Role of interferon-γ in murine cytomegalovirus infection

Role of interferon-7 in murine cytomegalovirus infection CLAIRE POMEROY, DAVID DELONG, CONNIE CLABOTS, PAULRICIPUTI,AND GREGORY A. FILICE LEXINGTON, KENTUCKY,AND MINNEAPOLIS,MINNESOTA

Interferon-y has well-documented antiviral and immunomodulatory activity, but its role in the control of cytomegalovirus (CMV) infection is not well studied. In a mouse model of murine CMV (MCMV) disease, interferon-,/concentrations in serum but not in bronchoalveolar lavage fluid increased in response to viral infection. Serum interferon-,/levels peaked at day 2 in the relatively resistant C57BL/6 mice, and, in contrast, did not peak until day 6 in susceptible BALB/c mice. Mice genetically lacking interferon-5, (GKO) were more susceptible to MCMV, although strain differences persisted, with C57BL/6 GKO mice experiencing less severe MCMV disease than BALB/c GKO mice. Treatment of MCMV-infected BALB/c mice with exogenous interferon-./starting 2 days after viral infection had a modest protective effect at lower interferon-5' doses (104 units), but interferon-5' therapy markedly increased morbidity and mortality when higher doses (I 05 units) were used. We conclude that interferon-'/plays a significant role in host response to MCMV and that the cytokine has dose- and time-dependent beneficial and adverse effects. (J Lab Clin Med 1998; 132:124-33)

Abbreviations: BAL = b r o n c h o a l v e o l a r

l a v a g e ; C M V = cytomegalovirus; ELISA = enzyme-linked i m m u n o s o r b e n t assay; GKO = y-interferon knock-out; M C M V = murine c y t o m e g a l o v i r u s ; PFU = p l a q u e - f o r m i n g unit

C

ytomegalovirus is a major cause of morbidity and mortality, especially in immunocompromised patients, and improved therapies are urgently needed. In some forms of CMV disease, especially pneumonia, the disease progresses despite aggressive antiviral therapy, suggesting that immunomodulatory strategies might be advantageous. Interferon-y is a cytokine with antimicrobial and

From the Division of InfectiousDiseases, Department of Medicine, University of Kentucky, Lexington; the Department of Veterans Affairs Medical Center, Lexington;the InfectiousDisease Section, Medicine and Research Services, Department of Veterans Affairs Medical Center, Minneapolis; and the Department of Medicine, University of Minnesota Medical School,Minneapolis. Supported by the Department of VeteransAffairs and the Minnesota Medical Foundation. Submitted for publication June 27, 1997;revision submitted February 23, 1998;acceptedFebruary 26, 1998. Reprint requests: Claire Pomeroy, MD, Associate Professor and Chief, Divisionof InfectiousDiseases, RoomMN672, University of Kentucky Medical Center, 800 Rose Street, Lexington, KY 405360084. 0022-2143/98 $5.00 + 0 5/1/90761

124

immunomodulatory activity critical to host defense against a variety of infectious agents. 1 However, little is known about its role in CMV disease. Mouse models of CMV have proved useful in defining immune responses to the virus. Although type I interferons (c~ and 13) have been well studied in MCMV infection, 2 investigations of interferon-y are more limited.3, 4 Therefore we studied the role of interferon- 7 during MCMV infection in genetically susceptible (BALB/c) and resistant (C57BL/6) mice. 5 We first assayed interferon-y levels during MCMV infection in serum and BAL fluid. Then, to define the functional importance of the cytokine, we determined the course of MCMV disease in mice in which the gene for interferon-y had been deleted (GKO mice). Finally, to determine the potential impact of immunomodulatory therapy, we studied the impact of parenteral interferon-7 on the course of MCMV.

METHODS Mice. Specific pathogen-free 6- to 8-week-old BALB/c AnNHsd and C57BL/6 female mice (average weight of 20 g) were purchased from Harlan Sprague Dawley, Indianapolis,

J Lab Clin Med Volume 132, Number 2

Pomeroy

BALB/c mice

125

C57BL/6 mice

II

400 q I

300 *Death O

2...

E'-"

200

rJ~

100

0

2

4

6/7

9

11

0

2

4

0

2

4

6/7

9

11

0

2

4

6/7

9

11

Days After M C M V Injection 8,0,)0 PFU of M C M V

15,000 PFU of M C M V

8,000 PFU of M C M V

15,000 PFU of M C M V

Fig. 1. Serum interferon-~/concentrations of mice during MCMV infection. Interferon-~/(pg/mL +_.SD) production in serum was measured by ELISA after injection of BALB/c mice and C57BL/6 mice with varying doses of MCMV. *P < .05 as compared with baseline (day 0) concentrations of the cytokine.

