Content uploaded by Anneline Nansen
Author content
All content in this area was uploaded by Anneline Nansen on May 04, 2015
Content may be subject to copyright.
of June 7, 2013.
This information is current as
βαLipopolysaccharide: Central Role of IFN-
Viral Infection Causes Rapid Sensitization to
Anneline Nansen and Allan Randrup Thomsen
http://www.jimmunol.org/content/166/2/982
2001; 166:982-988; ;J Immunol
References http://www.jimmunol.org/content/166/2/982.full#ref-list-1
, 25 of which you can access for free at: cites 45 articlesThis article
Subscriptions http://jimmunol.org/subscriptions is online at: The Journal of ImmunologyInformation about subscribing to
Permissions http://www.aai.org/ji/copyright.html
Submit copyright permission requests at:
Email Alerts http://jimmunol.org/cgi/alerts/etoc
Receive free email-alerts when new articles cite this article. Sign up at:
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Immunologists All rights reserved.
Copyright © 2001 by The American Association of
9650 Rockville Pike, Bethesda, MD 20814-3994.
The American Association of Immunologists, Inc.,
is published twice each month byThe Journal of Immunology
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
Viral Infection Causes Rapid Sensitization to
Lipopolysaccharide: Central Role of IFN-
␣
1
Anneline Nansen and Allan Randrup Thomsen
2
LPS is the major active agent in the pathogenesis of Gram-negative septic shock. In this report we have studied the influence of
concurrent viral infection on the outcome of LPS-induced shock. We find that infection with vesicular stomatitis virus sensitizes
mice to LPS at an early time point following infection. Treatment of mice with the chemical IFN inducer, polyinosinic:polycytidylic
acid, has a similar effect. This hypersensitivity to LPS correlated with hyperproduction of TNF-
␣
in vivo. The cellular and
molecular mechanisms underlying this phenomenon were investigated using Ab-depleted and gene-targeted mice. Our results
revealed that while NK cell depletion and elimination of IFN-
␥
partially protected against the sensitizing effects of vesicular
stomatitis virus and polyinosinic:polycytidylic acid, the most striking effect was observed in IFN-
␣
R-deficient mice. Thus hy-
perproduction of TNF-
␣
was completely abrogated in IFN-
␣
R-deficient mice, indicating that the principal mechanism under-
lying rapid virus-induced sensitization to LPS is an IFN-
␣
-mediated priming of mice for an augmented production of TNF-
␣
in
response to LPS. This conclusion was further supported by the finding that pretreatment of mice with rIFN-
␣
mimicked the
effect of viral infection. In conclusion, our results reveal a previously unrecognized proinflammatory effect of IFN-
␣
and point
to a new pathway through which viral infection may influence the outcome of concurrent bacterial infection. The Journal of
Immunology, 2001, 166: 982–988.
The Gram-negative bacterial wall constituent, endotoxin
(LPS), is the major active agent in the pathogenesis of
septic shock. A shock-like state can be induced by a sin-
gle injection of LPS into animals (1). Most of the toxic effects of
LPS are mediated through the release and action of macrophage-
derived inflammatory cytokines (1–3). TNF-
␣
appears to consti-
tute a central element in the pathogenesis as indicated by the rel-
ative resistance to LPS-induced toxicity in mice lacking the p55
TNFR, the TNF-
␣
molecule, or producing high levels of soluble
TNFR1 fusion protein (4–6).
Another important regulator of LPS-induced pathology is IFN-
␥
(3, 7), the involvement of which is supported by several lines of
evidence. Administration of IFN-
␥
or neutralizing Abs to IFN-
␥
has been shown to modify the lethal outcome of several forms of
endotoxic shock and Gram-negative bacterial infections, and ex-
periments with IFN-
␥
R-deficient mice have revealed that these
mice are relatively resistant to LPS-induced shock (2, 8, 9). It is
generally believed that one aspect of the contribution of IFN-
␥
in
LPS-induced shock consists in priming monocytes/macrophages
by inducing the expression of receptors for TNF-
␣
on the surface
of these cells, which in turn enables autocrine binding of TNF-
␣
and subsequent activation of monocytes/macrophages (10–13). In
addition to the early effect exerted by IFN-
␥
in LPS-induced
shock, IFN-
␥
also seems to promote LPS-induced lethality by
more late-acting mechanisms (2). In mice, a generalized shock
syndrome, known as the generalized Shwartzman reaction, can be
elicited by two consecutive injections of LPS (14, 15). A priming
dose of LPS is injected into the footpad and followed after 24 h by
an i.v. challenge injection of LPS. After the last challenge, which
is nonlethal per se, the mice develop a generalized shock syndrome
and die within 48 h. IFN-
␥
has been shown to be a critical com-
ponent in the priming phase of the generalized Shwartzman reac-
tion. Thus local injection of LPS is believed to trigger release of
IL-12 and IL-15 that synergizes in inducing NK cells to produce
IFN-
␥
(16), which in turn primes macrophages for activation.
Upon subsequent exposure to LPS, the primed macrophages be-
come hyperactivated and produce large amounts of TNF-
␣
and
IL-1 (17, 18). Thus endogenous IFN-
␥
may sensitize mice to oth-
erwise nonlethal doses of LPS.
The involvement of IFN-
␥
in the pathogenesis of endotoxic
shock prompted us in a previous study to investigate whether pro-
duction of this cytokine in the context of viral infection could alter
antibacterial host responses through a modified cytokine network
(19). Our results revealed that systemic infection of mice with the non-
cytopathogenic lymphocytic choriomeningitis virus (LCMV)
3
sensitized
mice to low amounts of LPS, and that this hypersensitivity correlated with
hyperproduction of TNF-
␣
. Hyperproduction of TNF-
␣
was found
to be temporally correlated with virus-induced T cell-dependent
production of IFN-
␥
, thus only a marginally increased IFN-
␥
and
TNF-
␣
production was observed in T cell-deficient nude (nu/nu)
mice and in mice infected with vesicular stomatitis virus (VSV), a
virus that induces less extensive T cell activation than does
LCMV. Furthermore, LCMV infection was found to be much less
efficient in priming IFN-
␥
-deficient mice for hyperproduction of
TNF-
␣
(19), and neutralization of the latter cytokine markedly
protected against a lethal outcome (20). Notably, however, our
Institute of Medical Microbiology and Immunology, University of Copenhagen,
Copenhagen, Denmark
Received for publication May 30, 2000. Accepted for publication October 26, 2000.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported in part by the Danish Medical Research Council, the
Biotechnology Center for Cellular Communication, the Beckett Foundation, and
Gerda and Aage Haensch’s Foundation. A.N. is the recipient of a Ph.D. scholarship
from the Faculty of Health Sciences, University of Copenhagen.
