Abstract
purpose. Herpes simplex virus (HSV)-1 infection of the murine cornea is known to
stimulate a vigorous interleukin (IL)-6 response, but whether this
pleiotropic cytokine is an essential participant in corneal
inflammation is unclear. This study was designed to compare the early
inflammatory response in IL-6 gene-deficient mice to that in wild-type
hosts.
methods. Gene knockout and wild-type mice (C57BL/6 background) were infected
intracorneally with HSV-1 (strain RE) and observed through clinical
examination and immunohistochemistry for the development of corneal
opacity. Virus corneal titers were determined by standard plaque assay
on Vero cells. Cytokine and chemokine levels in corneal lysates were
measured with commercial ELISA kits.
results. Corneal opacity in IL-6−/− mice was substantially
diminished in comparison with IL-6+/+ hosts 24 to 48 hours
after intracorneal viral infection, and corneal levels of (MIP)-2 and
MIP-1α were significantly reduced. Local administration of IL-6 at
the time of infection restored corneal opacity and chemokine levels to
that of wild-type hosts. Antibody neutralization of endogenous IL-6 in
IL-6+/+ animals reduced corneal opacity scores and MIP-2
levels to that of IL-6−/− mice. Ex vivo studies with
excised corneal buttons revealed that uninfected IL-6−/− corneas injected with IL-6 produced MIP-2 and MIP-1α at levels
comparable to that seen in IL-6+/+ hosts.
conclusions. Collectively, these results suggest that IL-6 promotes corneal
inflammation by acting in an autocrine–paracrine fashion to induce
resident corneal cells to make MIP-2 and MIP-1α, which in turn
recruit neutrophils to the virus infection site.
Interleukin-6 (IL-6) is a cytokine that is rapidly produced
at local tissue sites after disruption of homeostasis due to trauma or
infection.
1 It is functionally diverse, influencing the
growth, differentiation, and activation of cells and their expression
of proteins principally in inflammation, immune responses, and
hematopoiesis. To evaluate the inflammatory role of IL-6, studies have
been conducted in gene-disrupted mice. It has been found that
antigen-induced arthritis is absent or substantially reduced in the
IL-6
−/− mice.
2 3 Such mice are
also resistant to experimental autoimmune
encephalomyelitis.
4 5 Administration of exogenous IL-6
restores susceptibility.
4 Other studies have shown that
less tissue injury is observed in turpentine-
6 or
carrageenan-treated IL-6
−/− hosts
7 and that treatment with an anti-IL-6 receptor (IL-6R), mAb is
protective in a murine model of colitis.
8 Thus, IL-6 can
be an important promoter of inflammation.
However, IL-6 can also function as an anti-inflammatory cytokine. For
example, Xing et al.
9 reported that acute inflammatory
responses in animal models of endotoxic lung or endotoxemia were more
severe in IL-6
−/− mice. Administration of
recombinant IL-6 reduced circulating levels of proinflammatory
mediators. IL-6 can also produce an anti-inflammatory effect by
inducing glucocorticoids and natural antagonists of IL-1α (reviewed
in Ref.
10 ). Additionally, IL-6 may exist as a
bystander, neither actively promoting nor inhibiting
inflammation.
11
Ongoing studies in our laboratory have sought to identify and
characterize the mediators that participate in the inflammatory cascade
in the herpes simplex virus (HSV)-1–infected murine cornea. In this
model, the inflammatory response can progress over a 14- to 21-day
period to severe stromal keratitis, causing permanent
blindness.
12 13 14 As part of the early warning alarm
system, IL-6 appears rapidly after both virus
15 and
bacterial infection.
16 17 18 Indeed, it is one of the more
abundantly produced cytokines and can persist as inflammation
progresses.
IL-6 is made by a wide variety of cell types, including those found in
ocular tissue. We have shown that IL-1α and TNF-α can induce human
corneal fibroblasts and epithelial cells to synthesize and secrete
IL-6.
19 Additionally, studies in excised mouse corneas
have revealed that IL-6 mRNA induced by endogenous IL-1α is found in
all three resident corneal cell types and that the corneal epithelial
layer is the most abundant producer of protein.
20 Kanangat
et al.
21 have reported that HSV-1 infection selectively
upregulates IL-6 gene expression in the murine epithelial-like cell
line EMT-6.
Although IL-6 is quickly produced after HSV-1 corneal infection, it is
not known whether it is a necessary component of the inflammatory
cascade. We, therefore, initiated studies to investigate whether IL-6
may be involved in the modulation of chemokine expression. We focused
on macrophage inflammatory protein (MIP)-2 and MIP-1α, because these
chemokines are known to rapidly recruit neutrophils into the cornea
after HSV-1 infection.
