TREATMENT WITH ANTI-TUMOR NECROSIS FACTOR a DOES NOT INFLUENCE THE IMMUNE PATHOLOGICAL RESPONSE AGAINST LYMPHOCYTIC CHORIOMENINGITIS VIRUS T.P. Leist,*j’ and R.M. Zinkernagel The role of tumor necrosis factor alpha (TNF LY)in the immunopathological events induced by infection with lymphocytic choriomeningitis (LCM) virus (LCMV) was assessed by treatment of C57Bl/6 mice with a sheep antibody to murine TNF LYantiserum to strongly interfere with anti-Listeria host defense. However, despite its effectiveness in Listeria infections in vivo, antibody to TNF Q!used at 6 x lo4 neutralizing units per day subcutaneously had no detectable influence on the kinetics of maturation of antiviral cytotoxic T-cell activity, inflammatory processes, or clearance of virus. First, onset and severity of LCMV-induced hepatitis, as assessed by cytotoxic T-cell activity, viral titers in the liver, serum liver enzyme values, and histology, were not detectably affected by antibody to TNF (Y. Second, incidence of lethal LCM disease after intracerebral infection and the kinetics of the primary footpad swelling reaction observed after local foot inoculation were not altered by anti-TNF ar antibody treatment. From the data presented we conclude that TNF cuas assayed by in vivo therapy with a polyclonal anti-TNF (Y antibody plays no detectable role in the host reaction against LCMV.
o 1990by W.B. SaundersCompany.
There is accumulating evidence that macrophagederived tumor necrosis factor (TNF) a! is crucially involved in many tissue injuries (Cerami and Beutler, 1988). Studies utilizing neutralizing antibodies against this cytokine revealed its importance in resistance to bacterial infections during the T-cell independent phase of the host response (Hauser et al., 1989; Havell, 1987; Nakane et al., 1988). In viral infections the role of macrophages and monocytes in mediating inflammation and causing tissue damage are only poorly understood (Allison, 1974; Blanden, 1974; Lehmann-Grube et al., 1987; Thomsen and Marker, 1983; Marker and Thomsen, 1987). In this study we evaluated whether TNF (Y influenced disease parameters in murine LCMV infections with the use of a specific polyclonal sheep antiTNF (Y antiserum that did not detectably cross-react with Interferon (IFN) CX,p, or y when tested in vitro.
From the Institut ftir Pathologie Universittitsspital Zurich, 8091 Zurich, Switzerland. *To whom reprint requests should be sent. tPresent address: Institute of Microbiology and Immunology, University of California-Los Angeles School of Medicine, Los Angeles, CA, 90024- 1147. 0 1990 by W.B. Saunders Company. 1043-4666/90/0201-0006$5.00/O KEY WORDS: tumor necrosis factor; lymphocytic chorioeningitis virus CYTOKINE,
Vol. 2, No. 1 (January),
1990: pp 29-34
MATERIALS
AND
METHODS
Mice Six- to eight-week-old
inbred C57B1/6
(H-2b) and out-
bred ICR (H-2q) mice of either sex were obtained from the breeding colony of the Institut fiir Zuchthygiene, Tierspital, Zurich, Switzerland.
Antisera and Treatment of Animals For the production of anti-TNF (Yantiserum, a sheep was initially immunized intramuscularly with 100 Kg of recombinant murine TNF (Y titering 1 to 2 x 10’ U/mg (Biogent, Gent, Belgium) in Freund’s complete adjuvant. The sheep was boosted five times in the course of seven months with the same amount of recombinant TNF LYpresented in Freund’s incomplete adjuvant. With this immunization procedure, neutralizing titers of 3 x lo5 neutralizing U/ml were obtained. The serum was partially purified by ammonium sulfate precipitation. Doses of 6 x IO4 neutralizing U of each anti-TNF (Y antiserum were given in a 0.2 ml balanced salt solution for the entire period indicated for each individual experiment. Control mice were injected with a similarly treated preimmunization serum. Mice were treated daily with subcutaneous injections into the right flank or locally into the right thigh. Animals were infected with the indicated doses of LCMV or Listeria monocytogenes either intravenously, intracerebrally, or alternatively into the footpad. 29
30 1 Leist
and Zinkernagel
viruses Lymphocytic choriomeningitis virus (LCMV) WE was originally obtained from F. Lehmann-Grube, Hamburg (Lehmann-Grube, 1982). Viral dilutions were prepared in minimal essential medium containing 1 to 5% of heat inactivated fetal calf serum. Plaque forming units (pfu) were determined on L cellsas described by Lehmann-Grube et al. (1985).
Culturing and Enumeration of Bacteria A seedof Listeria monocytogenes was originally obtained from R.V. Blanden, Australian National University, Canberra (Blanden and Langman, 1972) and was maintained in a virulent state by passagein mice. The lethal dose for 50% of the animals was about 1.5 x 10’ colony forming units (cfu) for C57B1/6 mice. A frozen stock was used to prepare a fresh 12 to 16 h culture in Trypticase soy broth (BBL Microbiology System, Cockeysville, MD) for each experiment and the infectious dose was assessedretrospectively by plating each inoculum. Viable bacterial counts were determined in individual spleensand livers as described by Mackaness (Mackaness, 1962).
Determination of Footpad Swelling Reaction and LCMV Titers in Organs Mice were injected into the footpads with either viral preparations or organ homogenates. Livers were homogenized in glass tubes with Teflon pestles in 4 ml of medium. Because of toxicity for the cell monolayers, low viral titers could not be determined by plaque assaysin liver homogenates diluted 1:lO and 1:lOO. Therefore, all viral dilutions were performed as follows. The frozen and thawed samples were serially diluted lo-fold in medium before 30-~1aliquots were injected into four footpads. Swelling of the footpads was measured from day 5 to 15 after injection (Lehmann-Grube, 1982; Hotchin, 1971; Zinkernagel et al., 1985) with a spring-loaded caliber (Kroplein, Schluchtern, Hessen, FRG), and the greatest dilution that still gave a positive reaction in at least two of four footpads was multiplied by the initial dilution factor and taken as the ID,, of LCMV. In comparative assaysone pfu of LCMV WE corresponded to about 10 in vivo infectious doses(Lehmann-Grube et al., 1985).
LCM Disease in anti-TNF cr Treated Mice C57B1/6 mice treated with anti-TNF a antiserum were infected intracerebrally with 100 pfu of LCMV WE and the incidence of death due to the induced T-cell immune response was monitored (Lehmann-Grube, 1982; Hotchin, 197 1; Zinkernagel et al., 1985).
