current status of meningococcal group b vaccine candidates ... · other extreme ofonly a benign...

CLINICAL MICROBIOLOGY REVIEWS, OCt. 1994, p. 559-575 Vol. 70893-8512/94/$04.00+0

Current Status of Meningococcal Group B Vaccine Candidates:Capsular or Noncapsular?

J. DIAZ ROMERO AND I. M. OUTSCHOORN*Unidad de Respuesta Immune, Centro Nacional de Biologia Celular y Retrovirus,

Instituto de Salud Carlos III, Majadahonda, Madrid 28220, Spain

INTRODUCTION ................................................... 559Meningococcal Meningitis and Associated Serogroups.................................................. 559Strategies Used in Developing a Group B Vaccine .................................................. 560

NONCAPSULAR VACCINE CANDIDATES .................................................. 560Noncapsular Surface Antigens: the Main OMPs .................................................. 560IRPs...................................................561LOS..........................561Lip, Pili, and CtrA.................................................. 563Clinical Trials: OM .................................................. 563Other Noncapsular Antigens: IgA Proteases .................................................. 563

THE CAPSULE.........................564Presence of Polya2-8 in Bacterial and Animal Cells.................................................. 564Molecular Mechanisms of Expression .................................................. 565Capsule Structure .................................................. 566Barrier Function ................................................... 567Role of Anticapsular Antibodies in Protection against Infection .................................................. 567Specific Antibody Recognition of Polya2-8NeuNAc .................................................. 568Strategies To Increase Response ................................................... 568

CONCLUSIONS ................................................... 569ACKNOWLEDGMENTS .................................................. 569DD1.T1IRR.L.E.NR. 7^

INTRODUCTION

Meningococcal Meningitis and Associated Serogroups

Meningococcal meningitis caused by Neisseria meningitidis iscurrently an important health problem in both developed anddeveloping countries. In the latter, estimates of 330,000 casesannually result in 35,000 deaths (172). This gram-negativediplococcus is a specific human pathogen, although it ispossible to induce the disease in monkeys by intraspinalinjection of pure cultures (180). Virulent strains, i.e., those thathave been isolated from the bloodstream or cerebrospinalfluid, are invariably capsulated. This capsule surrounds eachorganism and enables it to avoid host defenses (124). Infectionis spread by droplets in suspension, and the dispersion of themicrobe is followed by colonization of the nasopharynx. Thecarrier state is relatively common (5 to 10% depending on theseason) and only occasionally progresses to symptomatic dis-ease (15, 124).The organism can cause a variety of clinical syndromes

ranging from fulminating septicemia with septic shock andlittle or no meningitis, multiple organ failure, and death to theother extreme of only a benign meningococcemia and varioustypes of arthritis described as being associated with meningo-coccal infection (186). The most habitual presentation ismeningitis, the commonest form of which is produced byinvasion of the bloodstream by the meningococcus across theblood-brain barrier (the meninges). The organisms are thencapable of multiplying in the cerebrospinal fluid, causing

* Corresponding author. Phone: 34 1 638-1111, ext. 246, or 638-8079. Fax: 638-1611 or -8206 or 639-1859.

cerebral inflammation that often leads to mental retardation ordeath. The mechanism by which the disease progresses from anasymptomatic carrier state to invasive infection is unknown.This is partly due to the lack of a suitable animal model inwhich the various stages of human disease can be reproduced.Various animal species, including monkeys, guinea pigs, rab-bits, rats, mice, and chicken embryos, have been used in thestudy of meningococcal pathogenesis (17). The models ofintraperitoneal infection in mice and in newborn rats havebeen widely used in studies of passive protection in group Bmeningococcal vaccine evaluations (185, 224). Although thesemodels have the advantage of simplicity, they are inadequatefor reproducing the route of infection in humans. Recently, amodel of intranasal infection of newborn mice has beendeveloped, but even though it more closely approximates thenatural route, this model is still a limited one for the study ofmeningitis (131). Predisposing factors to disease developmentsuch as influenza virus infection (92, 170) or the failure tosecrete blood group substances of the ABO system (28, 240)have been identified from epidemiological data. The course ofthe disease can be very rapid, from the appearance of strikingsymptoms or diagnosis to death within 24 h.Three species of organisms, N. meningitidis, Haemophilus

influenzae, and Streptococcus pneumoniae, cause approximately75% of the cases of bacterial meningitis and share the commoncharacteristic of having polysaccharide capsules that play afundamental role in pathogenesis (28). The N. meningitidiscapsule is made up of high-molecular-weight anionic polysac-charides that form the basis of the serological classification ofthe species into distinct serogroups: A, B, C, 29-E, H, I, K, L,W-135, X, Y, and Z (19, 201). Meningococcal disease is anendemic as well as worldwide epidemic illness. The endemic

559

7, No. 4

xupir JDJXUPJI,4 %-ria ............................................................................................................................................................aoy

on March 29, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

560 DLAZ ROMERO AND OUTSCHOORN

form has an incidence of one to five cases per 100,000population anually, while epidemics with peaks of 500 casesper 100,000 have occurred (124). Although the proportion ofmeningococcal disease caused by individual serogroups variesamong countries, more than 95% is caused by strains ofserogroups A, B, C, Y, and W-135 (224). More than 90% of thestrains isolated from patients belong to serogroups A, B, andC, and group B alone accounts for 50 to 70% of the cases.Serogroups X, Z, and 29-E are only rarely associated withdisease, and patients with infections caused by strains of theseserogroups generally have some form of immunodeficiency(100). A high proportion of meningococcal isolates from thenasopharynx are not serogroupable. In one study of nasopha-ryngeal spread within a population in which the causativeorganism belonged to the B:15:P1.16 (serogroup:serotype:subtype) strain (35), approximately half of the strains identifi-able as the epidemic strain by the presence of serotype- andsubtype-specific antigens (see below) could not be groupedserologically. Apparently, certain meningococci tend to losethe ability to express capsules during nasopharyngeal coloni-zation. Noncapsulated strains can be easily eliminated bynormal human serum and hence are normally incapable ofcausing systemic disease. Such a strain can cause meningitisunder exceptional circumstances, e.g., C6 deficiency (92a).However, some evidence exists indicating that the polysaccha-ride capsule impedes fusion with mucosal cells in such a waythat the diminished presence of polysaccharide on the cellsurface could help nasopharyngeal colonization (195, 219).

Substantial evidence exists supporting the unique role of thecomplement system in the prevention of meningococcal dis-ease (44, 45, 59), which is the most habitual infection incomplement-deficient individuals. The frequency of hereditarycomplement deficiency states in patients with neisserial diseaseis 5 to 10% (59). Meningococcal disease is as characteristic ofindividuals with deficiencies in one of the alternative pathwayproteins (properdin or factor D) as it is in individuals withdeficiencies in one of the terminal complement components. Inthe former instance, they are incapable of activating comple-ment yet possess an intact classical pathway (150, 190),whereas in the latter, they cannot assemble the membraneattack complex and express complement-dependent bacteri-cidal activity (13). In both groups, about 60% of the individualsdevelop infections during adolescence. In the group sufferingfrom alternative pathway deficiencies, however, a high mortal-ity rate (50%) results, and those who survive have infrequentrecurrence. The group with the deficiency in terminal comple-ment components has a lower mortality rate, but almost halfsuffer from recurrent infections.

Hence, the importance of an intact alternative complementpathway and complement-dependent bactericidal activity asbasic defense mechanisms against N. meningitidis infectioncannot be overstated. The effectiveness of meningococcalvaccines should therefore correlate with the induction andpersistence of complement-dependent bactericidal antibodies.

Strategies Used in Developing a Group B VaccineAt present, the group B meningococcus is the commonest

cause of bacterial meningitis in industrialized countries afterH. influenzae type b, which could soon disappear when avaccine (22, 42, 209) comes into general use. Although groupA and C strains are the principal causes of epidemic bacterialmeningitis in the meningitis belt of Africa and in China (2,222), group B strains have been recognized as being responsi-ble for the high disease prevalence maintained in several areasof hyperendemicity, including Norway, the Faroe Islands, and

some parts of England (173, 225, 228). In Spain, serogroup Bis responsible for endemic meningitis in Galicia and has beenthe predominant serogroup in the last epidemic waves (91% ofthe cases in the 1980 epidemic) (58, 180).

Meningococcal vaccines against the A and C serogroupswere developed at the end of the 1960s (80, 230). These weremade of high-molecular-weight purified polysaccharides de-rived from each organism group. This polysaccharide is a linearhomopolymer which, in the case of group A, is made up of0-acetylated residues in the 3 position of mannosamine-6-phosphate linked o(1-6) and, in the case of group C, is a linearhomopolymer made up of 0-acetylated residues in position 7and/or 8 (in O-acetyl-positive form) or made up of nonacety-lated (O-acetyl-negative) N-acetyl neuraminic acid linked ct(2-9). The efficiency and safety of these vaccines have beendemonstrated in a number of controlled field trials and inepidemics, although the vaccine is ineffective in children underage 2 (group C) or in infants under 6 months of age (group A),and the long-term protective effects of vaccination with acombined group A and C polysaccharide vaccine have beenunsatisfactory (37, 41). The W-135 and Y polysaccharides havenow been incorporated into a tetravalent vaccine, leavingserogroup B as the only pathogenic meningococcus for whichno vaccine is currently available.

Part of the problem in obtaining a serogroup B vaccine isthe lack of immunogenicity of the capsular polysaccharide inhumans. This polysaccharide is a homopolymer made up ofN-acetyl neuraminic acid residues in a(2-8) linkage (124).Among the alternatives studied for developing a useful vaccinefor serogroup B have been the following: using noncapsularantigens, either present on the cell surface or secreted; increas-ing the immunogenicity of the capsular polysaccharide; andinducing anticapsular antibodies by using structurally similarimmunogens (81a; cf. reference 52).

NONCAPSULAR VACCINE CANDIDATES

Noncapsular Surface Antigens: the Main OMPsAll gram-negative bacteria have a cellular envelope consist-

ing of two lipid layers on either side of a semirigid peptidogly-can layer. The outermost one, named the outer membrane, ismade up of a complex mixture of lipopolysaccharides (LPS)and proteins embedded in a lipid bilayer. The outer membranecontains three to five main outer membrane proteins (OMPs)subdivided into five different structural classes that apparentlycorrelate with their molecular weights in sodium dodecylsulfate (SDS)-polyacrylamide gels of 46,000 + 1,000, 41,000 ±1,000, 38,000 ± 1,000, 33,000 ± 1,000, and 28,000 ± 1,000 andare designated classes 1 to 5, respectively (66).

