Immune responses to viral antigens in man and their relevance to vaccine development: Memorandum from a WHO Meeting*

KMemoranda are state- Les Mimorandums ments concerning the exposent les conclu- Memoranda conclusions or recom- sions et recomman- mendations of certain dations de certaines WHO scientific meet- re'unions scientifiquesAIey7nm oran du11 ms / / ings; they are signed de l'OMS; ils sontby the participants in signes par les partici- the meeting. pants d ces reunions. Bulletin ofthe World Health Organization, 61(6): 935-940 (1983) © World Health Organization 1983 Immune responses to viral antigens in man and their relevance to vaccine development: Memorandum from a WHO Meeting* A meeting of an informal working group on immune responses to viral antigens was convened in Amsterdam on 15-18 November 1982 to discuss recent developments in this fieldand toformulateproposalsfor international collaborative studies relevant to the needs of the WHO Programme on Vaccine Development. The meeting concluded that the time was ripe for collaborative studies on a wide range of aspects of the immune responses to viral antigens in man. Areas for collaborative effort were defined. These included (a) the role of virus-specific cytotoxic T lymphocytes in virus clearance following infection, (b) correlation of changes in natural killer cells and virus replication, virus virulence, and induction of cytotoxic T Iymphocyte, (c) induction of interferons and other mediators of immune responses, and (d) role of individual viral antigens in immune responses. This Memorandum summarizes the discussions and recommendations of the group. Improvements in the prevention and control of viral, chlamydial and rickettsial diseases of public health importance remain an important aspect of WHO activities. Previous meetings dealing with these infections have emphasized the need for a better understanding of the mechanisms underlying natur- ally acquired and vaccine-induced immunity. The past few years have seen several important scientific developments in the understanding of mechanisms and parameters of immunity to viral infections in man (1). There is increasing evidence that cell-mediated immunological processes, acting alone or in conjunc- tion with specific antibody, may contribute to protec- tive immunity and potentially to immunopathological processes. In addition, recently acquired knowledge of the molecular biology of viruses (2) and the devel- opment of new biotechnologies, especially the tech- niques for manipulation of defined coding sequences of DNA, their controlled expression in eukaryotic * This Memorandum was drafted by the signatories listed on page 940, on the occasion of a meeting held in Amsterdam on 15-18 November 1982. Requests for reprints should be addressed to the Director, Division of Communicable Diseases, World Health Organ- ization, 1211 Geneva 27, Switzerland. A French translation of this Memorandum will appear in a later issue of the Bulletin. and prokaryotic host cells, and the production of monoclonal antibodies by cell fusion techniques, have provided the stimulus for the present high level of scientific activity in the development of viral vaccines. Although in some countries effective prevention and control of certain viral diseases are provided by currently available vaccines of proven safety and effi- ciency, in many areas of the world viral infections are still responsible for a large proportion of the total incidence of communicable disease. There is thus a clear need for improvements in the prevention of viral diseases by vaccines and many laboratories are now becoming involved in studies in vaccine development and in the mechanisms of immunity to viral diseases. The evaluation of immune responses by all available methods is especially important in the light of recent expectations for the development of novel vaccines by new biotechnological approaches, especially re- combinant DNA technology and synthetic peptide chemistry. Detailed information on the immuno- logical properties of these products will be required to determine their efficacy and safety. In addition, relatively few direct comparisons have been made 4358 -935- MEMORANDUM between parameters of immunogenesis associated with live or inactivated vaccines, or between inacti- vated vaccines composed of whole virus particles or subviral structures. It is clear, however, that for cer- tain viruses, e.g., measles inactivated virus, vaccines may not be effective unless they include all the essen- tial immunogens of the virus. The results of the precise measurements of im- munological responses to viral vaccines or to natural infection in man may differ between laboratories as a result of variations in techniques and the reagents employed. Consequently, it is considered that pro- gress in this field would be greatly enhanced by international