The Establishment and Diversification of Epidemic-Associated Serogroup W Meningococcus in the African Meningitis Belt, 1994 to 2012

Abstract

Epidemics of invasive meningococcal disease (IMD) caused by meningococcal serogroup A have been eliminated from the sub-Saharan African so-called "meningitis belt" by the meningococcal A conjugate vaccine (MACV), and yet, other serogroups continue to cause epidemics. Neisseria meningitidis serogroup W remains a major cause of disease in the region, with most isolates belonging to clonal complex 11 (CC11). Here, the genetic variation within and between epidemic-associated strains was assessed by sequencing the genomes of 92 N. meningitidis serogroup W isolates collected between 1994 and 2012 from both sporadic and epidemic IMD cases, 85 being from selected meningitis belt countries. The sequenced isolates belonged to either CC175 (n = 9) or CC11 (n = 83). The CC11 N. meningitidis serogroup W isolates belonged to a single lineage comprising four major phylogenetic subclades. Separate CC11 N. meningitidis serogroup W subclades were associated with the 2002 and 2012 Burkina Faso epidemics. The subclade associated with the 2012 epidemic included isolates found in Burkina Faso and Mali during 2011 and 2012, which descended from a strain very similar to the Hajj (Islamic pilgrimage to Mecca)-related Saudi Arabian outbreak strain from 2000. The phylogeny of isolates from 2012 reflected their geographic origin within Burkina Faso, with isolates from the Malian border region being closely related to the isolates from Mali. Evidence of ongoing evolution, international transmission, and strain replacement stresses the importance of maintaining N. meningitidis surveillance in Africa following the MACV implementation. IMPORTANCE Meningococcal disease (meningitis and bloodstream infections) threatens millions of people across the meningitis belt of sub-Saharan Africa. A vaccine introduced in 2010 protects against Africa's then-most common cause of meningococcal disease, N. meningitidis serogroup A. However, other serogroups continue to cause epidemics in the region-including serogroup W. The rapid identification of strains that have been associated with prior outbreaks can improve the assessment of outbreak risk and enable timely preparation of public health responses, including vaccination. Phylogenetic analysis of newly sequenced serogroup W strains isolated from 1994 to 2012 identified two groups of strains linked to large epidemics in Burkina Faso, one being descended from a strain that caused an outbreak during the Hajj pilgrimage in 2000. We find that applying whole-genome sequencing to meningococcal disease surveillance collections improves the discrimination among strains, even within a single nation-wide epidemic, which can be used to better understand pathogen spread.

Keywords: Africa; Neisseria meningitidis; disease outbreaks; epidemiology; evolution; meningitis; meningococcus.

FIG 1

Maximum-likelihood phylogeny of 86 CC11 N. meningitidis serogroup W isolates, rooted on NM3683. Branches with less than 85% bootstrap support are collapsed. Scale bar represents one substitution per 10,000 bases in the whole-genome alignment. Isolates are labeled with country and year of isolation if available. Isolates from the 2002 Burkina Faso epidemic and the 2000 Hajj-related Saudi Arabian outbreak are in boldface. Clades are labeled at their roots and described in the text. Subclade IVa is presented in Fig. 2.

FIG 2

Maximum-likelihood phylogeny of 55 isolates in subclade IVa, rooted on subclade IV. Branches with less than 70% bootstrap support are collapsed. Scale bar represents one substitution per 10,000 bases in the whole-genome alignment. Isolates are labeled with country and year of isolation, with Burkina Faso National Reference Laboratories that provided isolates listed in parentheses. Isolates from Burkina Faso 2012 are in boldface. Subclades mentioned in the text are labeled at their roots.

FIG 3

Frequencies of fHbp alleles in each subclade of CC11. Each bar represents a different subclade, colored by the proportion of isolates from that subclade that carry each allele, with the number of isolates written to the left. The bar for subclade IV does not include isolates from subclade IVa. Each allele encodes a different protein, and protein subvariants are listed in the legend.