22 The experimental approach was to
compare the inflammatory response elicited by HSV-1 intracorneal
infection in IL-6 gene–deficient mice with that of their wild-type
counterparts. In this study, IL-6 was a critical component of the early
inflammatory response. Our data indicate that IL-6 enhanced the
production of MIP-2 and MIP-1α, chemokines important in the
recruitment of neutrophils into the HSV-1–infected cornea.
Four-week-old IL-6 gene–deficient mice and matched control mice
(C57Bl/6 mice) were obtained from Jackson Laboratories (Bar Harbor,
ME). All animals were cared for in accordance with the ARVO Statement
for the Use of Animals in Ophthalmic and Vision Research and in
compliance with federal, state, and local regulations.
Rat IgG mAb to mouse IL-6 and murine recombinant IL-1α were
purchased from Biosource International (Camarillo, CA); recombinant
murine (rm)IL-6 from Endogen, Inc. (Woburn, MA); and hamster
IgG monoclonal antibody to mouse IL-1α from R&D Systems (Minneapolis,
MN). Rat RB6-8C5 monoclonal antibody (mAb), a gift from Robert Coffman
(DNAX Research Institute, Palo Alto, CA), was prepared as
previously described.
23 RB6-8C5 mAb reacts with the Ly-6G
antigen, which is expressed on neutrophils and other granulocytes at a
level that directly correlates to the differentiation and maturation
stage of the cell.
Intracorneal injection was accomplished by first puncturing the
corneal epithelium wall with a 30-gauge disposable needle. A 30-cm
32-gauge stainless steel needle attached to a dispenser (Hamilton,
Reno, NV) was then threaded into the corneal stroma and 1.0 μL of
solution containing the appropriate dose of HSV-1 strain RE
(103–105 plaque-forming
units [PFU]) was injected. In some experiments the 1.0-μL inoculum
contained a mixture of HSV-1 (105 PFU/μL)
admixed with mAb to IL-6 (5 μg/μL) or rmIL-6 (50 ng/μL). These
mixtures were prepared immediately before use and directly
administered. Pilot studies established that neither addition of IL-6
nor antibody to IL-6 impaired virus infectivity. Subconjunctival
antibody injections were performed by using a 2-cm 32-gauge needle and
syringe (Hamilton) to penetrate the perivascular region of the
conjunctiva and deliver 4 μL into the subconjunctival space. To score
corneal opacity, eyes were graded on a scale of 0 to +4 by visual
observation, using a dissecting biomicroscope. Eyes were graded as
follows: 0, clear cornea; +1, slight corneal haze; +2, moderate corneal
opacity; +3, severe corneal opacity with visible iris; +4, severe
corneal opacity with iris not visible. Eyes were examined in a coded
manner, with the reader unaware of the treatment administered.
After intracorneal injection, the corneas were immediately
excised, trimmed with a 2-mm trephine, and incubated individually in
200 to 250 μL RPMI-1640 medium in a polypropylene tube at 37°C in
5% CO2 for 10 hours. The medium, with or
without, corneas was then frozen at −70°C.
Samples containing corneas were thawed, minced, and processed
for 80 seconds (Tissue Tearer; Biospec Products, Bartlesville, OK),
sonicated for 20 seconds, and clarified by centrifugation at
150g for 10 minutes, thus producing clarified corneal
lysates. Samples (clarified corneal lysates or supernatant) were
assayed by ELISA for MIP-2, MIP-1α, and IL-6. The MIP-2 kit (assay
sensitivity, 1.5 pg/mL), and MIP-1α kit (assay sensitivity, 1.5
pg/mL), were purchased from R&D Systems (Minneapolis, MN). The IL-6 kit
(assay sensitivity, 7.0 pg/mL) were purchased from Endogen, Inc.
Immunohistochemical staining was performed using a slight
modification of the procedure of Hendricks et al.
24 Infected eyes were enucleated and embedded in optimal cutting
temperature compound (Tissue Tek; Sakura Finetek, Torrance, CA) and
6-μm sections were cut at −20°C. The sections were fixed in cold
acetone for 10 minutes and then blocked with normal goat serum. The
primary antibody, RB6-8C5 (5 μg/mL) was applied to the tissue for 1
hour at room temperature. The sections were then washed and stained
using the streptavidin-biotin complex immunoperoxidase staining
procedure, according to the manufacturer’s protocol (Zymed
Laboratories, South San Francisco, CA). The slides were washed in
distilled water and counterstained in Harris hematoxylin for 3 minutes.
