IL-22-Producing RORγt-Dependent Innate Lymphoid Cells Play a Novel Protective Role in Murine Acute Hepatitis

Retinoid-related orphan receptor (ROR) γt is known to be related to the development and function of various immunological compartments in the liver, such as Th17 cells, natural killer T (NKT) cells, and innate lymphoid cells (ILCs). We evaluated the roles of RORγt-expressing cells in mouse acute hepatitis model using RORγt deficient (RORγt−/−) mice and RAG-2 and RORγt double deficient (RAG-2−/− × RORγt−/−) mice. Acute hepatitis was induced in mice by injection with carbon tetrachloride (CCl4), to investigate the regulation of liver inflammation by RORγt-expressing cells. We detected RORC expression in three compartments, CD4+ T cells, NKT cells, and lineage marker-negative SCA-1+Thy1high ILCs, of the liver of wild type (WT) mice. CCl4-treated RORγt−/− mice developed liver damage in spite of lack of RORγt-dependent cells, but with reduced infiltration of macrophages compared with WT mice. In this regard, ILCs were significantly decreased in RAG-2−/− × RORγt−/− mice that lacked T and NKT cells. Surprisingly, RAG-2−/− × RORγt−/− mice developed significantly severer CCl4-induced hepatitis compared with RAG-2−/− mice, in accordance with the fact that hepatic ILCs failed to produce IL-22. Lastly, anti-Thy1 monoclonal antibody (mAb), but not anti-NK1.1 mAb or anti-asialo GM1 Ab administration exacerbated liver damage in RAG-2−/− mice with the depletion of liver ILCs. Collectively, hepatic RORγt-dependent ILCs play a part of protective roles in hepatic immune response in mice.


Introduction
Retinoid-related orphan receptor ct (RORct) is a transcription factor that regulates a variety of immunological processes [1][2][3] and has an indispensable role in the development of Th17 cells [1]. Activated Th17 cells secrete a variety of IL-17 family cytokines including IL-17A, IL-21, and IL-22, which promotes tissue inflammation by induction of other proinflammatory mediators and the recruitment of leukocytes to sites of inflammation [4]. Among IL-17 family cytokines, the role of IL-22 in inflammatory responses is unclear owing to contrary data suggesting pro-or antiinflammatory functions in distinct tissues [5][6][7][8][9]. However, during the pathogenesis of acute hepatitis models, IL-22 produced by Th17 cells is thought to have a protective role by preventing tissue injury [7][8][9].
The development of all T cells and NKT cells depend on RORct to some extent, as it is expressed by CD4 and CD8 double positive thymocytes [10,11]. Con A-induced acute hepatitis is a lymphocytemediated hepatitis model in rodents [12], largely dependent on NKT cell secretion of IFN-c, TNF-a, and IL-4 [13][14][15][16].
Furthermore, RORct is essential for generation of lymphoid tissue inducer (LTi) cells, which are critically involved in the development of secondary lymphoid tissues, such as lymph nodes, Peyer's patches, and cryptopatches [3]. Recent studies reported that various subtypes of RORct + innate lymphoid cells (ILCs), including LTi cells, producing IL-17A, IL-22 and/or IFN-c have various roles in innate immune responses, lymphoid tissue formation, and tissue remodeling [2,17].
A1though the roles of RORct-dependent Th17 cells or NKT cells in the development of murine acute hepatitis models have been clarified, those of RORct dependent ILCs have not been investigated. To clarify the roles of RORct-dependent ILCs in the development of acute hepatitis, we induced CCl 4 -hepatitis in RORct 2/2 and RORct 2/2 6RAG-2 2/2 mice.