All meningococcal strains express a protein of either class 2or class 3, but never both simultaneously, and therefore aregrouped in a single class, 2/3 (91, 147). These proteins functionas porins, are homologous in sequence to gonococcal proteinsIA (class 3) and IB (class 2), and form the basis of meningo-coccal typing. Approximately 20 different serotypes exist withinN. meningitidis serogroups B and C on the basis of theimmunological differences between the proteins of class 2/3,while serogroup A reacts mainly with monoclonal antibodiesspecific for only two serotypes, 4 and 21 (226). The epitopesdefining serotypes of the class 2 and 3 proteins seem to have astrong conformational component. Two variable regions havebeen proposed for the class 3 proteins that are exposed on thecell surface and are associated with serotype specificity (234).With newborn rats in an animal protection model (185), ithas been demonstrated that monoclonal antibodies directed

CLIN. MICROBIOL. REV.


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

STATUS OF MENINGOCOCCAL GROUP B VACCINE CANDIDATES 561

against class 2/3 proteins were only occasionally protective andbactericidal.

Class 1 proteins, for which porin function with slight cationselectivity has been demonstrated (206, 207), are generallypresent in most N. meningitidis serotypes, but they may beabsent. Considerable variation in their level of expressionexists, and the expression rate can sometimes be reducedduring the course of an infection (66, 107). In contrast withother OMPs, class 1 has no known equivalent in the gonococ-cus. Antigenic variation in this protein defines the subtype of ameningococcal strain (56, 200). To date, monoclonal antibodypreparations against 15 different subtypes have been produced,and certain class 1 proteins have been demonstrated to bind totwo different subtype-specific monoclonal antibodies (227).Analogous to the class 3 proteins, the amino acid sequences ofdifferent class 1 proteins show two main variable regions wherethe subtype-specific epitopes are located (18, 152). Initially,these epitopes appeared to be linear in nature and inaccessibleto specific antibodies in the intact cells of some strains (227),but the most recent studies using cyclic peptides point to theconformational nature of the class 1 epitopes (39). Althoughantibodies to class 1 proteins were protective in the newbornrat model mentioned above, this protection appeared to betype specific (185).

Class 5 proteins, which are analogous to gonococcal proteinII, are modified by heat in a characteristic manner. Proteinssolubilized at 37°C migrate faster in SDS-polyacrylamide gelelectrophoresis (PAGE) than those solubilized at 100°C (6).These class 5 proteins show not only interstrain heterogeneitybut also two types of intrastrain variability: (i) phase variationin levels of expression; and (ii) antigenic variations, i.e.,simultaneous expression of multiple proteins. Up to fourdifferent variants of class 5 proteins have been observed in asingle strain of N. meningitidis, and a single strain can simul-taneously express none to four of them (2). Specific bacteri-cidal antibodies against exposed surface epitopes of class 5 areperhaps too specific to be considered relevant for their inclu-sion in a hypothetical vaccine.

Recently, a new OMP has been described, termed class Sc(3, 57). Of similar molecular weight to the rest of the class 5proteins, it bears only weak sequence homology to the othersbut has a conserved interstrain sequence. An important rolein the invasion and colonization of the epithelial mucosa hasbeen suggested for this protein (219). Specific human mono-clonal antibodies show bactericidal activity against meningo-coccal strains that express high protein Sc levels but fail to doso with the laboratory strain B385 (57). Most virulent menin-gococcal isolates are probably not susceptible to this mecha-nism of killing, because they express only small or undetectableamounts of 5c (174).The class 4 protein appears to be very stable and conserved

in all meningococcal strains. It bears considerable homology togonococcal protein III and to a portion of the OMP, OmpA, ofEscherichia coli. Monoclonal antibodies directed against pro-tein III also react with the class 4 protein; however, not only dothese have no bactericidal activity, but they inhibit the bacte-ricidal effect mediated by antibodies directed against otherouter membrane antigens (147).

IRPs

Iron-regulated proteins (IRPs) are another important com-ponent of the meningococcal outer membrane. The ability of abacterial pathogen to acquire iron is critical for pathogenesis,and the meningococcus is no exception. An example is theincreased lethality of meningococci in a murine infection

model after the administration of iron-containing compounds(38). The availability of inorganic iron in the human host isextremely low (144). Among the main iron sources are hemo-globin and the iron transport proteins, transferrin and lacto-ferrin (21). The majority of bacterial pathogens use sidero-phores, soluble low-molecular-weight chelating agents that cancompete with host iron-binding proteins and internalize theiron after they bind to membrane receptors (144, 164). N.meningitidis seems incapable of synthesizing siderophores butsequesters iron directly from transferrin (21, 58, 84), lacto-ferrin (164), and hemoglobin (120) by receptor-mediatedprocesses. These receptors are OMPs inducible under iron-limiting conditions, i.e., IRPs.While the lactoferrin binding protein has been identified as

an IRP of 105 kDa specific for human lactoferrin, the trans-ferrin binding proteins (TBPs), also with human specificity (8),are very heterogeneous in both size and antigenicity (58, 84). ATBP1 with a molecular mass of 98 kDa which cannot bindtransferrin after SDS-PAGE has been described, as has aTBP2 of molecular mass 68 to 85 kDa which can bind humantransferrin after SDS-PAGE and electroblotting. Previously, a70- to 71-kDa protein that was not identical to TBP2 (149, 164,182) and a 36- to 37-kDa protein denominated major-IRPwhich also seemed to bind transferrin had been described(144). Recently, a hemoglobin-binding protein exposed on thebacterial surface and which appears not to be serogroupspecific has been identified (120). However, unlike other IRPs,it can bind to both human and bovine proteins, suggesting thatthe heme fraction is the ligand being recognized.

Studies of the antigenicity and immunogenicity of IRPs havemainly been based on the transferrin receptors. Analysis of the70-kDa protein indicates that it might be possible to elicitbactericidal antibodies that are highly strain specific, pointingto the antigenic variability of the exposed epitopes (164). TBP2has considerable molecular heterogeneity, but common anti-genic domains have been shown in TBP2s of N. meningitidisstrains and even N. gonorrhoeae and H. influenzae type b;however, no data regarding how these epitopes are exposed inthe intact organism are available (199). In this context, it hasbeen demonstrated that in vivo the human transferrin receptoris expressed on the meningococcal surface (7). Antibodies thatblock iron acquisition could potentially be protective withoutthe need for complement-dependent bactericidal activity (239).

LOS

Some nonprotein components of the outer membrane ofgram-negative bacteria are glycosylated, thus enabling interac-tion of the organism with the aqueous environment. Theglycolipids of the enterobacteria, which live in a bile acid-richenvironment, are called LPS, the hallmark of gram-negativeorganisms, and consist of three structural domains. The out-ermost one is the 0-specific polysaccharide or 0 antigen (156),which is long, hydrophilic, and neutral. It is made up ofrepetitive oligosaccharide (OS) units and carries the mainserological specificity. It helps to prevent dispersion of thebacterial lipid membranes by the bile acids (82). The 0-specificpolysaccharide is linked to a central OS, which in turn is linkedto a basal glycolipid (lipid A) anchored in the outer membraneand associated with the biological effects of LPS (endotoxin)toxicity that usually accompanies the bacteremia due to gram-negative organisms (156, 168). The glycolipids of Neisseria spp.,along with those of other bacteria that colonize mucosalsurfaces not bathed in bile acids (e.g., in the respiratory andgenital tracts), lack the 0-specific polysaccharide but instead

VOL. 7, 1994


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

562 DIAZ ROMERO AND OUTSCHOORN

Lacto-N-neotetraoseGIcNAc Gal GIc Hep

KDO

A

FIG. 1. Structure of L3 LOS. NeuNAc, sialic acid; Gal, galactose; GlcNAc, N-acetylglucosamine; Glc, glucose; Hep, heptose; KDO,3-deoxy-D-manno-2-octulosonic acid; PEtN, phosphoethanolamine. Adapted from reference 160 with permission of the publisher.

have an OS linked to lipid A that is termed lipooligosaccharide(LOS) (75, 117) (Fig. 1).Twelve to 13 different types of LOS, known as immunotypes,

have been defined among N. meningitidis isolates and charac-terized by using monoclonal antibodies (85, 213). Immuno-types are used to complete the serotyping system describedabove, and the complete serological classification of a menin-gococcal isolate includes the polysaccharide serogroup, theclass 2/3 protein serotype, the class 1 protein subtype, and theLOS immunotype (66). In parallel with this typing system, themeningococci can also be divided into electropherotypes bymultilocus enzyme electrophoresis (36, 179, 192, 226). Electro-pherotype typing examines electrophoretic differences in 10 to15 enzymes required for bacterial growth. The electrophero-types have been grouped in clones and clusters or electro-pherotype complexes (222, 223). This type of analysis hasrevealed that the genetic diversity of N. meningitidis is greaterthan that of any other bacterial species analyzed to date bythis method, although the serogroup A isolates are relativelyhomogeneous. The distribution of serotypes and electro-phoretic patterns of OMPs from different strains seem not tobe related to the clonal genetic structure of the population(36). The most recent method for characterizing meningococ-cal isolates is based on the restriction fragment length poly-morphisms within the PCR products of the porA gene, whichencodes the class 1 OMP (108). Use of this method confirmsthe great diversity displayed by N. meningitidis B isolates.The LOS from Neisseria strains are composed of one to six

components separable by SDS-PAGE; their relative molecular

masses have been estimated to be between 3.15 and 7.10 kDa(111). The heterogeneous electrophoretic mobility of the LOScomponents principally reflects differences in OS chemicalcomposition, which also accounts for the antigenic differences.The OS structures of some strains have now been shown to bebi- or triantennary structures with 2-keto-3-deoxy-D-manno-octulonic acid linked to lipid A at the reducing end and tocontain eight to ten sugar units with various amounts andlocalization of phosphoethanolamine (Fig. 1). At the nonre-ducing end of the longest antenna, eight of the immunotypeshave a structure present in the glycosphingolipids of theparagloboside series which is a precursor of the human bloodgroup antigens: lacto-N-neotetraose (LNnT) (133, 204). Thisstructure is polysialidated in human cell membranes, and themeningococcal species that produce sialic acid (serogroups B,C, W, and Y) also sialidate LNnT of their LOS (83, 160, 232).Three of the LOS immunotypes are prevalent (in decreasingorder): L3,7,9, L2, and Li containing the first two LNnT thathave not been shown to be immunodominant in L3,7,9 (212).Due to progress in organic chemical syntheses, it has beenpossible to prepare different structures that mimic the internalepitopes of these prevalent immunotypes, eliminating LNnT(210).With respect to possible vaccine preparation, only the OS

have generally been considerated adequate for this purposesince lipid A is responsible for inducing septic shock (168, 210,220), although recent preliminary studies have been carriedout with liposome-incorporated LOS (163), thus decreasingtoxicity. For a vaccine, the OS is coupled to protein carriers



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


(53) by conjugation procedures that preserve the phosphoeth-anolamine groups, which are important for OS immunogenic-ity (85, 211). Tetanus toxoid and meningococcal OMPs havebeen used for this purpose (214).The latest studies on OS conjugate vaccines have given

contradictory results (85, 214). By using an OS lacking LNnT,which was derived from a group A N. meningitidis strain andconjugated to tetanus toxoid (85), rabbit antisera with bacte-ricidal activity against the homologous A strain (L8,10,11) andagainst two other strains that share the L8 epitope wereobtained, although bactericidal activity for the last two strainswas low. From another study in which OS from two prevalentimmunotypes, L3,7,9 and L2, were used with different carriers,one can conclude that it would be difficult, if not impossible, toinduce anti-LOS bactericidal antibodies against the most prev-alent immunotypes in a heterogeneous population by using thistype of conjugate (214).

