Inhibition of the CXCL9-CXCR3 axis suppresses the progression of experimental apical periodontitis by blocking macrophage migration and activation

Apical periodontitis (AP) is an acute or chronic inflammatory disease caused by complex interactions between infected root canal and host immune system. It results in the induction of inflammatory mediators such as chemokines and cytokines leading to periapical tissue destruction. To understand the molecular pathogenesis of AP, we have investigated inflammatory-related genes that regulate AP development. We found here that macrophage-derived CXCL9, which acts through CXCR3, is recruited by progressed AP. The inhibition of CXCL9 by a CXCR3 antagonist reduced the lesion size in a mouse AP model with decreasing IL-1β, IL-6 and TNFα expression. The treatment of peritoneal macrophages with CXCL9 and LPS induced the transmigration and upregulation of osteoclastogenic cytokines such as IL-1β, IL-6 and matrix metalloprotease 2, a marker of activated macrophages. This suggests that the CXCL9-CXCR3 axis plays a crucial role in the development of AP, mediated by the migration and activation of macrophages for periapical tissue destruction. Our data thus show that CXCL9 regulates the functions of macrophages which contribute to AP pathogenesis, and that blocking CXCL9 suppresses AP progression. Knowledge of the principal factors involved in the progression of AP, and the identification of related inflammatory markers, may help to establish new therapeutic strategies.

TRAP staining. TRAP staining solution was prepared by adding the following solutions and reagents: Administration of CXCR3 antagonist in the mouse AP model. The AP mice received a gavage administration of SCH546738 (30mpk, diluted to 0.2 ml using 0.5% w/v methylcellulose) or a 0.5% w/v methylcellulose (vehicle) every 2 days from day 2 until the day before sacrifice. The mice without AP received a 0.5% w/v methylcellulose vehicle every 2 days from day 2 until the day before sacrifice and used as a control. The CXCR3 antagonist SCH546738 was purchased from MedChemExpress (Monmouth Junction, NJ).
Effect of CXCL9 on activation of peritoneal macrophages. Peritoneal macrophages were prepared as described previously 45 . Briefly, cell suspensions were divided into four groups in a 6-well plate, cultured for 1 h, and then washed twice to remove non-adherent cells and used as peritoneal macrophages. These cells were later treated with LPS (100 ng/mL), recombinant mouse (rm) CXCL9 (100 ng/mL, 492-MM; R&D Systems) or LPS (100 ng/mL) and rmCXCL9 (100 ng/mL) in the presence of 10% FBS for 24 h. After incubation, the cells were assayed by the gene expression analysis of Cxcl9, Cxcl10, Il1b, Il6, Mmp2, Tnf (primer lists: sTable 1), as described previously 44 .
Effect of CXCL9 on the chemotaxis of macrophages. To perform the migration assay, peritoneal macrophages were inoculated into a QCM Chemotaxis Assay 24-well plate (ECM506, EMD Millipore Corp, Billerica, MA), and then treated with rmCXCL9 (100 ng/mL), LPS (100 ng/mL) or rmCXCL9 and LPS in the presence of 10%FBS for 24 h. Quantification of the migrated cells was performed using a VersaMax Microplate Reader in accordance with the manufacturer's protocol (Molecular Devices, Sunnyvale, CA).

Statistical analysis.
The study results were all tested in triplicate. The means and standard deviations were thereby calculated and the statistical significance of the differences between groups were examined using oneway analysis of variance (ANOVA), followed by a Dunnett's multiple comparison test, or a multiple comparison test followed by a Tukey-Kramer test. Test results were considered significant at probability (p) values < 0.05.

