67K kinase in different tissues and
plasma of control and interferon-treated mice. Virology 120, 240-246.
25. HOVANESSIAN, A.G., and RIVIÈRE, Y. (1980). Interferon-mediated induction of 2-5A synthetase and
protein kinase in the liver and spleen of mice infected with Newcastle disease virus or injected with poly(I) ·

poly(C). Ann. Virol. (Inst. Pasteur) 131E, 501-516.

26. BUFFET-JANVRESSE, C, MAGARD, H., ROBERT, N., and HOVANESSIAN, A.G. (1983). Assay
and the levels of 2-5A-synthetase in lymphocytes of patients with viral, bacterial and autoimmune assay.
Ann. Immunol. (Inst. Pasteur) 134D, 247-258.
27. HOVANESSIAN, A.G., RIVIÈRE, Y., and KRUST, B. (1983). Assay of interferon-mediated protein ki¬
nase activity from plasma tissue extracts. Ann. Biochem. 129, 349-356.
28. GALABRU, J., ROBERT, N., BUFFET-JANVRESSE, C, RIVIÈRE, Y., and HOVANESSIAN, A.G.
(1985). Continuous production of interferon in normal mice: Effect of anti-interferon globulin, sex, age,
strain and environment on the levels of 2-5A synthetase and p67K kinase. J. Gen. Virol. 66, 711-718.
29. YOUN, J.K., HOVANESSIAN, A.G., RIVIÈRE, Y., GILBERT, H., and LACOUR, F. (1983). Enhance¬
ment of natural killer cell activity and 2-5A synthetase in mice treated with polyadenylic-polyuridilic acid.
Cell. Immunol. 79, 298-308.

26
 ACTION OF IFN IN RABIES VIRUS-INFECTED MICE

30. SARON, M.F., RIVIÈRE, Y., HOVANESSIAN, A.G., and GUILLON, J.C (1982). Chronic production
of interferon in carrier mice congenitally infected with lymphocytic choriomeningitis virus. Virology 117,
253-256.
31. BACH, J.F. (1975). The mode of action of immunosuppressive agents, in: Frontiers of Biology, Vol. 41.
A. Neuberger and E.L. Tatum (eds.). New York: Elsevier/North-Holland Biomedicai Press, pp. 173-225.
32. DE MAEYER, E. (1981). Interferon and the immune system: A review. In: 77ie Biology of the Interferon
System. E. De Maeyer, G. Galasso, and H. Schellekens (eds.). Amsterdam: Elsevier/North-Holland Bio-
medical Press, pp. 203-209.
33. GRESSER, I. (1977). Commentary: On the varied biologic effects of interferon. Cell Immun. 34, 406-415.
34. EPSTEIN, L.B. (1977). The effects of interferons on the immune response in vitro and in vivo, in: Inter¬
ferons and Their Actions. W. Stewart II (ed.). Cleveland: CRC Press, pp. 91-132.
35. NAKAMURA, M., MANSER, T., PEARSON, G.D.N., DALEY, M.J., and GEFTER, M.L. (1984).
Effect of IFN- on the immune response in vivo and on gene expression in vitro. Nature 307, 381-382.
36. IWASAKI, T. (1978). Experimental virus infections in nude mice, in: The Nude Mouse in Experimental
and Clinical Research. J. Fogh and B.C. Grovanella (eds.). New York: Academic Press, Inc., pp. 457-475.

Address reprint requests to:

Dr. Ara G. Hovanessian
Département de Virologie
Institut Pasteur
25 rue du Dr. Roux
75724 Paris Cedex 15, France

Received 1 August 1986/Accepted 15 October 1986

27