scientific collaboration. In particular, attempts should be made towards the development of research reagents and the standardization of methods used in studies of viral immunity. WHO has an impor- tant role in encouraging, coordinating and supporting the work in this field as an integral aspect of its programmes on the prevention and control of viral diseases and on vaccine development. It is recognized that information on immune mechanisms based entirely on animal models may, in some circum- stances, be of limited value and that clinical studies in man play an essential role. ASSAY PROCEDURES FOR ANTIBODY TO VIRAL ANTIGENS In the past, laboratory studies on immunity to viral infection, like other diagnostic and epidemiological investigations, have relied heavily on the detection and assay of specific antiviral antibodies. For a full understanding of the role of antibodies in immunity, careful analysis is required of the contribution to the immune response of each of the separate antigenic components of the virus. In addition, the specificity and sensitivity of the assay systems available for detecting antibodies should be known in respect of the individual antigenic components of the viruses and the class of immunoglobulin detected. Tests for virus neutralizing antibodies have been widely employed. Such assays are likely to remain of great importance because of their direct functional relevance to mechanisms of immunity. There are clearly limitations in the use of such assays for viruses that are not cultivable (or only poorly) or that are highly pathogenic for man. For these agents (e.g., hepatitis virus B), much reliance has been placed on the use of direct antibody-antigen binding assays. Other biologically based methods which detect anti- body reacting with viral surface antigens, such as haemagglutination-inhibition (HI) tests widely em- ployed for haemagglutinating viruses (e.g., influenza virus), are also of great value. However, it should be recognized that antibodies detected in this way will not always correlate with virus neutralizing activity or with immunity. Indeed, recent studies employing monoclonal antibodies to viral antigens revealed that antibodies reacting with a given viral antigen may be heterogeneous in their biological activities -the bio- logical properties of the antibody being determined by the precise region (antigenic site) on the antigen molecule with which the antibody reacts. Further- more, antibodies directed against the same sites, but which are of different immunoglobulin classes or sub- classes, may have different biological properties. Among the newer biological tests, single radial haemolysis (SRH) in gels has been of particular value for assays of antibody reacting with viral surface antigens, particularly for certain enveloped viruses, and is worthy of further development. This method has been used widely for assays of antibody to antigens of several viruses of medical importance, including the influenza, parainfluenza, rubella and measles viruses. In the case of influenza viruses, SRH may detect haemagglutinin-specific antibodies which are not detectable by conventional HI tests but which nevertheless are relevant to immunity. Recently a number of non-biological techniques dependent only on direct binding of antibodies to viral antigens have been developed. Included among these techniques are enzyme-linked immunoassays (ELISA) and radioimmunoassay (RIA) procedures and antigen-blocking methods employing single radial diffusion (SRD). These methods have the potential advantage of being highly sensitive. How- ever, it is important to recognize that such non-bio- logical tests may frequently detect antibodies that have no direct functional role in immunity. Where these tests are used, relevant information should be obtained on the specificity of each antibody assay in terms of the individual antigenic components of the virus. Furthermore, it is often possible to devise test systems specific to the antibody to a given antigenic component of a virus, or specific to the antibodies of a particular immunoglobulin class. The development and standardization of these methods and the role in immunity provided by the antibody which they detect is an important field for international collaborative work. ANTIBODY-DEPENDENT CELL-MEDIATED CYTOTOXICITY ASSAYS Another method of detecting potentially protective antibody is the use of antibody-dependent, cell- mediated cytotoxicity (ADCC) assays. The potential importance of this aspect of cellular immunity has been recently demonstrated in studies with Sindbis 936 IMMUNE RESPONSES TO VIRAL ANTIGENS virus. Passive administration to mice of certain monoclonal antibodies to Sindbis virus, with no detectable in vitro neutralizing activity, clearly prevented lethal Sindbis infection in these animals. The possibility exists therefore that