Cytoxicity Assay Activity of cytotoxic T cells was tested in a “Cr release assay (Zinkernagel et al., 1985) using MC57G (H-2b) and L929 (H-2’) as target cells. Groups of three to five adult mice were infected intravenously with either LCMV WE at time points indicated for the individual experiments prior to the measurement of cytotoxicity. Briefly, single cell suspensionsof spleen cellswere assayedin duplicate. Effector and target cells (1 x 104) were each added in 0.1 ml volumes to round-bottom
CYTOKINE,
Vol. 2, No.
1 (January1990:29-34)
96-well plates to yield effector to target ratios of 70:1, 23:1, and 8: 1. After mixing, the plates were spun at 1,500g for 6 min and incubated at 37OC in air containing 5% CO,. The test duration was 5 h. The percentage of the specific “Cr release was calculated as (experimental cpm - spontaneous cpm) x 100. maximal cpm - spontaneous cpm Spontaneous releasewas determined in supernatants of targets plus 0.1 ml medium and maximal release was measured with target cells incubated with 0.1 ml of 1 N HCl.
Determination of Serum Enzyme Concentrations Serum concentrations of aspartate aminotransferase (AST, EC.2.6.1.1), alanine aminotransferase (ALT, EC.2.6.1.2), glutamate dehydrogenase (GLDH, EC.1.4.1.3), and alkaline phosphatase (AP, EC.3.1.3.1) were determined on a Hitachi 705 Selective analyzer (Hitachi Ltd., Nacka Works, Japan) as described earlier (Zinkernagel et al., 1986).
RESULTS Characterization of the Sheepanti-TNF (Y Antibody Preparation As shown elsewhere (Hauser et al., 1989), antiTNF cr antiserum did not detectably cross-react with IFN (Y, /3, or y when tested in vitro. Anti-TNF cr
antiserum treatment of C57B1/6 mice infected intravenously with Listeria monocytogenes resulted in enhanced Listeria titers in both liver and spleen. Additional evidence is given in a later section that anti-TNF (Y antiserum injected subcutaneously enhances spreading and persistence of Listeria after peripheral infection into the footpad (Fig. 1).
Comparison of the EJkts of anti-TNF a on the Courseof Infections with LCMV and Listeria Earlier experiments using anti-IFN y antiserum had shown that local injection of the antiserum into the thigh revealed effects of Listeria infection in the corresponding footpad more readily than systemic application of the antiserum (Leist et al., 1988b). We analyzed the anti-TNF (Y antiserum under analogous conditions in mice infected with L. monocytogenes in the following protocol to evaluate effectiveness of the treatment. C57B1/6 mice were inoculated with 2 x lo5 plaque forming units (pfu) of LCMV WE or 120 colony forming units (cfu) of Listeria into the right hind footpad and treated subcutaneously with a daily injection of anti-TNF (Y antiserum into the right thigh. Anti-TNF a-treated animals developed antiviral cytotoxic T-cell activities in the liver and spleen that were not significantly different from those of normal serumtreated animals (Fig. 1C). Five days after inoculation, anti-LCMV cytotoxic T-ceil activities measured in liver
Effect of anti-TNF LYon antiviral host response / 31
pfu) in both organs but were able to control the spread of the bacterium (Fig. 1; a,b). While both number and size of the lesions observed in the livers and spleens of L. monocytogenes-infected mice that had received antiTNF LYantiserum were increased compared to controls, no difference could be observed between LCMV infected animals that were either injected with anti-TNF (Yantiserum or normal serum (not shown). DAY
Q =
DAY
5
DAY
4.5 -
5 b
3.5 -
z -2
2.5 -
9
1.5. DAY
%
100
4 iii ;
80
8
LIVER
I
Iii
G 2 (I) s
2
2
DAY
5
DAY
8
40 60
20 0 DAY
2
DAY
5
DAY
8
TIME AFTER INFECTION Figure 1. Effect of anti-TNFa on course of infections with LCMV and Listeria. C57B1/6 mice were locally infected with 2 x 10’ pfu of LCMV WE into the right hind footpad and were injected daily with 6 x lo4 NU of anti-TNF a antibody from day 0 to the day prior to testing of cytotoxic T-cell activities in livers and spleens.
and spleen of anti-TNF a-treated animals and controls were at background levels, and by day 8, levels of cytotoxicity were comparable for both experimental groups. In C57B1/6 mice inoculated with 120 pfu of Listeria into the right hind footpad, the difference between anti-TNF a-treated animals and the test groups was about lo- to 30-fold on day 2 and about 50- to loo-fold on day 5 after infection (Fig. 1; a,b). By day 8 most of the control mice had cleared the Listeria from the liver and spleen (three out of five for both normalserum treated and infected, but otherwise unmanipulated, animals), while all anti-TNF a-treated mice still had considerable titers of L. monocytogenes (~-3 log-
Kinetics of anti-LCMV Speci’c Cytotoxic T-cell Responsesin Liver and Spleen of anti-TNF a-Treated C57B1/6 Mice When cytotoxic T-cell responses were monitored in the spleen and liver of anti-TNF (Y antiserum-treated and normal-serum treated C57B1/6 mice intravenously inoculated with either 2 x lo5 or 2 x lo3 pfu of LCMV WE, no significant differences were observed between the experimental groups; one of three similar experiments is shown in Table 1. In mice inoculated with the greater virus dose, levels of cytotoxicity assessed in liver and spleen of both anti-TNF (Y antiserum-treated and normal-serum treated mice were highest six days after infection and declined thereafter. In C57B1/6 mice receiving the lower virus dose, levels of cytotoxicity in both experimental groups peaked on day 9 and were still high 12 days after infection; anti-TNF (Y antiserum treatment had no detectable effect. Conventional histology did not show any difference in the size of the lesions in anti-TNF m-treated animals and those of controls (not shown).
Serum Concentration of Liver Enzymes and LCMV Titers in anti-TNF a-Treated C57B1/6 Mice In both anti-TNF (Yantiserum-treated and normalserum treated C57B1/6 mice intravenously infected with 2 x lo5 pfu of LCMV WE, increased levels of AST, ALT, and GLDH peaking around 9 days after infection was observed. Since the findings are comparable for the three enzymes, only values for GLDH are indicated (Table 1). Serum enzyme concentrations were not significantly different for anti-TNF a-treated mice and controls. In mice initially infected with a loo-fold lower dose of virus (2 x lo3 pfu), no increase of liver enzymes in serum was observed in any of the groups. In addition, viral titers measured in the liver were not detectably influenced by the anti-TNF a! treatment (Table 1).
