Characterization of vaccine antigens of meningococcal serogroup W isolates from Ghana and Burkina Faso from 2003 to 2009

Neisseria meningitidis is a major cause of bacterial meningitis and a considerable health problem in the 25 countries of the ‘African Meningitis Belt’ that extends from Senegal in West Africa to Ethiopia in the East. Approximately 80% of cases of meningococcal meningitis in Africa have been caused by strains belonging to capsular serogroup A. After the introduction of a serogroup A conjugate polysaccharide vaccine, MenAfriVac ™, that began in December 2010, the incidence of meningitis due to serogroup A has markedly declined in this region. Currently, serogroup W of N. meningitidis accounts for the majority of cases. Vaccines based on sub-capsular antigens, such as Generalized Modules for Membrane Antigens (GMMA), are under investigation for use in Africa. To analyse the antigenic properties of a serogroup W wave of colonisation and disease, we investigated the molecular diversity of the protein vaccine antigens PorA, Neisserial Adhesin A (NadA), Neisserial heparin-binding antigen (NHBA) and factor H binding protein (fHbp) of 31 invasive and carriage serogroup W isolates collected as part of a longitudinal study from Ghana and Burkina Faso between 2003 and 2009. We found that the isolates all expressed fHbp variant 2 ID 22 or 23, differing from each other by only one amino acid, and a single PorA subtype of P1.5,2. Of the isolates, 49% had a functional nhbA gene and 100% had the nadA allele 3, which contained the insertion sequence IS1301 in five isolates. Of the W isolates tested, 41% had high fHbp expression when compared with a reference serogroup B strain, known to be a high expresser of fHbp variant 2. Our results indicate that in this collection of serogroup W isolates, there is limited antigenic diversification over time of vaccine candidate outer membrane proteins (OMP), thus making them promising candidates for inclusion in a protein-based vaccine against meningococcal meningitis for Africa.


Introduction
Neisseria meningitidis is a major cause of bacterial meningitis in the African Meningitis Belt 1 . Between 1993 and 2012, nearly 1 million suspected cases were reported with 100,000 deaths, and 80% of the cases were caused by serogroup A 2 . Following the introduction of the serogroup A polysaccharide conjugate vaccine MenAfriVac ™ in 2010, the incidence of group A disease decreased, but outbreaks of meningitis due to other meningococcal serogroups, in particular serogroup W, continue to occur 1,3 . Serogroup W was responsible for an epidemic of around 13,000 cases of meningitis in Burkina Faso in 2002 4 and contributed to a total of 639 deaths in 2012 in the same country 5 . Around 40% of infected people who develop sepsis die and survivors often suffer from limb loss, cognitive dysfunction, brain damage or visual impairment.
An approach towards developing a broadly-protective meningococcal vaccine for Africa is based on the use of subcapsular antigens included in GMMA (Generalized Modules for Membrane Antigens). GMMA are outer membrane blebs from bacteria genetically engineered to release large quantities of membrane vesicles, which are enriched in outer membrane proteins. Other strain modifications are included to increase safety and immunogenicity by the upregulation of immunogenic antigens 6,7 . GMMA from geneticallyengineered strains with up-regulated expression of meningococcal factor H binding protein (fHbp) have been shown to provide broad protection against African meningococcal isolates from different serogroups 7 . Other outer membrane antigens that have been shown to induce the production of bactericidal antibodies include PorA, Neisserial adhesin A (NadA) 8 and Neisserial heparin-binding antigen (NHBA) 9 .
To help determine the potential coverage of these antigens in a GMMA-based vaccine for Africa, we investigated their genetic diversity in serogroup W carriage and disease isolates from Burkina Faso and Ghana collected between 2003 and 2009. These two countries have suffered repeatedly from meningococcal meningitis outbreaks 4,10 . Focusing on isolates collected over a period of years from a defined geographic region provides the opportunity to monitor the dynamics, variation and diversity of surface-exposed antigens over time.  Table 3. PCR products were separated by gel electrophoresis using a 0.8% Tris base, acetic acid and EDTA (TAE) agarose gel (BioRad Laboratories, Hercules, USA). PCR products were purified using the PureLink PCR Purification Kit (Invitrogen) according to the manufacturer's instructions. The DNA amount was measured using the NanoDrop ND-1000 Spectrophotometer.

