MicroRNA-155-IFN-γ Feedback Loop in CD4+T Cells of Erosive type Oral Lichen Planus

Oral lichen planus (OLP) is a T cell-mediated immune disorder, and we have indicated a Th1-dominated immune response in OLP. MicroRNA-155 (miR-155) could promote Th1 cells polarization. The present study aims to determine the role of miR-155 in immune response of OLP. The expression of miR-155 and the target mRNA was tested by Real-Time PCR. The serum levels of IL-2, 4, 10 and IFN-γ were examined with ELISA. Furthermore, in vitro study was built to observe the function of miR-155 in erosive-type OLP (EOLP). Finally, we determined the expression and correlation of miR-155 and SOCS1 in EOLP CD4+ T cells. The results showed miR-155 was high related with the disease severities. Besides, serum IFN-γ was specifically increased in EOLP group, while IL-4 was decreased. In vitro studies showed miR-155 could reinforce IFN-γ signal transducer, and the induction of IFN-γ could also promote miR-155 expression in EOLP CD4+ T cells. In addition, miR-155 levels were negatively related with SOCS1 mRNA expression in EOLP CD4+ T cells. Our study revealed a positive miR-155- IFN-γ feedback loop in EOLP CD4+ T cell, which might contribute to the Th1-dominated immune response. Furthermore, miR-155 could be used for the evaluation and treatment of OLP.


Results
The levels of miR-155 and cytokines in peripheral blood of OLP. The expression of miR-155 increased in peripheral blood of EOLP patients compared with the control (p < 0.05), in addition, the expression of miR-155 in EOLP group was significant higher than that in NEOLP group (p < 0.05) (Fig. 1A). However, no difference was found between NEOLP group and the controls. Furthermore, the correlation analysis revealed that the miR-155 expression was highly related with the RAE scores which represented the severities of OLP (p < 0.01, r = 0.855) (Fig. 1B), and the correlation coefficient was much higher in EOLP patients (p < 0.01, r = 0.882) (Fig. 1C).
The cytokines profiles in the serum of OLP patients showed that just like the expression pattern of miR-155, the IFN-γ levels of EOLP group increased (p < 0.01), and were higher than that of NEOLP group (p < 0.05) (Fig. 2B). On the contrary, the IL-4 levels decreased in EOLP group (p < 0.01), and were much lower than that of NEOLP group (p < 0.05) ( Fig. 2A). Moreover, there was descend of IL-4 levels in NEOLP group compared with the control (p < 0.05) ( Fig. 2A). For the other two cytokines, the levels of IL-2 only increased in NEOLP group (p < 0.05) (Fig. 2D), IL-10 levels in the two OLP groups both declined (p < 0.05, p = 0.01), and there was no difference between them (Fig. 2C).
The impacts of miR-155 regulation on CD4 + T cell proliferation as well as the levels of IFN-γ and IL-4 in the supernatant. As shown in Fig. 3, when miR-155 was inhibited by antagomir-155, the proliferation of EOLP CD4 + T cell was declined at the time point of 24 h (p < 0.01) and 36 h (p < 0.01) compared with the negative control group (ANNC group). In addition, the levels of IFN-γ were decreased in the supernatant (p < 0.05), but the levels of IL-4 were increased (p < 0.05). when miR-155 was promoted by agomir-155, the proliferation of EOLP CD4 + T cell was raised at all the four time point (12 h, p < 0.05; 24 h, 36 h, 48 h, p < 0.01) compared with the negative control group (ANC group). Moreover, the levels of IFN-γ and IL-4 in the supernatant both decreased (IFN-γ , p < 0.05; IL-4, p < 0.01) (Fig. 4).

The levels of miR-155 in EOLP CD4 + T cell induced with IFN-γ. After induced by IFN-γ for
24 h, the expression of miR-155 in EOLP CD4 + T cell was increased compared with the blank control group (Fig. 5).
The expression of miR-155 and SOCS1 mRNA in EOLP CD4 + T cell. After browsing the potential targets, SOCS1 was found to have one and only one conserved site combined with miR-155. For the first batch of samples, the mean value of SOCS1 mRNA expression in EOLP CD4 + T cell was higher than that of healthy controls, but there was no statistical difference between them (data not shown). Owing to the insufficient of samples, we supplemented the sample number of both EOLP group and the control group. As shown in Fig. 6A,B, it demonstrated that in EOLP group, the expression of miR-155 of CD4 + T cell was increased (p < 0.01), but the expression of SOCS1 mRNA was decreased (p < 0.05). Furthermore, the correlation analysis revealed that the expression of miR-155 in CD4 + T cell was high related with the RAE scores of EOLP patients (p < 0.01), and the correlation coefficient reached up to 0.917 (Fig. 6C). In addition, it demonstrated a negative correlation between the SOCS1 mRNA expression and the miR-155 levels (p < 0.01, r = − 0.541) (Fig. 6D).

