Immunology and Cell Biology (1992) 70, 59-63
Viral escape from immune recognition: Multiple strategies of adenoviruses ARNO MULLBACHER Division of Cell Biology, John Curtin School ofMedical Research, Australian National University, Canberra, Australian Capital Territory, Australia. Summary Human adenovirases can cause persistent infections in man. The strategies of C type adenoviruses (types Ad2 and Ad5) to evade immune recognition are many, but all involve single early genetic regions (E3). Gene(s) within E3 have been shown to allow the adenovirus to avert cytokine-mediated apoptosis. Furthermore, the E3 region controls the cytotoxic T cell epitope (a major histocompatibility complex [MHC] class I molecule with an adenovirus-derived peptide) on the cell surface of infected cells. On the one hand the E3 gene product 19 kDa can bind to nascent class I MHC in the endoplasmic reticulum and thus prevent its transport to the cell surface, and on the other hand the E3 region down-regulates the Ela gene product, the immunodominant cytotoxic T cell determinant.
The immune system of vertebrates has developed a most sophisticated mechanism, as has been illustrated in this issue, to continually adapt to combat parasites. The only constraints placed on the generation of diversity of this powerful defence mechanism is 'self-tolerance, and immune response gene defects are the price to pay'.' By what means do parasites and especially viruses, with their limited 'blank genetic space', evade the host's immune recognition? The great diversity of viruses with respect to their genetic make-up, strategies of infection, replication, and host range suggests that these strategies will differ markedly between virus groups. Table 1 shows a selection of some of the identified strategies used by certain viruses to evade immune recognition. A large number of viruses escape from antibody-mediated immunity through what is called genetic drift.^ Influenza viruses, rota-
Table 1. Virus strategies to escape immune recognition Escape from antibody mediated immunity a) Viral neutralization Genetic drift (orthomyxo-, reo-, picornaviruses) Genetic shift (orthomyxo-, reoviruses) b) Inhibition of complement cascade (Pox viruses) Resistance to cytokines TNF cytolysis (adenoviruses) Lymphotoxin (adenoviruses) Escape from T cell-mediated immunity a) Selection of cytotoxic T cell-resistant variants (LCMV) b) Downregulation of T cell epitopes i) Inhibition of macromolecular synthesis of MHC antigens (herpes-, adeno-, rhabdo-, papoviruses) ii) Competition for P2 microglobulin (CMV)
Correspondence: Arno Mlillbacher, Division of Cell Biology, John Curtin School of Medical Research, Australian National University, PO Box 334, Canberra City, ACT 2601, Australia. Presented at the 4th Frank and Bobbie Fenner Conference in Medical Research on Genetic Variation and Selection in Microbiology and Immunology, held on 6-8 November 1991 at the John Curtin School of Medical Research, Canberra, Australia. Accepted for publication 6 November 1991.
60
A. MUllbacher
viruses and retroviruses (ie. HIV) and rhinoviruses are good examples, the latter showing the most diversity with over 100 serotypes identified. Most of the viruses that undergo extensive genetic drift are RNA or RNA/ DNA viruses with error-prone genetic replication mechanisms. Genetic shifts occur in viruses with segmented genomes as a result of reassortment. Another documented strategy of interfering with humoral immunity is by tampering with the complement cascade as has been shown for pox viruses.^ However, it is not clear what advantage this has given pox viruses per se. Evidence has been obtained that certain viruses have evolved means to escape the antiviral effects of cytokines. One example of this is adenoviruses' ability to interfere with tumour necrosis factor (TNF)- and lymphotoxin-mediated target cell lysis,'''^ most likely by averting apoptosis.^ One of the genes involved in this protection is encoded in the E3 region of the viral genome, a region involved in more than one way in the evasion of immune recognition. The strategies of evading direct T cellmediated immunity are more varied and diverse. First, similar to escape from antibodymediated immunity, it has been shown with virus-specific T cell receptor transgenic mice that selection of cytotoxic T cell (Tc)-resistant variants, with different antigenic specificity, can occur.^ However, the biological significance of this is not quite clear. Major histocompatibility complex (MHC) gene duplication and polymorphism are probably the evolutionary mechanisms to counter such possibilities. Besides evading T cell receptor (TCR) recognition through specific antigenic changes, a number of viruses have been shown to interfere with expression of the epitope, which is a molecular entity comprising an MHC antigen plus viral peptide that is recognized by T cells. Most extensively documented is the down-regulation of expression of MHC class I antigens, which are the signposts for Tc cells.^ This does occur by specific interference with macromolecular synthesis of MHC antigens, at the level of either transcription or translation, and is not solely due to a generalized shutdown of host protein synthesis.
