Production of immunoregulatory cytokines in clinically asymptomatic periapical lesions depends on the lesions size Zavisnost produkcije imunoregulatornih citokina u klini č ki asimptomatskim periapeksnim lezijama od veli č ine lezije

Bacground/Aim. Development of periapical lesions (PLs) involves a complex cross-talk between pathogenic microor-ganisms from the root canal and host immune mechanisms. In these processes proinflammatory cytokines are involved in stimulation of inflammation and osteodestructive mechanisms, whereas anti-inflammatory cytokines, with the immunoregulatory functions, have the opposite effects. How this balance is controlled is still the subject of numerous studies. The aim of this study was to examine whether the local production of interleukin (IL)-10, IL-27 and transforming and the compo-sition/phenotype of PL-ICs. Conclusion. This study sug-gests that IL-10, IL-27 and TGF- β may play different roles in suppression of unwanted immune/inflammatory reactions in asymptomatic PLs, depending on the extension of pathological process as judged by the size of lesions.

Spearman's correlation test. The values of p < 0.05 were considered to be significant. Results. The levels of IL-10 and TGF-β were significantly higher in the PL-ICs cultures of large-size lesions than in small ones (p < 0.01 and p < 0.05, respectively). In contrast, the levels of IL-27 were higher in the cultures of small-size lesions than in small ones (p < 0.05). Although the total number of PL-ICs and the proportion of mononuclear phagocytes were higher in the large-size PLs (p < 0.01 and p < 0.05, respectively), their main composition was not significantly different between the groups. The proportions of B cells/plasma cells (CD19/CD38 + cells), CD8 + T cells and CD14 + cells were significantly higher in the large-size PLs (p < 0.005; p < 0.05; p < 0.05, respectively). In contrast, the proportion of total T cells (CD3 + cells) was higher in the small-size lesions (p < 0.05). No statistically significant correlation was found between the levels of these cytokines and the composition/phenotype of PL-ICs. Conclusion. This study suggests that IL-10, IL-27 and TGF-β may play different roles in suppression of unwanted immune/inflammatory reactions in asymptomatic PLs, depending on the extension of pathological process as judged by the size of lesions.

Introduction
Periapical lesions (PLs), common chronical pathological processes of the oral cavity, are triggered by bacterial infection of pulpal and endodontic environment 1 . Their pathogenesis involves a complex host immune/inflammatory response to the bacteria and their products in order to eliminate the invading microorganisms. However, the same mechanisms may also result in the tissue injury, followed by the destruction of soft and mineralized tissues surrounding the root apex 2 . The breakdown of tissues is triggered by different host mediators, which independently or cooperatively stimulate proteolysis and bone resorption processes [1][2][3][4][5] . Simultaneously, the regulatory mechanisms are triggered with the aim to suppress the inflammation and tissue destruction and to restrict the PL development 3,4 .
PLs consist of granulation tissue, proliferating epithelium or cyst infiltrated by different inflammatory cells (ICs) 5 . The compostion of ICs as well as functional and phenotypic properties of both infiltrating and stromal cells depend on the activation status of PLs which is under control of a series of cytokines. Cytokines play a major role in the modulation of immune/inflammatory reactions within PLs, and are critical determinants of lesions outcome 2, 6 . In this context, it is believed that the T-helper 1 (Th1) immune response, mediated by interferon-γ (IFN-γ) is involved in the progression of PLs and bone destruction, whereas T-helper 2 (Th2) cytokines, such as interleukin 4 (IL-4), IL-5, IL-10 and IL-33, are described to be important for the humoral immune response and to limit or attenuate the tissue damage 2 . IL-17 may play a role in exacerbating inflammation 7 and osteolytic processes 8 within PLs. On the other hand, the Foxp3 + CD4 + CD25 + subset of T regulatory cells (Tregs) and Tr1 cells exert suppressive effects on inflammatory osteolysis, in which cytokines Transforming growth factor beta (TGF-β) and IL-10 seem to play a crucial role [9][10][11] . The functions of Th17 and Tregs cytokines are interconnected and the Th17/Tregs archetype was suggested to influence the PLs outcome 12, 13 through the balance between the production of the osteoclastogenic factor named a receptor activator of nuclear factor kappa-Β ligand (RANKL) and its antagonist osteoprotegerin (OPG) 6,14 . Our previous results showed that IL-27 may have both pro-inflammatory and immunoregulatory functions in PLs 15 .
Pro-inflammatory cytokines prevail in symptomatic PLs, whereas the production of immunoregulatory cytokines characterizes predominantly asymptomatic PLs 3, 6, 16 . Since restriction of inflammation within PLs may occur at different stages of their development, the aim of this study was to examine whether the local production of three immunoregulatory cytokines (IL-10, IL-27 and TGF-β) in human asymptomatic PLs depends on the lesion size and how the levels of these cytokines correlate with the cellular composition/phenotype of such PLs. Similar data have not been published yet.

