The pathogenetic aspects of measles virus infection: memorandum from a WHO meeting.

Memoranda/Memorandums The pathogenetic aspects of measles virus infection: Memorandum from a WHO meeting* Over the last three years considerable progress has been made in various areas related to measles virus and infection. A meeting to discuss the currently available data on the molecular biology of mea- sles virus, measles immunology, immunopathology, as well as animal models for measles infection, and to identify studies that need to be carried out towards developing new vaccines was organized jointly by WHO and the U.S. National Institutes of Health and held in Montreux, Switzerland, on 20-21 April 1993. This Memorandum summarizes the discussions and recommendations made by the participants. Introduction The high impact of measles on the lives of children, particularly in developing countries, highlights the urgent need to reduce the disease's incidence and * This Memorandum is based on the report of a WHO/NIH Meeting on the Pathogenetic Aspects of Measles Virus Infection, held in Montreux on 20-21 April 1993. The participants at the Meeting were: Dr M.A. Billeter, Zurich, Switzerland; Dr W. Belli- ni, Atlanta, GA, USA; Dr R.S. van Binnendijk, Bilthoven, Nether- lands; Dr P. Borrow, La Jolla, CA, USA; Dr F. Brown, Greenport, NY, USA; Dr F. Cutts, London, England; Dr G. Fennelly, Bronx, NY, USA; Dr B.G. Gellin, Baltimore, MD, USA; Dr D. Griffin, Bal- timore, MD, USA; Dr S.L. Katz, Durham, NC, USA; Dr D. Kola- kofsky, Geneva, Switzerland; Dr H.W. Kreth, Wurzburg, Germa- ny; Dr U.G. Liebert, Wurzburg, Germany; Dr R. Leopardi, Turku, Finland; Dr K. Mcintosh, Boston, MA, USA; Dr V. ter Meulen, Wurzburg, Germany; Dr P. Minor, Potters Bar, Herts, England; Dr S.A. Moyer, Gainesville, FL, USA; Dr B.R. Murphy, Bethesda, MD, USA; Dr D. Naniche, Lyon, France; Dr E. Norrby, Stock- holm, Sweden; Dr P.L. Ogra, Galveston, TX, USA; Dr A.D.M.E. Osterhaus, Bilthoven, Netherlands; Dr E. Paoletti, Troy, NY, USA; Dr B.K. Rima, Belfast, Northern Ireland; Dr S. Schneider- Schaulies, Wurzburg, Germany; Dr J.R. Stephenson, Porton Down, Salisbury, England; Dr T.F. Wild, Lyon, France; Obser- vers: Dr J. Crowley, Pearl River, NY, USA; Dr R. Gluck, Berne, Switzerland; Dr R.W.J. van der Heyden, Bilthoven, Netherlands; Dr H. Homann, Martinsried, Germany; Dr 0. Leroy, Marnes-la- Coquette, France; Dr S. Makino, Tokyo, Japan; Dr T. Nakaya- ma, Tokyo, Japan; Dr J.M. Meegan, Bethesda, MD, USA; Dr C. Rabourdin-Combe, Lyon, France; Dr P. Rota, Atlanta, GA, USA; Dr L. Roux, Geneva, Switzerland; WHO Secretariat: Dr C.J. Clements, Dr P.-H. Lambert, Dr Y. Pervikov (Secretary), Dr R. Scott. A r6sume of this article in French appears on pages 204-206. Requests for reprints should be sent to Microbiology and Immunology Support Services, World Health Organization, 1211 Geneva 27, Switzerland. Reprint No. 5468 mortality rates. The major components of the strate- gy to eradicate measles include attaining the maxi- mum level of control using current vaccines, expand- ing research in virology and immunology to fill the gaps in critical knowledge, and the development of improved vaccines. In 1989, following the recom- mendation of the WHO Expanded Programme on Immunization (EPI), measles was included on the list of priorities of the WHO/UNDP Programme for Vaccine Development (PVD). In 1990, PVD convened a meeting on the devel- opment of new vaccines against measles, and the recommendations made at this meeting were used in establishing the PVD strategic plan of activities. Over the last three years considerable progress has been made in some areas related to measles virus and infection. In view of this, PVD and the U.S. National Institutes of Health (NIH) held a meeting on 20-21 April in Montreux, Switzerland, to discuss the currently available information on the molecular biology of measles virus, measles immunology, and immunopathology, as well as animal models for measles infection, and to identify the studies that should be conducted towards developing new mea- sles vaccines. Molecular genetics of measles virus Studies of the basic molecular mechanisms of measles virus reproduction and its genetics are cur- rently ongoing and will be important for the long- term goals of understanding virus pathogenesis and the mechanism of attenuation of vaccine viruses. Measles virus contains a nucleocapsid with the Bulletin of the World Health Organization, 1994, 72 (2): 199-206 © World Health Organization 1994 199 Memorandum single-stranded, negative-strand-sense RNA genome (ca. 16kbase) encapsidated in the major nucleocapsid protein N. The virion also contains an RNA-depend- ent RNA polymerase, consisting of the viral L and P protein subunits, which catalyses