Mice infected with an adenovirus mutant in which the E3 region is deleted, including TNF-resistance genes, develop fatal liver pathology within 3–4 days after infection. At least 10-fold more wild-type virus was needed to cause comparable pathology. These results indicate that the E3 region is critically involved in modulating the pathogenesis of adenovirus infection and that TNF may play a role in liver damage. To explore the latter possibility, the course of disease was examined in infected mice lacking TNFR-I and/or TNFRII, TNF only, or both TNF and lymphotoxin-α. Only mice lacking both TNFRI and TNFRII were protected from the lethal affects of the mutant adenovirus. Mice deficient in TNF or TNF and lymphotoxin-α displayed the fatal pathology. This outcome is consistent with the existence of another related ligand that binds TNFRI/II to mediate liver damage during infection with this mutant.

Human adenovirus (Ad)4 is responsible for a range of clinical illnesses that most commonly affect the respiratory and gastrointestinal tracts (1). Human Ad is also of current interest as a potential vector for vaccine and gene therapy (2). The dsDNA genome of Ad is transcribed at early and late phases in relation to onset of DNA synthesis (3). The early (E) region E1A gene products are responsible for multiple biological functions, including susceptibility of virus-infected cells to the cytolytic effect of TNF (4, 5). Early region 3 (E3) of the Ad genome is not required for virus replication in vitro and in vivo (3), allowing deletion of E3 genes and insertion of foreign genes for application of Ad as a vector (1, 6). However, it is now evident that the E3 region codes for proteins that interact with the immune system of the host (reviewed in Refs. 7, 8), and whereas the E1A proteins mediate susceptibility to TNF (4), E3-14.7K, E3-10.4K, and E3-14.5K counteract the effect of TNF on human and mouse cells (9, 10). An additional protein outside the E3 region, E1B-19K also protects against TNF lysis of human cells (11). Thus, the actions of these proteins in concert serve to balance pathogenicity of the virus with the host response, achieving a benign infection.

TNF is a multifunctional cytokine produced by activated macrophages and lymphocytes, as well as other cell types (12). It is expressed in two forms, the 17-kDa soluble form, and its precursor, the 26-kDa membrane form, both of which are bioactive (13). TNF and its close relative, secreted homotrimeric lymphotoxin (LT)-α (LT-α3), bind to two receptors, TNFRI (p55) and TNFRII (p75) that have distinct molecular masses of 55–60 kDa and 75–80 kDa, respectively (12). Most cells express both receptor types, with TNFRI being constitutively expressed at a low level, whereas TNFRII is expressed in response to external stimuli (14, 15). There is ∼30% homology between the extracellular domains of both TNF receptors; however, there is no homology between the intracellular domains, suggesting that they interact with different signaling pathways (16, 17). Generally, activation of TNFRI is associated with cytotoxicity and induction of gene expression, whereas TNFRII expression is associated with thymocyte proliferation and T cell activation (18). Furthermore, TNFRII seems to be the dominant receptor for the 26-kDa membrane form of TNF (19).

Human Ad does not replicate in mouse cells; however, proteins of the early regions of the viral genome, including E1A and E3, are expressed. Previously, we reported the unexpected finding that dl327, an Ad5 mutant lacking the E3 region, caused fatal disease in mice (20). The present study explores the pathogenesis of dl327-induced disease in mice, concentrating in particular on the role of TNF, LT-α3, and their receptors.

The Ad5 wild-type (wt) and the deletion mutant dl327, which lacks most of the E3 region, were propagated in HeLa cells, as described previously (21). Virus stocks were titrated on HeLa cells using the immunofluorescent method, in which rabbit polyclonal anti-Ad5 serum and FITC-conjugated anti-rabbit Ig (Silenus Laboratory, Hawthorn, Australia) were used as primary and secondary Ab, respectively (21). Virus titer was expressed in infectious units per ml (IU/ml). In addition, titration of dl327 in vivo determined the lethal dose in mice. E1A and E3 19-kDa product expression of the two viruses used was confirmed by immunoprecipitation, as has been described (22). The E3 region, represented by E3 19K, was expressed in Ad5 wt-infected cells, but not in dl327-infected cells, and E1A was overexpressed in dl327-infected cells (23).

