Genetic Diversity of Meningococcal Serogroup B Vaccine Antigens among Carriage Isolates Collected from Students at Three Universities in the United States, 2015–2016

ABSTRACT Carriage evaluations were conducted during 2015 to 2016 at two U.S. universities in conjunction with the response to disease outbreaks caused by Neisseria meningitidis serogroup B and at a university where outbreak and response activities had not occurred. All eligible students at the two universities received the serogroup B meningococcal factor H binding protein vaccine (MenB-FHbp); 5.2% of students (181/3,509) at one university received MenB-4C. A total of 1,514 meningococcal carriage isolates were obtained from 8,905 oropharyngeal swabs from 7,001 unique participants. Whole-genome sequencing data were analyzed to understand MenB-FHbp’s impact on carriage and antigen genetic diversity and distribution. Of 1,422 isolates from carriers with known vaccination status (726 [51.0%] from MenB-FHbp-vaccinated, 42 [3.0%] from MenB-4C-vaccinated, and 654 [46.0%] from unvaccinated participants), 1,406 (98.9%) had intact fHbp alleles (716 from MenB-FHbp-vaccinated participants). Of 726 isolates from MenB-FHbp-vaccinated participants, 250 (34.4%) harbored FHbp peptides that may be covered by MenB-FHbp. Genogroup B was detected in 122/1,422 (8.6%) and 112/1,422 (7.9%) isolates from MenB-FHbp-vaccinated and unvaccinated participants, respectively. FHbp subfamily and peptide distributions between MenB-FHbp-vaccinated and unvaccinated participants were not statistically different. Eighteen of 161 MenB-FHbp-vaccinated repeat carriers (11.2%) acquired a new strain containing one or more new vaccine antigen peptides during multiple rounds of sample collection, which was not statistically different (P = 0.3176) from the unvaccinated repeat carriers (1/30; 3.3%). Our findings suggest that lack of MenB vaccine impact on carriage was not due to missing the intact fHbp gene; MenB-FHbp did not affect antigen genetic diversity and distribution during the study period.

ABSTRACT Carriage evaluations were conducted during 2015 to 2016 at two U.S. universities in conjunction with the response to disease outbreaks caused by Neisseria meningitidis serogroup B and at a university where outbreak and response activities had not occurred. All eligible students at the two universities received the serogroup B meningococcal factor H binding protein vaccine (MenB-FHbp); 5.2% of students (181/3,509) at one university received MenB-4C. A total of 1,514 meningococcal carriage isolates were obtained from 8,905 oropharyngeal swabs from 7,001 unique participants. Whole-genome sequencing data were analyzed to understand MenB-FHbp's impact on carriage and antigen genetic diversity and distribution. Of 1,422 isolates from carriers with known vaccination status (726 [51.0%] from MenB-FHbp-vaccinated, 42 [3.0%] from MenB-4C-vaccinated, and 654 [46.0%] from unvaccinated participants), 1,406 (98.9%) had intact fHbp alleles (716 from MenB-FHbp-vaccinated participants). Of 726 isolates from MenB-FHbp-vaccinated participants, 250 (34.4%) harbored FHbp peptides that may be covered by MenB-FHbp. Genogroup B was detected in 122/1,422 (8.6%) and 112/1,422 (7.9%) isolates from MenB-FHbp-vaccinated and unvaccinated participants, respectively. FHbp subfamily and peptide distributions between MenB-FHbp-vaccinated and unvaccinated participants were not statistically different. Eighteen of 161 MenB-FHbp-vaccinated repeat carriers (11.2%) acquired a new strain containing one or more new vaccine antigen peptides during multiple rounds of sample collection, which was not statistically different (P = 0.3176) from the unvaccinated repeat carriers (1/ 30; 3.3%). Our findings suggest that lack of MenB vaccine impact on carriage was not due to missing the intact fHbp gene; MenB-FHbp did not affect antigen genetic diversity and distribution during the study period. IMPORTANCE The impact of serogroup B meningococcal (MenB) vaccines on carriage is not completely understood. Using whole-genome sequencing data, we assessed the diversity and distribution of MenB vaccine antigens (particularly FHbp) among 1,514 meningococcal carriage isolates recovered from vaccinated and unvaccinated students at three U.S. universities, two of which underwent MenB-FHbp mass vaccination campaigns following meningococcal disease outbreaks. The majority of carriage isolates recovered from participants harbored intact fHbp genes, about half of which were recovered from MenB-FHbp-vaccinated participants. The distribution of vaccine antigen peptides was similar among carriage isolates recovered from vaccinated and unvaccinated participants, and almost all strains recovered from repeat I nvasive meningococcal disease (IMD) caused by Neisseria meningitidis (the meningococcus) is a major public health concern (1). This bacterium can be asymptomatically carried as a commensal in the human nasopharyngeal mucosa. Exposure to meningococci can lead to either carriage or, less commonly, to IMD. Carriage prevalence varies among studies conducted in different populations and is age related, peaking during late adolescence in developed countries (2,3). Risk factors, including behaviors linked with social mixing and smoking, have also been associated with carriage acquisition (3).
