The chronicles of Porphyromonas gingivalis : the microbium , the human oral epithelium and their interplay

The microbiota of the human oral mucosa consists of a myriad of bacterial species that normally exist in commensal harmony with the host. Porphyromonas gingivalis, an aetiological agent in severe forms of periodontitis (a chronic inflammatory disease), is a prominent component of the oral microbiome and a successful colonizer of the oral epithelium. This Gram-negative anaerobe can also exist within the host epithelium without the existence of overt disease. Gingival epithelial cells, the outer lining of the gingival mucosa, which function as an important part of the innate immune system, are among the first host cells colonized by P. gingivalis. This review describes recent studies implicating the co-existence and intracellular adaptation of the organism in these target host cells. Specifically, recent findings on the putative mechanisms of persistence, intercellular dissemination and opportunism are highlighted. These new findings may also represent an original and valuable model for mechanistic characterization of other successful host-adapted, self-limiting, persistent intracellular bacteria in human epithelial tissues.


Introduction
In the Divina Commedia, Dante describes his fantastic pilgrimage through Inferno, Purgatory and Paradise.The fascination with the characterization of Porphyromonas gingivalis infection in the last three decades appears to follow a similar path to the great master's journey in the brilliantly constructed and astonishingly complex gallery of characters.P. gingivalis, a Gram-negative anaerobe and successful colonizer of oral tissues, has notably been examined for its extremely pathogenic persona in a large number of studies involving many host cell types.In early studies, the organism was shown to possess a formidable array of virulence factors with the capacity to perturb host defence mechanisms and disintegrate structural components of the periodontal tissues.This has given a rather vitriolic perception of the microbe, placing it in the lower circles of Inferno.
Nonetheless, despite its pathogenic potential, it has been demonstrated recently that P. gingivalis can be subgingivally present and colonize the epithelium in the absence of overt disease (Colombo et al., 2006(Colombo et al., , 2007;;Rudney & Chen, 2006).As knowledge on the enormous diversity of the oral microbiome and the host cell biology of the epimucosal tissues has advanced, the understanding of the impact of P. gingivalis on the oral environment has caused a change in the status of P. gingivalis as well, moving it closer to Purgatory.In particular, our deepening understanding of the conditions of biological equilibrium between this microbe and the epithelium has encouraged some to reevaluate the pathogenic role of P. gingivalis in disease formation and perhaps redeem this organism's reputation.
Central to the latter perspective, this review is intended to enrich our current view of the adaptation and existence of this controversial oral bacterium within the host epithelium.Furthermore and specifically, recent findings on the putative mechanisms of persistence, opportunism and dissemination of P. gingivalis within the gingival epithelium will be highlighted.

Cause for concern
P. gingivalis, a prominent constituent of mature subgingival biofilms and a successful colonizer of oral mucosa, is recognized as one of the periodontal bacteria that assembles a prototypical polybacterial pathogenic consortium in severe and chronic forms of periodontitis, called the 'red complex' (Holt & Ebersole, 2005;Socransky & Haffajee, 1992).The organism has also been identified as a risk factor for coronary heart disease, pulmonary infections and pre-term, low birth weight deliveries (Herzberg & Weyer, 1998;Offenbacher et al., 2006;Scannapieco, 2006).Thus, until quite recently, most studies have focused on the pathogenic traits of the organism.A large body of evidence suggests that P. gingivalis produces an array of potential virulence factors.These include extracellular proteases (e.g.cysteine proteinases) that can cause modulation of the host immune response, attachment, and degradation or cleavage of host cell proteins and surface receptors (Curtis et al., 2005;Giacaman et al., 2007;Potempa et al., 2000;Sheets et al., 2008); a unique lipopolysaccharide (LPS) that interferes with host inflammatory response systems via innate and adaptive immunity (Darveau et al., 1998;Hajishengallis et al., 2006); adhesins such as fimbriae and haemagglutinins; and a putative invasin (haloacid dehalogenase family phosphoserine phosphatase).All these factors appear to play a role in successful invasion of host tissues (Lamont & Jenkinson, 2000;Tribble et al., 2006).Accordingly, the organism's multifaceted structural properties with their potentially damaging actions against various host cell types, including macrophages, neutrophils, fibroblasts, dendritic cells, endothelial cells, and oral epithelial and non-oral epithelial cells (HeLa derivatives KB and HEp-2 cell lines), have been significantly discussed in previous reviews (Cutler & Jotwani, 2006;Hajishengallis & Harokopakis, 2007;Kantarci & Van Dyke, 2002;Kinane et al., 2008;Rodrigues et al., 2008;Sheets et al., 2008).