Patients
The study was conducted on human PLs (n = 30) at the Military Medical Academy (MMA), Belgrade, Serbia after an approval from the Ethics Committee of MMA and obtaining an informed consent from the patients.The average age of the patients was 31 years (range: 19-59 years). The patients had no malignant or autoimmune diseases and did not use immunosuppressive /immunomodulatory drugs. In addition, the patients had not been treated with antibiotics one month before the PLs excision. The clinical part of the study was performed at the Department for Oral Surgery, Clinic for Stomatology, MMA, at the time of teeth extraction or apicoectomy. The immunological part of the study was performed at the Institute for Medical Research, MMA. PLs were radiographically diagnosed using the standard equipment for intraoral (Carestream CS 2200 rentgenaparature) and extra-oral (ortopantomography and Dental Cone Beam Computed Tomography -CBCT) radiography of the maxillofacial region. Intraoral radiographs were obtained using the OrtoDent digital diagnostic center, Belgrade, following the rules for patients positioning according to the manufacturer's instructions. The gained ortopantomography shots were calibrated by the Vatech program software for panoramic pictures (Ez dent-i), while the 3D CBCT analyses were done in the Ez 3D plus and Ez 3D-i Vatech program software. The size of radiolucent PLs on the radiographs and tomographs was analyzed by the above mentioned softwares and the smallest and largest diameters were measured. All PLs were clinically asymptomatic.
The lesions were divided according to their size into the small and large PLs. The small lesions (n = 12) were those whose greatest diameter was less than 3.5 mm. The lesions whose smallest diameter was higher than 4.5 mm were classified as the large lesions (n = 18). No further distinctions between specimens were made regarding sex, age, and etiology or tooth type. The examples of radiographs/tomographs of one small-and one large-size PLs are given on Figures 1 and 2. After extraction, PLs were immediately placed in a medium consisting of RPMI-1640 (Sigma, Munich, Germany) and antibiotics/antimycotics, and transported to the laboratory.

Preparation of inflammatory cells
The inflammatory cells from PLs (PL-ICs) were isolated using a procedure previously optimized by our research group 5,17 . Briefly, periapical tissue was placed in a Petri dish containing 1 mL RPMI-1640 medium and cut into 2-3 mm diameter pieces using a scalpel. The tissue was then digested for 15 min with 0.05% collagenase type IV (Sigma) and 0.02% DNAse (Sigma) in 10 mL RPMI-1640 medium in a cell incubator at 37ºC. After that, the tissue was pressed through a stainless-steel mesh using a syringe plunger, filtered and resuspended in RPMI-1640 medium containing 1 mm EDTA. The released cells were washed twice by centrifugation in the RPMI medium containing 0.5 Mm EDTA at room temperature (400 g for 10 min), and counted. The viability of cells, determined by Trypan Blue dye, was between 90% and 95%. After that, cytospins were prepared from each sample of PL-ICs using a cytocentrifuge (MPW-350, Poland) on the poly-l-lysine-coated glass slides.
The cytospins were stained with May-Grünwald-Giemsa or used for immunocytochemistry. The cellular composition was analyzed by using a light microscope (Olympus). The identification of cells was done by the morphological criteria. A minimum of 500 cells was analyzed on each cytospin. The results are given as percentages of cell subsets.

