Meningococcal B vaccine antigen FHbp variants among disease-causing Neisseria meningitidis B isolates, Italy, 2014–2017

Background Typing of Neisseria meningitidis isolates is crucial for the surveillance of invasive meningococcal disease (IMD). We performed a molecular epidemiology study of N. meningitidis serogroup B (MenB) causing IMD in Italy between 2014 and 2017 to describe circulating strains belonging to this serogroup, with particular regards to the two factor H-binding protein (FHbp) subfamilies present in the bivalent MenB vaccine. Materials and methods A total of 109 culture positive and 46 culture negative MenB samples were collected within the National Surveillance System (NSS) of IMD in Italy and molecularly analyzed by conventional methods. Results Overall, 71 MenB samples showed the FHbp subfamily A and 83 the subfamily B. The subfamily variants were differently distributed by age. The most frequent variants, A05 and B231, were associated with cc213 and cc162, respectively. All MenB with the FHbp A05 variant displayed the PorA P1.22,14 and 85.7% of them the FetA F5-5. The majority of MenB with the FHbp B231 variant showed the PorA P1.22,14 (65.4%) and 84.6%, the FetA F3-6. Conclusion MenB circulating in Italy were characterized by a remarkable association between clonal complex and FHbp variants, although a high degree of genetic diversity observed over time. A dynamic trend in clonal complexes distribution within MenB was detected. Our results stress the importance of continued meningococcal molecular surveillance to evaluate the potential vaccine coverage of the available MenB vaccines.


Introduction
In Italy, invasive meningococcal disease (IMD), which includes septicemia, meningitis, or both, develops in 0.28 persons/100,000 population in 2018 (http://old.iss.it/mabi/). Patients with IMD may have nonspecific symptoms early in the disease, but their condition can deteriorate rapidly. Therefore, vaccination represents the optimal strategy for the prevention of IMD [1,2]. N. meningitidis serogroup B (MenB) today represents the main serogroup circulating in several European countries [3] including Italy (http://old.iss.it/binary/mabi/cont/Interim_ Report_2018_finale.pdf) given the widespread use of recommended meningococcal C vaccination in the country [4]. The factor H-binding protein (FHbp), also referred as GNA1870 (Genome-Derived Neisseria Antigen 1870) or LP2086 (lipoprotein LP2086) [5][6][7], included in the 4CMenB vaccine [8,9] and in the bivalent MenB vaccine [10,11], induces serum bactericidal antibodies. FHbp, due to its ability to specifically bind to factor H of the human complement regulatory protein and to inhibit the alternative complement pathway [7,12], may improve the survival of N. meningitidis in human blood. Furthermore, FHbp, based on the variation of the amino acid sequence, can be divided into two subfamilies, A and B [10,13] with the 83%-99% sequence identity within and the 60-75% between subfamilies, while by other authors FHbp is classified in three variants (v.1, v.2 and v.3) [5].
The FHbp variants included in the MenB bivalent vaccine are B01 (v1.55) and A05 (v3.45); whereas, B24 (v1.1) variant is contained in the 4CMenB vaccine [7]. Several studies are ongoing to evaluate the ability of new MenB vaccine formulations to prevent the carriage status among adolescents and adults as a major contributing factor to herd protection in a population [14,15].
With the growing public health concern associated with meningococcal disease caused by MenB, it is important to investigate the molecular epidemiology of disease-causing N. meningitidis strains and to identify variants of FHbp from those N. meningitidis belonging to different clonal complexes.
In the present study, we investigated the two FHbp subfamilies present in the MenB bivalent vaccine. The percentage, the distribution, and the diversity patterns of FHbp variants were assessed among culture positive and culture negative MenB causing IMD in Italy from 2014 to 2017. The sequence type (ST), clonal complex (cc), PorA and FetA types and the age of the patients, were also included.

Surveillance of invasive meningococcal disease
The National Surveillance System (NSS) for IMD is based on mandatory reporting to the Ministry of Health and to the Italian Institute of Public Health (Istituto Superiore di Sanità, ISS, http://www.iss.it/mabi). The ISS, as National Reference Laboratory (NRL) is the coordinator of the NSS, as already reported [16].

Ethics statement
For this study, the samples were in respect of the ethical requirement and no patient identification information was presented in the study.
antisera (Thermo Scientific, Waltham, Massachusetts, USA) or by multiplex PCR on bacterial DNA extracted using the QiAmp mini kit (Qiagen, Hilden, Germany) [17], from an overnight culture or directly from clinical sample.

