INTRODUCTION
Serogroup B meningococci have a capsular polysaccharide that is antigenically similar to sialic acid polysaccharides expressed by human cells (
1). Because of limited immunogenicity of this bacterial polysaccharide “self-antigen” and the risk of eliciting cross-reacting autoantibodies, vaccine developers looked beyond the capsule for alternative vaccine antigens to promote protective immunity (reviewed in reference
2). Among the various protein antigens investigated, factor H (FH) binding protein (FHbp) elicited broad serum bactericidal activity and ultimately became an important antigen in two recently licensed meningococcal vaccines for prevention of serogroup B disease. These vaccines are referred to as MenB-4C (Bexsero; GlaxoSmithKline) (so named because the vaccine contains four protective component antigens) and MenB-FHbp (Trumenba; Pfizer), which contains only FHbp.
FH is a complement-downregulating protein present in high concentrations in serum (
3). FH binds to host cells to inhibit amplification of complement activity and thus helps to prevent damage to host tissues when the complement cascade is activated. Meningococci have evolved mechanisms to recruit human FH to the bacterial surface that enable the organism to downregulate complement activation and evade complement-mediated bacteriolysis (
4,
5). Among the several meningococcal FH ligands (
4), FHbp is the most important and is highly specific for human FH (
6) and for some nonhuman primate FH (
7).
The FHbp antigen in meningococcal serogroup B vaccines can form a complex with FH that affects recognition of the bacterial antigen by the host. For example, studies in human FH transgenic mice and rhesus macaques indicated that binding of FH to the FHbp vaccine antigen decreases anti-FHbp serum bactericidal antibody responses (
8–10). In theory, the interaction between host FH and the FHbp vaccine antigen also could elicit serum anti-FH autoantibodies. Experimental support for this hypothesis comes from a MenB-4C immunogenicity study in human FH transgenic mice in which a few mice developed serum IgM anti-FH antibodies (
9). Also, in infant rhesus macaques whose macaque FH binds to FHbp similarly to human FH, two MenB-4C-vaccinated animals in each of two studies developed transient serum IgG anti-FH antibodies (
11,
52).
Autoantibodies to FH have been causally implicated in rare diseases of complement dysregulation such as autoimmune atypical hemolytic uremic syndrome (aHUS) (
12,
13) and C3 glomerulopathies (C3G) (
14). There is no evidence to date that humans immunized with serogroup B vaccines are at higher risk of developing these diseases. However, given the plausibility of FHbp elicitation of anti-FH antibodies, we undertook the first study in humans to evaluate anti-FH autoreactivity after vaccination with a meningococcal serogroup B vaccine.
DISCUSSION
Both of the licensed MenB vaccines contain recombinant FHbp, which in previous studies in human FH transgenic mice and infant macaques elicited transient cross-reactive antibodies to human FH (
9,
11,
52). However, it is not known if autoantibodies to FH are elicited in humans immunized with MenB vaccines. To address this issue, we conducted a prospective multicenter immunogenicity study of the MenB-4C vaccine and found small but statistically significant increases in serum antibody reactivity to FH at 3 weeks after vaccination using either of the vaccination schedules tested. The findings were confirmed in analysis of the anti-FH reactivity of stored paired pre- and postimmunization sera from a previous MenB-4C immunogenicity study in adults (MenB-4C study 1) (
20) but not in analysis of the anti-FH reactivity of stored sera from three groups of adults given other meningococcal vaccines that did not contain recombinant FHbp. Conceivably, binding of host FH to the recombinant FHbp vaccine antigen in MenB-4C leads to conformational changes in FH, which are recognized as “foreign” by the immune system. The immunogenicity of FH in the complex also might be enhanced by adjuvant activity from the OMV component of the MenB-4C vaccine and/or the aluminum hydroxide adjuvant.
While statistically significant, the average increases in serum anti-FH OD405 following administration of the MenB-4C vaccine containing recombinant FHbp were much lower than in serum from patients with autoimmune aHUS, and the increased postimmunization values are within the range measured in unvaccinated sera. Overall, these small average increases in anti-FH after vaccination are therefore unlikely to pose an increased risk of anti-FH autoimmune diseases; however, they suggest that MenB-4C has the potential to perturb the ability of the immune system to recognize host FH.
Serum autoantibodies to FH have been found naturally in a small proportion of healthy persons. In one study, 1% of blood donors with a mean age of 43 years (range, 18 to 68) and 8% of older subjects with a mean age of 78 years (range, 48 to 92) had serum anti-FH antibody titers that were >2 SD above the mean (
21). In most cases, the antibodies did not appear to have deleterious effects (
22). However, anti-FH autoantibodies are implicated in the pathogenesis of certain human diseases involving complement dysregulation such as autoimmune aHUS (
17,
18,
23–25) and C3G (
17,
26).