Failure of anti-TNF (YTreatment to Influence Susceptibility to LCM or the Primary Footpad Swelling Reaction The influence of daily subcutaneous doses of antiTNF (Yantiserum on the course of LCM was tested in
CYTOKINE,
32 1 Leist and Zinkernagel
TABLE spleen.
1.
Treatment
with anti-TNF
u antiserum had no effect on kinetics of anti-LCMV
Vol. 2, No. 1 (January 1990: 29-34)
cytotoxic T-cell activity in liver and
C57Bl/6 mice were infected with 2 x 10’ or 2 x 10’ pfu of the WE isolate of LCMV at the time points indicated prior to testing. The cytotoxicity of the liver and spleen lymphocytes was tested on infected and uninfected MC57G (H-2”) target cells in a “Cr release assay. Spontaneous release was less than 18%. On uninfected MC57G target cells killing was less than 5%. Indicated values represent the means of three mice assayed individually. 96 Specific 5’Cr Release
Spleen
Liver Time after infection
Initial virus
Log,, PFU of Normal
Serum
Anti-TNP
Normal Serum
Anti-TNF
GLDH
* S.E.M. (U/I)
LCMV
Per Liver
dose
(PFU)
25
8
3b
25
8
3
70
23
8
70
23
8
Day 6 Day 9 Day 12
2 x lo5 2 x lo5
95 75
70 54
39 23 18
76 54 32
48 37 18
37 23
37
72 41 35
68 58
58
92 63 60
27 17
2 x lo5
51 24 26
11
25
13
19 11 7
80 f 20 790 + 120 1040 + 290
960 e 180 830+ 310
8.2 f 0.4 7.2 * 0.5 5.5 f 1.1
8.7 f 1.1 6.7 f 0.7 5.9 f 0.5
Day 6 Day 9
2 x lo3 2 x lo3 2 x 10’
42 77
17 63
8
53
23
12
68
33
41
83
74
52
83
67
87
69
54
85
62
43
67
42
14 42 18
67 84 53
49 73 27
25 48 16
15 f 5 25 r 9 20 * 7
15 *4 20 f 6 15 + 5
5.7 f 0.7 2.5 f 0.4 12.2
5.2 f 0.5 2.2 f 1.1 t2.2
Day 12
Normal
serum
Anti-TNF
Normal
110 * 30
serum
Anti-TNF
a
“animals were treated from day 0 to the day prior to testing. beffector to target cell ratio.
C57B1/6 mice that were inoculated intracerebrally with lo* pfu of LCMV WE. One group of mice was also injected intracerebrally with 30 ~1 of anti-TNF (Y antiserum on days 5 and 7 after infection. In either protocol treatment anti-TNF (Yantiserum did not prolong survival of infected animals (not shown). C57B1/6 mice were injected locally into the right footpad with 40 pfu of LCMV WE to measure the primary footpad swelling reaction. Mice treated with anti-TNF (Yantiserum into the right thigh developed swelling with the same kinetics as controls (Fig. 2). All mice shown in Fig. 2 were bled on day 10 after infection and the presence of virus was monitored; all mice had cleared the virus by that time (data not shown). DISCUSSION Despite its effects in infections with L. monocytogenes (Fig. 1; a,b), anti-TNF (Yantiserum treatment of C57B1/6 mice infected either locally or systemically with LCMV WE did not detectably affect virus-induced disease patterns when compared with controls. Mice inoculated intracerebrally with LCMV WE succumbed to LCM 8 to 10 days after inoculation irrespective of the treatment they received. LCM is associated with massive infiltration of inflammatory cells into the choriomeninges and is strictly CD8+ T-cell dependent (Hotchin, 1971; Cole, 1972; Doherty and Zinkernagel, 1974; Zinkernagel et al., 1985; Leist et al., 1987; Ceredig et al., 1987). Macrophages and monocytes make up a considerable subpopulation of the infiltrate (Schwendenmann et al., 1983; Allan et al., 1987). They and their resident counterpart in the brain, the microglia, are probably activated in view of the substantial amounts of IFN y that are found in the cerebrospinal
fluid after onset of LCM (Frei et al., 1988). It also has been known for some time that LCMV infection causes systemic activation of macrophages and monocytes revealed by increased bactericidal activity (Allison, 1974; Blanden, 1974). However, the demonstrated lack of an influence of anti-TNF (Yantiserum on the progression of LCM together with the earlier reported absence of TNF (Y from the cerebrospinal fluid even in the preterminal stage (Leist et al., 1988a) make it unlikely that this interleukin plays a significant role in this LCM. Overall these findings suggest that the recruited macro-
140 130 120 110
100 5
7 TIME
AFTER
9 INFECTION
11 (DAYS)
13
Figure 2. Primary footpad swelling reaction in anti-TNFa treated mice. C57B1/6 mice were locally infected with 40 pfu of LCMV WE into the footpad and were injected from day 0 to day 13 with daily doses of 6 x lo4 NU of anti-TNF a! antibody (Cl), normal serum (0) or did not receive any further treatment (+). Footpad thickness was monitored daily thereafter and compared with preinfection values. Indicated values are the measured increase of footpad thickness in percent of the preinfection value and are means of four individual footpads: S.E.M. was less than 8%.
