Human Umbilical Cord Mesenchymal Stem Cells Ameliorate Hepatic Stellate Cell Activation and Liver Fibrosis by Upregulating MicroRNA-455-3p through Suppression of p21-Activated Kinase-2

Mesenchymal stem cells (MSCs) were shown to have potential therapeutic effects for treatment of liver fibrosis, and dysregulated expression of microRNAs (miRNAs) played a pivotal role in the pathogenesis of liver fibrosis by regulating their downstream target genes. However, the mechanism by which MSCs affect the progression of liver fibrosis by regulating miRNA expression remains unclear. Here, we investigated whether human umbilical cord MSCs (HUC-MSCs) attenuated hepatic fibrosis by regulating miR-455-3p and its target gene. Significantly upregulated miRNA (miR-455-3p) was screened out by GEO datasets analysis and coculture HUC-MSCs with hepatic stellate cell (HSC) LX-2 cells. p21-activated kinase-2 (PAK2) was forecasted to be the target gene of miR-455-3p by bioinformatics analyses and confirmed by luciferase reporter assay. HUC-MSCs were transplanted into mice with carbon tetrachloride- (CCl4-) induced liver fibrosis, the result showed that HUC-MSC transplantation significantly ameliorated the severity of CCl4-induced liver fibrosis, attenuated collagen deposition, improved liver function by reducing the expression of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in serum, upregulated miR-455-3p, and suppressed PAK2 expression of liver tissue in mice. Taken together, our study suggests that HUC-MSCs inhibit the activation of HSCs and mouse CCl4-induced liver fibrosis by upregulation of miR-455-3p through targeting PAK2.


Introduction
Liver fibrosis and cirrhosis are common pathological processes of many chronic liver diseases, including chronic hepatitis B and C, alcoholic steatohepatitis, nonalcoholic steatohepatitis, and autoimmune hepatitis, which can eventually lead to liver failure or liver cancer. Hepatic stellate cells (HSCs) are widely accepted to be the fibrogenic cell type in the liver [1]. The characteristic of fibrosis is accompanied by the activation and proliferation of HSCs and increased production of excessive extracellular matrix [2,3]. Currently, there are still few strategies to reverse liver fibrosis except orthotopic liver transplantation, but organ donor shortage, complications of surgery, and immunosuppression limit its application [4]. Therefore, there is an urgent need to find a new alternative treatment strategy.
Mesenchymal stem cells (MSCs) are multipotent stem cells, which can be isolated from bone marrow, adipose tissue, umbilical cord, and other tissues [5,6]. There have been demonstrated that MSCs have an inflammatory tendency and can homing to the injured site to play an anti-inflammatory role [7,8]. Recently, accumulating evidences suggest that exogenous MSC treatment could ameliorate HSC activation and liver fibrosis, differentiate into hepatic cells, and promote the recovery of liver function [9,10].
MicroRNAs (miRNAs), a class of endogenous noncoding RNAs, approximately 22 nucleotides long, function as nega-tively regulate protein translation by binding to the 3 ′ untranslated region (3′ UTR) of mRNAs of their target gene [11]. Recent studies have shown that miRNAs regulate the activation of HSCs and the progress and development of liver fibrosis. For example, miR-125b promotes activation of HSCs and liver fibrosis by activating the ras homolog gene family, member A (RhoA) signaling [12], miR-26b-5p inhibits liver fibrogenesis by targeting platelet derived growth factor receptor-β (PDGFR-β) [13], and miR-455-3p alleviates HSC activation and prevents liver fibrosis by suppressing heat shock factor 1 (HSF1) expression [14]. However, the underlying mechanisms of MSCs alleviating HSC activation by regulating miRNA expression remains unknown.
In this study, we used two miRNA profile data (GSE33857 and GSE77271) about liver fibrosis which were downloaded from publicly available GEO datasets and screened out five common differentially expressed miRNAs (DEMs) (miR-342-3p, miR-125a-5p, miR-455-3p, miR-574-5p, and miR-193a-3p) which were upregulated or downregulated in experimental animals or human liver fibrosis samples. The result was verified by quantitative real-time reverse transcription-polymerase chain reaction (qRT-PCR) in activated HSCs induced by transforming growth factor β1 (TGFβ1). Subsequently, we cocultured activated HSCs with human umbilical cord mesenchymal stem cells (HUC-MSCs) and found that miR-455-3p expression was the most significant difference compared with the HSC group cultured alone. These findings were subsequently verified in mouse models of liver fibrosis treated with HUC-MSCs. Finally, our results demonstrated that HUC-MSC treatment ameliorated liver fibrosis by upregulating miR-455-3p through suppression of p21-activated kinase-2 (PAK2).