Six serogroups (A, B, C, W, X, and Y) cause most IMD cases worldwide, with B, C, and Y predominating in the United States (1,4), and unencapsulated N. meningitidis (nongroupable, or NG) is commonly associated with carriage (5)(6)(7)(8). Quadrivalent meningococcal conjugate vaccines confer protection against serogroups A, C, W, and Y and are routinely recommended for all U.S. adolescents aged 11 to 18 years as well as certain other individuals at increased meningococcal disease risk (9). Two proteinbased serogroup B meningococcal (MenB) vaccines were licensed in the United States in 2014 to 2015 (10). MenB-FHbp (Trumenba; also known as bivalent rLP2086; Pfizer) contains two FHbp (factor H binding protein) peptides, one from subfamily A (A05) and one from subfamily B (B01) (Pfizer nomenclature) (10,11), corresponding to peptides 3.45 and 1.55 (GlaxoSmithKline [GSK] nomenclature), respectively. MenB-4C (Bexsero [also known as 4CMenB]; GSK) has four components: FHbp B24 or 1.1, NhbA (neisserial heparin binding antigen) peptide 2 (p0002), NadA (neisserial adhesin A) peptide 3.8 (NadA-3.8), and outer membrane vesicle (OMV) from the N. meningitidis serogroup B strain NZ98/254 (derived from the MeNZB vaccine) with porin A variable region 2 (PorA-VR2) variant 4 as the major antigen (12,13). In the United States, MenB vaccines are recommended for adolescents aged 16 to 23 years based on shared clinical decision making; they are also recommended for individuals aged 10 years or older who have increased risk for N. meningitidis serogroup B disease because of specific underlying conditions, a serogroup B outbreak, or occupational exposure as a microbiologist (14). Although licensed to protect against serogroup B disease, the vaccine antigens are also present in meningococci of other serogroups and may protect against non-B meningococcal disease (15,16). While the long-term impact of MenB vaccines on meningococcal disease remains to be evaluated, collective evidence indicates that MenB vaccines do not have an impact on total or serogroup B carriage prevalence (7,8,17).
Outbreaks of serogroup B meningococcal disease were reported at 10 U.S. universities during 2013 to 2018 (18). Two outbreaks occurred in 2015 at universities in Rhode Island (RI-1) and Oregon (OR) (7,8). MenB vaccination campaigns were implemented at each university for outbreak control. All eligible students at RI-1 received MenB-FHbp (8). While the majority of the eligible students at OR received MenB-FHbp, 5.2% of students received MenB-4C (7). Carriage evaluations conducted at the universities during 2015 to 2016 showed no decrease in meningococcal carriage following vaccination (7,8). A third carriage evaluation was also conducted at a university near RI-1 (RI-2) as a reference population, where neither N. meningitidis serogroup B outbreaks nor MenB mass vaccination campaigns occurred (19). To understand the molecular mechanisms underlying the lack of impact of MenB-FHbp on carriage and the selective pressure of this vaccine on genetic diversity and prevalence of MenB vaccine antigens over time, we sequenced genomes of all carriage isolates collected from the three carriage evaluations and assessed potential changes in diversity and distribution of MenB vaccine antigens among vaccinated and unvaccinated students as well as changes in vaccine antigen profiles among students who participated in the carriage evaluations at multiple rounds of sample collection.