the transcription and replication of the nucleocapsid template. The envelope proteins H and F, responsible for adsorp- tion to cells and fusion, respectively, and M, consti- tute the remaining structural proteins of the virion. The synthesis of both measles virus mRNAs and free leader RNA has been detected in vitro, suggest- ing that the mechanism of measles transcription is analogous to that of other negative-strand RNA viruses. Furthermore, editing of measles virus P mRNA occurs during in vitro transcription with purified virus, suggesting that this activity is virus encoded. A significant advancement has been to establish a eukaryotic expression system for the measles virus L, P, and N genes, which encode the proteins requir- ed for viral transcription and replication. These cDNAs cloned in plasmids downstream of the T7 promoter are transfected into cells and expression is driven by infection with a vaccinia virus recombi- nant encoding T7 RNA polymerase. Full-length viral proteins are synthesized, and two protein complexes (N-P and P-L) are formed in the transfected cells, as assayed by co-immunoprecipitation with antibody specific for one protein. Studies are ongoing to understand the function and effect of mutations in the viral H and F proteins and to synthesize a genomic cDNA. With the vacci- nia-T7 expression system, the haemagglutinin (H) and fusion (F) protein coding regions cloned from persis- tent measles infection, cases of subacute sclerosing panencephalitis (SSPE), and of measles inclusion body encephalitis (MIBE) were analysed. The SSPE- derived H proteins were transported poorly to the cell surface, as reflected also by altered glycosylation and oligomerization patterns; nevertheless, the residual amount of H reaching the surface was sufficient for fusion helper function. The SSPE-derived F glycopro- teins appeared normal in terms of their fusion proper- ties, although the carboxyl, terminal cytoplasmic domains of these F variants were truncated. Interest- ingly, only certain combinations of H and F variants were competent to induce formation of syncytia, indi- cating that an intimate association of the two proteins is required to trigger the fusion process. The application of reverse genetics to nonseg- mented, negative-strand RNA viruses has not so far proved possible. Systems are being developed to establish the rescue of infectious measles virus from cloned DNA bearing the entire viral genomic sequence, into which either artificial alterations or naturally occurring mutations could be introduced. One approach is to use conventional plasmids repli- cating in the nucleus and expressing MV proteins and genomic RNA under the strong constitutive cytomegalovirus (CMV) promoter. The second type of vector would yield either the genomic or antige- nomic virus RNA with authentic 5'- and 3-ends mediated by T7 polymerase initiation and a hepatitis delta-virus ribozyme and T7 terminator to give cleavage, respectively. The in vitro synthesis of some (+)-sense genomic RNA in the latter system has been reported. Important for understanding measles virus patho- genesis has been the determination of sequences of genes of various wild-type and vaccine viruses. Mea- sles virus shows an evolutionary pattern in which a number of specific lineages have been identified. The classification of different virus strains into lin- eage groups is independent of the genes used to con- struct the database, hence no recombination appears to take place. The most variable coding region of the genome comprises the sequence that encodes the final 105 amino acids of the C terminus of the N pro- tein. This sequence has now been determined for over 35 strains, and these have been grouped into six lineage groups. One group contains the Edmonston vaccine strain and all the derived vaccines together with two other lineages circulating earlier. Three other lineages appear to be currently circulating in the world. The significance of these variations on immunology and pathogenesis remains to be deter- mined, but is important for vaccine development. Genetic variation of several types has been described for mutant viruses isolated from SSPE patients, and the functional significance of individual changes in particular proteins is the subject of cur- rent investigations. Similarly, the comparison of the entire Edmonston wild-type sequence with that of the vaccine virus is under way. Again, the functional significance of identified changes as important for attenuation will be assayed by expression of individ- ual protein and, when it becomes available, by analy- sis in a genomic cDNA. To determine the degree of genetic relatedness between a number of vaccine strains, the sequences of the H, F and N genes were determined