Immunological aspects of the prevention of viral diseases*

Uepdinate LepJ oin Articles in the Update series give a concise, authoritative, and up-to-date survey of the present position in the selectedfields, covering many different aspects of the biomedical sciences and public health. Most of s the articles are written by ce acknowledged experts on the red subject. les I Les articles de la rubrique Le pointfournissent un bilan concis et fiable de la situa- tion actuelle dans les do- maines consider0s, couvrant de nombreux aspects des sciences biomedicales et de la ante' publique. La plupart de es articles auront donc ete digOes par les specialistes plus autorises. Bulletin of the World Health Organization, 65 (1): 1-10 (1987) World Health Organization 1987 Immunological aspects of the prevention of viral diseases* WHO WORKSHOP This article summarizes the conclusions of a WHO Workshop which reviewed the role ofimmunological mechanisms in protection against and recoveryfrom virus diseases. Antibodies are importantforpreventing infectionfrom some viruses, e.g., influenza virus. In contrast, cell-mediated immunity is more important in the process of recovery from an already established infection. Furthermore, while anti- bodies are extremely specific in their action, cell-mediated immunity exhibits a greater degree of cross-reactivity. Thisfinding could be useful in designing potential vaccines, particularly attenuated virus vaccines using molecular biological methods. Such an approach could be valuableforpoliomyelitis vaccine, whose RNA sequence has now been determined. The possibility of using recombinant vaccinia viruses as a potential vectorfor other virus antigens is also discussed. Host response to viral infections occurs in two phases. Initially the response is largely non-antigen specific and involves stimulating the production of interferons (IFNs) and activating the natural killer (NK) cells. Virus-infected cells are more susceptible than normal cells to NK-lysis, and activated NK cells release mediators such as IFNs. Once viruses start replicating, the following immune responses are triggered in the second phase: production of humoral antibodies; activation of regulatory T cells (that can amplify or suppress effector functions); and T-cell-mediated immunity, effected by cytotoxic T cells that are able to lyse virus-infected cells or T cells that are able to induce delayed-type hypersensitivity (DTH). Regulatory T cells play a key role in controlling the action of both antibody-forming B cells and effector T cells, which, after antigenic or mitogenic stimulation, release B-cell and T-cell growth and maturation factors, respect- ively, e.g., BSF, BCGFII, and interleukin 2 (IL-2), as well as lymphokines, such as inter- feron gamma, and undefined chemotactic substances. In addition to its antiviral activity, interferon gamma has many functions, e.g., activation of macrophages, including * This article is based on the report of a WHO Workshop on Immunovirology held at the Pasteur Institute, Paris, on 3-6 December 1985. Requests for reprints should be addressed to the Director, Division of Communicable Diseases, World Health Organization, 1211 Geneva 27, Switzerland. A French translation of this article will appear in a later issue of the Bulletin. Ihe participants at the Workshop are listed on pp. 9-10. 4744 1-I UPDATE induction of interleukin 1 (IL-1) release, as well as enhancement of class I major histo- compatibility (MHC) antigens and induction of class II MHC antigen expression at the cell surface. In this way, the whole network of interactions involved in resistance to viral infections is induced. A considerable amount of work is still required to identify the immune response(s) that protect cells against viral infection or clear an intracellular virus once infection has occurred. Individual viruses differ in their requirements for protection. For example, the presence of neutralizing antibodies at the appropriate site of viral entry can frequently prevent infection; however, antibodies are often specific for certain viral components. Antibodies that neutralize viruses in vitro do not necessarily do so in vivo, and for many viruses T cells are required to clear an infection. A good viral vaccine should therefore stimulate a balanced immune response, leaving the host with long-term B-cell memory (for production of antibodies of appropriate