FHbp subfamily distribution among vaccinated and unvaccinated participants. MenB-FHbp was provided to all eligible students at RI-1 and most eligible students at OR; MenB-4C was provided to 5.2% (181 of 3,509) of students at OR. A total of 1,422 N. meningitidis isolates (646 from RI-1, 528 from OR, and 248 from RI-2) were recovered from participants with known vaccination status (Table S4) (Table 2). Overall, for all three universities, higher proportions of isolates contained FHbp B/v1 than FHbp A/v2-3 among all or non-B genogroups within both vaccinated and unvaccinated groups. In contrast, a higher proportion of genogroup B isolates contained FHbp A/v2-3 than B/v1 within both groups. There was no significant difference in FHbp subfamily distribution between vaccinated and unvaccinated groups among all genogroups, B genogroup,  FHbp peptides distribution and predicted strain coverage among vaccinated and unvaccinated participants. The distribution of FHbp peptides was assessed among carriage isolates from MenB-FHbp-vaccinated and unvaccinated participants at RI-1 and OR (Fig. 2). Overall, isolates containing FHbp B09/1.13 and B16/1.4 peptides were predominant among MenB-FHbp-vaccinated groups (215/646; 33.3% at RI-1 and 133/528; 25.2% at OR) and unvaccinated groups (95/646; 14.7% at RI-1 and 90/528; 17.1% at OR). Only eight genogroup B isolates contained either FHbp B09/1.13 or B16/1.4, six from MenB-FHbp-vaccinated (all at RI-1) and 2 from unvaccinated (one each at RI-1 and OR) participants. A small proportion of isolates either lacked the fHbp allele or contained a truncated fHbp allele, including nine isolates from MenB-FHbp-vaccinated participants and four isolates from unvaccinated participants at RI-1 and two isolates from OR (one each from MenB-FHbp-vaccinated and unvaccinated participants). There was no statistically significant difference in FHbp peptide distributions between vaccinated and unvaccinated participants among genogroup B isolates and non-B genogroups from each university.
While FHbp B01/1.55 was not detected in any isolate in this study, FHbp A05/3.45 was present in 12 of 726 (1.7%) isolates (11 genogroup B and one undetermined genogroup) from MenB-FHbp-vaccinated participants (nine from RI-1 and three from OR), all of which were from six MenB-FHbp-vaccinated participants (each received either one or two vaccine doses  one (3.3%) from RI-2 had two isolates that acquired a new strain genotype. The difference between vaccinated and unvaccinated repeat carriers who acquired a new strain was not statistically significant (P = 0.3176).

DISCUSSION
Analysis of whole-genome sequencing data has allowed us to understand MenB vaccine antigens and assess the impact of MenB-FHbp on their genetic diversity and distribution in U.S. meningococcal carriage. Almost all carriage isolates (99%) contained an intact fHbp gene, with 61% belonging to FHbp B/v1 peptides. These findings are in contrast to a previous study showing a higher proportion of FHbp A/v2-3 peptides among carriage isolates (25), but the proportion of genogroup B isolates in this prior study was greater than that found in our study. Similarly, the majority of carriage isolates (.99%) in our study contain an intact nhbA gene, with 99% of isolates harboring both fHbp and nhbA. In contrast, the nadA gene was detected in only a small proportion (7%) of carriage isolates. We found an equal proportion of carriage isolates with an intact fHbp gene recovered from vaccinated (95% received MenB-FHbp) and unvaccinated participants. Statistical analysis showed no significant difference in the distribution of FHbp subfamilies or peptides between vaccinated and unvaccinated populations, suggesting that the vaccine does not impact the overall FHbp diversity and distribution.
Comparative whole-genome analyses provide evidence of ongoing genetic shift and gene loss in Bordetella pertussis following the introduction of pertussis vaccines, possibly due to vaccine selective pressure (26). Our data suggest that MenB-FHbp vaccination did not exert significant selective pressure on MenB vaccine antigens in our study. We found that 63% of carriage isolates lacking or containing a truncated form of the fHbp allele were from MenB-FHbp-vaccinated participants. However, whether MenB-FHbp administration exerts a long-term impact similar to that observed with pertussis vaccines warrants further investigations. Most repeat carriers received MenB-FHbp. Almost all strains recovered from MenB-FHbp-vaccinated and unvaccinated repeat carriers retained the same vaccine antigen peptides. Only a small percentage of carriage isolates recovered from the vaccinated repeat carriers acquired a new strain with different vaccine antigen peptides. Among isolates from the repeat carriers, we observed significant differences in FHbp peptide distribution between vaccinated and unvaccinated groups. However, it could be due to the much higher proportion of isolates from MenB-FHbp-vaccinated than from unvaccinated repeat carriers.