for four Edmonston-derived and four non-Edmonston-derived measles vaccines isolated from 1954 to 1970. In gen- eral, the sequences from all of the vaccine strains were very similar but not identical, with two of the non-Edmonston-derived strains being slightly more divergent. Although some vaccine-specific changes were evident, the sequences of the vaccines were also similar to those of several wild-type viruses that were isolated before 1960, including a low-passage seed of the Edmonston strain. These data suggest that measles virus is remarkably stable in tissue cul- ture and that the genetic diversity in the more recent- WHO Bulletin OMS. Vol 72 1994200 Pathogenetic aspects of measles virus Infection ly isolated strains of measles virus was not present in viruses that circulated during the 1950s and 1960s. Measles virus-host cell interactions Measles virus infects many human and primate cells in vitro, but rarely cells derived from other species. The primary targets of virus replication in vivo are monocyte/macrophages, endothelial cells, and epi- thelial cells. Occasionally, neurons and other cells of the central nervous system are infected, and this may lead to persistent infection in the absence of an effective immune response (MIBE) and in a rare late-onset neurological disease occurring primarily in immunologically normal children infected at a young age (SSPE). In vitro, measles virus replicates lytical- ly in many cell types, but persistent infection can often be established. It is therefore important to iden- tify what determines the types of cells infected and the outcome of infection. Several advances have been made in understanding the interaction of measles virus with different types of host cell using in vitro analyses. A cellular receptor for measles virus, which is present on many types of cells, has been identified using a monoclonal antibody that inhibits the binding of virus to human cells. The antibody recognizes a glycoprotein of 57-67 kDa that has been identified as the human membrane cofactor protein (CD46), a member of the regulators of the complement-activa- tion gene cluster. CD 46 has many isoforms. Expres- sion of one isoform confers on murine cells the abil- ity to bind measles virus and to induce cell-cell fusion. Work is progressing to determine the role of different CD46 isoforms and whether virus binding to CD46 can lead to cell infection. Identification of the receptor introduces the possibility of developing a transgenic mouse as a practical animal model for the study of measles. The outcome of infection of a cell with measles virus may be cell lysis or persistent infection. Expression of mRNAs downstream of the N gene is reduced in human lymphocytic and monocytic cell lines compared with expression in the monkey kid- ney cells commonly used for in vitro propagation of virus. Comparison of the expression of mRNA in lytically infected and persistently infected leukocyte cell lines revealed no differences. However, expres- sion of the F and H glycoproteins is reduced during persistent infection. A similar pattern of reduced expression of the F and H glycoproteins can be induced by expression of MxA, a cellular protein induced by interferon. This binding raises the pos- sibility that MxA protein expression contributes to persistence of measles virus in certain cells. Monocytes are the primary peripheral blood mononuclear cell (PBMC) infected during measles; this leads to the production of viral RNA and pro- teins but little cell-free infectious virus. Monocyte function is altered by infection, resulting in normal or increased production of IL-1p, expression of increased MHC class II, and decreased tumour necro- sis factor-a. Infected monocytes allow presentation of measles virus antigens to T cells, but inhibit the presentation of other antigens. This change in mono- cyte function may contribute to the immune sup- pression that occurs during measles infection. Data were also presented suggesting that the interaction of infected cells that express viral pro- teins with uninfected lymphocytes can affect the pro- liferation, and therefore, function and activation, of lymphocytes through a transmembrane signalling pathway. Therefore, several mechanisms have been identified by which viral infection of leukocytes, particularly monocytes, may lead to suppression of immune responses to other antigens and infectious agents. Measles and immune responses Immune suppression has long been recognized to be a consequence of measles infection and a likely contributor to the increased susceptibility to second- ary bacterial and viral infections that characterizes the disease. At the same time, the immune response to measles virus is effective in clearing infection and in