Reason for hope
Even though a vast body of critical scholarship on P. gingivalis exists, the pathogenesis of infection still remains to be fully understood.Indeed, the microbial association with the oral epithelium, which is the main host tissue initially encountered by P. gingivalis, appears to be far more complex than previously believed.Recent studies have examined the composition of subgingival species in or on the sulcular gingival epithelial cells (GECs) obtained from periodontally healthy and diseased individuals.Using 16S rRNA probes in conjunction with confocal microscopy, these studies demonstrated substantial intracellular invasion of human gingival epithelium by oral bacteria, including P. gingivalis.Intriguingly, the results showed no significant differences in the percentage and levels of any particular species such as P. gingivalis between the epithelial samples from healthy and diseased subjects (Colombo et al., 2006(Colombo et al., , 2007)).
Similarly, separate studies analysing healthy human buccal epithelial cells for the characterization of intracellular bacteria found that P. gingivalis was extensively present in each sample (Rudney et al., 2005).It is also compelling to note that those epithelial cells harboured large masses of intracellular bacterial consortia resembling the polymicrobial nature of tooth-surface biofilm, and P. gingivalis formed only a small proportion of the bacterial consortia.In a further study (Rudney & Chen, 2006), epithelial samples were analysed for viability using markers of both cell membrane integrity and metabolic activity, which indicated no significant level of apoptosis or necrosis in epithelial cells heavily invaded by bacteria.The same study suggested that oral epithelium collected from healthy subjects, likely to have a natural tolerance for the polymicrobial intracellular flora, also contained P. gingivalis.These results show interesting parallels with in vitro research on P. gingivalis interactions with primary GECs, discussed below.Nevertheless, the contribution made by intracellular bacteria other than P. gingivalis in shaping the overall status of the oral epithelium needs to be considered.
Expanding the insight: the dynamics between the host-adapted microbe, P. gingivalis, and gingival epithelial cells Epithelial cells form an initial site of interaction with both resident microbiota and invading microbial pathogens.These cells function as an important arm of the immune response, as they have the ability to recognize, respond to and eliminate infection without detriment to the host.The microbiota of the oral mucosa consists of a myriad of bacterial species that normally exist in commensal harmony with the host (Jenkinson & Lamont, 2005;Marsh, 2006).In the gingival compartment, epithelial cells are the first host cells encountered by colonizing bacteria such as P. gingivalis.Studies on the interaction of the oral epithelium with the components of the inhabitant flora have revealed GECs that constitute a cellular interface to the colonizing microbiota in the gingival crevice and can sense and respond to the presence of bacteria, resulting in the secretion of innate immune effectors such as inflammatory cytokines and antimicrobial peptides (Hasegawa et al., 2008;Marshall, 2004;Nisapakultorn et al., 2001;Weinberg et al., 1998).
Likewise, mechanistic understanding of P. gingivalis colonization within oral epithelium has come largely from studies on human gingival epithelial cell infection performed in vitro using primary epithelial cultures of gingiva.Nearly a decade ago, these studies demonstrated that the organism is highly invasive and can rapidly enter primary cultures of human GECs.Furthermore, it was shown to be capable of intracellular replication (Belton et al., 1999;Lamont et al., 1995).Soon afterward, the mechanisms that orchestrate the successful internalization of the organism into GECs were delineated.It was shown that the adherence of P. gingivalis to GECs and the trigger event for subsequent invasion is primarily mediated by the binding of the major fimbriae to the b1 integrin receptor and by subsequent phosphorylation/activation of the putative integrin signalling proteins FAK (focal adhesion kinase) and paxillin, and the concurrent remodelling of the actin cytoskeleton (Yilmaz et al., 2002(Yilmaz et al., , 2003)).Later studies with transformed non-oral-epithelial cell lines such as HEp-2 confirmed these findings and suggested that the bacterial fimbriae-b 1 integrin receptor's association with cell lipid membrane rafts could mediate activation of the actin cytoskeleton reassembly, thereby allowing P. gingivalis to enter host cells (Nakagawa et al., 2002;Tsuda et al., 2008).