speci- ficity) as well as memory for T-helper (Th) cells and/or cytotoxic T cells. THE ROLE(S) OF CELLS INVOLVED IN NATURAL RESISTANCE TO VIRAL INFECTIONS The contributions made by various cells of the immune system to resistance to viral infection, recovery from viral challenge, and to the pathology of viral infections are complex and have only recently been studied. These cells may be classified in two groups, as follows. (1) Non-antigen-specific cytotoxic cells. Among cells of this category are natural killer (NK) lymphocytes, which are active in vitro against tumour cells (most of which bear Fc receptors and the surface antigens Leu 11 and OKM 1) and have increased killing activity against certain virus-infected target cells; killer lymphocytes bearing Fc receptors that are active in antibody-dependent cell-mediated cytotoxicity (most of which are also contained in the Leu 11 and OKM1 + subsets); and macrophages that are cytotoxic against some virus-infected cells, in addition to playing a role in antigen processing and presentation, as well as in secretion of inflammatory activators such as IL-1. All three types of cytotoxic effector cells discussed above secrete antiviral mediators, such as interferons, which have profound effects on the immune system. For example, interferon alfa is produced by virus- infected macrophages and by peripheral blood lymphocytes contained in NK subsets (as well as by DR + cells stimulated by viruses (1)) as well as in response to virus-infected cells, e.g., dengue virus and hepatitis virus A (2). Other mediators, such as tumour necrosis factor, that are secreted by infected cells may also contribute to their cytotoxic action. (2) Antigen-specific cytotoxic cells. MHC-restricted cytotoxic T lymphocytes (CTL) are antigen-specific, usually belong to the OKT8 + subset, and are restricted to class I MHC molecules; however, virus-specific killer T cells of the OKT4 + helper-inducer subset have also been detected in long-term cultures. Interferon gamma and other lymphokines are produced by CTL and Th cells after stimulation by virus-infected cells, e.g., influenza virus (3, 4), which activate macrophages, stimulate growth and maturation of progenitor cells and antibody-forming B cells, and initiate delayed hypersensitivity and immunopatho- logical reactions. Th cells only recognize virus antigen in combination with class II MHC molecules on the surface of antigen-presenting cells. Activated Th cells produce IL-2, which is required for the proliferation of antigen-specific CTL and other T lymphocytes. Interferons Interferon alfa and beta are antiviral cytokines produced by many types of cells 2 IMMUNOVIROLOGY following viral infection and elicit protection against a wide spectrum of viruses. These interferons also stimulate changes in immune function, e.g., augmentation of NK activity and enhancement of expression of class I MHC antigens; however, it is not clear whether they have an immunoregulatory role in vivo. In contrast, interferon gamma-which is structurally unrelated to interferon alfa or beta-is produced by specific immune stimu- lation and is more important as an immunomodulator than as an antiviral agent (5). Inter- feron gamma is a macrophage-activating factor, specifically enhances expression of class I and class II MHC antigens, and may potentiate the activities of non-specific NK cells. Whereas a mixture of interferon alfa and gamma has a synergic antiviral effect in vitro (6), its ability to potentiate antiviral and immunoregulatory effects in vivo has not been demonstrated. Interferon gamma itself may potentiate the destruction of virus-infected cells by lymphotoxin (7). Genetic factors governing the production of and response to interferons also influence resistance to and recovery from virus infections. Immunological memory Although there is some controversy surrounding the concept of T-cell memory, it is well established that the number of virus-specific T cells multiplies several-fold after virus priming (8). The role of antigen persistence in maintaining T-cell memory is not clear. Response of cytotoxic T cells was not accelerated by more than 24 hours when monitored 5 months after primary challenge with lymphocytic choriomeningitis virus. This is in accord with investigations of the cross-response of mice primed with vesicular stomatitis (Indiana) virus that were then challenged with the New Jersey