establishing lifelong immunity to reinfection. Fail- ure of the immune system to clear measles virus is associated with congenital and acquired cellular im- munity deficiencies. Occasionally, vaccine strains of live, attenuated measles virus given to children with severe combined immunodeficiency have led to pro- gressive fatal infection. Several advances have been made in understanding the inter-relationships between measles virus infection and the induction of virus clearance along with immune suppression. During the period when measles rash is present, measles-virus-specific CD8+ cytotoxic T cells are generated and can be detected in cerebrospinal fluid (CSF) as well as blood; simultaneously, the plasma levels of soluble CD8 increase. Memory MHC class- I-restricted cytotoxic cells can be detected in normal seropositive individuals, and the epitopes identified have been predicted using known peptide anchor sequences. To date, CD8 epitopes have been defined in the N, F, H and L proteins; in contrast, during per- sistent infection (SSPE) only CD4+ T cells can be found in blood or CSF. As with other virus infec- tions, induction of a CD8+ T cell response requires measles virus replication, while induction of CD4+ T cells does not. WHO Bulletin OMS. Vol 72 1994 201 Memorandum Measles-virus-specific MHC class-1I-restricted CD4 T cells are also induced during infection, and epitopes have been mapped in the F protein near the transmembrane segment. Persistently elevated plas- ma levels of soluble CD4 suggest that activation of this cell continues for a longer period of time than activation of CD8 T cells. CD4 T cells can be classi- fied functionally into type 1 and type 2, according to the pattems of lymphokines that they produce. Type- 1 CD4 cells characteristically produce interferon and IL-2 and elicit macrophage activation and delayed- type hypersensitivity reactions; type-2 CD4 cells characteristically produce IL-4, IL-5 and IL-10, pro- viding help for B-cell proliferation and differentia- tion, and counteracting the effects of interferon on macrophages. Studies of the lymphokines produced in vivo and by PBMCs cultured in vitro after measles infection suggest that type-2 CD4 cells secreting IL- 4 are activated for a prolonged period. The function- al characteristics of type-2 CD4 T cells in providing B-cell help, down-regulating macrophage activation, and suppressing type-I CD4 T cell function are con- sistent with the observations of increased IgE levels, decreased natural killer cell activity, decreased delayed-type hypersensitivity skin test reactivity, decreased lymphocyte proliferation, and robust pro- duction of measles-virus-specific antibodies after infection. Thus, immune suppression may also in part be the consequence of activation of type-2 CD4 T cells during measles infection. Animal models for measles research Data were presented on development of the macaque model for measles virus infection that was recently established in the Netherlands at the request of WHO. Cynomolgus monkeys (Macaca fascicularis) infected intratracheally with 1-104 TCID50 of MV- BIL, a measles virus recently isolated in the Nether- lands, showed extensive replication of the virus in lung macrophages, PBMCs, and pharyngeal cells. The onset and peak of the infection varied with the dose of virus used, and correlated with the develop- ment of specific IgM, IgA, and IgG serum antibo- dies. Lethargy, lack of appetite, and temperature rise were observed during viraemia. The same parameters were determined in mon- keys infected with wild-type Edmonston strain and Schwartz vaccine strain. Essentially the same kinet- ics were observed, but the numbers of infected cells and antibody titres were considerably lower. After challenge with the wild strain MV-BIL, all the mon- keys used in these studies proved to be protected. The potential use of this model for evaluation of the efficacy and safety of new generations of vaccines were discussed extensively. A model for the atypical measles syndrome is currently being developed. Experimental models of measles encephalitis in rats and mice have been established for studying the pathogenesis and immune regulation of acute and persistent central nervous system (CNS) infection by measles virus. After intracerebral infection, the dis- ease course is dependent on the age and strain of the animals. Susceptibility to measles encephalitis is immunogenetically controlled and correlates with MHC haplotypes. Molecular characterization of the replication revealed restriction of envelope gene expression similar to SSPE and MIBE. Intrinsic properties of brain cells account for this phenome- non, and the presence of certain antiviral antibodies acts synergistically via currently undefined