The slides were then examined under a light microscope, and pictures
were taken (BX50 camera; Olympus Optical Co., Tokyo, Japan).
Bone marrow (BM) was flushed from the femur and tibia with media
(RPMI-1640 with 5% newborn calf serum) using a syringe and 25-gauge
needle. The BM cells were centrifuged (1200 rpm for 5 minutes), washed
twice, layered over a gradient (Histopaque 1119 and 1077; Sigma, St.
Louis, MO), and centrifuged at 700
g for 30 minutes. The
enriched neutrophil layer was removed, washed twice in medium and
treated with red blood cell lysis buffer (Sigma). Contaminating
monocytes were depleted by adherence (30 minutes at 37°C) to a
polystyrene tissue culture plate
25 (Corning Glass Co.,
Corning, NY). Neutrophil purity was consistently more than 99%, as
assessed by staining (HEMA 3; Biochemical Sciences Inc., Swedesboro,
NJ) of cytospin slides (Shandon, Pittsburgh, PA).
For stimulation 1 × 106 neutrophils in 0.5
mL medium were placed in triplicate in 24-well tissue culture plates
(Corning Glass Co.). The wells were precoated with 0.5 mL newborn calf
serum per well for 1 hour at 37°C and then washed three times with
PBS. Stimulations were performed by the addition of rmIL-6 to wells at
a final concentration of 5 or 50 ng/mL. rmIL-1α was used at 1 or 10
ng/mL. Phorbol 12-myristae, 13 acetate (PMA; 30 ng/mL) served
as the positive control and medium only as the negative control. After
incubation for 8 hours at 37°C in 5% CO2 supernatants were removed, clarified by centrifugation, and assayed for
chemokine content by ELISA.
The Mann-Whitney test was used to determine significant
differences in the corneal opacity scores between treated and control
groups. Student’s t-test was used to evaluate mediator
responses between treated and control cells. The level of confidence at
which the results were judged significant was P <
0.05.
When HSV-1 is inoculated intracorneally a rapid inflammatory
response is elicited, which is evident by 24 hours and peaks at 48
hours after infection.
22 We investigated whether
disruption of the IL-6 gene would influence this initial response. It
was found that IL-6
−/− animals challenged with
virus over a 100-fold dose range consistently showed significantly less
corneal cloudiness than did IL-6
+/+ mice, as
determined by microscopic observation during the first 2 days
postinfection.
Figure 1 shows representative results from one of five independent experiments.
This striking disparity was confirmed histologically:
IL-6
−/− corneas were less swollen and contained
many fewer neutrophils positively stained with RB6-8C5 mAb than did
corneas from IL-6
+/+ hosts
(Fig. 2) . The difference in inflammation was transient; the opacity scores of
the two groups were no longer significantly different 4 days after
infection. The distinct disparity in corneal responsiveness observed
during the first 2 days could not be attributed to a difference in
HSV-1 replication as the titers of infectious virus in the wild-type
mice (1.0 × 10
5 ± 0.4 ×
10
5) were not significantly different
(
P > 0.7) from that observed in the ocular tissue of
knockout hosts (1.2 × 10
5 ± 0.5 ×
10
5). Similar virus titer results were obtained
in two additional independent experiments.
To evaluate whether exogenous IL-6 could restore heightened
corneal opacity, reconstitution experiments were performed in
IL-6
−/− mice. It was found that a single
inoculation of IL-6 administered locally, together with the virus
challenge inoculum restored corneal opacity to the levels seen in
IL-6
+/+ mice at 24 and 48 hours after infection
(
Fig. 3A ; only 48-hour results from a representative experiment are shown). No
corneal cloudiness was observed in animals given 50 ng IL-6 in the
absence of HSV-1. Administration of IL-6 together with the virus
significantly raised the levels of chemokines MIP-2
(Fig. 3B) and
MIP-1α
(Fig. 3C) relative to that seen in control mice treated with
virus only. Furthermore, the amounts of each mediator were comparable
to the levels found in the corneas of the IL-6
+/+ hosts and thus correlated with increased mean corneal opacity scores.
Effect of IL-6 Neutralizing Antibody on Corneal Opacity and MIP-2
Response in HSV-1–Infected IL-6+/+ Hosts
We next determined whether antibody neutralization of endogenous
IL-6 would influence HSV-1 induced corneal opacity in
IL-6
+/+ mice. IL-6 antibody was introduced
intracorneally at a concentration of 2.5 μg/μL when admixed with
the challenge virus. In preliminary studies, it was determined that a
concentration of anti-IL-6 antibody lower than 2.5 μg/μL was not
sufficient to neutralize the effect of IL-6 (data not shown).