Materials and Methods
2.1. Isolation, Culture, and Expansion of HUC-MSCs. HUC-MSCs were isolated as described previously [15]. Umbilical cords were donated by mothers who provided written informed consent for research purposes. The collection and isolation procedure were both approved by the Ethics Committee of Soochow University (Suzhou, China). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA), 1 mmol/L GlutaMAX™ (Gibco), and 100 IU/mL penicillin/streptomycin (Gibco) at 37°C in a 5% CO 2 -humidified atmosphere with medium changes every 2-3 days and were subcultured by using 0.25% trypsin-EDTA (Thermo Fisher Scientific, Waltham, MA, USA) upon reaching 80%-90% confluence. The ability to differentiate into osteoblasts and adipocytes and the surface antigen profile of HUC-MSCs were confirmed as previously reported [15,16] 2.3. HUC-MSCs/LX-2 Cocultures. Fully activated LX-2 cells were cultured alone or cocultured with HUC-MSCs using transwell inserts (24 mm diameter, 0.4-μm pore size, Corning, NY, USA) to assess the effects of HUC-MSCs on LX-2 cells. All coculture experiments were performed with LX-2 cells in the lower chamber and HUC-MSCs in the upper compartment using DMEM medium with 10% FBS. LX-2 cells were collected after culturing for 48 hours.

Cell Proliferation
Assays. The proliferation of LX-2 cells was measured using a Cell Counting Kit-8 (CCK8) assay (Dojindo Molecular Technologies, Inc., Kumamoto, Japan). In brief, LX-2 cells were seeded in the lower chamber of 6well plate using transwell inserts at a density of 5 × 10 4 cells per well, while in the coculture group the same number of HUC-MSCs was seeded in the upper compartment. After culturing for 48 hours, the upper department was removed, and the CCK8 reagent was added to the lower chamber containing LX-2 cells and incubated for additional 2 hours according to the manufacturer's protocol. Optical densities were measured using a microplate reader (SpectraMax M5; Molecular Devices, Sunnyvale, CA, USA) at a wavelength of 450 nm.
2.6. Luciferase Reporter Assay. The 3 ′ UTR sequence of human PAK2 that contained the putative wild-type (WT) or mutant (MUT) miR-455-3p binding site was inserted into the pGL3 luciferase vector (Promega, Madison, WI, USA). For the luciferase reporter experiments, LX-2 cells were cotransfected with the above constructed 3 ′ UTR luciferase vector and miR-455-3p mimics or NC using Lipofectamine™ 3000 reagent. Cells were harvested 48 hours after transfection for dual-luciferase reporter assay by a fluorescence detector (Promega). The relative firefly luciferase activity was normalized to Renilla luciferase activity.