ADCC mech- anisms are involved in such protection and these may act through the recognition of viral antigens on the membranes of infected cells by certain subclasses of IgG antibodies. These antibodies may escape detec- tion by other assays and research in this area should be encouraged. In particular, there is a need for pre- cise information on the nature of the virus-coded antigens involved in ADCC on the target cell mem- brane. This is an area where collaborative studies between virologists and immunologists are likely to be highly productive. Studies on ADCC responses should be done in parallel with assays of cytotoxic T lymphocytes, described below, with particular reference to the nature of the target cells involved in each system. CYTOTOXIC T LYMPHOCYTE ASSAYS IN MAN For certain groups of viruses, infection or immu- nization is known to stimulate populations of lymphocytes which are cytotoxic for virus-infected target cells and the process, unlike the ADCC response described above, is independent of anti- body. The cytotoxic T lymphocyte (CTL) responses of man for specific viruses are reviewed below. Influenza virus Cytotoxic T lymphocytes are generated from peripheral blood lymphocytes of man by in vitro incubation with virus antigen for 4-7 days. The CTL activity is specific for influenza virus antigens and histocompatibility antigens (HLA) on infected cells. Not all human donors exhibit a measurable CTL response and, indeed, evidence has been obtained in a number of laboratories that the proportion of responders has declined progressively over the past five years, to such an extent that 70% of donors gave responses in 1977, while only 25%o responded in 1982 in assays employing target cells infected with influenza virus A. This may be a reflection of the low level of epidemic activity of influenza virus A in the past few years. A major problem with this assay appears to be the preparation of the target cells. Despite many attempts, it is still rarely possible to achieve more than 50% lysis of the target cells. Both lymphocytes and lymphoblasts stimulated by phytohaemag- glutinin have been infected and used as targets but there seems to be no qualitative difference between them. Batch-to-batch variation between preparations of the same viruses used for infecting target cells has been observed by all investigators. There have been only occasional attempts to exchange materials between laboratories; however, the assays can be per- formed on cells that have been stored frozen. Collab- orative experiments, involving the exchange of target and effector cells, therefore appear to be feasible. Despite the technical difficulties, there are indi- cations that this area of research could be profitable. Recent studies in volunteers have shown an associ- ation between high levels of influenza-specific CTL activity and rapid clearing of virus after deliberate intranasal challenge with influenza virus. CTL activity cannot distinguish between different sub- types of influenza virus A, thus they might contribute some heterotypic immunity. Inoculation with inacti- vated or live influenza A (HINl) vaccines has been found to boost CTL memory, but for a duration of only a few months. The declining T cell response in uninoculated individuals, noted above, suggests that, after natural infection also, the CTL memory is limited in duration. The standardization of CTL assay methods should be attempted and improvements in target cell prep- arations are needed. Availability ofcommon reagents and techniques might encourage other laborat6ries to perform this assay and contribute to investigations of this aspect of immunity. International coordination and support are required for these collaborative experiments. Epstein-Barr virus (EBV) Cytotoxic T lymphocytes can be prepared from seropositive donors by the incubation of lymphocytes with autologous EBV-transformed lymphoblastoid cell lines. The main technical problem is the parallel induction of natural killer (NK) cell activity, but this can be avoided by extending the culture period and by cell separation procedures. Levels of CTL activity specific for EBV seem to remain constant in indi- viduals who have been infected with the virus. Cell lines, cultivated in the presence of T-cell growth factor, possess CTL activity specific for EBV, which remains stable for 3-6 months. Cells with EBV- specific CTL activity have also been cloned. Studies in experimental animals and in man have demonstrated that cytotoxic T lymphocytes can recognize foreign antigens, such as viruses or haptens, in conjunction with cell antigens of the major histocompatibility complex. Among the anti- gens recognized by human CTL are H-Y, dinitro- phenyl, influenza virus, human (alpha) herpesvirus, and cytomegalovirus in conjunction with self-anti- gens that are closely associated with the serologically