Discussion
The miRNAs in peripheral blood were easy to be collected and detected, thus, many researches aimed to find disease-related miRNAs in circulation to assist the diagnosis and study the remote and systemic   regulation of miRNA 16,17,[26][27][28] . Through screening the sera, Nylander E had ever drawn a pessimistic prediction that there was no specific LP-associated miRNA profile in peripheral blood 29 . However, in 2013, our group first reported that miR-125a was down-regulated in peripheral blood of OLP patients 5 .
In the present study, the result showed that miR-155 was up-regulated in peripheral blood of EOLP patients. Furthermore, miR-155 was found to be high related to the RAE scores of OLP patients (p < 0.01, r = 0.855), and the relationship coefficient was higher when the subjects was limited to EOLP patients (p < 0.01, r = 0.882). It implicated that miR-155 might be a positive candidate to assess the severities of EOLP patients.
We and some researchers had proved that there was chasm about the immune mechanisms of EOLP and NEOLP in peripheral blood, like the expression of transcription factor, the activation of T cells  was used to calculate the relative expression of miR-155 and SOCS1 mRNA. The significant differences between EOLP group and the healthy controls were tested by independent-samples t test, and results are represented as box plots. (C,D) Pearson's correlation test was performed, and statistical significance as well as the correlation coefficient is shown. Each dot plot represents a subject, and the correlation is fitted into a straight line. and the levels of cytokines. Furthermore, clinically, these two types of OLP showed great difference in treatment, outcome and the malignant risk 3,5-7 . As the inflammatory bowel disease had two predominant types of crohn's disease and ulcerative colitis, in our opinion, NEOLP and EOLP might be considered as two types of diseases but not just different manifestations of OLP 30 . The current data represented that the expression of miR-155 in EOLP group was higher than in NEOLP group; likewise, the level of IFN-γ increased in EOLP group, and was elevated compared with NEOLP group. But nearly the opposite to miR-155 or IFN-γ , the IL-4 level decreased in EOLP group, and was much lower than that of NEOLP group. The serum levels of IL-2 and IL-10 were distinguished between OLP groups with the controls, but no distinction was presented between EOLP group and NEOLP group. Our data further proved that NEOLP and EOLP were quite different in cytokine levels and miRNA expression, and that the immune response of EOLP tended to be much more Th1-dominated than that of NEOLP. In addition, it also revealed that the expression of miR-155 in peripheral blood might be associated with the levels of IFN-γ and IL-4.
IFN-γ could promote Th1 cell differentiation by induction of T-bet, a transcription factor critical to Th1 cell differentiation, and Th1 cells could produce IFN-γ , which formed a positive feedback to reinforce Th1-dominated immune response 10,11 . A second way in which IFN-γ was thought to contribute to Th1 immune response functioned by inhibiting the proliferation of Th2 as well as the secretion of IL-4 31 . IL-4 could play the same role in Th2-dominated immune response. Both IFN-γ and IL-4 should exert their function by binding to their receptors on the membrane and being assisted by JAKs/ STAT pathway (JAKs/STAT1 for IFN-γ and JAKs/STAT6 for IL-4) 32 . The receptor of IFN-γ (IFN-γ R) had two chains, IFN-γ Rα and IFN-γ Rβ , in which IFN-γ Rα was responsible for the combination with IFN-γ , while IFN-γ Rβ was related to the process of signal transduction, in addition, committed Th1 cells could regulate the expression of IFN-γ Rβ and help CD4 + T cells to resist the anti-proliferation effect of IFN-γ 31,33,34 . In most references, miR-155 was reported to inhibit the Th2 cell differentiation and the production of IL-4 by targeting c-MAF, a distinct transcription factor of Th2 cells which was crucial for IL-4 gene transcription 33 . In addition, miR-155 could enhance Th1 cell differentiation by targeting SOCS1, an inhibitor of JAKs/STAT1 pathway, and strengthen the signal transduction of IFN-γ 22,35 . However, there was still debating like Banerjee et al. announcing that miR-155 could target IFN-γ Rα to inhibit IFN-γ signal transduction 31 .