Lip, Pili, and CtrAIn addition to OMPs, IRPs, and LOS, other noncapsular

surface antigens considered possible candidates for a vaccinehave been Lip, the pili, and CtrA. Lip is a surface antigencommon to all pathogenic Neisseria species but absent innonpathogenic ones. This antigen, initially designated H.8(25), is a lipoprotein with a molecular mass of 18 to 30 kDa.Antibodies to Lip recognize both continuous and conforma-tional epitopes, but none have the capacity to promote com-plement-mediated bactericidal activity (26, 203).The pili are superficial appendages that N. meningitidis

organisms express on mucosal surfaces, and they facilitatebinding of the organism to host cells (196, 217, 218). The maincomponents of the meningococcal pili are repetitive polypep-tides called pilins, which contain epitopes common to bothmeningococci and gonococci. These common epitopes are notgenerally immunogenic after infection, probably because theyare inaccessible in the assembled pili of these organisms.Another alternative to the capsular antigen, but closely

related to it, has been identified through studies on expressionand translocation mechanisms of N. meningitidis B capsularpolysaccharide. It is an OMP designated CtrA which formspart of a complex system of capsular transport (70) (seebelow). CtrA has lipoprotein properties, is present in N.meningitidis and not in other Neisseria species (pathogenic orotherwise), and is highly conserved among the strains whosecapsules are composed of homo- or heteropolymeric sialicacid: B, C, W-135, and Y. Analysis of its secondary structurereveals that it is comparable to other OMPs that have porinfunction, and although other data relevant to its antigenicityare lacking, CtrA is under consideration as a potential candi-date for vaccine development.

Clinical Trials: OMYOver the last decade, the efficacy of meningococcal outer

membrane vesicle (OMV)-derived vaccines has been evaluatedin clinical trials. OMVs are liberated into the medium throughevaginations of the bacterial outer membrane during logarith-mic growth (47). These vesicles are depleted of LOS bydetergent treatment and can be combined with the meningo-coccal capsular polysaccharide for improved solubility (65).A vaccine made up of OMPs of the B:15:P1.3 strain,

noncovalently coupled to N. meningitidis C capsular polysac-charide and absorbed with aluminum hydroxide, was tested forefficacy in Iquique, Chile, during 1987 to 1989 (235). Thepreparation was practically free of LOS (0.1%), and theprotein component was present in the form of 250-kDa

multimeric aggregates instead of vesicles. The vaccine inducedmainly nonfunctional antibodies, some of which were directedagainst class 1 OMPs. An insufficient number of individualsdeveloped functional antibodies (trial efficacy was 51%), andthe antibody response did not persist for very long.With a vaccine prepared from a Cuban epidemic strain,

B:4:P1.15, clinical trials were carried out in Cuba during 1987to 1989 (189) and in Sao Paulo, Brazil, from 1989 to 1990 (43).The vaccine components are purified OMPs enriched with acomplex of high-molecular-mass (65 to 95 kDa) proteins whichhave not been characterized (239). The complex is nonco-valently bound to the N. meningitidis C capsular polysaccha-ride, is aluminum hydroxide absorbed, and contains approxi-mately 1% LOS. In the Cuban trial, the global efficacy was80%, while in Sao Paulo, where the case control efficacy studywas carried out for different age groups, this efficacy reached74% in children older than 4 years and decreased dramaticallyin younger ones.

Trials have been undertaken in Norway with an OMV-derived vaccine from strain B:15:P1.7,16 prepared by deoxy-cholate extraction (27, 173). The preparation contained class 1,3, and 5 OMPs absorbed on aluminum hydroxide and otherminor protein components. This vaccine contained 8% LOSand 12% deoxycholate but no substantial amount of meningo-coccal polysaccharide. A large proportion of the bactericidalantibodies stimulated by this vaccine seemed to have beendirected against protein Sc (174), whereas opsonic activity wasinduced against meningococcal strains that were nontypeableand had class 3 OMP (86). The vaccine efficacy among children12 to 16 years of age was only 57%, which was considered toolow for general use of the vaccine (27).A limitation of OMV as a vaccine against N. meningitidis B

is that the protection induced is serotype or subtype specific(214). Although some associations between certain serotypesand subtypes in epidemic outbreaks in particular areas havebeen established, during periods of low incidence the disease-causing strains are heterogeneous and an important percent-age are not serotypeable (18, 107, 205). On the other hand,strains prevalent in a particular geographic area do not neces-sarily coincide with those of another: the combinations ofserotype and subtype antigens 15:P1.16, 4:P1.15, and 2a:P1.2associated with epidemics in northern Europe are not foundamong isolates encountered in a recent study in Greece (205).Even within a region, changes among the prevalent serotypesmay occur (179).

In order to achieve broad protection for a particular geo-graphic area, the vaccine should contain 5 to 10 differentOMPs (214). The problem cannot be resolved by simplycombining different OMV preparations, since this would giverise to high LOS levels per vaccine dose with unacceptablerisks of secondary effects. Recently, recombinant DNA tech-nology has allowed the cloning and sequencing of structuralgenes of certain OMPs (239). Studies have now begun on geneconstructs for multivalent vaccines in which additional genescoding for class 1 OMPs are inserted into N. meningitidisstrains (207). Constructs producing two distinct class 1 proteinshave been made to date, and it now seems possible to increasethe number of inserts. The OMVs of strains with differentcombinations of three class 1 proteins could thus be combinedinto a potential multivalent vaccine against the most prevalentsubtypes.

Other Noncapsular Antigens: IgA Proteases

The immunoglobulin A (IgA) proteases are another cate-gory of noncapsular and nonsurface antigens that have been

VOL. 7, 1994


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


proposed as potential candidates for a vaccine that couldprotect not only against N. meningitidis B infection but alsoagainst other serogroups of the species and even other speciesthat produce this protease (74, 88a, 129). The IgA proteaseshave been identified as extracellular enzymes from pathogenicbacteria: N. gonorrhoeae, N. meningitidis, H. influenzae, S.pneumoniae, Streptococcus sanguis, and Streptococcus mitior(109, 114). All of these organisms, although diverse andtaxonomically unrelated, are implicated either in initiatingdisease via the mucous membranes, where IgA is present, or inthe onset of dental plaque. IgA proteases have been identifiedonly in species of the above genera generally considered to behuman pathogens. They share common properties, being metalion dependent, extracellular, proteolytic, and autoproteolyticmicrobial enzymes acting optimally at neutral pH. Their onlyknown substrate is IgAl, the main IgA isotype present inserum and upper respiratory tract secretions of humans (33,109), and gorilla and chimpanzee IgA (40, 171, 187). The mainisotype of colorectal secretions, IgA2, is resistant to the actionof these IgA proteases.IgA proteases break the only peptide bond in a duplicated

octapeptide found in the hinge region of the IgAl heavy chain,Thr-Pro-Pro-Thr-Pro-Ser-Pro-Ser (146), liberating intact mo-nomeric Faba fragments that totally retain their antigenbinding capacity and Fc or Fc-secretory component (109). N.meningitidis produces two mutually exclusive enzyme typeswithin the same strain: type 1, which breaks the a chainbetween Pro-237 and Ser-238; and type 2, which breaks thechain between Pro-235 and Thr-236. Although most serogroupA strains produce type 1 proteases and those of serogroups Xand Y produce type 2, in any particular serogroup, somestrains produce type 1 and others produce type 2. The majorityof the epidemic-associated meningococcal strains produce type1 (129).The specific roles played by IgAl and the IgA proteases in

the pathogenesis of mucosal infections are not yet clear. Thesecretory immunoglobulins (IgA) still act as a first line of hostmucosal defense, interfering with cell adhesion and microbialcolonization (87). IgA has antibody-dependent antibactericidalactivity mediated by monocytes in the absence of addedcomplement (130, 157). Through this Fca-dependent mecha-nism, the ability of local monocytes and macrophages to limitbacterial multiplication in mucosal tissues may be promoted,thereby decreasing the opportunity for systemic invasion bypathogenic strains. On the other hand, an excretory functionhas also been attributed to IgA (139, 140). Active transport ofIgA immune complexes by epithelial mucosal cells via transcy-tosis from the basal surface to the apex is mediated by thepolymeric immunoglobulin receptor whose outer domainmakes up the secretory component (116, 145). This seems animportant and efficient mechanism of antigen exclusion fromthe systemic circulation (104) and could also operate toexclude different pathogens, including capsulated bacteria(Fig. 2). IgA rupture in the hinge region could interfere withboth Fc and polymeric immunoglobulin receptor function. TheFaba and Fca. fragments resulting from proteolysis of secre-tory IgAl by proteases would be incapable of inducing bacte-rial agglutination and inhibiting bacterial adherence (33). Inaddition, the Faba fragments, through their linkage to bacte-rial surface epitopes, could protect bacteria by blocking theiraccess to intact antibody molecules.

Little is known regarding the immunogenicity and antigenic-ity of the IgA proteases. Some data exist suggesting thatpatients with infections caused by Neisseria spp. produceantibodies to IgA proteases, and a study in The Gambia duringa meningococcal epidemic (34) clearly showed such an anti-

body response during clinical disease or on nasopharyngealtransport. Polyclonal rabbit antibodies could recognize pro-teases of different gonococcal and meningococcal strains (29)and could inhibit enzyme activity of all meningococcal pro-teases irrespective of their type (129), a potentially significantfinding for the possible application of IgA proteases in aprotective vaccine.

