TRANSACTIONSOF THE iboullr %~IETY OF TROPICALMEDICINE AND HYGIENE (1991) 85, SUPPLEMENT1, 1%22

Serum P. Helena

antibodies M&e13

and bacterial

National

Public

Health

meningitis Institute,

Abstract

The role of serum antibodies in bacterial meningitis and the human defence against it is reviewed. There is good evidence from animal and human studies in vitro and in vivo showing that serum antibodies to the capsular polysaccharide of the bacteria play a major role in protection, whereas antibodies to other cell surface components vary in protective ability. A major problem associated with the use of capsular polysaccharides as vaccines has been the T cellindependent nature of polysaccharide antigens, and the often poor immune response of infants to such ant&ens. This Droblem has now been overcome bv covaent conjugation of the polysaccharide to prote&, resultinn in T cell-deoendent ant&ens. Vaccines produc& with this te&nology from the capsular polysaccharide of Haemophilus i$uenzae type b have already proven highly immunogenic and protective in young infants. The technology is being applied to other encapsulated bacteria. Introduction

Classical epidemiological studies have shown an inverse relationship between the incidence of meningitis at different ages and the presence of serum antibodies to the causative agent (FOTHERGILL & WRIGHT, 1933; G~LDSCHNEIDER et al., 1%9a). The inference has been that such antibodies protect from the disease and account for the low incidence of bacterial meningitis in adults. This gives rise to several questions. How would antibodies mediate protection? What bacterial components are important in this protection? How could we take advantage of this to reduce the incidence of bacterial meningitis? Mechanism

of protection

19

Bacterial meningitis is caused by several speciesof encapsulated bacteria. An important role of the capsule appears to be to protect the bacteria from phagocytosis; these bacteria are rapidly killed once inside polymorphonuclear leucocytes. Conversely, antibodies that bind to the bacterial surface opsonize them for phagocytosis via both Fc and complement receptors. The spleen is an important organ for phagocytosis of even weakly opsonized bacteria, and understandably lack of spleen function increases the requirement for serum antibodies for protection (HOSEA et d., 1981). A second host defence function, complementmediated killing, is also dependent on antibodies. In this oathwav. antibodv-initiated comolement activation has to &ceed thrbugh all its ste;s and leads to a final insertion of the comolex CS-C9 in the bacterial membrane. The Gram-pdsitive pneumococci are not susceptible to this mechanism because their thick layer of peptidoglycan prevents the access of this complex to the membrane. Neisseria speciesare very susceptible, and thus persons with defects in one of the terminal complement components (from C6 to C9)

Mannerheimintie

166, Helsinki,

Finland

are specifically sensitive to invasive Neisseria infections (PETERSEN et al., 1979). Both immunoglobulin (Ig) G and IgM antibody classes are efficient in these 2 antibody-mediated defence mechanisms?whereas IgA is not. In fact, IgA antibodies to bacterd cell surface components have been shown to inhibit the bactericidal and o~sonic activities of antibodies of other Ig classes(GRIF~ISS & BERTRAM, 1977). This antagonistic effect of IgA has been suggestedas an explanation of the apparent lack of protection of serum antibodies in some patients in whom meningitis developed in spite of substantial antibody titres in the acute phase serum (GOLDSCHNEIDER et al., 1969b). Specificity

of protective

antibodies

To opsonize or start complement-mediated lysis, antibodies have to bind to the bacterial surface: therefore the main surface component, the capsular polysaccharide, is also the main target of protective antibodies. The inverse correlation between disease incidence at different ages and serum antibodies, initiallv shown for bactericidal antibodies. is verv clear f& anti-capsular antibodies also (PELTOLA er aZ:, 1977). Monoclonal anticapsular antibodies protect very well in experimental infection models (SAUKKONEN & LEINONEN: 1988), and immunization with capsular polysaccharlde vaccines protects from disease, provided a satisfactory response is obtained (PELTOLA et al., 1977; GRIFFISS et al., 1987). Other surface components can also be protective, at least in experimental infection models (SAUKKONEN & LEINONEN, 1988; SAUKKONEN er al., 1989; YOTHER et al., 1982; MUNSON et al., 1983). A

prerequisite appears to be that the target of the antibody is exposed on the surface of intact bacteria. In such models, antibodies to the C polysaccharide of pneumococci of types that have a less well developed capsule, to lipopolysaccharide, and to exposed epitopes of certain outer membrane proteins of meningococci and Haemophilus injkenzae type b have been shown to be motective. The swcial interest in identifying such alternative targets for protective immunitv derives from the PTOUD B N. me&&i& and the difliculties in develo$ng-a vaccine ba&d on its capsular polysaccharide. The outer membrane protein of class I is currently the most promising candidate for a group B meningococcal vaccine (SAUKKONEN

Capsular

et al., 1989; MCGUINNESS

polysaccharides

as vaccines

et al., 1990).