Results
Development of a mouse AP model and analysis of the host-immune/inflammatory response. To  www.nature.com/scientificreports/ mice was commenced from 4 weeks until 12 weeks (sFig. 1a). The average length (RL) of the mesial root canal was 1.1 mm at 4 weeks and increased to 1.4 mm at 9 weeks. The mean root canal width (RW) was 0.27 mm at 4 weeks but decreased to 0.10 mm at 9 weeks (sFig. 1b). These results suggested that the open apices of the mesial and distal roots were completely closed by 9 weeks, indicating the completion of root development. Hence, we decided to use 10-week-old mice for all further experiments. It is well documented that AP is caused by oral bacterial infection which stimulates host inflammatory/ immune responses [1][2][3][4][5] . Sequential changes in inflammatory responses are required for bone destruction and we thus performed spatial temporal investigations during the onset of AP in our mouse model. We thus investigated lesion development and the inflammatory process in our mouse AP model (sFig. 2a), as described previously 42,43 . A 3D reconstruction image (axial, coronal and sagittal views) of the mesial root of the mandibular first molar revealed an apical radiolucent area (highlighted in yellow) on day 3 that gradually increased up to 28 days (Fig. 1a, sFig. 2b). The size of this area was 0.043 mm 3 on day 3 and progressively increased to 0.048 mm 3 and 0.075 mm 3 at days 14 and 28 after exposure, respectively. This radiolucent area increased slowly thereafter, but not statistically significantly, up to 56 days (Fig. 1b). Quantitative µCT analysis of the lesion (i.e. apical radiolucent area) volume showed a significant difference between the animals without exposure (W/O) and those with exposure after 28 days. In addition, inflammatory cell infiltration at the AP was clearly increased after 28 days (Fig. 1c) indicating the development of AP and consistent with previous findings 47,48 . We then performed immunostaining to investigate the extent of neutrophil infiltration at the periapical area following pulp exposure. Infiltration by Ly6g (a neutrophil marker) and MMP9 (an inflammatory marker) were observed until 14 days but these cells had disappeared by 28 days (Fig. 1d). During the maturation stage of the apical lesion, anti-Mac2 positive macrophages were detectable at 14 days after the exposure and could be observed at remarkably increased levels after 28 days www.nature.com/scientificreports/ (Fig. 1e). These results indicated that a neutrophil-based inflammatory response was active as an early process at 14 days and a macrophage-based reaction was functioning at 28 days after exposure. These results further suggested that an AP-initiated innate immune response against bacterial infection induces neutrophil infiltration at the initial stages and that macrophages are subsequently recruited during the development of the lesion.
Identification of CXCL9 as a major chemokine expressed in macrophages in the AP mouse model. We next conducted global gene expression analysis of the innate and adaptive immune systems in our AP model using an RT 2 PCR Array (PAMM-052ZR) to determine whether macrophage-based inflammation was involved in the development of AP. The results of our PCR Array analysis of inflammation-associated factors indicated an increased expression of inflammatory markers related to toll-like receptor signaling such as Nfkb1, Myd88, Tlr4, Cd14, and cytokines such as Tnf and Il6, until 28 days after exposure (Fig. 2a,b). These data indicated that macrophage migration accompanied by toll-like receptor signaling is involved in the development www.nature.com/scientificreports/ of periapical lesions. Hence, we speculated that the molecules which regulate macrophage migration and activation, such as chemokines and cytokines, affect periapical lesion formation.
To test this hypothesis, we performed PCR Array (PAMM-011Z) analysis of inflammatory cytokines and receptors. In addition to inflammatory markers including Il1b, chemokines have also been shown to be highly expressed after 14 days of left mandibular first molar pulp exposure (sFig. 3). Among these factors, we found a gene cluster of chemokines such as Ccl 5,8,12,19,Cxcl 1,5,9,10,12,13 and Ccr2 which was highly expressed at day 28, while no expression of anti-inflammatory cytokines such as IL-10 and TGFβ was observed. Interestingly, the expression pattern of these genes was almost identical to that of Il1b, which has been shown as a critical factor for AP 43 , suggesting that these genes could play a role in AP development (Fig. 2c). We then further confirmed the expression of these molecules using quantitative PCR analysis. Among these factors, CXCR3-and CCR3related chemokines such as Cxcl9, Ccl5 and 8 became increased at day 7 and were found to be further elevated at days 14 and 28 (Fig. 2d). Cxcl10 was also increased at days 7 and 14 but decreased at day 28. Cxcl9 was more highly expressed than the others and we performed immunohistochemical analysis to investigate the localization of these chemokines at periapical tissues. As shown in Fig. 2e, macrophage-derived CXCL9 and Mac2-positive macrophages were observed, suggesting that macrophages expressing CXCL9 are recruited during the onset of periapical lesions. Anti-inflammatory cytokines such as IL-10 and TGFβ were not detected (data not shown) whereas FoxP3, a marker for regulatory T cells, was found to be upregulated but quantitative PCR analysis did not confirm its expression pattern. We therefore did not investigate these genes further in our present analyses. The onset of AP was investigated up to 28 days in this study, and detailed analyses are required for more than 28 days to understand the expression of other inflammatory markers.