2
Address correspondence and reprint request to Dr. Allan Randrup Thomsen, Insti-
tute of Medical Microbiology and Immunology, University of Copenhagen, The Pa-
num Institute, 3C Blegdamsvej, DK-2200 Copenhagen N, Denmark. E-mail address:
A.R.Thomsen@immi.ku.dk
3
Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; poly
(I:C), polyinosinic:polycytidylic acid; VSV, vesicular stomatitis virus; nu/nu, nude;
MT/
MT, B cell deficient; rHuIFN-
␣
, recombinant human hybrid IFN-
␣
A/D.
Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
results revealed that LPS-induced production of TNF-
␣
was sig-
nificantly augmented in both LCMV-infected T cell-deficient
nu/nu mice and IFN-
␥
-deficient mice compared with levels found
in matched uninfected controls. Moreover, very similar results
were obtained when wild-type mice were challenged with LPS at
an early time-point after LCMV infection, before T cells have been
substantially activated. Together, these findings suggest that al-
though T cell-dependent IFN-
␥
production is important, it is not
the only mechanism through which viruses may prime for TNF-
␣
production. IFN-
␣
is a likely candidate in this respect, because
most viral infections including LCMV and VSV are known to
induce substantial production of IFN-
␣
(21–25). In virus-infected
immunocompetent mice, production of IFN-
␣
generally reaches
peak levels 1–3 days post infection (p.i.), and IFN-
␣
has been
shown to be a critical and indispensable component of early anti-
viral host defenses as mice made genetically deficient for IFN-
␣
R expression are often unable to control virus replication, and
may die (24).
IFN-
␣
has been found to synergize in vitro with subactivating
doses of LPS to activate macrophages for production of TNF-
␣
and to enhance their microbicidal activity (26, 27). Therefore, the
present study was undertaken to explore in greater detail whether
early virus-induced production of IFN-
␣
could sensitize mice to
LPS-induced shock through an augmented production of TNF-
␣
in
response to LPS. This was done by monitoring production of
TNF-
␣
and lethality following an i.p. injection of LPS into mice
preinfected with VSV or pretreated with the chemical IFN-inducer
polyinosinic:polycytidylic acid (poly(I:C)) (28, 29). The response
of a number of mutant mice with targeted defects in genes of
potential interest was analyzed and compared with wild-type mice.
We demonstrate in this study that virus-induced production of
IFN-
␣
may prime mice for an augmented production of TNF-
␣
in
response to LPS. This is evidenced by the findings that an aug-
mented production of TNF-
␣
occurs despite the lack of either NK
cells, T cells, B cells, or IFN-
␥
, but is completely abolished in
mice made genetically deficient for IFN-
␣
R expression. Further-
more, pretreatment of mice with a rIFN-
␣
also augmented TNF-
␣
production and increased the lethality. Thus this seems to be a
general phenomenon that may be induced by different viral infec-
tions and, therefore, may be regarded as a virus-induced analogue
to the bacterially induced generalized Shwartzman reaction.
Materials and Methods
Mice
C57BL/6, C57BL/6-nu/nu, and 129/Sv mice were obtained from Bomholt-
gaard (Ry, Denmark). Mice defective in IFN-
␣
R expression (IFN-
␣
R
⫺/⫺
) on a 129 background were derived from breeding pairs from B &
K Universal (North Humberside, U.K.). Mice deficient in production of
IFN-
␥
(IFN-
␥
⫺/⫺
) and B cell deficient (
MT/
MT) mice both on a C57BL
background were the progeny of breeding pairs obtained from The Jackson
Laboratory (Bar Harbor, ME) and the National Institutes of Health (Be-
thesda, MD), respectively.
MT/
MT mice were bred using heterozygous
female and homozygous males, and the offspring were selected by testing
sera in a sandwich ELISA for the presence of IgM Abs. Unless otherwise
specified, 6- to 8-wk-old female mice were used in all experiments, and
animals from outside sources were always allowed to acclimatize to the
local environment for at least 1 wk before use. All animals were housed
under specific pathogen-free conditions as validated by testing of sentinels
for unwanted infections according to Federation of European Laboratory
Animal Science Association guidelines; no such infections were detected.
Infection/pretreatment of mice
Where indicated, mutant and wild-type mice were infected with VSV of the
Indiana strain originally provided by K. Berg of the Institute of Medical
Microbiology and Immunology (Copenhagen, Denmark), and in one ex-
periment with LCMV of the Traub strain (30). In other experiments, mu-
tant and wild-type mice were treated with poly(I:C) (Sigma, St. Louis,
MO). Mice to be infected received a virus dose of 10
3
LD
50
LCMV or 10
6
PFU VSV in an i.v. injection of 300
l, whereas mice to be treated with
poly(I:C) were injected i.p. with 150
g poly(I:C) dissolved in 150
l
sterile PBS.
Recombinant human hybrid IFN-
␣
A/D (rHuIFN-
␣
)
rHuIFN-
␣
originating from Hoffmann-La Roche (Nutley, NJ) was kindly
provided by K. Berg (Institute of Medical Microbiology and Immunology).
This rHuIFN-
␣
has previously been tested in mice and found to exert the
expected effect (31, 32). Mice to be treated received 2 ⫻10
6
U i.p. the day
before LPS challenge.