Animal Models.
Animal experiments were performed according to the guidelines of Institutional Animal Care and Use Committee of Soochow University (Suzhou, China). Male C57BL/6 mice of 8-10 weeks age were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). Liver fibrosis mice were induced by carbon tetrachloride (CCl 4 ) (20% in olive oil, 5 μL/g) intraperitoneal administration twice a week for 12 weeks, while mice in the control group were treated with olive oil. Each experimental group contained 6 mice. Mice were divided into three groups: control group, CCl 4 group, and CCl 4 +HUC-MSC group. CCl 4 +HUC-MSC group mice were injected 5 × 10 5 HUC-MSCs in 0.3 mL phosphate-buffered saline (PBS) by the tail vein once a week 2 BioMed Research International at the 6th week after CCl 4 administration for six times, while mice in the control and CCl 4 groups were injected with only 0.3 mL PBS during the same period. At the end of the 12th week, venous blood was obtained from the retroorbital vein to measure liver serum markers. Then, all mice were sacrificed and liver lobes were obtained for further analyses.
2.8. Histology. Liver tissues were fixed in 4% paraformaldehyde and embedded in paraffin. Sections were cut at 5 μm thickness and stained with hematoxylin-eosin (H&E). Collagen deposition of sections was detected by Masson's trichrome staining and Sirius red staining using standard protocols.
2.9. Immunofluorescence and Immunohistochemistry. Immunofluorescence staining was performed to assess α-smooth muscle actin (α-SMA) and collagen 1α1 (Col1α1) on LX-2 cells as routine protocols. Briefly, LX-2 cells were seeded into 6-well plates and divided into three groups: LX-2 group, LX-2 cells cultured alone; LX-2+TGFβ1 group, LX-2 cells cultured alone and treated with TGFβ1 for activation; and LX-2 +HUC-MSCs group, LX-2 cells treated with TGFβ1 and then co-cultured with HUC-MSCs. After being cultured, the cells were washed twice with PBS and then fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 in PBS for 20 minutes at room temperature. After being washed with PBS, the cells were incubated with 10% goat serum in PBS for blocking nonspecific binding sites and then incubated overnight with primary antibodies specific to α-SMA After being incubated with DAPI (Beyotime, Shanghai, China) for staining nuclei, the images of cells were acquired using a fluorescence microscope (OlympusBX-53, Tokyo, Japan). For immunofluorescence staining and immunohistochemical analysis of liver tissues, paraffin-embedded tissue sections were dewaxed in xylene and dehydrated in a series of graded alcohols; heat-induced antigen retrieval was accomplished by microwaving in citric saline for 15 minutes. The sections were further blocked in PBS plus 0.025% Tween 20 with 2% bovine serum albumin. For immunofluorescence staining, the sections were incubated with primary antibodies specific for α-SMA (1 : 200; Abcam) and PAK2 (1 : 200; Abcam), then were further incubated with Dylight 649conjugated and Dylight 488-conjugated secondary antibody; finally, the samples were incubated with DAPI (Beyotime). For immunohistochemical analysis, after blocking, the sections were incubated with primary antibodies specific for α-SMA (1 : 200; Abcam), Col1α1 (1 : 200; Abcam), and PAK2 (1 : 100; Abcam) overnight at 4°C. After washing with PBS, the sections were incubated with horseradish peroxidase-(HRP-) conjugated secondary antibody for 1 hour at room temperature. Subsequently, the signal was visualized by 3,3 ′ -diaminobenzidine tetrachloride, and the nuclei were counterstained with hematoxylin.
2.10. Hepatic Hydroxyproline Content. The hepatic hydroxyproline level was measured using an assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China), according to the manufacturer's instructions.
2.11. Quantitative Real-Time PCR. Isolation of total RNA from cultured LX-2 cells or liver tissues was conducted using Trizol reagents (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized using the PrimeScript RT reagent kit (Takara, Shanghai, China). qRT-PCR was performed in triplicate using the SYBR Premix Ex Taq (Takara) on the ABI ViiA 7Dx real-time PCR system (Life Technologies, NY, USA). The expression levels of mRNA of interest and microRNA were calculated according to the comparative threshold cycle value (2 −ΔΔCt ) method and were normalized against β-actin or U6. All of the primers were synthesized by GENEWIZ (Suzhou, China). The primer sequences used in this study were as follows: α-SMA (human), forward: 2.12. Western Blot Analysis. Cells and liver tissues were lysed in RIPA buffer (Millipore, Billerica, MA, USA), and total proteins were extracted. The protein concentration was measured using a BCA protein assay kit (Beyotime, Nanjing, China). Proteins used for Western blot analysis were performed routinely; membranes were incubated with primary antibodies against α-SMA (1 : 1000; Abcam), Col1α1 (1 : 2000; Abcam), GAPDH (1 : 2500; Abcam), and PAK2 (1 : 2000; Abcam) overnight at 4°C. Then, HRP-conjugated goat anti-rabbit IgG (1 : 1000; Cell Signaling Technology, Beverly, MA, USA) secondary antibody was employed at 37°C for 1 hour. The immunobands were visualized using the chemiluminescence HRP substrate (Millipore). The protein expression was standardized to GAPDH and quantified using Gel-Pro Analyzer software (Media Cybernetics, Rockville, MD, USA).