Whole genome sequencing
For each culture positive sample, 1 ng of DNA was used for preparing libraries following the Nextera XT DNA protocol in Illumina MiSeq platform (kit v3, 600 cycles). A first quality check of the raw sequence data was performed using FastQC [18]. Reads were trimmed using the software Sickle [19] to maintain a Q score >25, and de novo assembly was carried out with the ABySS software version 1.5.2 (k parameter = 63) [20]. Contigs longer than 500 bp were selected using an ad hoc script and kept for further analysis.

Multilocus sequence typing (MLST), PorA, FetA and FHbp analysis
For culture positive samples, multilocus sequence typing (MLST), PorA, FetA and FHbp typing were identified in silico using WGS data through PubMLST.org database (http://pubmlst. org/neisseria/). For culture negative samples, seven housekeeping genes (abcZ, adk, aroE, fumc, pgm, pdhc and gdh) together with porA, fetA and fHbp genes were amplified referring to PCR conditions available on the PubMLST.org database or using primers and amplification parameters as already described [21]. PCR was performed using Veriti 96 well instrument (Applied Biosystem, Foster City, USA) or Mastercycler personal (Eppendorf, Hamburg, Germany). Multiple sequence and amino acid alignments were performed using Chromas version 1. 15 and Clustal Omega web-site (https://www.ebi.ac.uk/Tools/msa/clustalo/). The PubMLST.org database was used to identify allele types, the FHbp subfamilies A and B and their variants and to submit new sequences for fHbp gene. Moreover, the PubMLST.org database was used to submit new alleles of MLST loci and STs.

Meningococcus of serogroup B causing invasive disease
In the period 2014-2017, 174 MenB samples were sent to ISS within the NSS, of which 155 were analyzed: 19, in fact, were unsuitable for molecular analyses due to limited sample volume or to a low DNA concentration. Among the 155, 109 were culture positive and 46 culture negative. Seventy-three were from IMD cases presenting meningitis and 33 septicaemia, 21 meningitis and septicaemia. For 28 samples, the data was not available.

FHbp subfamilies A and B
FHbp subfamily A in 71 MenB and FHbp subfamily B in 83 MenB were identified. One sample resulted a hybrid, FHbp A/B subfamily.
B231 was identified through the period with a peak in 2016 (n = 12; Fig 2); all of them were associated with the cc162 (n = 26; Fig 3). One culture negative sample was not enough to obtain the cc result. The B03 variant was detected up to 2016, whilst B24 from 2014 to 2017 with the higher number in 2016 (n = 4; Fig 2). B03 and B24 variants associated mostly to cc41/ 44 (n = 9) and to cc32 (n = 5), respectively (Fig 3).
Within the subfamily B, FHbp variant B257 was here identified and deposited in the Neisseria org web site (http://pubmlst.org/neisseria/). Fig 4, panel I and II, shows the distribution of the subfamily A and B variants by age groups. The age groups were those indicated in the IMD NSS report (http://www.iss.it/mabi).

FHbp subfamily A and B variants by age groups
Variants represented by one sample (singleton) were not included. The majority of the subfamily A variants were detected among MenB collected from young adults and adults (age group 25-64 years, n = 29; 40.8%), including the main variants A05 (n = 5), A06 (n = 7) and A22 (n = 5). The second large group of subfamily A variant (n = 13; 18.3%) was identified among MenB samples from infants less than 1 year of age followed by children (age group 1-4 years, n = 8; 11.3%) (Fig 4I).
The majority of the subfamily B variants were detected among MenB collected from adolescents and young adult (age group 15-24 years, n = 23; 27.7%). In this age group, 7 of 23 showed the B24 and 5 of 23 the B231 variants.

FHbp subfamily A and B variants by PorA, FetA, cc and ST
The majority of MenB with the A22 variant belonged to cc41/44 (n = 7; 70%) of which ST-414 as the main ST (n = 4; 40%), (Table 1I). Several PorA and FetA alleles were found in association with A22 variant.