IN. Age-matched BALB/c and C57BL/6 GKO mice were purchased as breeder pairs from Harlan Sprague Dawley and bred in our institutional animal facility. All mice were housed in AAALAC-approved facilities and were periodically monitored to ensure that they remained free of significant murine pathogens. Animal studies were reviewed and approved by the Subcommittee on Animal Studies of the Minneapolis Veterans Affairs Medical Center and the Lexington Veterans Affairs Medical Center. Virus. The Smith strain of MCMV was maintained by salivary gland passage in BALB/c mice. Salivary gland homogenates (10% wt/vol) were prepared as described previously 6 and stored in RPMI (GIBCO Laboratories, Grand Island, NY) with 10% fetal calf serum at - 7 0 ° C. For these studies each preparation of virus was titrated on mouse embryo fibroblasts in second passage as described. 6 To establish M C M V infection, mice were injected intraperitoneally with varying sublethal doses of the virus] In this model, symptoms from MCMV infection begin about 2 days after injection, peak 4 to 8 days after injection (with death if a lethal dose of virus

is administered), and subside by 10 to 12 days after injection. Viral replication is detected in all major organs, and histopathologic analysis demonstrates hepatic necrosis and gastrointestinal inflammation but not pneumonitis. 8 Serum a n d BAL sampling. Blood was sampled by retroorbital puncture performed under methoxyflurane anesthesia. Blood was allowed to clot and was spun to collect serum. Serum was stored at -70 ° C. For the collection of BAL fluid, mice were killed and lungs were lavaged once with 1 ml of phosphate-buffered saline solution containing 0.6 mmol/L ethylenediaminetetraacetic acid, as previously described. 9 After centrifugation to remove BAL cells and contaminating red blood cells, the supernatant was stored at -70 ° C for future cytokine assays. 9 Cytokine assays. Levels of interferon-y in BAL fluid and serum were assayed by ELISA with a commercially available kit (Genzyme, Cambridge, MA). For this kit, 1 U of interferon-'/ is considered to be equivalent to 54 pg/ml, and the lower detection limit is 5 pg/ml, per the manufacturer's information. Quantitation of interferon-y in samples was performed

126

J Lab Clin Med August 1998

Pomeroy

by comparison of optical densities in samples with those of standard curves constructed for each study. Interferon treatment. Recombinant murine interferon-]' was a gift from Genentech, South San Francisco, CA. For preliminary kinetic studies, BALB/c mice were injected with 105 U of interferon-y intravenously or intrapefitoneally. For treatment studies, BALB/c mice were injected intraperitoneally with 104 or 105 U of interferon-'f starting 2 days after MCMV (or vehicle) injection and continued daily for the study duration. Mice were observed for mortality, and surviving mice were weighed daily. Moribund mice were killed. Statistical analysis. Studies of interferon-'/concentrations in serum and BAL fluid were performed with a minimum of 8 mice per treatment group per time point. Cytokine ELISA assays were performed in triplicate, and results represent the mean of results for the 8 mice. To determine whether interferon-y concentrations during MCMV infection (day 2, 4, 6/7, 9, or 11) were significantly greater than baseline (day 0), a 1-sample t test was performed. To compare differences between serum interferon-'/levels in BALB/c and C57BL/6 mice at day 2 and at day 4, two-tailed, two-sample tests corrected for unequal variances were used. For studies of MCMV infection in GKO mice, 6 to 9 mice were included in each study group (BALB/c normal, BALB/c GKO, C57BL/6 normal, and C57BL/6 GKO). To compare mortality in normal versus GKO mice, a Fisher's exact test for a 2 x 2 contingency table was used. For treatment studies, 12 mice were studied in each treatment group (total of 9 groups) in 2 separate experiments. To compare survival between treatment groups, a log rank statistic for comparing 2 survival curves was used. In all cases, a P value of < .05 was considered statistically significant. RESULTS Interferon- 7 production in mice with MCMV infection.

