T he R ole of A cquired I mmunity and P eriodontal D isease P rogression

Abstract

Our understanding of the pathogenesis in human periodontal diseases is limited by the lack of specific and sensitive tools or models to study the complex microbial challenges and their interactions with the host’s immune system. Recent advances in cellular and molecular biology research have demonstrated the importance of the acquired immune system not only in fighting the virulent periodontal pathogens but also in protecting the host from developing further devastating conditions in periodontal infections. The use of genetic knockout and immunodeficient mouse strains has shown that the acquired immune response—in particular, CD4⁺ T-cells—plays a pivotal role in controlling the ongoing infection, the immune/inflammatory responses, and the subsequent host’s tissue destruction. In particular, studies of the pathogen-specific CD4⁺ T-cell-mediated immunity have clarified the roles of: (i) the relative diverse immune repertoire involved in periodontal pathogenesis, (ii) the contribution of pathogen-associated Th1-Th2 cytokine expressions in periodontal disease progression, and (iii) micro-organism-triggered periodontal CD4⁺ T-cell-mediated osteoclastogenic factor, ‘RANK-L’, which is linked to the induction of alveolar bone destruction in situ. The present review will focus on some recent advances in the acquired immune responses involving B-cells, CD8⁺ T-cells, and CD4⁺ T-cells in the context of periodontal disease progression. New approaches will further facilitate our understanding of their underlying molecular mechanisms that may lead to the development of new treatment modalities for periodontal diseases and their associated complications.

Abbreviations used in the paper are as follows: Antibody, Ab; antigen, Ag; antigen-presenting cells, APC; Actinobacillus actinomycetemcomitans, A. actinomycetemcomitans or Aa; β₂-microglobulin, β₂m; cytotoxic CD8⁺ αβ T-lymphocytes, CTL; dendritic cells, DC; delayed-type hypersensitivity, DTH; immunoglobulin, Ig; Fc receptor, Fc-R; interferon-γ, IFN-γ; receptor activator of NF-κB ligand, RANK-L; molecular weight, MW; Porphyromonas gingivalis, P. gingivalis or Pg; localized juvenile periodontitis, LJP; lipopolysaccharide, LPS; mouse mammalian tumor virus, MMTV; non-obese diabetic and severe combined immunodeficiency mice, NOD/SCID mice; osteoclast, OC; T-helper cells, Th; superantigen, SAg; transforming growth factor-β, TGF-β; secretory-IgA, s-IgA; T-cell receptor, TCR; T cytotoxic-1 cells, Tc1; and T cytotoxic-2 cells, Tc2.

(1) Introduction

The periodontium that anchors the teeth to the jaws consists of the gingiva, periodontal ligament, cementum, and alveolar bone. It is normally in a balanced state with the periodontal microbiota in the dental plaque (biofilm). Human periodontal diseases (i.e., gingivitis, periodontitis) result from heterogeneous etiologies including complex biofilm in the subgingival microenvironment, social and behavior modulations, and genetic or epigenetic traits of the host, each of which is influenced and/or modulated by the host’s immune and inflammatory responses. As a result of the maturation and changes in the biofilm, mainly an increase in facultative anaerobic, Gram-negative micro-organisms (Socransky et al., 1998), early vascular changes occur in the periodontium, with exudation and migration of phagocytic cells, including neutrophils and monocytes/macrophages, into the junctional epithelium and gingival sulcus, resulting in initial gingival inflammation. These changes are accompanied by increases in the size of the connective tissue infiltrated by leukocytes, loss of perivascular collagen fibers, and proliferation of junctional epithelium. During the early stage, the inflammatory infiltrate is mostly T-cells, whereas in the established lesions, B-cells become the most common inflammatory cells (Page and Schroeder, 1976). These changes signify a local alteration of immunoregulatory events in the host. The resulting cellular and fluid exudates cause further breakdown of the adjacent connective tissue and epithelium, followed by proliferation, apical migration, and lateral extension of the junctional epithelium. All of these alterations contribute to periodontal pocket formation. The pathogenic species present in the subgingival biofilm release an array of virulence factors that can evade anti-bacterial host defense mechanisms and then cause damage to the host tissue via immune/inflammatory interactions, which typically consist of neutrophils, monocytes/macrophages, dendritic cells (DCs), T-cells, and predominantly IgG-producing plasma cells. As the disease proceeds to more advanced stages, tissue destruction involves significant alveolar bone resorption and continuing loss of the collagen needed for tissue attachment. Widespread manifestations of inflammatory and pathological responses associated with periods of quiescence and active exacerbation become evident (Nabers et al., 1988; Papapanou et al., 1989; Beck, 1996). Further study and understanding of the underlying pathological mechanisms await more sensitive measurement or methods that differentiate between active and quiescent disease stages.

Despite the production of many virulence factors—such as proteases, metabolic and toxic by-products, enzymes, and lipopolysaccharides (LPS) from subgingival micro-organisms—that can cause adverse and deleterious effects, it is now generally believed that much of the damage that occurs in perio-dontitis is the result of host immune and inflammatory responses to the invading pathogens. Further, this process has been intensively studied and is thought to involve the active expression of both catabolic cytokines and inflammatory mediators, including IL-1, IL-6, TNF-α, PGE₂, etc., through the activation of monocytes/macrophages, lymphocytes, fibro- blasts, and other cellular elements involved. These cytokines and inflammatory mediators are capable of acting alone or together to stimulate alveolar bone resorption and collagen destruction via tissue-derived matrix metalloproteinases (i.e., collagenases), a major pathway for the breakdown of bony and soft connective tissue associated with periodontal disease activity (Offenbacher, 1996; Zambon, 1996). However, the pathological mechanisms underlying the progression of gingivitis to early periodontitis lesions remain unclear and will require further investigations. Therefore, based on the emerging data, this paper will focus on the nature of the host’s acquired immune responses along with some biological processes thought to provide immune protection against periodontal infections during the disease progression.

First, in the past two decades, there has been considerable interest in the study of the roles of different innate immune cells involved in the periodontal diseases. For example, macrophages and monocytes become activated after encountering periodontal pathogens like Actinobacillus actinomycetemcomitans (A. actinomycetemcomitans) and Porphyromonas gingivalis (P. gingivalis), and are capable of producing pro-inflammatory cytokines (i.e., IL-1, TNF-α) for local tissue destruction and phagocytosis of the micro-organisms. They can then further activate naïve T- and/or B-cells for subsequent adaptive immune responses (Ebersole and Taubman, 1994; Zambon, 1996; Baker, 2000). Neutrophils, another first line of the host’s defense, have been suggested to play both protective and destructive roles in periodontal diseases. This was primarily based on the observations that: (i) individuals suffering from neutrophil disorders (i.e., leukocyte adhesion deficiency syndrome, Chédiak-Higashi syndrome, cyclic neutropenia, etc.) have an increased susceptibility to and severity of periodontal tissue destruction, and (ii) neutrophil-derived matrix metalloproteinases (i.e., MMP-2, MMP-9; Teng et al., 1992; McCulloch, 1994) and/or pro-inflammatory cytokines (i.e., IL-1, IL-1 receptor antagonist, TNF-α) contribute to tissue breakdown of the native and/or denatured constituents of the periodontal tissues. Although these innate immune cells, such as neutrophils and macrophages, could, under normal circumstances, be primarily protective in the first-line defense, they are unable to clear the ongoing microbial loads and continuous assault once acute infection and exacerbation take place. Further, other innate immune cells, such as γδ T-cells and NK T-cells, have been shown to express significant amounts of various cytokines in vitro and in vivo after initial contact with foreign invaders (Chomarat et al., 1994; Lundqvist et al., 1994; Gadue and Stein, 2001). It has been postulated that the cytokine milieu produced in the local tissue environment or draining lymph nodes is critical in regulating or guiding the subsequent development of the adaptive immunity.