defined HLA-A and -B histocompatibility antigens. In contrast, a significant portion of the EBV-specific 937 MEMORANDUM CTL activity, induced during natural infectious mononucleosis, appears to depend upon recognition of non-polymorphic antigens on HLA-A, -B and -C molecules in conjunction with EBV antigens. ipheral blood lymphocytes to mumps antigens. A study of CTL responses after mumps immunization or natural infection has not yet been done and could be very informative. Measles virus It has been postulated that cellular immune responses may play a role in recovery from measles virus infection and the maintenance of an immune state. In limited studies, peripheral blood lymphocyte preparations obtained from children during the acute phase of measles appear to contain cytotoxic T lym- phocytes which produce HLA-restricted lysis of measles-infected target cells. Although the biological role of measles virus- specific CTL activity remains unknown, it is likely that CTL could eliminate measles-infected cells in vivo, provided that such virus-infected cells express the relevant viral antigens on their surface. Abnor- malities in the generation or regulation of measles- immune CTL may contribute to the occurrence of subacute sclerosing panencephalitis and perhaps other diseases in which persistent measles infection may play a role. Approximately 12% of normal adults tested pro- duce significant CTL responses upon in vitro exposure of their lymphocytes to measles antigen. The cytotoxic response is measles virus-specific both in terms of stimulation and at the effector level. Studies of the specificity of CTL have indicated that the virus-specificity activity is self-specific. An analysis of the cellular requirements for the pro- duction of measles virus-specific CTL activity demonstrated that T cells and macrophages were sufficient for the generation of killer cells. Most of the cytotoxic effector activity was mediated by T cells of the T3 +, T4 and T8 + subpopulations. Mumps virus Cytotoxic T lymphocytes can aid in recovery from infection or they may contribute to immunopatho- logical aspects of an infection. It has been suggested that sensitized T cells, CTL, or cells mediating delayed type hypersensitivity (DTH), might play a role in certain virus infections such as lymphocytic choriomeningitis in mice, or mumps meningitis in man. Recently, CTL activity was demonstrated with lymphocytes isolated from the cerebrospinal fluid and blood of patients with mumps meningitis. The cytotoxic T lymphocytes were virus-specific and HLA-restricted. Mumps-specific CTL activity could be detected in 8 out of 10 healthy donors, who were immune to mumps, by the in vitro exposure of per- NATURAL KILLER CELL RESPONSES Natural killer (NK) cells are a class of cytotoxic leukocytes present in man and other mammals which non-specifically lyse various cell types, including virus-infected target cells. Their cytotoxic activity is markedly augmented by interferon. At present there is no direct evidence that these cells are essential for immunity or for recovery from viral infections. However, the information so far obtained from studies in infected experimental animals and in man (summarized below) suggests that this is a potentially important field for research, and that further studies should be performed on the NK-cell responses to viral infections and after immunization of humans. In experimental mice, susceptibility to alpha- herpesviral disease (herpes simplex infection) has been correlated with the level of NK-cell activity. Mice infected with influenza virus have increased levels ofNK activity early during the infection, which decrease by the time the CTL response is detected, and before antibody responses are seen. Humans infected with influenza, cytomegalovirus or dengue viruses show augmented NK-cell activities but the time of appearance of increased activity appears to vary. Following the administration of influenza virus to adult human volunteers, the NK activity and serum interferon levels increased by the third day after infection and declined by day 10. Infection with an experimental, attenuated dengue vaccine has been shown to stimulate NK activity by the tenth day after inoculation when viraemia and elevated serum inteferon levels were also seen in some volunteers. In other studies performed on recipients of bone marrow transplants, cytomegalovirus in- fection has been associated with an increase in NK activity, as well as in CTL activity in patients who subsequently recovered from infection. Increases in NK activity may at least in part be due to the stimulation of interferon production as a result of virus infection. There is evidence, however, to sug- gest that certain natural killer cells may specifically, or preferentially, kill certain virus-infected target cells. In addition, the degree of killing of commonly used human target cell lines, e.g., k562, appears to vary in different laboratories. Collaborative studies in this field will be of value in confirming the obser- vations and determining the basis for variations in the findings from different laboratories. 