Subsequent studies attempting to understand the bacterium's intracellular life in human primary GECs illustrated that P. gingivalis can replicate to high levels intracellularly and maintain viability for extended periods.Interestingly, despite the burden of large numbers of intracellular bacteria, infected GECs do not undergo apoptotic or necrotic death and are resistant to apoptosis determined by annexin-V/propidium iodide and TUNEL analysis (Yilmaz et al., 2004).Furthermore, P. gingivalis infection induces an anti-apoptotic phenotype in primary GECs by rendering the host cells resistant to cell death caused by potent proapoptotic agents (Nakhjiri et al., 2001;Yilmaz et al., 2004Yilmaz et al., , 2008)).A preliminary investigation determined that the organism could balance the expression of pro-apoptotic Bax and anti-apoptotic Bcl-2 in order to prevent gingival cell death (Nakhjiri et al., 2001).Shortly afterwards, a comprehensive study shed light on the phenotypic outcomes and the mechanisms that P. gingivalis may utilize to promote host cell survival.This seemed to largely involve mitochondria-dependent signalling via activation of the PI3-kinase/AKT survival pathway, inhibition of cytochrome c release, and mitochondrial membrane depolarization (Yilmaz et al., 2004).A later study also confirmed that P. gingivalis infection can affect several mitochondrial anti-apoptotic pathways, including the inhibition of proapoptotic caspase-3 activation through dual JAK/Stat and AKT signalling in GECs (Mao et al., 2007).In general, the induction or inhibition of apoptosis by bacteria appears to be specific to the host cell type, bacterial species, strain and duration of infection.Nevertheless, the inhibition of apoptosis is essentially observed in infections with intracellular bacteria, which reside in or pass through epithelial tissues during the course of infection.These include several persistent and highly host-adapted bacteria such as Mycobacterium tuberculosis, Helicobacter pylori, and Chlamydia and Neisseria species (Amieva et al., 2002;Byrne & Ojcius, 2004;Velmurugan et al., 2007).Thus, P. gingivalis embodies the qualities of a successful self-limiting and balanced pathogen, for it has evolved sophisticated mechanisms that enable it to invade, replicate and colonize in host tissues successfully.
The association of P. gingivalis with apoptosis has been examined in a variety of cell types.P. gingivalis induces apoptosis in Jurkat T-cells, a non-oral epithelial cancer line (KB cells), B cells and human gingival fibroblasts, but inhibits apoptosis in human monocytes, macrophages, neutrophils and primary GECs (Chen et al., 2001;Geatch et al., 1999;Gemmell et al., 1999;Hiroi et al., 1998;Nakhjiri et al., 2001;Ozaki & Hanazawa, 2001;Preshaw et al., 1999;Yilmaz et al., 2004).On the other hand, a recent in vitro study designed to examine host cell death, utilizing primary GECs and heat-killed P. gingivalis, suggested that the dead organism in elevated numbers could induce apoptosis without any necrosis, by upregulation of the NFkB pathway (Brozovic et al., 2006).Thus, the inhibition or induction of cell death by P. gingivalis appears to be complex, and closely associated with the host cell type, bacterial strain, initial inoculation, and the use of live (metabolically active) versus dead bacteria, or the presence of specific bacterial components (e.g.cysteine proteinases, LPS).The following section addresses newly identified anti-apoptotic properties and the intercellular dissemination mechanism used by the live organism and probably vital for survival of P. gingivalis in its favourite host tissue, oral epithelium.
Fresh insights: the genius of P. gingivalis in gingival epithelial cells Advances in the understanding of host cell biology during bacterial infections in the last decade strongly support the notion that the characterization of bidirectional interactions between the bacterial invader and the host could assist in determining the balance between health and disease.Accordingly, recent proteomics and genomics studies have revealed that P. gingivalis infection induces regulation of a massive number of distinctive P. gingivalis proteins and genes that appear to be important for the adaptation and survival of the organism in epithelial cells (Hosogi & Duncan, 2005;Xia et al., 2007;Zhang et al., 2005).