Ad5 wt was heat inactivated at 60°C for 1 h. Titration as described above confirmed the loss of infectivity. Live and heat-inactivated virus were used to determine distribution of viral DNA in mouse tissue.

HeLa cells and the mouse methylcholanthrane-induced fibrosarcoma cell line, HTG, were cultured in DMEM supplemented with 10% FBS in a 37°C humidified incubator with 5% CO2.

Female C57BL/6J (H-2b) and (B6 × 129)F1 (H-2b) mice were obtained from the specific pathogen-free breeding facilities of the Animal Services Division, the John Curtin School of Medical Research (Canberra, Australia). TNFRI-deficient mice (24) (obtained from H. Blüthmann, Hoffmann-La Roche, Basel, Switzerland) and TNFRII and TNFRI/RII double-deficient mice (generously provided by J. Peschon, Immunex Research and Development, Seattle, WA) (25) were used as indicated. All TNFR-deficient mice used in these studies were hybrids between C57BL/6J and 129/Sv. TNF and TNF/LT-α gene-targeted mice were both generated directly on the C57BL/6J background and have been described (26). Mice in which the gene for LT-α has been inactivated produce neither the secreted LT-α3 homotrimer, nor a functional LT-α1β2 membrane heterotrimer (27). Thus, the TNF/LT-α gene-targeted mice are broadly TNF as well as LT deficient. Unless otherwise specified, all mice used in these experiments were 8–16 wk of age.

For assessment of tissue tropism of the virus, mice were inoculated i.p. with 1 × 108 IU Ad5 wt, live or heat-inactivated, and various tissues were removed and used for DNA preparation. For assessment of the lethality of infection and histopathology, mice were inoculated i.p. with 1–8 × 108 IU of dl327 or up to 8 × 109 IU Ad5 wt. The dose variation between experiments was caused by variation in the dose of virus stock needed to cause death in normal control mice, which was determined before each experiment.

At the prescribed times postinoculation, necropsies were performed, in which segments of the liver were removed and immediately immersed in 10% neutral buffered Formalin. Samples of liver tissue were then embedded in paraffin for sectioning and staining with hematoxylin and eosin. Evaluation of histologic sections was performed in a blinded fashion.

Total DNA was purified from freshly collected or frozen organs of Ad-infected mice. Mouse lung, liver, kidney, spleen, heart, and trachea were homogenized individually and DNA purified using standard methods. A 1012-bp fragment was amplified from the E1 region by nested PCR, initially using the primer pair 5′-CCG GGA GGA TCC ATG AGA CAT ATT ATC TGC C and 3′-CAG CCA CCT CTA GAA CAT TCA TTC CCG, and then for the second pair using the same 5′ primer as above with 3′: CAC AGG TTG AAT TCT TAT GGC CTG GGG CGT T. The annealing temperature was 60°C for both reactions.

We had previously shown (28) that mouse H-2Kk-restricted Ad-5 immune cytotoxic T (Tc) cells lysed target cells infected with dl327 more efficiently than those infected with wt Ad5. The explanation for this phenomenon was not interference by the E3 19k product with MHC class I Ag expression, which affects several allelic products, but not H-2Kk (29). Instead, it was due to increased expression of E1A protein, which contains the H-2Kk-binding peptide determinant recognized by Tc cells (28). Thus, it was of interest to investigate whether dl327 was a better Tc cell immunogen than wt virus in inbred mice. However, we found that mice of several different strains (BALB/c, CBA/H, C3H.H2o, and C57BL/6J) injected with dl327 died 2 to 4 days postinfection at virus doses that, when wt virus was used, caused no mortality or pathology over a period of up to 28 days postinfection. At least 10-fold more wt virus was needed to cause comparable pathology. A representative experiment is shown in Fig. 1 for C57BL/6J mice. At two virus doses (1 × 108 and 2 × 108 IU), no animal infected with the wt Ad5 virus died. Infection with dl327 led to death for most animals by days 3 to 4. Significant morbidity occurred by day 2. Upon autopsy, gross liver damage was observed in dl327-injected animals. Fig. 2 shows histology of liver damage in C57BL/6J mice after 1, 2, and 3 days of infection with dl327. Severe damage can be seen from day 2 onward, with massive hepatocyte destruction, but with little evidence of inflammation or lymphocyte infiltration. No pathology was seen in kidney or lung.