DNA sequencing
The primers used for porA VR1 sequencing were 210 and 103L (  14 and aligned for the construction of phylogenetic trees using the maximum likelihood method with the general time-reversible model of evolution and correction for partial deletion of gaps (GenBank accession numbers: nadA1 FJ619641.1; nadA2 GQ302859.1; nadA3 JN166979.1; nadA4 FJ619644.1; nadA5 FJ619645.1). All trees were un-rooted.  Western blot analysis of fHbp expression in whole cell lysates For 17 isolates labelled with an asterisk in Table 1, Western blot analysis of the fHbp expression level in whole cell samples was performed as described by Seib et al. 15 . The strains were sub-cultured on GC agar and incubated overnight at 37°C with 5% CO 2 . A 7 mL aliquot of Mueller-Hinton broth (Becton Dickinson, Franklin Lakes, NJ, USA) supplemented with 0.25% glucose (Sigma-Aldrich) was inoculated with single colonies to an optical density at 600 nm (OD 600 ) of 0.12-0.16. The suspensions were incubated at 37°C with 5% CO 2 to an OD 600 of 0.6 corresponding to approximately 1.8×10 8 cfu/ml (exponential growth phase). The cells from 1 mL of culture were collected by centrifugation at 17,900 g for 5 min in a microcentrifuge (Eppendorf), re-suspended in 100 μL phosphate buffered saline (PBS) and heat inactivated in a water bath at 56°C for 1 hour. Protein concentrations of the lysates were determined using a Lowry protein assay kit (BioRad Laboratories, Hercules, USA) with bovine serum albumin (Sigma-Aldrich) as a standard.

Results
The serogroup W isolates studied are homogenous with respect to fHbp and porA sequence variants PorA is an immunodominant antigen in N. meningitidis, but multiple subtypes exist with little cross-protection between meningococci expressing different PorA subtypes. fHbp can be divided into three antigenic variants, each of which is divided into sub-variants. Individual sequences are classified by a peptide ID number. Within each variant group, cross-protection is observed 18 . From the typing analysis of the fHbp and porA genes, all serogroup W isolates tested expressed fHbp variant 2, ID 22 (isolates from Burkina Faso) or 23 (isolates from Ghana), which differ by one amino acid, and PorA subtype P1.5,2 ( Figure 1 and Supplemental File). Despite the limited number of isolates studied, these results suggest that between Burkina Faso and Ghana, which share a common border, there has been conservation of fHbp and PorA antigens among W isolates over a period of seven years.

The serogroup W isolates studied have intermediate or high fHbp expression
The level of fHbp protein expression can affect susceptibility of meningococci to anti-fHbp antibodies. High expressers of fHbp are generally more susceptible to killing than low expressers 19 . We measured fHbp expression in 17 isolates. We selected 4 out of 8 (50%) strains from Burkina Faso and 13 out 23 (56%) strains from Ghana for fHbp expression analysis. These were selected to cover isolates from different years including the oldest and newest strains. Within this group of strains selected, n=2 (50%) of the strains from Burkina Faso and n=7 (53%) of the strains from Ghana were case isolates, while the remainder were carriage isolates. We prepared whole cell extracts of the serogroup W test strains and the serogroup B reference strain and compared fHbp levels with defined amounts of a fHbp v.2 protein standard by Western blot and densitometry measurement (Dataset 1). Expression level of the reference serogroup B strain 8047 was set to 100% and levels of expression of the serogroup W strains were compared with the reference strain. Isolates with means below 33% of the reference strain were classified as low expressers while isolates with expression above 100% were categorized as high expressers. Those with mean fHbp expression between 33-100% were considered intermediate expressers.
The expression of fHbp among the W isolates was variable, ranging from 50-152%, compared to the reference serogroup B strain 8047, with 41% of the isolates expressing equal or higher levels of fHbp compared to the reference strain ( Figure 2). There was no significant difference in fHbp expression between case and carrier isolates studied (P=0.74, Mann Whitney U test). This indicates that levels of fHbp protein on the bacterial surface can vary among strains collected from a relatively small region and expressing the same fHbp ID.
The genomes of most serogroup W isolates studied contain nadA NadA and NHBA induce the production of bactericidal antibodies against N. meningitidis serogroup B strains. Wang et al. found that nadA was not present among a small number (n=13) of W isolates tested as part of an analysis of 896 serogroup B, C, Y and W isolates from the USA, while nhbA was present in 92% of W isolates 20 . Among the African W strains investigated in this study, the nadA allele 3 was present in 26/31 (84%) of isolates (Figure 1). Among the remaining five W isolates (1 case, 4 carrier isolate), PCR amplification across the nadA site gave a 2 kb product instead of the expected 1 kb product. Western blotting using whole cell lysate and polyclonal mouse anti-NadA allele 3 antibody indicated that these isolates did not express NadA (Dataset 1). Sequencing of this fragment confirmed the presence of the insertion sequence IS1301. This 842-bp mobile genetic element is known to cause a number of effects including insertions and deletions that result in silent mutations, knock-out of gene expression or regulation of downstreamlocated genes. For example, insertion of IS1301 into the capsular siaA gene mediates loss of encapsulation resulting in increased adherence and entry of meningococci into epithelial cells 21,22 . 17 out of 21 carrier (81%) and 9 out of 10 (90%) case isolates had a nadA gene. Previous reports found that nadA is present in about 50% of group B case isolates, but underrepresented in carrier isolates 23 . NhbA was present in 30/31 (94%) of the meningococcal isolates studied. However, genetic sequencing in these isolates revealed a stop codon for 15/30 isolates, which has not previously been reported. The alleles of the remaining strains were identified as allele 17 (Supplemental File) using the Neisseria typing database available at http://pubmlst.org/neisseria/NHBA/.

