Characterization of meningococcal carriage isolates from Greece by whole genome sequencing: Implications for 4CMenB vaccine implementation

Herd protection, resulting from the interruption of transmission and asymptomatic carriage, is an important element of the effectiveness of vaccines against the meningococcus. Whilst this has been well established for conjugate polysaccharide vaccines directed against the meningococcal capsule, two uncertainties surround the potential herd protection provided by the novel protein-based vaccines that are used in place of serogroup B (MenB) polysaccharide vaccines (i) the strain coverage of such vaccines against carried meningococci, which are highly diverse; and (ii) the generation of a protective immune response in the mucosa. These considerations are essential for realistic estimates of cost-effectiveness of new MenB vaccines. Here the first of these questions is addressed by the whole genome sequence (WGS) analysis of meningococci isolated from healthy military recruits and university students in Greece. The study included a total of 71 MenB isolates obtained from 1420 oropharyngeal single swab samples collected from military recruits and university students on voluntary basis, aged 18–26 years. In addition to WGS analysis to identify genetic lineage and vaccine antigen genes, including the Bexsero Antigen Sequence Type (BAST), the isolates were examined with the serological Meningococcal antigen Typing System (MATS) assay. Comparison of these data demonstrated that the carried meningococcal population was highly diverse with 38% of the carriage isolates showed expression of antigens matching those included in the 4CMenB vaccine. Our data may suggest a limited potential herd immunity to be expected and be driven by an impact on a subset of carriage isolates.


Introduction
Neisseria meningitidis, the meningococcus, despite its propensity to cause invasive meningococcal disease (IMD) principally meningitis and septicaemia worldwide, is an obligate PLOS ONE | https://doi.org/10.1371/journal.pone.0209919 December 28, 2018 1 / 13 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 In the early 21 st century, the proportion of MenB meningococci recovered in carriage studies ranged from 2.2% to 43.9% in Europe [3], with a study in Greece showing that MenB was predominant (34.9%) among groupable carriage isolates [14]. The aim of the present study was the characterisation of the 71 asymptomatically carried capsule group B (MenB) isolates obtained from healthy young adults by serogrouping/genogrouping, whole genome sequencing (WGS), and Meningococcal antigen Typing System (MATS) assay which is based on the serum antibody bactericidal activity rather than mucosal immunity, in order to identify the meningococcal genotypes circulating among military recruits and University students. Data on serogroup, sequence type (ST), antigen variability, Bexsero Antigen Sequence Type (BAST) and the MATS results established baseline data and indicated the possible impact of 4CMenB immunization on asymptomatic transmission in this setting.

Isolation and identification of N. meningitidis
The isolates examined were the serogroup B subset of isolates recovered from healthy young adults (military recruits or university students) enrolled in a previous carriage study [14]. All participants -in addition to the approval by the Ethics Commitee from the National School of Public Health-responded ad hoc to a structured self-administrated questionnaire and a written informed consent form was signed as previously described [14]. The pharyngeal swabs were immediately plated on New York City medium (OXOID LTD, Basingstoke, Hampshire, England) and incubated at 37˚C in the presence of 5% CO 2 . Cultured plates were examined at 24 and 48 hours for suspected Neisseria colonies. Identification procedures included Gram stain, oxidase test, and a rapid carbohydrate utilization test. All N. meningitidis identified colonies were stored at -70˚C in Heart Infusion Broth with 20% glycerol. In addition, the supernatant from a heat killed cell suspension was prepared and stored at -20˚C using 1μl of overnight culture suspended in 200μl of PCR grade water, vortexed, heated at 100˚C for 10 min, and centrifuged at 20,000 g for 12 min for further confirmation for the presence of the porA gene, by PCR amplification as described previously [15].

Capsular identification
Serogroup and genogroup was determined for all isolates. Serogrouping was carried out by slide agglutination test (Remel Europe Ltd. UK) according to manufacturer's instructions and genogrouping determined by the implementation of a multiplex PCR targeting specific capsule group genes, as described previously [16].