FIGURE 1.

Virulence assay of wt Ad5 and dl327 in C57BL/6 mice. Animals were infected with 1 or 2 × 108 IU wt Ad5 (□) or dl327 (○) i.p., and survival was monitored for 6 days postinfection.

FIGURE 1.

Virulence assay of wt Ad5 and dl327 in C57BL/6 mice. Animals were infected with 1 or 2 × 108 IU wt Ad5 (□) or dl327 (○) i.p., and survival was monitored for 6 days postinfection.

Close modal
FIGURE 2.

Histology of liver in C57BL/6 mice infected i.p. with 2 × 108 IU of dl327. At 24 h after infection (A), the liver appears normal. x250. At 48 h (B) and 72 h (C) after infection, there is massive hepatic cell death without inflammation. Magnification, ×750. Livers of Ad5 wt-infected (2 × 108 IU) mice were normal in appearance at 72 h (D). Magnification, ×750.

FIGURE 2.

Histology of liver in C57BL/6 mice infected i.p. with 2 × 108 IU of dl327. At 24 h after infection (A), the liver appears normal. x250. At 48 h (B) and 72 h (C) after infection, there is massive hepatic cell death without inflammation. Magnification, ×750. Livers of Ad5 wt-infected (2 × 108 IU) mice were normal in appearance at 72 h (D). Magnification, ×750.

Close modal

PCR analysis of DNA from various organs taken on day 3 postinoculation from mice infected with wt Ad5 showed the presence of viral DNA predominantly in the liver (Fig. 3), thus corroborating the observed liver pathology. Viral DNA was not detected in any of the organs derived from mice injected with heat-inactivated virus (data not shown), indicating the requirement for infectivity and specific liver tropism in this system rather than the liver functioning simply as a filter.

FIGURE 3.

Detection of viral DNA sequences in mouse tissue by nested PCR. Tissues of two individual mice infected with 1 × 108 IU of Ad5 wt are shown. The left-hand lane contains markers. Ad5 DNA was detected mainly in liver with lesser amounts in kidney and spleen.

FIGURE 3.

Detection of viral DNA sequences in mouse tissue by nested PCR. Tissues of two individual mice infected with 1 × 108 IU of Ad5 wt are shown. The left-hand lane contains markers. Ad5 DNA was detected mainly in liver with lesser amounts in kidney and spleen.

Close modal

The early onset of liver pathology and the lack of lymphocyte infiltration suggested that the damage was not caused by cell-mediated immunocytolysis. Rather, the kinetics of liver damage correlated with the kinetics of cytokine release in Ad-infected mice, in particular with maximum levels of TNF at 2 to 3 days postinfection (30). In view of the evidence that Ad E3 gene products counteract the effect of TNF during Ad infection in a mouse model, that TNF inhibits Ad replication (31, 32, 33), and that both TNF and LT-α lyse dl327-infected mouse cells in vitro (31, 32, 33), we suspected that the liver pathology was caused by TNF.