Discussion
Neisseria meningitidis of serogroup B represents an major challenge in the prevention and control of invasive meningococcal disease due to this serogroup [1, 2,4].
In this report, we provide the results of a molecular study conducted in Italy in a 4-years period, from 2014 to 2017, on the N. meningitidis strains of serogroup B causing invasive meningococcal disease and the vaccine antigen FHbp variants.
In Italy, in 2018, the incidence of MenB accounted for 0.12 cases per 100,000 inhabitants in the overall population, with a case fatality rate of 8.6%. A higher incidence of MenB was observed in the age group less than 1 year of age (1.53 cases per 100,000 inhabitants), in the age groups 10-14 (0.10 cases per 100,000 inhabitants) and 15-24 (0.25 cases per 100,000 inhabitants) years of age, data from the Italian National Surveillance System. An outbreak of MenB has described in Italy due to the switch from C to B of cc11 strain [22].
In this epidemiological situation due to its growing importance, vaccination against MenB should be considered. The National Immunization Plan (2017-2019) recommends this vaccination for infants less than one year of age and is offered free of charge [4]. In some Italian Regions, the recommendation of MenB vaccination has also been extended to other age groups (i. e. from 10 to 11 or from 11 to 12 or at 13 years of age) [23][24][25].
The main clonal complexes (cc s ) characterizing the MenB collected in Italy suggest a dynamic trend in their distribution. For example, an increase of MenB cc213 and cc162 has been observed, while MenB belonging to cc41/44 decreased from 40.7% in 2014 to 26% in 2017.
Overall, all MenB harbored the fHbp subfamilies A or B. Seventy-one (45.8%) MenB showed the FHbp subfamily A and 83 (53.5%) the subfamily B. One sample resulted a hybrid, FHbp A/B subfamily. The distribution of the FHbp subfamilies seems to be different from what described elsewhere. Recently, in Canada [26] 63% of MenB was associated with FHbp subfamily B as well as reported in the United States, Europe, New Zealand, and South Africa [27,28] where, combining all countries, the overall subfamily distribution was about 70% of subfamily B versus 30% subfamily A.
In this study, FHbp subfamily A was represented by 22 variants among the MenB analyzed; A05 (v3.45) variant was the most frequent (19.7%), followed by the A06 (14.1%) and A22 (14.1%). In particular, the A05 variant was the only variant identified among the meningococci belonging to cc213. As a reminder, this variant is one of two included in the bivalent MenB vaccine.
FHbp subfamily variants were distributed differently according to the age of the infected patients. MenB collected from young adults and adults (age group 25-64 years) showed a high percentage of subfamily A variants (40.8%), including major variants, A05, A06 and A22. A second large group of subfamily A variants was found between MenB from infants less than one year of age (18.3%) and children 1-4 years old (11.3%). Subfamily B appears to be more distributed across all age groups with slightly different in percentage by variant. Of note, the most frequent subfamily B variant, B231, was mainly present among MenB isolated from infant and children (52.9%).
Hereby, a remarkable association has been observed between clonal complexes and FHbp variants and as previously described [27,[29][30][31] in some cases also with PorA allele (e.g. 100% of A05 MenB belonged to cc213 and showed PorA P1.22,14 and the 100% of B231 MenB belonged to cc162). The opposite is not always true, as neither PorA nor MLST predict the FHbp variant. Furthermore, to strengthen the association between clonal complexes and FHbp variants, B24 was always identified with cc32 and the majority of B44 with cc269, as already reported [26,31].
The FetA variants were characterized by high antigenic diversity without any specific association with the FHbp variant.
One limitation concerns the analysis and subsequent evaluation of MenB as potentially susceptible to the MenB-FHbp vaccine as the results presented here are limited exclusively to the analysis of genetic data. In fact, the meningococcal antigen surface expression (MEASURE) assay to quantify FHbp expression, as a correlation to susceptibility to bactericidal killing [32], has not yet performed in the sample collection described. However, the presence of these variants among the Italian MenB and the sequence identity identify within the subfamilies (83%-99%) and across them (60-75%) [10,27], is an important result to be considered in the use of the bivalent MenB vaccine.
Although relatively rare, N. meningiditis infection has long been recognized as a significant public health concern due to the high rates of morbidity and mortality associated with invasive meningococcal disease. Furthermore, outbreaks due to this pathogen are not uncommon and MenB continues to be a significant cause of disease in adolescents and young adults. Dynamic changes may occur in some variants, highlighting the need for ongoing molecular surveillance to recognize the emergence and expansion of new clones, especially those with new variants, and to investigate and possibly predict vaccine failures.
To conclude, the main findings were: i) a high degree of genetic diversity within serogroup B meningococci over time; ii) the detection of one of two subfamily FHbp variants included in the bivalent vaccine; iii) most of the MenB were characterized by clonal complexes common through the world and by a correlation between clonal complexes and the FHbp variant.
The study highlighted the need for continuous surveillance of circulating MenB and to molecularly characterize the genes encoding vaccine antigens, i.e. FHbp variants, to follow temporal changes in their distribution for further evaluation of vaccine composition and vaccination programs.