In the case of adenoviruses, cell surface class I MHC antigen expression can be shown to be downregulated in a class I molecule-specific manner,'^'^ by retention of the molecules in the endoplasmic reticulum, thus preventing glycosylation in the golgi apparatus and subsequent transport to the cell surface. Finally, one additional strategy has evolved with human cytomegaloviruses (HCMV), which encode a MHC-like protein that can bind P2 microglobulin (P2m).''* This could make ^2^ less available for MHC class I expression on the cell surface, thus limiting Tc cell recognition of HCMV-infected cells. Reducing availability of P2m may also influence natural killer (NK) cell susceptibility, as it has been shown that ^2^^ 's involved in target cell lysis by virally-induced NK cells.'^•"' It may not be coincidental that all the viruses that have evolved strategies to evade T cell immune recognition can and do produce persistent infections. Here I will present evidence that adenoviruses have evolved an additional mechanism to evade T cell recognition. There are three good reasons to study adenoviruses. First, adenoviruses cause human diseases in their own right. Second, adenoviruses can cause persistent infections in humans and animals, and as adenovirus molecular biology is one of the best understood viral models, their use may offer advantages in learning more of such successful host-parasite relationships. Finally, adenoviruses have recently been identified as candidates for vaccine vectors. Should they ever become viable options to be used for human or veterinary use, a thorough understanding of their biology and behaviour in the host will be essential. We have previously established a model for analysing the murine Tc cell response to C type adenovirus type 5 (Ad5),'^ a summary of which is given in Table 2. The intriguing observation was that adenovirus-immune Tc cells Iysed target cells infected with the E3 deletion mutants more efficiently than when infected with wild-type (w.t.) Ad5 virus. The E3 gene of C type adenovirus encodes a number of open reading frames, and up to six E3-encoded polypeptides have been identified.'^ The longest, the E3 19 kDa protein, has been shown to selectively interfere with the cell surface expression of class I MHC
Adenovirus immune evasion Table 2. Summary of the murine Ad5 immune Tc response* 1. Murine Ad5 immune Tc cells can be generated in vivo and in vitro 2. Tc cells are MHC-class I restricted map to K end in H-2k map to D end in H-2b 3. The immunodominant viral determinant involved in H-2K'' Tc recognition maps to the left 1 3 0 a - a o f E l a 4. Deletions in E3 enhance Tc recognition * From MuUbacher et al. 1989.'^
antigens.'^ However it does not interfere with the expression of H-2K'^ MHC class I antigen, the restriction element for Ad5 immune Tc cells of the murine H-2'' haplotype employed in our studies. We confirmed this lack of inhibition of cell surface expression of H-2K using FACS analysis as well as lysability of w.t. Ad5-f5 E3 deletion mutant-infected target cells by alloreactive K''-specific Tc cells.'^ Two possibilities for E3 gene interference with Tc cell epitope formation were investigated: first, that E3 may interfere with the presentation of peptides rather than with class I MHC cell surface expression; and second, that E3 gene product(s) may interfere with the availability of adenovirus-specific peptide determinants.. The first proposition could be dismissed as we could show that Ad5 w.t.- and E3 deletion mutant-infected target cells were, after superinfection with vaccinia or influenza virus, equally lysable by vaccinia or influenza immune Tc cells, respectively. Thus adenovirus E3 gene product(s) did not interfere with the ability of class I MHC molecules to present vaccinia- or influenza virus-derived peptide determinants to Tc cells. Initial evidence that E3 gene product(s) interfere with the availability of the viral peptide determinant, which together with the class I MHC molecule makes up the Tc epitope, came from studies using different multiplicities of infection of Ad5 w.t. vs E3 deletion mutant viruses for target cells. At high multiplicities (< 100 i.u.) the differential of lysability between Ad5 w.t.- and E3 deletion mutant-infected targets disappeared. Ela