938 IMMUNE RESPONSES TO VIRAL ANTIGENS 939 ASSAYS OF SOLUBLE MEDIATORS INDUCED BY VIRUSES AND VIRAL VACCINES AS AN INDEX OF IMMUNITY AND IMMUNOPATHOLOGY There is now clear evidence for the existence of soluble immunoregulatory factors by which cells of the immune system may cooperate in the generation of immune responses of the host. Such responses include antibody production (helper and suppressor factors), activation of macrophages (macrophage- activating factors), growth factors for T cell prolifer- ation and generation of cytotoxic T cells (inter- leukins), and interferons (IFN).a The latter substances are able to exert a direct antiviral effect and possess immunoregulatory functions. Inter- ferons can activate all known types of cytotoxic cells, i.e., NK cells, ADCC cells, macrophages, and spe- cific T lymphocytes. On the other hand, mediators secreted by macrophages (monokines) may partici- pate in immunopathological phenomena associated with viral infections. Fever may be triggered by inter- leukin 1 (endogenous pyrogen). Prostaglandins and superoxide anions (02 -, H202) participate in immunosuppression induced by viruses. A role for plasminogen activator has been proposed in haemor- rhagic shock syndrome. The secretion of such medi- ators may be important for the efficacy and safety of vaccines and should be assessed when new vaccines are developed. An ideal vaccine should induce memory for both humoral and cell-mediated effectors of antiviral immunity. The former aspect may be assessed by studies of antibody production. The latter aspect is more difficult to define. Production of y-IFN by specifically stimulated lymphocytes may be a con- venient way to demonstrate T cell sensitization following infection or immunization. Only sensitized subjects produce this lymphokine when their leuko- cytes are stimulated in vitro with the relevant viral antigens. It is known that interferon production occurs at the final step of the cascade of interleukin secretion, at a stage where specific cytotoxic cells are being generated. Thus, interferon production may indicate that specific clones of T cells have been generated as a result of immunization. The results of IFN assays may provide more specific information on T cell sensitization than was available from assay systems previously employed, such as lymphocyte- proliferation responses. A programme of collaborative work on the role of mediators of immune responses using T lymphocytes stimulated with influenza virus, measles virus, and Epstein-Barr virus would be of considerable scientific value to the understanding of mechanisms of im- a Interferons (IFN) exist as three classes: ci-IFN, of which there are at least 14 potential subtypes, ,B-IFN and -y-IFN. munity to viruses. The species of IFN produced by such cultures should be distinguished by specific antisera. An additional area of research related to viral immune responses which should be investigated is the production of monokines by viruses or viral immune complexes. The following assays may be used to assess the production of monokines by viral antigens or vaccines: (i) Interleukin 1 may be titrated by measuring its ability to amplify the proliferation of thymocytes stimulated with suboptimal doses of phytohaemag- glutinin. (ii) Plasminogen activator may be titrated by its ability to lyse radioactive fibrin coated on plastic plates. (iii) Superoxide anions are best detected by chemi- luminescence techniques using luminol. RECOMMENDATIONS 1. There is an important scientific need for collab- orative studies on a wide range of aspects of the immune responses to viral antigens in man. Such activity would be highly relevant to the WHO pro- grammes on the prevention and control of virus dis- ease and on vaccine development. International co- ordination, as well as encouragement and support, is needed for these studies. 2. The studies should include analysis of all aspects of the host-response mechanism, e.g., humoral and cellular immune responses, together with the role of mediators (e.g., interferons). They should also include investigations of the role of the various viral components as antigens contributing to immunity or immunopathology. 3. The standardization of methods used in studies of viral immunity should be attempted and this aspect should be reflected in the design of appropriate collaborative investigations. 