Vaccines basedon the capsular polysaccharide have been developed, tested and used extensively for the three main causative speciesof bacterial meningitismeningococci, pneumococci and Huemophilus infuensue type b. Despite much success, problems still remain, as discussed below. The different polysaccharides differ in their ability to immunize (MAKELA et al., 1981). Some of them are

good immunogens in all age groups, e.g. the group A meningococcal and type 3, 4, 8 or 9 pneumococcal polysaccharides. At the other extreme, the type B meningococcal polysaccharide is a poor immunogen in all age groups, probably due to immunological tolerance based on the existence of very similar polysaccharide structures on some human glycoproteins (FINNE et al., 1983). However, many of the capsular polysaccharides are quite good immunogens in adults, but elicit very little antibody response in infants-good examples being the capsules of H. in&ensue tvw b, meninaococci of STOUD C. and pneumococc: of the so-ca&d pediatricgroups 6, 14, 19 and 23. This is of course a serious nroblem from a medical point of view, since me&&is and other infections causedby these bacteria are most prevalent at exactly the time of their lack of immunogenicity. A similar pattern of low or lacking antibody response to these polysaccharides is seen both after vaccination and after disease (KjiYHTY et al., 1981) caused by these bacteria, so the problem cannot be in the vaccine preparation but rather in the nature of the polysaccharide. It is, however, not possible in the present state of knowledge to tdentify the structural features of a polysaccharide that would make it a poor or a good immunogen in infancy. Antibodies elicited by immunization with capsular polysaccharide vaccines remain at an elevated level (compared to the natural level of antibodies in an unimmunized person) for varying periods of time, depending on the person’s age and condition. Thus pneumococcal antibodies may last for 10 years or more in immunized young adults, but disappear more rapidly in elderly persons as well as in splenectomized or immunocompromised persons (MUFSON et al., 1987; GIEBINK et al., 1984). Antibodies to the group A meningococcal polysaccharide remained elevated for less than 3 years if the vaccine had been given before 3 years or age, but for 5 years in children vaccinated when 12 years old (KAYHTY et al., 1980; RAUTONENet a!., 1986). Re-imrnunizatlon with a polysaccharide vaccine does not give rise to an increased secondary type of response. If the second dose is given rather soon, 3-6 months, after the first dose, the responsemay be even smaller than the responseof previously unimmunized persons of the same age; however, after a longer interval the responsewill be the sameas in previously unimmunized subjects (GOLD et al., 1975; KAYHTY et al., 1984). The immunoglobulin class distribution of the antibodies formed in response to a capsular polysaccharide vaccine contains all the three main classes IgA, IgG and IgM. The IgA antibodies are relatively short-lived. IgG is the main class of antibodies (KiiYHTY er al., 1981)later after immunization as well as in unimmunized persons. However, at all times IgM occupies a much larger share of the serum antipolysaccharide antibodies than of antibodies to protein antigens. All these characteristics of the antibody responseto capsular polysaccharides are suggestive of a T cellindependent nature of the polysaccharide antigens. This would mean that they can directly stimulate B cells to antibody production, but cannot stimulate T cells and therefore do not benefit from T cell help. Therefore, a memory cell population is not estab-

lished, leading to the observed lack of an increased secondary response and to a lack of maturation in antibody affinity. The observed T cell-dependent suppressive effect described in certain experimental models was apparently due to T cell response to the antibodies elicited (BAKERet al., 1982). The T cell independence could also explain the lack of response to many polysaccharide antigens in infancy: the direct stimulation of B cells is believed to require a late-maturing type of B cell not present at that time (HOWARD,1987). Polysaccharide-protein