cellular mechanisms on viral transcription and leads to anti- genic modulation in vitro and in vivo. Neurovirulent measles virus variants have also been selected in the presence of antiviral antibodies, leading to subacute encephalitis in rats. Studies on the role of different effector mecha- nisms responsible for the protective immunity of rodents against measles virus infection have shown that adoptive transfer of CD4+ T cells against the N, F and H proteins could protect rats and mice from encephalitis. Rats and mice infected with recombi- nant vaccinia viruses expressing the N, F or H pro- teins had complete protection. Finally, autoimmune CD4+ T cells from rats with measles-virus-induced encephalitis have been shown to induce experimental allergic encephalitis in non-infected recipient rats. Collectively, the data presented showed that infec- tion of mice and rats with rodent-adapted, neurotrop- ic measles viruses provides an attractive model to study the pathogenesis and immune regulation of measles virus CNS infection. Approaches to new measles virus vaccines Data were presented on the induction of immunity with recombinant vaccinia viruses (rVV) encoding one or more different measles virus antigens. Mice immunized with rVV-M and rVV-F proved to be protected. Similar experiments with rVV-MV in the canine distemper virus (CDV)-dog system showed that rVV-N, rVV-F, and rVV-H induced a certain level of protection against CDV infection. A role for cell-mediated immune responses was suggested. rVV-N and rVV-H induced measles-virus-specific cytotoxic T lymphocyte (CTL) activity in BALB/c and C3H mice. It was concluded that rVV-MV ful- fils a number of important parameters for a measles vaccine: induction of virus neutralizing and fusion inhibiting antibodies and of CTL, which eliminate virus-infected cells. WHO Bulletin OMS. Vol 72 1994202 Pathogenetic aspects of measles virus Infection Work is in progress on two highly attenuated poxvirus vectors: NYVAC and ALVAC. NYVAC is a strain of vaccinia virus derived from the Copen- hagen strain, whereas ALVAC was derived from an attenuated strain of canarypox virus. NYVAC- and ALVAC-based recombinants expressing the H and F proteins of measles virus have been developed. Pre- clinical studies in rodents and squirrel monkeys have shown that measles-virus-specific virus neutralizing, haemagglutination-inhibiting, and fusion-inhibiting antibodies were induced. Both single (H/F) and dou- ble (H+F) recombinants of ALVAC induced protec- tion against canine distemper in dogs. The potential of both NYVAC and ALVAC recombinants as future measles vaccines was discussed. Evaluation in the macaque model seems to be the appropriate next step. Defective recombinant adenoviruses as vectors for vaccination against measles have been studied. The N gene of measles virus has been inserted into a defective adenovirus under the control of the CMV IE promoter. This virus was shown to produce high levels of the N protein of the correct size in MRC5 cells, which assembled into large nucleocapsid-like structures. High levels of N-specific antibodies, an N-specific CMI response, and protection were induced in mice vaccinated with this virus. Attempts to insert genes coding for the H and F proteins into defective adenovirus vectors are currently being made. The advantages of using defective adenovirus vectors as human vaccines were summarized. Among these, the ability to stimulate mucosal immu- nity upon oral administration and the restricted host range of the attenuated viruses may be considered to be of major importance for their use as candidate vaccines against measles in humans. Experiments are in progress on the use of BCG that expresses measles virus antigens to prime mem- ory T cells as an immunization strategy. Firstly, the advantages of BCG as a vaccine vehicle for measles antigens were summarized. Of these, its safety, stability, low cost, and the possibility of oral admin- istration soon after birth in the presence of maternal antibodies seemed the most relevant for its future use in candidate measles vaccines. Recombinant BCG strains expressing the N protein have been tested in C3H and BALB/c mice and proved to be safe and stable in vivo. Proliferative T-cell responses to the N protein were induced, but no specific serum anti- bodies were detected. There was some degree of pro- tection against CNS measles virus infection in C3H mice. Higher rates of protection may be conferred by recombinant BCG strains secreting the measles virus N protein, e.g., as a fusion protein. Data were also presented supporting the use of the immune-stimulating complex (ISCOM) matrix as a candidate for the development of a measles subunit vaccine. An overview was provided on the results