The current study showed that in the supernatant, IFN-γ levels decreased while IL-4 levels increased in the presence of antagomir-155, and the proliferation activity of EOLP CD4 + T cells was abated at 24 h and 36 h post-transfection. This part seemingly agree on the mainstream attitude, as miR-155 was suppressed, the pent-up signal transduction of IFN-γ could not lead to Th1 cell differentiation, and the production of IFN-γ was decreased either. On the contrary, the weakening signal transduction of IFN-γ enhanced the IL-4 secretion and Th2 differentiation. The past work had demonstrated a Th1-dominated immune response in EOLP CD4 + T cells, and this may be why the proliferation activity was decreased at 24 h and 36 h. In the early 12 h, for the compensation of attenuated IFN-γ signal transduction, there was less effect on anti-proliferation of IFN-γ , thus, the proliferation activity showed no difference with the negative control group. However, for the last 12 h (36-48 h), the increasing Th2 cells gradually made up for the decreasing differentiation of Th1 cells, and the immune condition might have been changed.
In the presence of agomir-155, IL-4 levels decreased in the supernatant, and the proliferation activity of EOLP CD4 + T cells was strengthened in the whole process. It could be explained by the enhancement of IFN-γ signal transduction regulated by miR-155, which promoted the Th1 cell differentiation as well as the declination of IL-4 production. For the effect of IFN-γ on anti-proliferation, the proliferation activity of EOLP CD4 + T cells did not show at first timing point (p < 0.05) significant difference as that of the following three timing point (p < 0.01), when the effect was inhibited by regulating the expression of IFN-γ Rβ . However, it seemed hard to elaborate why the levels of IFN-γ in supernatant was decreased. A speculation was made that the growing EOLP CD4 + T cells recruited more and more IFN-γ combining with the IFN-γ Rα on the membrane to keep the enhancement of IFN-γ signal transduction, and the dissociated IFN-γ in the supernatant declined compared with the negative control group. Furthermore, we found that the induction of IFN-γ could also promote the expression of miR-155 in EOLP CD4 + T cells. All these results might reveal a positive feedback of IFN-γ signal transduction and miR-155 expression in EOLP CD4 + T cells: when miR-155 is over-expressed, the signal transduction of IFN-γ is unlimited, which activates the bic gene, and promotes the miR-155 expression. Same phenomenon was observed in innate immune response, but there was no explicit mechanism 36,37 . The feedback of IFN-γ and miR-155 might play an important role in the Th1-dominated immune response of EOLP.
The most possible target of miR-155 to promote the IFN-γ signal transduction was SOCS1, which had one and only one conservative binding sites with miR-155 (Targetscan) 22,31,35,38 . The data showed SOCS1 mRNA decreased in EOLP CD4 + T cells (p < 0, 05), and there was a clear negative correlation between SOCS1 mRNA and miR-155 expression (p < 0.01, r = − 0.541). These results might prove that through targeting SOCS1, miR-155 and IFN-γ formed a positive feedback in EOLP CD4 + T cells. By the way, the expression of miR-155 in EOLP CD4 + T cells represented an extremely high positive relationship with the RAE scores of EOLP patients (p < 0.01, r = 0.917).
In conclusion, we found miR-155 was highly expressed in the peripheral blood of EOLP patients, and was high related with the severity of OLP. Furthermore, a positive feedback loop of miR-155 and IFN-γ was found in EOLP CD4 + T cells, which might contribute of the Th1-dominated immune response in Scientific RepoRts | 5:16935 | DOI: 10.1038/srep16935 EOLP, and SOCS1 was considered to be the most possible target of miR-155 involved in the feedback loop.

