MiR30a Inhibits Inflammation by Targeting Klf14 in Autoimmune Hepatitis


 MiR30a plays diverse roles in inflammatory diseases, including autoimmune hepatitis (AIH). Klf14 is associated with the inflammation in AIH. We investigated whether miR30a exerts its regulatory function via Klf14. Concanavalin A (Con A)-induced AIH mice were infected with a miR30a agomir or antagomir. MiR30a expression was quantified using qRT-PCR. TargetScan and luciferase reporter assays were used to predict the relationship between miR30a and Klf14. Liver inflammation was evaluated by measuring serum alanine transaminase (ALT) and aspartate aminotransferase (AST) levels, performing histology, and measuring mRNA expressions of inflammatory cytokines and Klf14 by qRT-PCR, protein of Klf14 by western blotting, and Tregs by FACS. MiR30a was downregulated in the hepatocytes (HCs) of AIH mice, which was negatively associated with the liver inflammation. MiR30a overexpression alleviated the inflammation, whereas downregulation of endogenous miR30a aggravated it. The mRNA and protein level of Klf14 were inversely correlated with the miR30a expression. The luciferase reporter assay validated the relationship between Klf14 and miR30a. Moreover, the frequency of Tregs was consistently correlated with the expression of miR30a. MiR30a may play an essential role in AIH, and its ability to regulate the inflammatory responses may, at least partially, be mediated by targeting Klf14 to modulate Tregs.


Introduction
Autoimmune hepatitis (AIH) is a chronic disorder that can present acutely and is characterized by hepatocellular necrosis and in ammation 1 . The speci c mechanism underlying AIH remains unclear.
Concanavalin A (Con A)-induced hepatitis is a well-established animal model of AIH that has been used in numerous preclinical studies of AIH pathogenesis 2,3 . Accumulating evidence has demonstrated that miR30 plays diverse roles in in ammation 4 . In regards to liver diseases, decreased expression of miR30 in hepatocytes (HCs) of the brotic liver and upregulation of miR30 could prevent liver brosis and in ammation 5,6 . Based on our observation of the decreased expression of miR30a in AIH mice, we further investigated the roles of miR30a and its pathways involved in AIH.
Moreover, Krüppel-like factors (Klfs) regulate the pathogeneses of many diseases, including in ammatory disorders. Previous studies have reported the regulatory functions of Klf14 in liver diseases [7][8][9] . One previous study showed that Klf14 played a certain role in AIH by modulating regulatory T cells (Tregs) 10 . Impairment of Tregs is key to the development of AIH 11,12 . Coincidentally, Klf14 is also a potential target of miR30a. Thus, we investigated whether overexpression of miR30a could serve as a therapeutic tool to inhibit liver in ammation by modulating Klf14 expression to regulate Tregs frequency in a Con A-induced AIH mouse model.