Presently, it is unclear how the innate immune system interacts with acquired immune responses to enforce and/or generate overall anti-bacterial immunity and the subsequent cytokine profile (i.e., type-1 or type-2 cytokines) involved at the ‘onset’ of the disease or more stable lesions (i.e., gingivitis) before significant tissue and/or alveolar bone destruction takes place (Taubman and Kawai, 2001; Teng, 2002). On the contrary, it has recently been shown, in an experimental model of mouse arthritis, that the acquired immune response may have significant influence on the activation of the subsequent innate immune responses through immunoglobulins (Igs) of B-cells, Fc receptors (Fc-R), and the complement network (Ji et al., 2002). Alternatively, other cell types (i.e., monocytes/macrophages or resident mucosal cells; Hofbauer et al., 1999, 2000; Rani and MacDougall, 2000) or the local immune microenvironment (Teng et al., 1995) may be critically involved in ‘directing’ or ‘polarizing’ T-helper (Th) cell differentiation (i.e., Th1 or Th2) in periodontal inflammation and/or destruction. Recent studies suggest that the maturational state of DCs can greatly influence the differentiation pathway of Th1 or Th2 effectors, depending on the cytokines expressed (MacDonald and Pearce, 2002). This suggests that the early contact and interactions between the acquired immunity and different innate immune cells (including γδ T-, NK T-cells, monocytes/macrophages, and dendritic cells; Bachmann and Kopf, 2002) may have a drastic impact on the development of periodontal disease and/or its progression. It has been shown that numerous DCs are involved in various parts of the infected periodontal tissues (Jotwani et al., 2001; Cirrincione et al., 2002), and DCs are the natural adjuvants of immune responses (Gallucci et al., 1999). Changes in the local environment or chemokine and cytokine development may, in turn, reduce or shut down the immunopathology and/or tissue damage during microbial infection (Reis e Sousa et al., 1999). Thus, the role of innate immune responses in periodontal disease progression clearly deserves further study in the future.

It has been shown that both humoral and cell-mediated immune responses play important roles in the host defense against microbial infectious disease such as human periodontitis (Ebersole and Taubman, 1994). However, the contribution of the acquired immune cells in the progression of periodontal disease has long been controversial, with its exact role in the protection vs. destruction of the host’s periodontium being unclear (Klausen, 1991; Ebersole and Taubman, 1994; Zambon, 1996). Likewise, the results of periodontal studies of humans with congenital and acquired immunodeficiency are variable and difficult to interpret, due to various medical complications (Oshrain et al., 1983; Van Dyke, 1991; Dahlén et al., 1993). As presented above, during the infectious process, inappropriate immune and inflammatory responses can induce tissue damage or even undesired systemic reactions and outcomes (Teng et al., 2002). Recent advances in the field suggest that the periodontitis-associated immune mechanism(s) is a double-edged sword, with one edge fighting the invading pathogens and the other triggering tissue damage in the host. Therefore, the concept that the underlying immune mechanism(s) of periodontitis is mainly beneficial and/or protective to the host needs to be reconsidered and justified. In the present review, some recent findings regarding the role of acquired immunity (in the order of the B-cell-mediated humoral immunity, CD8⁺ cytotoxic T-cell-mediated immunity, and CD4⁺ helper T-cell-mediated immunity) in the context of periodontal disease progression will be discussed. In addition, the discussion on CD4⁺ helper T-cell-mediated immunity in periodontitis will include: (i) pathogen-reactive immune repertoire and superantigen (SAg) activity; (ii) Th cells and cytokine network; and (iii) Th cells and bone diseases: OPG/RANK-L.

(II) B-cell-mediated Humoral Immunity and Periodontitis

The precise role of humoral immunity in periodontal disease progression has not been fully elucidated, although the production of antibody (Ab) response has long been suggested to be beneficial to the host in fighting periodontal infections (Ebersole and Taubman, 1994). In general, after encountering invading pathogens, in particular extracellular micro-orga-nisms, antigen (Ag)-specific naïve B-cells undergo affinity maturation via clonal selection, somatic hypermutation, and Ig receptor editing (Janeway and Travers, 1997a). As a result, activated B-cells bearing the higher or highest affinity for Ag are rescued from apoptosis (MacLennan et al., 1992). Subsequently, these B-cells with mutated B-cell receptor genes, located in the germinal centers of regional lymph nodes, undergo further isotype-switching and differentiate into effector or memory B-cells. Certain cytokines such as interferon-γ (IFN-γ), interleukin-4 (IL-4), IL-5, IL-10, IL-12, transforming growth factor-β (TGF-β), anti-CD40-Ligand/CD40, and bacterial lipopolysaccharide (LPS) are potent stimulators of Ab class-switching (i.e., IgG; McGhee et al., 1999; Nahm et al., 1999). Both IgM and IgG2a (an IgG isotype) activate the complement cascade to generate C3b and iC3b, which opsonize bacteria for phagocytosis, and IgG2a production is enhanced by IFN-γ. Further, other Ab isotypes like IgG1, IgG2b, and IgA can bind to different Fc-R isoforms on macrophages, monocytes, and neutrophils via a separate internalization process to eliminate the pathogens. However, some bacteria cell wall compounds (including peptidoglycan, LPS, and lipoproteins) stimulate B-cells directly and non-specifically via cross-linking without the help of T-cells (called T-independent Ag), resulting in polyclonal B-cell activation (i.e., IgG; Tew et al., 1989). Most bacteria release T-dependent Ag, comprised mostly of protein Ag that can be processed and presented by antigen-presenting cells (APC), including B-cells, macrophages, and CD4⁺ Th cells. It has been suggested that the production of cytokine IL-1 by B-cells involves polyclonal B-cell activation by this type of Ag (Wanatabe et al., 1996). An early Ab-mediated control of periodontal infection through this process may limit the bacterial spread and/or associated tissue damage, thereby reducing the bacteria load for the subsequent acquired immune responses.

It is clear that Ab produced by activated memory B-cells that reside in the secondary lymphoid tissues are positive for surface-Ig, B220, CD19, CD21, MHC class II, and CD40L, and the terminally differentiated plasma cells that reside in bone marrow are positive for CD23, CD138, and Blimp-1 (Turner et al., 1994; Lin et al., 1997). These cells are capable of producing Abs constitutively and are able to opsonize the invading pathogens, promote the FcR-mediated phagocytosis, neutralize the bacterial toxins, trigger complement-mediated lysis and Ab-dependent cytotoxicity, and prevent bacterial entry at the mucosal surfaces—all of which enable the effective immune responses to provide the host protection from developing undesirable diseases (Nahm et al., 1999). Several studies have shown that periodontal pathogen-specific Igs are effective in reducing or inhibiting bacterial colonization, proliferation, and spread (Zambon, 1985; Ebersole, 1990; McArthur and Clark, 1993). It is also known that increased levels of local and/or systemic Ab production coincide with the history (exposure and/or bacterial load) of periodontal infection to specific micro-organisms (Ebersole et al., 2001). Subsequently, it was shown that the increased levels of specific Abs to one or more periodontal pathogens are generally associated with specific forms of periodontal disease (Ebersole, 1990). There is good evidence that most Igs, including secretory-IgA (s-IgA), IgA, and IgG, can be produced locally at the mucosal sites (Hagewald et al., 2000). In addition, systemic IgG also contributes to this part of the humoral immunity during periodontal infections (Kinane et al., 1999; Kinane and Lappin, 2001).