THE CAPSULE

Presence of Polya2-8 in Bacterial and Animal Cells

One of the commonest procaryotic outer cell surfaces is thecapsule, a very highly hydrated structure that can protect thecell against desiccation and form an outer barrier to bacterio-phages, toxins, and antibacterial agents (81). The physico-chemical characteristics of the capsule constitute a clear ad-vantage for capsulated cells in their natural environment (23).The majority of gram-negative bacterial capsules are made upof negatively charged hydrophilic polysaccharides (24). TheN-acetyl neuraminic acid (NeuNAc) forming a part of differentpolymers is a structure present in the capsule of severalbacteria (46). Apart from serogroup B, it is present in groupsC, Y, and W135 of N. meningitidis, in E. coli K9 and K92, andin types I, II, and III of the group B streptococci. The N.meningitidis B capsule is made up of a 200-residue homopoly-mer of NeuNAc in a(2-8) linkage (polyox2-8NeuNAc) (88, 125,162). This structure is both chemically and immunologicallyindistinguishable from that encountered in the capsule of E.coli Kl (105), an important cause of meningitis in newborns(30, 112, 169, 183). It is also present in Pasteurella haemolyticaserotype A2, an important veterinary pathogen (4), as well asin Moraxella nonliquefaciens, a normally nonpathogenic com-mensal recovered in about 17% of human nasal specimens (32,46). Vaccines based on polya2-8NeuNAc would have theunique advantage of being based on an epitope present in allN. meningitidis B and E. coli Kl strains (94).

This structure is also present on the surface of animal cells,where it has been proposed to act as a plasticity marker (177).The neural cellular adhesion molecule (NCAM) is a mem-brane glycoprotein that can promote cell-cell adhesion via ahomophilic binding mechanism (178). The molecule is presentin various forms differing in carbohydrate content. A highlysialidated form which is prevalent in embryonic and newbornbrains exists and is gradually replaced by a less sialidated formduring development. Antibodies against the N. meningitidis Bcapsule also recognized the highly sialidated form of NCAM(60, 66).One of the models proposed to explain the characteristic

adhesion-regulatory properties of polya2-8NeuNAc is that ofoccurrence through simple physical steric hindrance mecha-nisms and by charge repulsion, which regulates the intercellu-lar space. Polyct2-8NeuNAc will increase the space betweendifferent cell membranes, diminishing the levels of bindingbetween their receptors (1, 178, 233). Changes in the length ofthe polya2-8NeuNAc in NCAM modulate the adhesion prop-erties of this molecule during embryonic development (176).Another model used to explain the adhesion properties ofpolya2-8NeuNAc is one of ligand recognition and participa-tion in the transmission of intracellular signals (177, 181). It isof,interest to emphasize here the hypothesis that specific sialicpolymer binding sites are present in the brain (112). This couldexplain the bacterial tropism toward this organ when thisstructure is present in the capsule.

Polysialidated NCAM is also present in the developing



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


Ning

Dlmerc 1A

/.

0

FIG. 2. Hypothetical N. meningitidis (Nmg) excretion by specific IgA. (A) Dimeric IgA transport from the lamina propria to the lumenmediated by the polymeric Ig receptor. (B) Mucosal epithelium adhesion and absorption inhibition by N. meningitidis IgA. (C) IgA-mediated N.meningitidis excretion. Partially based on reference 140, with permission of the publisher.

kidney and in Wilms tumor, an embryonic type of kidneytumor (118, 176). Research on a wide range of tumorsindicates that many tumors of neuroendocrine origin expresspolysialidated NCAM. Among these are pheochromocytoma,small-cell carcinoma of the lung, neuroblastoma, rhabdomyo-sarcoma, pituitary tumors, teratomas, and medullary carci-noma of the thyroid (151, 176). However, it is difficult to assigna general role for polyax2-8NeuNAc as having both invasiveand metastatic tumor potential since many highly invasivehuman tumors which do not exhibit this surface marker exist,although the marker is also present in certain normal adulttissues. Polya2-8NeuNAc seems to be widely distributed in theadult rat brain, as much associated with NCAM as with thevoltage-dependent sodium channel a-subunit (241). Neverthe-less, under certain conditions, polya2-8NeuNAc together witha variety of other factors could contribute to potentiateinvasive growth and tumor metastasis (176).The view that a polya2-8NeuNAc-based vaccine could ini-

tiate autoimmune processes due to the presence of this struc-ture in NCAM (61) is not borne out by fact. Although severalcomplications have been reported to occur during recoveryfrom meningitis, including arthritis, pericarditis, and a case ofconus medullaris syndrome, these complications are thought to

result from the host immune response to the meningococcus(197), probably directed against a meningococcal stress proteinthat can induce antibodies cross-reactive with host determi-nants (16, 158). Vaccines containing serotype proteins andpolya2-8NeuNAc have been tested in animal models (127) andin humans, both children and adults (236), without harmfuleffects. Animals hyperimmune to N. meningitidis B do not showanomalous symptoms in spite of achieving high levels ofpolya2-8NeuNAc-specific antibody (9, 30). The absence ofdetectable autoimmune disease in a patient having a monoclo-nal gammopathy with extremely high levels of IgM specific forpolyox2-8NeuNAc, and which cross-reacted with denaturedDNA and polynucleotides (73, 102, 103), is another argumentin favor of anticapsular vaccine developmental efforts.

Molecular Mechanisms of Expression

The capsules of N. meningitidis B, E. coli Kl, and H.influenzae type b, the predominant serotypes associated withmeningitis and sepsis in newborns and young children, sharecommon genetic and biochemical properties; therefore, theyhave been classified in the same group of capsular polysaccha-rides, group II (31, 68).

VOL. 7, 1994


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


Capsular 0polysaccharide 0

CtrA

Outermembrane _J&

CtrB CtrC

Inner Q;?5membrane J jS

c" -00 --0O 00(MP-Ne.NAc fl8CtrD CtrD

NeuNAc UP{NeuNAc)m PL.NeuNAc)n

FIG. 3. Model for expression of capsular polysaccharides in N.meningitidis B. PEP, phosphoenolpyruvate. See text for other designa-tions. To standardize nomenclature, the following changes were madefor the polysialic acid biosynthesis gene products: ConB (correspond-ing to the NeuNAc condensing enzyme) was replaced by SiaC, SynB(corresponding to the CMP-NeuNAc synthetase) was replaced bySiaB, and SiaB (corresponding to the polyat2,8 sialyltransferase is nowtermed SiaD (55a). Adapted from reference 67 with permission of thepublisher.

The genes required for the biosynthesis and expression ofthe N. meningitidis B capsule are localized in a 24-kb chromo-somal moiety termed the meningococcus capsule gene complex(72). Five different regions can be defined within the capsulegene complex: A, B, C, D, and E. The genes that code for allof the enzymes needed for polymer synthesis are located in theA region (67, 208). This expresses four different proteins withmolecular masses of 38, 18.2, 38.3, and 54.4 kDa. The 38.3-kDaprotein is a condensing enzyme (ConB, now SiaC) that syn-thesizes the polysaccharide monomer by N-acetylman-nosamine (ManNAc) condensation with phosphoenolpyru-vate, while the 18.2-kDa protein has synthetase activity (SynB,now SiaB) transferring NeuNAc molecules to the activatedstate: CMP-NeuNAc. Both enzymes act within the cytoplasm(Fig. 3). In contrast, the polysialyl transferase activity (SiaB,now SiaD [55a]) assigned to the 54.4-kDa protein, whichpolymerizes CMP-NeuNAc to form polyoa2-8NeuNAc, is asso-ciated with the bacterial inner membrane. In addition, unde-caprenol has been described as the acceptor molecule neededfor polymerization (137). No enzymatic activity has beenassigned to the 38-kDa protein.The B and C regions of the capsule gene complex are

implicated in transport (67, 68). The capsules of gram-negativebacteria appear to be bound to a lipidic moiety which probablyacts as a hydrophobic anchor in the outer bacterial membrane(193). In N. meningitidis B, a phospholipid covalently bound tothe reducing end has this function (79, 186a). This phospho-lipid presumably acts also to anchor the polysaccharide to theinner membrane (67), a process considered fundamental intranslocation, and its incorporation into the polysaccharideseems to be mediated by the B region. This region codes fortwo proteins with molecular masses of 45.1 and 48.7 kDadesignated LipA and LipB according to their putative functionof phospholipid incorporation (71).The C region directs the transport process itself (67). This

region codes for four proteins with molecular masses of 41, 42,30, and 25 kDa: CtrA, CtrB, CtrC, and CtrD, respectively.

CtrA is located in the outer membrane and has a structuresimilar to that present in other proteins that have porinfunction (70). It has hence been proposed that CtrA may forma pore in the outer membrane through which the maturecapsular polysaccharide can pass (67). CtrB and CtrC are innermembrane proteins (67, 68), and it seems unlikely that theyform a pore in the inner membrane through which polysaccha-ride can pass from the cytoplasm to the periplasm. Rather,these proteins could work by an active transport mechanismsuch as the "flippase" (71, 90).

Regulatory functions in capsule expression have been as-signed to the D and E regions (67). Preliminary evidenceindicates that these regions may participate in LOS biosynthe-sis and assembly.

Capsule Structure

In contrast to its chemical composition and biological activ-ity, little is known regarding the formation and structuralorganization of the assembled macromolecules that make upthe capsule in live bacteria. This lack of understanding is duemainly to the biophysical properties of the capsule which havegiven rise to a series of technical drawbacks in its study. Morethan 95% of the capsule is water (23), and this hydrated naturemakes it difficult to examine capsules by electron microscopy.Upon embedding and dehydration, the capsular polymerscollapse easily. Various stabilizing procedures have been de-vised to avoid this structural collapse. Among the most widelyused are stabilization via cationic charges, the use of anticap-sular antibodies, and cryofixation followed by freeze substitu-tion.