Molecular characterization (MLST, PorA, FetA)
All 71 MenB isolates along with 56 of 127 meningococcal isolates belonging to a genogroup, were characterised by 'finetyping' (MLST and PorA and FetA typing), as described previously [17] using the PubMLST.org/neisseria database (http://pubmlst.org/neisseria/) [14]. Sequence types (ST) were defined and grouped into Clonal Complexes (ccs). PorA genotyping for variable regions 1 and 2 (VR1 and VR2) was performed as described previously [15] and compared with the variable sequences in the Neisseria PorA database (http://pubmlst.org/neisseria/PorA/ ). Similarly, the FetA variable region was also obtained for all the typable isolates, as previously described [18], and compared with variable sequences in the Neisseria FetA database (http:// pubmlst.org/neisseria/FetA/). Genomic DNA extraction was performed using GenElute Bacterial Genomic DNA Kit (SIGMA, Germany) following the manufacturer's instructions. Briefly, one microliter of 18-hour culture was suspended in 180 μl of Lysis Solution T and the extraction included the optional RNase A treatment step. The eluate was stored at -20˚C until sequencing. Whole genome libraries were created using Nextera XT DNA Library Preparation Kit (Illumina FC-131-1096), using Double Indexing Strategy (Nextera XT Index Kit v2 FC-131-200x) according to the manufacturer's instructions. Each genome assembly was annotated using the pubMLST. or/neisseria sequence definition database and included the loci defining the MLST, BAST, antigen finetyping, and cgMLST v1.0 schemes.

Meningococcal antigen typing system (MATS)
To determine the proportion of strains expected to be covered by 4CMenB all isolates were analyzed by MATS ELISA. MATS ELISA was carried out at the Meningococcal Reference Laboratory, Institut Pasteur (Paris, France), one of the accredited reference laboratories that participated in the MATS standardization process [20,21]. MATS ELISA values were calculated as antigen-specific relative potencies compared with MenB reference strains expressing each vaccine antigen [20,22].
Predicted coverage using MATS-PBT (Positive Bactericidal Threshold) was calculated as described previously using the threshold that was established for invasive isolates [20,21,22]. The presence of at least one antigen with a relative potency greater than its MATS-PBT relative potency value (0.012 for fHbp, 0.294 for NHBA and 0.009 for NadA) or the presence of PorA VR2 P1.4 (the VR2 present in the OMV-NZ component of 4CMenB) was considered sufficient for an isolate to be covered by 4CMenB. Strains that did not meet these criteria were considered not covered. Estimates of the 95% confidence intervals (95% CI) for the MATS-PBTs were derived on the basis of overall assay repeatability and reproducibility (0.014-0.031 for fHbp, 0.169-0.511 for NHBA, 0.004-0.019 for NadA) [22]. These intervals defined the 95% CI of strain coverage by 4CMenB.

Genetic diversity and association analysis
The diversity of each vaccine component was assessed using the Simpson's index of diversity (D) [19,23] and was calculated for each antigen using the Comparing Partitions Online Tool (http://www.comparingpartitions.info/index.php?link=Tool). A D index value near one indicated high diversity and values <1 indicated lower diversity. Cramer's V coefficient was used to assess the association of the BAST antigenic variants with clonal complex, using SPSS version 20, with values from 0, indicating no association, to 1, indicating complete match.

Capsular typing
Among the 180 N. meningitidis isolates recovered from 1420 pharyngeal swabs, 71 were identified as MenB either by slide agglutination test (45/71, 63.4%) or PCR (26/71; 36.6%) and included in the present study for further analysis. All 71 isolates possessed genes encoding the group B capsule (genogroup B). A complete csb gene (NEIS2161) sequence was assembled in 64 of the 71 (90%) isolates. In the remaining seven isolates an identifiable csb gene was present, although the cps region was incompletely assembled. Among the 26 capsule genogroup B PCR positive isolates not expressing serogroup B phenotypically, 18 (69%) presented sequence insertions or deletions switching off the gene expression (genetically phase variable 'off'), as were four isolates identified as serogroup B by the slide agglutination test.