It was shown that specific subclasses of Ig (i.e., human IgA and IgG4) can reduce inflammation during chronic bacterial infections at mucosal sites (Fanger et al., 1997; Sutterwala et al., 1998), and this can also be achieved by opsonic Abs (i.e., IgG1 and IgG₂), which enable the neutrophils to kill invading pathogens like A. actinomycetemcomitans and P. gingivalis (Baker and Wilson, 1989; McArthur et al., 1989). A recent study by Rajapakse et al.(2002) showed that immunization with proteinase-adhesin complexes of P. gingivalis reduces colonization of the subgingival crevice by P. gingivalis and subsequent periodontal bone loss in the rat periodontitis model, where disease severity correlates with concomitant increases in the Ag-specific IgG₂ and IgA and a decrease in the IgG4 responses. This finding suggests that a Th2-associated humoral response (i.e., IgG4 and IgA in humans) directed against some bacterial Ags could be “protective” against periodontal infections (Taubman and Kawai, 2001; Seymour and Gemmell, 2001; Fokkema et al., 2002). However, conflicting data also exist regarding the nature and degree of protection being exerted (Williams et al., 1985; Ebersole et al., 1987; Magnusson et al., 1991). It has been shown that local or systemic production of pathogen-specific Igs often did not correlate with the clinical stages of the disease, such as gingivitis and initial and/or established periodontitis (Ebersole and Taubman, 1994; Tinoco et al., 1997; Hagewald et al., 2000; Albandar et al., 2001). Further, this is also the case when bacterial colonization and Ab titers became significantly lowered after clinical treatments (Grbic et al., 1999). Indeed, it has been documented that long-lived plasma cells that do not require Ag stimulation for Ab production can contribute to mucosal immunity. This pre-existing Ab participates in the front-line defense and is likely a key protective molecule against microbial infections (Hibi and Dosch, 1986). It was shown that locally applied Ab against P. gingivalis was effective in reducing bacterial colonization; however, systemically generated Ab was much less effective (Okuda et al., 1988). This suggests that humoral immunity (i.e., s-IgA, IgG) and/or other immune mechanism(s) (i.e., innate immunity) are involved in the first-line defense during early periodontal infection. For example, s-IgA production is elevated as the levels of local exposure to pathogens increase, and it can block bacterial adhesion to host cells (Wold et al., 1990).

Most of the human clinical and some animal studies have shown that serum Ab titers do not correlate with the clinical stages of periodontitis (Ebersole and Taubman, 1994; Albandar et al., 2001) and/or alveolar bone destruction (Evans et al., 1994; Baker et al., 1999a). However, immunization with P. gingivalis fimbrial protein, which generates high titers of anti-fimbrial Ab, can prevent alveolar bone loss in gnotobiotic rat and non-human primates when subsequently infected orally with whole P. gingivalis (Klausen, 1991; McArthur and Clark, 1993). It has been reported that sera from periodontitis patients are of low titers/affinities and thus non-protective (Zambon, 1985) in that they do not synergize with opsonization of the invading pathogens for phagocytosis (Van Dyke and Hoop, 1990; Ebersole et al., 2001), despite the production of large quantities of IgG₂ in patients suffering from localized juvenile periodontitis (LJP, recently called aggressive periodontitis; Ebersole et al., 1982). In contrast, compared with the generalized form of juvenile periodontitis subjects, LJP patients frequently have higher “titers” of IgG₂ that react with high-molecular-weight (MW) LPS or carbohydrate Ags, and these Ab responses have been shown to be associated with “less” severe disease in LJP (Zambon, 1985; Ebersole and Taubman, 1994). Therefore, specific IgG₂ may appear to be an effective opsonin to facilitate FcγR-mediated phagocytosis of Ags and pathogens for host protection (Wilson et al., 1995; Wilson and Bronson, 1997). This suggests that triggering a high IgG₂ response might help control periodontal infection and localize the disease. This effect is further supported by the finding that a humoral Ab-promoting Th2 clone protects the host from A. actinomycetemcomitans-induced alveolar bone destruction in a rat model of experimental periodontitis (Yamashita et al., 1991). Thus, temporary depletion of Ab-producing B-cells in vivo may lead to a more exaggerated periodontal bone loss in mixed infections with A. viscosus and P. gingivalis (Klausen et al., 1989). Therefore, Ab-produ-cing B-cells may confer some degree of protection against alveolar bone loss in periodontal disease progression. Further, Th2-associated PGE₂ expression may be at least partly involved in regulating specific IgG₂ production (or other isotypes). Since it is still unclear how immunoregulation modulates the pathogenesis of periodontal diseases (Gao and Teng, 2002), further study to decipher the molecular mechanisms of the humoral Ab response that is protective in one subject and harmful in another is needed (Offenbacher, 1996).

Overall, it is unlikely that this protection from periodontal disease progression is solely accountable by humoral Ab immunity (Klausen, 1991; Ebersole at al., 2001), and it is rarely able to accomplish this task in the absence of cell-mediated mechanisms (Yoshi et al., 1987; Taubman and Kawai, 2001). B-cells are not required, at least in the animal model, for alveolar bone destruction. In most of the studies, active immunization protocols were implemented, and Ab-mediated protections against periodontal bone loss were observed (Klausen, 1991; McArthur and Clark, 1993; Rajapakse et al., 2002). However, the results obtained have not been consistent, and this effect may vary greatly depending upon the specific bacterial Ags, LPS, and/or different pathogenic species used in the studies (Chen et al., 1990; Ebersole et al., 1991; Lamster et al., 1998; Moritz et al., 1998). Furthermore, the results of a separate study (Teng et al., 1999, 2000) where a cell depletion approach was used in the humanized NOD/SCID mouse model suggest that human B-cells do not play a significant role in A. actinomycetemcomitans-induced alveolar bone destruction in vivo. Mechanistically, this is in accordance with the study by Baker et al.(1994), who showed that an immunodeficient state in SCID mice (without B- and T-cells) results in “decreased” periodontal bone loss, compared with that of the immuno-competent host when orally infected by P. gingivalis. In that study, it was the CD4⁺ T-cells, not B-cells, which were primarily involved in mediating alveolar bone loss in vivo (Baker et al., 1999a,b). Further, it is noteworthy that active immunization protocols were not used in both studies (Baker et al., 1994; Teng et al., 2000), and the different results generated by the use of different animal species and experimental protocols cannot be overlooked (i.e., from rats; Klausen, 1991). Collectively, these findings are consistent with previous notions that: (i) Abs can inhibit bacterial colonization in gingival crevices, and (ii) B-cell-deficient mice are more susceptible to bacterial infections following both primary and secondary immune responses (Teitelbaum et al., 1998; Hou et al., 2000; Mittrucker et al., 2000). Pathogen-reactive B-cells and Abs can contribute to the control of periodontal infections in the mucosa, although they are unable to complete the task without the help of other immunological armors. In addition, these findings can be used to underscore the failure of efficient Ab production, and/or their function(s) may be one of the reasons for the shift from stable gingivitis to destructive periodontitis. Further, the results of a recent study in humans showed that IgG titers against P. gingivalis infection were significantly lower in periodontitis sites than those in gingivitis sites (Mooney and Kinane, 1997). Thus, the quality of the humoral Ab response may somehow affect the progression of the periodontal infections.

To avoid the attack by host Abs, many invading pathogens express antigenic variations of their cell wall or capsule constituents, thus allowing the bacteria to escape recognition (i.e., low or cryptic antigenicity) and elimination by specific Abs generated in previous contacts or to prevent internalization and subsequent killing (phagocytosis) by neutrophils, monocytes, and macrophages. In addition, some periodontal pathogens produce “attacking” arsenals (i.e., proteases) that destroy IgA, IgG, and even B-cells to counteract the host’s immune protection (Spitznagel et al., 1995; Gronbaek Frandsen, 1999; Yun et al., 2001). Therefore, orchestrating the bacterial evasion, opsonization, phagocytosis, cell-mediated immune response, and regulation by the cytokine network could successfully lead to Ab-mediated immune protection during periodontal infection.

Collectively, periodontal Ab-producing B-cells are, initially, protective for the host (Hou et al., 2000). However, this protection may be inadequate or of insufficient longevity to deal with the continuous assault from biofilm and/or chronicity of the infectious disease, especially when bacterial evasion may have affected other immune armors. Meanwhile, the productive memory B-cells and Ab-producing plasma cells may eventually become “crippled” or “degenerated” and thus fail to halt the invading pathogens (i.e., bacterial evasion) that trigger further immune/inflammatory responses associated with tissue destruction (Ebersole et al., 2001). As a result, Ab-mediated protection is incapable of achieving total or sterile eradication of the invading periodontal pathogens and, hence, is sub-optimal. Thus, this “failure of protection” may tip the balance toward a more destructive process during disease pathogenesis. On the other hand, the role of “activated” B-cells serving as APC to activate naïve and/or memory T-cells may be taken over by other more professional APC, such as macrophages and dendritic cells (Jotwani et al., 2001), for cell-mediated immunity. That may be the reason why the humoral immune response to invading pathogens usually precedes, but does not prevent the onset or progression of, periodontal tissue destruction in vivo (Baker et al., 1999a). Therefore, Ig and B-cells alone, at the cellular level, do not usually confer full protection without the involvement of T-cell-mediated immunity. A recent study by Choi et al.(2001) showed that B-cells can produce RANK-L, TNF-α, IL-6, MIP-1α, and MCP-3, which are involved in the activation of osteoclastogenesis for bone resorption in vitro, and, in addition, that periodontal B-cells also secrete IL-1β, raising the possibility that B-cells by themselves may play an important role in periodontal bone destruction. Futher studies are required to understand the cellular and molecular mechanisms underlying the failing humoral immune system in the context of periodontal disease progression before one attempts to design active immune thera-peutics or vaccines for diagnostic and treatment needs.