conjugate

vaccines

A logical idea for improving the immunogenicity of polysaccharides was therefore their conjugation to protein in the hope of thus converting them to protein-like T cell-dependent antigens. Techniques for this were available from immunological studies with haptens and hapten-protein conjugates, and studies in experimental animals suggested that the expected result could be achieved with many polysaccharides. The vaccine that has served as a prototype of polysaccharide-protein conjugate vaccines is that for H. infhwzue type b. The capsular polysaccharide of H. in&en.zue type b was initially expected to provide a means of preventing meningitis causedby this bacterium in infants and young children, in whom it occurs at a high frequency before natural antibodies develop at the age of about 2 years (FOTHERGILL& WRIGHT, 1933; PELTOLA et al., 1977; SCHNEERSON et al., 1971). In a field trial the polysaccharide vaccine proved, however, disappointing at exactly this age: protection was good in 2-4 years old children, but absent under 18 months of age(PELTOLAet al., 1984). The poor protection was correlated precisely with poor immunogenicity of the vaccine under 18 months of age (PELTOLAet al., 1977). When this vaccine was licensed in the USA for use in children after their second birthday, quite a few vaccine failures were reported (OSTERHOLMet al., 1988): apparently the age limit is not as clearcut as initially implied, so that in some children the necessary B cell responsiveness matures later. There was, therefore, a pressing need to obtain a vaccine that would work in infancy and could thus be used very widely: H. in@znzue type b meningitis is an important and serious- disease of infants a&d young children world-wide (MUSSERet al., 1990). The first conjugate H. influenzue type b vaccine tested in field trials contained the nolvsaccharide linked to dinhtheria toxoid (ESKOLAit al., 1985). Other conjug’ates,in which the polysaccharide had been partially reduced in size and the protein carrier and the mode of linkage varied, have since been introduced and shown to have more or less similar relevant properties (EINHORNet al., 1986). Covalent linkage to protein seemsessential for these vaccines as, on the basis of animal experiments, is also the amount of polysaccharide relative to protem. The H. influenzae type b conjugates were indeed able to immunize young infants: 2 or 3 doses starting as early as 2 months of age gave rise to antibody concentrations not achieved with the polysaccharide vaccine below the ageof 2 years; with someconjugates even one dose appeared sufficient (ESKOLA et al., 1985; EINHORNet al., 1986). Most importantly, the

21 response could be boosted, i.e. a further dose several months later gave rise to a much increased secondary response. In further agreement with a T cell-dependent nature of the conjugate vaccines, the antibody responsecontained a higher proportion of IgG and the amount of IgGi relative to IgGr was increased: however, in these respects the response still appeared to be intermediate between responsesto proteins and those to polysaccharides (SEPPALA et al., 1988). Most importantly, the H. injluen.zae type b coniugate vaccines have proven efficacious in preventing meningitis and other invasive infections caused by this bacterium in infancy (ESKOLA er al., 1990; BLACKet al., 1991). The protective efficacy was of the order of 90% after the initial series of 2 or 3 injections in infancy, and 100% after the booster dose given about one year later. Furthermore, the nation-wide use of these vaccines in Finland has, in four years. practically eliminated invasive H. in$ueneae type b disease in children under 5 years of age (data to be published). Conclusions The demonstration that the principle of converting a T cell-independent polysaccharide vaccine that was ineffective in infancy to a T cell-dependent conjugate that was highly efficacious in infants actually worked is of fundamental importance because it proved that the problems associated with the polysaccharide vaccine were in fact due to its T cell-indeoendent nature. This finding opens the way for converting other, only partially effective, polysaccharide vaccines to more effective T cell-denendent coniuaates. The possibility of giving such conjugate ‘vaccines in infancy when a number of vaccines are routinely administered will greatly facilitate their use. The immunological memory elicited by the conjugates means that booster doses of the vaccine can be given to ensure adequate serum antibody levels through life. In many casesnatural colonization of the nasopharynx or intestine by the sameor cross-reactive bacteria will provide a natural booster and maintenance of protective immunity. Conjugates of meningococcal group A and C polysaccharides combined with the H. infkert.zue type b conjugate could thus soon become general meningitis vaccines to be incorporated in infant immunization programmes. Conjugates of the most common pneumococcal polysaccharides are expected to provide protection not only from meningitis but also from the much more common respiratory infections, pneumonia and otitis media, caused by pneumococci. The main problems with such conjugate vaccines envisaged at present are their cost and the question of whether or not the samecarrier protein could be used in all of them. These problems do not appear insurmountable, at least not in industrialized countries, and thus the problem of bacterial meningitis may be much reduced, in fact leaving only the group B meningococcus as an important future challenge. References Baker, P. J., Amsbaugh, D. F., Stashak, P. W., Caldes, G. & Prescott, B. (1982). Direct evidence for the involvement of T suppressor cells in the expression of low-dose paralysis to type III pneumococcal polysaccharide. Journal of Zmtnunobgy, 128, 1059-1062.