Discussion
At the time of writing IMD, particularly that caused by serogroup B meningococci, was incompletely controlled by immunisation, largely because of the antigenic and genetic diversity of Neisseria meningitidis. In particular, there were uncertainties as to the ability of the proteinbased vaccines to generate herd protection (immunity), especially given the diversity of meningococci meningococci isolated from asymptomatic carriers [1].
Previous studies have shown that the introduction of monovalent meningococcal group C polysaccharide conjugate vaccine (MCC) in national immunization programs reduced MenC IMD through herd protection, by reducing the transmission of the epidemic strain among asymptomatic carriers [11,12]. Similarly, the national introduction of a strain-specific outer membrane vesicle vaccine (MeNZB) in New Zealand reduced IMD cases from the epidemic clone, with limited evidence for cross-protection [26] with a potential impact on the acquisition of carriage of the epidemic strain [27]. Similar impact on the acquisition of carriage of an epidemic isolate was also suggested for another OMV vaccine (MenBVac) [28]. The impact of meningococcal vaccines on asymptomatic transmission is an important aspect in the evaluation of the impact of new meningococcal vaccines on disease, particularly their cost-effectiveness [29]. Recently, a multicomponent protein-based vaccine (4CMenB, Bexsero), designed to target MenB disease in the absence of a vaccine against the group B capsule, was licensed in many European countries including Greece; however, a carriage study in UK has shown that the potential broad additional effect on commensal Neisseria and non-disesae associated N. meningitidis cannot be predicted [14]. While no effect on carriage was observed 1 month after completion of the vaccine course, this effect was from 3 months after dose two of 4CMenB. This effect was observed for all meningococci but not specifically of serogroup B isolates. This global effect could be expected as this vaccine does not target the capsule but surface exposed proteins that can be shared by the isolates regardless the serogroup. However, an effect on serogroup B should be analysed on the basis of the expression data (e.g. MATS as performed in our study) but that still require a "correlate of protection" for carriage isolates.
No correlate of protection against acquisition of carriage at the mucosal surfaces was available to this study, and this prevented this issue being addressed directly, by assaying the mucosal immunity against meningococcal isolates; however, data from animal models have demonstrated the detection of immune response in the mucosal secretions for protein-based vaccines, which correlated with the prevention against intranasal colonization by isolates showing multiple matching with vaccine antigen [28,30]. Our data clearly show that the expression of antigens of the 4CMenB vaccine in carriage isolates as suggested by the MATS data.
Here, the 4CMenB vaccine antigens were assessed in a collection of MenB isolates obtained in Greece from asymptomatic carriage in young adults aged 17-26 years. The study included isolates obtained immediately before and after the introduction of the 4CMenB vaccine in Greece; however, this initial implementation in Greece was unlikely to have affected carriage in the age group sampled. Consequently, this study represented a pre-vaccine meningococcal carriage sample in an age group with high carriage [14,31]. It provides a reference sample for that will be essential for: (i) assessing the potential impact of the 4CMenB vaccine on carriage; (ii) monitoring the effect of vaccine use on antigen evolution among carried meningococci; and (iii) analyzing any genomic changes in the meningococcal population circulating in Greece. The application of WGS, in combination with the MATS assay, provided a combination of genotypic and phenotypic characterisation for these isolates. However, a correlate of protection against acquisition of carriage is still lacking to evaluate the WGS/MATS data from carriage isolates.
Among the isolates that possessed a cps region encoding the group B capsule (genogroup B, MenB), over a third (n = 26, 36.6%) were defined as phenotypically non-groupable (NG) by agglutination, suggesting down-regulation of capsule expression during carriage. This is consistent with many previous studies including a recent investigation in Italy [10], where 40% of group B carrier isolates were NG. A total of 21 (81%) of the NG isolates exhibited intact capsule operons, with 18 (69%) containing a csb gene predicted to be phase variable off. This demonstrates the importance of using WGS data to investigate such isolates, as this is the most practical and cost-effective means of determining the presence and expression status of the capsule operon. The majority of genogroup B isolates analysed in this study were