Periodontal and periapical diseases are the most common forms of destructive infectious/inflammatory diseases of the tooth-supporting tissues, and represent the most prevalent form of bone pathology. Periodontitis is initiated by bacteria harbored in the tooth-attached biofilm invading the surrounding epithelial and connective periodontal tissues, triggering a host immune/inflammatory response and the subsequent lesion development. Similarly, lesions of endodontic origin initiate as an immune/inflammatory response to a bacterial insult of the dental pulp, which can lead to necrosis, allowing the spread of the infection front to the periapical region and the consequent lesion formation. Although both conditions share a quite common etiology, it is noteworthy that based on bacterial infection alone, it is not possible to explain these complex pathological processes [1, 2, 3••]. Indeed, while any bacteria can essentially trigger a host response, the current paradigm of periodontal and periapical disease etiology states that specific bacteria (gram negative anaerobic rods) initiate and chronically sustain the host immune/inflammatory response, the nature and extent of which are ultimately responsible for the degree of irreversible tissue destruction and disease severity [4].

In this scenario, the interplay between the challenging microorganisms and resident and inflammatory host cells is thought to determine the disease outcome. Theoretically, if the host response is efficient in keeping the microorganisms spatially confined and in limited number, the tissue homeostasis will then be preserved. Conversely, if the host response is incapable of counteracting the bacterial challenge (where the presence of specific pathogens presenting virulence factors that allow tissue invasion or interfere with host’s defense mechanisms is supposed to be a critical event; as well the rupture of physical barriers that limit bacterial infiltration – such as enamel/dentin destruction), the result will be conversion to chronic inflammation, with destruction of the surrounding tissues as a consequence [5, 6].

The initial host cells’ innate response involves the recognition of microbial components (i.e., LPS, bacterial DNA, diacyl lipopeptides, peptidoglycan, etc.) as “danger signals” by pattern recognition receptors, such as the toll-like receptors (TLRs) [7]. TLRs are expressed by both resident cells and leukocytes in the periodontal/periapical environment, and upon their activation, an intracellular signaling cascade is stimulated, leading to activation of the transcription factors that mediate the cellular response and the subsequent production of inflammatory mediators, such as cytokines [8].

Cytokines exert their functions through binding to specific cellular receptors, therefore both cytokine and receptor expression can determine the nature of an individual cell response, and in a broader context, can impact on the intensity and duration of the inflammatory response, and ultimately the clinical outcome of periodontal/periapical disease. Over the last few decades we have accumulated a broad but somewhat superficial knowledge on how individual cytokines contribute to the development of these diseases. Recently, the application of new technology and analytical tools have revealed the multitude and complexity of cytokines, especially regarding the selective synthesis and secretion of effector cytokines by specific CD4+ T-lymphocyte subsets, and their participation in complex regulatory networks.

Numerous studies have intended to characterize the host response to infection in periodontal/periapical diseases; providing valuable information on the interactions between the bacterial biofilm and the responding host cells [911]. With the progress of our knowledge from purely descriptive observations towards more mechanistic findings, we are beginning to get a reasonable glimpse of the whole picture. In this context, the present review intends to present a chronological perspective on the development of the current hypothesis on the etiology of periodontal and periapical diseases, focusing on the specific contribution of cytokines and cytokine networks, and on how our increasing knowledge and insight into the immune processes is producing a more complete account of the extremely complex phenomena that characterizes these destructive diseases (Table 1, Figure 1).

Table 1 Cytokine networks of pro-inflammatory cells and T lymphocyte helper subsets
Fig. 1
figure 1

Defining features of pro-inflammatory and T lymphocyte helper cell subsets in periodontal and periapical lesions. Pro-inflammatory and T lymphocyte helper cell subsets present characteristic features (such as master switches, signature cytokines and classic functions) that define such cell subset in phenotypic and functional terms. When these cells are considered in the context of periodontal and periapical lesions, they have putative roles in the tissue destructive/protective responses, as well in the control of infection

Pro-inflammatory and Anti-inflammatory Cytokines in the Host Innate Immune Response

The initial investigations of immune response to bacterial infection in periodontal/periapical diseases demonstrated that both resident and infiltrating cells were able to recognize oral bacteria (and its antigens), and respond to them in a pro-inflammatory manner. In the case of periodontitis, the gingival epithelium is the first line of defense to infection and is routinely challenged by oral bacteria, constitutively secreting cytokines that attract a moderated but constant influx of immune cells (predominantly neutrophils), which arrest the bacterial growth and prevent infiltration [4, 12]. Epithelial cells respond to the periodontal infection by up-regulating pro-inflammatory cytokine/chemokine expression, thereby providing the stimuli for the leukocyte recruitment that will lead to the latter disease stages [13].