(III) CD8 Cytotoxic T-cell-mediated Immunity and Periodontitis

The role and contribution of classic cytotoxic CD8⁺ αβ T-lymphocytes (CTL) in periodontal disease progression are not well-understood. Several earlier studies have demonstrated that, in general, there is a decrease in CD4⁺ T-cells and an increase of CD8⁺ CTL (a depressed CD4/CD8 ratio) in the infected gingival tissues and/or human peripheral blood leukocytes (HuPBL) samples of periodontitis subjects, when compared with those of the healthy subjects (Suzuki et al., 1984; Stoufi et al., 1987; Kinane et al., 1989; Nagasawa et al., 1995). In addition, there appears to be a preferential usage of certain human T-cell-receptors (TCR; i.e., Vβ genes) among CTLs in specific pathogen-infected periodontal tissues (Gemmell et al., 1998). CTLs can be found next to the periodontal fibroblasts in deep gingival tissues, where direct cytopathic and degenerative changes occur (Page and Schroeder, 1976). These findings suggest that the CTL existing in the gingival tissues may be an integral part of immune/inflammatory response in periodontal infections (Seguier et al., 1999; Takeichi et al., 2000).

The development of CTL requires the recognition of foreign Ag peptides by MHC class I molecules (or HLA-A, B, C in humans). The short peptides (i.e., 5–15 amino acids) are generated in the “proteosome” machinery of the infected cells, then transported for Ag processing (7–9 a.a.) before being loaded onto the endoplasmic reticulum, where MHC class I molecules pair non-covalently with β₂-microglobulin (β₂m) prior to the final expression of these molecules on the cell surfaces of all nucleated cells. To date, most of the periodontal micro-organisms under study are known to be extra-cellular pathogens which are destined to MHC class II pathway within APC for Ag processing and presentation. However, foreign peptides that are presented by the MHC class I pathway are usually generated from endogenously synthesized proteins or proteins that are secreted into the cytoplasm (Pamer and Cresswell, 1998). It has been reported that some pathogens, including Salmonella enterica and Mycobacterium tuberculosis, can elicit CD8⁺ T-cells response through a phagosomal or phago-lysosomal pathway via different mechanisms (Shaible et al., 1999). It is currently unknown whether similar pathway(s) can occur in periodontal micro-organisms like A. actinomycetemcomitams, P. gingivalis, and Treponema denticola. It is noteworthy that these three particular periodontal pathogens have been suggested to invade (or trans-endocytose) gingival sulcular epithelia for survival or tissue invasion in vitro or in vivo (Lamont et al., 1995; Meyer and Fives-Taylor, 1997; Lux et al., 2001), and, as a result, CTL might be generated for potential effector function(s).

Two recent independent studies using the gene-targeted knockout approach and Ab-mediated cellular depletion, respectively, have come to the same conclusion: that CTL is not involved in alveolar bone destruction in the mouse model of experimental periodontitis. In the first study, the extent of alveolar bone loss detected in the immuno-competent mice inoculated with P. gingivalis was comparable with that observed in the β₂M gene-targeted knockout mice that do not develop CTL in vivo (Baker et al., 1999b). In the second study, CD8⁺ T-cells were depleted in vivo by Ab-mediated cytotoxicity before A. actinomycetemcomitans was orally inoculated into the humanized NOD/SCID mice, which received A. actinomycetemcomitans-reactive HuPBL from LJP patients for engraftment (Teng et al., 2000). CTL depletion in these mice did not result in any difference regarding alveolar bone loss compared with that of the HuPBL-NOD/SCID mice carrying the full HuPBL engraftment. However, CD4⁺ T-cell depletion was associated with reduced alveolar bone loss. Together, these findings demonstrate that CTL is not a major player in immune-mediated periodontal tissue destruction in vivo. Interestingly, activated CTL can express high levels of RANK-L (OPG-L, TRANCE, ODF), CD40L, and TNF-α after encountering specific Ags or mitogens in vivo, and they have been implicated in some human and mouse arthritic diseases (Kong et al., 1999b), but not in periodontal bone loss. Nevertheless, the potential role of CTL in complementing or aiding other immune protection mechanisms in periodontal disease has not been fully examined or addressed.

Recent studies have shown that naïve CD8⁺ T-cells, as is the case with CD4⁺ T-cells, can also differentiate into at least two subsets with distinct cytokine expression patterns: T cytotoxic-1 (Tc1) cells that secrete type-1 cytokines, including IL-2, TNF-α/β, and IFN-γ; and Tc2 cells that produce type-2 cytokines, including IL-4, IL-5, IL-6, and IL-10 (Mosmann et al., 1997). Both subsets are “cytotoxic” via the perforin/granzymes and/or Fas/Fas-L pathways (see description below), and both efficiently kill resting and activated B-cells and other infected nucleated cells in the host (Li et al., 1997). Both Tc1 and Tc2 induce similar inflammatory cell infiltrates; however, perforin-ablated Tc1 and Tc2 cells were still able to induce delayed-type hypersensitivity (DTH; a classic cell-mediated immune response), although at lower levels, suggesting that perforin-mediated cytotoxicity of CD8⁺ T-cells is not essential for a DTH response (Li et al., 1997). Recent studies have shown that there is a Tc2-like cytokine expression profile in the gingival tissues of periodontitis subjects (Wassenaar et al., 1996; Petit et al., 2001). This finding suggests a potential role of CTL in local immune regulation and promotion of humoral immune response during periodontal infections. Co-existence of predominantly IFN-γ-producing Tc1 cells in periodontally infected gingival tissues has also been reported (Takeichi et al., 2000). Since Tc1 and Tc2 also exhibit regulatory loops that inhibit the reciprocal response in vitro and in vivo, similar to Th1 and Th2 with cross-regulatory properties (Mosmann et al., 1997; Morel and Oriss, 1998), induction of an inappropriate Tc differentiation may render the host more susceptible to infection or disease (Shen et al., 1998; Takeichi et al., 2000). The influence and differential contribution of Tc1 and Tc2, if any, to periodontal disease progression remains unclear to date.

The anti-microbial effects of CD8⁺ CTL effectors can be accomplished by several mechanisms: (i) expression of cytokines such as IFN-γ and TNF-α that can activate macrophages for killing (innate immunity), exactly like CD4⁺ Th cells; and (ii) direct lysis of the infected target cells via either the perforin/granzyme pathway or the Fas/Fas-L pathway (Janeway and Travers, 1997b). In the former case, activated CTL are efficient producers of IFN-γ and are likely to initiate the innate immunity cascade via the IL-12 pathway and amplify downstream an acquired Th1 immune response in the host (see below). In parasitic (i.e., helminthes) and viral infections of mice, memory CTLs have been shown to express immunoregulatory properties by promoting IL-12 and Th1 protective immunity (Sedegah et al., 1994). Whether CD8⁺ T-cells exhibit a similar trait in periodontal infection with such an auxiliary pathway awaits further studies. In the latter situation, perforin can induce cell death by itself, but its function is mainly to introduce granzymes A and B (serine proteases produced by CTL and NK cells) into the infected target cells via the caspase cascades for apoptosis (Sarin et al., 1998; Metkar et al., 2002). However, the Fas/Fas-L pathway is primarily used to regulate lymphocyte activation (Shresta et al., 1998) and granulocyte reactivity (Jewett et al., 2000; Gamonal et al., 2001) and to prevent over-reactivity in both cell types. This process, although not presently well-understood, may be a potentially important step in the protection of periodontal tissues from damage due to microbial infection and associated inflammation (Gamonal et al., 2001). Recently, CTLs additional bactericidal activity has been shown to be mediated by granulysin, which is introduced into bacteria-containing phagosome via a perforin-dependent pathway (Pena et al., 1997; Stenger et al., 1998). Subsequently, macrophages and dendritic cells can take up Ag from cells undergoing apoptosis and/or apoptotic cells and induce an Ag-specific CD8⁺ CTL response (Albert et al., 1998). Although activation of apoptotic signaling pathways has been described in diseased periodontal tissues, there is lack of direct evidence that these events are directly associated with or the result of CTL interactions in situ (Sawa et al., 1999; Gamonal et al., 2001).

