Meningococcal Vaccines Directed at Capsular Group B

MenB-4C

The MenB-4C vaccine contains three recombinant proteins: FHbp subfamily B (as a fusion protein with GNA 2019), Neisseria Heparin binding antigen (NHba, as a fusion protein with GNA 1030), and Neisseria adhesin A (NadA), which are combined with a detergent-extracted OMV from strain NZ98/254 ( Fig. 40.2 ). A 0.5-mL dose contains 50 µg of each the three recombinant proteins and 25 µg of OMV. The antigens are adsorbed with 0.519 mg of Al ³⁺ (as aluminum hydroxide) ( Table 40.2 ). MenB-4C was licensed in the UK and member states of the European Union and European Economic Area in 2013 and approved for the age group 2 months to 55 years. The vaccine has been subsequently approved in Canada, Australia, Brazil, Argentina, Chile, and Uruguay and more than 30 other countries, including the United States. In the United States, MenB-4C is FDA approved for use in the age group 10 to 25 years.

MenB-FHbp

MenB-FHbp contains two lipidated recombinant FHbp sequence variants, one from each subfamily. ^, The 0.5-mL dose contains 60 µg of each FHbp variant (total of 120 µg of protein) (see Table 40.2 ). The antigens are adsorbed with 0.25 mg of Al ⁺³ (as aluminum phosphate). MenB-FHbp is approved as a two or three dose schedule in the United States, UK, the European Union member states plus Iceland, Liechtenstein, and Norway, for individuals 10 years and older. In the United States, the MenB-FHbp vaccine is approved for the ages 10 to 25 years. MenB-FHbp vaccine can be administered on a two-dose (0 and 6 months) schedule to healthy adolescents and young adults or as a tailored three-dose (0, 1–2, and 6 months) schedule for individuals at increased risk.

FHbp

First discovered by scientists at the former Chiron Vaccines in Siena, Italy, using genome mining, the protein was designated “genome-derived Neisseria antigen 1870” or “GNA1870.” Investigators at Wyeth Vaccines (now Pfizer) in New York independently discovered the same antigen using biochemical and immunologic approaches and called it “lipoprotein 2086” or “LP2086.” The protein was subsequently found by Sanjay Ram and colleagues to bind a human protein, called Factor H (FH), which downregulates complement activation (particularly the alternative pathway). The binding was shown to be specific for human and some nonhuman primate FH and enabled meningococci to survive in nonimmune serum. Rat FH does not bind to meningococci, and the organism is killed by 20% nonimmune infant rat serum after incubation of the bacteria for 60 minutes. The addition, however, of as little as 3 µg/mL of human FH was sufficient to downregulate the rat alternative pathway and “rescue” the bacteria from killing (typical concentrations of FH in human serum range from 200 to 600 µg/mL). These data provide one reason why meningococci naturally infect only humans; in humans, binding of FH to the bacteria allows N. meningitidis to evade host defenses. A human genome wide association study identified variants in the FH region associated with host susceptibility to meningococcal disease. Because of the importance of binding of FH to GNA1870/LP2086 in enabling the organism to evade complement, the various names of the antigen were changed to FHbp.

The FH molecule contains 20 domains. ^, Each domain contains approximately 60 amino acids and can be thought of as a series of individual beads on a string. Domains 1 to 4 are responsible for downregulating complement activation, and domains 6 and 7 contain the amino acid residues that interact with FHbp. The structure of FHbp in a complex with FH domains 6 and 7 is shown in Fig. 40.3 . The FHbp molecule contains two antiparallel β-barrel domains. The amino acid residues in FHbp that interact with FH are located on both the N-terminal and C-terminal portions of the molecule and appear to be the most important epitopes for eliciting bactericidal anti-FHbp antibodies. ^, In immunized mice whose mouse FH does not bind to the vaccine antigen, serum anti-FHbp antibodies are directed at epitopes located in the FH binding site, and the antibodies block FH binding. With less bound FH, there is less complement downregulation and the bacteria become more susceptible to complement-mediated bactericidal activity. ^, In immunized humans and nonhuman primates, binding of FH to the FHbp vaccine antigen appears to mask these epitopes, and the resulting anti-FHbp antibodies are largely directed to epitopes outside of the FH binding area. As a result, the antibodies do not block FH binding. ^{, ,}

As of Feburary 2021, 1750 alleles with 11,347 distinct FHbp amino acid sequence variants (peptide alleles) had been identified in the PubMLST database and assigned individual FHbp peptide identification (ID) numbers. Based on amino acid sequence similarity, these variants can be subclassified into three variant groups (1, 2, or 3) or two subfamilies (A and B). Variant group 1 corresponds to subfamily B, and variant groups 2 and 3 correspond to subfamily A. Within a subfamily, FHbp amino acid sequence identity ranges from 88% to 99%. Between subfamilies, sequence identity is approximately 60%. While exceptions have been noted, in general, antibodies directed at FHbp in subfamily A have greater bactericidal activity against strains expressing subfamily A FHbp, and vice versa for strains expressing FHbp in subfamily B. ^{, ,} Because of amino acid relatedness and extensive cross-reactivity between FHbp sequence variants in variant groups 2 and 3 (subfamily A), ^, this chapter uses the subfamily classification system.