Materials and Methods
Mice Eight-to 12-wk-old C57BL/6 (WT) mice were purchased from Japan CLEA (Tokyo, Japan). C57BL/6 background RAG-2deficient mice were obtained from Central Laboratories for Experimental Animals (Kawasaki, Japan). Mice with green fluorescent protein reporter complementary DNA knocked-in at the site for initiation of RORct translation on the C57BL/6 background (RORct 2/2 ) were kindly provided by Dr. D. Littman [18]. RAG-2 2/2 6 RORct 2/2 mice were obtained by crossing RAG-2 2/2 mice with RORct 2/2 mice. Mice were maintained under specific pathogen-free conditions in the Animal Care Facility of Keio University School of Medicine. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of Keio University School of Medicine. All surgery was performed under anesthesia, and all efforts were made to minimize suffering.

Preparation of hepatic mononuclear cells
Hepatic mononuclear cells (MNCs) were isolated from the liver as described previously [19]. Briefly, livers were perfused through the portal vein with PBS, then minced and passed through nylon mesh. The filtrate was centrifuged at 506g for 50 seconds and supernatant was collected and centrifuged. Cells were suspended in a Hanks' balance salt solution and overlaid on a Histopaque solution (Sigma-Aldrich, St. Louis, MO, USA). After centrifugation at 2000 rpm for 20 minutes, the cells were collected from the upper layer of the Histopaque.

Measurement of liver injury
Serum ALT levels were measured using the LDH-UV kinetic method (SRL Inc., Tokyo, Japan). Livers were fixed in 10% formalin and embedded in paraffin. Sections were stained with H&E and examined. Statistic evaluation was performed by using pathological score referring to the previous report [20]. Briefly, the overall degree or grade of necrosis was scored from 0 to 4 on the basis of the severity and distribution of the necrotic lesions and the number of lobes affected as follows: 0-None, 1-Minimal to Mild, 2-Moderate, 3-Marked, 4-Severe to Diffuse.

Flow cytometric analysis and cell sorting
Flow cytometric analysis was performed as described previously [19]. Briefly, after blocking with anti-FcR (CD16/32, BD Pharmingen, San Jose, CA, USA) for 15 minutes, the cells were incubated with fluorescence-labeled mAb at 4uC for 20 minutes. For intracellular staining, cells were stimulated for 3 h with mouse recombinant IL-23 (40 ng/mL, R&D Systems, Minneapolis, MN, USA). For the final 1.5 h, GolgiPlug (BD Pharmingen) was added. After surface staining, the cells were resuspended in Fixation/ Permeabilization solution (BD Pharmingen), and intracellular staining was performed. In terms of intracellular staining for transcriptional factor, Fixation/Permeabilization Concentrate (eBioscience) was used, and intracellular staining was performed without stimulation. The following mAbs were used: anti-CD11b

RT-qPCR
RNA was extracted from liver tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Complementary DNA was synthesized from extracted RNA using TaqManH Reverse Transcription Reagents (Applied Biosystems, Foster City, CA, USA). Reverse transcription was performed at 25uC for 10 min, 48uC for 30 min, and then 95uC for 5 min. Complementary DNA was analyzed by RT-qPCR using TaqManH Universal PCR Master Mix (Applied Biosystems) in Applied Biosystems StepOne TM /StepOnePlus TM (Applied Biosystems). The following probes were purchased from Applied Biosystems: Il6 (Mm00446190_m1), Rorc (Mm01261022_m1) and mouse b-actin. Relative quantification was achieved by normalizing to the value of the b-actin gene.

Statistical analysis
The results were expressed as the mean 6 standard error of mean (SEM). Groups of data were compared using the Student's ttest. In terms of the evaluating pathological score, data compared using the Mann-Whitney test. Differences were considered to be statistically significant when p,0.05. ***p,0.001, **p,0.01, *p,0.05. NS, not significantly different.