Effect of anti-TNF (Yon antiviral host response / 33
phage and monocyte-mediated infiltrate and the virus specific cytotoxic T cells do not mediate the inflammatory damage via TNF LY(Thomsen and Marker, 1983; Lehmann-Grube et al., 1987; Marker and Thomsen, 1987). These findings are paralleled by those obtained in the footpad injection experiments. Antibody to TNF a! had no detectable influence on either time course or magnitude of the footpad swelling reaction. In a recent study Doherty et al. (1989) reported that in LCMV-infected animals administration of TNF a! could cause a delay or an acceleration of death within a few hours, dependent on the severity of the preexisting inflammation. Compatible findings were reported earlier for endotoxin treatment (Barlow, 1964) during LCM. In view of (i) the inefficiency of anti-TNF CYantibody to modulate LCM and (ii) the absence of detectable amounts of TNF a in the cerebrospinal fluid at any time during an LCMV infection, it seems likely that the effect of endotoxin via TNF (Y, or of TNF n directly, reflects the actions of this cytokine in an established inflammatory process without revealing a significant role in the pathobiology of LCM. In this study, treatment with anti-TNF a antibody did not affect kinetics, amplitude, or virus dose dependency of the generation or the recruitment of cytotoxic T cells to the liver and spleen, as measured in “Cr release assays (Table 1, Fig. 1). Also, viral replication and the progression of inflammation and cell destruction, assayed by liver enzyme levels in serum or by histology (not shown), were not altered when compared to controls (Table 1). While contribution of T cells to the pathophysiology of LCMV infections is well documented, it is less clear to what extent macrophages and monocytes are involved in these processes (Thomsen and Marker, 1983; LehmannGrube et al., 1987; Marker and Thomsen, 1987). While this study cannot evaluate the overall role of macrophages and monocytes versus cytotoxic T cells in an anti-LCMV host response, the data presented suggest that their mode of action is not detectably mediated via TNF LY. Acknowledgments
The authors would like to thank J. Tavernier, Biogent, Gent, for continuous support with recombinant TNF cr; K. Frei, Section of Clinical Immunology, University Hospital, Zurich, for conventional TNF (Y assays, and E. Haenseler, Institute of Medical Chemistry, University Hospital, Zurich, for the determination of serum enzyme values. This work was supported by Grant No. 3.259-0.87 of the Swiss National Foundation. REFERENCES Allan JE, Dixon JE, Doherty PC (1987): Nature of the inflammatory process in the central nervous system of mice infected with lymphocytic choriomeningitis virus. Curr Top Microbial Immunol 134: 13 l143
Allison AC (1974): On the role of mononuclear phagocytes in immunity against viruses. Prog Med Virol 18:lS Barlow JL (1964): Hyperreactivity to endotoxin in mice infected with lymphocytic choriomeningitis virus, in Landy M, Braun W (eds) Bacteria1 Endotoxins. Quinn and Boden, Rahway, NJ Blanden RV, Langman RE (1972): Cell mediated immunity to bacterial infection in the mouse. Thymus derived cells as effecters of acquired resistance to Listeria monocytogenes. Stand J Immunol 1:379-391 Blanden RV (1974): T cell response to viral and bacterial infection. Transplant Rev 1956-88 Cerami A, Beutler B (1988): The role of cachectin/TNF in endotoxic shock and cachexia. Immunol Today 9:28-31 Ceredig R, Allan JE, Tabi Z, Lynch F, Doherty PC (1987): Phenotypic analysis of the inflammatory exudate in murine lymphocytic choriomeningitis. J Exp Med 1651539-1551 Cole GA, Nathanson N, Prendergast RA (1972): Requirement for O-bearing cells in lymphocytic choriomeningitis virus-induced central nervous system disease. Nature 238:335-337 Doherty PC, Zinkernagel RM (1974): T cell-mediated immunopathology in viral infections. Transplant Rev 19:89-120 Doherty PC, Allan JE, Clark IA (1989): Tumor necrosis factor inhibits the development of viral meningitis or induces rapid death depending on the severity of inflammation at time of administration. J Immunol142:3576-3580 Frei K, Leist TP, Meager A, Gallo P, Leppert D, Zinkernagel RM, Fontana A (1988): Production of B cell stimulatory factor-2 and interferon y in the central nervous system during viral meningitis and encephalitis. Evaluation in a murine model infection and in patients. J Exp Med 168:449-453 Hauser T, Zinkernagel RM, Leist TP (1989): Effects of anti-tumor necrosis is factor a on Listeria resistance in nude mice (Submitted) Have11EA (1987): Production of tumor necrosis factor during murine Listeriosis. J Immunol 139:4225-423 1 Hotchin J (1971): Persistent and slow virus infections. Monogr Virol 3, Karger (Basel). Lehmann-Grube F (1982): Lymphocytic choriomeningitis virus, in Foster HL, Small JD, Fox JG (eds) The Mouse in Biomedical Research, Vol. II, Diseases, Academic, San Diego, CA. p 23 1-266 Lehmann-Grube F, Assmann U, LSliger C, Moskophidis D, Lohler J (1985): Mechanism of recovery from acute virus infection. I. Role of T lymphocytes in the clearance of lymphocytic choriomeningitis virus from spleens of mice. J Immunol 134:608-615 Lehmann-Grube F, Krenz I, Krahnert T, Schwachenwald R, Moskophidis D, Lijhler J, Villeda-Posada C (1987): Mechanism of recovery of acute virus infection. IV. Questionable role of mononuclear phagocytes in clearance of lymphocytic choriomeningitis virus from spleens of mice. J Immunol 138:2282-2289 Leist TP, Cobbold SP, Waldmann H, Aguet M, Zinkernagel RM (1987): Functional analysis of T lymphocyte subsets in antiviral host defense. J Immunol 138:2278-2281 Leist TP, Frei K, Kam-Hansen S, Zinkernagel RM, Fontana A (1988a): Tumor necrosis factor a in cerebrospinal fluid during bacterial but not viral meningitis. Evaluation in murine model infections and in patients. J Exp Med 167:1743-1748 Leist TP, Heuchel R, Zinkernagel RM (1988b): Increased bactericidal macrophage activity induced by immunological stimuli is dependent on interferon-y. Interference of anti-IFN y but not anti-IFN (Y/P with modulation of macrophage activity caused by lymphocytic choriomeningitis virus infection or systemic graft-vs.host reactions. Eur J Immunol 18:1295-1298 Mackaness GBM (1962): Cellular resistance to infection. J Exp Med 116:381 Marker 0, Thomsen AR (1987): Clearance of virus by T lymphocytes mediating delayed type hypersensitivity. Curr Top Microbial Immuno1 134:145-184
34 / Leist and Zinkernagel
Nakane A, Minagawa T, Kato K (1988): Endogenous tumor necrosis factor cachectin is essential to host resistance against Listeria monocytogenes infection. Infect Immun 56:2563-2569 Schwendenmann G, Lohler J, Lehmann-Grube F (1983): Evidence for cytotoxic T-lymphocyte-target cell interaction in brains of mice infected intracerebrally with lymphocytic choriomeningitis virus. ActaNeuropathol (Berl) 61:183-195 Thomsen AR, Marker 0 (1983): Studies on the role of mononuclear phagocytes in resistance to acute lymphocytic choriomeningitis virus infection. &and J Immunol18:271-277
CYTOKINE,
Vol. 2, No. 1 (January 1990: 29-34)
Zinkernagel RM, Leist T, Hengartner H, Althage A (1985): Susceptibility to lymphocytic choriomeningitis virus isolates correlates directly with early and high cytotoxic T cell activity, as well as with footpad swelling reaction, and all three are regulated by H-2D. J Exp Med 162:2125-2141 Zinkernagel RM, Haenseler E, Leist T, Cerny A, Hengartner H, Althage A (1986): T cell-mediated hepatitis in mice infected with lymphocytic choriomeningitis virus. Liver cell destruction by H-2 class I-restricted virus-specific cytotoxic T cells as a physiological correlate of the %-release assay? J Exp Med 1641075-1092
o 1990by W.B. SaundersCompany.