61
is transcribed and translated prior to E3 gene products, and this suggested that high virus input provided plateau levels of Ela products (the viral protein containing the immunodominant peptide determinant) for optimal Tc recognition. Investigating the effects of the E3 gene on transcription of Ela using dot-blot hybridization experiments, no evidence could be obtained for transcriptional control by E3. However, when quantitating Ela protein products in Ad5 w.t.- or E3 deletion mutantinfected target cells by either using Western blot analysis or immunoprecipitation, eight to 10-fold lower amounts were found in Ad5 w.t.-infected cells. Thus E3 gene product(s) modulate the synthesis of the Ela gene products containing the Ad5 immune Tc immunodominant peptide determinants. The mechanism by which this occurs, and which E3 gene product(s) are involved, are dealt with in an accompanying article.'^° Thus the presence of the E3 gene allows the virus to escape T cell-mediated immune surveillance by down-regulating either the host (MHC class I) or the virus (Ela) molecules (or both) that make up the epitope recognized by the T cell receptor. However, recent experiments suggest that the E3 gene confers a selective advantage on adenovirus by rather more indirect means (unpubl. obs.) Using E3 deletion mutants to infect animals and test for the ability to generate strong Ad5-specific Tc responses, it was found that the virus lacking E3 was much more virulent in mouse strains that were susceptible to both types of immune modulation by E3 (down-regulation of class I MHC and down-regulation of Ela). This was initially very surprising, as one would have anticipated a stronger Tc response leading to faster recovery. However, very recent experiments suggest that in the absence of E3 genes, the Tc response to adenovirus is of such magnitude that it leads to immunopathology and death of the host, an outcome not beneficial to the virus. It is not yet clear as to what mediates death in E3 deletion mutant-infected animals: direct cytotoxicity, as in the classic LCMV model ;^' effects mediated by cytokines; or a combination of both.
62
A. Miillbacher
Thus the E3 gene of adenovirus modulates the host's Tc response, on the one hand limiting elimination of virus-infected cells by Tc, and on the other, preventing Tc-mediated immunopathology. This work, beside giving some insight into persistent infection by adenovirus, has implications for the use of adenoviruses as vaccine vectors. At present the E3 gene, deletions of which do not affect virus growth in vitro, is the preferred location for insertion of additional foreign cDNA. However, deletions in E3 may, as outlined above, change the virulence of the virus in as yet unforeseen ways, and thus render them unsuitable for vaccines.
8.
9.
10.
11.
Acknowledgements I would like to thank Dr G. L. Ada and Dr R. V. Blanden for critically reading the manuscript.
12.
References 13. 1. Miillbacher, A. 1981. Natural tolerance a model for Ir gene effects in the cytotoxic T cell response to H-Y. Transplantation 78: 58-61. 2. Laver, W. G. and Webster, R. G. 1968. Selection of antigenic mutants of influenza viruses. Isolation and peptide mapping of their hemagglutinating proteins. Virology 34: 193-102. 3. Kotwal, G. J., Isaacs, S. N., McKenzie, R., Frank, M. M. and Moss, B. 1990. Inhibition of the complement cascade by the major secretary protein of vaccinia virus. Science 250: 827-830. 4. Gooding, L. R., Elmore, L. W., Tollefson, A. E , Brady, A. and Wold, W. S. M. 1988. A 14 700 MW protein from the E3 region of adenovirus inhibits cytolysis by tumor necrosis factor. Cell 53: 341-346. 5. Gooding, L. R., Aquino, L., Duerksen-Hughes, P. J., Day, D., Horton, T. M., Yei, S. and Wold, W. S. M. 1991. The Elb 19 000 molecularweight protein of group C adenoviruses prevents tumor necrosis factor cytolysis in human cells but not of mouse cells. J. Virol. 65: 3083-3094. 6. Pircher, H., Moskophidis, D., Rohrer, U., Burki, K., Hengartner, and Zinkernagel, R. M. 1990. Viral escape by selection of cytotoxic T cell-resistant virus variants in vivo. Nature 346: 629-633. 7. Hecht, T. T. and Summers, D. F. 1972. Effect
14.