Results
CD4 + T cells, NKT Cells and ILCs in the Liver Expressed RORct, and ILCs Were Decreased in RAG-2 2/2 6 RORct 2/ 2 mice To determine the expression of RORct in the liver of normal mice, we isolated the following MNC subsets: CD11b + TCRb 2 macrophages, CD11c + TCRb 2 dendritic cells (DCs), CD4 + TCRb + CD4 T cells, CD8 + TCRb + CD8 T cells, NK1.1 + TCRb 2 NK cells, NK1.1 + TCRb + NKT cells and Lin 2 SCA-1 + Thy1 high ILCs, and evaluated the level of RORC expression by RT-qPCR. To obtain ILCs from the liver, we used anti-CD3e mAb in addition to the lineage markers [2], to further exclude T and NKT cells. Thus, using the isolated liver cell populations we detected the expression of RORct in CD4 T cells, NKT cells, and ILCs, and observed the expression of RORct in ILCs was much higher than in CD4 T cells and NKT cells (Fig. 1A).

RORct 2/2 Mice Develop Carbon Tetrachloride (CCl 4 )induced Acute Hepatitis
To assess the role of RORct dependent cells including NKT cells, Th17cells and hepatic RORct + ILCs in hepatitis, we used a carbon tetrachloride (CCl 4 )-induced hepatitis model in WT and RORct 2/2 mice. CCl 4 administration directly damages hepatocytes and induces hepatitis. [21] IL-22 produced by Th17 cells has been reported to act protectively as an epithelial repair factor in the CCl 4 -induced hepatitis model [7][8][9]. However, there was no significant difference in the serum ALT levels between WT and RORct 2/2 mice after CCl 4 treatment ( Fig. 2A). Histological assessment showed a similar level of liver injury in CCl 4 -administered WT and RORct 2/2 mice (Fig. 2B). Interestingly, although the infiltration of macrophages is essential for the development of this model [19], the rate and number of macrophages were significantly decreased in RORct 2/2 mice compared with WT mice after CCl 4 injection (Fig. 2C). Furthermore, both the frequency and absolute number of NKT cells in the liver of CCl 4 -administered RORct 2/2 mice were also significantly lower than those of CCl 4 -administered WT mice (Fig. 2D). We also found that aGalCer, the specific simulator for NKT cells could induce the infiltration of inflammatory macrophage into the liver, and that phenomenon was disappeared in RORct 2/2 mice (Fig. S1). The development of the similar degree of hepatitis in RORct 2/2 mice following CCl 4 administration compared with WT mice despite a significant decrease in the number of pathological macrophages and NKT cells indicate the existence of some RORct-dependent protective cells besides RORct-dependent pathological NKT cells in this model.

Hepatic RORct + ILCs Are Different from Conventional NK cells which Aren't Related to the Development of CCl4induced Hepatitis
To further evaluate the possible involvement of NK cells in this model of CCl 4 -induced hepatitis, anti-asialo GM1 Ab was administered to RAG-2 2/2 mice 1 day and 3 hours prior to before CCl 4 -injection to deplete conventional NK cells. As the result, there was no difference between PBS treated-and antiasialo GM1 Ab treated-RAG-2 2/2 mice as to not only the level of  serum ALT, but also the degree of histological liver injury ( Fig. 5A  and 5B). Absolute cell number of macrophages was also not significantly affected by the anti-asialo GM1 Ab treatment (Fig. 5D), although NK cells were efficiently depleted (Fig. 5E), indicating that NK cells are not essential for the development of this model. Furthermore, the injection of anti-asialo GM1 Ab did not deplete hepatic RORct + ILCs (Fig. 5E). This result suggested that hepatic RORct + ILCs are different cell population from NK cells.