represented by clonal complexes 41/44cc, 35cc, 32cc, 213cc, 269cc, and 162cc. This was also consistent with previous observations from a variety of high-income countries [10,32]. As the relationship between asymptomatic carriage and the development of invasive disease remains incompletely understood, it is important to collect isolates belonging to hyperinvasive genotypes worldwide. The peptide sequences of the principal 4CMenB vaccine antigens, summarized by the BAST type, were extracted from the assembled WGS data using the tools integrated into the https://pubmlst.org/neisseria/ [19]. As has been reported previously for disease isolates [19,33], each of the vaccine antigens exhibited extensive sequence variation among the 71 genogroup B isolates, with NHBA being the most diverse antigen (" Table 1"). The strong correlation of cc with both BAST and individual antigen variant, such that in the absence of antigen data cc could be used as a surrogate for likely cross protection, was consistent with several studies of invasive disease isolates [19,33,34]. Similarly consistent with studies of IMD isolates, exact matches to the vaccine antigen variants were rare: PorA VR2 (P1.4) was found in only one isolate belonging to 41/44 cc; NHBA peptide 2 was found in five isolates (7.6%); the fHbp target peptide 1 and NadA target peptide 8 were not present in this collection of isolates. The prevalence of PorA VR2 P1.4 (1, 1.4%) was considerably lower to that was found among invasive isolates (7%) in Greece [35]. Consequently, any impact of the 4CMenB vaccine on carriage isolates will depend on immunological cross protection of the type assessed in the MATS assay.
All three fHbp variant families were present among the 71 carriage isolates. Consistent with the findings of other carriage studies, variant family 2 was more abundant than variant family 1 [10,32]. No cross reactivity of the 4CMenB vaccine antibodies with isolates containing variant family 2 or 3 peptides, was observed ("S1 Table"); however, 90% of the isolates containing fHbp variant family 1 peptides were MATS RP positive ("S1 Table"). Similar data indicating the expression of fHbp family 1 among 95% of carriage isolates harbouring this variant were reported from France using ELISA [36]. There were 26 different peptides for NHBA, and only 5 isolates (7.7%) contained the 4CMenB vaccine variant, peptide 2. The presence of the NHBA peptide 2 among group B meningococci from carriers is 30%, lower than the presence of this peptide among IMD group B isolates in Greece (10.1%) [35]. This is in contrast with a recent carriage study from Spain, which found comparable levels of NHBA peptide 2 among carriers and IMD associated group B isolates of 3% and 4% for carrier and invasive group B isolates respectively [32].
While BAST data demonstrated a high diversity of the fHbp, NHBA, NadA, and PorA antigens, this was mainly due to fHbp and NHBA, and a limited number of BASTs occurred at high frequency [37]. BAST-1 (fHbp-1, NHBA-2, NadA-8, PorA P1.7-2,4), corresponding to the 4CMenB vaccine components, was absent from this dataset which was consistent with the fact that the vaccine formulation was assembled from meningococci from multiple ccs. MATS estimated the overall potential 4CMenB vaccine coverage at 38.6%, which was substantially lower than the estimated coverage for MenB IMD isolates (89.2%) from Greece for 1999-2010 [35]. A total of eight isolates (11.4%) predicted to be covered by MATS contained capsule group B-encoding cps regions but were non-groupable (NG) by serological methods. One limitation of this study was that no correlation had been established between the level of antibodies needed for protection at the mucosal level and the levels of expression of the antigens targeted by the 4CMenB.
In conclusion, this study demonstrates the utility of WGS data in the characterisation of meningococcal carriage isolates to assess the prevalence of vaccine antigens, including the presence and expression of polysaccharide and protein antigens. The BAST antigens found among carried isolates were highly diverse, which was reflected by a low level of predicted cross-protection of 4CMenB against carried MenB isolates. As the majority of carriage MenB isolates do not belong to hyperinvasive linages, this observation does not necessarily preclude the use of 4CMenB to generate herd immunity against IMD. To affect disease rates, it will be important to disrupt the transmission of hyperinvasive meningococci; indeed, it may well be preferable that the transmission of less invasive variants of any meningococcal group is not impacted by immunisation. At the time of writing, a number of large-scale vaccination and carriage investigations were underway and the data generated from these, using approaches similar to those reported here, will provide a definitive answer as to the impact of this vaccine on meningococcal transmission.