MenB-FHbp contains two lipidated recombinant FHbp sequence variants, one from each subfamily. The lipidation is thought to act as an adjuvant and increase immunogenicity. ^, The subfamily A FHbp sequence variant in this vaccine is ID 45 (called A05 by Pfizer), and the subfamily B variant is ID 55 (also called B01) ( Fig. 40.4 ). MenB-4C contains a subfamily B FHbp variant (ID 1or B24) that is not lipidated, and three other antigens capable of eliciting serum bactericidal activity (see Fig. 40.2 ). The FHbp antigen in MenB-4C is a fusion protein with a second antigen called GNA 2091. GNA 2091 was initially thought to contribute to vaccine protection and was combined with FHbp to facilitate vaccine production. The antigen was subsequently found to be localized on the periplasmic side of the outer membrane and did not elicit serum bactericidal activity. GSK considers GNA 2091 an “accessory protein” in the MenB-4C vaccine, and the presence of GNA 2091 in the FHbp fusion protein may increase FHbp stability and immunogenicity.

FHbp vaccines elicit broad serum bactericidal antibody responses in mouse, rabbit, and nonhuman primate models. ^{, , , , ,} Both manufacturers (GSK and Pfizer) have proposed that a match between the FHbp subfamily (or variant group) of the vaccine and the strain, and with expression of strain antigen above a certain threshold, then the isolate can be considered covered by the vaccine. ^, However, the story is more nuanced. ^, For example, within a subfamily there can be 10-fold or greater differences in susceptibility against mutants with different FHbp amino acid sequence variants and unexpected cross-reactive bactericidal activity against subfamily A with the FHbp subfamily B antigen in MenB-4C. Also, Pajon et al. noted the effect of decreasing strain antigen expression on lowering serum bactericidal activity of mice immunized with a recombinant FHbp vaccine ID 1 that matched the FHbp in the test strains. Note that many wild-type meningococcal strains causing disease have similar FHbp expression to the mutant used in this study with 50% decreased FHbp expression. ^{, , , ,} Intrinsic strain resistance to anti-FHbp bactericidal activity also may be mediated by binding of FH to two other meningococcal FH ligands, NspA and the porin protein PorB, which can downregulate complement. These experimental observations underscore the challenges for predicting strain coverage by the antibodies elicited by FHbp vaccines.

Rarely (∼1 : 1000), invasive meningococcal isolates with absent FHbp genes or frameshift mutations have been identified that do not express a functional FHbp. These isolates express other FH ligands, such as NspA or PorB, and are completely resistant to anti-FHbp bactericidal activity. To date, vaccination with FHbp-containing vaccines has not selected for emergence of FHbp-null strains, possibly because selection would occur mainly on infected mucosal surfaces and MenB vaccines appear to have minimal effect on decreasing nasopharyngeal colonization.

Neisseria Adhesin A

MenB-4C contains a subfamily B FHbp antigen. MenB-4C coverage against meningococcal strains with subfamily A FHbp is therefore largely dependent on protective antibodies elicited by the three other vaccine antigens. One of these is NadA, which is a surface-exposed lipoprotein discovered by genome mining and designated GNA1994. Present in the outer membrane as a homotrimer, NadA has structural similarities to a family of autotransporter adhesins and may promote Neisseria adhesion to epithelial cells (hence the name NadA). In experimental animals and humans, recombinant NadA vaccines elicited serum bactericidal antibodies. ^,

As of February 2021, there were 322 NadA alleles with 130 amino acid sequence variants and other changes such as IS element insertions or internal stop codons in the public database. Many of the differences are accounted for by phase variable on and off configurations, frameshifts and deletions. In a classification system updated in 2014, NadA was subdivided into three variant groups based on amino acid similarity (NadA-1, NadA-2/3, and NadA-4/5). MenB-4C contains a peptide variant from group 2/3, which elicits cross-reactive antibodies to NadA-1 but not NadA-4/5.