Normally resistant cells become sensitive to TNF when infected with Ad E3 deletion mutants. This sensitivity is dependent on E1A (4, 5). E1A is expressed in Ad-infected mouse cells in vivo because E1A-specific cytotoxic T cells can be generated (28). Although we have no evidence for elevated E1A expression in dl327-infected compared with Ad5-infected mice in vivo, such elevation has been shown to occur in mouse cells in vitro infected with E3 deletion mutants, including dl327 (22, 23). Elevated levels of E1A in vivo could possibly be responsible for high sensitivity of hepatocytes of dl327-infected mice to TNF. However, normal hepatocytes in vivo do not divide and results in vitro (5), showing a need for DNA synthesis for increased sensitivity to TNF in E1A-expressing cells may not be applicable. No evidence of mitosis in hepatocytes was seen in liver of dl327-infected mice before development of liver pathology. The lack of expression of E3 proteins that inhibit TNF-mediated cytolysis may be the key issue (9).

Mice deficient in one or both of the TNF receptors or their ligands, TNF and LT-α3, were infected, and Ad-mediated liver pathology was assessed. Table I shows that mice deficient in TNFRI alone incurred the same level of morbidity and mortality as wt mice after infection with dl327. Similarly, a majority of mice deficient in TNFRII alone were not protected from the increased virulence of infection by dl327 (Table II). Both TNFRI-deficient and TNFRII-deficient mice had liver pathology similar to that shown in Fig. 2, B and C. However, when TNFR double knockout mice (TNFRI/II) were used, no deaths occurred under the same conditions (Table II). This result indicates that signaling via either TNFRI or TNFRII is sufficient to mediate severe liver damage in dl327-infected mice, and that protection is afforded only when mice are deficient for both receptors.

Table I.

Effect of Ad dl327a infection on liver in TNFRI-deficient and wt mice

Number of Moribund or Dead Mice/Total Number per Group
TNFRI deficientWild typeb
Day 2 p.i. (moribund) 2 /3 2 /3 
Day 3 p.i. (dead) 4 /4 3 /3 
Number of Moribund or Dead Mice/Total Number per Group
TNFRI deficientWild typeb
Day 2 p.i. (moribund) 2 /3 2 /3 
Day 3 p.i. (dead) 4 /4 3 /3 
a

dl327 was used at 5 × 108 IU injected i.p. This dose was lethal for normal mice as determined with this virus stock immediately prior to the experiment.

b

Wild type refers to (B6 × 129)-F1 hybrid mice.

Table II.

Effect of Ad dl327a on TNFRII-deficient and TNFRI/RII-deficient mice

Expt.Number of Dead Miceb /Total Number per Experiment
TNFRII deficientTNFRI/RII deficient
4 /5 0 /5 
3 /5 0 /5 
3a 2 /5 0 /5 
3bc 0 /5 0 /5 
Expt.Number of Dead Miceb /Total Number per Experiment
TNFRII deficientTNFRI/RII deficient
4 /5 0 /5 
3 /5 0 /5 
3a 2 /5 0 /5 
3bc 0 /5 0 /5 
a

dl327 lethal dose was determined in wt mice immediately prior to each experiment and ranged from 2 × 108 IU to 8 × 108 injected i.p.

b

Mouse deaths occurred between days 2 and 4 p.i. The experiment was terminated on day 5 p.i.

c

Ad5 wt was used in this experiment at the same dose as dl327 (4 × 108 IU).

The question then arises as to which of the known TNFR ligands, TNF or LT-α3, is responsible for liver damage. To determine whether TNF, LT, or both are responsible for the Ad-induced liver damage, C57BL/6J strain TNF- and TNF/LT-α-deficient mice (26) were infected with dl327 and examined. dl327 was titrated in normal C57BL/6J before experimentation. Results with TNF-deficient and TNF/LT-α-deficient mice are shown in Table III. In the absence of TNF or both TNF and LT-α, the animals were not protected and behaved like normal C57BL/6J animals. Comparable severe liver pathology occurred in both groups of deficient mice and was similar in nature and magnitude to that in normal C57BL/6J mice (Fig. 2, B and C).

Table III.

Effect of Ad dl 327a on TNF-deficient and TNF/LTα-deficient mice

Expt.Number of Dead Miceb /Total Number per Experiment
TNF deficientTNF/LTα deficient
4 /5 ND 
3 /4 3 /4 
5 /7 6 /6 
Expt.Number of Dead Miceb /Total Number per Experiment
TNF deficientTNF/LTα deficient
4 /5 ND 
3 /4 3 /4 
5 /7 6 /6 
a

dl327 was used at 8 × 108 IU injected i.p. This dose was lethal for normal mice with this virus stock immediately prior to these experiments.

b

Mouse deaths occurred between days 2 and 4 p.i. The experiment was terminated on day 5 p.i.

The observation that mice deficient for both TNF and LT-α were susceptible to the lethal effects of dl327 was unexpected. We have discussed the unlikely possibility that elevated E1A levels directly mediate liver pathology. An interesting alternative explanation is that a ligand other than TNF or LT-α released during Ad infection is responsible for liver damage and death in these mice. However, the conundrum is that liver pathology failed to occur in the absence of TNFRI and TNFRII, and yet no ligands other than TNF and the LT-α3 homotrimer are known to bind to either of these receptors ( (12). Therefore, the use of the Ad dl327 mutant may have revealed the existence of an alternative ligand for these receptors. This possibility is strengthened by the recent demonstration of LIGHT (34), a cell membrane-associated ligand produced by activated T cells that has strongest homology to LT-β, although it is distinct from this as well as LT-α. Most interesting in the context of the current study is that LIGHT binds not only to the LT-β previously thought to bind predominantly the membrane-associated LT-α1β2 ligand, but also to the TNFR family member, herpes virus entry mediator, present on T lymphocytes. Thus, it seems that viruses such as Ad and herpes may stimulate infected tissue cells or cells of the immune system to produce additional ligands within the TNF superfamily. The role that TNF family ligands may play in virus-induced tissue damage was highlighted in a recent study by Orange and colleagues, who demonstrated a TNF-mediated component to viral pathogenesis during murine CMV infection in mice (35) by neutralization of TNF directly using anti-TNF Ab. Their results indicated a requirement for TNF in development of necrotic but not inflammatory foci in the liver.

We thank Dr. M. Griffith for help with PCR and Drs. M. Lobigs and A. Braithwaite for many discussions.

4

Abbreviations used in this paper: AD, adenovirus; E, early; LT, lymphotoxin; Tc, cytotoxic T; wt, wild type.

1
Horwitz, M. S..
1990
. Adenoviruses. B. N. Fields, and D. M. Knipe, eds.
Virology
2nd Ed.
1723
-1740. Raven Press, New York.
2
Graham, F. L..
1990
. Adenoviruses as expression vectors and recombinant vaccines.
Trends Biotechnol.
8
:
85
3
Horwitz, M. S. 1990. Adenoviridae and their replication. In Virology, 2nd Ed. B. N. Fields and D. M. Knipe, eds. Raven Press, New York, pp. 1679–1721.
4
Duereksen-Hughes, P., W. S. M. Wold, L. R. Gooding.
1989
. Adenovirus E1A renders infected cells sensitive to cytolysis by tumor necrosis factor.
J. Immunol.
143
:
4193
5
Shisler, J., P. Duerksen-Hughes, T. M. Hermiston, W. S. Wold, L. R. Gooding.
1996
. Induction of susceptibility to tumor necrosis factor by E1A is dependent on binding to either p300 or p105-Rb and induction of DNA synthesis.
J. Virol.
70
:
68
6
Wesseling, J. G., G.-J. Godeke, V. E. C. Schijns, L. Prevec, F. L. Graham, M. C. Horzinek, P. J. M. Rottier.
1993
. Mouse hepatitis virus spike and mucleocapsid proteins expressed by adenovirus vectors protect mice against a lethal infection.
J. Gen. Virol.
74
:
2061
7
Wold, W. S. M., L. R. Gooding.
1991
. Region E3 of adenovirus: a cassette of genes involved in host immunosurveillance and virus-cell interactions.
Virology
184
:
1
8
Hayder, H., A. Müllbacher.
1996
. Molecular basis of immune evasion strategies by adenovirus.
Immunol. Cell Biol.
74
:
504
9
Gooding, L. R., I. O. Sofola, A. E. Tollefson, P. Duerksen-Hughes, W. S. Wold.
1990
. The adeonvirus E3-14.7K protein is a general inhibitor of tumor necrosis factor-mediated cytolysis.
J. Immunol.
145
:
3080
10
Gooding, L., T. S. Ranhein, A. E. Tollefspm, L. Aquine, P. Duerksen-Hughes, T. M. Horton, W. S. M. Wold.
1991
. The 10,400- and 14,500-dalton proteins encoded by region E3 of adenovirus function together to protect many but not all mouse cell lines against lysis by tumor necrosis factor.
J. Virol.
65
:
4114
11
Gooding, L. R., L. Aquino, P. J. Duerksen-Hughes, D. Day, T. M. Horton, S. Yei, W. S. M. Wold.
1991
. The E1B 19,000-molecular-weight protein.
J. Virol.
65
:
3083
12
Bazzoni, F., B. Beutler.
1996
. The tumor necrosis factor ligand and receptor families.
N. Engl. J. Med.
334
:
1717
13
Perez, C., I. Albert, K. DeFay, N. Zachariades, L. Gooding, M. Kriegler.
1990
. A nonsecretable cell surface mutant of tumor necrosis factor (TNF) kills by cell-cell contact.
Cell
63
:
251
14
Brockhaus, M., H. J. Schoenfeld, E. J. Schlaeger, W. Hunziker, W. Lesslauer, H. Loetscher.
1990
. Identification of two types of tumor necrosis factor receptors on human cell lines by monoclonal antibodies.
Proc. Natl. Acad. Sci. USA
87
:
3127
15
Thoma, B., M. Grell, K. Pfizenmaier, P. Scheurich.
1990
. Identification of a 60-kD tumor necrosis factor (TNF) receptor as the major signal transducing component in TNF responses.
J. Exp. Med.
172
:
1019
16
Lewis, M., L. A. Tartaglia, G. L. Bennett, G. C. Rice, G. H. Wong, C. Y. Chen, D. V. Goeddel.
1991
. Cloning and expression of cDNAs for two distinct murine tumor necrosis factor receptors demonstrate one receptor is species specific.
Proc. Natl. Acad. Sci. USA
88
:
2830
17
Itoh, N., S. Yonehara, A. Ishii, M. Yonehara, M. Mizushima, A. Hase, Y. Seto, S. Nagata.
1991
. The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis.
Cell
66
:
233
18
Tartaglia, L. A., R. F. Weber, I. S. Figari, C. Reynolds, M. A. Palladino, Jr, D. V. Goeddel.
1991
. The two different receptors for tumor necrosis factor mediate distinct cellular responses.
Proc. Natl. Acad. Sci. USA
88
:
9292
19
Grell, M., E. Douni, H. Wajant, M. Löhden, M. Clauss, B. Maxeiner, S. Georgopoulos, W. Lesslauer, G. Kollias, K. Pfizenmaier, P. Scheurich.
1995