Black, S., Shinefield, H. & Fireman, B. (1991). Efficacy in infancy of oligosaccharide conjugate Haemophilus injluenzae type b vaccine (HbOC) in a US population of 61,080 ~bdben. Pedtamc Znfectwus Dtseases Journal, 10, 9?Einhorn, M. S., Weinberg, G. A., Anderson, E. L., Granoff, P. D. & Granoff, D. M. (1986). Immunogenicity in infants of Haemophilus injkenzae type b polysaccharide in a conjugate vaccine with Neisseriu meiringiridis outer-membrane protein. Lancez, ii, 299-302. Eskola, J., K;iyhty, H. & Peltola, H. (1985). Antibody levels achieved in infants by course of Haemophilus injkenzae type b polysaccharide/diptheria toxoid conjugate vaccine. Lancet, i, 1184-1186. Eskola, J., K;iyhty, H. & Takala, A. K. (1990). A randomized prospective field trial of a conjugate vaccine in the protection of infants and young children against invasive Haemophilus influenzae type b disease. New England Journal of Medicine, 323, 1381-1387. Finne, J., Leinonen, M. & Miikela, P. H. (1983). Antigenic similarities benveen brain componentsand bacteria causing meningitis: implications for vaccine development and pathogenesis. Lancet, ii, 355-357. Fothergill, L. D. & Wright, J. (1933). The relation of age incidence to the bactericidal power of blood against the causal organism. Journal of Zmmunology, 24, 273-284. Giebink, G. S., Le, C. T. SCSchiffman, G. (1984). Decline of serum antibody in splenectomixed children after vaccination with pneumococcal capsular polysaccharides. %wnul of Pediatrics. 105. 57G-584. Gold, R., Lepow, M.. L. ik Goldschneider, I. (1975). Clinical evaluanon of group A and group C meningococCalpolysaccharide vaccinesin infants. 3oumal of Clinical Znvestigation, 56, 1536-1547. Goldschneider, I., Gotschlich, E. C. & Artenstein, M. S. (1%9a). Human immunity to the meningococcus. I. The role of humoral antibodies. Jouml of Experimental Medicine, 129, 1307-1326. Goldschneider, I., Gotshclich, E. C. & Artenstein, M. S. (1969b). Human immunity to the meningococcus. II. Development of natural immunity. Journul of Experimental Medicine, 129, 1327-1348. Griffiss, J. McL. & Bertram, M. A. (1977). Immunoepidemiology of meningococcal disease in military recruits: II. Blocking of serum bactericidal activity by circulating IgA early in the course of invasive disease. Journal of Znfectious Diseases, 136, 733-739. Griffiss, J. McL., Apicella, M. A., Greenwood, B. & M&e+ P. H. (1987). Vaccines against encapsulated F;terl;8global agenda. Remews of Znfectwus Deceases, Hose:, S. W.,‘Brown, E. J. & Hamburger, M. I. (1981). Opsonic requirements for intravascular clearance after splenectomy. New England 3ouwd of Medicine, 304, 24525n -.- -_-. Howard, J. G. (1987). T cell-independent responses to polysaccharides: their nature and delayed ontogeny. In: Towards Better Carbohvdrate Vaccines. Bell. R. & Torrigiani, G. (editors). &Chester, UK: John-Wiley (for the World Health Organization), pp. 221-231. Kiiyhty, H., Karanko, V., Peltola, H., Sama, S. & Miikela, P. H. (1980). Serum antibodies to capsular polysaccharide vaccine of group A Neisseria meningitidis followed for three years in infants and children. 3oumal of Infectious Diseases, 142, 861-868. Kiiyhty, H., Jousimies-Somer, H., Peltola, H. & Maehi, P. H. (1981). Antibody response to capsular polysaccharides of groups A and C Neisseria meningitidis and Haemophilus iy?uenzae type b during bacteremic disease. Journal of Znfectious Diseases, 143, 32-41. KByhty, H., Karanko, V. & Peltola, H. (1984). Serum armbodies after vaccination with Haemophilus in&enaae type b polysaccharide and response to reimmunixationno evidence of immunological tolerance or memory. Pediam’cs, 74, 857-865.