A specialized set of innate immune cells, CD8⁺ γδ T-cells that respond to epithelial surface CD1 molecules (a group of non-polymorphic MHC molecules encoded by genes located outside MHC loci), can recognize glycolipid Ag (i.e., lipoarabinomannan, glucose monomycolate) from a Gram-negative bacterium without Ag processing. These T-cells are able to secrete various cytokines (IFN-γ, IL-10, TNF-α, TGF-β, and IL-6) and exert cytolytic activity (Chomarat et al., 1994; Lundqvist et al., 1994). Thus, it has been postulated that, via cytokine expression and cytolytic activity, CD8⁺ γδ T-cells can modulate the development of adaptive immune response (Mak and Ferrick, 1998; Sugita and Brenner, 2000). It has been shown that the number of activated intra-epithelial CTL increases in the peripheral blood and local gingival tissues in periodontitis subjects (Nagai et al., 1993; Lundqvist et al., 1994; Gemmell and Seymour, 1995; Takeichi et al., 2000). However, their actual contribution in the linkage of innate immunity to acquired immune responses under normal conditions and during periodontal infection is still not clear and will require further study.

In conclusion, CD8⁺ CTL do not directly participate in periodontal tissue destruction during disease progression; however, they may play some protective roles in fighting periodontal pathogens by generating important cytokines for both innate and adaptive immune responses and by cytolytic killing of bacteria-infected or -damaged tissues and cells.

(IV) CD4 T-cell-mediated Immunity and Periodontitis

Recent advances in molecular and cell immunology have led to a better understanding of how the acquired immune response—in particular, CD4⁺ Th cells—works to battle microbial infections in the periodontal and oral tissues. It is now known that Th cells are the pivotal cell type in modulating and regulating the acquired immune system, composed of B-cell-mediated humoral immunity and CD4⁺ Th and CD8⁺ CTL-induced cell-mediated immunity. For example, Th cells are required for Ab maturation and specific Ig isotype-switching in B-cells, while B-cells, as APC, can present exogenous Ag to induce Ag-specific T-cell responses. In addition, Th cells can modulate Ag-specific CTL activity in a specific local microenvironment (Keene and Forman, 1982). The generation of specific CD4⁺ Th cells is initiated by the recognition of longer peptides (12–25 a.a. in length) present in the binding grooves of MHC class II molecules (in humans: HLA-DR, DQ, DP) on the surfaces of specialized APC such as macrophages, dendritic cells, and activated B-cells. The presenting peptides are usually derived from exogenous sources, such as extracellular periodontal micro-organisms and their shed antigens and toxins. They are physically processed in acidified late-endosomal or lysosomal vesicles by proteases (i.e., aminopeptidases and cathepsins; Pieters, 1999) before binding to empty MHC class II molecules. They must then undergo final transportation to the cell surface for activation of Th cells. Further, for full activation of naïve T-cells through signaling of tri-molecular complex TCR/MHC-peptide (called signal 1), it is now clear that co-stimulation (called signal 2) is absolutely required. Typically, APC express co-stimulatory molecules such as B7-1 (CD80), B7-2 (CD86), and CD40, which will engage T-cells and amplify TCR signaling (signal 1) through their corresponding receptors (CD28, CD40-L). CD28 and CD40-L are the best-characterized signal 2 molecules that mediate co-stimulation in T-cell activation. Engagement of CD40 and CD40L provides an indirect signal mediated through the up-regulation of other co-stimulatory molecules, such as B7s, on the surfaces of APC. Interaction between B7s and CD28 delivers a direct stimulus to T-cells. B7-1 and B7-2 have been reported to promote Th1 and Th2 responses, respectively (Seder and Paul, 1994). Other molecules—including LFA-1, ICAM-1, 2, 3, and gelactin—have been shown to engage in co-stimulation activity (Janeway and Travers, 1997b; Taubman and Kawai, 2001).

It has been shown that fine-tuning of the co-stimulation can lead to up- or down-regulation of T-cell activation through CD28 (a positive signal 2) or CTLA-4 (a negative signal 2; Alegre et al., 2001). Binding of CTLA-4 to B7 molecules plays an essential role in limiting the actively proliferating responses of activated T-cells to Ags and B7s on the surfaces of APC. Kawai et al.(2000) recently showed that, in vivo, infusion of soluble CTLA-4-Ig to antagonize CD28-mediated signaling through B7 molecules abolished the inflammatory Th1 immune response (see below) associated with periodontal bone destruction in a rat model of experimental periodontitis. This suggests that periodontal inflammation may be under the control of general pri-ming and activation schemes of cell-mediated immunity and co-stimulatory signaling. Professional APC—such as activated B-cells, DCs, and macrophages—express MHC class II and class I molecules after being activated/primed. DCs reside in mucosal surfaces or epithelium (i.e., Langerhans cells), and they can transport Ag to local lymph nodes to initiate a primary immune response. DCs are the most powerful APC for priming naïve T-cells and triggering Th1 immunity against microbial infections. Macrophages can also modulate APC activity by “enhancing” or “suppressing” the immune response, depending on the nature of the stimuli (Hu and Klempner, 1997).

Earlier studies concerning the role of cell-mediated immunity in periodontal disease progression demonstrated very mixed results and with different variables involved (for review, see Klausen, 1991; McArthur and Clark, 1993). These conflicting data have hindered any conclusion regarding which component(s) or cell type(s) is/are critically responsible for periodontal disease progression. It was then suggested that a generalized or selective immune suppression can aggravate periodontal disease, and the immunological status of the host at the time of infection with periodontal pathogens seemed important for the development of periodontitis (Brandtzaeg and Kraus, 1965; Klausen, 1991). It has now become clearer that alveolar bone destruction, the hallmark of periodontitis, is mediated by elevated osteoclast activity during disease pathogenesis. Recent studies also show that interactions between the cell-mediated immune responses—in particular, CD4⁺ T-cells, bone cells, and the microenvironment—are critical for the development of alveolar bone loss (Theill et al., 2002; see below). The discussion on the contributions of Th cells to periodontal disease progression will be presented in the following order: (a) pathogen-reactive immune repertoire and superantigen activity, (b) Th cells and cytokine network, and (c) Th cells and bone diseases: OPG/RANK-L interaction.

(A) Pathogen-reactive immune repertoire and superantigen (SAg) activity in periodontitis

The T-cell recognizes Ags via its heterodimeric TCR composed of an α-chain and a β-chain (Vα and Vβ). Immune diversity of T-cell recognition is generated by the recombination of variable (V), diversity (D), joining (J), and constant (C) segments of the TCR genes. This process is further facilitated by the N region in addition to non-germline encoded nucleotides at the V-(D)-J junctions (Davis and Bjorkman, 1988). Ags are generally recognized by the binding site of the V regions of both TCR Vα and Vβ genes (Davis and Bjorkman, 1988). Fine specificity of Ag recognition has been shown to be determined by the complementary-determining region 3 of the TCR genes (Davis and Bjorkman, 1988; Boitel et al., 1992). It is generally recognized that each T-cell retains a distinct clonotype of its TCR responsible for Ag-specific T-cell response.