Immunocytochemistry
For immunostaining, anti-CD3, -CD4, -CD8, -CD14, -CD19, -CD38 and -HLA DR unconjugated monoclonal antibodies (mAbs) were obtained from Serotec, Oxford, UK. Rabbit anti-mouse unconjugated-and peroxidase conjugated-Ig, as well asan alkaline phosphatase anti-alkaline phosphatase (APAAP) complex were purchased from DAKO, Copenhagen, Denmark. Cytospins were fixed with 2% pararosaniline (Sigma) for 2 min at room temperature, washed with phosphate-buffered saline (PBS) for 10 min, blocked with rabbit serum for 20 min and washed with PBS. Cytospins were incubated with the primary antibodies for 60 min at room temperature followed by washing with PBS. When the staining was further processed for the immunoperoxidase method, the slides were blocked with 0.3%H 2 O 2 in PBS for 20 min and then incubated with appropriate dilution of antimouse peroxidase conjugated Ig. The immunoperoxidase reaction was developed with diaminobenzidine. When the immunostaining was performed with the immunoalkaline phosphatase method, the slides were incubated with anti-mouse unconjugated Ig and then with APAAP. The reaction was finished by using Fast Red, as a substrate. Both enzyme substrates were obtained from Sigma. The controls were the samples incubated with an irrelevant mAb, mouse anti-rat CD4 (OX-38), Serotec, non-reactive with human cells. To identify B cells and plasma cells, cytospins were stained with the combination of anti-CD19 mAband anti-CD38 mAb. T cells were identified based on the positivity with anti-CD 3 mAb and subsets of T cells based on the staining with anti-CD4 and anti-CD8 mAbs, respectively. CD14 was a marker of mononuclear phagocytes, whereas HLA-DR was a marker of antigen-presenting cells. Cytospins were analyzed by light microscopy. On each cytospin at least 500 cells were counted. The percentages of positive cells were determined on the basis of total counted cells.

Cell cultures
PL-ICs were cultivated in 96-wells, with the roundbottomed plates (ICN, Costa Mesa, CA) (1 × 10 5 cells/well, 200 µL) in the complete culture medium consisted of RPMI-1640 medium supplemented with 10% fetal calf serum (FCS) (Sigma) and the standard culture solutions of antibiotics 17 . Phorbol myristate acetate (PMA) (20 ng/mL) (Sigma) and Ca 2+ ionophore (A 23187, 1 M) (Sigma) were used for the cell stimulation 18 . After 24 h, the cell supernatants were collected, centrifuged and frozen at -70ºC until the levels of cytokines were determined.

Cytokine assays
The concentrations of IL-10, IL-27 and TGF-β in culture supernatants were detected by using the specific ELISA kits (R&D, Minneapolis, USA) following the instructions of the manufacturer. The levels of cytokines were determined on the basis of standard curve, constructed by known concentrations of these cytokines. The results are presented as pg/mL.

Statistical analysis
Statistical analysis was performed by using the Student t-test and Spearman's correlation test. The values of p < 0.05 were considered to be significant.

Results
The first aim of this study was to examine the composition of PL-ICs isolated from clinically asymptomatic lesions divided according to their size. As presented in Table 1, the total number of PL-ICs from the small-size lesions was lower compared to the number of PL-ICs isolated from the large size lesions (p < 0.01). However, their main composition was not significantly different, except that the large-size lesions contained a higher proportion of mononuclear phagocytic cells (p < 0.05).
The results related to the production of three immunoregulatory cytokines in the cultures of PL-ICs were presented in Figure 3. The levels of IL-10 and TGF-β were significantly higher in the PL-ICs cultures of larger-size lesions (p < 0.01 and p < 0.05, respectively). In contrast, the level of IL-27 was higher in the cultures of small-size lesions (p < 0.05).