While the genomic data in this study showed that a high proportion of carriage isolates harbored an intact fHbp gene, FHbp surface expression has not been evaluated among these isolates. The antigen surface expression level and its accessibility to bactericidal antibodies are essential factors that can impact vaccine effectiveness (27)(28)(29). The genetic diversity and expression of FHbp may affect the bactericidal activity of FHbp-specific antibodies. It is conceivable that a lack of vaccine impact on carriage is due to the low level of antigen expression. A carriage study conducted in France among individuals aged 1 to 25 years showed that FHbp expression was significantly lower in carriage isolates than in invasive isolates; 32% of all carriage isolates tested had no detectable FHbp (25). Further investigations on the level of vaccine antigen surface expression in the carriage isolates from our study may shed light on mechanisms underlying the lack of impact of MenB vaccines on carriage. While induction of mucosal antibody responses has been demonstrated for meningococcal serogroup C polysaccharide-conjugate vaccines, thereby reducing the carriage and improving the level of herd protection (30)(31)(32), this effect has not been confidently demonstrated for MenB vaccines. Prior research suggests that a mucosal mode of administration is required to achieve an adequate immune response within the mucosal environment using MenB vaccines (33)(34)(35). Finally, although MenB vaccine-induced bactericidal activity is a potential surrogate marker for immunity after vaccination against invasive meningococcal diseases, its correlation with meningococcal carriage remains unclear.
We recently showed that most of the carriage isolates belong to CC198 and CC1157, with a very low proportion of carriage isolates in this study belonging to hyperinvasive lineages such as CC32, CC41/44, and CC11 (36), consistent with a previous report (25). Therefore, MenB vaccine antigens that are typically found in isolates of hyperinvasive lineages were detected in only a small proportion of carriage isolates. Previous studies indicated that 56 to 60% of invasive serogroup B isolates collected during 2000 to 2014 in the United States contain FHbp B/v1, with FHbp B24/1.1 included in MenB-4C being the most prevalent (33 to 34%) (37,38); this particular peptide was also present in the meningococcal outbreak strains from RI-1 (ST-9069) and OR (ST-32/CC32). In our study, FHbp B/v1 peptides were carried by 25% of genogroup B carriage isolates, with FHbp B24/1.1 being detected in 8% of genogroup B isolates. FHbp A05/3.45, included in MenB-FHbp, was detected in 9% of genogroup B isolates, while FHbp B01/1.55 was absent from all carriage isolates analyzed in this study. Similarly, these two FHbp peptides were either absent or present at a low prevalence among invasive isolates (37). Additionally, NhbA p0002, included in MenB-4C, showed greater prevalence among invasive N. meningitidis serogroup B isolates (9 to 11%) (37, 38) than among genogroup B carriage isolates (2%).
Overall, similar FHbp peptide distributions were observed among carriage isolates from both vaccinated and unvaccinated participants during the study period. Additional investigations over a longer period are warranted to adequately evaluate the vaccine impact on carriage.

MATERIALS AND METHODS
Data collection. Isolates included in this study were collected during February 2015 to March 2016 from meningococcal carriage evaluations at three U.S. universities (RI-1, RI-2, and OR) following vaccination using standardized methods as previously described (7,8,19). Briefly, four carriage evaluation rounds were conducted in conjunction with mass vaccination campaigns at RI-1 (February, April, and September 2015 and March 2016) and OR (March, May, and October 2015 and February 2016); two carriage evaluation rounds (March and April 2015) were conducted at RI-2, which is located in the same city as RI-1 and where no N. meningitidis serogroup B outbreak and response activities occurred. A total of 1,514 meningococcal isolates (650 from RI-1, 616 from OR, and 248 from RI-2) were recovered from 8,905 oropharyngeal swabs collected from 7,001 unique individuals (2,014 from RI-1, 3,509 from OR, and 1,478 from RI-2). N. meningitidis species was determined using real-time PCR targeting sodC (39), and serogroup was determined using slide agglutination serogrouping (SASG) (40,41). A total of 1,587 individuals (615 from RI-1, 613 from OR, and 359 from RI-2) participated in multiple evaluation rounds (repeat participants); 348 were identified to be N. meningitidis carriers, with 154 carrying N. meningitidis in only one round and 194 carrying N. meningitidis in at least 2 rounds (repeat carriers).