Several recent studies have assessed the potential immune repertoire of human periodontitis patients by PCR analyses with single-strand conformation polymorphism and nucleotide sequencing and/or flow-cytometry analyses. The results have shown that there is a relatively limited TCR immune repertoire involved, probably reflecting a limited number of T-cell specificities at the local site of periodontal inflammation (Mathur et al., 1995; Nakajima et al., 1996; Yamazaki et al., 1996; Geatch et al., 1997; Gemmell et al., 1997; Berglundh et al., 1998; Preshaw et al., 1998). This suggests that T-cells infiltrating periodontitis lesions are responding to a restricted number of bacterial Ags or epitopes during the disease processes. In contrast, in a separate, larger-scale analysis of TCR (30) Vα and (22) Vβ genes by quantitative PCR, it was found that: (i) there is more than limited or biased (i.e., > oligoclonal: relatively diverse) CD4⁺ TCR immune repertoire used by A. actinomycetemcomitans-reactive human periodontal CD4⁺ T-cells; and (ii) there are a few dominant TCR genes shared by the majority of LJP patients (Gao and Teng, 2002). This suggests that limited numbers of microbial Ags or epitopes are commonly recognized by different patients with similar disease characteristics (i.e., LJP). These observations differ slightly from those found in “general” periodontal infections, where most of the studies lacked well-defined microbial specificities for the immune repertoire analyzed. One possible explanation is that the total numbers of TCR genes involved in the above study (Gao and Teng, 2002) were greater than those of the other studies. Nevertheless, the issue of human immune repertoire associated with specific periodontitis will await further study with more careful and systemic analysis of T-cells at the clonal level.

Interestingly, some periodontal micro-organisms, including Prevotella intermedia and A. actinomycetemcomitans, have been suggested to possess SAg activity (Mathur et al., 1995; Zadeh et al., 1996). Unlike conventional Ags, SAg (including bacterial SEA, SEB, TSSC-1, etc.) and mouse mammalian tumor virus (MMTV) are known to bind to the outer surfaces of the MHC-II binding grooves and certain specific TCR Vβ genes independent of TCR Vα gene usage, thereby activating a larger population of T-cells (from 2 to 30%, regardless of their MHC and peptide specificity) than that processed by conventional Ag (Li et al., 1999). The net result of this massive T-cell activation by SAg includes over-production of T-cell-derived cytokines, bystander activation of other non-specific immune cells, and, subsequently, a down-regulation of specific immune responses (i.e., anergy induction; MacDonald et al., 1993; Ulrich et al., 1995). However, the results of the studies regarding SAg activity in periodontal micro-organisms have shown some conflicting results. Several studies have shown that there is a rather biased immune TCR Vβ gene expression in the mononuclear cells of the infected periodontal tissues (Mathur et al., 1995; Nakajima et al., 1996; Yamazaki et al., 1996; Geatch et al., 1997; Gemmell et al., 1997; Berglundh et al., 1998; Preshaw et al., 1998), suggesting that some periodontal pathogens may possess certain SAg activity (Mathur et al., 1995; Nakajima et al., 1996; Zadeh et al., 2001). Conversely, others have shown no SAg-like biased TCR Vβ expression (Mathur et al., 1995; Petit and Stashenko, 1996). Furthermore, by comparison of immune reactivity pattern and TCR Vα and Vβ gene usage in A. actinomycetemcomitans-reactive human periodontal CD4⁺ T-cells of the infected tissues from LJP patients and in a humanized mouse model of experimental periodontitis, it was found that there is no evidence of T-cell anergy, and that the reactive CD4⁺ T-cells express the functional activation marker CD25 (IL-2R_α chain), CD69, and IL-2 (Teng et al., 2000; Gao and Teng, 2002). In addition, there is a more restricted TCR Vα gene usage than Vβ in the responding T-cell pool, suggesting that at least A. actinomycetemcomitans possesses no SAg characteristics. Additional studies have also come to the same conclusion about this particular and some other periodontal micro-organisms (Mathur et al., 1995; Petit and Stashenko, 1996; Gemmell et al., 1998, 1999). It is important that both cellular and molecular criteria regarding SAg characteristics be considered and applied before one can draw any definitive conclusion.

(B) Th cells and cytokine network in periodontitis

After being properly primed and activated, naïve CD4⁺ Th cells differentiate into effector and/or memory phenotypes that can be functionally categorized into several subsets based on the cytokines expressed: Th1 cells produce IL-2, IFN-γ, and TNF-α/β (type-1 cytokines); and Th2 cells produce IL-4, IL-5, IL-6, IL-10, and IL-13 (type-2 cytokines; Seder and Paul, 1994). The intermediate “Th0” cells express IL-2 and low levels of mixture of both Th1 and Th2 cytokines like IL-3, IFN-γ, and IL-4 (Paliard et al., 1988; Street et al., 1990). Meanwhile, other more discrete cytokine expression patterns, such as IL-2, IL-4, and IL-5 but not IFN-γ (Street et al., 1990), and IFN-γ and IL-10 but not IL-4, have also been reported (Assenmacher et al., 1994). Interestingly, there are reports describing unique “Th3” cells expressing TGF-β but not IL-4 (Chen et al., 1994). Whether these particular cells may represent a surrogate lineage of Th development, similar to other Th2-like phenotypes described (Teng et al., 1995; Hozumi et al., 1998), or the newly described “regulatory” T-cells (Tr-1; Groux et al., 1997) that express IL-10 and/or TGF-β, but not IL-4, remains to be further investigated. More recently, different “phenotypic” Tr cells have been described as being CD4⁺CD25⁺ or CD3⁺CD4^-CD8^- alpha/beta^-TCR⁺ T-cells (Field et al., 2001; Zhang ZX et al., 2001). Tr cells with different phenotypes have been shown to be implicated in autoimmune colitis, type-1 diabetes, poly-organ autoimmune diseases, and induction of self-tolerance (Chatenoud et al., 2001; Singh et al., 2001).

Multiple factors can influence the development of Th subsets, including Ag dose, MHC molecules, the strength of peptide signals, and different levels of co-stimulatory molecule expression. Among these, the most critical factor for the development of Th1 or Th2 phenotype is the cytokines available in the local microenvironment, both in vitro and in vivo (Seder and Paul, 1994). For instance, IL-4 is absolutely required for committing Th2 development, and IFN-γ is required for Th1 development. Further, IL-4 and IFN-γ can cross-regulate Th1 and Th2 differentiation while amplifying their own development and growth (Kelso, 1995), respectively. This cross-regulatory mutual antagonism serves to redirect the immune response toward a phenotype or effector function appropriate to resolve the microbial challenges. Thus, Th1/Th2 dichotomy and/or effectors provide a cytokine-based framework for understanding CD4⁺ Th heterogeneity during development toward more pathogenic (or destructive) vs. beneficial (healing) immune responses. However, the results of studies of the contributions of Th1/Th2 or other cytokine profiles to the pathogenesis of periodontal disease have been shown to be conflicting and controversial.

It has been postulated that specific Th1 vs. Th2 cytokine profiles may be associated with susceptibility to periodontal infection (Seymour et al., 1993; Bartova et al., 2000; Ebersole et al., 2000; Seymour and Gemmell, 2001). However, many recent studies showed mixed results with respect to the cytokines expressed by HuPBL and/or gingival mononuclear cells of patients with chronic adult or early-onset periodontitis (Fujihashi et al., 1991; Manhart et al., 1994; Tokoro et al., 1997; Gemmell and Seymour, 1998; Sigusch et al., 1998). However, these studies lack clear information about the clinical stage of disease and associated microbial profiles for specificities, and those considerations may explain the lack of agreement in the results obtained. Further, it was shown that some A. actino-mycetemcomitans-specific Th1 or Th2 cells generated in vitro could home to rat periodontal tissues and mediate protection or destruction of alveolar bone, respectively (Yamashita et al., 1991; Eastcott et al., 1994; Ebersole and Taubman, 1994; Kawai et al., 2000). This suggests that the cytokines expressed by CD4⁺ T-cells may indeed play a significant role in host defense against periodontal infection. The molecular mechanism of Th1 vs. Th2 subsets in “regulating” or “modulating” periodontal disease progression is not fully understood, although it is likely that there is a very complex host-microbial interaction in periodontal infections. Several studies have shown that there are more abundant Th2 than Th1 cells in various diseased periodontal tissues (Yamazaki et al., 1995, 1997; Sigusch et al., 1998; Lappin et al., 2001). These findings are consistent with the early observation that B-cells (humoral immunity) are predominantly enriched in “established” periodontal lesions (Page and Schroeder, 1976), while T-cells (cell-mediated immunity) are more predominant at the “early” stage of the disease. Thus, directing or influencing Th1 vs. Th2 differentiation at the early stage of the infection could have some impact on the development and/or progression of the periodontal disease.