A CXCR3 antagonist reduces the AP lesion volume. To investigate whether CXCL9 plays a role in periapical lesion formation, an antagonist of this cytokine was administrated into the mouse AP model via oral gavage and evaluated by µCT and histological analysis. SCH546738 is a small molecule non-competitive CXCR3 antagonist with a far higher affinity than AMG487 and binds to CXCR3 at an allosteric site and changes its conformation 49 .
Sagittal sections of µCT images revealed smaller periapical lesion formation at the apex of the mesial and distal roots in the SCH546738 treatment group (Fig. 3a). Quantitative analysis indicated that this CXCR3 antagonist significantly inhibited lesion formation compared with vehicle (Fig. 3b). HE staining analysis further revealed that bone resorption at periapical tissues was reduced in CXCR3 antagonist treatment group compared with the vehicle control. A higher magnification view confirmed that CXCR3 antagonist administration resulted in fewer inflammatory cell infiltrations (arrows) in the periapical lesion compared with the vehicle (Fig. 3c). Quantitative PCR analysis indicated that the SCH546738 treatment did not affect the expression of Cxcl9 or 10 but that a significant reduction of inflammatory cytokines such as Il1b, Il6 and Tnf was also caused by this antagonist (Fig. 3d). HE staining showed fewer inflammatory cell infiltration and immunohistochemical analysis confirmed that fewer CD11b-positive inflammatory cells infiltrations were evident in the SCH546738 treatment group (Fig. 4a). In contrast, neither inflammatory cell infiltration nor bone resorption were affected in the control group. Quantitative analysis of CD11b-positive cells confirmed that the CXCR3 antagonist significantly inhibited inflammatory cell migration compared with the vehicle treatment (Fig. 4b). Taken together, these results indicated that a CXCL9-CXCR3 axis plays an important functional role in the onset of periapical lesions through the activation of macrophages.

CXCL9 induces macrophage migration and activation.
We further evaluate the role of CXCL9 in the migration and activation of macrophages using mouse peritoneal macrophages and a human monocytic leukemia cell line (THP-1). A chemotaxis assay was first performed to evaluate macrophage migration following exposure to rmCXCL9. The results indicated that macrophages had migrated across the membrane under the influence of rmCXCL9 in a similar manner to those in a positive-control LPS group (Fig. 5a). This migration ability was enhanced by the combination of rmCXCL9 and LPS treatments (Fig. 5b). We next performed a macrophage inflammatory assay using quantitative PCR to evaluate the role of CXCL9 in the activation of these immune cells. Quantitative PCR analysis revealed that an rmCXCL9 single treatment had no obvious effects on the expression levels of inflammatory markers such as Cxcl9, 10, Il1b, Il6, Tnf and Mmp2. Notably however, Cxcl10, Il1b, Mmp2 and Il6 showed high expression upon exposure to a combination of rmCXCL9 and LPS (Fig. 5c). The macrophage-activation effects of rmCXCL9 were confirmed in the THP-1 cell line in which inflammatory markers such as CXCL9, IL1B, IL6, TNF, and MMP2 are significantly induced. The expression of these markers was inhibited by treatment with SCH546738, a CXCR3 antagonist.
To next investigate the regulation of macrophage activation via the CXCL9-CXCR3 axis, we applied rmCXCL9 and the CXCR3 antagonist to THP-1 cells after Phorbol12-myristate13-acetate (PMA) treatment. The results supported a regulatory role of CXCL9 in the migration of macrophages and in lesion development in AP (sFig. 4). Quantitative PCR analysis indicated that treatment of the THP-1 cells with rmCXCL9 upregulated the inflammatory markers associated with macrophages such as IL1B, IL6, TNF and MMP2. These markers were in turn suppressed by the CXCR3 antagonist.
Finally, we functionally tested whether rmCXCL9 directed the differentiation of macrophages into osteoclasts. Immunostaining of cathepsin K (Fig. 6a) or TRAP (Fig. 6b) staining revealed an inhibition of osteoclast recruitment in the periapical lesion upon exposure to the SCH546738 CXCR3 antagonist. An osteoclast differentiation assay using M-CSF dependent bone marrow macrophages (MDBMs) indicated an absence of TRAP-positive multinucleated like cells in the M-CSF/CXCL9 group similar to that of treatment with M-CSF only (negative control) (Fig. 6c). In contrast, multinucleated cells were clearly evident when MBDMs were stimulated with M-CSF/RANKL (positive control), but this was similar to cells treated with M-CSF/RANKL/CXCL9 (Fig. 6c).   www.nature.com/scientificreports/ activated macrophages to induce osteoclastogenic cytokines and promote the differentiation of osteoclast precursor cells into osteoclasts (Fig. 6d). We next investigated the expression of CXCR3 during osteoclast differentiation. The results showed that osteoclast markers such as cathepsin K (Ctsk), calcitonin receptor (Calcr), nuclear factor of activated T-cells (Nfatc1), osteoclast stimulatory transmembrane protein (Oc-stamp) and matrix metalloproteinase 9 (Mmp9) become highly expressed after 3 days of culture, but the relative expression of Cxcr3 (6 >) was lower compared with those of Ctsk (200 >), Calcr (60 >), Nfatc1(25 >), OC-Stamp(20 >) and Mmp9 (700 >) (sFig. 5).