Coming to grips with the bacterium's favouring intracellular life and prolonging host cell survival forces one to examine whether P. gingivalis resides and multiplies exclusively in the cytosol of the same cell throughout the infection or if it spreads to neighbouring cells to propagate.A recent study revealed that the suppression of GEC death and the promotion of cell survival by P. gingivalis appears to set off the bacterium's capacity to disseminate directly from cell to cell through an actin cytoskeleton-dependent process, where the cortical actin filaments are assembled to form membranous projections to transmit the bacteria to adjacent host cells (Yilmaz et al., 2006).Exploitation of the host-cell actin cytoskeleton thus appears to be a common mechanism for spreading of persistent intracellular bacteria.The intercellular movement of bacteria has been postulated to facilitate both local and systemic spread while avoiding the host immune response (Gouin et al., 1999(Gouin et al., , 2005)).The capacity of P. gingivalis to disseminate intercellularly appears to be acquired relatively late in the intracellular infectious process: little cell-to-cell spreading can be detected in early periods of the infection, but the spreading process gains momentum after 24 h.This was determined by novel flow-cytofluorimetry and fluorescence microscopy techniques (Yilmaz et al., 2006) (Fig. 1a, b).It is tempting to suggest that inducing resistance to apoptosis could represent a coordinated plan utilized by P. gingivalis to acquire adequate time to adapt and express necessary effector molecules to colonize and spread in the host tissues without provoking a full-bodied immune response.
Several potential mechanisms have been postulated for the ability of P. gingivalis to invade into deeper layers of gingival epithelium in order to gain access to subepithelial tissues, such as through transcellular or paracellular pathways (McCormick, 2003).The extracellular cysteine proteinases of P. gingivalis, RgpA, RgpB and Kgp, have been implicated in the breakdown of epithelial transmembrane proteins, E-cadherin, b 1 integrin and occludin, which are essential components of epithelial function.These have been shown to form a mechanical barrier against microbes studied in the Madin-Darby canine kidney (MDCK) cell line and the HeLa derivative KB cancer cell line (Chen et al., 2001;Katz et al., 2000).Similarly, a recent study suggested that transmission of the bacterium between KB cells occurred when the bacterium released itself into the paracellular space before it infected the adjacent cell (Li et al., 2008).Interestingly, a recent whole-cell quantitative proteomic study examining the change from extracellular to intracellular life for P. gingivalis demonstrated that the production of several proteases, including the classical virulence factors RgpA, RgpB and Kgp, was decreased during the infection of GECs (Xia et al., 2007).
In the context of new information concerning co-existence of P. gingivalis with GECs, a recent in vitro infection study performed with primary GECs provided a significant step forward (Yilmaz et al., 2008).It identified the secretion of a putative P. gingivalis effector/anti-apoptotic molecule known as P.g.-NDK, a distant homologue of nucleoside diphosphate kinase (NDK).NDK has also been shown to be produced by successfully opportunistic and persistent bacteria such as Mycobacterium tuberculosis and Pseudomonas aeruginosa (Kamath et al., 2000;Zaborina et al., 1999).Yilmaz et al. (2008) showed that the NDK, an ecto-ATPase produced by intracellular P. gingivalis, can interfere with the P2X 7 receptor-dependent apoptosis of GECs by consuming ATP e and preventing activation of P2X 7 receptors by ATP ligation.It has been previously recognized that ATP e ligation of P2X 7 receptors on macrophages triggers a variety of cellular effects, including activation of an inflammasome, secretion of IL-1b, mediation between the killing of intracellular bacteria, and the induction of host cell death by apoptosis and/or necrosis.Therefore, ATP released from infected cells or stressed cells at sites of inflammation is viewed as a generic 'danger signal' that can alert the innate immune system to the presence of infection (Khakh & North, 2006;Mariathasan & Monack, 2007;Schwiebert & Zsembery, 2003;Surprenant et al., 1996).Conversely, these special signalling events can be modulated by opportunistic intracellular bacteria to promote their persistence in the host cells (Coutinho-Silva et al., 2001;Franchi et al., 2007;Zaborina et al., 1999).Thus, the P. gingivalis ATPscavenging enzyme, NDK, is secreted during infection and can modify host cell biology by suppressing cell death and allowing the organism to survive and proliferate for protracted periods in the gingival epithelium and possibly disseminate in the face of host immunity.
It was of further interest to investigate whether there is a positive association between the inhibition of host epithelial cell death and the induction of host cell proliferation.A very recent proteomic study examined long periods of P. gingivalis infection for the impact on the primary GEC cell cycle (Kuboniwa et al., 2008).The proteomic analysis illustrated that the infection induces broad-based changes in the level and phosphorylation status of proteins that exert multi-level control on the host cell cycle.The cell-cycle pathways involving cyclins, PI3K, and p53, were found to be notably affected.Cytofluorimetry analysis confirmed the predicted phenotype, showing that primary GECs infected with P. gingivalis had accelerated host cell proliferation in a fimbriaedependent manner (Kuboniwa et al., 2008).In addition, three-dimensional confocal scanning fluorescence microscopy has shown that 24 h infected primary GECs harbouring high numbers of intracellular P. gingivalis undergo successful mitosis (O ¨.Yilmaz, unpublished data; Fig. 2).A promising implication of these findings is that successful persistence may require the inhibition of host cell death (prolonged cell survival) and the induction of the host cell cycle (increased cell proliferation) to occur simultaneously in order to balance the host biology and environment.This may enable the colonizing bacterium to effectively adapt to a changing milieu only when there is an equilibrium provided by the negative regulation of physiological host cell death.This hypothesis may explain the in vivo findings with oral epithelial cells showing remarkable tolerance to a vast number of intracellular bacteria, as those cells show no significant levels of cell death (Rudney & Chen, 2006).