Cryofixation instantly freezes the capsular polymers in theirfully hydrated and extended state, and during substitution,water is slowly replaced by organic solvent (81). By thistechnique, the capsule of E. coli Ki has been observed toconsist of two layers, an inner and an outer one (10). Thefibrous material of the inner layer forms a dense stratum whichis probably made up of a series of thick bundles set perpen-dicular to the outer membrane surface. The outer layer ismade up of a netlike assembly of randomly oriented fine fibers.The complete capsule is about 10 nm thick. This thickness isnotably less than that of other capsules: e.g., Klebsiella pneu-moniae, with a similar layout, is about 160 nm thick (10).However when the E. coli Kl capsule was stabilized withantibodies, much higher values for the capsular thickness wereobtained, i.e., >300 nm (115). The phenomenon of capsularswelling with specific antibody treatment has long been known(23, 141), but it is difficult to explain this 30-fold differencesolely in these terms. At present, no electron microscopic dataexist regarding the N. meningitidis B capsule, and possibleextrapolation of E. coli Kl data should be made with caution.The capsule in vivo is heterogeneous due to its capacity to

absorb a variety of ions and macromolecules. The influence ofdivalent cations in the recognition of polyoa2-8NeuNAc byspecific antibodies can be shown by hemagglutination assays,using purified polya2-8NeuNAc as a sensitizer. The addition ofCa2+ (0.1 M CaCl2) to the assay buffer can block the hemag-glutination of erythrocytes sensitized with a polyot2-8NeuNAc-specific monoclonal antibody (52a). The influence of divalentcations in epitope expression has also been shown in LOS (82).Although the capsule seems to be a uniform structural

entity, it is probable that a mixture of capsular material andLPS is produced (23). In fact, the formation of 0 antigen-capsular polysaccharide complexes in E. coli KI by noncova-lent interactions has been described (153). It has also beenshown that Ca2' and Mg2+ salts of LPS form extended



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


NUCLEIC ACIDSICULTURE AGE

PROTEINSLOS

5 4 3 2 1 0

% CONTAMINANTS

lillRelative inhibitory power

IgM anticapsular antibodies or IgG specific for noncapsularantigens can increase alternative pathway function, althoughby different mechanisms. While the binding of IgG to noncap-sular antigens increases both C3 and factor B deposition on thebacterial surface, IgM promotes lysis independent of thequantities of C3 and factor B deposited and hence is moreeffective than IgG (93). One hypothesis to explain this differ-ence is that IgM can alter both charge and tertiary structure ofthe capsule and thus expose sites previously protected fromlysis or IgM can redirect deposition of complement fromnonlytic to lytic sites.

FIG. 4. Correlation between nucleic acid content of purified polya2-8NeuNAc and relative inhibitory power in an hemagglutination inhi-bition assay.

ribbonlike structures (106), which could constitute the backingover which dense fibers of the inner layer of the E. coli Klcapsule are laid down. In addition, during purification ofcapsular material, polysaccharide and nucleic acid were ob-served in close association (23). By using capsular polysaccha-ride purified from N. meningitidis B cultures at different growthtimes (48) to sensitize erythrocytes, a correlation was estab-lished between the nucleic acid content of purified polysaccha-ride and the capacity to inhibit hemagglutination by an anti-polya2-8NeuNAc horse polyclonal serum (9, 52a) (Fig. 4). Ithas long been known that RNA, being an acidic molecule, canbind noncovalently to other acidic molecules via divalentcations to form stable complexes (78). On the other hand,high-molecular-weight polyphosphates (200 residues long) arefound closely bound to the surface of different Neisseria species(202). Hence, nucleic acids as well as polyphosphate, which hashigh affinity for divalent and trivalent cations, could act ascapsular backing through interactions with polya2-8NeuNAcand mediated by the divalent cations.

Barrier Function

Evidence obtained from studies with other bacteria indicatethat the capsule may not function as a barrier in immunereactions with the cell surface (141), but this is an improbablegeneralization. For some organisms, the capsular polysaccha-ride seems to function as a physical barrier whose presencecould impede access by phagocytes to opsonic fragmentsbound to subcapsular structures (135). In fact, there is muchevidence to suggest that the E. coli Kl and N. meningitidis Bcapsules play such a role.The E. coli Ki capsule seems to block both agglutination by

anti-O antibodies (198, 216) and direct binding of antibodies tothe internal LPS core (76, 188). In N. meningitidis B, thecapsule blocks agglutination by wheat germ agglutinin, which isspecific for N-acetylglucosamine present in LOS (62). In N.meningitidis B isolates from cerebrospinal fluid, the epitope ofthe OMP P3.15 is observed to be poorly marked by specificantibodies on electron microscopic examination (221). Thissuggests steric hindrance by other surface molecules, thecapsule being an obvious candidate.

However, the best evidence for barrier function comes fromthe capsule's inherent capacity to block the alternative com-plement pathway. Many studies have shown that the N.meningitidis B capsule, like the E. coli Kl capsule, is animportant virulence factor (5, 112, 138). In the absence ofantibodies, the ability of N. meningitidis B to activate thealternative pathway is a function of its cell-associated sialicacid capsule, which impedes bactericidal activity (93). Specific

Role of Anticapsular Antibodies in Protectionagainst Infection

It has been postulated that antibodies to the N. meningitidisB capsule will not be protective in humans because they do notinduce bactericidal activity in the presence of homologouscomplement (237, 239). In contrast, when passive protectionmodels are employed, mouse monoclonal antibodies to thecapsule are clearly protective against N. meningitidis B infec-tion (122, 169). This is so even though recent experimentsusing a complement fixation microassay (51) show that mousemonoclonal antibodies fail to lyse polya2-8NeuNAc-sensitizederythrocytes in the presence of homologous complement (52a).However, in a chicken embryo model, absorption with polya2-8NeuNAc of antisera prepared by hyperimmunization withdifferent N. meningitidis B serotypes completely eliminates theprotective activity (63). In view of these data, what mecha-nism(s) may be implicated in protection by anticapsular anti-bodies?

First, we should consider the role of anticapsular antibodiesin N. meningitidis B phagocytosis. The relative importance ofserum bactericidal activity in relation to neutrophil-mediatedmeningococcal destruction has not been precisely defined (123,175). It has been shown that nonimmune human serum has thecapacity to opsonize N. meningitidis B cells, which in turn arephagocytized by polymorphonuclear leukocytes (175, 225).Both antibodies and complement are needed for this process,and although the specificity of the antibodies has not beenconclusively demonstrated, data exist that point directly to thecapsule. The need for activation of the classical pathway forboth opsonization and phagocytosis, as well as the demonstra-tion of C3 deposition on the capsular surface (225), suggeststhat the C3 deposition is mediated by anticapsular antibodyactivation of the classical pathway. In addition, the efficiencyand kinetics of opsonization are similar in different strainsindependent of serotype (225), and the presence of low titersof anticapsular antibodies in most adults has been described(121, 124, 237). This mechanism, dependent on the classicalpathway and phagocytes, can function at low antibody titersand is probably sufficient to prevent dissemination of theinfectious inoculum.Above a critical threshold of anticapsular antibodies, a

second mechanism is brought into play: subcapsular C3 depo-sition dependent on anticapsular antibody. This deposition ismediated by the alternative pathway due to the effect describedabove of inactivation of the anticomplementary capacity of thecapsule by anticapsular antibodies (93). The resulting activa-tion of the membrane attack complex would lead to bacteriallysis. This bimodal behavior to different antibody concentra-tions has been described for other capsulated bacteria thathave anticomplementary surface properties (55).Yet a third mechanism could exist by which anticapsular

antibodies could play a protective role for N. meningitidis B.The same mechanism that enables subcapsular C3 deposition

4I848b

16b

VOL. 7, 1994

a


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


by the alternative pathway could facilitate the access of anti-bodies directed against habitually masked subcapsular anti-gens. Thus, antibodies could exert bactericidal action via theclassical pathway, and the synergistic effect of the protectivefunction of capsular and noncapsular antibody in E. coli Kland N. meningitidis B could be explained (63, 113).

Specific Antibody Recognition of Polya2-8NeuNAc

An obstacle to the study of the polya2-8NeuNAc-anti-polyoa2-8NeuNAc interactions has been the difficulty in obtain-ing specific antisera and monoclonal antibodies (69, 148),although at present IgM (142, 148, 169, 237) and IgG (69, 194)monoclonal antibodies are both available. The traditionalconcept of antibody recognition of polysaccharide antigenswhereby only six or seven residues are needed to form theantibody binding epitope (14, 101, 155) is not in agreementwith known data regarding polyot2-8NeuNAc-anti-polyot2-8NeuNAc interactions (89, 99, 124).The conformation of a linear homopolymer of a single sugar

joined by a single type of glycosidic linkage is determined bythe linkage which controls the position of each sugar ring inrelation to the adjacent bound residues (167). The repetitivenature of the linkage pattern results in the regular, periodicstructure of the polymer. Many studies have been carried outto try to determine the three-dimensional structure of polyot2-8NeuNAc by nuclear magnetic resonance. There are studiesthat suggest that the polymer adopts a permanent helicalstructure in solution (231). Others, however, postulate that thepredominant conformation of polysaccharide in solution is arandom coil with only the occasional presence of an extendedhelical structure (35), the epitope recognized by anti-polyct2-8NeuNAc antibodies.

Smaller OS of the same repetitive structure can completelyinhibit antibody binding to the native polysaccharide at highconcentrations (99, 166, 194), and it is even possible to inhibitthis binding with NeuNAc (49, 166), indicating that the sameantibodies bind to antigens smaller than the native polysaccha-ride, although with lower affinity. The presence of conforma-tional epitopes could explain how a specific antibody responseto a polysaccharide with repetitive structural units identical tohost OS could be obtained even though cross-reactions be-tween antipolysaccharide antibodies and host antigens are notproduced (229). This situation occurs with sialogangliosidesand sialoglycoproteins containing short chains of NeuNAcbound by a(2-8) linkages that are ubiquitous on host cellsurfaces (124), but this explanation does not eliminate thepossibility of cross-reactivity with the long-chain polya2-8Neu-NAc present in NCAM in vivo.

In addition to the conformational requirements, anothertype of restriction exists in the interaction between polya2-8NeuNAc and its specific antibodies, i.e., temperature. Anti-polyot2-8NeuNAc antibodies show markedly reduced aviditywhen the temperature is increased from 4 to 37°C (134). Thisdifference in avidity is still considerable for reactions attemperatures over a shorter range, e.g., between 22 and 370C(184). The decrease in avidity at 37°C can be translated into anantigen concentration dependence on antibody binding (50).Only surfaces with a high polyoa2-8NeuNAc density, like N.meningitidis B or E. coli Kl capsules, are recognized by thespecific antibody at 37°C, whereas in vivo, host cells with thisstructure do not fulfill these recognition requirements. Thepolya2-8NeuNAc of NCAM could, however, be recognized byin vitro assays carried out at room temperature or 4°C (1, 60,61).

2.5 -

2 *

o 1.5-

0.5

5 4 3 2-Log serum dilution

FIG. 5. Total (i), IgM (*), and IgG (X) responses to capsularpolysaccharide of N. meningitidis B as measured by enzyme-linkedimmunosorbent assay (49). For controls (+, El, and 0), wells coatedwith bovine serum albumin were used.