FH consists of 20 domains, referred to as short consensus repeats (SCR). In autoimmune aHUS, the anti-FH autoantibodies are reported to be primarily directed at SCR 19 and 20 in the C-terminal region of the molecule (
16,
27), which bind self surfaces and the C3d portion of C3b. Antibodies directed at this region can decrease FH function as measured in hemolytic assays (
16–18). In autoimmune aHUS, the anti-FH antibodies have been reported to decrease FH function (
18). With impaired ability of FH to downregulate complement activation (
17,
18), unchecked complement activation can lead to onset of renal disease and hemolytic anemia characteristic of aHUS. While C3G also is associated with serum autoantibodies to FH (
17), the antibodies are reported to be of lower avidity for FH than in patients with autoimmune aHUS (
23), to be directed at different FH epitope specificities located in the N-terminal region (
23), and typically not to affect FH functional activity as measured in the hemolytic assay (
17). Nevertheless, the anti-FH autoantibodies appear to have some functional consequences with respect to progression of renal disease in patients with C3G, perhaps in concert with other antibodies such as C3 nephritic factor (
17).
In our prospective study, three MenB-4C-vaccinated subjects showed increases in serum anti-FH binding after vaccination that approached the anti-FH titers present in sera from two patients with autoimmune aHUS. However, the increased titers after vaccination were transient and returned to baseline or near baseline by 4 to 5 months. Further, none of the sera with elevated titers had impaired FH function measured in the hemolytic assay. In historical MenB-4C study 2, 10 MenB-vaccinated college students with stored postimmunization sera (9.6%) had elevated anti-FH antibody at 6 weeks to 2 months postimmunization versus 1 (2%) unvaccinated student. Important limitations of this study were a lack of preimmunization sera to determine whether the anti-FH antibodies were present before vaccination and a lack of additional follow-up sera to determine antibody persistence. However, only 1 of the 11 serum samples with elevated anti-FH antibody activity from vaccinated students had low FH function in the hemolytic assay, which suggested that most of the subjects with elevated serum anti-FH antibody reactivity had minimal risk of developing autoimmune aHUS. However, the risk of C3G in this population is not known since serum anti-FH antibodies in patients with C3G are reported not to affect the level of FH function measured by the hemolytic assay (
17).
As noted above, increased serum anti-FH reactivity was not observed in stored paired pre- and postimmunization sera from three other historical studies of other meningococcal vaccines that did not contain significant levels of recombinant FHbp (
28–30). Two of these studies investigated OMV vaccines prepared from group B strain H44/76, which naturally expresses FHbp. Interestingly, one individual given an OMV vaccine exhibited a significant increase in serum reactivity to FH (
Fig. 4D). OMVs are treated with detergents to decrease endotoxin content. The treatment also removes other detergent-soluble molecules, including FHbp, which is a lipoprotein, and FHbp is reported to be a very minor protein in the detergent-extracted outer membrane vesicle (dOMV) prepared from strain NZ98/254 (<0.045 μg per 25-μg dose [
31]) or from strain H44/76 (0.01 to 0.025 μg per 25-μg dose [
32,
33]). Whether this small amount (less than 0.05 μg per 25-μg human dose) is sufficient to recruit FH and elicit anti-FH antibody response is not known. Our data indicated that the subjects given OMV vaccines did not as a group show increases in anti-FH OD
405 after vaccination (
P = 0.14 and 0.71). Thus, there was no statistical evidence that vaccination evoked antibody to FH, which is consistent with our hypothesis that the potential for eliciting anti-FH antibodies is highest when FHbp is present in high amounts (i.e., 50 μg of the recombinant fusion protein in MenB-4C). While it is not possible to draw conclusions from a single case of elevated anti-FH titers in an OMV-vaccinated subject, there is a possibility that host FH complexed with residual FHbp in the OMV might have been responsible.
As of February 2018, an estimated 20 million doses of MenB-4C had been distributed worldwide (
https://www.gsk.com/en-gb/media/press-releases/gsk-s-meningitis-b-vaccine-bexsero-receives-breakthrough-therapy-designation-from-us-fda-for-prevention-of-invasive-meningococcal-disease-in-children-2-10-years-of-age/; accessed 20 March 2019). To date, there have been no published reports suggesting elevated incidences of autoimmune aHUS or C3G after MenB-4C vaccination. However, four cases of nephrotic syndrome were identified recently in children ages 2 to 5 years who had been vaccinated with MenB-4C during a mass immunization campaign in Quebec, Canada (
34). The nephrotic syndrome was considered to be idiopathic based on therapeutic responses to steroid therapy. However, none of the cases had a biopsy performed or serum anti-FH antibody level determined, and one of the cases relapsed and required long-term immunosuppression therapy. Given the relative rarity of nephrotic syndrome in the population, with an estimated annual incidence of 17.7 per 100,000, the four cases were considered by the authors to represent “a potential vaccine safety signal.”