In vivo depletion of NK cells
Mice were depleted of NK cells 1 day before treatment with poly(I:C) or
infection with VSV. NK cells were depleted by an i.v. injection of 50
l
purified rabbit anti-asialo G
M1
Abs (Wako Pure Chemicals, Osaka, Japan)
diluted in 300
l PBS. Control mice received an i.v. injection of 300
l
PBS. NK cell depletion was confirmed by testing NK cell cytotoxic activity
in a standard (51) Cr release assay, using NK cell sensitive YAC-1 cells as
targets (30).
LPS challenge
Mice were challenged i.p. with LPS from Escherichia coli serotype 055:B5
(Sigma), and unless otherwise specified, the challenge dose was 50
g/
mouse. Virus-infected mice were challenged with LPS on day 1 p.i. (VSV)
and in one experiment on day 8 p.i. (LCMV), whereas mice injected with
poly(I:C) were challenged with LPS 1 day after treatment. Control groups
of mice were included in all experiments and consisted of mice that re-
ceived LPS alone or were injected i.p. with 150
l PBS 1 day before LPS
challenge (poly(I:C) experiments).
Lethality experiments
In most experiments mortality was recorded 24 and 48 h after LPS
challenge.
Quantitation of cytokine levels in serum
Cytokine concentrations in sera were determined using a sandwich ELISA.
The following ELISA kits (Endogen, Cambridge, MA) were used in this
study: TNF-
␣
, IL-6, IL-1
␣
, and IL-1

. The assays were run according to
the manufacturer’s instructions, and cytokine levels in sera were calculated
by comparison with a standard curve generated using recombinant cyto-
kine; the limit of detection for all cytokines was 15 pg/ml.
Results
Sensitization to LPS in VSV-infected and poly(I:C)-treated
animals
It has previously been found that LCMV infection is associated
with markedly increased sensitivity to LPS (19, 20). Although the
infected mice were most susceptible to LPS during the height of
the adaptive response, significantly increased lethality was ob-
served already during the innate response. To examine whether
rapid sensitization to LPS could be induced by viruses other than
LCMV, we examined the susceptibility to LPS of mice infected
with VSV or treated with poly(I:C). Briefly, lethality experiments
were conducted in mice infected with VSV or treated with
poly(I:C) 1 day before LPS challenge. All mice that were chal-
lenged with LPS received an i.p. challenge dose of 50
g/mouse,
and mortality was recorded after 24 and 48 h (Table I).
All mice infected with VSV 1 day before LPS challenge died
within 24 h and very similar results were obtained in mice injected
with poly(I:C) 1 day before LPS challenge; approx. 2/3 of these
mice died within 24 h, and the remaining died within 48 h. None
of the mice that were infected with VSV, treated with poly(I:C), or
challenged with LPS alone died during the 48-h observation
period.
The outcome of these lethality experiments together with the
previous finding that mice infected with LCMV 3 days before LPS
983The Journal of Immunology
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
challenge also succumb within 48 h, strongly suggest that early
virus-induced sensitization to LPS is a general phenomenon that
may be induced by different viral infections.
Production of TNF-
␣
in VSV-infected and poly(I:C)-treated mice
after LPS challenge in vivo
To see whether the enhanced sensitivity to LPS in VSV-infected
and poly(I:C)-treated mice was reflected in the production of
TNF-
␣
, serum levels of TNF-
␣
were determined 1.5 h after LPS
challenge (Table I). As expected, augmented serum levels of
TNF-
␣
, which clearly exceeded those found in mice infected with
VSV, treated with poly(I:C), or challenged with LPS alone, were
detected in mice infected with VSV or treated with poly(I:C) 1 day
before LPS challenge. Thus enhanced susceptibility to LPS-in-
duced shock correlated with a 5- to 10-fold increase in the pro-
duction of TNF-
␣
in virus-infected or poly(I:C)-treated mice. In
contrast, we did not find the production of IL-1
␣
, IL-1

, or IL-6 to
be substantially augmented (⬍2-fold).
Comparison between early and late virus-induced sensitization
to LPS
To characterize this rapidly induced sensitization to LPS in greater
detail, we compared the response to LPS at this stage to the late
LCMV-induced sensitization to LPS in which situation the most
important priming mechanism has recently been shown to be the
production of IFN-
␥
by virus-activated T cells (19, 20). Briefly,
LPS-induced mortality and production of TNF-
␣
in response to
two different doses of LPS were compared between mice treated
with poly(I:C) 1 day before LPS challenge and mice infected with
LCMV 8 days before LPS challenge reflecting the time-point of
maximal T cell activation (Fig. 1). A clearly dose-dependent pro-
duction of TNF-
␣
was observed in both poly(I:C)-treated mice and
LCMV-infected mice. Furthermore, irrespective of the LPS chal-
lenge dose used, serum levels of TNF-
␣
were clearly elevated in
mice treated with poly(I:C) or infected with LCMV compared with
the levels measured in mice that received LPS alone. However, the
serum levels of TNF-
␣
in response to either challenge dose of LPS
were generally three to five times lower in mice treated with
poly(I:C) than in mice infected with LCMV 8 days earlier. With
respect to LPS-induced mortality, our experiments revealed that
mice treated with poly(I:C) 1 day before LPS challenge died
within 48 h when challenged with 50
g LPS, but survived chal-
lenge with 2
g LPS. In contrast, all mice infected 8 days earlier
with LCMV died within 24 h when challenged with either dose of
LPS; none of the mice challenged with LPS alone died.
From the results outlined above it is evident that mice are less
susceptible to LPS during the innate phase of the host response
than seen later in the infection; thus a relatively high challenge
dose of LPS was required to elicit lethal amounts of TNF-
␣
in
mice treated with poly(I:C). Nevertheless, poly(I:C) did prime
mice for an augmented production of TNF-
␣
even in response to
the lower LPS dose.