15.
16.
of vesicular stomatitis virus infection on the histocompatibility antigen of L cells. J. Virol. 10: 578-585. Rogers, M. J., Gooding, L. R., Margulies, D. H. and Evans, G. A. Analysis of a defect in the H-2 genes of SV40 transformed C3H fibroblasts that do not express H-2K''. / Immunol. 130: 2418-2422. Jennings, S. R., Rice, P. L., Kloszewski, E. D., Anderson, R. W., Thompson, D. L. and Tevethia, S. S. 1985. Effect of herpes simplex virus types 1 and 2 on surface expression of class I major histocompatibility complex antigens on infected cells./. Virol. 56: 757-766. Paabo, S., Severinsson, L., Andersson, M., Martens, I., Nilsson and Peterson, P. 1989. Adenovirus proteins and MHC expression. Adv. Cancer Res. 52: 151-163. Mellencamp, M. W., O'Brien, P. C. M. and Stevenson, J. R. Pseudorabies virus-induced suppression of major histocompatibility complex class I antigen expression. Virology 65: 3365-3368. Burgert, H-G. and Kvist, S. 1987. The E 3 / 19K protein of adenovirus type 2 binds to the domains of histocompatibility antigens required for CTL recognition. EMBO J. 6: 2019-2016. Rawle, F. C , Tollefson, A. E., Wold, W. S. M. and Gooding, L. R. Mouse anti-adenovirus cytotoxic T lymphocytes. Inhibition of lysis by E3 gpl9K but Not E3 14.7K'. / Immmwt. 143: 2031-2037. Grundy, J. E., McKeating, J. A. and Griffiths, P. D. 1982. Cytomegalovirus strain AD 169 binds P2 microglobulin in vitro after release from cells./ Cen. Virol. 68: 777-784. Mullbacher, A. and King, N. J. C. 1989. Target cell lysis by Natural killer cells is influenced by P2 microglobulin expression. Scand.J. Immunol. 30: 21-29. Blanden, R. V. 1989. Possible factors contributing to persistence of cytomegalovirus infection. In The Immune Response to Viral Infections.
B. A. Askonas, B. Moss, G. Torringiani, and S. Gorini (ed.), pp. 17-36. Plenum Publishing Corporation, New York. 17. Mullbacher, A., Bellett, A. J. D., and Tha Hla, R. 1989. The murine cellular immune response to adenovirus-type 5. Immunol. Cell Biol. 67: 31-39. 18. Gooding, L. R. and Wold, W. S. M. 1990. Molecular mechanisms by which adenoviruses counteract antiviral immune defences. Rev. Immunol. 10: 53-71. 19. Zhang, X. L., Bellett, A. J. D., Tha Hla, R , Braithwaite, A. W. and Mullbacher, A. 1991. Adenovirus E3 gene products interfere with
Adenovirus immune evasion the expression ofthe cytolytic T cell immunodominant Ela antigen. Virology 180: 199-106. 20. Zhang, X., Bellett, A. J. D., MuUbacher, A. and Braithwaite, A. W. 1992. Downregulation of Ela expression by E3 gene products in
63
Group C adenoviruses. Immutwl. Cell Biol. 70: 65-71. 21. Doherty, P. C , and Zinkernagel, R. M. 1974. T cell-mediated immunopathology in viral iniections. Transplant Rev. 19: 89-\20.