Results
Establishment of an AIH model and endogenous miR30a expression in AIH mice Successful establishment of AIH in mice is veri ed in Fig. 1A-C. After administration of Con A, serum ALT and AST levels increased signi cantly as compared to those in the control group (Fig. 1A). Fig. 1B shows H&E staining of the liver in both the control (Ctr) group and Con A-induced (Con A) group. We observed signi cant disruptions in liver structure and lymphocyte in ltration in the liver of the Con A group. Furthermore, we detected higher mRNA expression of proin ammatory cytokines (CD68, TNF-α, IL-1β, and IL-6) by qRT-PCR in the Con A group (Fig. 1C). The effect of Con A on hepatic in ammatory injury was consistent with previous ndings.
To evaluate whether miR30a was aberrantly expressed in the HCs of AIH mice, the expression of miR30a was determined. MiR30a expression was signi cantly downregulated in the HCs of Con A-induced AIH mice. Fold changes in miR30a expression were detected using qRT-PCR and presented as ratios to U6B and RNU44 expression (Fig. 1D). The results showed that miR30a expression in the HCs of AIH mice was signi cantly lower than that in the control mice, indicating that in ammation suppressed endogenous miR30a expression in AIH.
Overexpression and downregulation of miR30a regulated hepatic in ammation in AIH We established the overexpression and downregulation of miR30a in AIH mice. After pretreatment with miR30a agomir followed by Con A induction, we performed qRT-PCR to con rm the overexpression of miR30a. As shown in Fig. 2A, the miR30a agomir increased the expression of miR30a to a substantially high level (Con A vs. agomir + Con A, ***p <0.001), whereas the miR30a agomir negative control group showed a similar miR30a expression level (Con A vs. agomir NC + Con A, not signi cant [NS]). We then established the blockage of endogenous miR30a expression in the HCs of AIH mice. After transfecting miR30a antagomir followed by Con A induction, the miR30a expression level dramatically decreased compared with that of control group (Con A vs. antagomir + Con A, **p <0.01), whereas the negative control group showed similar miR30a expression levels (Con A vs. antagomir NC + Con A, NS). These data con rmed that HCs were successfully modulated with miR30a agomir and antagomir.
We then tested the in uence of overexpression or downregulation of miR30a on hepatic in ammation in the AIH model. As shown in Fig. 2B, serum levels of ALT and AST were measured in the different groups. ALT and AST decreased back to a signi cantly low level in the agomir + Con A group (Con A vs. agomir + Con A, ****p < 0.0001), whereas they rose even higher when the endogenous expression of miR30a was further suppressed in the antagomir + Con A group (Con A vs. antagomir + Con A, **p < 0.01, *p < 0.05). In addition, the mRNA levels of proin ammatory cytokines (CD68, TNF-α, IL-1β, and IL-6) were signi cantly decreased in the miR30a agomir + Con A group, whereas they increased back to a high level in the miR30a antagomir + Con A group (Fig. 2C). The hepatic in ammation caused by Con A and the effects of miR30a agomir or antagomir treatment were further con rmed by the liver histology results (Fig. 2D).
These data suggest that overexpression of miR30a could alleviate HCs in ammation, whereas downregulation of endogenous miR30a expression aggravated HCs in ammation in AIH mice.
The interaction between miR30a and Klf14 in AIH TargetScan 7.0 predicted Klf14 to be a target of miR30a. Therefore, we tested the mRNA and protein levels of Klf14 in the HCs of AIH mice transfected with miR30a agomir or antagomir. The results in Fig.  3A show that Con A induction caused a signi cant increase in the expression of Klf14 mRNA (Ctr vs. Con A, **p <0.01). However, Klf14 expression signi cantly decreased back to a low level in miR30a overexpressing HCs (Con A vs. agomir + Con A, ***p < 0.001), whereas it was upregulated in the miR30a downregulated HCs (Con A vs. antagomir + Con A, *p <0.05). Similar results were also observed for Klf14 protein levels (Fig. 3B). The expression of miR30a was inversely correlated with Klf14 expression, indicating that miR30a might regulate in ammation in AIH by targeting Klf14. To validate this assumption, a luciferase reporter assay was performed to con rm the relationship between Klf14 and miR30a. The putative miR30a binding site in the 3′-UTR of Klf14 gene is shown in Fig. 3C, and D. The results showed that miR30a agomir signi cantly decreased the luciferase activity of Klf14 3′-UTR, but had little effect on the Klf14 3′-UTR mutant, which con rmed that Klf14 was a target of miR30a, and that miR30a expression was negatively correlated with Klf14 expression.

MiR30a regulates HCs in ammation by targeting Klf14 to modulate Tregs frequency
In order to examine the interactions among miR30a, Klf14 and Tregs, we sought to explore the changes in the frequency of Tregs in different groups by FACS. As shown in Fig. 4, after the administration of Con A, we observed a signi cantly decreased frequency of Tregs in liver MNCs (Ctr vs. Con A, ***p < 0.001). After overexpression of miR30a by agomir, frequency of Tregs increased back to a substantially high level (Con A vs. agomir + Con A, ***p < 0.001), with mRNA and protein levels of Klf14 decreasing back (see the above mentioned Fig. 3A, B). Conversely, in miR30a-downregulated AIH mice, reduced Tregs frequency (Con A vs. antagomir + Con A, *p < 0.05) was accompanied by increased Klf14 expression (see the above mentioned Fig. 3A, B). These results indicate that miR30a exerts its anti-in ammatory effect by targeting Klf14 to modulate Tregs in AIH.