There is accumulating evidence that macrophagederived tumor necrosis factor (TNF) a! is crucially involved in many tissue injuries (Cerami and Beutler, 1988). Studies utilizing neutralizing antibodies against this cytokine revealed its importance in resistance to bacterial infections during the T-cell independent phase of the host response (Hauser et al., 1989; Havell, 1987; Nakane et al., 1988). In viral infections the role of macrophages and monocytes in mediating inflammation and causing tissue damage are only poorly understood (Allison, 1974; Blanden, 1974; Lehmann-Grube et al., 1987; Thomsen and Marker, 1983; Marker and Thomsen, 1987). In this study we evaluated whether TNF (Y influenced disease parameters in murine LCMV infections with the use of a specific polyclonal sheep antiTNF (Y antiserum that did not detectably cross-react with Interferon (IFN) CX,p, or y when tested in vitro.
From the Institut ftir Pathologie Universittitsspital Zurich, 8091 Zurich, Switzerland. *To whom reprint requests should be sent. tPresent address: Institute of Microbiology and Immunology, University of California-Los Angeles School of Medicine, Los Angeles, CA, 90024- 1147. 0 1990 by W.B. Saunders Company. 1043-4666/90/0201-0006$5.00/O KEY WORDS: tumor necrosis factor; lymphocytic chorioeningitis virus CYTOKINE,
Vol. 2, No. 1 (January),
1990: pp 29-34
MATERIALS
AND
METHODS
Mice Six- to eight-week-old
inbred C57B1/6
(H-2b) and out-
bred ICR (H-2q) mice of either sex were obtained from the breeding colony of the Institut fiir Zuchthygiene, Tierspital, Zurich, Switzerland.
Antisera and Treatment of Animals For the production of anti-TNF (Yantiserum, a sheep was initially immunized intramuscularly with 100 Kg of recombinant murine TNF (Y titering 1 to 2 x 10’ U/mg (Biogent, Gent, Belgium) in Freund’s complete adjuvant. The sheep was boosted five times in the course of seven months with the same amount of recombinant TNF LYpresented in Freund’s incomplete adjuvant. With this immunization procedure, neutralizing titers of 3 x lo5 neutralizing U/ml were obtained. The serum was partially purified by ammonium sulfate precipitation. Doses of 6 x IO4 neutralizing U of each anti-TNF (Y antiserum were given in a 0.2 ml balanced salt solution for the entire period indicated for each individual experiment. Control mice were injected with a similarly treated preimmunization serum. Mice were treated daily with subcutaneous injections into the right flank or locally into the right thigh. Animals were infected with the indicated doses of LCMV or Listeria monocytogenes either intravenously, intracerebrally, or alternatively into the footpad. 29
30 1 Leist
and Zinkernagel
viruses Lymphocytic choriomeningitis virus (LCMV) WE was originally obtained from F. Lehmann-Grube, Hamburg (Lehmann-Grube, 1982). Viral dilutions were prepared in minimal essential medium containing 1 to 5% of heat inactivated fetal calf serum. Plaque forming units (pfu) were determined on L cellsas described by Lehmann-Grube et al. (1985).
Culturing and Enumeration of Bacteria A seedof Listeria monocytogenes was originally obtained from R.V. Blanden, Australian National University, Canberra (Blanden and Langman, 1972) and was maintained in a virulent state by passagein mice. The lethal dose for 50% of the animals was about 1.5 x 10’ colony forming units (cfu) for C57B1/6 mice. A frozen stock was used to prepare a fresh 12 to 16 h culture in Trypticase soy broth (BBL Microbiology System, Cockeysville, MD) for each experiment and the infectious dose was assessedretrospectively by plating each inoculum. Viable bacterial counts were determined in individual spleensand livers as described by Mackaness (Mackaness, 1962).
Determination of Footpad Swelling Reaction and LCMV Titers in Organs Mice were injected into the footpads with either viral preparations or organ homogenates. Livers were homogenized in glass tubes with Teflon pestles in 4 ml of medium. Because of toxicity for the cell monolayers, low viral titers could not be determined by plaque assaysin liver homogenates diluted 1:lO and 1:lOO. Therefore, all viral dilutions were performed as follows. The frozen and thawed samples were serially diluted lo-fold in medium before 30-~1aliquots were injected into four footpads. Swelling of the footpads was measured from day 5 to 15 after injection (Lehmann-Grube, 1982; Hotchin, 1971; Zinkernagel et al., 1985) with a spring-loaded caliber (Kroplein, Schluchtern, Hessen, FRG), and the greatest dilution that still gave a positive reaction in at least two of four footpads was multiplied by the initial dilution factor and taken as the ID,, of LCMV. In comparative assaysone pfu of LCMV WE corresponded to about 10 in vivo infectious doses(Lehmann-Grube et al., 1985).
LCM Disease in anti-TNF cr Treated Mice C57B1/6 mice treated with anti-TNF a antiserum were infected intracerebrally with 100 pfu of LCMV WE and the incidence of death due to the induced T-cell immune response was monitored (Lehmann-Grube, 1982; Hotchin, 197 1; Zinkernagel et al., 1985).
Cytoxicity Assay Activity of cytotoxic T cells was tested in a “Cr release assay (Zinkernagel et al., 1985) using MC57G (H-2b) and L929 (H-2’) as target cells. Groups of three to five adult mice were infected intravenously with either LCMV WE at time points indicated for the individual experiments prior to the measurement of cytotoxicity. Briefly, single cell suspensionsof spleen cellswere assayedin duplicate. Effector and target cells (1 x 104) were each added in 0.1 ml volumes to round-bottom
CYTOKINE,
Vol. 2, No.