To determine the role of interferon-], in host response to M C M V infection, we studied interferon-3' concentrations during M C M V infection in 2 strains of mice with differing susceptibilities to the virus: relatively resistant C57BL/6 mice (lethal dose for 50% of mice = ~1 x 105 PFU) and relatively susceptible BALB/c mice (lethal dose for 50% of mice = -1 x 104 PFU). Because larger doses of M C M V are needed to produce clinically significant disease in C57BL/6 mice than to produce this level of disease in BALB/c mice, we studied mice given amounts of M C M V capable of causing severe but sublethal disease in C57BL/6 mice (15,000 PFU) and BALB/c mice (8000 PFU). For comparison, we also studied C57BL/6 mice injected with 8000 PFU of M C M V (resulting in no apparent clinical disease) and BALB/c mice injected with 15,000 PFU of M C M V (resulting in death after 5 days). Interferon-3' concentrations were assayed in the serum and B A L fluid collected 0, 2, 4, and 6 or 7, 9, and 11 days later.

Table I. MCMV infection in immunocompetent and GKO mice Dose of MCMV (PFU)

BALB~ Mice None 1075 2150 4300 8600 C57BL/6 Mice None 3583 7167 14,333 28,667 57,333

Mortality rate (%) Normal GKO

Max weight loss (%) Normal GKO

O 0 0 0 0

0 0 25 50 75*

None 3 4 7 6

None 16 10 16 24

0 0 0 0 11 40

0 0 0 0 78* 60

None None None 8 8 16

None None 3 8 17 16

*For mortality, P < ,05 (GKO versus normal mice),

Interferon-7 was clearly produced in response to M C M V infection, and there were significant differences between BALB/c and C57BL/6 mice in the timing of increases in interferon- 7 in serum (Fig. 1). As shown, serum interferon- 7 levels peaked at day 2 in resistant C57BL/6 mice but not until day 6 in susceptible BALB/c mice. On day 2, serum interferon- 7 levels in C57BL/6 mice given 15,000 PFU of M C M V were 300.1 _+ 33.5 pg/mL and were only 142.2 _+ 6.3 pg/mL in BALB/c mice given 15,000 PFU of M C M V (P < .01). By day 4, serum interferon- 7 levels in these C57BL/6 mice were already decreasing, to 41.4 + 31.8 pg/mL, whereas interferon- 7 levels in the BALB/c mice were increasing, to 226.9 + 15.6 pg/mL (P < .001). Thereafter, BALB/c mice given 15,000 PFU of M C M V died, reflecting their increased suceptibility to MCMV, whereas C57BL/6 mice continued to recover and had serum interferon- 7 levels that continued to return toward baseline undetectable levels. In contrast, when mice were treated with 8000 PFU of MCMV, C57BL/6 mice had minimal signs of clinical illness and lower serum interferon- 7 levels. Indeed, serum interferon-,/ levels in C57BL/6 mice were not detectable at days 6/7, 9, and 11. BALB/c mice given 8000 PFU of M C M V were sicker than C57BL/6 mice given comparable doses of virus and had higher serum interferon- 7 levels (Fig. 1). Thus it appears that the timing of interferon- 7 production may be an important factor associated with the differing suceptibility o f BALB/c and C57BL/6 mice to M C M V infection. B A L levels of interferon- 7 were comparable in BALB/c and C57BL/6 mice and did not increase to levels above baseline at any time during M C M V infection in either strain of mice (data not shown). This was con-

J Lab Clin Med Volume 132, Number 2

Pomeroy

Intravenous

Intraperitoneal 10 5

Serum

127

BAL

Serum

BAL

10 4

g: o

~J

10 3

10 2

I,m,I

10 1

10 0 -0

4 6

24 48

'0

4 6

24 48

0

4 6

24 48

-0

4 6

24 48

Hours after injection Fig. 2. Interferon-y concentrations in serum and BAL of mice treated with parenteral interferon-y. Interferon-T

levels (pg/mL_+SD) in serum and BAL fluid were measured by ELISA after intravenous and intraperitoneal injection of recombinantinterferon-yinto uninfected BALB/cmice.

sistent with the observation in this mouse model that systemic illness, but not pneumonitis, develops in animals. MCMV infection in immunocompetent versus GKO mice. To determine whether the observed interferon-y

response to MCMV was functionally significant, we studied the course of MCMV infection in interferon-y knock-out (GKO) mice. As shown in Table I, the mortality rate was higher in GKO mice than in their immunocompetent counterparts for both BALB/c and C57BL/6 mice (P = .03 for GKO versus normal BALB/c mice given 8600 PFU of MCMV, and P = .015 for GKO versus C57BL/6 mice given 28,667 PFU of MCMV). Corresponding to the higher mortality rates, BALB/c GKO mice had greater maximum weight loss at each dose of MCMV tested (Table I). Similarly, a trend toward greater weight loss was observed in C57BL/6 GKO mice as compared with immunocompetent C57BL/6 mice (Table I). These results suggest that an interferon-y response is beneficial to the host. However, significant strain differences persisted, with