Discussion
This study showed that the expression of miR30a was inhibited by in ammation in AIH mice, accompanied by changes in Klf14 expression and Tregs frequency. MiR30a plays an essential role in liver injury in AIH. Upon further regulation of miR30a expression in AIH mice by miR30a agomir and antagomir, expected trends were found in the changes in Klf14 expression and Tregs frequency. Interestingly, Klf14 was predicted and con rmed to be a direct target of miR30a. These results suggest that miR30a may play an essential role in the regulation of the in ammatory responses in AIH, at least in part, via Klf14 regulation of Tregs.
One previous study has already shown that Klf14 knockout (KO) could promote the differentiation of CD4 + cells to adaptive Tregs more readily and enhanced Tregs suppressor function both in vitro and in vivo via chromatin remodeling at the FOXP3 TSDR. In addition, KLF14 KO mice were resistant to experimentally induced colitis 13 . These results indicate that Klf14 serves as an important regulator of Tregs differentiation and biological function. In our study, we also observed an elevated Klf14 level and decreased Tregs frequency in AIH mice; meanwhile, when the Klf14 expression was downregulated by miR30a agomir, the ratio of Tregs was signi cantly rescued, which was consistent with the above study. However, another study concluded that transfection of Klf14 by recombinant adenoviral vector protected the liver by increasing the frequency of Tregs in AIH mice 10 . These contradictory ndings require further investigation. More importantly, the underlying molecular mechanism by which Klf4 in HCs in uences the frequency of Tregs remains to be elucidated.
MiR30a is widely known as a tumor suppressor that participates in tumorigenesis [19][20][21][22] . Although the role of miR30a in tumors has been well established, its role in AIH warrants further study. Geng et al. found that the herbal extract thymoquinone exerted an anti-brotic effect by upregulating miR30a expression, indicating the anti-brotic potential of miR30a in the liver 23 . Here, we aimed to explore the anti-liver in ammation potential of miR30a transfection in AIH, which is considered to be a liver disease relevant to autoimmunity and in ammation. We found that in ammation inhibited miR30a expression in AIH mice, and overexpression or downregulation of miR30a could in uence the in ammatory responses. These effects of miR30a were accompanied by the expected changes in Klf14 expression and Tregs frequency.
Based on these results, we inferred that miR30a might exert its protective effects on AIH by targeting Klf14 and increasing the number of Tregs. Together, our ndings provide new insights into the mechanisms of AIH and indicate that miR30a is an effective therapeutic target for AIH or other Tregsrelated liver diseases.
Notably, miR30a does not always play a protective role against diseases, indicating that this role may be disease-speci c. A meta-analysis showed that an increase in miR30a expression was associated with a high incidence of contrast-induced nephropathy 24 . A recent study by Wang et al. showed that miR30a is associated with blood-brain barrier damage in acute cerebral ischemia 25 . The above evidence reminds us to pay attention to miR30a-related pathogenesis of different diseases. Both contrast-induced nephropathy and acute cerebral ischemia are ischemia diseases 26, 27 . In contrast, AIH is closely related to in amation and immune-mediated liver injury, and miR30a plays a protective role against in ammation and immune-mediated liver injury, as shown in our present study. Therefore, it is reasonable to assume that miR30a may represent as an ideal target for the the treatment of in ammation and immunemediated liver diseases.
As was shown in our analysis using TargetScan, miR30a can regulate hundreds of downstream genes in mice. This raises the question: are there any other downstream genes and pathways that are regulated by miR30a in AIH? Further, is there any discrepancy in the function of miR30a among species, as we only employed an AIH mouse model in the present study? The mechanisms of action of miR30a in AIH may be complicated. Further studies are needed to tailor miRNA function to achieve the desired therapeutic effect in a speci c disease.

Conclusions
Herein, we employed an AIH model to examine the potential of miR30a as a therapeutic tool to eliminate liver in ammation by modulating Klf14 to regulate Tregs. In conclusion, our data show that miR30a might exert its protective effect by regulating the downstream Klf14 gene, leading to the changed pattern of Tregs in AIH mice.