Using an experimental mouse model to study A. actinomycetemcomitans-specific periodontitis, investigators have recently showed that periodontal tissue destruction (alveolar bone loss) is strongly associated with both Th1- and Th2-related cytokine expressions in vivo (Teng, 2002), similar to the findings reported by Fujihashi et al.(1996) and Prabhu et al.(1996). It is noteworthy that there is abundant Th1-associated cytokine (TNF-α, IFN-γ) production during “active” alveolar bone destruction in which Th2-associated cytokines (in particular, IL-10 and TGF-β) are also highly expressed. This finding is consistent with those of another recent in situ study by Kinane and Lappin (2001) wherein both Th1- and Th2-associated cytokine signals were up-regulated in active periodontitis tissues. This is also the case in the study by Baker et al.(1999b), using genetic knock-out mouse strains infected by P. gingivalis, where both Th1- (i.e., IFN-γ) and Th2- (i.e., IL-6) associated cytokines appeared to be involved in alveolar bone destruction. One interpretation of these observations is that while Th1 effectors producing inflammatory cytokines can mediate tissue destruction, anti-inflammatory cytokines (i.e., IL-10, TGF-β) produced by Th2 cells are involved in tissue homeostasis during inflammatory destruction and subsequent repair or remodeling processes (Roberts et al., 1997; Sasaki et al., 2000; Kinane and Lappin, 2001). On the other hand, subtle differences in cytokine balance may result in progression to periodontal destruction over time; or, alternatively, different effectors with different cytokine expressions may signify different stages of periodontal disease development (Infante-Duarte and Kamradt, 1999). In fact, it is likely that, at the single-cell level, specific microbial Ag(s) recognized by the activated CD4⁺ Th cells can express both type-1 inflammatory cytokines and type-2 anti-inflammatory cytokines, depending on its overall “potency” in the local environment during the disease process. Indeed, our recent studies suggest that this could be the case where the same pathogenic Ag of A. actinomycetemcomitans associated with alveolar bone destruction in vivo (Gao and Teng, 2002; Teng, 2002) triggers both type-1 and type-2 cytokine expressions (Kelso, 1995; Teng Y-TA, personal communication). Th1 or Th2 effectors can individually or in combination regulate the host defense against periodontal infection. Conversely, modulation of Th1-/Th2-associated cytokine expressions may be part of the characteristics originating from specific periodontal pathogen(s) during infection (Choi et al., 2000; Gemmell et al., 2002). Further, these Th effectors with memory phenotype (Kawai et al., 1998) can efficiently transmigrate through the endothelium and home to local inflamed tissues to bring about tissue destruction (Kawai et al., 2000). Thus, blocking or deficiency of lymphocyte transmigration or their homing molecules, such as P- or E-selectins (Niederman et al., 2001), chemokines (Butcher and Picker, 1996), and chemokine receptors (Sallusto et al., 1998), may affect their homing ability and the subsequent development of pathogen-specific Th cells critically involved in cell-mediated and anti-bacterial immune responses. However, the molecular mechanisms of defining the “pathogenic potency” of specific Th1 vs. Th2 effectors for periodontal disease progression in vivo remain to be determined. Genetic studies with different Th1 and Th2 knockout mouse strains may provide more definitive information.

Cytokine IL-12 produced by monocytes, macrophages, neutrophils, and DCs during early innate immune responses promotes Th1 development, growth, and survival, and ablation of original IL-12 stimulation favors Th2 differentiation (Hsieh et al., 1993; Kalinski et al., 1999). Recent studies on human and mouse DCs suggest that different LPS moieties prepared from P. gingivalis and E. coli can polarize the immune response to Th2 or Th1 differentiation pathway through different IL-10/IL-12 ratios, respectively (Jotwani et al., 2001; Pulendran et al., 2001). Different microbial compounds (i.e., LPS from anaerobes) representing various pathogenic species can induce maturation of DCs through toll-like receptors (TLRs) (i.e., for TLR-4), evolutionarily conserved from Drosophila Toll molecules (Akira et al., 2001). Further investigation is required to determine whether signaling via specific TLRs would preferentially induce Th1 development (Jotwani et al., 2001; Pulendran et al., 2001; Schnare et al., 2001). Further, this IL-12-mediated Th1 development is often achieved via an IFN-γ-dependent and, to a lesser degree, by IFN-γ-independent mechanisms (Brewington et al., 2001). Therefore, DCs can be tuned to express certain “biasing” cytokines (i.e., IL-12, IL-10) to promote or to minimize the development of cytokine-mediated immunopathology (Reis e Sousa et al., 1999; Jotwani et al., 2001). Further, IL-12 released by macrophages stimulates NK cells and T-cells to produce IFN-γ, which in turn augments IL-12 secretion, thereby generating a positive feedback loop (Smeltz et al., 2002). Another cytokine, IL-18, also produced by monocytes and many other cell types, plays a role in CD4⁺ Th1 cell differentiation and promotion of T-cell proliferation and IFN-γ expression (Kashiwamura et al., 2002). To date, the contributions of other Th1-related cytokines—including IL-15, IL-18, and IL-23—and Th2-related cytokines—including IL-4, IL-5, IL-10, IL-13, TGF-β, and PGE₂—in periodontal disease progression are still unclear, and their roles need to be studied further (see Baker, 2000; Seymour and Gemmell, 2001; Taubman and Kawai, 2001).

(C) Th cells and bone diseases: OPG/RANK-L in periodontitis

The TNF family molecule RANK-L (receptor activator of NF-κB ligand; or OPG-L, TRANCE, ODF) and its receptor RANK have been shown to be the key regulators of bone remodeling and are directly involved in the differentiation, activation, and survival of osteoclasts (OC) and OC precursors (Wong et al., 1997; Lacey et al., 1998). RANK-L and its receptor RANK signaling pathway are essential for T-cell/dendritic-cell interactions, dendritic cell survival, lymph node formation, and organogenesis (Anderson et al., 1997; Kong et al., 1999a). RANK-L is expressed by bone marrow stromal cells, osteoblasts, and chondrocytes on cell surfaces as a membrane-bound trimer, and its release is proteolytically activated by the metalloprotease-disintegrin TNF-α convertase (Lum et al., 1999). It has been shown, at least in vitro, that RANK-L can activate mature OC and OC precursors for osteoclastogenesis in the presence of colony-stimulating factor-1 (CSF-1); however, CSF-1 expression is not essential for OC differentiation (Lacey et al., 1998; Yasuda et al., 1998). RANK is expressed by hematopoietic OC progenitors, mature OC, chondrocytes, monocyte/macrophages, and dendritic cells (Anderson et al., 1997; Hsu et al., 1999). Genetic mutations of RANK-L and RANK demonstrate the same defective phenotypes in OC development and severe osteopetrosis, suggesting that both molecules are essential for osteoclastogenesis (Kong et al., 1999a; Li et al., 2000). OPG (osteoprotegerin), a TNF-receptor family molecule, also known as osteoclastogenesis inhibitory factor, functions as a soluble decoy receptor to RANK-L and competes for RANK-L-RANK binding (Simonet et al., 1997). Thus, OPG is a natural inhibitor of OC maturation and activation in vivo and in vitro (Lacey et al., 1998; Kong et al., 1999a). OPG transgenic mice are osteopetrotic, with normal tooth development but with defective endosteal OC activity, and OPG-deficient mice are severely osteopetrotic, with fragile bone integrity (Simonet et al., 1997). RANK-L and OPG are regulated by several hormones (i.e., vitamin D, estrogens, glucocorticoids) and cytokines (i.e., TNF-α, IL-1, IL-4, IL-6, IL-11, IL-17; Yasuda et al., 1998; Theill et al., 2002). Recent studies have shown that OPG, RANK-L, and RANK interaction forms an intricate regulatory triad in osteoporosis and osteolytic diseases such as arthritis, periodontal disease, tumor-associated bone metastasis (Honoré et al., 2000; Teng et al., 2000; Hofbauer and Heufelder, 2001; Theill et al., 2002), and even mammary gland lactation (Fata et al., 2000). The ratio of RANK-L/OPG is critically involved in regulating and directing osteoclastogenic and/or osteoblastogenic development (Theill et al., 2002).