Materials and Methods
This experiments followed the principles outlined in the Declaration of Helsinki in the use of human samples and were approved by the Ethics Committee of School and Hospital of Stomatology, Wuhan University with approved NO 2011051. All participating subjects gave their informed consents.

Patients and controls.
The patients involved in this study were clinically and pathologically diagnosed as OLP according to the definition of OLP made by the WHO 4 . In term of the manifestation, they were divided into two groups: erosive type OLP (EOLP) group and non-erosive type OLP (NEOLP) group 39 . Age and gender matched healthy volunteers were recruited as the controls. In the first stage, 10 EOLP patients, 10 NEOLP patients and 10 controls was recruited, and in the second stage, 7 more EOLP patients and 3 more controls were replenished. Table 1 displays the clinical details. The subjects neither had any systemic disorders (such as cardiovascular disease, diabetes mellitus, etc) nor any soft tissue lesions in the oral mucosa. Smokers and severe alcoholics were excluded. Besides, patients on immunotherapy, receiving any medical treatment of OLP (local or systematic) within 3 months or having medicines affecting RNA synthesis and transcription in 6 months should not be included. All patients recruited in this study had been treated as needed following the sample collection.
Evaluation of the severities of OLP patients. RAE (reticular, atrophic and erosive) scoring system recommended by our previous study, which has shown much practicality and efficiency, was used to assess the severity of OLP in different clinical forms ( Serum and EOLP CD4 + T cell isolation. Fourteen milliliter peripheral blood sample was drawn from each subject. Two milliliter blood (without anticoagulation) was kept in room temperature for 1 h, and then centrifuged at 3000rpm for 10 min, and stored the supernatant serum in − 20 °C. Another ten milliliter blood (with anticoagulation) was diluted with equivalent phosphate buffers (PBS) and the sample was transferred to the centrifuge tubes with lymphocytes separation medium (tbdscience Biotech Ltd, Tianjin, China). Isolation of peripheral blood mononuclear cells from EOLP patients was performed by Ficoll-Paque density gradient centrifugation, and then a human CD4 T lymphocyte enrichment set-DM

IFN-γ inducement and miR-155 regulation in EOLP CD4 + T cell. Recombinant Human IFN-γ
(PeproTech, NJ, USA) was reconstituted in water to a concentration of 1.0 mg/ml. Agomir-155 (a modified RNA oligomer specifically increase miR-155 activity), antagomir-155 (a modified antisense RNA oligomer specifically reduce miR-155 activity) and their negative control reagents were prepared as 20 μ mol/L working solution (Shanghai GenePharma Co.,Ltd, Shanghai, China ). For each sample, added 2 ml EOLP CD4 + T cells cultured solution (cell viability > 95%) to two wells of 12-well plate to make sure each well containing 2 × 10 5 CD4 + T cells, meanwhile, sixteen wells in 96-well plate were filled with 6000 CD4 + T cells, and expanded the volume to 200 μ l with Opti-MEM ® I Reduced-Serum Medium (Gibco ® Life Technologies, CA, USA). Then 10 ng Recombinant Human IFN-γ was appended into one well with CD4 + T cells in 12-well plate, and agomir-155, antagomir-155 as well as their negative control regents were transfected into the CD4 + T cells of the sixteen wells in 96-well plate with proper Lipofectamine 2000 Reagent (Life Technologies, CA, USA) to make sure that the concentration of agomir-155 and its negative control regents was 50 nM, and for antagomir-155 and its negative control regents, the concentration was 100 nM. The IFN-γ induced CD4 + T cells were marked as IFN-γ induced group, and CD4 + T cells transfected with agomir-155, antagomir-155 and their negative control regents were marked as A group, AN group, ANC group, ANNC group, respectively. After 24 h, IFN-γ induced CD4 + T cells as well as the supernatants from A group, AN group, ANC group and ANNC group were collected. The threshold cycle (Ct) of three replicates and two replicates for internal control per sample was used to calculate 2 −ΔΔCT .
Enzyme-linked immunosorbent assay (ELISA). ELISA kit (R&D Systems, Minneapolis, MN, USA) was used to test the levels of interleukin 2 (IL-2), IL-4, IL-10 and IFN-γ in serum as well as the supernatant of CD4 + T cells according to the manufacturer's instructions.
Statistical analyses. Data were presented as mean ± SD, and statistical significance was defined as p value < 0.05. The One-Way ANOVA and Pearson's correlation test were performed by SPSS 13.0 for windows software (SPSS, Inc, Chicago, USA).