Strategies To Increase Response

In general, the immune response to purified polysaccharideis thymus independent; IgM antibodies are produced, and noincreased response is observed on hyperimmunization (53,124). In contrast to the majority of capsulated bacteria, E. coliKi and N. meningitidis B capsules are poorly immunogenicwhether administered as formalin-inactivated organisms or aspurified polysaccharide (51, 154). It is possible, however, toobtain high-titer anticapsular antisera in horses by repeatedimmunization with formalinized N. meningitidis B (9).The immune response to a polysaccharide can also be

increased by injecting it (or its constituent OS) coupledcovalently to protein carriers (53, 124, 132). This processgenerally transforms the antigen from a thymus-independentto a thymus-dependent one, resulting in a stronger responsethan to the polysaccharide alone. It also gives rise to IgGantibodies and memory cells, as has been demonstrated for themeningococcal group A and C capsular polysaccharides (41),H. influenzae type b (12), pneumococcus (161, 215), or thegroup B streptococcus (119, 159).

Tetanus toxoid has been among the most frequent proteincarriers used to prepare polyao2-8NeuNAc conjugates withwhole polysaccharide (97) or derived OS (165). In spite of thisapproach to convert the N. meningitidis A and C conjugatesinto T-dependent immunogens and thus to obtain high titers ofspecific antipolysaccharide antibodies, the immunogenicity ofpolya2-8NeuNAc was not increased.The optimal carriers for increasing the anticapsular response

in conjunction with unmodified polya2-8NeuNAc are theOMPs, which form a noncovalent complex with the polya2-8NeuNAc probably across the lipid moiety of the polysaccha-ride (143). This is the case even though different couplingmethods can give rise to a variable polysaccharide response(64, 127, 238). Although it is possible to obtain high titers ofspecific polya2-8NeuNAc antibodies in mice immunized withthe OMP complex, the antibody is restricted to the IgM isotype(49, 126, 143) (Fig. 5). The complex has been shown to protectmice (143) against N. meningitidis B and newborn rabbitswhose mothers were immunized with complex against E. coliKl (123). In humans, the anti-polya.2-8NeuNAc antibodiesobtained with the complex were biologically active in a mousepassive protection model (126).Another strategy to improve the immunogenicity of polyao2-

8NeuNAc has been to chemically modify it by replacing the



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


N-acetyl groups of the NeuNAc residues with N-propionylgroups, leaving the specific antigenicity unaltered (98). Al-though the modified polysaccharide is still poorly immuno-genic, its conjugation to tetanus toxoid gives rise to muchhigher titers of anticapsular antibodies in mice than thehomologous conjugate with unmodified polysaccharide (94).The response obtained with this conjugate includes (i) IgGantibodies of the IgGl isotype, which bound to purifiedpolyNeuNAcoa2-8 and were not bactericidal; and (ii) a secondantibody population of the IgG2a and IgG2b isotypes, whichwere both bactericidal and efficient in mouse passive protec-tion tests and which bound polya2-8NeuNAc only when it wascell associated or coupled via a spacer arm to an immunoad-sorbent column (20, 95, 96).A final approach to elicit anticapsular antibodies against N.

meningitidis B is to use polysaccharides with similar structures.Cross-reactivity between N. meningitidis and N. lactamica ledto the proposal that antigenically similar capsules existed inboth organisms (136). However, further studies invalidated thishypothesis and demonstrated that the cross-reactivity wasderived from an epitope present in the LOS (110). The mostpromising work in this direction has been carried out with E.coli K92. Its capsule is a homopolymer of NeuNAc linkedalternately a(2-8) and a(2-9) (128), and this structure has beenused before to induce antibodies to the N. meningitidis Ccapsule (77). The use of E. coli K92 polysaccharide-tetanustoxoid conjugates has been demonstrated to induce anti-polya2-8NeuNAc as well as anti-polyot2-9NeuNAc antibodiesof IgG isotypes (184). This cross-reactivity has the addedadvantage of being able to potentially constitute a protectivevaccine against N. meningitidis B and C and E. coli Ki in-fections.

CONCLUSIONS

The search for an ideal antigen as a potential candidate foran N. meningitidis B vaccine has made the outer surface of thisorganism one of the most widely studied cell coats. Not onlyhave the results advanced our understanding of the differentmechanisms of associated virulence and pathogenicity, but theresearch techniques developed have often been applicable tothe development of vaccines that protect against other capsu-lated bacterial pathogens. Hence, one of the H. influenzae typeb vaccines consists of a capsular polysaccharide conjugate tomeningococcal OMPs that are used for their special immuno-genic properties (11, 54).

For an antigen to be considered a vaccine candidate for theentire serogroup B, two fundamental requirements must befulfilled: it must be highly conserved among the differentstrains, and it must induce bactericidal antibodies. The latter isof particular importance in protection against N. meningitidis(191). To date, it does not seem that these characteristics canbe found in a single candidate vaccine, so a multifacetedapproach to the problem is needed in order to meet this goal.

ACKNOWLEDGMENTS

We thank Elena Primo for help with bibliographic and databasefacilities of the Instituto Carlos III libraries.

REFERENCES1. Acheson, A., J. L. Sunshine, and U. Rutishauser. 1991. NCAM

polysialic acid can regulate both cell-cell and cell-substrateinteractions. J. Cell Biol. 114:143-153.

2. Achtman, M. 1991. Clonal properties of meningococci fromepidemic meningitis. Trans. R. Soc. Trop. Med. Hyg. 85(Suppl.1):24-31.

3. Achtman, M., M. Neibert, B. A. Crowe, W. Strittmatter, B.Kusecek, E. Weyse, M. J. Walsh, B. Slawig, G. Morelli, A. Moll,and M. Blake. 1988. Purification and characterization of eightclass 5 outer membrane protein variants from a clone of Neisseriameningitidis serogroup A. J. Exp. Med. 168:507-525.

4. Adlam, C., J. M. Knights, A. Mugridge, J. M. Williams, and J. C.Lindon. 1987. Production of colominic acid by Pasteurella hae-molytica serotype A2 organisms. FEMS Microbiol. Lett. 42:23-25.

5. Ahmed, H., and H. J. Gabius. 1989. Purification and properties ofa Ca2+-independent sialic acid-binding lectin from human pla-centa with preferential affinity to 0-acetylsialic acids. J. Biol.Chem. 264:18673-18678.

6. Aho, E. L. 1989. Molecular biology of class 5 outer membraneproteins of Neisseria meningitidis. Microb. Pathog. 7:249-253.

7. Ala'Aldeen, D. A., N. B. Powell, R. A. Wall, and S. P. Borriello.1993. Localization of the meningococcal receptors for humantransferrin. Infect. Immun. 61:751-759.

8. Alcantara, J., R. H. Yu, and A. B. Schryvers. 1993. The region ofhuman transferrin involved in binding to bacterial transferrinreceptors is localized in the C-lobe. Mol. Microbiol. 8:1135-1143.

9. Allen, P. Z., M. Glode, R. Schneerson, and J. B. Robbins. 1982.Identification of immunoglobulin heavy-chain isotypes of specificantibodies of horse 46 group B meningococcal antiserum. J. Clin.Microbiol. 15:324-329.

10. Amako, K., Y. Meno, and A. Takade. 1988. Fine structures of thecapsules of Kiebsiella pneumoniae and Escherichia coli Kl. J.Bacteriol. 170:4960-4962.

11. Ambrosino, D. M., D. Bolon, H. Collard, R. Van Etten, M. V.Kanchana, and R. W. Finberg. 1992. Effect of Haemophilusinfluenzae polysaccharide outer membrane protein complex con-jugate vaccine on macrophages. J. Immunol. 149:3978-3983.

12. Ambrosino, D. M., S. K. Sood, M. C. Lee, D. Chen, H. R. Collard,D. L. Bolon, C. Johnson, and R. S. Daum. 1992. IgGl, IgG2 andIgM responses to two Haemophilus influenzae type b conjugatevaccines in young infants. Pediatr. Infect. Dis. J. 11:855-859.

13. Andreoni, J., H. Kayhty, and P. Densen. 1993. Vaccination andthe role of capsular polysaccharide antibody in prevention ofrecurrent meningococcal disease in the late complement compo-nent-deficient individuals. J. Infect. Dis. 168:227-231.

14. Arakatsu, Y., G. Ashwell, and E. A. Kabat. 1966. Immunochemi-cal studies on dextrans. V. Specificity and cross-reactivity withdextrans of the antibodies formed in rabbits to isomaltonic andisomaltotrionic acids coupled to bovine serum albumin. J. Immu-nol. 97:858-866.

15. Arakere, G., and C. E. Frasch. 1991. Specificity of antibodies toO-acetyl-positive and O-acetyl-negative group C meningococcalpolysaccharides in sera from vaccinees and carriers. Infect.Immun. 59:4349-4356.

16. Arakere, G., M. Kessel, N. Nguyen, and C. E. Frasch. 1993.Characterization of a stress protein from group B Neisseriameningitidis. J. Bacteriol. 175:3664-3668.

17. Arko, R J. 1989. Animal models from pathogenic Neisseniaspecies. Clin. Microbiol. Rev. 2:S56-S59.

18. Ashton, F. E., L. Mancino, A. J. Ryan, J. T. Poolman, H.Abdillahi, and W. D. Zollinger. 1991. Serotypes and subtypes ofNeisseria meningitidis serogroup B strains associated with menin-gococcal disease in Canada, 1977-1989. Can. J. Microbiol. 37:613-617.

19. Ashton, F. E., A. Ryan, B. Diena, and H. J. Jennings. 1983. A newserogroup (L) of Neissena meningitidis. J. Clin. Microbiol. 17:722-727.

20. Ashton, F. E., J. A. Ryan, F. Michon, and H. J. Jennings. 1989.Protective efficacy of mouse serum to the N-propionyl derivativeof meningococcal group B polysaccharide. Microb. Pathog.6:455-458.

21. Banerjee Bhatnagar, N., and C. E. Frasch. 1990. Expression ofNeisseria meningitidis iron-regulated outer membrane proteins,including a 70-kilodalton transferrin receptor, and their potentialfor use as vaccines. Infect. Immun. 58:2875-2881.

22. Barington, T., C. Heilmann, and V. Andersen. 1990. Quantitationof antibody-secreting cells in the blood after vaccination withHaemophilus influenzae type b conjugate vaccine. Scand. J.