Following the confirmation of CXCR3 expression in osteoclast progenitors, we next investigated the effects of rmCXCL9 on osteoclast differentiation at 0, 3, and 5 days. The results showed that the number of osteoclasts cultured in M-CSF/RANKL did not increase upon treatment with rmCXCL9 at 0, 3 and 5 days (sFig. 6a,b). www.nature.com/scientificreports/ These data suggested that CXCL9 is recruited and activates macrophages to induce osteoclastogenic cytokines and promote the differentiation of osteoclast precursor cells into osteoclasts (Fig. 6d).

Discussion
AP is a chronic inflammatory disease caused by oral bacterial infection. AP is primarily treated by disinfection of the root canal, which is carried out to reduce the immune and inflammatory reactions and thereby promote wound healing of the periapical tissue [1][2][3][4][5] . Anti-inflammatory approaches represent an alternative strategy to promote healing in therapy-resistant AP cases. A mouse AP model has been extensively used to investigate the molecular pathogenesis of this disease, the effects of medications, and the systemic conditions underlying the development of AP 42,43,47,48 . In our current study, we used the AP mouse model to screen for inflammatoryrelated molecules and identify highly expressed markers during AP development. We found from our analyses www.nature.com/scientificreports/ that CXCL9, a small cytokine belonging to the CXC chemokine family, and its receptor CXCR3, play a role in chemotaxis induction and macrophage activation, processes which in turn recruit and activate osteoclasts during AP development. Notably, the blocking of CXCR3 using an antagonist suppressed AP development through the inhibition of macrophage activation and migration. Thus, our current data provide new insights into the future development of anti-inflammatory therapeutic regimens targeting the CXCL9-CXCR3 axis that promote accelerated healing and reduce inflammatory bone destruction in cases of AP (Fig. 7). A growing body of literature has indicated that significant efforts have been made using mouse models to better understand and characterize the signaling pathways in inflammatory bone resorption in AP, and also the cells involved in this disorder 42,43,47,48 . It has been suggested that toll like receptors (TLR) are crucial factors in the activation of the innate immune system in response to primary endodontic infections. A prior study by Daniel Rider et al. reported that TLR2-deficient mice show progression of AP with increasing CD14-positive macrophages, whereas a TLR2/4 double deficiency rescued the progression of AP in the mice, indicating that multiple TLR and innate immune cells are involved in the regulation of AP 48 . Additionally, various studies have proposed that TLR signaling promotes the induction of proinflammatory cytokines such as IL-1β, IL-6 and TNFα that are highly expressed in both an experimentally induced mice AP model and in human periapical lesions www.nature.com/scientificreports/ that promote bone destruction 17,18,42,43,47,48 . In our present study, we confirmed that TLR signaling-related genes, including Tlr4, Nfkb, Myd88, Cd14, Il6 and Tnf, are gradually increased in the AP mouse model following induction of the infection and become highly expressed when lesion formation accompanied by bone destruction is established. Our present results have indicated that bacterial infection from an infected root canal stimulates the recruitment of innate immune cells such as macrophages through the activation of TLR signaling, which leads to a progression of AP. A number of studies investigating the pathogenesis of AP in mouse models have identified an influx of neutrophils at an early stage and that macrophages along with osteoclasts are recruited to the periapical area and cause bone destruction as the lesion progress 42,43,47,48 . Similarly, our current immunobiological data confirmed that neutrophil recruitment was clearly evident during the early immune responses to AP, as a result of the high expression of chemokines (Cxcl15, Ccl1, 3,4,9,11,22,24) and inflammatory cytokines (Tnf, Il3, 4,5,7,13,11,16,17α,17β). As the lesion