After the initial response of epithelial cells to the microbes, the permeability of the junctional epithelium is increased by the inflammatory mediators, and allows the contact of bacteria (and/or their products) with gingival connective tissue, where most of the host/microbe interaction events occur. Indeed, gingival fibroblasts (GF), the predominant cell type in gingival connective tissue, possess the capacity to respond to infection with the secretion of pro-inflammatory mediators that trigger, and to some extent sustain, the initial steps of the immune response [14].

In the periapical lesion, the spread of the infection front across the root apex/periodontium results in the direct contact of the infecting microbes with the periodontal ligament fibroblasts, that although not routinely exposed to bacterial challenge, also have the capability to respond to infection with cytokine production, participating in the initial recruitment phase of the host response similar to GFs [3••]. It is important to note that leukocytes can also play a role in the initial steps of host response to oral infections, since neutrophils and resident macrophages can be found in clinically healthy connective periodontal, pulpal and periapical tissues.

Within the pro-inflammatory mediators thought to coordinate the inflammatory cell migration in periodontal/periapical tissues, TNF-α, IL-1, and IL-8 have been extensively investigated. IL-1 and TNF-α share several pro-inflammatory properties, paramount among them the capacity to activate endothelial cells to provide the signals required for leukocyte diapedesis [3••]. Indeed, these pro-inflammatory cytokines induce the production of chemokines. These selectively recruit leukocytes from the peripheral circulation to the site of infection via specific ligand-receptor interaction that ultimately triggers integrin-dependent adhesion; cytoskeletal re-arrangement to facilitate extravasation and migration, as well as the binding and detachment of cells from their substrate [5]. In the cell migration framework, IL-8 (CXCL8) is the prototypical and firstly identified member of the chemokine family [15]; specifically acting in the chemoattraction of neutrophils, which form a sub-epithelial barrier that exerts a potent microbicidal action by means of its secretory functions (reactive oxygen species and bactericidal proteins), and acts as a unified phagocytic apparatus [16]. IL-8 production can be directly induced by TLR stimulation, or indirectly via TNF-α and/or IL-1 [17]. In periodontitis patients, IL-8 level is significantly increased compared with gingivitis-affected and healthy subjects [18], and seems to correlate with the periodontal status of the patients before and after treatment [19, 20]. While the early neutrophil migration is under the control of IL-8, the developing chronic inflammatory response also involves the migration of macrophages, where CCL2 (MCP-1) appears as a master regulator of monocyte mobilization [21, 22]. The level of CCL2 is increased in the saliva and gingival crevicular fluid (GCF) of periodontitis patients, and correlates with periodontitis severity [23]. It has also been reported that CCL2 levels are increased in the GCF and serum of periodontitis-affected smokers compared with non-smoking patients, suggesting that this chemokine could be involved in the worsening of the periodontal condition characteristic of the former [24]. Similar to periodontitis, IL-8 and CCL2 can be described as coordinating the respective migration of neutrophils and macrophages during the development of periapical lesions [25, 26].

Once in the periodontal/periapical tissues, neutrophils and macrophages seem to act along epithelial and connective tissue cells towards the maintenance and amplification of the inflammatory/immune response. Interestingly, both neutrophils and macrophages can be a significant source of TNF-α and IL-1, generating a positive loop towards response amplification [27].