In summary, the lack of reports of autoimmune aHUS or C3G is reassuring, but these diseases are rare in the population and onset is likely to be delayed for months or years beyond vaccination and may require a secondary trigger (
35). Therefore, an association with vaccination will be difficult to ascertain (
36). Also, it is possible that only certain subgroups of the population are at increased risk of developing anti-FH related autoimmune diseases, such as persons with preexisting autoimmune diseases or healthy persons with a deficiency of complement factor H (CFH)-related proteins such as occurs in ∼ 6% of the population (
37), and are associated with autoantibodies to FH and aHUS (
24,
26,
38,
39). The risk of these diseases also may be heightened in the presence of rare variants in the FH gene at functional sites or in the context of naturally low serum FH levels (
40–42) if these persons should also acquire anti-FH autoantibodies. Thus, there remains a need for continued surveillance for diseases associated with anti-FH autoantibodies in larger populations and possibly for performing case-control studies in immunized and unimmunized persons to determine relative risk.
Finally, our studies of serum anti-FH reactivity were limited to adults immunized with the recommended two doses of MenB-4C. Similar studies are needed in infants and children given the recommended two-dose or three-dose schedule used in Europe and in adults immunized with the MenB-FHbp vaccine (Trumenba), which contains two lipidated FHbp antigens (
43,
44). Studies also are needed on serum anti-FH antibody in adults given booster doses of MenB-4C or MenB-FHbp, since serum bactericidal titers after the recommended 2-dose or 3-dose schedules can decline within a year to levels below those considered protective (
43,
45,
46), and booster doses may be needed, which may also boost anti-FH reactivity. This is particularly true for adults at increased risk of developing meningococcal disease such as laboratory workers with occupational exposure to meningococci (
47) and patients with aHUS or paroxysmal nocturnal hemoglobinuria who are undergoing complement inhibition therapy (
48–50).
ACKNOWLEDGMENTS
We are grateful to Serena Giuntini, who established the serum anti-FH assays at the UCSF Benioff Children’s Hospital Oakland, Oakland, CA, and did preliminary studies on sera from MenB-4C-vaccinated subjects. The control sera from two patients with aHUS and autoantibodies to FH were kindly provided by David Kavanagh and Kevin Marchbank of Newcastle University, United Kingdom. Matthew Eggleston, National Jewish Health Complement Laboratory, Denver, CO, performed the anti-FH functional hemolytic assay. We thank Dr. John Atkinson, Washington University School of Medicine, St. Louis, for critical review of the manuscript.
This investigation was supported by research grants R01 AI046464 (D.M.G.), R01 AI114701 (D.M.G. and P.T.B.), and R01 AI099125 (P.T.B.) from the National Institute of Allergy and Infectious Diseases, NIH. The work was performed in a facility funded by the Research Facilities Improvement Program grant C06 RR016226 from the National Center for Research Resources, NIH. The CIRN study was funded by the Canadian Institutes of Health Research.
The study was conceived and designed by J.M.L., D.M.G., and P.T.B.; data were acquired by K.S., J.M.L., S.G., C.Q., C.D., M.G., and D.M.G.; data were analyzed and interpreted by K.S., P.T.B., J.M.L., S.G., C.Q., C.D., Q.L., M.G., and D.M.G.; drafting of the article was performed by and critical intellectual content was provided by K.S., P.T.B., J.M.L., S.G., C.Q., C.D., Q.L., M.G., and D.M.G.; final approval of the version to be submitted was provided by K.S., P.T.B., J.M.L., S.G., C.Q., C.D., Q.L., M.G., and D.M.G.
D.M.G. and P.T.B. are inventors on patent applications or on issued patents in the area of meningococcal vaccines. These include mutant FHbp antigens with low levels of binding to FH. Rights to these inventions have been assigned to UCSF Benioff Children’s Hospital Oakland. D.M.G. receives consultation fees from serving on advisory boards of Allopex LLC and Sanofi Pasteur for vaccines unrelated to the work described in this article. J.M.L. holds the Canadian Institutes of Health Research – GlaxoSmithKline Chair in Pediatric Vaccinology, and her institution has received funding from Pfizer and GSK for meningococcal vaccine studies. S.G. receives research funding and consulting fees from Merck and GSK, research funding from VBI Vaccines and Meridian Bioscience, and consulting fees from the Infectious Diseases Research Institute. M.G., K.S., Q.L., C.D., and C.Q. declare no conflicts of interest. We were solely responsible for the investigation design, data analysis, and writing of the manuscript. All of us attest we meet the ICMJE criteria for authorship.