progressed, we detected the dominant expression of specific chemokines such as Cxcl9, Cxcl10, Ccl5 and Ccl8, which can recruit macrophages, and clear evidence of osteoclast presentation in the lesion area. Macrophages are important phagocytic cells in the innate immune system and secrete proinflammatory cytokines such as IL-1β, IL-6, TNFα to induce the differentiation of osteoclast precursors derived from circulating macrophages. Our findings indicate that specific chemokines involved in macrophage recruitment and activation underlie the eventual progression in AP to tissue destruction. A previous study has reported in this regard that that inhibition of macrophage activation using an anti-CD14 antibody suppressed the AP lesion size in TLR2 KO mice 48 . These findings collectively support the notion that the migration and activation of macrophages induced by chemokines such as CXCL9 contribute to AP lesion formation and eventual alveolar bone destruction via osteoclast recruitment and activation. Recent evidence has suggested a strong correlation between the expression of chemokines and chemokineinduced macrophage recruitment in inflammatory conditions [29][30][31][32] . A number of studies have confirmed the presence of various chemokine receptors such as CCR1, CCR2, CCR3, CCR5, CXCR1, CXCR3 and chemokines such as CCL2, CCL5, CXCL10, CXCL12 in AP lesions from both humans and mice 38,39 . In our present study, we found that numerous chemokines and their receptors are expressed during the development of AP. However, specific chemokines such as CXCL9, CXCL10 and their receptor CXCR3 were found to be involved in macrophage recruitment and activation in our current mouse AP model experiments. It has also been proposed that the axis of CXCL9, CXCL10 and CXCR3 functions primarily in immune cell migration, differentiation, and in the activation of cytotoxic T cells, natural killer T cells and macrophages. The authors of that study concluded that the CXCL9, CXCL10 and CXCR3 expression induced by IFNγ in M1 macrophages causes the expression of inflammatory cytokines including IL-1β, IL-6 and TNFα and results in the progression of tissue destruction in various disorders such as metabolic diseases, asthma, allergies, atherosclerosis, fibrosis, reduced wound healing, Figure 7. Model for the inflammatory mechanism mediated by the CXCL9,10-CXCR3 axis during AP progression. The production of CXCL9 and10 and their receptor CXCR3 is induced by bacterial infection in the root canal. These molecules induce chemotaxis, promote the differentiation of macrophages, and produce osteoclastogenic cytokines to promote bone destruction by activating osteoclasts during the development of AP. A CXCR3 blockade suppresses the tissue destruction caused by an AP lesion through the inhibition of macrophage activation and migration. www.nature.com/scientificreports/ and autoimmunity 50 . Indeed, it has been reported also that CXCL9 and CXCL10 act as macrophage chemoattractants in arthritis, chronic cardiac inflammation, and cancer, thus implicating these chemokines in inflammatory disease [34][35][36][37] . In addition, the accumulation and activation of tissue resident macrophages, such as perivascular macrophages, has been shown to be dependent on CXCR3 and CXCL10 during the arterial remodeling triggered by hemodynamic stress 51 . In contrast with earlier findings, another study reported that CXCR3 signaling suppress M2 macrophage polarization in breast cancer tumors 52 . From this, we speculate that the activation of M1 macrophages by CXCR3 signaling promotes the inflammatory cytokine expression that plays a role in the onset of AP. Further studies are needed however to more precisely elucidate the detailed mechanisms by which the CXCL9, CXCL10 and CXCR3 axis controls macrophage recruitment and activation. Our current cluster gene analysis and quantitative PCR data indicated a high expression of the cytokines Il1b, Il6, Tnf and chemokines Cxcl9, Cxcl10, Ccl5 and Ccl8, in parallel with the presence of macrophages. These macrophages were possibly of an M1 phenotype as they were found to be associated with pro-inflammatory cytokines.