The role of NK cells
Production of IFN-
␥
by NK cells has been shown to play an es-
sential role in lethal LPS-induced Shwartzman reaction in mice
(18). Moreover, during the course of a viral infection IFN-
␥
may
be produced not only by virus-activated T cells but also by virus-
activated NK cells (33, 34), and NK cell activation can be ob-
served in mice infected with VSV or pretreated with poly(I:C) as
evidenced by augmented ex vivo cytotoxicity (data not shown). To
find out whether early virus-induced priming of mice for an aug-
mented production of TNF-
␣
was dependent on NK cells, we ex-
amined the susceptibility to LPS in mice depleted of NK cells and
either preinfected with VSV or pretreated with poly(I:C) 1 day
before LPS challenge. As shown in Fig. 2, depletion of NK cells
did not eliminate VSV- and poly(I:C)-dependent sensitization for
an augmented production of TNF-
␣
. Partial protection against a
lethal outcome was observed; this may reflect both the reduced
TNF-
␣
response in some mice as well as a role for NK cells in later
stages of LPS-induced shock.
The role of T cells
It has previously been found that both viruses and poly(I:C) can
activate CD44
high
T cells, a subset characterized by the capacity to
FIGURE 1. Comparison between early and late virus-induced sensiti-
zation to LPS. Groups of B6 mice were injected i.p. with poly(I:C) 1 day
before LPS challenge, infected i.v. with LCMV 8 days before LPS chal-
lenge, or left untreated. Mice were challenged with either 2 or 50
g
LPS/mouse and the serum levels of TNF-
␣
were determined 1.5 h after
challenge with LPS. The shapes represent individual mice.
Table I. Sensitization to LPS in mice treated with poly(I:C) or infected with VSV
Treatment
a
TNF-
␣
(ng/ml)
b
Mortality, No. Dead/Total (%)
c
24 h 48 h
Poly(I:C) LPS 151.6 15/22 (68) 22/22 (100)
PBS LPS 25.9 0/14 (0) 0/14 (0)
Poly(I:C) None ⬍1.8 0/3 (0) 0/3 (0)
VSV LPS 143.1 6/6 (100) 6/6 (100)
None LPS 9.1 0/6 (0) 1/6 (17)
VSV None ⬍0.3 0/6 (0) 0/6 (0)
a
Mice were treated i.p. with 150
g poly(I:C) or infected i.v. with VSV (2 ⫻10
6
PFU) 1 day before challenge with 50
g LPS/mouse.
b
Animals were eyebled 1.5 h after challenge with LPS for determination of TNF-
␣
serum levels. TNF-
␣
levels shown are medians of the respective groups.
c
Mortality was recorded 24 and 48 h after LPS challenge.
984 ROLE FOR TYPE I IFN IN VIRUS-INDUCED SENSITIZATION TO LPS
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
produce IFN-
␥
(35, 36). To examine the possible influence of early
T cell-dependent production of IFN-
␥
in the early sensitization to
LPS, we evaluated the susceptibility to LPS in T cell-deficient
nu/nu mice infected with VSV or treated with poly(I:C) 1 day
before LPS challenge and in nu/nu mice challenged with LPS
alone, and compared the observed responses to those in corre-
sponding groups of wild-type mice (Fig. 2). Although the priming
effect was less consistent in nu/nu mice than in similarly treated
wild-type animals, the overall picture is clearly that T cells are not
mandatory. We assume that the greater variability in the response
of nu/nu mice reflects the well-known tendency toward an altered
stage of macrophage activation in these mice.
The role of B cells
LPS (endotoxin) is well known for its capacity to activate B cells
that may produce IFN-
␥
under certain conditions (37). To examine
whether early virus-induced priming of mice for an increased sus-
ceptibility to LPS involved B cells, we examined the response of
B cell-deficient
MT mice treated with poly(I:C) 1 day before LPS
challenge with respect to lethality and production of TNF-
␣
.As
shown in Fig. 2, the LPS-induced production of TNF-
␣
was as
high in poly(I:C)-treated
MT mice as in wild-type mice, and all
pretreated mice succumbed to the LPS challenge. Thus early virus-
induced sensitization to LPS does not require the presence of B
cells.
The role of IFN-
␥
IFN-
␥
is considered to be the most potent macrophage-activating
cytokine and may be produced by many cell types. To determine
the role of IFN-
␥
in early virus-induced sensitization to LPS, we
examined the susceptibility to LPS in IFN-
␥
-deficient (IFN-
␥
⫺/⫺
)
mice treated with poly(I:C) or infected with VSV 1 day before LPS
challenge. As shown in Fig. 2, IFN-
␥
⫺/⫺
mice were primed for an
augmented production of TNF-
␣
and most succumbed to LPS
challenge when pretreated with poly(I:C) or infected with VSV.
However, TNF-
␣
levels in serum were generally lower than found
in similarly pretreated wild-type mice. This finding is consistent
with the results obtained following NK cell depletion and indicate
that NK cell produced IFN-
␥
augments the early sensitization to
viral infection although this cytokine is not absolutely essential. In
this respect it may be important to note that the relative resistance
of IFN-
␥
R-deficient mice to LPS-induced shock has been shown to
reflect a diminished expression of receptors for LPS on monocytes
and macrophages in these mutants (9). As both poly(I:C) and VSV
are allowed only a short action time in our experimental setup, we
cannot rule out the possibility that the defect in IFN-
␥
⫺/⫺
mice
partially reflects deficient expression of LPS receptors on the sur-
face of monocytes and macrophages causing a generally reduced
responsiveness toward LPS.