22 M&e& P. H., Leinonen, M., Pukander, J. & Karma, P. (1981). A study of the pneumococcal vaccine in prevention of clinically acute attacks of recurrent otitis-media. :lz$eu~ of Infectious Diseases, 3, supplement, S124McGuinhess, B., Barlow, A. K. & Clarke, I. N. (1990). Deduced amino acid sequencesof class 1 protein (porA) from three strains of Neisseria meningitidis. 3ouw1al of Experimental Medicine, 171, 1871-1882. Mufson, M. A., Krause, H. E., Schiffman, G. & Hughey, D. F. (1987). Pneumococcal antibody levels one decade after immunization of healthy adults. AmericanJoutvual of

Medicine, 310, 1566-1569. Petersten, B. H., b,, T. J., Snydermann, R. & Brooks, G. F. (1979). Neisserrameningiridisand Neisseriagonowheeae bacteremia associated with C6, C7, or C8 deficiency. Annals of Zntemal Medicine, 90, 917-920. Rautonen, N., Pelkonen, J., Sipinen, S., Kiiyhty, H. & Maehi, 0. (1986). Isotvpe concentrations of human antibodies to‘ group A meningococcal polysaccharide. ‘fournal of Immunoloev. 137. 2670-2675.

Munson, R. S., Jr, Shenep, J. L., Barenkamp, S. J. & Granoff, D. M. (1983). Puritication and comparison of outer membrane protein P2 from Haemophilus influenzae gdzt4isolates. 3ournal of Clinical Investigation, 72,

Saukkonen, K. & Leino%i, M.*(1988). Infant rat model for meningococcal group B infection. In: Gonococci and Meningococci, Poolman, J. T. et al. (editors). Dordrecht, The Netherlands: Kluwer, pp. 815-820. Saukkonen, K., Leinonen, M., Abdillabi, H. & Poolman, J. T. (1989). Comparative evaluation of potential components for group B meningococcal vaccine by passive protection in the infant rat and in z&o bactericidal assay.

Musser, J. M., Kroll, J. S. & Granoff, D.. M. (1990). Global genetic structure and molecular epidemiology of encapsulate Haemophilus iybenzae. Reviews of Infectious Diseases, 12, 75-l 11. Osterholm, M. T., Rambeck, J. H. & White, K. E. (1988). Lack of efficacy of Haemophilus b polysaccharide vaccine in Minnesota. journal of the American Medical Associatian, 260, 1421-1428. _ Peltola, HJ., K;iyhty, H. & Sivonen, A. (1977). Haemophilus injuenzae type b capsular polysaccharide vaccine in children: a double-blind field trial of 100,090vaccinees 3 months to 5 years of age in Finland. Pediatrics, 60, 730-737. __ _ Peltola, H., Kiiyhty, H. & Virtanen, M. (1984). Prevention of Haemophilus injIuenaae type b infections with the capsular polysaccharide vaccine. New EnglandJoumal of

Schneerson,R., Rodriques, L. P. & Parke, J. C., Jr (1971). Immunny to disease caused by Haemophilus influenzcre type b: II. Specificity and some biological characteristics of ‘natural’, infection-acquired, and immunization-induced antibodies to the capsular polysaccharide of Haemophilus influenzae type b. 3ournal of Immunology, 107, 1081-1089. Seppitla, I., Sarvas, H., Makelii, O., Mattila, P., Eskola, J. & KByhty, H. (1988). Human antibody responsesto two conjugate vaccines of Haemophilus in@naae type b saccharidesand diphtheria toxin. ScandinavianJournul of Immunalagy, 28, 471479. Yother, J., Forman, C., Gray, B. M. & Briles, D. E. (1982). Protection of mice from infection with Streptococcus pneumoniae by anti-phosphocholine antibody. Infection and Immunity, 36, 184-188.

Medical Sciences, 293, 279-284.

Vaccine, 7, 325-328.