Additional antibody was administered subconjunctivally at 12-hour
intervals, and corneas were examined at 48 hours after infection.
Figure 4A shows results from a representative experiment in which corneal
opacity scores in IL-6
+/+ hosts after anti-IL-6
antibody treatment were significantly reduced compared with IgG-treated
control animals and indeed were comparable to that seen in
IL-6
−/− mice. In addition, corneal MIP-2
concentrations were also significantly reduced to levels observed in
IL-6
−/− mice
(Fig. 4B) .
We postulated that HSV-1 infection induces resident corneal cells
to synthesize and secrete IL-6. This hypothesis was investigated using
excised corneas to avoid production of mediators by infiltrating
leukocytes.
Figure 5 shows that after HSV-1 infection, corneal tissue from
IL-6
+/+ hosts had high levels of IL-6, whereas
none was seen in virus-infected IL-6
−/− corneas. Significantly, MIP-2 levels in infected
IL-6
+/+ corneas were 10-fold higher than that
seen in infected corneas without a functional IL-6 gene. This suggests
that IL-6 induces resident corneal cells to make a neutrophil
chemoattractant. Support for this prediction comes from experiments in
which antibody neutralization of IL-6 in HSV-1–infected
IL-6
+/+ corneas abrogated production of MIP-2 and
MIP-1α
(Fig. 6) .
Additionally we found that administration of exogenous IL-6 to
uninfected IL-6
−/− corneal tissue at a 10- to
50-ng dose stimulated a 24- to 45-fold increase in MIP-2 (
Fig. 7A ) and an 8- to 23-fold increase in MIP-1α
(Fig. 7B) . These amounts
were comparable to that stimulated by IL-6 in uninfected wild-type
corneas.
The IL-6 produced by resident corneal cells could also induce
neutrophils recruited to the HSV-1 infection site to produce
chemokines. Accordingly, we tested whether IL-6 stimulates MIP-2 or
MIP-1α synthesis in these cells.
Figure 8 depicts representative data from one of four independent experiments.
Five or 50 ng/mL IL-6 added to bone marrow–derived neutrophils did not
enhance production of MIP-2 or MIP-1α above background levels (a less
than twofold increase). In contrast, IL-1α stimulated a two- to
sixfold increase in MIP-1α. MIP-2 production was also enhanced by
IL-1α in a dose-dependent manner, although the amounts generated were
substantially lower (18- to 31-fold) than MIP-1α. Neutrophils
stimulated with 30 ng/mL PMA produced on average 209 pg/mL MIP-2 and
2198 pg/mL MIP-1α. Incubation of neutrophils with both IL-6 and
MIP-1α did not elevate chemokine production above that seen with
IL-1α alone.
IL-6 failed to stimulate neutrophils purified from the bone marrow of
IL-6−/− mice to produce MIP-2 or MIP-1α when
tested as described. However, IL-1α at 10 ng/mL induced a six- and
ninefold increase in MIP-2 and MIP-1α, respectively, over background
levels. The inability of IL-6 to induce chemokine production by
neutrophils was also observed when the incubation period was extended
from 10 to 18 hours.
In this study, evidence showed IL-6 played a prominent role in
initiating corneal inflammation after HSV-1 infection. Mice with a
disrupted IL-6 gene displayed substantially reduced corneal opacity 24
to 48 hours after intracorneal infection in comparison with their
wild-type counterparts, as judged both clinically and histologically. A
single 50-ng dose of IL-6 administered locally at the time of infection
restored corneal disease to wild-type levels. Furthermore, antibody
neutralization of endogenous IL-6 in IL-6
+/+ hosts reduced corneal inflammation to that observed in
IL-6
−/− animals. The recent report that IL-6
abolishes immune privilege in eyes with endotoxin-induced
uveitis
26 also supports the view that IL-6 is
proinflammatory in ocular tissue.
There are several mechanisms by which HSV-1 infection could induce
IL-6. One is that virus infection may directly activate IL-6 gene
expression in corneal cells.
21 Alternatively, virus
infection may induce IL-1α, which in turn induces IL-6. This latter
possibility is supported by earlier studies that showed that synthesis
of IL-6 stimulated by mechanical trauma in excised corneal buttons was
abrogated by antibody to IL-1α but not antibodies to IL-1β or
TNF-α.