C57BL/6 GKO mice able to tolerate higher doses of virus than BALB/c GKO mice, suggesting that additional host factors and/or immune responses also contribute to host response to MCMV. Interferon-y treatment of mice with MCMV infection. We hypothesized that treatment with recombinant murine interferon-y might protect susceptible BALB/c mice during MCMV infection. We elected to initiate interferon- 7 treatment 2 days after MCMV injection in BALB/c mice to simulate the timing of interferon- 7 production in resistant C57BL/6 mice. Preliminary studies were performed to determine the appropriate dose of interferon-y to use for treatment studies. In initial experiments, immunocompetent BALB/c mice were injected with a single dose of interferon-y (1 x 105 U) by the intravenous and the intraperitoneal route, and levels of interferon-y were measured in serum and BAL fluid over time (0, 4, 6, 24, and 48 hours after injection). Interferon- 7 levels in serum, but not BAL fluid, increased significantly by 4 and 6 hours after interfer-

128

J Lab Clin Med August 1998

Pomeroy

1 _

0

0

~ ~,.• 9

o

7.500 PFU of MCMV

m

°~

80 70

I

MCMV Only

j

b*"*O

50

t

+ 10,000 U Interferon

t t

4O t,., ¢1

'~1"'"~1/~'"'~'"',~

6O

t t

30 2O

$

t t t

10 0

n

0

•

1

•

/

. . . . . . .

n

2

4

3

+ 100,000 U Interferon

"?0-,"0-,-'(3

5

6

7

8

9

10

Days After MCMV Infection

,oo

80 1

%t_~.

7°1

b

60 1

=,

"t

301 20 -j lO4

"Et"",~'"'a... ~

~ ~ "~ .

0 1

•

0

,

•

1

,

•

2

,

•

,

3

•

\

. . . .

4

5

6

.

.

10.000 PFU of MCMV

°~-.-.A + lO,OOOu Interferon

MC.VOn,y

.

~O'v'O"*~ 7

) + 100,000 U Interferon

8

9

10

Days After MCMV Infection

100 |

"d

90

..>

80 70 60

t

".

~

~'%.

'-',

50 40 I.t

,%

\

~'~

3o 20

~

°,,, •

0

I

1

•

i

2

•

~

I

I

4

5

'~¢'"'/~'"'/~,

k

_

_

"0,,,. ~ "0''0-'0-'0

10, 0

15.000 PFU of MCMV

"~I-.-'A...

'

I

6

,

I

7

•

I

8

•

I

9

+10,000 U Interferon

_ •

MCMV Only + 100,000 U Interferon

I

10

Days After MCMV Infection

Fig. 3. Effect of interferon-7 therapy on mortality rate during MCMV infection. Mortality rate observed in BALB/c mice infected with 7500 PFU, 10,000 PFU, or 15,000 PFU of MCMV and treated with daily intraperitoneal injections of no interferon-y(El), 104 U interferon-T(A), or 105 U interferon-y(O) starting 2 days after viral injection.

on-y injection by both routes (Fig. 2). However, interferon-y levels had returned to values similar to baseline by 24 and 48 hours after injection (Fig. 2). On the basis of these preliminary studies, doses of 104 U and 105 U interferon- 7 delivered every day were used for subsequent treatment studies. These doses are

in the range used previously to treat M C M V and other infectious diseases in mice and were similar to those produced naturally in resistant mice in response to M C M V infection. Because serum levels attained by the intraperitoneal and intravenous route were comparable, intraperitoneal delivery was chosen for ease of admin-

J L a b Clin M e d V o l u m e 132, N u m b e r 2

Pomeroy

15 10 5

ol

o

=l.q

7,500 PFU of MCMV

-10 -15

+ 10,000 O Interferon

-20 -25

MCMV Only

"J*,'~..~_ ~

-30

-35 -40

;, 0

gh

2

,.,, ,,,,Q.~

,.,.;

,.,

4

7

5

+ 100,000 U Interferon

Death

, 8

'1'o

9

Days After MCMV Infection

J= .Jl

10,000 PFU of MCMV

~J

15-,

MCMV Only

10~ 5 0 -5 -10' ¢@

-15" -20"

~J

-25 -." + 10,000 U Interferon

-30 -. -35 -. ¢J /,,,.

+ 100,000 U Interferon

-40

.

,

0

.

,

1

.

,

2

.

,

.

3

.

4

.

.

.

5

.

.

,

6

7

.

.

.