obtained with ISCOMs for the immunogenic presen- tation of viral antigens, and points in favour of the use of such complexes for the generation of a candi- date measles vaccine were summarized. Among these the most relevant were the potent induction of long-lasting biologically active antibody responses to the F and H proteins (also in the presence of passive- ly transferred antibodies), the induction of specific CD4+ and CD8+ T cell responses, and the possibil- ities for local application. Experiments in different model systems (MV-mouse, CDV-dog/seal) indicate that ISCOM preparations containing the F and H proteins of morbilliviruses are potent inducers of protective immunity. Experiments to evaluate the potential of MV-ISCOM to induce specific B- and T- cell responses and protective immunity in macaques are in progress. Special attention will be given to the comparison of the induction of protective immunity in macaques with MV versus avipox recombinant viruses expressing the F and H proteins. General discussion and recommendations The discussion focused on major issues concerning immune responses to measles virus infection, the role of subclinical infection in measles transmission, adverse events associated with measles and measles vaccination, the diagnosis of measles in developing countries, and alternative vaccination strategies. There was general agreement that information is incomplete on the immune response and measles virus infection. In particular, cell-mediated immune response and its role in the recovery from infection, protection from reinfection, and lifelong immunity are poorly understood. The following are important areas for future research: the relative importance of envelope and internal viral proteins as antigens for cell-mediated immunity, and the specific immunological features of infection by measles virus in infants. The same priority should be given to the study of atypical mea- sles associated with the use of inactivated vaccine. An understanding of the pathogenetic mechanisms leading to atypical measles is essential for the evalu- ation of any new subunit vaccine; in this context, the monkey model is currently the only system in which this aspect can be properly studied. Complications of measles, such as measles encephalitis (ME) or SSPE, were discussed in rela- tion to the use of new measles virus isolates as vac- cines and new routes of administration. Although a theoretical risk cannot be excluded, the observation of a significant decrease in CNS complications after WHO Bulletin OMS. Vol 72 1994 203 Memorandum the introduction of widespread measles vaccination in the USA and elsewhere, indicates that the current immunization procedures appear not to promote ME or SSPE. Therefore, it was felt that studies on patho- genesis and factors involved in the establishment of measles-virus-associated CNS diseases do not require a special research emphasis in the context of vaccine development, although they should be evaluated in postmarketing surveillance studies. A discussion on how to improve measles vacci- nation produced some controversy. Although the cur- rently available live, attenuated measles virus vac- cines are safe and effective, they will not achieve the goal of reducing the number of measles cases world- wide by 90% in 1995, with a concomitant reduction of associated mortality from 1-25% to 0.01%. This predicted failure is related to the inhibition of live vaccines in young children owing to the presence of maternal antibodies and the observation that the present measles vaccines do not always induce life- long immunity. The basis of lifelong immunity to measles after wild-type infection is unclear. So far, no evidence of measles virus persistence has been found, although antigen persistence or the presence of longlife memory cell populations and polyclonal activation of virus-specific immune responses are possible mechanisms. Even though the current vaccines, when given properly, prevent measles in industrialized countries, they are only of limited efficacy in infants under 1 year of age in developing countries. Although it has been shown previously that mass immunization cam- paigns do control virus spread, and therefore indi- rectly protect small infants, this strategy will only be effective in the short-term. The participants therefore agreed that there is a need for new measles vaccines which, in particular, would be useful for vaccination at 2-6 months of age, in the presence of maternal antibodies. Administration of such vaccines may lead to efficient priming of the immune system and, in combination with subsequent booster immuniza- tion using either novel vaccine or existing live, atten- uated vaccine, may provide new strategies for mea- sles