Taken together, the pro-inflammatory cytokines can account for a significant extent of the tissue destruction associated with periodontal/periapical diseases, since they induce matrix metalloproteinases (MMPs) and receptor activators of nuclear factor kappa-B ligand (RANKL), the main mediators of soft and hard tissue destruction, respectively. Indeed, both soft connective tissue and bone integrity depend on the balance between resorption/deposition. The overexpression of pro-inflammatory cytokines in periodontal and periapical chronic inflammation interferes with the physiologic equilibrium, leading to pathologic tissue loss [2].

Even though pro-inflammatory cytokines have been classically related to destructive events, they also exert pivotal roles in the control of infection. Surprisingly, only recently has the “protective” nature of cytokines against infection been systematically investigated. TNF-α has a central role in the control of periodontal/periapical infections, and its deficiency generates a severe pathogen clearance impairment, increasing the systemic acute phase response. Indeed, TNF-α is a prototypical phagocyte activator, mediating both myeloperoxidase (MPO) and inducible nitric oxide synthase (iNOS) expression, respectively the major antimicrobial effectors of neutrophils and macrophages [28]. Accordingly, the lack of iNOS results in an increased periapical abscess formation following endodontic infection [29].

While the pro-inflammatory response reveals a dual nature (destructive while protective against infection), it is also important to remember that there are other cytokines involved in the overall cytokine network to prevent/attenuate the side effects of an exacerbated response. In this context, it is worthwhile to highlight the functions of IL-10, that antagonizes pro-inflammatory cytokine secretion and signaling, down-regulates MMP activity (directly and indirectly via tissue inhibitors of MMPs), and suppresses RANKL-mediated osteoclastogenesis (via osteoprotegerin) [30, 31]. These properties point to IL-10 as a “protective” mediator in periodontal/periapical lesions, although we must bear in mind that the protective/destructive dichotomy is a partial and oversimplified interpretation (albeit instructive) of a more complex phenomenon, given the dual role of host response previously discussed [2].

After the initial innate response characterized by the balance between pro-inflammatory vs. anti-inflammatory cytokines, the selective migration of specific T lymphocyte subsets drives the transition to the adaptive immune response; where T helper lymphocytes (Th) play a critical role in orchestrating the response. As described in the sequence, the discovery of Th subpopulations dramatically changed the understanding of cytokine networks beyond the pro-inflammatory vs. anti-inflammatory framework.

The Th1/Th2 Balance Paradigm

The recognition in the late 1980s of two distinct (Th1 and Th2) clonal linages of Th lymphocytes, defined by unique cytokine secretory patterns, was a cornerstone in the understanding of immunological responses and the development of the current paradigms of immune/inflammatory disease etiology [32, 33]. Th1 cells mediate immune responses against intracellular pathogens and are particularly important in infection control. A dendritic cell-derived cytokine, IL-12, is the principal cytokine responsible for Th1 lineage commitment and differentiation, which involves the activation of the key transcription factor T-bet [34]. The prototypic cytokine product of Th1 lymphocytes is IFN-γ, involved in the activation of macrophage microbicidal functions [35, 36]. In parallel with Th1 cell discovery, their opposing mates Th2 were also unraveled. Th2 cells mediate the host defense against extracellular parasites and secrete a wide array of cytokines, including IL-4, IL-5, IL-9, IL-13, and IL-25. IL-4 is the main positive feedback signal for Th2 lineage commitment via induction of the GATA-3 transcription factor [37]. The selective polarization to Th1 or Th2 lineage is dependent upon the cytokine milieu in which the antigens are presented to the naïve T CD4+ cells, in that the polarization is a mutually exclusive event that also generates subsequent positive feedback to the respective lineage polarization [38].

The Th1/Th2 dichotomy paradigm was quickly replaced by a more realistic Th1/Th2 balance hypothesis, where the distinctive features of immune response could be explained by the predominance of one clonal Th subset over the other, comprising a theoretical framework to investigate T cell-mediated immunological responses and related diseases, including periodontal/periapical diseases. The Th1 predominance hypothesis in periodontal/periapical diseases was supported by experimental evidence of the overexpression of IFN-γ in gingival mononuclear cells isolated from periodontitis patients [39], and by the quantitative dominance of IFN-γ over IL-4 in the lesional tissue/fluid of periodontal/periapical disease patients [21, 40]. The inner rationale for the Th1 predominance hypothesis of periodontal destruction was that the exacerbated activation of neutrophils and macrophages by Th1 cytokines interfered with, and amplified, the pro-inflammatory activity of phagocytes and even of resident fibroblasts, and ultimately increased local levels of MMPs and RANKL, leading to augmented soft tissue and bone destruction [4143]. While potentially destructive, IFN-γ-mediated Th1 responses are essential to avoid pathogen dissemination. Indeed, IFN-KO mice challenged by A. actinomycetemcomitans oral infection presented a less severe periodontitis phenotype, but on the other hand suffered a widespread (and often lethal) systemic infection, associated with lower MPO and iNOS production [44].