AIH animal model and miRNA transfection
Eight-week-old male C57BL/6 mice were obtained from Vital River (Beijing, China). AIH in mice was induced by intravenous injection of a single dose of freshly prepared Con A (15 mg/kg, Sigma-Aldrich Chemical Co., China). After Con A administration, blood samples were collected for measurement of plasma alanine transaminase (ALT) and aspartate aminotransferase (AST) levels using mouse AST ELISA kits (Biotron Diagnostics, Hemet, CA, USA), according to the manufacturer's instructions. MiR30a expression is regulated in vivo by miR30a agomir or antagomir. MiR30a agomir, antagomir, or their negative controls (all form RiboBio, Guangzhou, China) were injected via the tail vein in mice 72 h before Con A administration (n = 5 animals per group).
All experimental animal protocols followed the regulations for the Administration of Affairs Concerning Experimental Animals (AdminReg, China) and were approved by the Ethics Committee of Animal Experiments and monitored by the Department of Experimental Animals of the Third A liated Hospital of Sun Yat-Sen University LingNan Hospital. All the authors con rmed that the experiments complied with the ARRIVE guidelines.
TargetScan and luciferase reporter assay TargetScan 7.0 was used to predict the target genes interacting with miR30a. A list of candidate miR30a targets was obtained. Among them, Klf14 appeared to be a promising target, as it was implicated in the pathology of in ammation in the liver.
To con rm the relationship between miR30a and Klf14, fragments of the 3′-UTR (Wt) containing the binding site of miR30a, or a 3′-UTR mutant (Mut) of Klf14 were cloned into pMIRREPORTTM luciferase vectors (Huayueyang, Beijing, China). HCs were co-transfected with the luciferase reporter vector, Renilla luciferase control vector (pRL-hTK), and the miR30a agomir or negative control using LipofectamineTM 2000 (Invitrogen, Carlsbad, CA, USA). Luciferase assays were performed 48 h after transfection using the dual-luciferase reporter assay system (Promega, San Luis Obispo, CA, USA) according to the manufacturer's protocol. Firefly luciferase activity was normalized to Renilla luciferase activity.

Cell isolation and culture
Primary mouse HCs and liver mononuclear cells (MNCs) were isolated as previously described 11 . Brie y, mouse livers were perfused with liberase TM solution (Sigma-Aldrich, St. Louis, MO, USA) after euthanasia, ltered through a 70 μm nylon cell strainer (BD Falcon, Franklin Lakes, NJ, USA), and centrifuged at 20× g for 5 min. The pellets were suspended and placed on the surface of 30% Percoll solution, centrifuged at 1000 × g for 10 min at 4 °C and washed once with phosphate-buffered saline (PBS). HCs were used for mRNA/miRNA isolation or protein extraction. Supernatants containing MNCs were collected, resuspended in 30% Percoll, and gently overlaid onto 70% Percoll. After centrifugation at 1000 × g for 30 min, liver MNCs were harvested from the interphase, washed twice with PBS, and then resuspended for further uorescence-activated cell sorting (FACS) analysis.

RNA isolation and quantitative reverse transcriptionpolymerase chain reaction
Total miRNAs from HCs were isolated using the mirVana™ miRNA isolation kit (Ambion, Austin, TX, USA) following the manufacturer's instructions. RNA was isolated from liver tissue after exosome treatment using TRIzol reagent (Life Technologies, MD, USA) and digested with DNase I (Fermentas, Glen Burnie, MD, USA) according to the manufacturer's protocol. Subsequently, total RNA was reverse-transcribed to cDNA using a RevertAid First Stand cDNA Kit (Thermo Fisher Scienti c, Rutherford, NJ, USA). The expression levels of miR30a, Klf14, and in ammatory cytokines were quanti ed by quantitative reverse transcription-polymerase chain reaction (qRT-PCR) using a 7500 Fast Real-Time PCR System (Applied Biosystems, Frederick, MD, USA) and were normalized to the expression of RNU6B (U6B), RNU44, or GAPDH. All samples were prepared and analyzed three times individually. The PCR primers used for the genes of interest are listed in Supplementary Table S1 online.