Secular views of a grand topic: relevance of P. gingivalis to disease A compelling amount of information indicates that P.
exerts a high level of control on the differential expression of various critically important molecules in order to successfully adapt and persist in the host epithelium.At the same time, the organism elicits an array of gingival epithelial cell responses through highly structured and regulated signalling networks, which have been discussed in detail above.What formerly may have been isolated peaks in P. gingivalis and oral epithelium research have now resolved into a detailed landscape.Thus one can now distinguish the full panorama of dynamic connections between the bacterium and the host.The local anatomy, physiology and ecology of the host have been shown to play an important biological role, essential for the haemostasis of epimucosal tissues, namely the junctional epithelium.This is the epithelial component of the gingival sulcus and it serves as a major site of and the initial interactive interface for P. gingivalis colonization in the human oral cavity (Bosshardt & Lang, 2005;Schroeder & Listgarten, 1997).The junctional epithelium lacks differentiation and keratinization, and is known for its high cellular turnover.Desquamation occurs apically toward the bottom of the sulcus, while mitotic activity is very high in the suprabasal layers of the junctional epithelium called DAT cells.These are directly attached to the tooth and face the tooth surface (Bosshardt & Lang, 2005).Intriguingly, these features of the junctional epithelium present a logical route for the colonization strategies identified in undifferentiated primary epithelial cell cultures of gingiva (Yilmaz et al., 2006).Hence, the surface layers (suprabasal-DAT cells) invaded by the bacteria could disseminate in the junctional epithelium intercellularly to the bottom of the sulcus during the regular process of proliferation.Further, the ability of P. gingivalis to prolong host cell survival and enhance the proliferation of the infected host cells would allow the organism to establish itself in the gingival epithelium, and to contribute to disease when other host and microbial factors are conducive to the process of tissue destruction.
Based on the above information, we can logically hypothesize that P. gingivalis displays the characteristics of a bacterium that has evolutionarily co-existed with its host and that there is a balance between the organism and the oral mucosa shaped by this long-standing association.Moreover, the fact that P. gingivalis possesses a fascinating variety of potential virulence factors supports the increasingly popular view that commensal bacteria can become parasitic, amphibiotic or symbiotic depending on the context (local microenvironment and systemic host factors).Thus, how pathogenesis or susceptibility to inflammatory diseases such as periodontitis is facilitated can be truly appreciated by deciphering the co-evolved microbe-host biphasic interaction and the complex interrelationships between the oral microbiome and the host.While such a multi-faceted approach presents many remaining challenges, it certainly provides vast opportunities to answer long-standing questions.

Fig. 1 .
Fig. 1.(a) Intercellular translocation of P. gingivalis through actin fibres at 24 h post-infection visualized by immunofluorescence microscopy.The actin cytoskeleton is stained red and P. gingivalis is green.DAPI (blue) was used to stain the nuclei to determine the localization of P. gingivalis in the cytoplasm.The white arrow indicates actin projections and bacterial translocation between the host cells.Bar, 10 mm.[This panel is reproduced from Yilmaz et al. (2006) with permission from the American Society for Microbiology.](b) P. gingivalis spreading visualized by immunofluorescence microscopy.After 24 h co-incubation of initially 24 hinfected cells (blue) with initially uninfected cells (green), transmission of P. gingivalis to newly infected cells (green) can be seen, displaying red-labelled P. gingivalis in their cytosol.White arrows indicate newly infected cells.Bar, 10 mm.

Fig. 2 .
Fig. 2. Three-dimensional confocal scanning fluorescence microscopy showing a 24 h-infected primary GEC (actin, red; nuclei, blue) with high numbers of intracellular P. gingivalis (green) undergoing successful mitosis.White arrows indicate the infected cell producing two daughter nuclei.Bar 10 mm.