In an experimental animal model that mimics LJP, oral challenge of the HuPBL-engrafted NOD/SCID mice with A. actinomycetemcomitans, a well-known Gram-negative anaerobic micro-organism causing LJP, led to functional activation of the human CD4⁺ T-cells in the periodontium and triggered local alveolar bone destruction. Further, A. actinomycetemcomitans stimulation of CD4⁺ T-cells triggers the production of RANK-L, leading to osteoclast activation and osteoclastogenesis (see Fig. 1). In vivo inhibition of RANK-L function with its decoy receptor OPG significantly reduces both alveolar bone destruction after microbial challenge and the numbers of osteoclasts at the sites of local periodontal inflammation (Teng et al., 2000). This suggests that micro-organism-triggered induction of RANK-L expression on CD4⁺ T-cells and RANK-L-mediated osteoclast activation and bone loss represent the molecular explanation for the alveolar bone destruction observed in periodontal infections (Taubman and Kawai, 2001; Teng et al., 2002). Since the inhibition of alveolar bone loss is not complete in this animal model, the role of other molecules or auxiliary factors would need to be further investigated. (A schematic overview of the OPG/RANK-L:RANK interaction proposed in periodontal infection is shown in Fig. 2.) Moreover, it is known that the expression of pro-inflammatory cytokines (i.e., IL-1, TNF-α) can also regulate the balance of RANK-L/OPG in the bone microenvironment and/or mesenchymal tissues adjacent to bone (Hofbauer et al., 1999, 2000), thus contributing to bone destruction independent of cell-mediated immunity. Additionally, the potential contributions of local mucosal cells, dental cells, and/or phagocytes (i.e., monocyte/macrophages; Kawashima and Stashenko, 1999; Rani et al., 2000) in the early stages of infection/inflammation have yet to be defined in human periodontitis.

As previously mentioned, the pro-inflammatory cytokines IL-1 and TNF-α have been shown to be very important players in periodontal bone loss (De Nardin, 2001; Laine et al., 2001; Engebretson et al., 2002), and there is evidence that OC can be activated in a RANK-L-independent pathway via pro-inflammatory cytokines such as TNF-α and IL-1 (Azuma et al., 2000; Kobayashi et al., 2000; Zhang YH et al., 2001). Injection of soluble receptors to IL-1 and TNF-α markedly reduces alveolar bone loss in the non-human primate model of experimental periodontitis (Assuma et al., 1998). Further, RANK-L-deficient mice do not respond well to TNF-α- and IL-1-mediated bone resorption (Kobayashi et al., 2000), and OPG can inhibit bone destruction in the rat arthritis and mouse periodontitis models (Kong et al., 1999b; Teng et al., 2000), suggesting that OPG/RANK-L signaling is significantly and more critically involved in OC differentiation and activation. However, as long as OC is present and subjected to activation signals, both TNF-α and IL-1 can induce bone resorption and RANK-L production (Azuma et al., 2000). These findings suggest that two collateral independent pathways might co-exist. Interestingly, in the above experimental periodontitis model, RANK-L-mediated alveolar bone loss is associated with the presence of both Th1- and Th2-associated cytokines (Teng, 2002), suggesting a clear microbial-specific and host-factor-specific cell-mediated immunity for periodontal destruction in human periodontal disease such as LJP. The clinical significance of OPG/RANK-L in periodontal diagnosis awaits further study, and inhibition of RANK-L via OPG may have therapeutic value in preventing alveolar bone loss in human periodontal disease.

(V) Conclusion

Host immune-pathogen interaction is a complex process, which is intimately regulated by the acquired immunity to deal with immediate vs. long-lasting assaults imposed by ever-growing and -invading micro-organisms or pathogens in periodontal pockets. Recent studies suggest that there is much less tissue (i.e., alveolar bone) destruction when the host’s acquired immunity is absent or deficient during periodontal infection, compared with that of the immunocompetent host. It is apparent, based on the documented evidence, that B-cell-mediated humoral immunity is primarily protective in nature and is relatively neutral in mediating protection against periodontal tissue destruction as the microbial assault and tissue inflammation continue. Further, it requires the participation of cell-mediated immunity—in particular, CD4⁺ T-cells and/or other granulocytic lineages and/or innate immune cells—to perform all of its functions properly. CD8⁺ CTL do not contribute directly to periodontal destruction in situ, but may be important in killing or removing bacteria-infected host cells. Intriguingly, pathogen-specific CD4⁺ Th cells are both protective, by indu-cing B-cell-mediated humoral immune and CTL responses, and destructive, by producing RANK-L (and other inflammatory and/or pro-inflammatory cytokines), which modulates OC and OC precursor activity for osteoclastogenesis. In situ, the latter is associated with a mixed Th1- and Th2-associated cytokine profile of a relatively diverse immune repertoire.

To date, there is compelling evidence to suggest that the immune repertoire, in particular, CD4⁺ Th cells reactive and specific to periodontal pathogens at local sites, has an immense influence on the subsequent development of the host’s anti-bacterial immunity and breakdown of tissue integrity such as alveolar bone. A new paradigm of periodontal pathogenesis is now emerging toward the “acquired immunity-osteology” interactions (see Fig. 2) to enable investigators to explore and understand its mechanistic and molecular links in periodontal disease pathogenesis. It remains to be determined, at the genetic level, if type-1 cytokines or Th1 cells are essential for the destructive process, and if type-2 cytokines or Th2 cells contribute to the homeostatic and anti-inflammatory protective responses or vice versa. Further study of the genetic factors, initiation, triggering events, effector functions, and memory effects of T-cells at the single-cell and molecular levels will help us to understand these complex regulatory cascades involved in disease pathogenesis and will facilitate the generation of useful diagnostics and even therapeutic applications.

Figure 1.

Expression of RANK-L on A. actinomycetemcomitans-reactive periodontal CD4⁺ T-cells detected by flow cytometry and PCR analyses. The left panel (FACS analysis of RANK-L expression on cell surface by immunostaining to OPG-FITC conjugates): (a) un-stimulated HuPBL-derived CD4⁺ T-cells as background control; (b) isotypic Ig control for background staining; (c) HuPBL-derived CD4⁺ T-cells activated by anti-h-αβTCR plus anti-CD28 Mabs as positive control for RANK-L expression; and (d) periodontal CD4⁺ T-cells stained with OPG-FITC for cell-surface expression of RANK-L. The right panel (RT-PCR by the use of specific primers for RANK-L): (1) A. actinomycetemcomitans (Aa)-reactive periodontal CD4⁺ T-cells from HuPBL-NOD/SCID mice which received HuPBL samples from one LJP subject (called LJP 2); (2) Aa-reactive periodontal CD4⁺ T-cells from HuPBL-NOD/SCID mice which received HuPBL samples from another LJP subject (called LJP 1); (3) naïve unstimulated HuPBL-derived CD4⁺ T-cells; and (4) HuPBL-derived CD4⁺ T-cells activated by anti-h-αβTCR plus anti-CD28 mAbs as positive control for RANK-L expression. The experimental procedures have been described previously (Teng et al., 2000; Teng, 2002).

Figure 2.

Interaction of CD4⁺ T-cells and alveolar bone remodeling in the periodontium. After being challenged by periodontal pathogens through antigen processing and presentation (via APC-dendritic cells, macrophages, activated B-cells, etc.), these micro-organisms-specific periodontal CD4⁺ T-cells become activated and produce RANK-L molecules. RANK-L can activate osteoclasts (OC) directly to induce bone resorption and also induce the differentiation, survival, and activation of OC precursors. OPG, the natural decoy receptor of RANK-L produced by the stromal cells (i.e., osteoblasts, chondrocytes), can compete for the binding to RANK-L molecules, thereby counteracting the effects of RANK-L/RANK signaling on OC and OC precursors.