1 (January1990:29-34)
96-well plates to yield effector to target ratios of 70:1, 23:1, and 8: 1. After mixing, the plates were spun at 1,500g for 6 min and incubated at 37OC in air containing 5% CO,. The test duration was 5 h. The percentage of the specific “Cr release was calculated as (experimental cpm - spontaneous cpm) x 100. maximal cpm - spontaneous cpm Spontaneous releasewas determined in supernatants of targets plus 0.1 ml medium and maximal release was measured with target cells incubated with 0.1 ml of 1 N HCl.
Determination of Serum Enzyme Concentrations Serum concentrations of aspartate aminotransferase (AST, EC.2.6.1.1), alanine aminotransferase (ALT, EC.2.6.1.2), glutamate dehydrogenase (GLDH, EC.1.4.1.3), and alkaline phosphatase (AP, EC.3.1.3.1) were determined on a Hitachi 705 Selective analyzer (Hitachi Ltd., Nacka Works, Japan) as described earlier (Zinkernagel et al., 1986).
RESULTS Characterization of the Sheepanti-TNF (Y Antibody Preparation As shown elsewhere (Hauser et al., 1989), antiTNF cr antiserum did not detectably cross-react with IFN (Y, /3, or y when tested in vitro. Anti-TNF cr
antiserum treatment of C57B1/6 mice infected intravenously with Listeria monocytogenes resulted in enhanced Listeria titers in both liver and spleen. Additional evidence is given in a later section that anti-TNF (Y antiserum injected subcutaneously enhances spreading and persistence of Listeria after peripheral infection into the footpad (Fig. 1).
Comparison of the EJkts of anti-TNF a on the Courseof Infections with LCMV and Listeria Earlier experiments using anti-IFN y antiserum had shown that local injection of the antiserum into the thigh revealed effects of Listeria infection in the corresponding footpad more readily than systemic application of the antiserum (Leist et al., 1988b). We analyzed the anti-TNF (Y antiserum under analogous conditions in mice infected with L. monocytogenes in the following protocol to evaluate effectiveness of the treatment. C57B1/6 mice were inoculated with 2 x lo5 plaque forming units (pfu) of LCMV WE or 120 colony forming units (cfu) of Listeria into the right hind footpad and treated subcutaneously with a daily injection of anti-TNF (Y antiserum into the right thigh. Anti-TNF a-treated animals developed antiviral cytotoxic T-cell activities in the liver and spleen that were not significantly different from those of normal serumtreated animals (Fig. 1C). Five days after inoculation, anti-LCMV cytotoxic T-ceil activities measured in liver
Effect of anti-TNF LYon antiviral host response / 31
pfu) in both organs but were able to control the spread of the bacterium (Fig. 1; a,b). While both number and size of the lesions observed in the livers and spleens of L. monocytogenes-infected mice that had received antiTNF LYantiserum were increased compared to controls, no difference could be observed between LCMV infected animals that were either injected with anti-TNF (Yantiserum or normal serum (not shown). DAY
Q =
DAY
5
DAY
4.5 -
5 b
3.5 -
z -2
2.5 -
9
1.5. DAY
%
100
4 iii ;
80
8
LIVER
I
Iii
G 2 (I) s
2
2
DAY
5
DAY
8
40 60
20 0 DAY
2
DAY
5
DAY
8
TIME AFTER INFECTION Figure 1. Effect of anti-TNFa on course of infections with LCMV and Listeria. C57B1/6 mice were locally infected with 2 x 10’ pfu of LCMV WE into the right hind footpad and were injected daily with 6 x lo4 NU of anti-TNF a antibody from day 0 to the day prior to testing of cytotoxic T-cell activities in livers and spleens.
and spleen of anti-TNF a-treated animals and controls were at background levels, and by day 8, levels of cytotoxicity were comparable for both experimental groups. In C57B1/6 mice inoculated with 120 pfu of Listeria into the right hind footpad, the difference between anti-TNF a-treated animals and the test groups was about lo- to 30-fold on day 2 and about 50- to loo-fold on day 5 after infection (Fig. 1; a,b). By day 8 most of the control mice had cleared the Listeria from the liver and spleen (three out of five for both normalserum treated and infected, but otherwise unmanipulated, animals), while all anti-TNF a-treated mice still had considerable titers of L. monocytogenes (~-3 log-
Kinetics of anti-LCMV Speci’c Cytotoxic T-cell Responsesin Liver and Spleen of anti-TNF a-Treated C57B1/6 Mice When cytotoxic T-cell responses were monitored in the spleen and liver of anti-TNF (Y antiserum-treated and normal-serum treated C57B1/6 mice intravenously inoculated with either 2 x lo5 or 2 x lo3 pfu of LCMV WE, no significant differences were observed between the experimental groups; one of three similar experiments is shown in Table 1. In mice inoculated with the greater virus dose, levels of cytotoxicity assessed in liver and spleen of both anti-TNF (Y antiserum-treated and normal-serum treated mice were highest six days after infection and declined thereafter. In C57B1/6 mice receiving the lower virus dose, levels of cytotoxicity in both experimental groups peaked on day 9 and were still high 12 days after infection; anti-TNF (Y antiserum treatment had no detectable effect. Conventional histology did not show any difference in the size of the lesions in anti-TNF m-treated animals and those of controls (not shown).
Serum Concentration of Liver Enzymes and LCMV Titers in anti-TNF a-Treated C57B1/6 Mice In both anti-TNF (Yantiserum-treated and normalserum treated C57B1/6 mice intravenously infected with 2 x lo5 pfu of LCMV WE, increased levels of AST, ALT, and GLDH peaking around 9 days after infection was observed. Since the findings are comparable for the three enzymes, only values for GLDH are indicated (Table 1). Serum enzyme concentrations were not significantly different for anti-TNF a-treated mice and controls. In mice initially infected with a loo-fold lower dose of virus (2 x lo3 pfu), no increase of liver enzymes in serum was observed in any of the groups. In addition, viral titers measured in the liver were not detectably influenced by the anti-TNF a! treatment (Table 1).
Failure of anti-TNF (YTreatment to Influence Susceptibility to LCM or the Primary Footpad Swelling Reaction The influence of daily subcutaneous doses of antiTNF (Yantiserum on the course of LCM was tested in
CYTOKINE,
32 1 Leist and Zinkernagel
TABLE spleen.
1.