Conversely, other studies supported the Th2 predominance hypothesis in periodontal/periapical diseases, based on the fact that Th lymphocytes isolated from inflamed gingival tissues predominantly produced IL-4 over IFN-γ, after non-antigen specific stimulation [45]. Also, the initial descriptions of established periodontal lesions as a “B-cell lesion” [46] were used to support the possible Th2 role in periodontitis progression in the view of the Th2/B-cell cooperative axis. In this framework, the Th2-stimulated local production of antibodies would be a protective factor, albeit not fully supported by experimental evidence. However, some authors consider that Th2 polarization may indicate an impaired adaptive immune response, in which Th2 cytokines inhibit Th1 polarization (and consequently the phagocytic immune mechanisms), due to inherent host characteristics, or by evasion strategies mediated by virulence factors of periodontopathic bacteria [4749]. When the pro-inflammatory versus anti-inflammatory and Th1 vs. Th2 frameworks are considered in a global context, evidence points to a cooperative role of Th1 and pro-inflammatory cytokines in the amplification of tissue destructive events in gingival/periapical tissues; while Th2 mediators seem to cooperate with anti-inflammatory cytokines, given the similarity between the properties of IL-10 and IL-4 [50, 51].

Interestingly, conflicting data were readily available in the literature to support either the Th1 or Th2 predominance hypotheses. The heterogeneity of experimental design, disease definitions, time of sample collection, stimulation protocols, and analytical methods used in the different studies could explain some of these differences in outcome [51]. Additionally, the discovery and recognition of new T helper cell lineages overcame the traditional, and quite limited, Th1/Th2 balance hypothesis that restrained the full understanding of the cytokine networks operating in periodontal/periapical lesion development [50]. It will be discussed below how the expansion of the Th family has led to a new level of complexity of cytokine networks, where multiple sets of T cells can reciprocally inhibit or stimulate the functions of another Th set and by doing so, orchestrate the outcome of adaptive immune responses.

Th17 and Treg

When the relatively simple Th1/Th2 dichotomy and balance theories proved insufficient to fully explain the pathogenesis of a series of immune response-related conditions, the identification of additional Th subsets brought a new perspective, and additional complexity, to the evolving cytokine networks. In 2003, it was demonstrated that IL-23 could drive a population of T cells to express the transcription factor RORγT and to secrete IL-17 [52, 53]. These cells were later named Th17 and are able to secrete IL-17, IL-21 and IL-22, with IL-17 being the cytokine which characterizes the effector functions. IL-17 was found to be largely expressed in inflamed periodontal/periapical lesions [43, 54, 55]. Our group reported overexpression of IL-17, IL-23, IL-6, TGF-β and IL-1β in the gingival tissues of periodontitis patients compared to healthy controls, as well as augmented levels of both IL-17 and RANKL in alveolar bone from periodontitis-affected subjects [56]. These results provided evidence that the inflammatory context characteristic of periodontal disease (i.e., IL-1β, IL-6, and TGF-β) could drive the polarization of the Th17 subset in human periodontitis. Th17 cells are supposed to contribute to the development of periodontal/periapical lesions directly up-regulating MMPs, and indirectly by an inflammation amplification loop stimulating the secretion of the classic pro-inflammatory cytokines by other cell types [57, 58].