The role of IFN-
␣
A common denominator of viral infection and treatment with
poly(I:C) is rapid induction of high amounts of IFN-
␣
. In vitro,
endogenous IFN-
␣
has previously been shown to synergize in an
autocrine manner with subactivating doses of LPS to activate mac-
rophages for production of TNF-
␣
and to enhance their microbi-
cidal activity (27). Furthermore, in a recent in vivo study it was
shown that regulation of type 2 NO synthase expression during the
innate immune response in mice to Leishmania major is dependent
on IFN-
␣
(38). Therefore, to investigate whether IFN-
␣
was
pivotal for the early virus-induced increase in susceptibility to
LPS, we examined the priming effect of poly(I:C) and VSV in
IFN-
␣
R-deficient mice (IFN-
␣
R
⫺/⫺
) (Fig. 3). Because this ge-
netic defect has not been bred onto a C57BL/6 background, we had
to use mice of the 129 Sv strain for this part of our analysis. As
wild-type (129) mice were found to be substantially more sensitive
to the toxicity of LPS (also unprimed mice died within 48 h),
lethality was not well suited as a discriminating parameter al-
though we did consistently see accelerated mortality following
priming. However, analysis of TNF-
␣
levels gave an unequivocal
readout. Thus as previously seen with C57BL/6 mice, wild-type
(129) mice were primed for an augmented TNF-
␣
production in
response to LPS when treated with poly(I:C) or infected with VSV
1 day before LPS challenge. In contrast, the priming capacity of
FIGURE 3. The role of IFN-
␣
in early virus-induced sensitization to
LPS. Wild-type (129) mice and IFN-
␣
R-deficient mice (IFN-
␣
R
⫺/⫺
)
were injected i.p. with poly(I:C) or infected with VSV 1 day before LPS
challenge. Control groups of wild-type (129) mice or IFN-
␣
R
⫺/⫺
mice
were injected i.p. with PBS 1 day before LPS challenge (poly(I:C)) or left
untreated (VSV). Mice were challenged with 50
g LPS/mouse. Serum
levels of TNF-
␣
were determined 1.5 h after challenge with LPS; the
shapes represent individual mice.
FIGURE 2. The role of NK cells, T cells, B cells, and IFN-
␥
in early
virus-induced sensitization to LPS. Wild-type (B6), NK cell-depleted (B6),
T cell-deficient (nu/nu), B cell-deficient (
MT/
MT), and IFN-
␥
-deficient
(IFN-
␥
⫺/⫺
) mice were injected i.p. with poly(I:C) 1 day before LPS chal-
lenge or infected i.v. with VSV 1 day before LPS challenge, respectively.
Mice were challenged with 50
g LPS/mouse. Control groups consisted of
wild-type or mutant mice that received LPS alone. Serum levels of TNF-
␣
were determined 1.5 h after challenge with LPS; the shapes represent in-
dividual mice. Mortality was registered for 48 h after LPS challenge.
985The Journal of Immunology
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
both poly(I:C) and VSV was completely abolished in IFN-
␣
R
⫺/⫺
mice and the production of TNF-
␣
in pretreated/infected IFN-
␣
R
⫺/⫺
mice did not exceed the production induced in mutant
mice challenged with LPS alone. This result clearly demonstrates
that early virus-induced priming for an augmented production of
TNF-
␣
in response to LPS requires IFN-
␣
. To our knowledge
this in vivo priming capacity of IFN-
␣
for production of TNF-
␣
has not been described before.
Pretreatment with rIFN-
␣
sensitizes mice to LPS induced shock
Finally, to test whether IFN-
␣
sufficed for an augmented re-
sponse to LPS, wild-type (C57BL/6) mice were pretreated with
rHuIFN-
␣
to which murine cells are responsive (31, 32, 39); a
matched control group received PBS only. Twenty-four hours later
all mice were challenged with LPS and 1.5 h later TNF-
␣
levels in
serum were analyzed, and mortality over the next 48 h was reg-
istered. As evident from Fig. 4, pretreatment with rIFN-
␣
signif-
icantly increased both LPS-induced TNF-
␣
production and
mortality.
Discussion
In this report we have demonstrated and analyzed virus-induced
sensitization to LPS resulting from triggering of innate defense
mechanisms. We find that viral infection as well as a well known
IFN inducer substantially increases the sensitivity of mice to the
toxic effects of LPS. A central mediator in this sensitization is
IFN-
␣
as evidenced by the following observations. First, as al-
ready mentioned pretreatment of mice with the IFN-inducer
poly(I:C) mimics the effect of viral infection. Second, priming is
completely abolished in IFN-
␣
R
⫺/⫺
mice. Finally, pretreatment
with rHuIFN-
␣
known to work on murine cells also increases the
sensitivity to LPS. Notably, IFN-
␥
was not essential for early vi-
rus-induced sensitization to LPS, demonstrating that the mecha-
nism underlying this phenomenon differs from the mechanism un-
derlying late virus-induced sensitization to LPS, which we have
shown to be dependent on production of IFN-
␥
by virus-activated
T cells (19). A common feature of either scenario is the priming of
virus-infected mice for an augmented production of TNF-
␣
in re-
sponse to LPS and an increased susceptibility to endotoxic shock.
The increased susceptibility to LPS in virus-infected and poly-
(I:C)-treated mice as revealed in lethality experiments always cor-
related with an augmented production of TNF-
␣
. Given the central
role of TNF-
␣
in LPS-induced shock, and the fact that injection of
this cytokine causes an essentially identical syndrome (40, 41), it
is reasonable to assume that the augmented production of TNF-
␣
plays an important pathogenic role. However, this does not ex-
clude the critical participation of other cytokines in virus-induced
sensitization to LPS either during the priming phase or as
comediators.
Production of IFN-
␥
by NK cells has been shown to play an
essential role in the lethal LPS-induced Shwartzman reaction in
mice (18), and recently production of IFN-
␥
by NK cells has also
been shown to be implicated in virus-induced sensitization to LPS
(20). In the present set-up, depletion of NK cells did not eliminate
the priming effect on LPS-induced production of TNF-
␣
, indicat-
ing that production of IFN-
␥
by NK cells is not essential for early
virus-induced priming of mice for an augmented production of
TNF-
␣
.
IFN-
␣
has also been shown to activate CD44
high
T cells that
are known to have a high capacity for production of IFN-
␥
(35,
36). Our experiments with nu/nu mice revealed that early virus-
induced priming for an enhanced production of TNF-
␣
may occur
despite the lack of functional T cells. The results obtained with NK
cell-depleted and nu/nu mice were both supported by the finding
that IFN-
␥
⫺/⫺
mice produced enhanced amounts of TNF-
␣
when
infected with VSV or treated with poly(I:C). However, we did find
some reduction in the amounts of TNF-
␣
produced in primed IFN-
␥
⫺/⫺
mice compared with similarly treated wild types. This could
suggest that IFN-
␥
augmented IFN-
␣
-dependent priming. Alter-
natively, deficient expression of relevant receptors for LPS may
limit the general sensitivity in these mice thus reducing the basic
set point for TNF-
␣
production (9).
