Cellular and molecular biology of Neisseria meningitidis colonization and invasive disease

Abstract

The human species is the only natural host of Neisseria meningitidis, an important cause of bacterial meningitis globally, and, despite its association with devastating diseases, N. meningitidis is a commensal organism found frequently in the respiratory tract of healthy individuals. To date, antibiotic resistance is relatively uncommon in N. meningitidis isolates but, due to the rapid onset of disease in susceptible hosts, the mortality rate remains approx. 10%. Additionally, patients who survive meningococcal disease often endure numerous debilitating sequelae. N. meningitidis strains are classified primarily into serogroups based on the type of polysaccharide capsule expressed. In total, 13 serogroups have been described; however, the majority of disease is caused by strains belonging to one of only five serogroups. Although vaccines have been developed against some of these, a universal meningococcal vaccine remains a challenge due to successful immune evasion strategies of the organism, including mimicry of host structures as well as frequent antigenic variation. N. meningitidis express a range of virulence factors including capsular polysaccharide, lipopolysaccharide and a number of surface-expressed adhesive proteins. Variation of these surface structures is necessary for meningococci to evade killing by host defence mechanisms. Nonetheless, adhesion to host cells and tissues needs to be maintained to enable colonization and ensure bacterial survival in the niche. The aims of the present review are to provide a brief outline of meningococcal carriage, disease and burden to society. With this background, we discuss several bacterial strategies that may enable its survival in the human respiratory tract during colonization and in the blood during infection. We also examine several known meningococcal adhesion mechanisms and conclude with a section on the potential processes that may operate in vivo as meningococci progress from the respiratory niche through the blood to reach the central nervous system.

Figure 1. Schematic diagram showing structural organization of N. meningitidis LPS and some important determinants of immunotypes

The membrane-located part of meningococcal LPS comprises lipid A bound to two KDO and two heptose (HepI and HepII) moieties. Extended from the membrane are the structurally variable α- and β-chains of LPS. The different immunotypes of LPS are determined by variations both in the HepI α-chain extensions (Glc, Gal, GlcNAc, Gal and sialic acid) and the HepII β-chain extensions (GlcNAc, PEA and Glc). These arise as a result of phase variation in a number of genes shown and account for the antigenic variation of meningococcal LPS (see [48] for details). PEA may be added at position 3 or 6 on HepII by two distinct enzymes encoded by lpt3 and lpt6 [143]. The α-chain structures of the L3, L7 and L8 immunotypes are indicated. LNnT and silaylation: the immunotypes L7/L9, L2/L4/L5 contain identical α-chains terminating in LNnT (Gal-GlcNAc-Gal-Glc structure bound to HepI) which can be sialylated. The L7 immunotype, found in serogroup B and C isolates, refers to the immunotype lacking sialic acid, whereas L3 is sialylated. Phase variation of the lgtA gene gives rise to the immunotype L8, which cannot be sialylated [48].

Figure 2. Pili of N. meningitidis

(A) Transmission electron micrograph of negatively stained preparations of a meningococcal strain (M. Virji and D.J.P. Ferguson, unpublished work). Bundles of hair-like pilus filaments stretch several micrometres from bacterial surfaces (arrows). The scale bar represents 0.5 μm. (B) Schematic diagram indicating the relative cellular location of the gene products involved in pilus biogenesis. Several of these proteins (PilD, F, M, N, O and P) are implicated in the early stages of pilus synthesis, others (PilC, G, H, I, J, K, and W) may be necessary for functional maturation of the pilus. The pilus is extruded through the outer-membrane pore formed by PilQ. The remainder of the proteins play roles in pilus function. For instance, PilF and PilT (both inner-membrane-associated ATPases) appear to have opposing roles in pilus extension and retraction, and control pilus-associated functions, including twitching motility. PilX has been reported to be involved in bacterial aggregation and could have a role in colonization through promotion of microcolony formation by N. meningitidis (based on [60]). (C) Ribbon diagram of the three-dimensional structure of a pilin monomer of strain C311 (M. Virji and A. Hadfield, unpublished work) based on that of N. gonorrhoeae pilin [144]. The asterisk shows the structure and position of the unusual glycan modification (digalactosyl 2,4-diacetamido-2,4,6-trideoxyhexose) of the pilin [69]. Pili are made up of multiple pilin monomers stacked in a helical array, such that each helical turn is made up of five monomers (shown in a schematic representation of the fibre cross-section, bottom right).

Figure 3. Structures of the β-barrel outer-membrane proteins NspA and Opc

(A) The ribbon diagram shown represents an Opa-like eight-stranded β-barrel with four surface-exposed loops. The structure is derived from the co-ordinates of an Opa-like molecule, NspA [82], and was kindly provided by Professor Leo Brady (Department of Biochemistry, University of Bristol, Bristol, U.K.). The transmembrane domains are highly conserved between NspA and Opa proteins. However, the flexible surface loops are dissimilar and, in Opa proteins, the first loop is semivariable (SV), whereas the second and third loops are more extensively variable (designated hypervariable: HV1 and HV2). (B) Structure of N. meningitidis Opc protein presented as a ribbon diagram. Opc is a ten-stranded β-barrel presenting five largely invariant surface-exposed loops. This Figure was kindly provided by Professor Jeremy Derrick (Department of Biomolecular Sciences, University of Manchester Institute of Science and Technology, Manchester, U.K.) and is reprinted by permission from MacMillan Publishers Ltd: Nature Reviews Microbiology [23], copyright (2009) (http://www.nature.com/nrmicro/index.html).

Figure 4. Schematic overview of meningococcal interactions at the epithelial barrier of the nasopharynx and the mode of barrier penetration

Encapsulated (and possibly acapsulate) meningococci inhaled via respiratory droplets must first adhere to the epithelium within the nasopharynx to avoid removal by innate immune mechanisms such as mucus clearance. Pili extending beyond the capsule are considered to mediate the primary interaction with epithelial cells. Capsule down-modulation (or up-regulation of host receptors during inflammatory condition) allows interactions between outer-membrane proteins and their cognate host receptors. For example, Opa proteins may bind to CEACAMs and HSPGs, and Opc proteins can interact with HSPGs and, via vitronectin and fibronectin, to their integrin receptors. Although some minor adhesins such as NhhA have been shown to interact with HSPGs, the receptors targeted by MspA, App and NadA remain to be determined. Engagement of CEACAMs, integrins and HSPGs can result in meningococcal internalization by epithelial cells (1) by triggering a variety of host cell signalling mechanisms. Meningococci can be found in subepithelial tissue (2) in healthy individuals thus cellular entry or otherwise traversal across the epithelium may not be an unusual event. In addition, the fact that meningococci can interact with subcellular proteins such as α-actinin may also lend some support to this notion, although the role of this interaction in vivo remains unclear. On crossing the epithelial barrier, meningococci are able to interact further with proteins of the extracellular matrix including fibronectin (Fn) and vitronectin (Vn). Internalized bacteria may also migrate back to the apical surface for transmission to a new host (3). An animated version of the Figure is available at http://www.ClinSci.org/cs/118/0547/cs1180547add.htm.

Figure 5. Meningococcal entry into and survival within the vasculature

Capillaries in close proximity to mucosal epithelial tissues are a possible point of entry into the blood for N. meningitidis. In vivo, meningococci initially encounter the basolateral surface of endothelial cells and need to traverse in a basal to apical direction to enter the vasculature. The in vitro studies, however, do not allow easy examination of basal interactions as cultured cells present their apical surfaces to the media. Both integrins and HSPGs are known to be expressed on the basolateral surface of endothelial cells and, hence, are likely targets for vascular penetration. However, it should be noted that these receptors are also expressed apically and are also probably involved in the exit from the bloodstream. Once in the blood, only capsulate meningococci appear to survive; whether acapsulate bacteria arising naturally can survive in any microenvironment is not known. In addition, meningococci are able to bind to a number of negative regulators of complement such as C4bp, factor H and vitronectin. Acquisition of such factors could lead to decreased complement-mediated killing in vivo. Interactions with vascular cells via protein adhesins and their cognate receptors and via LPS–TLR4 provoke an inflammatory response leading to cytokine release and cellular damage. This could increase further cell barrier penetration and leakage, which accounts for the damage and clinical symptoms observed during meningococcal sepsis, typified in latter stages by a petechial rash (see inset; from meningitis.org). LPS has also been shown to be toxic for human endothelial cells in vitro [40].

Figure 6. Meningococcal penetration of the BBB and interaction with the meninges leading to meningitis

In areas of low blood flow and, hence, low shear stress, meningococci have been shown to adhere to the vasculature within the brain [139]. In addition, cytoskeletal re-arrangements leading to lipid microdomain formation could facilitate resistance to shearing forces once bacteria are bound [145]. Whether bacterial transmigration in this niche is predominantly by transcytosis across intact cell barriers or requires damage to the barrier (for example, due to the action of pro-inflammatory cytokines) for ease of passage is unknown, although more than one mechanism may operate. Once the BBB is breached, meningococcal interaction with cells lining the leptomeninges leads to the release of pro-inflammatory cytokines [9,108,142], provoking an inflammatory reaction resulting in meningitis.