Anti-Thy1, but not Anti-NK1.1, mAb Treatment Depletes Hepatic ILCs and Exacerbates CCl 4 -induced Hepatitis
To confirm the novel protective role of RORct-expressing ILCs in the CCl 4 hepatitis model, RAG-2 2/2 mice were treated with anti-Thy1 mAb or anti-NK1.1 mAb to deplete ILCs, as previously shown in murine models of acute colitis [22], and then administered CCl 4 . As shown in Fig. 6A, CCl4-induced serum ALT levels were markedly increased by the anti-Thy1 mAb treatment, but not the anti-NK1.1 mAb treatment. This was concordant with the histological extent of liver damage (Fig. 6B), and an increase in the rate and absolute cell number of macrophages in the liver of anti-Thy1 mAb-treated RAG-2 2/2 mice (Fig. 6C). Although anti-NK1.1 mAb-treated RAG-2 2/2 mice had a higher frequency of macrophages, owing to NK cell depletion, there was no difference in the absolute cell numbers among these groups (Fig. 6C). We confirmed that anti-NK1.1 mAb effectively depleted almost all NK cells in the liver, while anti-Thy1 mAb substantially but not completely depleted NK cells (Fig. 6D). Furthermore, in sharp contrast to anti-NK1.1 mAb treatment, ILCs were completely depleted by anti-Thy1 mAb treatment, whereas anti-NK1.1 mAb treatment increased the relative ratio and absolute cell number of ILCs (Fig. 6E). The effects of anti-NK1.1 mAb treatment in RAG2 2/2 mice were consistent with those of anti-asialo GM1 Ab that depletes conventional NK cells (Fig. 5).