8 9 0 1

Days After MCMV Infection

.mr

1

15,000 PFU of MCMV o

.2

-5

MCMV Only

°

+ 10,000 U Interferon

:2s

cr,.~....ZX

-30 -aS

h °'

-40 ¢J

-"-

"'"O-'O"'~*"O"O . 0 1

,

•

, 2

• ;

4

, 5

.

, 6

.

,

, 7

. 4

, 9

Days After MCMV Infection

+ 100,000 U Interferon

"1'0

Fig. 4. Effect of interferon-y therapy on weight loss during MCMV infection. Weight loss in BALB/c mice infected with 7500 PFU, 10,000 PFU, or 15,000 PFU of MCMV and treated with daily intraperitoneal injections of no interferon-y (C1), 104 U interferon-y (A), or 105 U interferon-y (©) starting 2 days after viral injection.

129

130

J Lab Clin Med August 1998

Pomeroy

istration. For interferon-'/treatment studies, mortality and weight loss (as an indicator of disease severity) were recorded for BALB/c mice injected with 3 different doses of MCMV (7.5 x 103 PFU, 1 x 104 PFU, and 1.5 x 104 PFU) and then treated with 104 U, 105 U, or no interferon-'/. Treatment with 104 U of interferon-'/appeared protective, with a trend toward a lower mortality rate (P = .33 for 7500 PFU, P = .13 for 10,000 PFU, a n d P = .30 for 15,000 PFU when comparing the mortality rate in mice given MCMV alone versus those receiving MCMV plus 104 U interferon-'/; Fig. 3). Analogously, treatment with 104 U of interferon-'/attenuated weight loss after infection with the highest dose of MCMV (Fig. 4). Unexpectedly, treatment with 105 U of interferon-]( was clearly not protective and in fact increased the mortality rate (P = .001 for 7500 PFU, P = .04 for 10,000 PFU, and P = .07 for 15,000 PFU when comparing the mortality rate in mice given MCMV alone versus that in mice receiving MCMV plus 105 U of interferon-'/; Fig. 3). At the lower dose of MCMV from which only 10% of non-interferon-treated animals died, treatment with 105 U of interferon-'/dramatically increased the mortality rate. Thus the effect of interferon-'/was dose dependent, with lower doses tending to provide protection and higher doses producing a deleterious effect with statistically significant increased morbidity and mortality rates. The observation of increased mortality rate in animals treated with 105 U of MCMV could be explained by either a direct toxic effect of interferon-'/or by the possibility that interferon-]' exacerbated the MCMV infection or the associated inflammatory response (or both). To rule out a direct toxic effect of interferon-'/, we performed additional experiments in which mice received either 104 U or 105 U of interferon-'/alone, 7500 PFU of MCMV alone, or 7500 PFU of MCMV plus 105 U of interferon-'/(Table II). We confirmed that BALB/c mice given a sublethal dose of MCMV suffered a significantly increased mortality rate after treatment with 105 U of interferon-'/. However, no mortality was observed in uninfected mice treated with comparable doses of interferon-'/(Table II). Thus a direct toxic effect of interferon-], cannot explain the observation of increased mortality rate in animals infected with MCMV and treated with interferon-'/. DISCUSSION

We report that interferon-'/was produced in response to MCMV disease and that different patterns of production occurred in mice of differing susceptibility to the viral infection. Specifically, serum interferon-'/levels peaked earlier (day 2 after viral infection) in resis-