20
Our present data provide evidence that IL-6 promoted induction of
neutrophil chemoattractants. Initial evidence for this conclusion comes
from in vivo studies that show that IL-6 administered to infected
corneas of IL-6
−/− mice elevates MIP-2 and
MIP-1α to wild-type levels and that antibody neutralization of IL-6
locally in wild-type corneas reduces MIP-2 to
IL-6
−/− levels. It is worth noting that our
follow-up ex vivo studies demonstrated that uninfected as well as
infected resident cells in corneal tissue exposed to IL-6 produced
MIP-2 and MIP-1α. Thus, IL-6 induction of chemokines does not require
that the corneal cells be infected. Collectively, a likely scenario
based on our findings is that HSV-1 corneal infection initiates an
inflammatory cascade in which IL-1α induces IL-6 and then these two
cytokines through autocrine–paracrine action stimulate resident
corneal cells to produce neutrophil chemoattractants. An earlier report
showing that IL-1α inoculated into the mouse cornea results in the
production of MIP-2
22 is compatible with this conclusion.
In contrast to resident corneal cells, exposure of purified mouse
neutrophils to recombinant IL-6 caused little or no elevation of MIP-2
or MIP-1α. However, IL-1α induced substantial levels of MIP-1α
and more modest but significant amounts of MIP-2. This suggests that
resident cells in the cornea are the principal targets of IL-6, whereas
IL-1α acts on both resident corneal cells and infiltrating
leukocytes. Whether IL-6 can induce infected corneal cells or
infiltrating neutrophils to make and secrete other proinflammatory
mediators remains to be determined. Additionally, IL-6 may also
influence the effector responses of neutrophils.
27
It has been reported that IL-10 is produced constitutively in the mouse
cornea and acts to antagonize development of virus-induced
inflammation.
28 It may be speculated that IL-6 enhances
corneal disease because it suppresses endogenous IL-10 expression.
However, in three independent experiments we found that administration
of IL-6 to IL-6
−/− mice did not alter IL-10
corneal levels (Fenton and Lausch, unpublished observations,
2001). In contrast, antibody neutralization of endogenous IL-10
significantly enhanced production of IL-6, MIP-2, and
MIP-1α.
28 We conclude that HSV-1 infection of
IL-6–deficient mice did not result in upregulation of the
anti-inflammatory cytokine IL-10 as was seen in
Candida
albicans infection.
29
IL-6
−/− mice have been observed to produce a
lower antibody response or a reduced cytotoxic T-lymphocyte response to
certain selected viral pathogens
30 but not
others.
31 We found that although HSV-1 ocular titers in
IL-6
−/− mice were similar to those of
IL-6
+/+ hosts at 48 hours after infection 8 of 15
(53%) showed development of fatal encephalitis, whereas only 1 of 15
(7%) IL-6
+/+ mice died. These results are in
agreement with the report of LeBlanc et al.
32 Neutrophils
and mononuclear cells are known to help limit HSV-1 growth and spread
to the central nervous system after ocular infection.
23 33 Their diminished early recruitment due to reduced MIP-2 and MIP-1α
production would be expected to increase
IL-6
−/− host susceptibility to HSV-1. Our
results may also help to explain the impaired neutrophil recruitment
and increased susceptibility seen in IL-6
−/− mice infected with
Listeria monocytogenes 34 and
C. albicans. 29 In addition, or alternatively,
the stronger inflammatory response in the IL-6
+/+ cornea may generate a cytokine milieu that favors a greater, and thus
more protective, T helper cell-1 response.
35 36
In summary, our findings showed IL-6 to be an important participant in
the cytokine cascade triggered by HSV-1 corneal infection. By inducing
neutrophil chemoattractants IL-6 amplified the inflammatory response.
Our data are in agreement with the report of Smith et
al.,
37 wherein IL-6 was associated with MIP-1α
expression in a bleomycin-induced lung injury model. Also, Romano et
al.
35 observed that IL-6
−/− mice
exhibit reduced chemokine production and defective leukocyte
recruitment in subcutaneous air pouches injected with inflammatory
stimuli. This defect could be reversed by IL-6 administration. In
follow-up studies IL-6 itself did not induce human endothelial cells,
which have no IL-6R, to make chemokine, but IL-6–soluble
(s)IL-6R complexes were active. Collectively, these studies and ours
indicate that chemokine production can be induced in distinctly
different cell types by IL-6 per se, or by IL-6–sIL-6R complexes. This
suggests that IL-6 in one form or another may promote chemokine
expression in a variety of tissues. Thus, IL-6 represents a potential
therapeutic target for suppressing ocular inflammation.