Viral escape from immune recognition: Multiple strategies of adenoviruses ARNO MULLBACHER Division of Cell Biology, John Curtin School ofMedical Research, Australian National University, Canberra, Australian Capital Territory, Australia. Summary Human adenovirases can cause persistent infections in man. The strategies of C type adenoviruses (types Ad2 and Ad5) to evade immune recognition are many, but all involve single early genetic regions (E3). Gene(s) within E3 have been shown to allow the adenovirus to avert cytokine-mediated apoptosis. Furthermore, the E3 region controls the cytotoxic T cell epitope (a major histocompatibility complex [MHC] class I molecule with an adenovirus-derived peptide) on the cell surface of infected cells. On the one hand the E3 gene product 19 kDa can bind to nascent class I MHC in the endoplasmic reticulum and thus prevent its transport to the cell surface, and on the other hand the E3 region down-regulates the Ela gene product, the immunodominant cytotoxic T cell determinant.
The immune system of vertebrates has developed a most sophisticated mechanism, as has been illustrated in this issue, to continually adapt to combat parasites. The only constraints placed on the generation of diversity of this powerful defence mechanism is 'self-tolerance, and immune response gene defects are the price to pay'.' By what means do parasites and especially viruses, with their limited 'blank genetic space', evade the host's immune recognition? The great diversity of viruses with respect to their genetic make-up, strategies of infection, replication, and host range suggests that these strategies will differ markedly between virus groups. Table 1 shows a selection of some of the identified strategies used by certain viruses to evade immune recognition. A large number of viruses escape from antibody-mediated immunity through what is called genetic drift.^ Influenza viruses, rota-
Table 1. Virus strategies to escape immune recognition Escape from antibody mediated immunity a) Viral neutralization Genetic drift (orthomyxo-, reo-, picornaviruses) Genetic shift (orthomyxo-, reoviruses) b) Inhibition of complement cascade (Pox viruses) Resistance to cytokines TNF cytolysis (adenoviruses) Lymphotoxin (adenoviruses) Escape from T cell-mediated immunity a) Selection of cytotoxic T cell-resistant variants (LCMV) b) Downregulation of T cell epitopes i) Inhibition of macromolecular synthesis of MHC antigens (herpes-, adeno-, rhabdo-, papoviruses) ii) Competition for P2 microglobulin (CMV)
Correspondence: Arno Mlillbacher, Division of Cell Biology, John Curtin School of Medical Research, Australian National University, PO Box 334, Canberra City, ACT 2601, Australia. Presented at the 4th Frank and Bobbie Fenner Conference in Medical Research on Genetic Variation and Selection in Microbiology and Immunology, held on 6-8 November 1991 at the John Curtin School of Medical Research, Canberra, Australia. Accepted for publication 6 November 1991.
60
A. MUllbacher
viruses and retroviruses (ie. HIV) and rhinoviruses are good examples, the latter showing the most diversity with over 100 serotypes identified. Most of the viruses that undergo extensive genetic drift are RNA or RNA/ DNA viruses with error-prone genetic replication mechanisms. Genetic shifts occur in viruses with segmented genomes as a result of reassortment. Another documented strategy of interfering with humoral immunity is by tampering with the complement cascade as has been shown for pox viruses.^ However, it is not clear what advantage this has given pox viruses per se. Evidence has been obtained that certain viruses have evolved means to escape the antiviral effects of cytokines. One example of this is adenoviruses' ability to interfere with tumour necrosis factor (TNF)- and lymphotoxin-mediated target cell lysis,'''^ most likely by averting apoptosis.^ One of the genes involved in this protection is encoded in the E3 region of the viral genome, a region involved in more than one way in the evasion of immune recognition. The strategies of evading direct T cellmediated immunity are more varied and diverse. First, similar to escape from antibodymediated immunity, it has been shown with virus-specific T cell receptor transgenic mice that selection of cytotoxic T cell (Tc)-resistant variants, with different antigenic specificity, can occur.^ However, the biological significance of this is not quite clear. Major histocompatibility complex (MHC) gene duplication and polymorphism are probably the evolutionary mechanisms to counter such possibilities. Besides evading T cell receptor (TCR) recognition through specific antigenic changes, a number of viruses have been shown to interfere with expression of the epitope, which is a molecular entity comprising an MHC antigen plus viral peptide that is recognized by T cells. Most extensively documented is the down-regulation of expression of MHC class I antigens, which are the signposts for Tc cells.^ This does occur by specific interference with macromolecular synthesis of MHC antigens, at the level of either transcription or translation, and is not solely due to a generalized shutdown of host protein synthesis.
In the case of adenoviruses, cell surface class I MHC antigen expression can be shown to be downregulated in a class I molecule-specific manner,'^'^ by retention of the molecules in the endoplasmic reticulum, thus preventing glycosylation in the golgi apparatus and subsequent transport to the cell surface. Finally, one additional strategy has evolved with human cytomegaloviruses (HCMV), which encode a MHC-like protein that can bind P2 microglobulin (P2m).''* This could make ^2^ less available for MHC class I expression on the cell surface, thus limiting Tc cell recognition of HCMV-infected cells. Reducing availability of P2m may also influence natural killer (NK) cell susceptibility, as it has been shown that ^2^^ 's involved in target cell lysis by virally-induced NK cells.'^•"' It may not be coincidental that all the viruses that have evolved strategies to evade T cell immune recognition can and do produce persistent infections. Here I will present evidence that adenoviruses have evolved an additional mechanism to evade T cell recognition. There are three good reasons to study adenoviruses. First, adenoviruses cause human diseases in their own right. Second, adenoviruses can cause persistent infections in humans and animals, and as adenovirus molecular biology is one of the best understood viral models, their use may offer advantages in learning more of such successful host-parasite relationships. Finally, adenoviruses have recently been identified as candidates for vaccine vectors. Should they ever become viable options to be used for human or veterinary use, a thorough understanding of their biology and behaviour in the host will be essential. We have previously established a model for analysing the murine Tc cell response to C type adenovirus type 5 (Ad5),'^ a summary of which is given in Table 2. The intriguing observation was that adenovirus-immune Tc cells Iysed target cells infected with the E3 deletion mutants more efficiently than when infected with wild-type (w.t.) Ad5 virus. The E3 gene of C type adenovirus encodes a number of open reading frames, and up to six E3-encoded polypeptides have been identified.'^ The longest, the E3 19 kDa protein, has been shown to selectively interfere with the cell surface expression of class I MHC
Adenovirus immune evasion Table 2. Summary of the murine Ad5 immune Tc response* 1. Murine Ad5 immune Tc cells can be generated in vivo and in vitro 2. Tc cells are MHC-class I restricted map to K end in H-2k map to D end in H-2b 3. The immunodominant viral determinant involved in H-2K'' Tc recognition maps to the left 1 3 0 a - a o f E l a 4. Deletions in E3 enhance Tc recognition * From MuUbacher et al. 1989.'^
antigens.'^ However it does not interfere with the expression of H-2K'^ MHC class I antigen, the restriction element for Ad5 immune Tc cells of the murine H-2'' haplotype employed in our studies. We confirmed this lack of inhibition of cell surface expression of H-2K using FACS analysis as well as lysability of w.t. Ad5-f5 E3 deletion mutant-infected target cells by alloreactive K''-specific Tc cells.'^ Two possibilities for E3 gene interference with Tc cell epitope formation were investigated: first, that E3 may interfere with the presentation of peptides rather than with class I MHC cell surface expression; and second, that E3 gene product(s) may interfere with the availability of adenovirus-specific peptide determinants.. The first proposition could be dismissed as we could show that Ad5 w.t.- and E3 deletion mutant-infected target cells were, after superinfection with vaccinia or influenza virus, equally lysable by vaccinia or influenza immune Tc cells, respectively. Thus adenovirus E3 gene product(s) did not interfere with the ability of class I MHC molecules to present vaccinia- or influenza virus-derived peptide determinants to Tc cells. Initial evidence that E3 gene product(s) interfere with the availability of the viral peptide determinant, which together with the class I MHC molecule makes up the Tc epitope, came from studies using different multiplicities of infection of Ad5 w.t. vs E3 deletion mutant viruses for target cells. At high multiplicities (< 100 i.u.) the differential of lysability between Ad5 w.t.- and E3 deletion mutant-infected targets disappeared. Ela