B cells may proliferate in vivo in response to LPS and may
produce IFN-
␥
in vivo when stimulated with IL-18 and IL-12 (37).
Our experiments revealed that B cells are not required in early
virus-induced sensitization to LPS because poly(I:C)-treated
MT/
MT mice were primed for an augmented LPS-induced pro-
duction of TNF-
␣
.
In the generalized Shwartzman reaction it is well established
that the local injection of LPS leads to production of IFN-
␥
by NK
cells. IFN-
␥
then primes macrophages for activation, and upon
subsequent exposure to LPS the primed macrophages become hy-
peractivated and produce enhanced amounts of TNF-
␣
(18). In
early virus-induced sensitization to LPS, IFN-
␣
could partly
work by inducing NK cells to produce IFN-
␥
. However, because
absence of NK cells and IFN-
␥
did not prevent the priming effect
of viral infection, other mechanism must also be important. There-
fore, it is likely that macrophages are also primed by IFN-
␣
in an
autocrine manner. This assumption is supported by in vitro studies.
Thus, Influenza A infected human macrophages have been shown
to produce dramatically elevated levels of TNF-
␣
when cocultured
with LPS, and because infected macrophages produced IFN-
␣
,
the authors concluded that priming of infected macrophages by
this cytokine preconditioned these cells to respond to LPS (42).
Moreover, macrophage-synthesized IFN-
␣
can augment NO pro-
duction in an autocrine fashion in cultures stimulated with sub-
activating doses of LPS (26, 42). In line with these observations a
recent in vivo study has revealed that IFN-
␣
regulates early NO
production in Leishmania major-infected mice and that this effect
is important for parasite containment (38).
The mechanism of IFN-
␣
-mediated priming in vivo is likely to
involve the Kupffer cells of the liver, which comprises the largest
fixed macrophage population in the mammalian body. Kupffer
FIGURE 4. Pretreatment with rHuIFN-
␣
sensitizes mice to LPS. Wild-
type (C57BL/6) mice were given either 2 ⫻10
6
units of rHuIFN-
␣
or PBS
i.p. 1 day before LPS challenge (50
g/mouse). Serum levels of TNF-
␣
were determined 1.5 h after challenge with LPS; the shapes represent in-
dividual mice. Mortality was registered for 48 h after LPS challenge. The
difference between IFN-
␣
-pretreated and sham-treated mice was statisti-
cally significant with regard to both parameters (p⬍0.02; Mann-Whitney
rank sum test for cytokine levels and Fisher’s exact test for mortality).
986 ROLE FOR TYPE I IFN IN VIRUS-INDUCED SENSITIZATION TO LPS
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
cells have been shown in vitro to produce TNF-
␣
in response to
endotoxin and are also able to produce IFN-
␣
(43). However, by
use of the RNase protection assay we have not been able to con-
vincingly demonstrate increased levels of mRNA coding for any of
the major proinflammatory cytokines including TNF-
␣
in the liver
and spleen of mice treated with poly(I:C) 1 day earlier (data not
shown). Therefore, rather than acting as a promotor of transcrip-
tion of proinflammatory cytokines, IFN-
␣
-mediated priming may
act as to increase LPS responsiveness of macrophages by inducing
receptors for TNF-
␣
and LPS (CD14) on the surface of these cells.
This assumption is supported by the finding that elevated serum
levels of soluble TNF-
␣
R type II can be detected in wild-type
mice, but not in IFN-
␣
R
⫺/⫺
mice, during the innate host re-
sponse toward different viral infections (44).
The proinflammatory action of IFN-
␣
described in this report
seemingly contradicts previous studies in which IFN-
␣
has been
reported to down-regulate inflammatory responses to LPS. Sys-
temic administration of rIFN-
␣
20 min after i.p. challenge of mice
with a lethal dose of LPS reduced the LPS-induced mortality by
almost 90% and interestingly, some protection was observed even
if rIFN-
␣
was administered 1 h before LPS challenge (43). IFN-
␣
has also been reported to down-regulate local inflammation, thus
the development of the local LPS-induced footpad swelling reac-
tion was suppressed in mice treated systemically with a natural
mixture of IFN-
␣
or rIFN-
␣
after local LPS challenge (45). The
anti-inflammatory effect of systemic IFN-
␣
on LPS-induced im-
mune responses is suggested to be mediated directly through
down-regulation of TNF-
␣
or more indirectly through regulation
of one or several molecules that mediate the inhibitory effect.
However, our results clearly demonstrate a proinflammatory effect
of IFN-
␣
. This apparent discrepancy might relate to the timing of
IFN-
␣
relative to LPS administration. In the above cited study,
IFN-
␣
was given either immediately after or just before LPS
challenge, which may not have allowed sufficient time for IFN-
␣
to exert a priming activity. Also the environment in which the
animals are maintained may critically influence the responsiveness
to IFN-
␣
.
Based on their immune modulatory effects IFNs are currently
being used as treatment against a wide variety of diseases (46, 47).
Thus IFN-
␣
is used to treat patients with certain cancers and
chronic viral diseases, whereas IFN-

is used to treat patients suf-
fering from the autoimmune disease multiple sclerosis. In general,
these treatments have beneficial effects but pathology induced by
IFN treatment has also been reported (46, 47). In view of our
results, it is possible that IFN treatment could influence suscepti-
bility to LPS-induced pathology in patients undergoing IFN treat-
ment, but this obviously needs to be explored.
In conclusion, our results demonstrate that like LPS and bacte-
rial DNA, virus infections may cause rapid sensitization of mice to
LPS-induced shock. Consequently, early virus-induced sensitiza-
tion to LPS may be regarded as a virus-induced analogue of the
generalized Shwartzman reaction. However, while LPS sensitizes
mice through NK cell production of IFN-
␥
, viruses act primarily
through induction of IFN-
␣
. To our knowledge in vivo priming
by IFN-
␣
for production of TNF-
␣
has not been reported before.
Acknowledgments
We thank Grethe Tho¨rner Andersen and Lone Malte for skillful technical
assistance.