4. There are several definitive areas of collab- orative work, as outlined below. A. Virus-specific cytotoxic T lymphocytes (i) Further collaborative studies are needed to con- firm that CTL memory is associated with virus clearance following natural influenza virus infection. The response to different types and strains of vaccine needs further exploration. The response to natural infection, including primary influenza infections in children, needs to be measured. Frozen effector cells and standard viruses to be used for stimulation of the effectors should be exchanged between laboratories. Standardization and, where possible, simplification of the procedure 940 MEMORANDUM for measuring CTL activity should be attempted. HLA typing of lymphocytes should be included in the studies. The collaborative studies should include the use of standard stocks of virus in the preparation of target cells. Since there is evidence that influenza viruses prepared in eggs and in cell cultures may possess different characteristics, viruses from both sources should be included. (ii) Exchange of frozen EBV-specific CTL should be undertaken in future collaborative studies. The studies should include an assessment of the role of CTL in the EBV-associated malignancies such as nasopharyngeal carcinoma and Burkitt's lymphoma. As with influenza-specific CTL assays, HLA typing would form an integral part of such studies. (iii) Further studies of CTL responses to measles virus would provide a basis for an improved understanding of the mechanism of immunity. Specific areas of interest include studies of subacute sclerosing panencephalitis, atypical measles infec- tions, measles encephalitis, and virus persistence. (iv) Studies of CTL responses after mumps immu- nization and natural infection should be performed. B. Natural killer cells (i) Collaborative attempts should be made to stan- dardize the methods for measuring the levels of natural killer cells during viral infections or after immunization. Methods of standardizing the assays should be assessed, e.g., the use of frozen versus fresh donor cells as effectors. Various target cells, in- cluding k562 or EBV-transformed B cell lines, should be compared in collaborative studies. (ii) Studies should be performed in collaborating laboratories to assess the changes in natural killer cell activity observed before and after immunization with viral vaccines. These studies should include placebos and should be carried out using coded labelling. The results of these assays will be correlated with evidence of virus replication, virus virulence, and interferon and CTL induction from clinical studies. C. Interferons and other mediators of immune responses Comparisons should be made of the levels of natural killer cell activity with the level of CTL activity and interferon production during the cyto- toxicity experiments. Viruses stimulate the pro- duction of interferons which can augment NK- and CTL-mediated lysis, and virus-stimulated CTL may themselves secrete interferon. Frozen effector cells (NK and CTL) and target cells should be exchanged between collaborating labora- tories for studies on virus-induced CTL and NK activities (see A and B above). The supernatant fluids containing interferon and other potential mediators should be exchanged for collaborative study. D. Role ofADCC in immunity Collaborative studies should be performed on the potential role of ADCC in immune responses and the immunopathological reactions to viral antigens. A suitable in vitro system for such studies would be target cells infected with suitable viruses, e.g., human (alpha) herpesvirus, yellow fever virus. E. Role of individual viral antigens in immune responses (i) The precise antigenic determinants involved in humoral antibody responses should be compared with those involved in CTL responses. A valuable model system for these studies is influenza. (ii) The levels of CTL activity should be correlated with the serological immune status of the donors in respect of the structural and non-structural antigens of influenza viruses. * * * F. Assaad, Division of Communicable Diseases, World Health Organization, Geneva, Switzerland T. Bektimirov, Virus Diseases, World Health Organ- ization, Geneva, Switzerland F. Ennis, University of Massachusetts Medical School, Worcester, MA, USA K. Lucas, Central Laboratory of the Netherlands, Red Cross Blood Transfusion Centre, Plesman- laan, Amsterdam, Netherlands A. McMichael, MRC Immunology Unit, John Rad- cliffe Hospital, Oxford, England G. C. Schild, National Institute for Biological Standards and Control, London, England G. Torrigiani, Immunology, World Health Organ- ization, Geneva, Switzerland J. L. Virelizier, Unite d'Immuno-Hematologie, Hopital des Enfants Malades, Paris, France REFERENCES 1. WHO Technical Report Series, No. 693, 1983 (Viral vaccines and antiviral drugs: report of a WHO Scien- tific Group). 2. The role of genetic and molecular characterization of viruses in relation to influenza surveillance and epi- demiology: a WHO Memorandum. Bulletin of the World Health Organization, 59: 875-879 (1981).