strain of this virus (9). Primary infec- tion of the mice with the Indiana strain did not improve their immune response to the New Jersey strain, or vice versa, as indicated by the level of IgG neutralizing antibodies. In contrast, mice primed with the Indiana strain exhibited a secondary response to common determinants or DNP present on the New Jersey strain. These results suggest that even an excess of primed Th cells are unable to increase the response of B cells to a new, unique determinant. Influenza virus Both humoral and cell-mediated immunity against influenza virus can be distinguished. Humoral immunity is of crucial importance in prophylaxis, and vaccines should ideally stimulate formation of neutralizing antibodies. However, these antibodies are directed against variable regions of the surface glycoproteins, particularly haemagglutinin (HA). Vaccines against influenza virus derived from strains that circulated in a previous year therefore offer only partial or no protection in subsequent years against "drifted" strains. Low response levels to such vaccines may also arise because of genetic factors (e.g., histo- compatibility antigen groups), natural or artificial immunosuppression (cancer, renal diseases, malaria, or chemotherapy), or old age. Continuing assessment of the efficacy of current and future vaccines against influenza is paramount for the development of improved prophylaxis, as is also investigation of immunomodulation by adjuvants, lymphokines, and thymic factors as a means of enhancing immune responsiveness. It is clear, however, that humoral IgG or IgA response predominantly depends upon activation of Th cells and the subsequent production of helper factors; and the mechanism of activation of immune response by viral antigens should therefore be studied. Cell-mediated immunity, as exemplified by CTL, is involved in recovery from infection. There is extensive experimental evidence for such immunity in animals, but only circum- stantial evidence in man. The influenza virus has been investigated in this respect, since MHC-restricted influenza-specific CTL have been detected in humans and their activity found to be higher after vaccination (10). Infection induces CTL responses that are 3 UPDATE usually cross-reactive with influenza virus A subtypes but a minor population of CTL was subtype-specific (11). Thus, any protective effect induced by vaccines to influenza virus A should be cross-protective. It is therefore important to identify the viral components recognized by the A virus cross-reactive CTL and determine how they can be most effectively induced. Such lymphocytes recognize virus antigen in man only in conjunction with a class I MHC antigen (human leukocyte antigen A or B) on the cell surface, but the nature of this interaction has not yet been elucidated. Nevertheless, recent work suggests that one antigen receptor molecule should recognize both antigens. The components of the influenza virus A recognized by CTL can be identified using recombinant DNA techniques. For example, direct transfection of cDNA copies of single influenza virus A genes into mouse lymphocytes indicates that, while HA is recognized by a minor population of subtype-specific CTL, nucleoprotein is an important target antigen for CTL that are cross-reactive to the virus. Similar findings have been made using recom- binant vaccinia virus into which single influenza genes have been inserted, and this technique has also revealed that not only mouse but also human CTL from several, but not all, donors respond to nucleoprotein (12). Stimulation of CTL with purified virus proteins has also been reported. Early studies with HA were equivocal, but recent studies on a genetically engineered HA2-NS1 hybrid (NS1 = non-structural protein 1) suggest that sub- type-specific CTL can be stimulated in this way. Similarly, purified nucleoprotein has been used to prime mice for cross-reactive CTL or to boost the latter in vitro as effectively as infective virus (15). Furthermore, HA2-NS1 or nucleoprotein (NP) can bind to uninfected cells that then serve as targets for HA2-NS1- and NP-specific CTL, respectively (13, 14). Recently it has been reported that two short synthetic nucleoprotein peptides bind to uninfected target cells, which are then recognized by virus-specific CTL (14). Two other peptide epitopes have been similarly identified: one associated with HWDb in