Interestingly, shortly after the discovery of Th17 cells, studies characterized their interplay with the previously known Th1 and Th2 subsets, since both Th1 and Th2 cytokines were recognized to antagonize IL-17 secretion and Th17 polarization [59]. Most notably, it was demonstrated that IFN-γ neutralization worsened the tissue damage in a delayed-type hypersensitivity model, providing the first evidence that the Th1 subset could arrest the tissue destruction mediated by IL-17, reversing the classical pro-inflammatory and pro-destructive stereotype assigned to Th1 lymphocytes [52, 60]. When a possible Th1/Th17 interplay is considered in the periodontal context, it was observed that progressive periodontal sites demonstrated overexpression of RANKL, IL-1β, IL-17 and IFN-γ compared with inactive sites, and positive correlations exist between RANKL/IL-17, RORγT/RANKL, and RORγT/IL-17. These studies indicate the putative involvement of the Th17/IL-17/RANKL axis in periodontitis progression [61, 62]. Interestingly, unpublished data from our group derived from molecular analysis of active periapical lesions demonstrate an inverse correlation between IFN-γ and IL-17 levels, suggesting that Th1 and Th17 pathways may operate independently (and not cooperatively) in the evolution of periapical lesions.

Switching to the immunosuppressive front, a new Th lineage with broad immunosuppressive properties was characterized, including the inhibition of Th1, Th2 and Th17 polarization [63]. This subset was named T regulatory (Tregs) and it was defined by the expression of the phenotypic markers CTLA-4, IL-10, TGF-β, GITR, CD103, and CD45RO; and Foxp3 was shown to be the master transcriptional regulator for Tregs differentiation [38]. Tregs were associated with high expression of the regulatory cytokines IL-10 and TGF-β in inflamed periodontal tissues [31], and their presence was associated with decreased disease severity [64]. Conversely, some studies have reported an overexpression of Foxp3, T-bet, RANKL, IL-17, IL-1β, and IFN-γ in human active periodontal lesions, compared with inactive ones [61, 62]. Other studies have shown that IL-10-expressing cells outnumber the pro-inflammatory cytokine-expressing cells in human periapical lesions [65], and that the proportion of Foxp3+ cells are more prevalent in periapical granulomas and less prevalent in residual radicular cysts [66]. While data from human samples is conflicting regarding the role of Tregs in periodontal/periapical lesion pathogenesis, experimental data links the presence of Tregs with the attenuation of disease progression rate, such findings being confirmed in a cause-effect manner by disabling Tregs with anti-GITR antibodies [64]. Accordingly, in a recent report we demonstrated that the selective recruitment of Tregs with a controlled release system of the chemokine CCL22 (a known Treg chemoattractant) is effective in reducing the clinical measures of inflammation and bone loss in canine/murine models of periodontitis. Most interestingly, while Tregs have been associated with the impairment of protective immunity in certain diseases, the Tregs local enrichment treatment did not increase the bacterial load in gingival tissues, and did not raise the levels of serum inflammation markers [67••]. This new approach to the periodontal therapy by immune modulation is an excellent example of how immune research could translate to the clinical setting in the near future.

Th9, Th22 and T Follicular Helper

Even more recently, Th9, Th22 cells, and T follicular helper subsets (Tfh) were discovered and shown to interact with the previously known Th subpopulations in the modulation of inflammatory/immune responses [68].

Th9 cells characteristically produce IL-9, initially designated as a Th2 cytokine, which exerts pro-inflammatory or anti-inflammatory activities by modulating Tregs and/or Th17 cell development and function [69]. Th9 cells produce and secrete TNF-α and are involved in pro-inflammatory amplifying loops in many skin diseases [70]. Th22 cells produce IL-22, which can exert pro-inflammatory effects by a synergistic action with classic pro-inflammatory mediators such as TNF-α and IL-17 [71, 72]. IL-22 can also directly up-regulate RANKL expression and therefore induce osteoclastogenesis [73]. We showed that both Th9 and Th22 cytokines are expressed in human and experimental periapical lesions, where they supposedly contribute to the lesion stability [74••]. However, no mention of the role of these cytokines in periodontal disease can be found in the literature.