Table II. Mortality in BALB/c mice with MCMV

infection a n d / o r interferon-y treatment Group

n

Mortality rate

MCMV Alone (7500 PFU) Interferon-yalone 104 U

4

0%

105 U

MCMV plus intefferon-y 105 U

4

0%

5

0%

4

100%

tant C57BL/6 mice than in susceptible BALB/c mice (day 6). Our results are consistent with previous reports that interferon-'/is produced during MCMV infection. Orange et al.11,12 demonstrated that interferon-'/production by NK cells characterizes the early host response to MCMV infection in C57BL/6 mice. In their studies and in ours, interferon-'/production in C57BL/6 mice peaked 2 days after MCMV infection, consistent with a model in which the cytokine is produced by NK cells as part of innate immunity before a specific T lymphocyte response. In another study, BALB/c mice also produced interferon-,/in response to MCMV infection; serum levels peaked 2 days after lethal MCMV infection, although levels were only measured for the first 4 days. 13 This contrasts with our results demonstrating that serum interferon-'/levels did not peak until day 6 or 7 in susceptible BALB/c mice. These discrepant results may be explained by differences in the dose of virus administered, differences in the animal models, or other differences in the study design. To our knowledge, our studies are the first to directly compare interferon-'/production in the two strains of mice and to document that interferon-'/ responses differ between these susceptible and resistant mouse strains. We conclude that although peak serum levels of interferon-'/are comparable in susceptible and resistant mice, differences in the timing of interferon-'/production characterize host immunity to MCMV. Differences in the production of interferon-'/by susceptible and resistant mice in response to MCMV infection were associated with differences in the pathogenesis of disease observed in the two mouse strains. At comparable levels of viral infection, more-severe clinical illness with greater weight loss developed in BALB/c mice and mortality rate increased when results in these mice were compared with those in C57BL/6 mice. Given our observations that the resistant mice produce interferon-'/ early, we speculate that the cytokine may contribute to host immunity, facilitating the initial control of viral replication and reducing disease manifestations in C57BL/6 mice. In contrast, in

J Lab Clin Med Volume 132, Number 2

BALB/c mice, the observed delay in interferon-]( production may contribute to their greater susceptibility to the virus. We hypothesize that the absence of interferon-]( might allow ongoing viral replication during the first few days after infection, resulting in high viral loads in multiple organs. Later, at day 6, when interferon-]( is produced in the susceptible mice, the resulting inflammation in critical organs may contribute to disease manifestations, especially the splenic necrosis and hepatic damage observed in these animals. Future studies assessing viral replication will be necessary to confirm this hypothesis. The cellular source of interferon-]( in response to MCMV infection remains incompletely defined. Interferon-]( can be produced by NK cells, CD4 Thl cells, and CD8 lymphocytes, and it is known that host defense against MCMV is characterized by early NK cell activity and later T cell-mediated responses. 11 In our studies, interferon-]( was detectable in the serum of MCMVinfected C57BL/6 mice within 2 days of infection. The rapidity of cytokine production would suggest that NK cells (rather than specific T cells) are the most likely source of interferon-](. However, in BALB/c mice, interferon-]( was not detectable until 6 days after MCMV infection. This observation raises the possibility that the cellular source of interferon-]( in BALB/c mice might be T lymphocytes as part of specific immunity, or alternatively, that NK cells may produce interferon]( but only after a delay. This latter hypothesis is consistent with previous observations that BALB/c mice appear to have a less-vigorous NK cell response than do resistant strains of mice. 14 To delineate the functional significance of interferon]( in host defense against MCMV infection, we studied MCMV disease in mice genetically altered to be deficient in interferon-7 (GKO mice). Both C57BL/6 GKO and BALB/c GKO mice were significantly more susceptible to MCMV than their immunocompetent counterparts. These findings suggest that the observed changes in serum interferon-]( levels are functionally important in host response against MCMV infection. Nevertheless, we found that strain differences persisted--that is, C57BL/6 GKO mice remained more resistant to MCMV than BALB/c GKO mice. We conclude that interferon-]( is an important factor contributing to host defense against MCMV infection but that other strain-specific differences are important and that, not surprisingly, host immunity against this virus is multifactorial. To further study the role of interferon-]( in the immune response to MCMV infection, we determined the effect of treatment with interferon-]( in susceptible BALB/c mice. We originally hypothesized that interferon-]( would protect susceptible mice from MCMV infection. To simulate the production of interferon-](