mice and the other with human leukocyte antigen B37 in a human donor. The above-mentioned results offer new ways of stimulating cell-mediated immunity with vaccines. Recombinant vaccinia viruses could be used for this purpose, but are unlikely to be acceptable for trials in volunteers. On the other hand, purified internal proteins, e.g., nucleoproteins, elicit protective T-cell responses in mice, although to a lesser extent than intranasal influenza infection, and these might form the basis of new vaccines for use in humans (12). Adoptive transfer of nucleoprotein-specific CTL to infected mice also induces protection against lethal infection with or limits replication of influenza virus in the lungs and trachea (15). Since T-cell recognition of peptides in conjunction with MHC molecules is influenced by genetic differences, a given peptide is unlikely to be of general value for stimulating cell-mediated immunity in a polymorphic population. In mice and humans NK-cell activity is augmented shortly after infection with influenza virus, and this is associated with an increase in the serum levels of interferons (16, 17). However, the role of these cytotoxic effector cells in the mechanism of resistance and recovery from infection remains unclear. Preliminary evidence suggests that high levels of Th cells can be harmful. For example, adoptive transfer of Th clones that are specific for HA enhances replication of influenza virus in the lung and induces lung consolidation. This suggests that activation of Th cells, while largely responsible for removal of virus-infected cells, can also induce immuno- pathology and that vaccines have to ensure a balanced immune response. Respiratory syncytial virus Respiratory syncytial virus (RSV) is the most frequent cause of lower respiratory infections in infants and is especially associated with bronchiolitis. Isolated first in 1956, the virus has not undergone clinically significant antigenic variation (although serological 4 IMMUNOVIROLOGY differences have been detected), and is therefore a prime candidate for control by immuniz- ation. Early attempts to develop an inactivated vaccine were, however, abandoned when a formalin-inactivated preparation appeared to potentiate subsequent natural disease rather than induce a protective response, despite induction of neutralizing antibodies and strong delayed-type hypersensitivity. Also, development of a live vaccine has been hampered by difficulties in balancing reduced virulence and immunogenicity and by problems in attaining complete genetic stability. Whereas systemic paramyxovirus infections, such as mumps and measles, usually afford lifelong immunity, infections caused by parainfluenza viruses or RSV lead to transient and incomplete protection. Although childhood bronchiolitis is probably the most severe manifestation of RSV, repeated infections occur throughout life that are associated with prolonged pulmonary abnormalities in young adults and are frequently fatal in the elderly (18). The role played by antibodies in protecting against RSV and in clearing established infections with this virus has been confirmed by studies involving passive transfer of poly- clonal and monoclonal antibodies to small animals. The mechanism of protection is not yet known, but it is unlikely to require the involvement of complement since mice that were deficient in complement component C5 were protected against infection. No evidence has yet emerged that monoclonal antibodies to nucleoprotein, phosphoprotein, or membrane protein of RSV offer protection against infection with the virus. Intraperitoneal inoculation of mice with 50 yg of monoclonal antibody to the fusion (F) protein at the height of infection clears RSV from the lungs within 6 hours, which illustrates the prophy- lactic and therapeutic potential of monoclonal antibodies against the virus. This result also underlines the value of antibody in protecting against RSV infection, and gives an indication of the important epitopes that should be included in potentially successful vaccines (19). Although virus-neutralizing antibodies with certain specificities appear to offer protection against infection with RSV, circulating antibody titres do not correlate well with resistance to infection with the virus. Studies of the role of T-cell subpopulations in protection, of antigen recognition patterns, and of the requirements for induction