In addition, our current in vitro data revealed an enhanced migration of macrophages under stimulation by both CXCL9 and LPS. Although the responses of THP-1 and peritoneal macrophages to CXCL9 were not identical, both of these cell types showed a high expression of Il1b, Il6 and Tnf by quantitative PCR analysis, which was consistent with our in vivo data. These data collectively suggest that the CXCL9-CXCR3 axis contributes to the macrophage migration and activation that induce the expression of pro-inflammatory cytokines necessary for AP progression and development. Much work has been conducted to date on the potential impact of chemokines on osteoclast activation and the important role of this in the lesion formation and bone destruction that occurs in AP 45,53 . More recent evidence has demonstrated increased levels of CXCL8, CXCL9, CXCL10, and CCL20 in rheumatoid arthritis, indicating the importance of chemokines in inflammatory bone diseases 53 . Among these factors, CXCL10 indirectly induces osteoclast genesis by promoting RANKL and TNFα expression in activated CD4 + T cells. Reciprocally, RANKL also induces CXCL10 expression in osteoclast precursors. In addition, anti-CXCL10 antibodies suppress bone destruction in a collagen-induced arthritis model 36 . Similarly, we found in our present experiments that CXCL9 does not directly promote osteoclast differentiation but activates macrophages which secrete the cytokines (IL-1β, IL-6 and TNFα) that can stimulate osteoclastogenic associated bone loss. It has been well established that RANKL is induced in AP models and our current model showed this expression pattern also. Although we need to further investigate whether CXCL9 accelerates LPS induced osteoclastgenesis 54 , our current data strongly suggest that a CXCL9-CXCR3 axis enhances osteoclastogenesis via specific cytokines released from activated macrophages.
Considering the complex interactions involving CXCL9 secreted from immune cells as a result of root canal bacteria that regulate the functions of the macrophages responsible for periapical inflammation and bone destruction, we investigated the effect of CXCL9 and its mechanism of action in the pathogenesis of AP. The abundant expression of CXCL9 during the development of AP make it a potential therapeutic target. Our mouse AP model was treated with a CXCR3 antagonist that blocks the activity of CXCL9,10 and 11 and we then evaluated the expression of the macrophage-derived proinflammatory cytokines involved in bone resorption. The CXCR3 antagonist decreased the levels of the inflammatory cytokines produced by macrophages, indicating the importance of this chemokine as a target molecule in the protective immune response. The inhibition of periapical lesions by the CXCR3 antagonist possibly modulated inflammatory cytokine expression in these cells, which in turn prevented bone destruction by inhibiting matrix metalloproteinase and recruiting osteoclasts. Indeed, CXCR3 antagonism has been associated with inflammatory and autoimmune diseases such as rheumatoid arthritis, multiple sclerosis, inflammatory bowel disease, systemic lupus erythematosus, chronic obstructive pulmonary disease, psoriasis, organ transplant rejection and graft-vs-host disease, asthma, and type 1 diabetes 50 . CXCR3 is predominantly expressed on Th1 CD4 + T cells and CD8 + cytotoxic T cells, and activates signaling molecules including Gαi protein and β-arrestin that can initiate intracellular signaling events, cell migration and proinflammatory cytokine production 50 . Hence, the CXCR3 receptor is a target molecule for the treatment of various inflammatory diseases. Our present data confirm that the progression of inflammatory bone destruction induced by oral bacterial infection can be suppressed by antagonizing CXCR3. A recent study has also demonstrated that the progression of periodontal disease, an inflammatory disease similar to AP, was inhibited by a blockade of CXCR3 in both deficient mice, and via treatment with the AMG487 antagonist 55 . In our present study, we used SCH546738, a CXCR3 chemokine antagonist similar to AMG487, which is approved by the FDA and has a well demonstrated efficacy in various preclinical models of inflammatory diseases 49 . Our results confirmed that SCH546738 significantly reduces the extent of the apical tissue destruction in an AP mouse model by decreasing the levels of inflammatory cytokines, and hence macrophage recruitment and activation, resulting in reduced osteoclast activation. A recent study has reported faster healing of periapical tissues when a TNFα antibody was used as an anti-inflammatory therapy in inflammatory bowel disease patients, who are considered to be at high risk of therapy-resistant AP 12 . These results strongly suggest that anti-inflammatory approaches involving a blockade of chemokine receptors such as CXCR3 would be a beneficial approach to treating therapy resistant AP. In this study, we only investigated the effect of CXCR3 antagonist on systemic administration but this approach is not appropriate for human translation. Further studies are necessary to develop a local drug delivery system into AP lesion for clinical translation.
In conclusion, our current findings provide new insights into the role of CXCL9 in the macrophage responses and functions in AP and suggest that an intervention involving a pharmacologic blockade of CXCL9 activity is a potential therapeutic approach for therapy resistant AP. CXCR3 antagonists can significantly suppress proinflammatory cytokines during the pathogenesis of AP and have the potential to reduce the periapical lesion size in this disease. CXCL9-induced macrophage migration and activation promote inflammatory cytokine expression, which in turn activates osteoclasts and promotes bone resorption. The acquired/adaptive immune responses play an important role in the development of therapy resistant AP in patients with other systemic diseases, whilst the innate immune response is mainly responsible in the mouse AP model, as shown in our present study. Further www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.