Treatment
with anti-TNF
u antiserum had no effect on kinetics of anti-LCMV
Vol. 2, No. 1 (January 1990: 29-34)
cytotoxic T-cell activity in liver and
C57Bl/6 mice were infected with 2 x 10’ or 2 x 10’ pfu of the WE isolate of LCMV at the time points indicated prior to testing. The cytotoxicity of the liver and spleen lymphocytes was tested on infected and uninfected MC57G (H-2”) target cells in a “Cr release assay. Spontaneous release was less than 18%. On uninfected MC57G target cells killing was less than 5%. Indicated values represent the means of three mice assayed individually. 96 Specific 5’Cr Release
Spleen
Liver Time after infection
Initial virus
Log,, PFU of Normal
Serum
Anti-TNP
Normal Serum
Anti-TNF
GLDH
* S.E.M. (U/I)
LCMV
Per Liver
dose
(PFU)
25
8
3b
25
8
3
70
23
8
70
23
8
Day 6 Day 9 Day 12
2 x lo5 2 x lo5
95 75
70 54
39 23 18
76 54 32
48 37 18
37 23
37
72 41 35
68 58
58
92 63 60
27 17
2 x lo5
51 24 26
11
25
13
19 11 7
80 f 20 790 + 120 1040 + 290
960 e 180 830+ 310
8.2 f 0.4 7.2 * 0.5 5.5 f 1.1
8.7 f 1.1 6.7 f 0.7 5.9 f 0.5
Day 6 Day 9
2 x lo3 2 x lo3 2 x 10’
42 77
17 63
8
53
23
12
68
33
41
83
74
52
83
67
87
69
54
85
62
43
67
42
14 42 18
67 84 53
49 73 27
25 48 16
15 f 5 25 r 9 20 * 7
15 *4 20 f 6 15 + 5
5.7 f 0.7 2.5 f 0.4 12.2
5.2 f 0.5 2.2 f 1.1 t2.2
Day 12
Normal
serum
Anti-TNF
Normal
110 * 30
serum
Anti-TNF
a
“animals were treated from day 0 to the day prior to testing. beffector to target cell ratio.
C57B1/6 mice that were inoculated intracerebrally with lo* pfu of LCMV WE. One group of mice was also injected intracerebrally with 30 ~1 of anti-TNF (Y antiserum on days 5 and 7 after infection. In either protocol treatment anti-TNF (Yantiserum did not prolong survival of infected animals (not shown). C57B1/6 mice were injected locally into the right footpad with 40 pfu of LCMV WE to measure the primary footpad swelling reaction. Mice treated with anti-TNF (Yantiserum into the right thigh developed swelling with the same kinetics as controls (Fig. 2). All mice shown in Fig. 2 were bled on day 10 after infection and the presence of virus was monitored; all mice had cleared the virus by that time (data not shown). DISCUSSION Despite its effects in infections with L. monocytogenes (Fig. 1; a,b), anti-TNF (Yantiserum treatment of C57B1/6 mice infected either locally or systemically with LCMV WE did not detectably affect virus-induced disease patterns when compared with controls. Mice inoculated intracerebrally with LCMV WE succumbed to LCM 8 to 10 days after inoculation irrespective of the treatment they received. LCM is associated with massive infiltration of inflammatory cells into the choriomeninges and is strictly CD8+ T-cell dependent (Hotchin, 1971; Cole, 1972; Doherty and Zinkernagel, 1974; Zinkernagel et al., 1985; Leist et al., 1987; Ceredig et al., 1987). Macrophages and monocytes make up a considerable subpopulation of the infiltrate (Schwendenmann et al., 1983; Allan et al., 1987). They and their resident counterpart in the brain, the microglia, are probably activated in view of the substantial amounts of IFN y that are found in the cerebrospinal
fluid after onset of LCM (Frei et al., 1988). It also has been known for some time that LCMV infection causes systemic activation of macrophages and monocytes revealed by increased bactericidal activity (Allison, 1974; Blanden, 1974). However, the demonstrated lack of an influence of anti-TNF (Yantiserum on the progression of LCM together with the earlier reported absence of TNF (Y from the cerebrospinal fluid even in the preterminal stage (Leist et al., 1988a) make it unlikely that this interleukin plays a significant role in this LCM. Overall these findings suggest that the recruited macro-
140 130 120 110
100 5
7 TIME
AFTER
9 INFECTION
11 (DAYS)
13
Figure 2. Primary footpad swelling reaction in anti-TNFa treated mice. C57B1/6 mice were locally infected with 40 pfu of LCMV WE into the footpad and were injected from day 0 to day 13 with daily doses of 6 x lo4 NU of anti-TNF a! antibody (Cl), normal serum (0) or did not receive any further treatment (+). Footpad thickness was monitored daily thereafter and compared with preinfection values. Indicated values are the measured increase of footpad thickness in percent of the preinfection value and are means of four individual footpads: S.E.M. was less than 8%.