of protective T-cell responses are therefore being carried out. In mice, intranasal RSV infections generate cross-reactive CTL memory cells that can be stimulated in vitro to generate H-2-restricted CTL (17). As reported previously for Sendai virus (and unlike influenza virus), RSV that is inactivated by ultraviolet radiation effectively primes mice for CTL memory and can form target cells, presumably because of the strong fusion (F) protein of the virus. Splenic T-memory cells primed by intranasal RSV infection, but not normal T cells, clear RSV from the lungs of irradiated or of athymic infected nude mice, which continue to shed virus in the absence of transferred primed T cells. Cytotoxic T lymphocytes do not distinguish between serologically different RSV. Studies analogous to those in progress for influenza virus -selection of antigen-specific T-cell clones, transfection of target cells by RSV genes, and construction of vaccinia recombinants to further define the role and viral specificity of RSV-antigen-specific CTL or Th in clearing RSV infection with the virus-are under way. The effects of maternal mouse antibodies on CTL priming following infection of mouse neonates are also being examined. This is important in view of the very early susceptibility of human infants to RSV infection and the serious reinfection problems in young children; the results may provide a basis for developing an effective vaccine. Poliovirus-molecular basis of antigenicity Studies of the molecular basis of the immunogenicity and of the neurovirulence of 5 UPDATE human poliovirus 3 have yielded information that is of importance for the control of inactivated poliovirus vaccines and for the control of attenuated poliovirus vaccines. Recombinant-DNA techniques may also facilitate the construction of safer, genetically more stable attenuated vaccines. Three antigenic sites have been mapped on poliovirus 3. These are located in the major capsid protein VP1 at position 89-100 from the N-terminus (site 1); in VP2 at residues 164-172 from the N-terminus (site 2); and in a complex of VP3 and VP1 consisting of residues 58-60 and 70-74 of VP3 and residues at positions 286-290 of VP1 (site 3). Site 1 is the most common target for neutralizing antibodies, while sites 2 and 3 have been found only in hyperimmune animals or animals immunized with antigenically modified virus (either a naturally occurring "drifted" strain or a virus treated with trypsin, which specifi- cally destroys site 1 and reduces the immunogenicity, but not the infectivity, of the virus). The immunodominance of the major antigenic site indicates that the standard test for inactivated poliomyelitis vaccines (an immunopotency test in guinea pigs) probably only monitors the reaction of the animal to site 1 but not to the other sites that may be induced in a fully immunized human subject. The local immune response to intestinal infection with poliovirus may be directed against the subsidiary sites, since site 1 may be destroyed by intestinal proteases. This suggests that the inactivated, injectable Salk vaccines may engender a different spectrum of antibodies to the live attenuated Sabin vaccines and confer a different type of immunity. In contrast to poliovirus 3, infection with poliovirus 1 appears only to induce antibodies to sites 2 and 3. This may also be significant in the epidemiology of the disease and in the standardization of inactivated vaccines against poliovirus 1. Recent isolates of poliovirus from Finland, where cases of poliomyelitis occurred in a population immunized with inac- tivated vaccine, were neutralized incompletely by polyclonal sera and by monoclonal anti- bodies to "classical" strains of poliovirus 3 and possessed 1-3 amino acid substitutions at site 1 on VP1. Synthetic oligopeptides containing the amino acid sequence of the antigenic site of VP1 of poliovirus 3 (Sabin strain) induce neutralizing antibodies in rabbits that were immunized with peptide coupled to carrier protein and administered Freund's adjuvant (21-25). Retroviruses With the advent of human T-cell leukaemia, acquired immunodeficiency syndrome (AIDS), and the recent discovery that both diseases are caused by retroviruses, a worldwide effort has been mounted to develop vaccines to prevent their spread and treat those already afflicted. The genome of human immunodeficiency virus (HIV), the virus responsible for AIDS, differs