These evaluations were considered public health evaluations and did not require CDC institutional review for human subjects' protection. Both evaluations were covered under project determination numbers 2015-6436 and 2015-6442 for "Evaluation of Neisseria meningitidis serogroup B carriage in institutional settings during an outbreak." Molecular characterization. Bacterial genomes of all 1,514 confirmed meningococcal isolates were further characterized with whole-genome sequencing using Illumina platforms (HiSeq2500 or MiSeq; San Diego, CA). Illumina reads were trimmed with cutadapt (42) to remove adaptor sequences and reads below a quality score of 28 (Q28) and 75 bp. De novo short-read assembly was carried out using SPAdes 3.7.0 (43) with the "careful" option. The capsular genogroup of each isolate was determined based on serogroup-specific genes (44); isolates that contained capsule genes but lacked any identifiable serogroupspecific capsule gene were deemed to have an "undetermined genogroup." Isolates that lack the whole-capsule locus are defined as capsule null (cnl) (44). Genome sequence assemblies were used in subsequent BLAST searches (45) against the PubMLST Neisseria allele database (46) to determine clonal complex (CC)/ sequence type (ST), MenB vaccine antigens (FHbp, NhbA, and NadA), and fine typing peptides (FetA and PorA) (47). All unique peptides of MenB vaccine antigens, including PorA, were assigned PubMLST peptide allele identifiers (IDs) as described previously (12,48,49).
Data analysis. While 1,514 carriage isolates were obtained, only one isolate from each participant who carried the same meningococcal strain in more than one round (repeat carriers) was included in the analysis to assess the MenB vaccine antigen diversity and distribution among the carriage isolates circulating in the three universities, resulting in a total of 1,337 isolates.
To compare MenB vaccine antigen distribution between vaccinated (received $1 dose of MenB-FHbp or MenB-4C 14 days prior to sample collection) and unvaccinated participants, all isolates from repeat carriers were included even if they appeared to be the same strain genotype (CC:ST:PorA:FetA); however, participants with unknown MenB vaccination records were excluded from the analysis (92 isolates from RI-1 and OR), resulting in a total of 1,422 carriage isolates. For phylogenetic analysis of MenB vaccine antigens, multiple-sequence alignment of peptides was generated by ClustalW (50) with CLC Genomics Workbench 7. Networks were created by SplitsTree, v 4.0 (51), with default parameters.
Statistical analysis. Statistical analysis was performed using SAS version 9.4 (SAS Institute, Cary, NC). Fisher's exact test was used to assess changes in the distribution of FHbp subfamilies and peptides in isolates recovered from vaccinated and unvaccinated participants, stratified by university and whether the isolate was N. meningitidis serogroup B or non-N. meningitidis serogroup B. Isolates negative for both subfamilies A and B were excluded from this test due to being a low proportion (0.9%) of the overall isolate collection. Chi-squared goodness-of-fit tests were used to test for changes in distribution among N. meningitidis serogroup B and non-N. meningitidis serogroup B isolates. A Bonferroni correction was applied to correct for multiple comparisons. To assess proportions between vaccinated and unvaccinated repeat carriers who acquired a new meningococcal strain, a Fisher's exact test was performed; FHbp peptide distributions in all isolates and genogroup B isolates among these groups were analyzed using a likelihood ratio chi-squared test.
Data availability. Sequence reads are available under NCBI BioProject no. PRJNA533315.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. FIG S1, PDF file, 0.2 MB.

ACKNOWLEDGMENTS
We are grateful to states and local Health Departments who assisted in the collection of carriage specimens. We thank the Biotechnology Core Facility at Centers for Disease Control and Prevention (CDC) for their assistance in whole-genome sequencing of isolates included in this work. We also thank members of the CDC Meningitis and Vaccine Preventable Diseases Branch for their support and feedback and Mary Ann Hall for providing scientific editing.
The findings and conclusions in this report are those of the authors and do not necessarily represent the CDC's official position.