immunization. In the discussion relating to the development of novel vaccines the participants agreed that in the absence of attenuation markers it would probably not be advisable to use any of the currently circulating measles virus strains as new live, attenuated vaccine. Opinions were expressed that it would probably be worthwhile to test in an animal model the highly attenuated poxvirus vectors expressing measles envelope genes as well as the ISCOM matrix as candidates for the development of measles subunit vaccine. In both cases preliminary data presented during the meeting were very encour- aging, and it was suggested that these two candidate vaccines should be immediately tested in the monkey model. However, it was emphasized that other vec- tors expressing measles virus antigens, e.g., adenovi- rus, BCG, etc., should be further developed. It was emphasized that the diagnosis of measles in developing countries is inhibited by the lack of simple laboratory tests that are quick, inexpensive, and appropriate to the climatic conditions (thermo- stable) for field trials. The availability of such assays would be important for epidemiological studies and a prerequisite for eradicating measles. Overall, the participants concluded that measles is a potentially eradicable disease; however, to achieve this goal, major efforts have to be undertaken in virology, immunology, and epidemiology. More- over, in addition to the maximum use of current vac- cines, novel vaccines should be developed in con- junction with improved diagnostic tests for the confirmation of measles. However, without a bal- anced investment strategy and a coordination of research activities the goal of global control of mea- sles will not be achieved. Resume Aspects pathog6netiques de l'infection par le virus de la rougeole: Mfmorandum d'une reunion de I'OMS La r6union organisee par l'OMS a Montreux les 20 et 21 avril 1993 a permis d'6valuer les r6cents progres r6alis6s dans le domaine de la recherche sur le virus de la rougeole qui, depuis 1989, est l'une des priorit6s du programme OMS/PNUD pour le d6veloppement de vaccins Genetique moleculaire du virus de la rougeole Bien que le g6nome du virus de la rougeole soit entierement s6quence, la technologie de la g6n6- tique inverse n'a, jusqu'a pr6sent, pas permis la pr6paration d'un clone infectieux. Toutefois, l'utili- sation de diff6rents systemes d'expression permet d'ores et d6ja de produire des prot6ines virales recombinantes. L'analyse du g6nome de diffe- rentes souches du virus de la rougeole, provenant de plusieurs zones geographiques et de diff6- rentes 6pidemies, a mis en 6vidence un proces- sus 6volutif permettant d'6tablir un certain nombre de lignages distincts. De plus, cette analyse montre que la proteine la plus variable est la region C-terminale de la proteine N. WHO Bulletin OMS. Vol 72 1994204 Pathogenetic aspects of measles virus infection Interaction virus de la rougeole/cellule hote Le virus de la rougeole infecte, in vitro, principale- ment les cellules humaines et de primates mais rarement les cellules d'autres especes. Bien que la plupart des infections soient lytiques, il est facile de mettre en 6vidence des infections persis- tantes et il est donc important de comprendre les m6canismes r6gulant ces deux types possibles d'infection. L'expression des antigenes viraux est reduite dans certaines lign6es cellulaires et en presence de certaines cytokines. L'infection virale peut modifier la fonction cellulaire, en particulier celle des monocytes, et contribuer, par exemple, a l'immunosuppression. De faqon similaire, des prot6ines exprim6es dans les cellules infectees peuvent affecter la prolif6ration de cellules non infect6es via, semble-t-il, la transduction d'un signal transmembranaire. Recemment, un recepteur cellulaire pour le virus de la rougeole a ete identifie. II s'agit de la molecule CD46, proteine membranaire apparte- nant a la famille des regulateurs de I'activation du compl6ment. Cette d6couverte rend possible le d6veloppement d'une souris transgenique comme modele animal pour l'etude de la rougeole. Rougeole et reponse immunitaire Au cours de l'infection rougeoleuse, des cellules T sp6cifiques CD4+ ou CD8+ sont produites et l'activation des premieres continuerait plus long- temps que celle des deuxiemes. Les 6tudes concernant le profil des cytokines synthetisees tardivement lors de l'infection virale suggerent que les cellules T CD4+ activees sont de type 2. Ceci corrobore l'importance de la r6ponse en anticorps existant a cette meme p6riode. Les cellules T memoires CD8+, dont l'activit6 est restreinte par les mol6cules du complexe majeur d'histocompatibilit6 de classe 1, sont mi- ses en 6vidence chez des individus s6ropositifs. De plus des 6pitopes