Western blotting
Total protein was extracted from HCs using Radioimmunoprecipitation buffer. Western blot analysis was performed as previously described 28 . Proteins were loaded into gels, subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene di uoride membranes. The membranes were incubated with anti-Klf14 polyclonal antibody (rabbit anti-mouse, 1:1000, PA5-23784, Invitrogen, Carlsbad, CA, USA), anti-β-actin Polyclonal Antibody (Rabbit anti mouse, 1:4000, #4967, CST, Beverly, MA, USA) with gentle agitation overnight at 4 ˚C. The membranes were washed three times for 5 min each with TBST (containing 0.1% Tween-20) and incubated with horseradish-peroxidaseconjugated anti-Rabbit IgG (Invitrogen) followed by chemiluminescence detection using Pierce TM ECL western blotting substrate (Thermo Fisher Scienti c, Rutherford, NJ, USA) according to the manufacturer's protocol. The membranes were subsequently analyzed using "Quantity One" software (Bio-Rad Laboratories, Hercules, CA, USA).

Histopathology
After sacri ce, the livers of individual mice were perfused with ice-cold PBS to remove blood components.
Livers were cut into 2 × 4 × 4 mm 3 sections, xed in 4% paraformldehyde, and embedded in para n. Subsequently, 5 μm slices were then cut at various depths in the tissue sections, stained with hematoxylin-eosin (H&E) and examined under light microscopy to determine the level of in ammation.

Flow cytometric analysis
Flow cytometry was conducted using a BD FACScalibur device (BD Biosciences, USA) and analyzed with FCS express V3. After washing with Hank's buffer devoid of Ca 2+ and Mg 2+ , 5 × 10 5 liver MNCs were blocked using in 1% bovine serum albumin at 4 °C for 30 min. Unfractionated cells were stained with allophycocyanin-conjugated anti-CD4 and PE-conjugated anti-CD25 (eBioscience, San Diego, CA, USA) monoclonal antibodies. Cells were incubated at 4 °C in the dark for 30 min, washed with PBS/1% fetal bovine serum, resuspended, and analyzed by ow cytometry using a Becton Dickinson uorescent activated cell sorter (FACSCantoII, Becton Dickinson Immunocytochemistry Systems, San Jose, CA, USA).
FACS-Diva software was used for analysis. A minimum of 2 × 10 4 gated events was acquired for each sample.
Statistical analysis SPSS software (version 20.0) was used for all statistical analyses. Data are expressed as the mean ± standard error. All results presented represent data collected from at least three independent experiments. Results were analyzed by using the Student's t-test. Statistical signi cance was set at p < 0.05.

Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Author contributions X Yuan made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data. S Pan and M Li had been involved in drafting the manuscript or revising it critically for important intellectual content. K Zhang agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work were appropriately investigated and resolved. X Bi had given nal approval of the version to be published. All authors read and approved the nal manuscript. Figure 1 Establishment of a mouse autoimmune hepatitis (AIH) model and endogenous miR30a expression in AIH mice. A. Serum levels of alanine transaminase (ALT) and aspartate aminotransferase (AST) were measured. Results are shown as mean ± SD (***p <0.001, ****p < 0.0001). B. Liver tissue was stained with hematoxylin and eosin (H&E). Signi cant disruptions of liver structure and lymphocyte in ltration in the liver were observed in the Con A group. C. mRNA levels of proin ammatory cytokines (CD68, TNF-α, IL-1β, and IL-6) increased signi cantly after Con A induction (**p < 0.01, ***p <0.001). D. Fold changes in miR30a expression as ratios to U6B (A) or RNU44 (B) expression were detected (**p < 0.01, ***p <0.001). (Ctr group: control group; Con A group: Con A-induced group) Figure 2 Effects of miR30a on the in ammation of HCs in AIH mice. Cells were transfected with miR30a agomir/antagomir and their negative controls followed by Con A induction. A. MiR30a mRNA levels as ratios to U6B levels were assessed. (Con A vs. agomir + Con A, ***p < 0.001; Con A vs. antagomir + Con A,