Fig. 3 -The levels of interleukin (IL)-10, IL-27 and transforming growth factor (TGF)-β in culture supernatants of asymptomatic periapical lesions (PL)-inflammatory cells (PL-ICs). Values are given as mean ± standard deviation for n = 12 (PL-ICs from the small-size lesions) or n = 18 (PL-ICs from the large-size lesions). * = p < 0.05; ** = p < 0.01, compared to the values of smallsize (PLs).
In order to explain whether these differences are associated with the cell subset composition in PLs, phenotypical analysis of PL-ICs was performed on the cytospin preparations by using two immunocytochemistry methods. Table 2 and Figures 4 and 5 show that the proportion of B cells/plasma cells (CD19/CD38 + cells), the subset of T cells (CD8 + ) and CD14 + cells were higher in the large-size lesions (p < 0.005; p < 0.05; p < 0.05, respectively). In contrast, the proportion of total T cells (CD3 + cells) was higher in the small-size lesions (p < 0.05). No significant differences were observed in the proportion of CD4 + T cells and HLA-DR + cells.
When the levels of cytokines in each group of lesions were correlated with the cellular composition of these lesions, no statistically significant correlation was found (data not shown).

Discussion
This is the first study investigating the production of immunoregulatory cytokines (IL-10, IL-27 and TGF-β) in clinically asymptomatic PLs, which were analyzed according to the size of PLs. IL-10 and TGF-β were implicated in the suppressive mechanisms mediated by different PL cells, including CD4 + CD25 + Foxp3 + Tregs 19 . In contrast, IL-27 was identified as both pro-inflammatory and immunosuppressive cytokine in PLs 15 . We decided to study the clinically asymptomatic lesions based on a large number of data showing that immunosupressive mechanisms are more operative in this type of lesions 3,4,6,20 . In addition, the large-size lesions are usulally an advanced stage of the PL development, characterized as large granulomas or cysts. In contrast, the smallsize lesions usually represent an early stage of PL induction 6,20 . The division into these two groups of PLs was arbitrary and based on the size of our PL collection. To avoid possible errors in the PL measurement 21 we excluded from the study those PLs whose greatest diameter was higher than 3.5 mm and PLs whose smallest diameter was less than 4.5 mm.
Both types of PLs had similar cellular content, composed predominantly of mononuclear infiltrating cells, which is a typical hallmark of asymptomatic lesions, whereas the proportion of granulocytes which are characteristic for symptomatic lesions (acute phase of PL development or PL exacerbation) was lower 3,17,20 . Our results showed that ICs from the large-size PLs produced higher levels of IL-10 and TGFβ than ICs from the small-size PLs. This finding correlates with a higher proportion of mononuclear phagocytic cells, which are an important source of these immunosuppressive cytokines 22 . Except macrophages, IL-10 is produced by dendritic cells, different subsets of T and B cells and various innate immune cells 23 . TGF-β is additionaly secreted by different stromal cells 24 and both cytokines are produced by Tregs 25 . In our previous paper 19 , we showed positive correlations between the frequency of Tregs and the levels of IL-10 and TGF-β in culture supernatants of mononuclear cells (MNC) isolated from PLs. It is interesting that all Tregs, isolated from PLs, expressed IL-10, but only a half of them were TGF-β-positive 19 . Based on all these data, it can be supposed that the proportion of Tregs in the large-size PLs is higher than that in the small-size lesions, but this hypothesis, which has not been tested so far, needs to be proved in further experiments.
IL-10 and TGF-β have the potent anti-inflammatory properties that plays a central role in limiting the host immune response to pathogens, thereby preventing the tissue damage, including osteolysis in PLs 3, 7, 20 . Dysregulation of IL-10 is associated with the enhanced immunopathological processes in response to infection and increased risk for the development of autoimmune diseases 23 . It is interesting that the systemic administration of IL-10 for autoimmune therapy was shown to be paradoxically proinflammatory, whereas the localized IL-10 delivery proved to be therapeutic 26  mediates its immunosuppressive activity by the heterodimeric IL-10 receptors (IL-10R1, IL-10R2), which is expressed at varying degrees in many cell types, especially in monocytes and macrophages. The ligation of receptor is followed by the activation of Janus kinase/signal transducers and activators of transcription (JAK/STAT) signaling pathway, and subsequent transcription of various immunomodulatory genes, which, in turn, inhibit the production of pro-inflammatory mediators, stimulate the release of anti-inflammatory