Pomeroy

131

observed in resistant mice, we initiated cytokine therapy at day 2 after viral infection. We found a dose-dependent response to interferon-]( treatment. Specifically, the lower dose of 104 U of interferon-]( was associated with protection against MCMV disease whereas the higher dose of 105 U of interferon-]( resulted in increased morbidity and mortality. To explain this, we hypothesize that interferon-]( has two major effects: (1) to limit viral replication, (2) to stimulate NK cell killing of virus-infected cells. In this scenario, limiting viral replication would decrease morbidity, whereas tissue damage to virus-infected cells would damage host cells. If so, the overall impact on the host could be beneficial or deleterious, as was observed in our studies, depending on which of the two effects predominates. An analogous situation of both negative and positive effects of interferon-]( on regulation of CMV in vivo has been reported in the rat model.15,16 Alternatively, the deleterious effects of interferon-]( could result if high doses of interferon-]( up-regulated viral replication, as demonstrated in model systems of other viruses such as HIV and other microbes including Trypanosoma brucei. 18 Previous studies of the impact of interferon-]( treatment on MCMV infection are limited. In vitro, interferon-]( decreases replication of MCMV in microglial cell culture, 19 has limited antiviral activity in fibroblast cultures, 20 and has been shown to have both inhibitory and stimulatory effects on rat CMV-infected cell cultures, depending on the target cell and the amount of cytokine added to the cell culture. 15J6 In one previous report, interferon-]( was found to protect Swiss mice against MCMV infection if immunomodulatory therapy was administered before viral infection but not if therapy was delayed until the day of infection.4 In rats immunosuppressed with lethal irradiation, interferon-]( therapy initiated before rat CMV infection was associated with protection, as evidenced by decreased viral replication and enhanced survival.15,16 Taken together, these previous studies suggest that interferon-]( can protect against CMV disease, but possibly only if it is administered before viral infection. Further, our results now demonstrate that delayed therapy may actually be deleterious. However, it is clear that although interferon-]( is critical to host defense, it is not necessary or sufficient in and of itself. Our finding that C56BL/6 GKO mice remain more resistant to MCMV than do BALB/c GKO mice highlights the multi-component nature of antiviral immunity. Previous investigators have emphasized that other elements of the immune system must be intact. For example, interferon-]( treatment of mice did not decrease viral replication in the salivary glands of irradiated or CD4 celldepleted animals and failed to enhance the anti-viral capacity of MCMV-primed spleen cells transferred to immunodepleted recipient animals.20,21 Further, interfer-

132

Pomeroy

on- 7 therapy did not replace CD8+ cell effector function to clear virus from lungs and spleen of MCMV-infected animals. 22 Depletion of i n t e r f e r o n - ' / w i t h monoclonal antibodies resulted in decreased survival and decreased viral clearance from the salivary gland (where CD4+ cells are important) but not from lung and spleen (where CD8+ cells are important), suggesting that CD8+ cells, but not CD4+ cells, can compensate for lack of interferon-`/and allow the host to successfully clear MCMV. 20,21 Treatment with monoclonal antibody against interferon-~/also decreased the ability of transferred immune spleen cells to protect against viral replication in salivary glands but not in the lung and spleen. 20 W h e n CD8+ spleen cells were used in these transfer studies, anti-interferon-y therapy attenuated viral clearance, 22 suggesting that interferon-y plays an augmentative role in the function of CD8+ cell-mediated viral clearance. Taken together, these reports suggest a protective role for the cytokine in these model systems. However, in other studies, rats treated with anti-interferon-y monoclonal antibody had no increase in mortality rate and had decreased viral replication in the spleen, suggesting that endogenous interferon-,/ m a y actually increase viral replication in the rat model.15, ~6 We conclude that interferon-,/has dose- and time-dependent functions in the immune response to M C M V infection; some of these may be beneficial, whereas others are harmful to the host. Our studies and those o f others demonstrating that i n t e r f e r o n - , / c a n have adverse as well as beneficial effects in hosts infected with viruses and other microbes raise potential concerns about the therapeutic use o f interferon-,/in i m m u n o c o m p r o m i s e d patients, including those with HIV infection or chronic granulomatous disease. 23,24 If intefferon-y therapy in immunocompromised patients has deleterious consequences similar to those observed in our mouse model, interferon-,/therapy may actually exacerbate underlying C M V infections and adversely affect the p a t i e n t ' s overall m e d i c a l course. Further, our findings suggest that cytokine therapy may be more beneficial if administered prophylactically and that delaying initiation of therapy until after a diagnosis of C M V disease has been established m a y abrogate therapeutic efficacy. We suggest that future studies of interferon-,/therapy in immunocompromised patients should include assessment of the possible exacerbation of CMV or other herpesvirus infections. We thank Suzanne Humphreys and Whitney Gatterdam for technical assistance and Dr. Richard Kryscio for statistical consultation. REFERENCES

1. Murray HW. Interferon-gamma, the activated macrophage, and host defense against microbial challenge. Ann Intern Med 1988;108:595-608.