T ont ete identifies au niveau des prot6ines N, F, H et L. Par contre, seules des cellules T CD4+ sont mises en 6vidence lors d'infections persistantes. Modeles animaux Des resultats concernant l'infection de macaques par le virus de la rougeole ont ete pr6sentes. Chez ces singes, l'infection par le virus de type sauvage ou de type vaccinal est suivie d'une replication virale au niveau des macrophages pul- monaires, des monocytes du sang peripherique et des cellules pharyngees, et d'une synthese d'lgM, d'igA et d'IgG specifiques. Bien que la cin6tique de la reponse immunitaire soit la meme quel que soit le type de virus ayant servi a l'immunisation, le nombre de cellules infectees et le titre des anti- corps sont consid6rablement plus bas dans le cas d'une infection par la souche vaccinale. Cette "faible" reponse immunitaire suffit a proteger tous les singes contre une 6preuve par le virus rougeo- leux de la souche sauvage. Des modeles exp6rimentaux d'encephalite rougeoleuse chez des rats et des souris ont 6t6 6tablis en vue de l'6tude de la pathog6nese et de la regulation de la reponse immunitaire lors d'infections aigues et persistantes du systeme nerveux central. Les r6sultats present6s d6mon- trent la faisabilite de ce type d'etude a I'aide de ces animaux. Approches pour la mise au point de nouveaux vaccins contre la rougeole Des 6tudes concernant l'induction de I'immunit6 avec des vaccins recombinants d6riv6s du virus de la vaccine (VV) exprimant un ou plusieurs anti- genes du virus de la rougeole ont ete pr6sent6es. L'induction d'un certain nombre de parametres clefs, comme l'apparition d'anticorps neutralisant et inhibant la fusion et de lymphocytes T cyto- toxiques, a ete montr6e. Des recombinants du virus de la vaccine plus att6nues (NYVAC) ainsi qu'un recombinant du canarypoxvirus (ALVAC) sont 6galement capables d'induire, de maniere effica- ce, des anticorps contre le virus de la rougeole. L'ad6novirus est un vecteur potentiel qui pre- sente I'avantage de stimuler I'immunite au niveau des muqueuses. Les r6sultats montrent que, dans ce systeme, la prot6ine N induit une reponse en anticorps et une r6ponse a m6diation cellulaire. Le BCG est 6galement un vecteur possible et est considere comme capable de stimuler les cel- lules T m6moires. Des 6tudes pr6liminaires, chez des souris, montrent 1'existence d'une r6ponse cellulaire induite par la prot6ine N et I'absence d'une r6ponse en anticorps au niveau du s6rum. Les r6sultats concernant l'immunogenicite des antigenes viraux contenus dans des ISCOMS ont ete pr6sentes. Ils montrent que ce conditionne- ment permet d'obtenir une r6ponse T specifique CD4+ et CD8+ et une production d'anticorps anti- F et anti-H. De plus, cette reponse en anticorps est observee meme en pr6sence d'anticorps transf6r6s passivement. Discussion generale et recommandations La discussion a e centree sur les resultats majeurs concernant les diff6rentes reponses immunitaires suite a une infection par le virus de la rougeole, le r6le des infections subcliniques dans la transmission de la rougeole, les complica- WHO Bulletin OMS. Vol 72 1994 205 Memorandum tions de la rougeole et de la vaccination antirou- geoleuse, le diagnostic de la rougeole dans les pays en developpement et les strat6gies vacci- nales de remplacement. La connaissance des me- canismes pathog6niques conduisant a des cas de rougeole atypiques est consid6r6e comme indis- pensable pour l'evaluation de nouveaux vaccins. Une discussion controversee a pris place sur la politique future en matiere vaccinale, favorisant soit les vaccins vivants att6nu6s disponibles, soit une nouvelle g6n6ration de vaccins. Cependant, sans le developpement de ces derniers, il semble impossible d'atteindre le but fix6, soit la r6duction de 90% des cas de rougeole en 1995. Le groupe s'est declare favorable a la mise au point de nou- veaux vaccins pouvant etre administr6s t6t dans la vie (2-6 mois). Les vaccins possibles a I'heure actuelle sont des vecteurs tels que le poxvirus att6nue ou les ISCOMS. Cependant, il faut encou- rager la recherche d'autres vecteurs. En r6sum6, le groupe a conclu que la rougeo- le est une maladie dont l'eradication est th6ori- quement possible. Afin d'atteindre ce but, des efforts considerables doivent etre realises dans les domaines virologique, immunologique et 6pi- demiologique. En outre, au-dela de l'utilisation maximale des vaccins d6ja existants, de nou- veaux vaccins doivent etre mis au point en con- jonction avec les tests de diagnostic am6lior6s permettant la confirmation de la rougeole. Cepen- dant, sans une strat6gie d'investissement 6quili- br6e et une coordination des activit6s de recher- che, le probleme global de la rougeole ne pourra etre maltris6. WHO Bulletin OMS. Vol 72 1994206