Discussion
The current study demonstrated that RORct-dependent hepatic ILCs played a protective role on murine acute liver injury hepatic MNCs from CCl 4 -injected RAG-2 2/2 or RAG-2 2/2 6RORct 2/2 mice 12 h after CCl 4 injection was determined by RT-qPCR. Data show the mean 6 SEM (n = 4/group). Data are representative of three independent experiments. doi:10.1371/journal.pone.0062853.g003 in IL-22 dependent manner. Fig. 7 gives a schematic view of the proposed roles of RORct-dependent hepatic ILCs in acute hepatitis models revealed in this study.
Previous reports have demonstrated the importance of RORctdependent cells in various immune responses, as RORct is essential for the development of lymphoid tissues and innate and adaptive immune cells such as Th17 cells, invariant or IL-17A producing NKT cells and ILCs [1,2,10,11,17,23]. To investigate the roles of RORct-dependent cells in the liver, we focused on RORct expression in hepatic MNCs subsets. We found that hepatic ILCs highly expressed RORct in addition to CD4 + T cells and NKT cells, and the number of hepatic ILC cells was considerably decreased in the liver of RAG-2 2/2 6 RORct 2/2 mice. This suggests the possibility that RORct-dependent hepatic ILCs are involved in hepatic immune responses. Notably, previous reports have emphasized that IL-22 produced by Th17 cells can act protectively in various acute hepatitis models, although NKT cells are pathogenic in the development of Con A-induced hepatitis [13,16]. Furthermore, we found that stimulated NKT cells induced the accumulation of inflammatory macrophages in the liver.
To investigate the roles of RORct-dependent cells in the pathogenesis of acute liver injury, we first performed CCl 4induced acute hepatitis model in WT and RORct 2/2 mice. Administration of a single dose of CCl 4 , characterized by the accumulation of macrophages and NKT cells, [19,21] caused a severe acute hepatitis both in WT and RORct 2/2 mice. Development of a similar degree of hepatitis following CCl 4 administration irrespective of the significant decrease in the number of macrophages and NKT cells in the injured liver of CCl 4 -injected RORct 2/2 mice compared with WT mice indicated that some RORct-dependent pathological and protective  immune cells keep a delicate balance in this model. As a candidate of RORct-dependent protective cells, we focused on hepatic ILCs that highly express RORct. In this regard, a Lin 2 SCA-1 + Thy1 high ILC subset expressing IL-17A, IFN-c, and IL-22 was shown to reside in the liver of H. hepaticus-infected mice, but their functional roles were not determined [2,22]. We identified the existence of these hepatic cells also in the normal condition without H. hepaticus-infection, and clarified that they have a similar feature with previously reported intestinal ILCs as to the expression of some surface markers [2]. We next performed CCl 4 -induced acute hepatitis model using RORct 2/2 and RAG-2 2/2 6 RORct 2/2 mice to investigate the roles of RORct-dependent cells in the liver except Th17 cells and NKT cells. As expected, CCl 4 -adminstered RAG2 2/2 6RORct 2/2 mice showed a severer hepatitis compared with CCl 4 -adminstered RAG-2 2/2 mice. Interestingly, hepatic MNCs from CCl 4 -treated RAG-2 2/2 mice consistently expressed IL-22, although they lack Th17 cells previously known as the source of IL-22 in the development of murine acute hepatitis models [7,9]. On the other hand, hepatic MNCs from CCl 4 -treated RAG-2 2/2 6 RORct 2/2 mice lacked IL-22 expression. Supporting these results, only Lin 2 SCA-1 + Thy1 high hepatic ILCs expressed IL-22 in the liver CCl 4 -treated RAG-2 2/2 mice, and hepatic ILCs obtained from the liver of CCl 4 -treated RAG-2 2/2 mice, but not those from CCl 4 -treated RAG-2 2/2 6 RORct 2/2 mice, produced IL-22 in response to IL-23 in vitro. Furthermore, the number of RORct + hepatic ILCs was also decreased, and exogenous IL-22 administration protected RAG-2 2/2 6 RORct 2/2 mice from hepatitis. This suggests a novel protective function of IL-22-expressing hepatic ILCs against acute liver injury. This was also confirmed using depleting mAbs. Hepatic RORct + ILCs were depleted by anti-Thy1 mAb, but not by anti-NK1.1 mAb or anti-asialo GM1 Ab, in the liver of CCl 4 -administrated RAG-2 2/2 mice causing a severer hepatitis. These results indicated that hepatic RORct + ILCs are different from conventional NK cells and act protectively against liver injury. Although hepatic RORct + ILCs produce IFN-c which is one of the pathological factors in CCl4-induced hepatitis, [24] that function of hepatic RORct + ILCs seems to be not critical for the development of hepatitis since there was no difference between the livers from CCl 4 -administrated RAG-2 2/2 mice and RAG-2 2/2 6 RORct 2/2 mice in terms of the expression of IFN-c.
We also noticed that RAG-2 2/2 6RORct 2/2 mice developed severer hepatitis following CCl 4 administration when compared with RORct 2/2 mice. These results may indicate that other tissue-protective cell subsets in B cells, T cells and/or NKT cells can ameliorate CCl 4 -induced tissue damage, although further study is needed to address this issue in the future.
The functions of some hepatic immunocompetent cells, such as T cells, NKT cells and macrophages have been believed to include defense against microbial pathogens or viral infection in the liver, [25][26][27] but can also inhibit the growth of cancer [28]. These cells mediate these functions by secreting inflammatory cytokines, like IFN-c and TNF-a, to induce cytotoxicity in infected cells and tumor cells [25][26][27][28]. Accordingly, those inflammatory cells also induce liver injury when activated excessively, and cause acute hepatitis as shown in some murine acute hepatitis models [19,29]. IL-22 produced by RORct + hepatic ILCs may counteract inflammatory cells-mediated liver injury because IL-22 is a survival factor for hepatocytes by preventing and repairing liver damage [7]. RORct is critically involved in the development and function of hepatic RORct + ILCs, also is important for Th17 cell development [1,2,4,10,11,17], thus Th17 cells may also play a role in the pathogenesis of liver inflammation, although this is still controversial [9,30,31]. However, as to the source of IL-22, Th17 cells certainly have a protective role against liver injury [9,23]. Although hepatic RORct + ILCs also have an ability to produce IL-22, hepatic RORct + ILCs may be able to act at an early innate immune response stage of liver injury compared with Th17 cells since they exist also in the naïve condition.
We conclude that RORct + hepatic ILCs have a pivotal protective function against liver injury via IL-22 production. Therefore, RORct + hepatic ILCs may be involved in clinical hepatitis and represent a candidate for a target of novel treatment for clinical hepatitis since IL-22 can prevent and repair liver damage.