References
1. Glauser, M. P., G. Zanetti, J. D. Baumgartner, and J. Cohen. 1991. Septic shock:
pathogenesis. Lancet 338:732.
2. Heinzel, F. P. 1990. The role of IFN-
␥
in the pathology of experimental endo-
toxemia. J. Immunol. 145:2920.
3. Doherty, G. M., J. R. Lange, H. N. Langstein, H. R. Alexander, C. M. Buresh, and
J. A. Norton. 1992. Evidence for IFN-
␥
as a mediator of the lethality of endotoxin
and tumor necrosis factor-
␣
.J. Immunol. 149:1666.
4. Rothe, J., W. Lesslauer, H. Lotscher, Y. Lang, P. Koebel, F. Kontgen, A. Althage,
R. Zinkernagel, M. Steinmetz, and H. Bluethmann. 1993. Mice lacking the tu-
mour necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly
susceptible to infection by Listeria monocytogenes. Nature 364:798.
5. Pasparakis, M., L. Alexopoulou, V. Episkopou, and G. Kollias. 1996. Immune
and inflammatory responses in TNF-
␣
-deficient mice: a critical requirement for
TNF-
␣
in the formation of primary B cell follicles, follicular dendritic cell net-
works and germinal centers, and in the maturation of the humoral immune re-
sponse. J. Exp. Med. 184:1397.
6. Garcia, I., Y. Miyazaki, K. Araki, M. Araki, R. Lucas, G. E. Grau, G. Milon,
Y. Belkaid, C. Montixi, and W. Lesslauer. 1995. Transgenic mice expressing
high levels of soluble TNF-R1 fusion protein are protected from lethal septic
shock and cerebral malaria, and are highly sensitive to Listeria monocytogenes
and Leishmania major infections. Eur. J. Immunol. 25:2401.
7. Kohler, J., D. Heumann, G. Garotta, D. LeRoy, S. Bailat, C. Barras,
J. D. Baumgartner, and M. P. Glauser. 1993. IFN-
␥
involvement in the severity
of Gram-negative infections in mice. J. Immunol. 151:916.
8. Silva, A. T., and J. Cohen. 1992. Role of interferon-
␥
in experimental Gram-
negative sepsis. J. Infect. Dis. 166:331.
9. Car, B. D., V. M. Eng, B. Schnyder, L. Ozmen, S. Huang, P. Gallay,
D. Heumann, M. Aguet, and B. Ryffel. 1994. Interferon-
␥
receptor deficient mice
are resistant to endotoxic shock. J. Exp. Med. 179:1437.
10. Stout, R. D. 1993. Macrophage activation by T cells: cognate and non-cognate
signals. Curr. Opin. Immunol 5:398.
11. Williams, J. G., G. J. Jurkovich, G. B. Hahnel, and R. V. Maier. 1992. Macro-
phage priming by interferon
␥
: a selective process with potentially harmful ef-
fects. J. Leukocyte Biol. 52:579.
12. Tsujimoto, M., Y. K. Yip, and J. Vilcek. 1986. Interferon-
␥
enhances expression
of cellular receptors for tumor necrosis factor. J. Immunol. 136:2441.
13. Liew, F. Y., Y. Li, and S. Millott. 1990. Tumor necrosis factor-
␣
synergizes with
IFN-
␥
in mediating killing of Leishmania major through the induction of nitric
oxide. J. Immunol. 145:4306.
14. Ozmen, L., M. Pericin, J. Hakimi, R. A. Chizzonite, M. Wysocka, G. Trinchieri,
M. Gately, and G. Garotta. 1994. Interleukin-12, interferon-
␥
, and tumor necrosis
factor-
␣
are the key cytokines of the generalized Shwartzman reaction. J. Exp.
Med. 180:907.
15. Billiau, A. 1988.
␥
-Interferon: the match that lights the fire? Immunol. Today
9:37.
16. Fehniger, T. A., H. Yu, M. A. Cooper, K. Suzuki, M. H. Shah, and
M. A. Caligiuri. 2000. Cutting edge: IL-15 costimulates the generalized Shwartz-
man reaction and innate immune IFN-
␥
production in vivo. J. Immunol. 164:
1643.
17. Cowdery, J. S., J. H. Chace, A. K. Yi, and A. M. Krieg. 1996. Bacterial DNA
induces NK cells to produce IFN-
␥
in vivo and increases the toxicity of lipo-
polysaccharides. J. Immunol. 156:4570.
18. Heremans, H., C. Dillen, J. van Damme, and A. Billiau. 1994. Essential role for
natural killer cells in the lethal lipopolysaccharide-induced Shwartzman-like re-
action in mice. Eur. J. Immunol. 24:1155.
19. Nansen, A., J. P. Christensen, O. Marker, and A. R. Thomsen. 1997. Sensitization
to lipopolysaccharide in mice with asymptomatic viral infection: role of T cell-
dependent production of interferon-
␥
.J. Infect. Dis. 176:151.
20. Nguyen, K. B., and C. A. Biron. 1999. Synergism for cytokine-mediated disease
during concurrent endotoxin and viral challenges: roles for NK and T cell IFN-
␥
production. J. Immunol. 162:5238.
21. Bro-Jorgensen, K., and S. Knudtzon. 1977. Changes in hemopoiesis during the
course of acute LCM virus infection in mice. Blood 49:47.
22. Schattner, A., A. Meshorer, and D. Wallach. 1983. Involvement of interferon in
virus-induced lymphopenia. Cell Immunol. 79:11.
23. Gresser, I., M. G. Tovey, C. Maury, and M. T. Bandu. 1976. Role of interferon
in the pathogenesis of virus diseases in mice as demonstrated by the use of
anti-interferon serum. II. Studies with herpes simplex, Moloney sarcoma, vesic-
ular stomatitis, Newcastle disease, and influenza viruses. J. Exp. Med. 144:1316.
24. Muller, U., U. Steinhoff, L. F. Reis, S. Hemmi, J. Pavlovic, R. M. Zinkernagel,
and M. Aguet. 1994. Functional role of type I and type II interferons in antiviral
defense. Science 264:1918.