Discussion
Since the introduction of a meningococcal A polysaccharide conjugate vaccine MenAfriVac ® in the African Meningitis Belt, outbreaks of meningitis caused by non-serogroup A meningococci, particularly W, are occurring with increased frequency. The development of a protein-based vaccine that can provide broad protection is an attractive prospect.  19 . It has been suggested that sparse distribution of antigens on the bacterial surface impedes cross-linking of two IgG anti-fHbp antibodies to correctly spaced epitopes 26 . Consequently, the antibodies cannot engage the complement protein C1q, preventing activation of the classical complement pathway. In the present study, the serogroup W isolates were found to express medium to high levels of fHbp when compared with a serogroup B strain known to express naturally relatively high levels of fHbp 17 , and there was no significant difference in expression between case and carrier isolates. This, together with the conservation of the fHbp ID among carrier and case isolates indicates that both carrier and case isolates could be targets of vaccine-induced anti-fHbp antibodies. NadA has emerged as an important protein for adhesion and invasion, and has been shown to elicit bactericidal antibodies 8 . In this study, the presence of the nadA gene in most case and carrier strains isolated from African countries suggests that NadA could be a potentially important vaccine antigen to be included in a GMMA vaccine for Africa.

Conclusion
This longitudinal study of meningococcal serogroup W isolates from two African countries, together with the findings of other studies, suggests that there is limited antigenic variation of meningococcal outer membrane proteins that induce bactericidal antibodies. These findings support a strategy of using protein-based vaccines, such as GMMA, to prevent meningococcal meningitis in Africa caused by serogroup W. Author contributions OK and CAM conceived the study. OK designed the experiments. EI carried out the research. EI, OK and CAM wrote the manuscript. AH, AS and GP provided the group W isolates studied. All authors were involved in the reviewing of the manuscript and have agreed to its final content.

Data availability
Competing interests OK and CAM are both employees of the Novartis Vaccines Institute for Global Health. CAM has received grant support from GlaxoSmithKline.

Grant information
This work was supported by an EU FP7 Marie Curie Actions Industry Academia Partnerships and Pathways (IAPP) Consortium Programme, entitled GENDRIVAX (Genome-driven vaccine development for bacterial infections).

Georgina Tzanakaki
National Meningococcal Reference Laboratory, National School of Public Health, Athens, Greece In this manuscript, Ispasanie describes the characterization of vaccine antigens of meningococcal et al. serogroup W isolates from both carriers and patients in Ghana and Burkina Faso.
The work seems very interesting and adds up to the previous knowledge with information on the limited antigenic variation of meningococcal OMPs that induce antibodies. The manuscript is well written and the methodology is well defined.

Minor comments:
The authors state in the discussion section, that after the introduction of the meningococcal A vaccine outbreaks caused particularly by serogroup W are occurring with increased frequency. To the reader's surprise, only 31 isolates were studied (21 carrier and 10 patient strains) within a seven year period (2003)(2004)(2005)(2006)(2007)(2008)(2009). Maybe the authors would consider adding more information of the numbers of serogroup W incidence or number of cases in both countries and the reason why they choose only this limited number for characterization of the vaccine antigens.

I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. Competing Interests: