Molecular Microbiology (1999) 32(6), 1124–1132

MicroReview Interactions of pathogenic Neisseria with host cells. Is it possible to assemble the puzzle? Xavier Nassif,* Ce´line Pujol, Philippe Morand and Emmanuel Euge`ne INSERM U411, Faculte´ de Me´decine Necker-Enfants Malades, Universite´ Rene´ Descartes, Paris, France. Summary

Neisseria meningitidis and Neisseria gonorrhoeae are human pathogens that have to interact with mucosa and/or cellular barriers for their life cycles to progress. Even though they both give rise to dramatically different diseases, the use of in vitro models has shown that most of the mechanisms mediating cellular interactions are common to N. meningitidis and N. gonorrhoeae. This suggests that bacterial cell interactions may be essential not only for pathogenesis but also for other aspects of the bacterial life cycle that are common to both N. meningitidis and N. gonorrhoeae. This manuscript will review the most recent developments concerning the mechanisms mediating cellular interaction of pathogenic Neisseria and will then try to put them into the perspective of pathogenesis and bacterial life cycle. Introduction

Neisseria meningitidis (MC) and Neisseria gonorrhoeae (GC) are two human extracellular pathogens that have to interact with cellular barriers for their pathogenesis. Neisseria meningitidis colonizes the nasopharynx and eventually spreads into the bloodstream before crossing the blood–brain barrier to induce meningitis. Neisseria gonorrhoeae colonizes and invades the epithelium of the genitourinary tract leading to a localized inflammatory process that is rarely followed by bacteraemia. For these pathogens, which infect only humans and do not survive in the environment, the ability to be pathogenic is a minor aspect of their life cycle. It is likely that a large part of their genome is devoted to coding for bacterial attributes necessary for colonization and survival in their niche, Received 12 January, 1999; revised 2 March, 1999; accepted 5 March, 1999. *For correspondence. E-mail [email protected]; Tel. (þ33) 1 406 15678; Fax (þ33) 1 406 15592. Q 1999 Blackwell Science Ltd

which is the nasopharynx for N. meningitidis and the genitourinary tract of asymptomatic carriers for N. gonorrhoeae. The use of in vitro models has allowed much progress to be made for identifying bacterial components and cellular receptors involved in these interactions. Most of these are common to both pathogenic Neisseria. This is consistent with the fact that N. meningitidis and N. gonorrhoeae belong to the same genospecies (Guibourdenche et al., 1986) and share many common genetic, biochemical and antigenic features. However, the role of these factors in vivo is mainly unknown. The goal of this review is to emphasize recent findings concerning molecular aspects of the interactions between pathogenic Neisseria and host cells, and then to address the possible role these attributes could have in the life cycle of these pathogens.

Type IV pili Type IV pili are filamentous structures emanating from the bacterial surface composed of protein subunits. They are of paramount importance to the pathogenic process, as revealed by the fact that primary cultures of clinical isolates of pathogenic Neisseria are invariably piliated. In vitro, their role in promoting adhesion to epithelial and endothelial cells is essential and has been well established. In the case of encapsulated meningococci, pili seem to be the main attribute involved in the initiation of MC cell interaction, and the adhesion of non-piliated encapsulated bacteria with mammalian cells remain limited (Nassif et al., 1994). Furthermore, after pilusmediated interaction of encapsulated MC, few bacteria are internalized, and most of them remain as extracellular adherent pathogens. Production of pili is also associated with a number of other phenotypes, such as high level competence for transformation by exogenous DNA, bacterial autoagglutination and twitching motility. Pilus biogenesis as well as the description of the pilus-associated proteins responsible for pilus-mediated adhesion will be reviewed extensively in other articles in this series. The current model suggests that 110 kDa PilC molecules are located in pilus fibres and carry a cell-binding domain (Rudel et al., 1995). PilC molecules are therefore viewed as adhesins responsible for pilus-mediated adhesion. In

Pathogenic Neisseria interactions with host cells 1125 GC, two pilC alleles have been identified both of which are adhesive. In contrast, different strains of MC have one or two pilC loci. However, when two PilC proteins can be produced, only one, designated PilC1, is adhesive; the other (PilC2) is unable to promote pilus-mediated adhesion (Nassif et al., 1994). The reason for this discrepancy is unknown, and could be because of either the lack of a cell-binding domain or that these non-adhesive PilC2 proteins are unable to migrate in the pili. Extensive comparison of the primary sequences of various PilC proteins does not help to explain this phenomenon. Transcription of the pilC1 adhesive allele in MC is under tight regulation. Initial interactions of MC with host cells upregulate transcription of pilC1 but not pilC2. This upregulation is essential for obtaining a full adhesive phenotype (Taha et al., 1998), demonstrating that the interaction of MC with cells leads to a cross-talk between the bacteria and the cells. The pilin is the pilus subunit. It is believed to be incapable of interacting directly with eukaryotic cells but rather to play an essential role as the fibre scaffold. However, some pilin variants are more efficient than others in enhancing bacterial cell interactions, owing to the ability of these variants to favour agglutination of pili, which then form large bundles (Marceau et al., 1995; Marceau and Nassif, 1999). Two mechanisms may explain the enhancement of adhesiveness by bundled pili: (i) aggregative pili increase bacteria–bacteria interactions and (ii) bundled pili could reinforce the interactions of the adhesin with its eukaryotic receptor because several adhesin molecules are present at the extremity of a bundle. The mechanism by which some pilins induce pilus aggregation is not known, but comparison of the primary sequence of low and high adhesive derivatives suggests that localized charge modifications might play a role in the bundling process (Parge et al., 1995). A major feature of pathogenic Neisseria pilins is that they are glycosylated. An O-linked galactose, alpha-1,3 GlcNAc, was first identified in GC pilin (Parge et al., 1995) and then in the pilin of one MC strain (Marceau et al., 1998). In another MC strain, the sugar identified is a trisaccharide, a digalactosyl 2,4-diacetamido-2,4,6-trideoxyhexose (Stimson et al., 1995). Glycosylation is not required for pilus biogenesis, and in fact non-glycosylated pilins are responsible for the formation of pili that tend to aggregate more and to form larger bundles than those obtained with glycosylated pilins. Consistent with this is the larger proportion of soluble, truncated monomers of pilin produced by strains having a glycosylated pilin than that observed when pilin is non-glycosylated (Marceau et al., 1998). The fact that glycosylation increases the amount of soluble pilin monomers suggests that unassembled pilin molecules could have a function. A cell-binding domain is present in the constant region of the pilin monomer (Marceau et al., 1995). According to the pilin structure, this Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

site would not be accessible in the pilus fibre. One may speculate that soluble monomers of pilin could signal to cells via this cell-binding domain and have an effect independent of their role as the building block of pili. Other post-translational modifications have been reported for pilin: a glycerophosphate (Stimson et al., 1996) and a phosphorylcholine epitope are linked to pilin (Weiser et al., 1998). The role of these modifications has not yet been explored. Recently, a phosphate has been shown to be linked to Ser-68. This phosphate promotes straighter and/or less bundled pili by charge modification (Forest et al., 1999). The complement regulatory protein (CD46) has recently been recognized as a pilus receptor for pathogenic Neisseria (Ka¨llstro¨m and Jonsson, 1998). Piliated Neisseria are unable to bind cells of non-human origin, such as CHO cells; however, they are capable of binding to CHO cells expressing human CD46. Attachment of bacteria is blocked by monoclonal antibodies against CD46 and by recombinant CD46 protein produced in Escherichia coli. The pilus-associated molecule responsible for this attachment has not been identified, but the adhesive forms of PilC are the best candidates. The consequences of this initial attachment between pili and CD46 remain to be assessed, but this interaction is believed to send a signal to the host cells; however, it is not efficient at causing the internalization of the bacteria inside the cells. The discovery that CD46 is a pilus receptor will lead to a better understanding of the cellular events after pilus-mediated interactions and of the exact role of pilus-mediated adhesion in the pathogenesis of both MC and GC. Class 5 (Opa) and Opc proteins The Opa associated (class 5) proteins are basic outer membrane proteins with a molecular weight of < 28 kDa. They are expressed in most MC and all GC and are structurally similar in both species. They form a family of proteins, each being encoded by a distinct opa gene. Opa proteins are subject to phase variation because of the occurrence of reversible sequence variation in the 58 coding region of their structural genes. A single gonococcal strain may carry up to 12 opa genes, whereas MC strains have fewer loci (three or four). Opc proteins are outer membrane proteins that have similar size and physicochemical properties as the Opa proteins. However, they are structurally distinct and have only limited sequence similarity. Opc also undergoes phase variation; this process is due to transcriptional regulation (Sarkari et al., 1994). The transcription of opc may vary from zero to the intermediate or high level. Even though an opc -like gene has recently been described in GC, the expression of an Opc protein is restricted to a subset of MC strains.

1126

X. Nassif, C. Pujol, P. Morand and E. Euge`ne

Table 1. The MS11 Opa proteins repertoire. Opa proteinsa,b

rOpaa,c

Molecules recognizedd

Phenotype

OpaA

Opa50

Heparan sulphate proteoglycan Vitronectin Fibronectin

Invasion of epithelial cells

Opa57

CD66a, CD66c, CD66d, CD66e

OpaC

Opa52

CD66a, CD66c, CD66d, CD66e

OpaD

Opa56

CD66a, CD66e

Uptake by professional phagocytes in an opsonin-independent pathway

OpaE

Opa57

CD66a, CD66e

Transcytosis across a tight-junction-forming monolayer of epithelial cells

OpaG

Opa58

CD66a, CD66c, CD66d, CD66e

Increase adhesion to TNF-a-activated HUVECs

OpaH

Opa59

CD66a, CD66e

OpaI

Opa60

CD66a, CD66c, CD66d, CD66e

OpaJ

Opa51

CD66a, CD66e

OpaK

Opa53

CD66a, CD66e (þ /¹)

OpaB e

a. This comparison is similar to the one published by Bos et al. (1997). b. Designation of naturally occurring Opa variants by Swanson et al. (1988). c. Designation of recombinant Opa variants according to Kupsch et al. (1993). d. The Opa specification for CD66 receptor is according to Gray-Owen et al. (1997b) and Bos et al. (1997). e. OpaC is bifunctional, and interacts as well with heparan sulphate (Chen et al., 1997).

Opa proteins in Neisseria gonorrhoeae The role of Opa proteins as mediating cellular interactions has been extensively studied in GC. Much of the more recent studies were performed using non-piliated derivatives of strain MS11, producing different Opa proteins. The MS11 Opa repertoire and the corresponding nomenclature is shown in Table 1. In this strain, the Opa proteins can be divided in two groups according to the type of molecules they recognize on the surface of the mammalian cells. The first group is represented by OpaA (Opa50 ), which binds to a heparan sulphate proteoglycan of the syndecan family (van Putten and Paul, 1995), thus resulting in adhesion and internalization of gonococci by some epithelial cell lines such as Chang, Hec1B, HeLa and ME180. This entry process resembles phagocytosis rather than macropinocytosis and involves a signalling pathway requiring stimulation of the phosphatidylcholine specific phospholipase and acidic shingomyelinase (Grassme´ et al., 1997). In some cell lines such as CHO, Hep-2 and HeLa, this proteoglycan-mediated adherence is not sufficient to promote entry of OpaA-expressing gonococci. An additional interaction between OpaA and a serum protein is required for invasion (Duensing and van Putten, 1997; Gomez-Duarte et al., 1997; van Putten et al., 1998a). This interaction involves fibronectin for Hep2 and HeLa cells, and vitronectin for CHO cells. Fibronectin or vitronectin bind specifically to OpaA proteins and to the corresponding integrin via the RGD sequence, thus bridging the gonococci to the host cells. In this case, the entry process results from a concerted action of the

heparan sulphate proteoglycan receptors and the integrin receptors. Protein kinase C (PKC) inhibitors block vitronectin-mediated invasion, suggesting that this latter internalization pathway possibly involves PKC (Dehio et al., 1998). Like OpaA, OpaC also has heparan sulphate-binding properties (Chen et al., 1997). The second group of Opa proteins encompasses the other 10 alleles that mediate interaction with eukaryotic cells via CD66 (Chen et al., 1997; Gray-Owen et al., 1997a). The CD66 family includes four different forms: the CD66a (biliary glycoprotein; BGP), CD66b (CEA gene family member 6; CGM6), CD66c (non-specific cross-reactive antigen; NCA), and CD66d (CEA gene family member 1; CGM1). CD66b is not bound by any of the Opa proteins, whereas the other members of this family are differentially recognized by the Opa proteins (see Table 1) (Bos et al., 1997; Gray-Owen et al., 1997b). Recognition does not involve the carbohydrate moiety of CD66, and the CD66 receptor specificity exhibited by neisserial Opa variants is controlled by determinants in the CD66 N-domain (Bos et al., 1998). The CD66 molecules are produced differentially by polymorphonuclear, epithelial and endothelial cells. Initially, Opa proteins, which do not bind to surface proteoglycans, were identified by their ability to produce an interaction with polymorphonuclear leucocytes (PMN), thus causing opsonin-independent uptake by professional phagocytes (Fischer and Rest, 1988). This CD66-mediated phagocytosis requires a Src-like tyrosine kinase- and Rac1dependent signalling pathway (Hauck et al., 1998). The different specificities of Opa proteins for CD66 receptor Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

Pathogenic Neisseria interactions with host cells 1127 suggest that these proteins mediate different cellular responses depending on the cell type and the pattern of CD66 molecules that are present on the surface. This is consistent with the fact that an enhanced oxidative burst is triggered in granulocytic cells by Opa proteins that bind CD66a. In addition to promoting interactions with PMN, Opa-mediated binding to CD66a increases GC adherence to tumour necrosis factor (TNF)-a-activated human umbilical vein epithelial cells (HUVECs) in vitro (Gray-Owen et al., 1997b). Opa proteins binding to CD66 allow trancytosis of GC across a monolayer of polarized T84 epithelial cells (Wang et al., 1998). Recently, another possible ligand for Opa proteins was identified by using the yeast two-hybrid system. Williams et al. (1998) showed that one Opa protein binds to pyruvate kinase. Subsequently, these authors have demonstrated that intracellular gonococci bind pyruvate kinase and require host pyruvate for growth. A similar function for meningococcal Opa proteins has not yet been reported.

Table 2. Classification of neisserial porins.a

Neisseria meningitidis Neisseria gonorrhoeae

porAb

porB c

Present Pseudogene

porB1a or porB1b porB2 or porB3

a. This table is according to Feavers and Maiden (1998). b. A porA gene is present only in N. meningitidis. In this latter species two porins PorA and PorB may be concomittantly expressed. A porA pseudogene has recently been described in GC, but no PorA protein is produced. c. Four porB alleles have been described: porB1a, porB1b, porB2 and porB3. PorB1a and PorB1b are found only in GC and are mutually exclusive, therefore a strain produces either one of these porins. These proteins were formerly designated P.IA and P.IB. PorB2 and PorB3 are present in MC and are also mutually exclusive, the proteins encoded by these alleles were formerly designated class 2 and class 3.

above), especially when considering that a MC equivalent of GC OpaA has not yet been found. However, in strains lacking Opc, MC ability to bind to proteoglycan may be carried by some other components, including an Opa protein.

Opa and Opc proteins in Neisseria meningitidis

Porins

Opc and class 5 proteins mediate an interaction between meningococci and eukaryotic cells, only in a non-encapsulated background. In capsulated bacteria, Opa and Opc do not seem to affect bacterial interactions with host cells. In addition class 5- and Opc-mediated cellular interactions are also profoundly inhibited by lipooligosaccharide (LOS) sialylation (Virji et al., 1992,1993; de Vries et al., 1998). Opa proteins allow MC interaction with epithelial cells, PMN and to a lesser extent with endothelial cells. It has been established that Opa proteins bind members of the CD66 family, thus mediating adhesion to and invasion of epithelial cells (Virji et al., 1996a,b). In the case of interactions with PMN, binding to CD66 leads to phagocytic elimination of Opa-producing MC. Opc significantly increases adhesion and invasion of bacteria into both Chang cells and HUVECs (Virji et al., 1992), however, only when present at high levels (Sarkari et al., 1994). Opc causes adherence of non-piliated MC and promotes interaction of unencapsulated variants, not only with human cells but also, to a lesser degree, with cells from other mammals. Cell-surface proteoglycans of cultured epithelial cells have recently been shown to be the receptors for Opc (de Vries et al., 1998). However, the exact nature of the proteoglycan receptors recognized by these adhesins/invasins is unknown. Opc has also been reported to facilitate adhesion to and invasion of endothelial cells through the binding of vitronectin in a trimolecular complex. The Opc-producing meningococci interact with vitronectin and use it to attach to the integrin avb3 on the apical surface of the endothelial cells (Virji et al., 1994). This raises the question of functional similarities between meningococcal Opc and gonococcal OpaA (see

Pathogenic Neisseria produce two porins designated PorA and PorB (Table 2). Neisserial porins have been shown to translocate spontaneously as functional voltage-gated ion channels into plasma membranes of eukaryotic cells, causing a transient change in membrane potential and interference with cell signalling (Ulmer et al., 1992). Furthermore, the porB1a gene of GC but not porB1b was shown to promote bacterial invasiveness in an Opa¹ background (van Putten et al., 1998b). This pathway is clearly independent of uptake conferred by Opa proteins. This PorB1a-mediated cellular invasion may be related to the more invasive behaviour of strains expressing PorB1a in natural infections. In addition, data showing that in GC PorB is involved in OpaA-mediated invasion have recently been reported (Bauer et al., 1999). The role of porins as promoting bacterial entry has not been addressed in MC, but it has been shown that PorA and PorB proteins are capable of nucleating actin (Giardina et al., 1998), suggesting that the translocated Neisseria encoded porins are involved in host cell actin reorganization during infection.

Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

Other bacterial components Several other bacterial components modulate bacteria– cell interactions. Sialylation of LOS can modulate Opaand/or Opc-mediated interactions negatively (van Putten, 1993; de Vries et al., 1998). Similar observations were made with the MC capsule (Nassif et al., 1994). Other adhesins apart from those reported above have also been described. For example, GC produces a 36 kDa outer membrane protein with a binding specificity

1128

X. Nassif, C. Pujol, P. Morand and E. Euge`ne

for a gangliotetraosylceramide. Several loci of both pathogenic Neisseria recognizing glycolipids and allowing binding to human epithelial cells have been identified (Paruchuri et al., 1990; Schwan et al., 1998). In addition, an unidentified inducible gonoccocal adhesin that binds the luteotropin receptor has been described (Spence and Clark, 1997). Regarding intracellular survival, IgA1 protease, which is believed to be crucial for mucosal colonization, has recently been shown to be important for intracellular survival. Hence, it cleaves the LAMP1 protein and therefore prevents phagolysosomal fusion, thus playing a role in the ability of pathogenic Neisseria to survive inside epithelial cells (Lin et al., 1997; Ayala et al., 1998). Is it possible to assemble the puzzle?

Gonorrhoea The events that led to the development of gonorrhoea were initially worked out using ex vivo organ culture models (McGee et al., 1981; McGee et al., 1983; Harvey et al., 1997). Bacteria were shown to adhere to the microvilli of non-ciliated cells of the epithelium. They induce focal polymerization of actin, often accompanied by microvilli extensions, enter the apical pole of epithelial cells and transcytose to the basolateral side of the cells. Once in the subepithelial space, they induce the inflammatory process responsible for the disease symptoms. It is tempting to speculate that the initial contact is established by bacterial pili and is then followed by a tight adherence via Opa proteins. This Opa-mediated interaction could promote actin polymerization followed by transcytosis through epithelial cells. Considering the data obtained using a monolayer of epithelial cells (Wang et al., 1998), binding of Opa protein to CD66 is the most likely pathway responsible for this transcytosis. In this model, pili are colonization factors. The role of OpaA remains unclear, especially considering that heparan sulphate-containing receptors are located at the basolateral surface of polarized cells. Consequently, OpaA-mediated entry cannot be the driving force for transcytosis. On the other hand, it could be a way of amplifying the invasion of cells, once the inflammatory process has started and has led to the breakdown of the mucosal barrier. The above model does not take into account several experimental observations. Piliated Opa¹ gonococci induce cytoskeletal rearrangements (Merz and So, 1997), cross a monolayer of epithelial cells within 48 h (Merz et al., 1996) and upregulate the induction of inflammatory cytokines by epithelial cells (Naumann et al., 1997). Taken together, these observations strongly suggest that the cellular events induced by pilus-mediated adhesion are sufficient to induce the inflammatory process. This is consistent with the data obtained using a human challenge model, which show that the occurrence of the inflammatory

process occurs with piliated strains (Cannon et al., 1998). These observations argue in favour of a more sophisticated role for pili than just being colonization factors. Thus, after pilus-mediated adhesion, a signal might be transduced to epithelial cells that leads to the induction of cytokine production and subsequent development of the inflammatory process. The recruitment of inflammatory cells to the subepithelial space may be responsible for the breakdown of the mucosal barrier and the passage of GC into this space. In this model, the role of Opamediated transcytosis and Opa-mediated cellular invasion in the events that lead to inflammation is unclear. However, in vivo, N. gonorrhoeae invades urethral epithelial cells (Apicella et al., 1996). These intracellular bacteria could correspond to the resistant forms that are protected from the bactericidal activity of the extracellular fluids. Opa-mediated invasion may therefore be a way to increase duration of the carrier state in asymptomatic patients, and subsequently to favour the transmission of the bacteria. As far as porins are concerned, their role in nucleating actin could play a major role in the cytoskeletal modifications responsible for bacterial entry. As pointed out above, the IgA protease might be essential in intracellular survival and therefore play a major role in persistence and dissemination. This is consistent with the data obtained using the human challenge model, which show that GC mutants unable to produce IgA protease were unaffected in their ability to cause infection (Cannon et al., 1998). In the near future, the use of the human challenge model should allow the identification of the role of other bacterial factors mediating cellular interactions in gonorrhoea.

Meningococcal diseases A major point in the life cycle of MC is its ability to colonize the nasopharynx asymptomatically. The mechanism of this colonization is unknown. MC strains colonizing the nasopharynx are mostly non- or poorly capsulated. Therefore, the factors available for cellular interactions are similar to those found in GC, and thus the mechanisms of colonization could be very similar. The question why these factors that are responsible for an inflammatory process in the genitourinary tract do not induce a similar process in the throat remain unanswered. One major difference is that the urethral epithelium is usually sterile, unlike the nasopharynx, and the signalling that occurs after bacterial colonization and that is responsible for the occurrence of the inflammatory process in the urethra may not happen in the nasopharynx. In the case of heavily capsulated pathogenic strains, the bacteria may be able to colonize the nasopharynx efficiently only with the help of bundle-forming pili. The mechanism by which these bacteria cross the mucosal barrier is unknown; a possible Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

Pathogenic Neisseria interactions with host cells 1129 Fig. 1. The interaction of N. meningitidis with an epithelial monolayer is a two-step process. During the first step of localized adherence, pilus-mediated adhesion is probably critical and thus requires the presence of one PilC adhesive allele. In addition, to get colonies on the surface of the cells, a pilin variant responsible for the formation of bundled pili needs to be produced. The second step corresponds to diffuse adherence. At this stage N. meningitidis has lost its pili. PilT, a cytoplasmic nucleotide-binding protein, is required to disperse bacteria from the localized to the diffuse adherence pattern, and to induce the loss of piliation and intimate attachment (Pujol et al., 1999). The component(s) necessary for intimate attachment are unknown.

way is via the M cells present at the tonsillar sites. It is conceivable that transcytosis through these cells may be an efficient route of invasion for MC, as described for enteric pathogens (Jones et al., 1994). Another situation in which MC interacts with a cellular barrier occurs when it crosses the blood–brain barrier. At this stage MC is capsulated and pili seem to be the only bacterial factor that can mediate any cellular interaction. The blood–brain barrier is composed of epithelial and/or endothelial cells that have tight junctions, which limits paracellular flux. Recent in vivo data have demonstrated that MC is capable of interacting with these structures and that the movement across these monolayers requires pilus-mediated adhesion and, more specifically, correlates with the presence of an adhesive form of PilC (Pron et al., 1997). However, a better understanding of all of the steps involved in the movement of MC across the blood–brain barrier will require the development of in vitro models. Because the main cellular characteristic of the blood–brain barrier is the existence of tight junctions that limit paracellular flux, the in vitro models used to study this step should be either a monolayer of tight junction-forming endothelial cells and/or epithelial cells. Even though considerable progress has been made in our understanding of the blood–brain barrier, models using human brain endothelial cells are not yet available. This has led to the use of epithelial monolayers of human polarized cells with organized tight junctions similar to those responsible for the blood–brain barrier. Using such a model, which most likely reflects only some aspects of the movement of MC across the blood–brain barrier, MC has been shown to cross a monolayer via the transcellular route without destroying the intercellular junctions (Pujol et al., 1997). It should be pointed out that even though bacteria are not been seen between cells, this does not exclude the possibility that this route is used. Initially, the adhesion of MC is localized, resulting in the formation of clumps of bacteria on the apical surface of the monolayer. Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

The bacteria then spread onto the surface of the cells as the clumps disappear and are replaced by a monolayer of MC covering the cells, thus realizing a diffuse adherence phenotype. At this stage, bacteria adhere intimately and firmly to the apical membrane and in some places are found intracellularly. Knowledge of the bacterial factors involved in all these steps is very incomplete (Fig. 1). As mentioned, pilusmediated adhesion is probably most critical during the first step of localized adherence. At the diffuse adherence step, MC has lost its pili. There are several possible explanations for this process. Firstly, cross-talk between bacteria and cells could lead to the downregulation of genes involved in pilus biogenesis. Secondly, pili could be retracted, thus bringing into contact the outer membrane of the bacteria and the cellular plasmic membrane. The loss of piliation after the localized adherence step suggests that pili per se are not involved in the intimate attachment observed during the diffuse adherence. The best candidates for this function would be the Opa/Opc proteins. However, several observations are not consistent with a role for these molecules in this intimate attachment: (i) opc isolates are capable of intimate attachment (Pujol et al., 1997); (ii) a naturally occurring Opa¹ variant is also able to undergo intimate attachment (Pujol et al., 1997). The role of porins in this process remains to be assessed. It should be pointed out that an inoculum of non-piliated bacteria is incapable of diffuse adherence, suggesting that the initial phase of localized adherence is necessary for the production of some as yet unidentified factors that lead to a diffuse adherence phenotype. Furthermore, there is evidence that PilT, a cytoplasmic nucleotidebinding protein, involved in pilus retraction, is required to disperse bacteria from the localized to the diffuse adherence pattern, and to induce the loss of piliation and intimate attachment (Pujol et al., 1999). Pili could be seen as sensory organs that, once they recognize their receptor on the surface of the cells, transduce a signal to the bacteria,

1130

X. Nassif, C. Pujol, P. Morand and E. Euge`ne

which in turn upregulate the expression of some unidentified component(s) necessary for intimate attachment and expression for the signalling events linked to this step. The loss of piliation at the diffuse adherence step could result from the same signalling event. The cellular mechanisms driving the transcytosis of bacteria remain unexplored. One of the most fascinating aspects of the study of the pathogenesis of N. meningitidis and N. gonorrhoeae is that they have evolved common mechanisms of interaction with cells, even though they clearly have a different pathogenesis. The question of whether additional factors are responsible for the specificity of the pathogenesis of each bacteria is unanswered and will benefit from the availability of the sequence of the meningococcal genome and its comparison with that of N. gonorrhoeae. On the other hand, the similarity in the bacterial attributes evolved by these pathogens suggests that bacterial cell interactions may be essential not only for pathogenesis but also for some other aspects of the life cycle of pathogenic Neisseria -like survival and dissemination. Acknowledgements We are grateful to C. Tinsley for careful reading of this manuscript. We wish to thank the anonymous reviewers for their helpful suggestions in improving this manuscript. The work in the laboratory of XN is supported by INSERM, Universite´ Rene´ Descartes Paris 5, and the Fondation pour la Recherche Me´dicale.

References Apicella, M.A., Ketterer, M., Lee, F.K., Zhou, D., Rice, P.A., and Blake, M.S. (1996) The pathogenesis of gonococcal urethritis in men: confocal and immunoelectron microscopic analysis of urethral exudates from men infected with Neisseria gonorrhoeae. J Infect Dis 173: 636–646. Ayala, P., Lin, L., Hopper, S., Fukuda, M., and So, M. (1998) Infection of epithelial cells by pathogenic neisseriae reduces the levels of multiple lysosomal constituents. Infect Immun 66: 5001–5007. Bauer, F.J., Rudel, T., Stein, M., and Meyer, T.F. (1999) Mutagenesis of the Neisseria gonorrhoeae porin reduces invasion in epithelial cells and enhances phagocyte responsiveness. Mol Microbiol 31: 903–913. Bos, M.P., Hogan, D., and Belland, R.J. (1997b) Selection of Opaþ Neisseria gonorrhoeae by limited availability of normal human serum. Infect Immun 65: 645–650. Bos, M.P., Kuroki, M., Krop-Watorek, A., Hogan, D., and Belland, R.J. (1998) CD66 receptor specificity exhibited by neisserial Opa variants is controlled by protein determinants in CD66 N-domains. Proc Natl Acad Sci USA 95: 9584–9589. Cannon, J.G., Johannsen, D., Hobbs, J.D., Hoffman, N., Dempsey, J.A.F., Johnston, D., et al. (1998) Studies of the human challenge model of gonococcal infection: population dynamics of gonococci during experimental infection,

and infectivity of isogenic mutants. In Abstracts of the Eleventh International Pathogenic Neisseria Conference. Nassif, X., Quentin-Millet, M.J., and Taha, M.K. (eds). Paris: EDIMARK, pp. 48. Chen, T., Grunert, F., Medina-Marino, A., and Gotschlich, E.C. (1997) Several carcinoembryonic antigens (CD66) serve as receptors for goncoccal opacity proteins. J Exp Med 185: 1557–1564. Dehio, M., Gomez-Duarte, O.G., Dehio, C., and Meyer, T.F. (1998) Vitronectin-dependent invasion of epithelial cells by Neisseria gonorrhoeae involves mv integrin receptors. FEBS Lett 424: 84–88. Duensing, T.D., and van Putten, J.P.M. (1997) Vitronectin mediates internalization of Neisseria gonorrhoeae by Chinese hamster ovary cells. Infect Immun 65: 964–970. Feavers, I.M., and Maiden, M.C.J. (1998) A gonococcal porA pseudogene: implications for understanding the evolution and pathogenicity of Neisseria gonorrhoeae. Mol Microbiol 30: 647–656. Fischer, S.H., and Rest, R.F. (1988) Gonococci possessing only certain P.II outer membrane proteins interact with human neutrophils. Infect Immun 56: 1574–1579. Forest, K., Dunham, S.T., Koomey, M., and Tainer, J. (1999). Crystallographic structure reveals phosphorylated pilin from Neisseria : phosphoserine sites modify type IV pilus surface chemistry and fibre morphology. Mol Microbiol 31: 743–752. Giardina, P.C., Wen, K.K., Williams, R., Lubaroff, D., Blake, M.S., Rubenstein, P.A., and Apicella, M.A. (1998) Neisseria -encoded porins influence actin dynamics, in vitro. In Abstracts of the Eleventh International Pathogenic Neisseria Conference Nassif, X., Quentin-Millet, M.J., and Taha, M.K. (eds). Paris: EDK, p. 36. Gomez-Duarte, O.G., Dehio, M., Guzman, C.A., Chhatwal, G.S., Dehio, C., and Meyer, T.F. (1997) Binding of vitronectin to Opa-expressing Neisseria gonorrhoeae mediates invasion of HeLa cells. Infect Immun 65: 3857–3866. Grassme´, H., Gulbins, E., Brenner, B., Ferlinz, K., Sandhoff, K., Harzer, K., et al. (1997) Acidic sphingomyelinase mediates entry of N. gonorrhoeae into nonphagocytic cells. Cell 91: 605–615. Gray-Owen, S.D., Dehio, C., Haude, A., Grunert, F., and Meyer, T.F. (1997a) CD66 carcinoembryonic antigens mediate interactions between Opa-expressing Neisseria gonorrhoeae and human polymorphonuclear phagocytes. EMBO J 16: 3435–3445. Gray-Owen, S.D., Lorenzen, D.R., Haude, A., Meyer, T.F., and Dehio, C. (1997b) Differential Opa specificities for CD66 receptors influence tissue interactions and cellular response to Neisseria gonorrhoeae. Mol Microbiol 26: 971–980. Guibourdenche, M., Popoff, M.Y., and Riou, J.Y. (1986) Deoxyribonucleic acid relatedness among Neisseria gonorrhoeae. N. meningitidis, N. lactamica, N. cinerea and Neisseria polysaccharea. Ann Inst Pasteur/Microbiol 137B: 177–185. Harvey, H.A., Ketterer, M.R., Preston, A., Lubaroff, D., Williams, R., and Apicella, M.A. (1997) Ultrastructural analysis of primary human urethral epithelial cell cultures infected with Neisseria gonorrhoeae. Infect Immun 65: 2420–2427. Hauck, C.R., Meyer, T.F., Lang, F., and Gulbins, E. (1998) Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

Pathogenic Neisseria interactions with host cells 1131 CD66-mediated phagocytosis of Opa52 Neisseria gonorrhoeae requires a Src-like tyrosine kinase- and Rac1dependent signalling pathway. EMBO J 17: 443–454. Jones, B.D., Ghori, N., and Falkow, S. (1994) Salmonella typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer’s patches. J Exp Med 180: 15–23. Ka¨llstro¨m, H., and Jonsson, A.-B. (1998) Characterization of the region downstream of the pilus biogenesis gene pilC1 in Neisseria gonorrhoeae. Biochim Biophys Acta 1397: 137–140. Kupsch, E.-M., Knepper, B., Kuroki, T., Heuer, I., and Meyer, T.F. (1993) Variable opacity (Opa) outer membrane proteins account for the cell tropisms displayed by Neisseria gonorrhoeae for human leukocytes and epithelial cells. EMBO J 12: 641–650. Lin, L., Ayala, P., Larson, J., Mulks, M., Fukuda, M., Carisson, S.R., et al. (1997) The Neisseria type 2 IgA1 protease cleaves LAMP1 and promotes survival of bacteria within epithelial cells. Mol Microbiol 24: 1083–1094. McGee, Z., Johnson, A., and Taylor-Robinson, D. (1981) Pathogenic Mechanisms of Neisseria gonorrhoeae: observations on damage to human fallopian tubes in organ culture by Bonococci of colony type 1 or type 4. J Infect Dis 143: 413–422. McGee, Z., Stephens, D., Hoffman, L., Schlech, W. III, and Horn, R. (1983) Mechanisms of mucosal invasion by pathogenic Neisseria. Rev Infect Dis 5: S708–S714. Marceau, M., and Nassif, X. (1999) S-pilin production by Neisseria gonorrhoeae does not require post-traductional modification of Ser63, as is the case for Neisseria meningitidis. J Bacteriol 181: 656–661. Marceau, M., Beretti, J.-L., and Nassif, X. (1995) High adhesiveness of encapsulated Neisseria meningitidis to epithelial cells is associated with the formation of bundles of pili. Mol Microbiol 17: 855–863. Marceau, M., Forest, K., Be´retti, J.-L., Tainer, J., and Nassif, X. (1998) Consequences of the loss of O-linked glycosylation of meningococcal type IV pilin on piliation and pilusmediated adhesion. Mol Microbiol 27: 705–715. Merz, A.J., and So, M. (1997) Attachment of piliated, Opa (-) and Opc (-) gonococci and meningococci to epithelial cells elicits cortical actin rearrangements and clustering of tyrosine-phosphorylated proteins. Infect Immun 65: 4341– 4349. Merz, A.J., Rifenbery, D.B., Arvidson, C.G., and So, M. (1996) Traversal of a polarized epithelium by pathogenic Neisseriae: facilitation by type IV pili and maintenance of epithelial barrier function. Mol Med 2: 745–754. Nassif, X., Beretti, J.-L., Lowy, J., Stenberg, P., O’Gaora, P., Pfeifer, J., et al. (1994) Roles of pilin and PilC in adhesion of Neisseria meningitidis to human epithelial and endothelial cells. Proc Natl Acad Sci, USA 91: 3769–3773. Naumann, M., Wessler, S., Bartsch, C., Wieland, B., and Meyer, T.F. (1997) Neisseria gonorrhoeae epithelial cell interaction leads to the activation of the transcription factors nuclear factor kappaB and activator protein 1 and the induction of inflammatory cytokines. J Exp Med 186: 247–258. Parge, H.E., Forest, K.T., Hickey, M.J., Christensen, D.A., Getzoff, E.D., and Tainer, J.A. (1995) Structure of the Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132

fibre-forming protein pilin at 2.6 A resolution. Nature 378: 32–38. Paruchuri, D.K., Seifert, H.S., Ajioka, R.S., Karlson, K.-A., and So, M. (1990) Identification and characterization of a Neisseria gonorrhoeae gene encoding a glycolipid-binding adhesin. Proc Natl Acad Sci USA 87: 333–337. Pron, B., Taha, M.-K., Rambaud, C., Fournet, J.-C., Pattey, N., Monnet, J.-P., et al. (1997) Interaction of Neisseria meningitidis with the components of the blood–brain barrier correlates with an increased expression of PilC. J Infect Dis 176: 1285–1292. Pujol, C., Euge`ne, E., de Saint Martin, L., and Nassif, X. (1997) Interaction of Neisseria meningitidis with a polarized monolayer of epithelial cells. Infect Immun 65: 4836–4842. Pujol, C., Euge`ne, E., Marceau, M., and Nassif, X. (1999) The meningococcal PilT protein is required for induction of intimate attachment to epithelial cells following pilusmediated adhesion. Proc Natl Acad Sci USA 96: 4017– 4022. van Putten, J.P.M. (1993) Phase variation of lipopolysaccharide directs interconversion of invasive and immuno-resistant phenotypes of Neisseria gonorrhoeae. EMBO J 12: 4043–4051. van Putten, J.P., and Paul, S.M. (1995) Binding of syndecan like cell surface proteoglycan receptors is required for Neisseria gonorrhoeae entry into human mucosal cells. EMBO J 14: 2144–2154. van Putten, J.P.M., Duensing, T.D., and Cole, R.L. (1998a) Entry of OpaAþ gonococci into HEp-2 cells requires concerted action of glycosaminoglycans, fibronectin and integrin receptors. Mol Microbiol 29: 369–379. van Putten, J.P.M., Duensing, T.D., and Carlson, J. (1998b) Gonococcal invasion of epithelial cells driven by PIA., a bacterial ion channel with GTP binding properties. J Exp Med 188: 941–952. Rudel, T., Scheuerpflug, I., and Meyer, T.F. (1995) Neisseria PilC protein identified as type-4 pilus tip-located adhesin. Nature 373: 357–362. Sarkari, J., Pandt, N., Moxon, E.R., and Achtman, M. (1994) Variable expression of the Opc outer membrane protein in Neisseria meningitidis is caused by size variation of a promoter containing poly-cytidine. Mol Microbiol 13: 207–217. Schwan, E.T., Eickerja¨ger, S., Fischer, E., Maier, J., Manning, P.A., and Meyer, T.F. (1998) Identification and characterization of genomic loci of both pathogenic Neisseria mediating independent recognition of glycolipids and binding to human epithelial cells. In Abstracts of the Eleventh International Pathogenic Neisseria Conference Nassif, X., Quentin-Millet, M.J., and Taha, M.K. (eds). Paris: EDK, p. 387. Spence, J.M.J.C.R., C., and Clark, V.L. (1997) A proposed role for the lutropin receptor in contact-inducible gonococcal invasion of Hec1B cells. Infect Immun 65: 3736– 3742. Stimson, E., Virji, M., Barker, S., Panico, M., Blench, I., Saunder, J., et al. (1996) Discovery of a novel protein modification: alpha-glycerophosphate is a substituent of meningococcal pilin. Biochem J 316: 29–33. Stimson, E., Virji, M., Makepeace, K., Dell, A., Morris, H.R.,

1132

X. Nassif, C. Pujol, P. Morand and E. Euge`ne

Payne, G., et al. (1995) Meningococcal pilin: a glycoprotein substituted with digalactosyl 2,4-diacetamido-2,4,6-trideoxyhexose. Mol Microbiol 17: 1201–1214. Swanson, J., Barrera, O., Sola, J., and Boslego, J. (1988) Expression of outer membrane protein ii by gonococci in experimental Gonorrhea. J Exp Med 168: 2121–2129. Taha, M.-K., Morand, P., Pereira, Y., Euge`ne, E., Giorgini, D., Larribe, M., and Nassif, X. (1998) Pilus-mediated adhesion of Neisseria meningitidis – the essential role of cell contact-dependent transcriptional upregulation of the PilC1 protein. Mol Microbiol 28: 1153–1163. Ulmer, J.B., Burke, C.J., Shi, C., Friedman, A., Donnelly, J.J., and Liu, M.A. (1992) Pore formation and mitogenicity in blood cells by the class 2 protein of Neisseria meningitidis. J Biol Chem 267: 19266–19271. Virji, M., Makepeace, K., Ferguson, D.J.P., Achtman, M., Sarkari, J., and Moxon, E.R. (1992) Expression of the Opc protein correlates with invasion of epithelial and endothelial cells by Neisseria meningitidis. Mol Microbiol 6: 2785–2795. Virji, M., Makepeace, K., Ferguson, D.J.P., Achtman, M., and Moxon, E.R. (1993) Meningococcal Opa and Opc proteins: their role in colonization and invasion of human epithelial and endothelial cells. Mol Microbiol 10: 499–510. Virji, M., Makepeace, K., and Moxon, E.R. (1994) Distinct mechanisms of interactions of Opc-expressing meningococci at apical and basolateral surfaces of human endothelial cells; the role of integrins in apical interactions. Mol Microbiol 14: 173–184.

Virji, M., Makepeace, K., Ferguson, D.J.P., and Watt, S.M. (1996a) Carcinoembryonic antigens (CD66) on epithelial cells and neutrophils are receptors for Opa proteins of pathogenic neisseriae. Mol Microbiol 22: 941–950. Virji, M., Watt, S.M., Barker, S., Makepeace, K., and Doyonnas, R. (1996b) The N-domain of the humain CD66a adhesion molecule is a target for Opa proteins of Neisseria meningitidis and Neisseria gonorrhoeae. Mol Microbiol 22: 929–939. de Vries, F.P., Cole, R., Dankert, J., Frosch, M., and van Putten, J.P.M. (1998) Neisseria meningitidis producing the Opc adhesin binds epithelial cell proteoglycan receptors. Mol Microbiol 27: 1203–1212. Wang, J., Gray-Owen, S.D., Knorre, A., Meyer, T.F., and Dehio, C. (1998) Opa binding to cellular CD66 receptors mediates the transcellular traversal of Neisseria gonorrhoeae across polarized T84 epithelial cell monolayers. Mol Microbiol 30: 657–671. Weiser, J.N., Goldberg, J.B., Pan, N., Wilson, L., and Virji, M. (1998) The phosphorylcholine epitope undergoes phase variation on a 43-kilodalton protein in Pseudomonas aeruginosa and on Pili of Neisseria meningitidis and Neisseria gonorrhoeae. Infect Immun 66: 4263–4267. Williams, J.M., Chen, G.-C., Zhu, L., and Rest, R.F. (1998) Using the yeast two-hybrid system to identify human epithelial cell proteins that bind gonococcal Opa proteins: intracellular gonococci bind pyruvate kinase via their Opa proteins and require host pyruvate for growth. Mol Microbiol 27: 171–186.

Q 1999 Blackwell Science Ltd, Molecular Microbiology, 32, 1124–1132