significantly from that of HTLV-I, the causative agent of adult T-cell leukaemia, and resembles that of the sheep Visna virus (a lentivirus). Lentiviruses cause damage to the nervous system through progressive demyelination of nerve cells, and it is now known that HIV can also infect brain tissue, thus probably accounting for the deterioration of the nervous systems of AIDS sufferers. The long-term prospect of neurological complications developing in later life in those infected with HIV, but showing no overt signs of AIDS, is a cause for concern. HIV exhibits a selective tropism for normal T4 (helper) lymphocytes and is directly cyto- pathic towards these cells. The cytopathological changes are only observed in activated T4 cells, but other mechanisms whereby these cells are inhibited or destroyed by the virus have been proposed. For example, in common with feline leukaemia virus, some HIV-specific proteins may be responsible for the immunosuppression; however, no such proteins associated with HIV have yet been isolated, and the destruction of T4 cells may be caused by active autoimmunity (26). Association of HIV proteins with the T4 antigen of uninfected T4 cells could provoke general destruction of such cells by classical cellular and 6 IMMUNOVIROLOGY antibody-mediated cytotoxicity. The immune deficiency associated with HIV infection may therefore result from the direct cytopathic effect of the virus and/or active autoimmune destruction of T4 lympho- cytes, possibly augmented by repeated stimulation of the immune system by antigen or alloantigens. This mechanism should be borne in mind in attempts to develop a vaccine. Also, viral proteins or synthetic polypeptides that associate with the T4 antigen may have immunopathological side-effects. Critical evaluation of less obvious viral antigens as potential vaccines for protection against infection with HIV is therefore necessary. Cytomegalovirus and human (alpha) herpesvirus 3 Normal adult recipients of live, attenuated Towne-strain cytomegalovirus (CMV) vaccine develop complement-fixing, anticomplementary, immunofluorescent, and neutralizing antibodies, as well as demonstrate CMV-specific lymphocyte proliferation. Double-blind, controlled trials of renal transplant recipients indicated that almost all indi- viduals immunized with CMV vaccine prior to transplantation seroconvert, but that their geometric mean titres are significantly lower than those of normal vaccinees. About one- third of immunized renal transplant candidates fail to undergo CMV-specific lymphocyte proliferation; nevertheless, immunized, high-risk recipients (originally seronegative individuals who receive kidneys from seropositive donors) are significantly protected from severe CMV disease relative to controls administered a placebo (27). In a separate study, renal transplant recipients who developed natural killer cell and specific CTL responses experienced less severe CMV disease than recipients who failed to develop such responses (28). Normal children and adults immunized with the OKA strain of vaccine against (alpha) herpesvirus 3 develop fluorescent antibodies to membrane antigens and neutralizing anti- bodies, while proliferation of lymphocytes specific to the virus and delayed cutaneous hypersensitivity also occur (29). In general, these responses are good indicators of protection; however, a few individuals, including some with natural infection, developed mild varicella after exposure, despite the presence of antibodies and/or specific cellular responses. The place of immunology in vaccine research and development Immunization should produce a balanced immunological response in the host. On the one hand, it should stimulate adequately immune responses (largely dependent on T-cell activation) to establish protective, long-term humoral and cell-mediated immunity. On the other hand, stimulation of immune responses that lead to immunopathology and other undesirable sequelae should be avoided. At present, however, it is not known how to influence this balance for vaccines against viruses. Information about virus structure and antigenicity helps in predicting which antigenic determinants are responsible for inducing neutralizing antibodies. So far, however, vaccines based on viral proteins or synthetic polypeptides bearing these antigenic deter- minants have mostly failed to elicit full protective humoral immunity. This may be due, in part, to lack of T-cell activation, since the epitopes "seen" by B and T cells are not necessarily the same; furthermore, genetic differences have been detected in the process of recognizing single epitopes and peptides. The inclusion in vaccines of viral proteins other than those with epitopes that induce neutralizing antibodies may provide better immune protection. T cells that respond in vivo or in vitro to "non-neutralizing" antigens may both enhance B-cell production of neutralizing antibody and induce cross-reactive responses from CTL of potential value in recovery from viral infections. The efficacy of viral protein vaccines could possibly be increased by using adjuvants, but, because of undesired side- 7 8 UPDATE effects, these have been used only to a limited extent for immunizing humans. Identifi- cation of new, less toxic adjuvants or carriers for isolated pure viral proteins or poly- peptides is therefore important, and the recent synthesis of non-toxic muramyl dipeptide derivatives with adjuvant activity, i.e., they induce synthesis of IL-1, which triggers T-cell activation, is promising. The presence of circulating neutralizing antibodies is insufficient to protect against re- infection by some viruses, e.g., RSV. For other viruses, e.g., hepatitis virus B and HIV, the presence of specific antibodies not only fails to clear the virus, but may also compound the pathology of the disease. Few if any general rules therefore apply to the development of viral vaccines, and, for this purpose, each virus must be treated separately. Cell-mediated immunity is critical for the resolution of many viral infections, but its contribution to protective immunity following vaccination, especially in the long term, remains uncertain. T lymphocytes do not usually recognize the viral antigenic determinants responsible for the induction of neutralizing antibodies. There is a growing body of evidence that viral protein determinants are recognized by T lymphocytes, but it is probable that only a few determinants will prove helpful for inclusion in vaccines. REFERENCES 1. PERUSSIA, B. ET AL. A leukocyte subset bearing HLA-DR antigens is responsible for the in vitro alpha interferon production in response to viruses. Natural immunity and cell growth regulation, 4: 120-137 (1985). 2. KURANE, I. ET AL. Human lymphocyte responses to hepatitis A virus-infected cells. Journal of immunology, 135: 2140-2144 (1985). 3. ENNIS, F. A. & MEAGER, A. Immune interferon produced to high levels by antigenic stimulation of human lymphocytes with influenza virus. Journal ofexperimental medicine, 154: 1279-1289 (1981). 4. MORRIS, A. G. ET AL. Immune interferon release when a cloned cytotoxic T-cell line meets its correct influenza-infected target cell. Nature, 295: 150-152 (1982). 5. DE MAEYER-GUIGNARD, J. & DE MAEYER, E. In: Gressor, I., ed. Interferon, vol. 6. New York, Academic Press, 1985, pp. 69-91. 6. CZARNIECKI, C. W. ET AL. Synergistic antiviral and antiproliferative activities of Escherichia coli-derived human alpha, beta, and gamma interferons. Journal of virology, 49: 490-496 (1984). 7. ADERKA, D. ET AL. Increase of vulnerability to lymphotoxin in cells infected by vesicular stomatitis virus and its further augmentation by interferon. Cellular immunology, 92: 218-225 (1985). 8. ASKONAS, B. A. ET AL. Cytotoxic T-memory cells in virus infection and the specificity of helper T cells. Immunology, 45: 79-84 (1982). 9. ZINKERNAGEL, R. M. ET AL. On the role of viruses in the evolution of immune responses. British medical bulletin, 41: 92-97 (1985). 10. MCMICHAEL, A. J. & ASKONAS, B. A. Influenza-virus-specific cytotoxic T-cells in man, induction and properties of the cytotoxic cell. European journal of immunology, 8: 705-7 10 (1978). 11. TOWNSEND, A. R. M. & MCMICHAEL, A. J. Specification of cytotoxic T lymphocytes stimu- lated with influenza virus studied in mice and humans. Progress in allergy, 36: 10-43 (1985). 12. MCMICHAEL, A. J. ET AL. Recognition of influenza A virus nucleoprotein by human cytotoxic T lymphocytes. Journal of general virology, 67: 719-726 (1986). 13. YAMADA, A. ET AL. Influenza virus subtype-specific cytotoxic T lymphocytes lyse target cells coated with a protein produced in E. coli. Journal of experimental medicine, 162: 1720-1725 (1985). 14. TOWNSEND, A. R. M. ET AL. The epitopes of influenza nucleoprotein recognized by cytotoxic IMMUNOVIROLOGY 9 T lymphocytes can be defined with short synthetic peptides. Cell, 44: 959-968 (1986). 15. TAYLOR, P. H. & ASKONAS, B. A. Influenza nucleoprotein-specific cytotoxic T cell clones are protective in vivo. Immunology, 58: 417-420 (1986). 16. LEUNG, K. N. & ADA, G. L. Induction of natural killer cells during immune influenza virus infection. Immunobiology, 100: 352-366 (1981). 17. ENNIS, F. A. ET AL. Interferon induction and increased natural killer cell activity in influenza infections in man. Lancet, 2: 891-893 (1981). 18. STOTT, E. J. & TAYLOR, G. Respiratory syncytial virus. Archives of virology, 84: 1-52 (1985). 19. STOTT, E. J. A comparison of three vaccines against respiratory syncytial virus in calves. Journal of hygiene, 93: 251-261 (1984). 20. PRINCE, G. A. ET AL. Immunoprophylaxis and immunotherapy of respiratory syncytial virus infection in the cotton rat. Virus research, 3: 193-201 (1985). 21. ALMOND, J. W. ET AL. New poliovirus vaccines: a molecular approach. Vaccine, 2: 177-184 (1984). 22. EVANS, D. M. A. ET AL. Critical role of an eight-amino-acid sequence of VPI in neutralization of poliovirus type 3. Nature, 304: 459-462 (1983). 23. EVANS, D. M. A. ET AL. Increased neurovirulence associated with a single nucleotide change in a noncoding region of Sabin-type 3 polio vaccine genome. Nature, 314: 548-550 (1985). 24. FERGUSON, M. ET AL. Induction by synthetic peptides of broadly reactive, type-specific neutralizing antibody to poliovirus type 3. Virology, 143: 505-515 (1985). 25. MINOR, P. D. ET AL. Location and primary structure of a major antigenic site for poliovirus neutralization. Nature, 301: 674-679 (1983). 26. KLATZMANN, D. & MONTAGNIER, L. Approaches to AIDS therapy. Nature, 319: 10-11 (1986). 27. PLOTKIN, S. A. ET AL. Towne-vaccine-induced prevention of cytomegalovirus disease after renal transplants. Lancet, 1: 528-530 (1984). 28. ROOK, A. H. ET AL. Virus-specific cytotoxic lymphocyte responses are predictive of the outcome of cytomegalovirus infection of renal transplant recipients. Transplantation proceedings, 16: 1466-1469 (1984). 29. ARBETER, A. M. ET AL. Varicella vaccine trials in healthy children. American journal of diseases of children, 138: 434-438 (1984). * * LIST OF PARTICIPANTS A. C. Allison, Syntex Research, Institute of Biological Sciences, Palo Alto, CA, USA B. A. Askonas, MRC National Institute for Medical Research, London, England M. M. Barme, Institut Pasteur, Paris, France F. Barre-Senoussi, Department of Virology, Institut Pasteur, Paris, France M. Brahic, Department of Virology, Institut Pasteur, Paris, France J. C. Chermann, Department of Virology, Institut Pasteur, Paris, France F. A. Ennis, Department of Medicine, University of Massachusetts Medical School, Worcester, MA, USA M. Girard, Department of Virology, Institut Pasteur, Paris, France C. Hannoun, Department of Virology, Institut Pasteur, Paris, France N. Horaud, Department of Virology, Institut Pasteur, Paris, France 0. Jarrett, Department of Veterinary Pathology, University of Glasgow, Glasgow, Scotland W. Jordan, National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA C. J. Lucas, Netherlands Red Cross Blood Transfusion Centre, Amsterdam, Netherlands 10 UPDATE A. Meager, National Institute for Biological Standards and Control, London, England (Rapporteur) A. McMichael, John Radcliffe Hospital, University of Oxford, Oxford, England L. Montagnier, Department of Virology, Institut Pasteur, Paris, France R. Petrov, Institute of Immunology, USSR Ministry of Public Health, Moscow, USSR C. R. Pringle, Department of Biological Sciences, University of Warwick, Coventry, England G. C. Schild, National Institute for Biological Standards and Control, London, England (Rapporteur) S. Starr, The Children's Hospital of Philadelphia, University of Pennsylvania, Phila- delphia, PA, USA E. J. Stott, Agricultural and Food Research Council, Compton, Berkshire, England J. L. Sullivan, Department of Pediatrics, University of Massachusetts Medical School, Worcester, MA, USA J. L. Virelizier, Laboratory of Viral Immunology, Institut Pasteur, Paris, France R. Zinkernagel, Institute for Pathology, University of Zurich, Zurich, Switzerland WHO Secretariat F. Assaad, Division of Communicable Diseases, World Health Organization, Geneva, Switzerland G. Torrigiani, Microbiology and Immunology Support Services, World Health Organiz- ation, Geneva, Switzerland