. The transmembrane form of tumor necrosis factor is the prime activating ligand of the 80 kDa tumor necrosis factor receptor.
Cell
83
:
793
20
Baithwaite, A. W., A. J. D. Bellett, G. E. Blair, A. Müllbacher, X. Zhang.
1993
. Human adenoviruses: genetics and immunobiology. D. B. Thomas, Jr, ed.
Viruses and the Cellular Immune Response
389
-428. Marcel Dekker, New York.
21
Braithwaite, A. W., B. F. Cheetham, P. Li, C. R. Parish, L. K. Waldron-Stevens, A. J. D. Bellett.
1983
. Adenovirus-induced alterations of the cell growth cycle: a requirement for expression of E1A but not of E1B.
J. Virol.
45
:
192
22
Zhang, X., A. J. D. Bellett, R. Tha Hla, T. Voss, A. Müllbacher, A. W. Braithwaite.
1994
. Down-regulation of human adenovirus E1a by E3 gene products: evidence for translational control of E1a by E3 4.5K and/or E3 10.4K products.
J. Gen. Virol.
75
:
1943
23
Zhang, X. L., A. J. D. Bellett, R. Tha Hla, A. W. Braithwaite, A. Müllbacher.
1991
. Adenovirus E3 gene products interfere with the expression of the cytolytic T cell immunodominant E1a antigen.
Virology
180
:
199
24
Rothe, J., L. Werner, H. Lötscher, Y. Lang, P. Koebel, F. Köntgen, A. Althage, R. Zinkernagel, M. Stienmetz, H. Bluethmann.
1993
. Mice lacking the tumor necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeriamonocytogenes.
Nature
364
:
798
25
Zheng, L., G. Fisher, R. Miller, J. Peschon, D. H. Lynch, M. J. Lenardo.
1995
. Induction of apoptosis in mature T cells by tumor necrosis factor.
Nature
377
:
348
26
Körner, H., M. Cook, D. S. Riminton, F. A. Lemcker, R. M. Hoek, B. Ledermann, F. Köntgen, B. Fazekas de St Groth, J. D. Sedgwick.
1997
. Distinct roles for lymphotoxin-α and tumor necrosis factor in organogenesis and spatial organization of lymphoid tissue.
Eur. J. Immunol.
27
:
2600
27
Williams-Abbott, L., B. N. Walter, T. C. Cheung, C. R. Goh, A. G. Porter, C. F. Ware.
1997
. The lymphotoxin-α (LTα) subunit is essential for the assembly, but not for the receptor specificity, of the membrane-anchored LTα1β2 heterotrimeric ligand.
J. Biol. Chem.
272
:
19451
28
Müllbacher, A., A. J. D. Bellett, R. Tha Hla.
1989
. The murine cellular immune response to adenovirus-type 5.
Immunol. Cell Biol.
67
:
31
29
Burgert, H. G., S. Kvist.
1985
. An adenovirus type 2 glycoprotein blocks cell surface expression of human histocompatibility class I antigens.
Cell
41
:
987
30
Ginsberg, H. S., L. L. Moldawer, P. B. Seghal, M. Redington, P. L. Kilian, R. M. Chanock, G. A. Prince.
1991
. A mouse model for investigating the molecular pathogenesis of adenovirus pneumonia.
Proc. Natl. Acad. Sci. USA
88
:
1651
31
Mestan, J., W. Digel, S. Mittnacht, H. Hillen, D. Blohm, A. Möller, H. Jacobsen, H. Kirchner.
1986
. Antiviral effects of recombinant tumor necrosis factor in vitro.
Nature
323
:
816
32
Mestan, J., M. Brockhaus, H. Kirchner, H. Jacobsen.
1988
. Antiviral activity of tumor necrosis factor: synergism with interferons and induction of oligo-2′,5′-adenylate synthetase.
J. Gen. Virol.
69
:
3113
33
Wong, G. H. W., D. V. Goeddel.
1986
. Tumor necrosis factors α and β inhibit virus replication and synergize with interferons.
Nature
323
:
819
34
Mauri, D. N., R. Ebner, R. I. Montgomery, K. D. Kochel, T. C. Cheung, G.-L. Yu, S. Ruben, M. Murphy, R. J. Eisenberg, G. H. Cohen, P. G. Spear, C. F. Ware.
1998
. LIGHT, a new member of the TNF superfamily, and lymphotoxin α are ligands for herpesvirus entry mediator.
Immunity
8
:
21
35
Orange, J. S., T. P. Salazar-Mather, S. M. Opal, C. A. Biron.
1997
. Mechanisms for virus-induced liver disease: tumor necrosis factor-mediated pathology independent of natural killer and T cells during murine cytomegalovirus infection.
J. Virol.
71
:
9248