Effect of anti-TNF (Yon antiviral host response / 33
phage and monocyte-mediated infiltrate and the virus specific cytotoxic T cells do not mediate the inflammatory damage via TNF LY(Thomsen and Marker, 1983; Lehmann-Grube et al., 1987; Marker and Thomsen, 1987). These findings are paralleled by those obtained in the footpad injection experiments. Antibody to TNF a! had no detectable influence on either time course or magnitude of the footpad swelling reaction. In a recent study Doherty et al. (1989) reported that in LCMV-infected animals administration of TNF a! could cause a delay or an acceleration of death within a few hours, dependent on the severity of the preexisting inflammation. Compatible findings were reported earlier for endotoxin treatment (Barlow, 1964) during LCM. In view of (i) the inefficiency of anti-TNF CYantibody to modulate LCM and (ii) the absence of detectable amounts of TNF a in the cerebrospinal fluid at any time during an LCMV infection, it seems likely that the effect of endotoxin via TNF (Y, or of TNF n directly, reflects the actions of this cytokine in an established inflammatory process without revealing a significant role in the pathobiology of LCM. In this study, treatment with anti-TNF a antibody did not affect kinetics, amplitude, or virus dose dependency of the generation or the recruitment of cytotoxic T cells to the liver and spleen, as measured in “Cr release assays (Table 1, Fig. 1). Also, viral replication and the progression of inflammation and cell destruction, assayed by liver enzyme levels in serum or by histology (not shown), were not altered when compared to controls (Table 1). While contribution of T cells to the pathophysiology of LCMV infections is well documented, it is less clear to what extent macrophages and monocytes are involved in these processes (Thomsen and Marker, 1983; LehmannGrube et al., 1987; Marker and Thomsen, 1987). While this study cannot evaluate the overall role of macrophages and monocytes versus cytotoxic T cells in an anti-LCMV host response, the data presented suggest that their mode of action is not detectably mediated via TNF LY. Acknowledgments
The authors would like to thank J. Tavernier, Biogent, Gent, for continuous support with recombinant TNF cr; K. Frei, Section of Clinical Immunology, University Hospital, Zurich, for conventional TNF (Y assays, and E. Haenseler, Institute of Medical Chemistry, University Hospital, Zurich, for the determination of serum enzyme values. This work was supported by Grant No. 3.259-0.87 of the Swiss National Foundation. REFERENCES Allan JE, Dixon JE, Doherty PC (1987): Nature of the inflammatory process in the central nervous system of mice infected with lymphocytic choriomeningitis virus. Curr Top Microbial Immunol 134: 13 l143
Allison AC (1974): On the role of mononuclear phagocytes in immunity against viruses. Prog Med Virol 18:lS Barlow JL (1964): Hyperreactivity to endotoxin in mice infected with lymphocytic choriomeningitis virus, in Landy M, Braun W (eds) Bacteria1 Endotoxins. Quinn and Boden, Rahway, NJ Blanden RV, Langman RE (1972): Cell mediated immunity to bacterial infection in the mouse. Thymus derived cells as effecters of acquired resistance to Listeria monocytogenes. Stand J Immunol 1:379-391 Blanden RV (1974): T cell response to viral and bacterial infection. Transplant Rev 1956-88 Cerami A, Beutler B (1988): The role of cachectin/TNF in endotoxic shock and cachexia. Immunol Today 9:28-31 Ceredig R, Allan JE, Tabi Z, Lynch F, Doherty PC (1987): Phenotypic analysis of the inflammatory exudate in murine lymphocytic choriomeningitis. J Exp Med 1651539-1551 Cole GA, Nathanson N, Prendergast RA (1972): Requirement for O-bearing cells in lymphocytic choriomeningitis virus-induced central nervous system disease. Nature 238:335-337 Doherty PC, Zinkernagel RM (1974): T cell-mediated immunopathology in viral infections. Transplant Rev 19:89-120 Doherty PC, Allan JE, Clark IA (1989): Tumor necrosis factor inhibits the development of viral meningitis or induces rapid death depending on the severity of inflammation at time of administration. J Immunol142:3576-3580 Frei K, Leist TP, Meager A, Gallo P, Leppert D, Zinkernagel RM, Fontana A (1988): Production of B cell stimulatory factor-2 and interferon y in the central nervous system during viral meningitis and encephalitis. Evaluation in a murine model infection and in patients. J Exp Med 168:449-453 Hauser T, Zinkernagel RM, Leist TP (1989): Effects of anti-tumor necrosis is factor a on Listeria resistance in nude mice (Submitted) Have11EA (1987): Production of tumor necrosis factor during murine Listeriosis. J Immunol 139:4225-423 1 Hotchin J (1971): Persistent and slow virus infections. Monogr Virol 3, Karger (Basel). Lehmann-Grube F (1982): Lymphocytic choriomeningitis virus, in Foster HL, Small JD, Fox JG (eds) The Mouse in Biomedical Research, Vol. II, Diseases, Academic, San Diego, CA. p 23 1-266 Lehmann-Grube F, Assmann U, LSliger C, Moskophidis D, Lohler J (1985): Mechanism of recovery from acute virus infection. I. Role of T lymphocytes in the clearance of lymphocytic choriomeningitis virus from spleens of mice. J Immunol 134:608-615 Lehmann-Grube F, Krenz I, Krahnert T, Schwachenwald R, Moskophidis D, Lijhler J, Villeda-Posada C (1987): Mechanism of recovery of acute virus infection. IV. Questionable role of mononuclear phagocytes in clearance of lymphocytic choriomeningitis virus from spleens of mice. J Immunol 138:2282-2289 Leist TP, Cobbold SP, Waldmann H, Aguet M, Zinkernagel RM (1987): Functional analysis of T lymphocyte subsets in antiviral host defense. J Immunol 138:2278-2281 Leist TP, Frei K, Kam-Hansen S, Zinkernagel RM, Fontana A (1988a): Tumor necrosis factor a in cerebrospinal fluid during bacterial but not viral meningitis. Evaluation in murine model infections and in patients. J Exp Med 167:1743-1748 Leist TP, Heuchel R, Zinkernagel RM (1988b): Increased bactericidal macrophage activity induced by immunological stimuli is dependent on interferon-y. Interference of anti-IFN y but not anti-IFN (Y/P with modulation of macrophage activity caused by lymphocytic choriomeningitis virus infection or systemic graft-vs.host reactions. Eur J Immunol 18:1295-1298 Mackaness GBM (1962): Cellular resistance to infection. J Exp Med 116:381 Marker 0, Thomsen AR (1987): Clearance of virus by T lymphocytes mediating delayed type hypersensitivity. Curr Top Microbial Immuno1 134:145-184
34 / Leist and Zinkernagel
Nakane A, Minagawa T, Kato K (1988): Endogenous tumor necrosis factor cachectin is essential to host resistance against Listeria monocytogenes infection. Infect Immun 56:2563-2569 Schwendenmann G, Lohler J, Lehmann-Grube F (1983): Evidence for cytotoxic T-lymphocyte-target cell interaction in brains of mice infected intracerebrally with lymphocytic choriomeningitis virus. ActaNeuropathol (Berl) 61:183-195 Thomsen AR, Marker 0 (1983): Studies on the role of mononuclear phagocytes in resistance to acute lymphocytic choriomeningitis virus infection. &and J Immunol18:271-277
CYTOKINE,
Vol. 2, No. 1 (January 1990: 29-34)
Zinkernagel RM, Leist T, Hengartner H, Althage A (1985): Susceptibility to lymphocytic choriomeningitis virus isolates correlates directly with early and high cytotoxic T cell activity, as well as with footpad swelling reaction, and all three are regulated by H-2D. J Exp Med 162:2125-2141 Zinkernagel RM, Haenseler E, Leist T, Cerny A, Hengartner H, Althage A (1986): T cell-mediated hepatitis in mice infected with lymphocytic choriomeningitis virus. Liver cell destruction by H-2 class I-restricted virus-specific cytotoxic T cells as a physiological correlate of the %-release assay? J Exp Med 1641075-1092