An important limitation of NadA as a vaccine antigen is that only about 20% to 40% of group B strains have genes coding for the protein (primarily strains from clonal complexes STs 8, 11, and 32). As noted above, NadA genes can be interrupted or frame shifted. Furthermore, the isolates for which anti-NadA antibodies might be most important for protection after MenB-4C vaccination (isolates with FHbp subfamily A, which is not present in the vaccine) infrequently have nadA genes (6%–19%). NadA expression levels also are largely controlled by a transcriptional regulator NadR (Neisseria adhesin A regulator). Expression of NadA is repressed in many strains by NadR, which can result experimentally in resistance to anti-NadA serum bactericidal activity. NadA expression can be induced by certain small molecules in saliva that inhibit NadR repression. In an ex vivo model of human whole blood infection, meningococci did not activate genes encoding for NadA. Nevertheless, expression of NadA during human meningococcal disease is sufficient to induce anti-NadA antibodies in convalescent serum. In short, MenB-4C can induce anti-NadA bactericidal activity against some NadA-positive meningococcal strains (such as strain 5/99, which has naturally high NadA expression and is used by GSK to assess anti-NadA serum bactericidal responses). However, for many nadA positive strains, expression of the protein is low under in vitro growth conditions used to measure bactericidal activity, and some NadA-positive strains are relatively resistant to serum bactericidal activity elicited by MenB-4C in adults (see below). ^, However, the assay conditions for measurement of serum bactericidal antibody are not optimum for expression of NadA and it is possible that the assays underestimate NadA as a protective antigen. Defining the actual contribution of anti-NadA antibodies to MenB-4C vaccine coverage therefore remains challenging.

Neisseria Heparin-Binding Antigen

NHba is a surface-exposed lipoprotein that also was discovered by genome mining and designated GNA 2132. Subsequently, the protein was found to bind heparin in vitro, which correlated with increased survival of unencapsulated meningococci in human serum. While the role of heparin binding as a surrogate for interactions of N. meningitidis with host polyanions and pathogenesis remains uncertain, the name of the protein was changed to NHba to reflect its heparin-binding function. Almost all MenB isolates harbor the NHba gene. As of February 2021, there were 1409 NHba amino acid sequence variants in the public database. To date, these variants have not been classified into related families because of the lack of clear relationships between the respective amino acid sequence variants to antigenic cross-reactivity. For classification, each NHba sequence variant is designated with a specific peptide ID. MenB-4C contains NHba peptide 2, which is part of a fusion protein with a second antigen, GNA 1030 (see Fig. 40.2 ). GNA1030 was initially thought to contribute to vaccine-induced protection and was combined with NHba to facilitate vaccine production. However, GNA1030 was subsequently found to be a ubiquinone-8 binding protein and not to elicit serum bactericidal activity. As is the case for GNA 2091 (the second antigen in the FHbp fusion protein), GSK considers GNA 1030 an “accessory protein” in the NHba antigen. The NHba fusion protein expresses the expected epitopes found in the native NHba and appears to be stable.

Defining the role of NHba in MenB-4C coverage has been challenging. One study found an approximate 100-fold difference in NHba expression across isolates. In some strains, an endogeonous protease, NalP, cleaves NHba, which releases a C-terminal 22-kDa fragment that may contain epitopes recognized by anti-NHba antibody. NHba also was found to be cleaved by protease activity of human lactoferrin. Although the investigators concluded that proteolytic cleavage of NHba did not affect anti-NHba bactericidal response, the cidal activity was measured with rabbit complement. Because rabbit FH does not bind to meningococci, the anti-NHba activation of complement was not downregulated, which would have falsely increased bactericidal activity measured with human complement. Early studies of mice immunized with a recombinant NHba vaccine given with Freund’s adjuvant showed that serum anti-NHba antibodies were bactericidal primarily with rabbit complement (only two selected strains were killed with human complement). GSK also has had difficulty finding a suitable meningococcal strain for measuring anti-NHba–specific serum bactericidal activity after MenB-4C vaccination. As a result, anti-NHba bactericidal activity was not measured in the early studies of the immunogenicity of MenB-4C in humans. ^, Also, anti-NHba bactericidal activity was not included in the human immunogenicity dataset for U.S. licensure of MenB-4C. ^,

Subsequently, Rossi et al. investigated serum bactericidal activity elicited by MenB-4C against different serogroup B strains responsible for two outbreaks on U.S. college campuses. Both outbreak strains were susceptible to bactericidal activity of sera from rhesus macaques or a human immunized with MenB-4C. This result was predicted based on high expression of two of the vaccine antigens in both strains, subfamily B FHbp and NHba, and moderate expression of a third vaccine antigen, NadA, in one of the strains. Depletion of anti-FHbp antibodies from the postimmunization sera, however, resulted in loss of all vaccine-induced bactericidal activity. Thus, despite strain NHba expression and high titers of anti-NHba antibody in the FHbp-adsorbed sera, there was no detectable anti-NHba bactericidal activity measured with human complement. The basis for this resistance is unknown and requires further investigation. These data, together with data on limited serum bactericidal responses of MenB-4C-vaccinated adults against some strains expressing NHba but mismatched for the other three MenB-4C antigens, underscore the difficulty of predicting the contribution of anti-NHba antibodies to protection elicited by MenB-4C. The MenB-4C NHba peptide-2 has been demonstrated to induce cross-reactive serum bactericidal antibodies to some NHba peptides but not others, but anti-NHba antibody that lacks bactericidal activity individually against a strain may contribute to bactericidal activity elicited by MenB-4C in concert with anti-FHbp antibody ^,