VOL. 7, 1994


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


Immunol. 31:515-522.23. Bayer, M. E. 1990. Visualization of the bacterial polysaccharide

capsule. Curr. Top. Microbiol. Immunol. 150:129-157.24. Bayer, M. E., and H. Thurow. 1977. Polysaccharide capsule of

Escherichia coli: microscope study of its size, structure, and sitesof synthesis. J. Bacteriol. 130:911-936.

25. Bhattacharjee, A. K., E. E. Moran, J. S. Ray, and W. D. Zollinger.1988. Purification and characterization of H.8 antigen from groupB Neisseria meningitidis. Infect. Immun. 56:773-778.

26. Bhattacharjee, A. K., E. E. Moran, and W. D. Zollinger. 1990.Antibodies to meningococcal H.8 (Lip) antigen fail to showbactericidal activity. Can. J. Microbiol. 36:117-122.

27. Bjune, G., E. A. H0iby, J. K. Gr0nnesby, 0. Arnesen, J. H.Fredriksen, A. Halstensen, E. Holten, A.-K. Lindbak, H. N0kleby,E. Rosenqvist, L. K. Solberg, 0. Closs, J. Eng, L. 0. Fr0holm, A.Lystad, L. S. Bakketeig, and B. Hareide. 1991. Effect of outermembrane vesicle vaccine against group B meningococcal diseasein Norway. Lancet 338:1093-1096.

28. Blackwell, C. C., K. Jonsdottir, D. M. Weir, M. F. Hanson, K. A.Cartwright, J. Stewart, D. Jones, and I. Mohammed. 1989. Bloodgroup, secretor status and susceptibility to bacterial meningitis.FEMS Microbiol. Immunol. 1:351-356.

29. Blake, M. S., and C. Eastby. 1991. Studies on the gonococcalIgAl protease. II. Improved methods of enzyme purification andproduction of monoclonal antibodies to the enzyme. J. Immunol.Methods 144:215-221.

30. Bortolussi, R., and P. Ferrieri. 1980. Protection against Esche-richia coli Kl infection in newborn rats by antibody to Klcapsular polysaccharide. Infect. Immun. 28:111-117.

31. Boulnois, G. J., and K. Jann. 1989. Bacterial polysaccharidecapsule synthesis, export and evolution of structural diversity.Mol. Microbiol. 3:1819-1823.

32. Bovre, K., K Bryn, 0. Closs, N. Hagen, and L. 0. Froholm. 1983.Surface polysaccharide of Moraxella non-liquefaciens identical toNeisseria meningitidis group B capsular polysaccharide. A chem-ical and immunological investigation. NIPH (Natl. Inst. PublicHealth) Ann. (Oslo) 6:65-73.

33. Brandtzaeg, P. 1992. Humoral immune response patterns ofhuman mucosae: induction and relation to bacterial respiratorytract infections. J. Infect. Dis. 165(Suppl. 1):S167-S176.

34. Brooks, G. F., C. J. Lammel, M. S. Blake, B. Kusecek, and M.Achtman. 1992. Antibodies against IgAl protease are stimulatedboth by clinical disease and asymptomatic carriage of serogroupA Neisseria meningitidis. J. Infect. Dis. 166:1316-1321.

35. Cartwright, K. A., J. M. Stuart, D. M. Jones, and N. D. Noah.1987. The Stonehouse survey: nasopharyngeal carriage of menin-gococci and Neisseria lactamica. Epidemiol. Infect. 99:591-601.

36. Caugant, D. A., L. F. Mocca, C. E. Frasch, L. 0. Froholm, W. D.Zollinger, and R. K. Selander. 1987. Genetic structure of Neis-seria meningitidis populations in relation to serogroup, serotype,and outer membrane protein pattern. J. Bacteriol. 169:2781-2792.

37. Ceesay, S. J., S. J. Allen, A. Menon, J. E. Todd, K. Cham, G. M.Carlone, S. H. Turner, L. L. Gheesling, W. DeWitt, B. D.Plikaytis, and B. Greenwood. 1993. Decline in the meningococcalantibody levels in African children 5 years after vaccination andthe lack of an effect of booster immunization. J. Infect. Dis.167:1212-1216.

38. Chen, C. Y., C. A. Genco, J. P. Rock, and S. A. Morse. 1989.Physiology and metabolism of Neisseria gonorrhoeae and Neisseniameningitidis: implications for pathogenesis. Clin. Microbiol. Rev.2:S35-S40.

39. Christodoulides, M., B. T. McGuinness, and J. E. Heckels. 1993.Immunization with synthetic peptides containing epitopes of theclass 1 outer-membrane protein of Neisseria meningitidis: produc-tion of bactericidal antibodies on immunization with a cyclicpeptide. J. Gen. Microbiol. 139:1729-1738.

40. Cole, M. F., and C. A. Hale. 1991. Cleavage of chimpanzeesecretory immunoglobulin A by Haemophilus influenzae IgAlprotease. Microb. Pathog. 11:39-46.

41. Costantino, P., S. Viti, A. Podda, M. A. Velmonte, L. Nencioni,and R. Rappuoli. 1992. Development and phase 1 clinical testingof a conjugate vaccine against meningococcus A and meningo-

coccus C. Vaccine 10:691-698.42. Decker, M. D., K. M. Edwards, R Bradley, and P. Palmer. 1992.

Comparative trial in infants of four conjugate Haemophilusinfluenzae type b vaccines. J. Pediatr. 120:184-189.

43. de Moraes, J. C., B. A. Perkins, M. C. C. Camargo, N. T. R.Hidalgo, H. A. Barbosa, C. T. Sacchi, I. M. L. Gral, V. L. Gattas,H. deG. Vasconcelos, B. D. Plikaytis, J. D. Wenger, and C. V.Broome. 1992. Protective efficacy of a serogroup B meningococ-cal vaccine in Sao Paulo, Brazil. Lancet 340:1074-1078.

44. Densen, P. 1989. Interaction of complement with Neissenia men-ingitidis and Neisseria gonorrhoeae. Clin. Microbiol. Rev. 2:S11-S17.

45. Densen, P. 1991. Complement deficiencies and meningococcaldisease. Clin. Exp. Immunol. 86(Suppl. 1):57-62.

46. Devi, S. J. N., R. Schneerson, W. Egan, W. F. Vann, J. B. Robbins,and J. Shiloach. 1991. Identity between polysaccharide antigensof Moraxella nonliquefaciens, group B Neisseria meningitidis, andEscherichia coli Kl (non-O acetylated). Infect. Immun. 59:732-736.

47. Devoe, I. W., and J. E. Gilchrist. 1973. Release of endotoxin inthe form of cell wall blebs during in vitro growth of Neisseriameningitidis. J. Exp. Med. 138:1156-1167.

48. Diaz Romero, J. 1989. Purificaci6n y caracterizaci6n del polisa-carido capsular de Neisseria meningitidis serogrupo B. Proc. XIICongr. Nac. Soc. Espan. Microbiol. I1:104.

49. Diaz Romero, J., and I. Outschoorn. 1993. Selective biotinylationof Neisseria meningitidis group B capsular polysaccharide andapplication in an improved ELISA for the detection of specificantibodies. J. Immunol. Methods 160:35-47.

50. Diaz Romero, J., and I. Outschoorn. 1993. Detection of antibod-ies to Neisseria meningitidis serogroup B capsular polysaccharideby a liquid-phase ELISA, p. 39-48. In J. Roth, U. Rutishauser,and F. A. Troy (ed.), Polysialic acid: from microbes to man.Birkhauser Verlag, Basel.

51. Diaz Romero, J., and I. Outschoorn. 1993. An amplified photo-metric microassay for complement fixation. Ann. Biol. Clin. 51:533.

52. Diaz Romero, J., and I. Outschoorn. A comparative study of theimmune response to poly-[a(2-8)-N-acetyl-neuraminic acid]. InJ. M. McGhee, J. Mestecky, and H. Tlaskalova (ed.), Recentadvances in mucosal immunology, in press. Plenum Press, NewYork.

52a.Diaz Romero, J., and I. Outschoorn. Unpublished data.53. Dick, W. E. J., and M. Beurret. 1989. Glycoconjugates of

bacterial carbohydrate antigens. A survey and consideration ofdesign and preparation factors. Contrib. Microbiol. Immunol.10:48-114.

54. Donnelly, J. J., R R. Deck, and M. A. Liu. 1990. Immunogenicityof a Haemophilus influenzae polysaccharide-Neisseria meningitidisouter membrane protein complex conjugate vaccine. J. Immunol.145:3071-3079.

55. Edwards, M. S., A. Nicholson-Weller, C. J. Baker, and D. L.Kasper. 1980. The role of specific antibody in alternative com-plement pathway-mediated opsonophagocytosis of type III groupB Streptococcus. J. Exp. Med. 151:1275-1287.

55a.Edwards, U., A. Muller, S. Hammerschmidt, and M. Frosch.Molecular analysis of the biosynthesis pathway of the a2,8polysialic acid capsule by Neisseria meningitidis. Mol. Microbiol.,in press.

56. Feavers, I. M., A. B. Heath, J. A. Bygraves, and M. C. Maiden.1992. Role of horizontal genetic exchange in the antigenicvariation of the class 1 outer membrane protein of Neisseriameningitidis. Mol. Microbiol. 6:489-495.

57. Fernandez de Cossio, M. E., M. Ohlin, M. Llano, B. Selander, S.Cruz, J. del Valle, and C. A. Borrebaeck. 1992. Human mono-clonal antibodies against an epitope on the class Sc outermembrane protein common to many pathogenic strains of Neis-seria meningitidis. J. Infect. Dis. 166:1322-1328.

58. Ferron, L., C. M. Ferreiros, M. T. Criado, and M. Pintor. 1992.Immunogenicity and antigenic heterogeneity of a human trans-ferrin-binding protein in Neisseria meningitidis. Infect. Immun.60:2887-2892.

59. Figueroa, J. E., and P. Densen. 1991. Infectious diseases associ-



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


ated with complement deficiencies. Clin. Microbiol. Rev. 4:359-395.

60. Finne, J., D. Bitter Suermann, C. Goridis, and U. Finne. 1987. AnIgG monoclonal antibody to group B meningococci cross-reactswith developmentally regulated polysialic acid units of glycopro-teins in neural and extraneural tissues. J. Immunol. 138:4402-4407.

61. Finne, J., M. Leinonen, and P. H. Makela. 1983. Antigenicsimilarities between brain components and bacteria causingmeningitis. Implications for vaccine development and pathogen-esis. Lancet ii:355-357.

62. Frasch, C. E. 1980. Role of lipopolysaccharide in wheat germagglutinin-mediated agglutination of Neisseria meningitidis andNeisseria gonorrhoeae. J. Clin. Microbiol. 12:498-501.