61
is transcribed and translated prior to E3 gene products, and this suggested that high virus input provided plateau levels of Ela products (the viral protein containing the immunodominant peptide determinant) for optimal Tc recognition. Investigating the effects of the E3 gene on transcription of Ela using dot-blot hybridization experiments, no evidence could be obtained for transcriptional control by E3. However, when quantitating Ela protein products in Ad5 w.t.- or E3 deletion mutantinfected target cells by either using Western blot analysis or immunoprecipitation, eight to 10-fold lower amounts were found in Ad5 w.t.-infected cells. Thus E3 gene product(s) modulate the synthesis of the Ela gene products containing the Ad5 immune Tc immunodominant peptide determinants. The mechanism by which this occurs, and which E3 gene product(s) are involved, are dealt with in an accompanying article.'^° Thus the presence of the E3 gene allows the virus to escape T cell-mediated immune surveillance by down-regulating either the host (MHC class I) or the virus (Ela) molecules (or both) that make up the epitope recognized by the T cell receptor. However, recent experiments suggest that the E3 gene confers a selective advantage on adenovirus by rather more indirect means (unpubl. obs.) Using E3 deletion mutants to infect animals and test for the ability to generate strong Ad5-specific Tc responses, it was found that the virus lacking E3 was much more virulent in mouse strains that were susceptible to both types of immune modulation by E3 (down-regulation of class I MHC and down-regulation of Ela). This was initially very surprising, as one would have anticipated a stronger Tc response leading to faster recovery. However, very recent experiments suggest that in the absence of E3 genes, the Tc response to adenovirus is of such magnitude that it leads to immunopathology and death of the host, an outcome not beneficial to the virus. It is not yet clear as to what mediates death in E3 deletion mutant-infected animals: direct cytotoxicity, as in the classic LCMV model ;^' effects mediated by cytokines; or a combination of both.
62
A. Miillbacher
Thus the E3 gene of adenovirus modulates the host's Tc response, on the one hand limiting elimination of virus-infected cells by Tc, and on the other, preventing Tc-mediated immunopathology. This work, beside giving some insight into persistent infection by adenovirus, has implications for the use of adenoviruses as vaccine vectors. At present the E3 gene, deletions of which do not affect virus growth in vitro, is the preferred location for insertion of additional foreign cDNA. However, deletions in E3 may, as outlined above, change the virulence of the virus in as yet unforeseen ways, and thus render them unsuitable for vaccines.
8.
9.
10.
11.
Acknowledgements I would like to thank Dr G. L. Ada and Dr R. V. Blanden for critically reading the manuscript.
12.
References 13. 1. Miillbacher, A. 1981. Natural tolerance a model for Ir gene effects in the cytotoxic T cell response to H-Y. Transplantation 78: 58-61. 2. Laver, W. G. and Webster, R. G. 1968. Selection of antigenic mutants of influenza viruses. Isolation and peptide mapping of their hemagglutinating proteins. Virology 34: 193-102. 3. Kotwal, G. J., Isaacs, S. N., McKenzie, R., Frank, M. M. and Moss, B. 1990. Inhibition of the complement cascade by the major secretary protein of vaccinia virus. Science 250: 827-830. 4. Gooding, L. R., Elmore, L. W., Tollefson, A. E , Brady, A. and Wold, W. S. M. 1988. A 14 700 MW protein from the E3 region of adenovirus inhibits cytolysis by tumor necrosis factor. Cell 53: 341-346. 5. Gooding, L. R., Aquino, L., Duerksen-Hughes, P. J., Day, D., Horton, T. M., Yei, S. and Wold, W. S. M. 1991. The Elb 19 000 molecularweight protein of group C adenoviruses prevents tumor necrosis factor cytolysis in human cells but not of mouse cells. J. Virol. 65: 3083-3094. 6. Pircher, H., Moskophidis, D., Rohrer, U., Burki, K., Hengartner, and Zinkernagel, R. M. 1990. Viral escape by selection of cytotoxic T cell-resistant virus variants in vivo. Nature 346: 629-633. 7. Hecht, T. T. and Summers, D. F. 1972. Effect
14.