Finally, it is important to mention Tfh, a CD4+ T cell subset found in the B-cell follicles of secondary (and feasibly tertiary) lymphoid organs [75], described as a major contributor to B cell-mediated antibody responses and an important source of IL-21 [76]. IL-21 is a pleiotropic cytokine highly expressed in gingival biopsies of chronic periodontitis [77, 78], and it has been implicated in osteoclastogenesis and bone resorption [79], as well as in the development of Th17 cells [80]. It has also been implicated in the inhibition of Tregs polarization in low-grade chronic inflammatory disorders, such as obesity [81]. Nevertheless, no mentions regarding the possible role of Tfh in either periapical lesions or periodontal disease can be found in the literature so far. Indeed, due to the relative novelty surrounding Th9, Th22, and Tfh subsets, along with the lack of specific studies focused on their patterns of expression (and possible role) in periodontal/periapical lesions pathogenesis, it is not possible to make any definitive conclusion regarding their individual and collective role in the overall cytokine network.

Cytokine Networks: from Classic T-helper Prototypes to the T-cell Plasticity Concept

Conventionally, the differentiation of naive CD4+ T cells into subsets has been considered as an irreversible event, defined by the expression of selective signature cytokines and a single master transcription factor [38]. However, with the discovery of new Th subsets our understanding of the fate of T-cell identities has changed. It has been recently established that the same cytokines can be expressed by more than one polarized Th subset, and that they can also change their phenotype, characterizing a phenomenon described as plasticity [82, 83••]. As an example, IL-10 was once thought to be a Th2 exclusive cytokine, but it is now clear that IL-10 can be produced by multiple Th subsets, such as Th1, Th2, Tregs, and Th17 cells [83••, 84]. Similarly, IL-9 is preferentially produced by the Th9 subset [85], but it can also can be produced by Th2, Th17, and Tregs [69, 86]. Another interesting plastic characteristic of Th cells is the recently discovered possibility of simultaneous expression of the Th17 and Tregs transcription factors (RORγT and Foxp3) in the same cell [87]. There are many other examples of Th plastic features and the complexity of these findings calls into question whether Th cell subsets are more appropriately viewed as a “work in progress”, rather than a terminally differentiated cell [83••]. Nevertheless, it is noteworthy that the majority of the experimental data regarding Th cell differentiation and plasticity is derived from in vitro studies employing extreme (and possibly artificial) polarizing stimuli to obtain homogenous populations [88], so extrapolations to the clinical setting must be made carefully.

With the perspective of the plastic capabilities of Th lymphocytes in mind, it is possible to hypothesize that the controversies that regularly arise, regarding the functions of all Th subsets in the pathogenesis of periodontal/periapical diseases could merely be the expression of the continuous metamorphosis of the Th subpopulations along disease stages, possibly as an adaptive reflex to the changing environmental conditions and immunological necessities, in the battlefront of periodontal infection. More powerful analytical tools will be required in order to comprehensibly reappraise our current data and understandings in light of this new viewpoint.

Concluding Remarks

Currently, great efforts are being made to integrate the overwhelming amount of new data regarding cytokine networks and their involvement in the overall regulation of inflammatory immune responses to periodontal/periapical infection. The discovery and characterization of Th subsets provided the conceptual framework to evaluate complicated data, helping in the development of explanatory models for the intricate immunological process underlying the pathogenesis of periodontal/periapical diseases. As ingeniously portrayed by John Conway in his famous cellular automaton “Game of Life”, complex systems could emerge from a limited set of simple rules, and the rules are often impossible to infer from the observation of the working system [89]; the focus of our efforts should therefore be to unveil the underlying rules that govern the biological systems, and not to get caught in the trap of using advanced technology simply to further describe these complex phenomena. For the last 25 years, the Th1/Th2 and the Th17/Tregs paradigms have provided researchers an invaluable intellectual framework to interpret experimental data. With the refinement of research techniques our apparently solid models are changing faster than ever before, and the scope of our understanding should expand accordingly. In trying to keep up with this growing knowledge we must apply the lessons learned and integrate the new concepts and ideas emerging from research data in more complex and holistic models, understanding our current notions and paradigms as flexible ones. The unveiling of Th plasticity opens a new opportunity for revisiting our ideas of cytokine networks in immunological processes; instead of rejecting the fresh notions and attempting to protect the established paradigms, we should embrace them and be willing to reinterpret all our accumulated data under this new light.