molecules such as interleukin-1 receptor antagonist (IL-1RA), soluble tumor necrosis factor-α (TNF-α) receptor, and interleukin IL-27 and down-regulate antigen presentation and phagocytosis 23,26 . Additionally, through the release or via physical interactions with T cells, IL-10 can directly or indirectly enhance the Treg function 26 , but it is also able to activate a certain type of immune cells, including the B cells, and to stimulate their proliferation. Such finding and association of IL-10 with the Th2 immune response are in agreement with a higher percentage of CD19 + /CD38 + B cells/plasma cells and with previous findings that the advanced stage of PL development is associated with the predominance of humoral immune response 3,20,23 .
TGF-β participates in several important pathophysiological processes and plays a bidirectional activity in the immune regulation. Its dual role in modulating the macrophage function is rapidly gaining recognition. TGF-β functions as a macrophage suppressing agent and as a monocyte activator 24 . It also exerts the inhibitory and stimulatory effects on the bone marrow cells, stimulates the chemotactic activity of osteoblasts and supports the differentiation of mesenchymal stem cells into osteoblasts and chondroblasts 24,27 . In our previous paper, we showed that the production of TGF-β by PL-ICs is upregulated by the mesenchymal stem cells isolated from PLs 28 . TGF-β also stimulates the production of extracellular matrix and collagen type I, the molecules which are of key importance for the promotion of tissue healing 29 . It is known that healing is a characteristic of advanced stage of the PL development 20 , a finding which is in correlation with our present results.
IL-27 is a relatively new cytokine member of the IL-6/IL-12 family, exerting both pro-inflammatory and anti-inflammatory properties. It consists of the Epstein-Barr virusinduced gene 3 (EBI3) and p28 subunits and acts through the IL-27 receptor complex formed by WSX-1 and gp130 subunits. DCs and activated macrophages are its main source 30 . In the PLs endothelial cells 15 as well as the mast cells 31 are also IL-27 + . It is well documented that IL-27 plays an important role in the initial immune response by stimulating the production of Th1 cytokines, which are important for the development of granulomatous diseases (tuberculosis, sarcoidosis and Crohn's disease) 32 . However, IL-27 has general inhibitory effects on the activity of Th1, Th2, Th17 and regulatory T cells 30 . The inhibitory effect of IL-27 on osteoclasts 33 is in accordance with this hypothesis, suggesting the role of this cytokine in the restriction of unwanted immune responses and bone destruction in PLs. In our previous experiment, we showed that the production of IL-27 by PL-MNC, especially in symptomatic PLs, was a significantly higher compared to its level in the cultures of peripheral blood MNC and correlated with the frequency of CD14 + and CD3 + cells. Exogenous IL-27 stimulated Th1 and downregulated the Th17 cytokine production by PL-MNC from symptomatic PLs, but the downregulated Th1 and Th2 responses in asymptomatic PLs suggested its complex biological functions in PLs 15 .
We showed in this study for the first time that the small-size asymptomatic PLs produce the higher levels of IL-27 than the large-size PLs. These results suggest that the immunoregulatory mechanisms are differently regulated during the progression of PL development, having in mind the opposite production of IL-10 and TGF-β. Our results also suggest that, during initiation of the PL development, IL-27 may have a more important down-modulatory role in suppressing the production of Th1 and Th2 cytokines, than the other two cytokines may have. However, under the certain conditions, IL-27 can stimulate the Th1 response, which is mediated by the activated CD4 + T-cells, recently migrated into the periapical tissue. This finding was already observed in our previous paper, in which we showed that exogenous IL-27 stimulated the production of IFN-gama by PL-MNC, especially during the exacerbation phase of the PL development 15 .

Conclusion
The results of this study, performed on clinically asymptomatic PLs, showed that the small-size lesions differ from the large-size lesions in terms of IL-10, IL-27 and TGF-β production. ICs from the small-size PLs produce significantly much higher quantity of IL-27 than the large-size PLs. It was the opposite with IL-10 and TGF-β. Cumulatively, it can be hypothesized that these immunoregulatory cytokines may play different roles in the suppression of unwanted immune/inflammatory responses in asymptomatic PLs, depending on the extension of the pathological process.

Acknowledgement
The study was supported by the project (grant No.

Declaration of interest
Authors declare no conflict of interest.