15.
16.
of vesicular stomatitis virus infection on the histocompatibility antigen of L cells. J. Virol. 10: 578-585. Rogers, M. J., Gooding, L. R., Margulies, D. H. and Evans, G. A. Analysis of a defect in the H-2 genes of SV40 transformed C3H fibroblasts that do not express H-2K''. / Immunol. 130: 2418-2422. Jennings, S. R., Rice, P. L., Kloszewski, E. D., Anderson, R. W., Thompson, D. L. and Tevethia, S. S. 1985. Effect of herpes simplex virus types 1 and 2 on surface expression of class I major histocompatibility complex antigens on infected cells./. Virol. 56: 757-766. Paabo, S., Severinsson, L., Andersson, M., Martens, I., Nilsson and Peterson, P. 1989. Adenovirus proteins and MHC expression. Adv. Cancer Res. 52: 151-163. Mellencamp, M. W., O'Brien, P. C. M. and Stevenson, J. R. Pseudorabies virus-induced suppression of major histocompatibility complex class I antigen expression. Virology 65: 3365-3368. Burgert, H-G. and Kvist, S. 1987. The E 3 / 19K protein of adenovirus type 2 binds to the domains of histocompatibility antigens required for CTL recognition. EMBO J. 6: 2019-2016. Rawle, F. C , Tollefson, A. E., Wold, W. S. M. and Gooding, L. R. Mouse anti-adenovirus cytotoxic T lymphocytes. Inhibition of lysis by E3 gpl9K but Not E3 14.7K'. / Immmwt. 143: 2031-2037. Grundy, J. E., McKeating, J. A. and Griffiths, P. D. 1982. Cytomegalovirus strain AD 169 binds P2 microglobulin in vitro after release from cells./ Cen. Virol. 68: 777-784. Mullbacher, A. and King, N. J. C. 1989. Target cell lysis by Natural killer cells is influenced by P2 microglobulin expression. Scand.J. Immunol. 30: 21-29. Blanden, R. V. 1989. Possible factors contributing to persistence of cytomegalovirus infection. In The Immune Response to Viral Infections.
B. A. Askonas, B. Moss, G. Torringiani, and S. Gorini (ed.), pp. 17-36. Plenum Publishing Corporation, New York. 17. Mullbacher, A., Bellett, A. J. D., and Tha Hla, R. 1989. The murine cellular immune response to adenovirus-type 5. Immunol. Cell Biol. 67: 31-39. 18. Gooding, L. R. and Wold, W. S. M. 1990. Molecular mechanisms by which adenoviruses counteract antiviral immune defences. Rev. Immunol. 10: 53-71. 19. Zhang, X. L., Bellett, A. J. D., Tha Hla, R , Braithwaite, A. W. and Mullbacher, A. 1991. Adenovirus E3 gene products interfere with
Adenovirus immune evasion the expression ofthe cytolytic T cell immunodominant Ela antigen. Virology 180: 199-106. 20. Zhang, X., Bellett, A. J. D., MuUbacher, A. and Braithwaite, A. W. 1992. Downregulation of Ela expression by E3 gene products in
63
Group C adenoviruses. Immutwl. Cell Biol. 70: 65-71. 21. Doherty, P. C , and Zinkernagel, R. M. 1974. T cell-mediated immunopathology in viral iniections. Transplant Rev. 19: 89-\20.