25. Thomsen, A. R., and C. J. Phau. 1993. Influence of host genes on the outcome of
murine lymphocytic choriomeningitis virus infection. In The Arenaviridae.
M. S. Salvato, ed. Plenum Press, New York, pp. 199–224.
26. Pace, J. L., S. W. Russell, P. A. LeBlanc, and D. M. Murasko. 1985. Comparative
effects of various classes of mouse interferons on macrophage activation for
tumor cell killing. J. Immunol. 134:977.
27. Zhou, A., Z. Chen, J. A. Rummage, H. Jiang, M. Kolosov, I. Kolosova,
C. A. Stewart, and R. W. Leu. 1995. Exogenous interferon-
␥
induces endogenous
synthesis of interferon-
␣
and -

by murine macrophages for induction of nitric
oxide synthase. J. Interferon Cytokine Res. 15:897.
28. Korngold, R., K. J. Blank, and D. M. Murasko. 1983. Effect of interferon on
thoracic duct lymphocyte output: induction with either poly I:poly C or vaccinia
virus. J. Immunol. 130:2236.
29. Field, A. K., A. A. Tytell, G. P. Lampson, and M. R. Hilleman. 1967. Inducers
of interferon and host resistance. II. Multistranded synthetic polynucleotide com-
plexes. Proc. Natl. Acad. Sci. USA 58:1004.
30. Marker, O., and M. Volkert. 1973. Studies on cell-mediated immunity to lym-
phocytic choriomeningitis virus in mice. J. Exp. Med. 137:1511.
987The Journal of Immunology
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
31. Brunda, M. J., D. Bellantoni, and V. Sulich. 1987. In vivo anti-tumor activity of
combinations of interferon-
␣
and interleukin-2 in a murine model. Correlation of
efficacy with the induction of cytotoxic cells resembling natural killer cells. Int.
J. Cancer 40:365.
32. Brunda, M. J., D. Rosenbaum, and L. Stern. 1984. Inhibition of experimentally-
induced murine metastases by recombinant
␣
interferon: correlation between the
modulatory effect of interferon treatment on natural killer cell activity and inhi-
bition of metastases. Int. J. Cancer 34:421.
33. Welsh, R. M., C. A. Biron, J. F. Bukowski, K. W. McIntyre, and H. Yang. 1984.
Role of natural killer cells in virus infections of mice. Surv. Synth. Pathol. Res.
3:409.
34. Orange, J. S., and C. A. Biron. 1996. An absolute and restricted requirement for
IL-12 in natural killer cell IFN-
␥
production and antiviral defense: studies of
natural killer and T cell responses in contrasting viral infections. J. Immunol.
156:1138.
35. Tough, D. F., and J. Sprent. 1996. Viruses and T cell turnover: evidence for
bystander proliferation. Immunol. Rev. 150:129.
36. Ernst, D. N., W. O. Weigle, D. J. Noonan, D. N. McQuitty, and M. V. Hobbs.
1993. The age-associated increase in IFN-
␥
synthesis by mouse CD8
⫹
T cells
correlates with shifts in the frequencies of cell subsets defined by membrane
CD44, CD45RB, 3G11, and MEL-14 expression. J. Immunol. 151:575.
37. Yoshimoto, T., H. Okamura, Y. I. Tagawa, Y. Iwakura, and K. Nakanishi. 1997.
Interleukin 18 together with interleukin 12 inhibits IgE production by induction
of interferon-
␥
production from activated B cells. Proc. Natl. Acad. Sci. USA
94:3948.
38. Diefenbach, A., H. Schindler, N. Donhauser, E. Lorenz, T. Laskay,
J. MacMicking, M. Rollinghoff, I. Gresser, and C. Bogdan. 1998. Type 1 inter-
feron (IFN
␣
/

) and type 2 nitric oxide synthase regulate the innate immune
response to a protozoan parasite. Immunity. 8:77.
39. Brunda, M. J., D. Tarnowski, and V. Davatelis. 1986. Interaction of recombinant
interferons with recombinant interleukin-2: differential effects on natural killer
cell activity and interleukin-2-activated killer cells. Int. J. Cancer 37:787.
40. Gutierrez-Ramos, J. C., and H. Bluethmann. 1997. Molecules and mechanisms
operating in septic shock: lessons from knockout mice. Immunol. Today 18:329.
41. Brouckaert, P., and W. Fiers. 1996. Tumor necrosis factor and the systemic inflam-
matory response syndrome. In Pathology of Septic Shock. E. T. Rietschel and H.
Wagner, eds. Springer, Berlin, pp. 167–187.
42. Peschke, T., A. Bender, M. Nain, and D. Gemsa. 1993. Role of macrophage
cytokines in influenza A virus infections. Immunobiology 189:340.
43. Tzung, S. P., T. C. Mahl, P. Lance, V. Andersen, and S. A. Cohen. 1992. Inter-
feron-
␣
prevents endotoxin-induced mortality in mice. Eur. J. Immunol. 22:3097.
44. Bartholdy, C., A. Nansen, O. Marker, and A. R. Thomsen. 1999. Soluble tumour
necrosis factor (TNF)-receptor levels in serum as markers of anti-viral host re-
activity. Clin. Exp. Immunol 116:299.
45. Heremans, H., R. Dijkmans, H. Sobis, F. Vandekerckhove, and A. Billiau. 1987.
Regulation by interferons of the local inflammatory response to bacterial lipo-
polysaccharide. J. Immunol. 138:4175.
46. Belardelli, F., and I. Gresser. 1996. The neglected role of type I interferon in the
T-cell response: implications for its clinical use. Immunol. Today 17:369.
47. Weinstock-Guttman, B., R. M. Ransohoff, R. P. Kinkel, and R. A. Rudick. 1995.
The interferons: biological effects, mechanisms of action, and use in multiple
sclerosis. Ann. Neurol. 37:7.
988 ROLE FOR TYPE I IFN IN VIRUS-INDUCED SENSITIZATION TO LPS
by guest on June 7, 2013http://www.jimmunol.org/Downloaded from