J Lab Clin Med Aug ust 1998

2. Grundy JE, Trapman J, Allan JE, Shellam GR, Melief CJ. Evidence for a protective role of interferon in resistance to murine cytomegalovirus and its control by non-H2-1inked genes. Infect Immun 1982;37:143-50. 3. Leist TE Eppler M, Zinkemagel RM. Enhanced vires replication and inhibition of lymphocytic choriomeningitis virus disease in anti-gamma interferon-treated mice. J Virol 1989;63:2813-9. 4. Fennie EH, Lie YS, Low MA, Gribling P, Anderson KE Reduced mortality in murine cytomegalovirus infected mice following prophylactic murine interferon-y treatment. Antiviral Res 1988;10:27-39. 5. Grundy JE, Mackenzie JS, Stanley NE Influence of H-2 and non-H-2 genes on resistance to murine cytomegalovirus infection. Infect Immun 1981;32:277-86. 6. Pomeroy C, Hilleren PJ, Jordan MC. Latent murine cytomegalovirus DNA in splenic stromal cells of mice. J Virol 1991;65:3330-4. 7. Pomeroy C, Kline S, Jordan MC, Filice GA. Reactivation of Toxoplasma gondii by cytomegalovirus in mice: antimicrobial activities of macrophages. J Infect Dis 1989; 160:305-11. 8. Selgrade MK, Collier AM, Saxton L, Daniels MJ, Graham JA. Comparison of the pathogenesis of murine cytomegalovirus in lung and liver following intraperitoneal or intratracheal infection. J Gen Virol 1984;65:515-23. 9. Pomeroy C, Filice GA, Hitt JA, Jordan MC. Cytomegalovirus-induced reactivation of Toxoplasmagondii pneumonia in mice: lung lymphocyte phenotypes and suppressor function. J Infect Dis 1992; 166:677-81. 10. McCabe RE, Lufi BJ, Remington JS. Effect of murine interferon-'/on murine toxoplasmosis. J Infect Dis 1984;150: 961-2. l 1. Orange JS, Wang B, Terhorst C, Biron CA. Requirement for natural killer cell produced interferon "/in defense against routine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration. J Exp Med 1995; 182:1045 -56. 12. Orange JS, Biron CA. An absolute and restricted requirement for IL-12 in NK cell IFN-y production and antiviral defense. J Immunol 1996; 156:1138-42. 13. Shanley JD, Goff E, Debs RJ, Forman SJ. The role of tumor necrosis factor-c( in acute murine cytomegalovims infection in BALB/c mice. J Infect Dis 1994;169:1088-91. 14. Bancroft GJ, Shellam GR, Chalmer JE. Genetic influences on the augmentation of natural killer cells during murine cytomegalovirus infection: correlation with patterns of resistance. J Immunol 1981; 126:988-94. 15. Haagmans BL, Schijns VECG, Vender Eertwegh AJM, Classen E, Horzinck MC. Role of tumor necrosis factor during cytomegalovirus infection in immunosuppressed rats: activation of vires replication. In: Mossman T, Abbas A, editors. New advances on cytokines. New York: Raven Press, 1992;277-82. 16. Haagmans BL, van der Meide PH, Stals FS, van den Eertwegh AJ, Claassen E, Bruggeman CA, et al. Suppression of rat cytomegalovirus replication by antibodies against gamma interferon. J Virol 1994;68:2305-12. 17. Biswas P, Poli G, Kinter AL, Justement JS, Stanley SK, Maury WJ, et al. Interferon y induces the expression of HIV in persistently infected promonocytic cells (U1) and redirects the production of virions to intracytoplasmic vacuoles in phorbol myristate acetate-differentiated U1 cells. J Exp Med 1992;176:739-50. 18. Olsson T, Bakhiet M, Edlund C, Hojeberg B, Van der Meide

J Lab Clin Med Volume 132, Number 2

PH, Kristensson K. Bidirectional activating signals between Trypanosoma brucei and CD8+ T ceils: a trypanosomereleased factor triggers interferon-gamma production that stimulates parasite growth. Eur J Immunol 1991 ;21:2447-54. 19. Schut RL, Gekker G, Hu S, et al. Cytomegalovirus replication in routine microglial cell cultures: suppression of permissive infection by interferon-gamma. J Infect Dis 1994; 169:1092-6. 20. Lucin R Pavic I, Polic B, Jonjic S, Koszinowski UH. Gamma interferon-dependent clearance of cytomegalovirus infection in salivary glands. J Virol 1992;66:1977-84. 21. Pavic I, Polic B, Crnkovic I, Lucin R Jonjic S, Koszinowski

Pomeroy

133

UH. Participation of endogenous tumor necrosis factor c~in host resistance to cytomegalovirus infection. J Gen Virol 1993;74:2215-23. 22. Hengel H, Lucin P, Jonjic S, Ruppert T, Koszinowski UH. Restoration of cytomegalovirus antigen presentation by gamma interferon combats viral escape. J Virol 1994;68:28997. 23. Heagy W, Groopman J, Schindler J, Finberg R. Use of interferon-gamma in patients with AIDS. Journal of Acquired Immune Deficiency Syndrome 1990;3:584-90. 24. Gallin JI. Interferon-'/in the management of infectious diseases. Ann Intern Med 1995;123:216-24.