63. Frasch, C. E., L. Parkes, R. M. McNelis, and E. C. Gotschlich.1976. Protection against group B meningococcal disease. I.Comparison of group-specific and type-specific protection in thechick embryo model. J. Exp. Med. 144:319-329.

64. Frasch, C. E., and M. S. Peppler. 1982. Protection against groupB Neisseria meningitidis disease: preparation of soluble proteinand protein-polysaccharide immunogens. Infect. Immun. 37:271-280.

65. Frasch, C. E., J. M. Zahradnik, L. Y. Wang, L. F. Mocca, andC. M. Tsai. 1988. Antibody response of adults to an aluminumhydroxide-adsorbed Neisseria meningitidis serotype 2b protein-group B polysaccharide vaccine. J. Infect. Dis. 158:710-718.

66. Frasch, C. E., W. D. Zollinger, and J. T. Poolman. 1985. Serotypeantigens of Neisseria meningitidis and a proposed scheme fordesignation of serotypes. Rev. Infect. Dis. 7:504-510.

67. Frosch, M., and U. Edwards. 1993. Molecular mechanisms ofcapsule expression in Neisseria meningitidis serogroup B, p.49-57. In J. Roth, U. Rutishauser, and F. A. Troy (ed.), Polysialicacid. From microbes to man. Birkhauser Verlag, Basel.

68. Frosch, M., U. Edwards, K. Bousset, B. Krausse, and C. Weis-gerber. 1991. Evidence for a common molecular origin of thecapsule gene loci in gram-negative bacteria expressing group IIcapsular polysaccharides. Mol. Microbiol. 5:1251-1263.

69. Frosch, M., I. Gorgen, G. J. Boulnois, K. N. Timmis, and D.Bitter Suermann. 1985. NZB mouse system for production ofmonoclonal antibodies to weak bacterial antigens: isolation of anIgG antibody to the polysaccharide capsules of Escherichia coliKl and group B meningococci. Proc. Natl. Acad. Sci. USA 82:1194-1198.

70. Frosch, M., D. Muller, K. Bousset, and A. Muller. 1992. Con-served outer membrane protein of Neisseria meningitidis involvedin capsule expression. Infect. Immun. 60:798-803.

71. Frosch, M., and A. Muller. 1993. Phospolipid substitution ofcapsular polysaccharides and mechanisms of capsule formationin Neisseria meningitidis. Mol. Microbiol. 8:483-493.

72. Frosch, M., C. Weisgerber, and T. F. Meyer. 1989. Molecularcharacterization and expression in Escherichia coli of the genecomplex encoding the polysaccharide capsule of Neisseria men-ingitidis group B. Proc. Natl. Acad. Sci. USA 86:1699-1673.

73. Gawinowicz, M. A., G. Merlini, S. Birken, E. F. Osserman, andE. A. Kabat. 1991. Amino acid sequence of the FV region of ahuman monoclonal IgM (NOV) with specificity for the capsularpolysaccharide of the group B meningococcus and of Escherichiacoli Kl, which cross-reacts with polynucleotides and with dena-tured DNA. J. Immunol. 147:915-920.

74. Gebran, S. J., E. L. Romano, H. A. Pons, L. Cariani, and A. N.Soyano. 1992. A modified colorimetric method for the measure-ment of phagocytosis and antibody-dependent cell cytotoxicityusing 2,7-diaminofluorene. J. Immunol. Methods 151:255-260.

75. Gibson, B. W., W. M. Melaugh, N. J. Phillips, M. A. Apicella,A. A. Campagnari, and J. M. Grifliss. 1993. Investigation of thestructural heterogeneity of lipooligosaccharides from pathogenicHaemophilus and Neisseria species and of R-type lipopolysaccha-rides from Salmonella typhimurium by electrospray mass specto-metry. J. Bacteriol. 175:2702-2712.

76. Gigliotti, F., and J. L. Shenep. 1985. Failure of monoclonalantibodies to core glicolipid to bind intact strains of Escherichiacoli. J. Infect. Dis. 151:1005-1011.

77. Glode, M. P., E. B. Lewin, A. Sutton, C. T. Le, E. C. Gotschlich,

and J. B. Robbins. 1979. Comparative immunogenicity of vac-cines prepared from capsular polysaccharides of group C Neisse-ria meningitidis 0-acetyl-positive and 0-acetyl-negative variantsand Eschenichia coli K92 in adult volunteers. J. Infect. Dis. 139:52-59.

78. Goodman, J. W., G. E. Roelants, and V. S. Byers. 1973. RNA-antigen complexes: mechanism of formation and the testing of apostulated mode of action. Ann. N. Y. Acad. Sci. 207:288-300.

79. Gotschlich, E. C., B. A. Fraser, 0. Nishimura, J. B. Robbins, andT. Y. Liu. 1981. Lipid on capsular polysaccharides of gram-negative bacteria. J. Biol. Chem. 256:8915-8921.

80. Gotschlich, E. C., T. Y. Liu, and M. S. Artenstein. 1969. Humanimmunity to the meningococcus. III. Preparation and immuno-chemical properties of the group A, group B, and group Cmeningococcal polysaccharides. J. Exp. Med. 129:1349-1365.

81. Graham, L. L. 1992. Freeze-substitution studies of bacteria.Electron Microsc. Rev. 5:77-103.

81a.Gregoriadis, G., B. McCormack, Z. Wang, and R. Lifely. 1993.Polysialic acids: potential in drug delivery. FEBS Lett. 315:271-276.

82. Grifliss, J. M., H. Schneider, R. E. Mandrell, R. Yamasaki, G. A.Jarvis, J. J. Kim, B. W. Gibson, R. Hamadeh, and M. A. Apicella.1988. Lipooligosaccharides: the principal glycolipids of the neis-serial outer membrane. Rev. Infect. Dis. 10(Suppl. 2):S287-S295.

83. Griffiss, J. M., R. Yamasaki, M. Estabrook, and J. J. Kim. 1991.Meningococcal molecular mimicry and the search for an idealvaccine. Trans. R. Soc. Trop. Med. Hyg. 85(Suppl. 1):32-36.

84. Griffiths, E., P. Stevenson, and A. Ray. 1990. Antigenic andmolecular heterogeneity of the transferrin-binding protein ofNeisseria meningitidis. FEMS Microbiol. Lett. 57:31-36.

85. Gu, X. X., and C. M. Tsai. 1993. Preparation, characterization,and immunogenicity of meningococcal lipooligosaccharide-de-rived oligosaccharide-protein conjugate. Infect. Immun. 61:1873-1880.

86. Guttormsen, H. K., R Bjerknes, A. Halstensen, A. Naess, E. A.Hoiby, and C. 0. Solberg. 1993. Cross-reacting serum opsonins tomeningococci after vaccination. J. Infect. Dis. 167:1314-1319.

87. Hajishengallis, G., E. Nikolova, and M. W. Russell. 1992. Inhi-bition of Streptococcus mutans adherence to saliva-coated hy-droxyapatite by human secretory immunoglobulin A (S-IgA)antibodies to cell surface protein antigen 1/11: reversal by IgAlprotease cleavage. Infect. Immun. 60:5057-5064.

88. Hallenbeck, P. C., F. Yu, and F. A. Troy. 1987. Rapid separationof oligomers of polysialic acid by high-performance liquid chro-matography. Anal. Biochem. 161:181-186.

88a.Halter, R., J. Pohlner, and T. F. Meyer. 1989. Mosaic-likeorganization of IgA protease genes in N. gonorrhoeae generatedby horizontal genetic exchange in vivo. EMBO J. 8:2737-2744.

89. Hayrinen, J., D. Bitter Suermann, and J. Finne. 1989. Interactionof meningococcal group B monoclonal antibody and its Fabfragment with alpha 2-8-linked sialic acid polymers: requirementof a long oligosaccharide segment for binding. Mol. Immunol.26:523-529.

90. Higgins, C. F., and M. M. Gottesman. 1992. Is the multidrugtransporter a flippase? Trends Biochem. Sci. 17:18-21.

91. Hitchcock, P. J. 1989. Unified nomenclature for pathogenicNeisseria species. Clin. Microbiol. Rev. 2:S64-S65.

92. Hubert, B., L. Watier, P. Garnerin, and S. Richardson. 1992.Meningococcal disease and influenza-like syndrome: a new ap-proach to an old question. J. Infect. Dis. 166:542-545.

92a.Humell, D. S., L. F. Mocca, C. E. Frasch, J. A. Winkelstein,H. J. J. Jean-Baptiste, J. A. Canas, and R J. Leggiadro. 1987.Meningitis caused by a nonencapsulated strain of Neisseriameningitidis in twin infants with a C6 deficiency. J. Infect. Dis.155:815-818.

93. Jarvis, G. A., and N. A. Vedros. 1987. Sialic acid of group BNeisseria meningitidis regulates alternative complement pathwayactivation. Infect. Immun. 55:174-180.

94. Jennings, H. J. 1989. The capsular polysaccharide of group BNeisseria meningitidis as a vehicle for vaccine development.Contrib. Microbiol. Immunol. 10:151-165.

95. Jennings, H. J., A. Gamian, and F. E. Ashton. 1987. N-propio-nylated group B meningococcal polysaccharide mimics a unique

VOL. 7, 1994


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


epitope on group B Neissenia meningitidis. J. Exp. Med. 165:1207-1211.

96. Jennings, H. J., A. Gamian, F. Michon, and F. E. Ashton. 1989.Unique intermolecular bactericidal epitope involving the homo-sialopolysaccharide capsule on the cell surface of group BNeisseria meningitidis and Escherichia coli Kl. J. Immunol. 142:3585-3591.

97. Jennings, H. J., and C. Lugowski. 1981. Immunochemistry ofgroups A, B and C meningococcal polysaccharide-tetanus toxoidconjugates. J. Immunol. 127:1011-1018.

98. Jennings, H. J., R. Roy, and A. Gamian. 1986. Induction ofmeningococcal group B polysaccharide-specific IgG antibodies inmice by using an N-propionylated B polysaccharide-tetanustoxoid conjugate vaccine. J. Immunol. 137:1708-1713.

99. Jennings, H. J., R. Roy, and F. Michon. 1985. Determinantspecificities of the groups B and C polysaccharides of Neisseriameningitidis. J. Immunol. 134:2651-2657.

100. Jones, D. M., and A. E. Jephcott. 1991. Neisseria, Brahamella,Moraxella, and Kingella, p.

current status of meningococcal group b vaccine candidates ... · other extreme ofonly a benign...

Documents