BioMed Research International
2.13. Statistical Analysis. The statistical analyses were performed by using GraphPad Prism 6 (GraphPad Software Inc., San Diego, CA, USA). Experimental data are expressed as the mean ± SD. Statistical significance was determined using the two-tailed Student's t-test for the comparison of two groups. For comparison among multiple groups, oneway analysis of variance (ANOVA) was performed. A value of P < 0:05 was considered statistically different.

HUC-MSCs Inhibit Proliferation and Activation of HSCs.
To ascertain whether HUC-MSC transplantation affects the proliferation of HSCs, human HSC cell line LX-2 cells were cocultured with HUC-MSCs using a coculture transwell chamber. CCK-8 assays were performed to measure the cell proliferation of LX-2 cells. For activation, LX-2 cells were incubated with TGFβ1. The results showed that TGFβ1 treatment significantly promoted LX-2 cell proliferation, while HUC-MSCs notably reversed this effect (Figure 1(a)). To investigate whether HUC-MSCs could influence HSC activation, LX-2 cells were cultured in the absence of serum for starvation and activated with TGFβ1. The expression of α-SMA and Col1α1 (two markers of activated HSCs) mRNA and protein in TGFβ1 treatment group was highly increased, as determined by qRT-PCR, Western blot, and immunostaining analyses compared with LX-2 cultured in the absence of TGFβ1, which indicated successful activation of LX-2 cells (Figures 1(b)-1(f)). Meanwhile, coculture with HUC-MSCs for 48 hours led to an obvious downregulation of α-SMA and Col1α1 mRNA expression, as determined by qRT-PCR (Figure 1(b)), and the inhibition effect of HUC-MSCs on HSC activation was in a dose-dependent manner ( Supplementary Figure 1(a) and 1(b)). The expression of α-SMA and Col1α1 was also downregulated confirmed by repeated Western blot and immunostaining analysis, suggesting that TGFβ1-mediated activation of LX-2 cells was inhibited by HUC-MSCs (Figures 1(c)-1(f)).

HUC-MSCs Promote Inactivation of LX-2 Cells by
Upregulating miR-455-3p. Increasing evidence demonstrated that miRNAs regulate a wide range of cellular processes, including cell proliferation, differentiation, and apoptosis. In addition, aberrant miRNA expression may associate with the development of multiple diseases, including liver fibrosis.

miR-455-3p Regulates PAK2 Expression by Directly
Targeting the 3 ′ UTR of Its mRNA. We then investigated the underlying mechanism by which miR-455-3p regulates HSC activation and to identify the downstream target genes of miR-455-3p, and we performed bioinformatics analyses using Microt4, miRanda, miRDB, RNA22, and TargetScan databases and focused on PAK2, an activated receptor of TGFβ. The 3′ UTR of PAK2 mRNA contained a complementary site for the seed region of miR-455-3p (Figure 3(a)). Moreover, PAK2 has been reported to be a downstream target gene of miR-455-3p in human osteoblasts and in human colon adenocarcinoma cells [17,18]. However, whether miR-455-3p regulates PAK2 expression in HSCs has not been confirmed yet. Therefore, to further demonstrate whether PAK2 was a direct target of miR-455-3p and mediated the suppressive effect of miR-455-3p on the profibrotic phenotype, the two putative miR-455-3p target sites (wild-type (WT) or mutant (MUT) 3 ′ UTR target sequences) were cloned into a luciferase reporter vector. Luciferase activity of WT 3 ′ UTR of PAK2 was significantly inhibited in LX-2 cells transfected with miR-455-3p (Figure 3(b)); no change of luciferase activity was observed after the cotransfection of miR-455-3p with MUT 3 ′ UTR of PAK2 (Figure 3(b)). To further investigate whether miR-455-3p is able to successfully modulate PAK2 expression in HSCs, LX2 cells were transfected with miR-455-3p mimics or NC, we found that overexpression of miR-455-3p strongly attenuated the expression of PAK2 protein compared with the controls (Figures 3(c)  and 3(d)). These findings suggested that miR-455-3p may repress the expression of PAK2 by directly targeting the 3′ UTR of PAK2.

HUC-MSCs Attenuate the Severity of Liver Fibrosis in
Mice. To further investigate the role of HUC-MSCs in hepatic fibrosis in vivo, HUC-MSCs were injected to mice once a week after 6 weeks of CCl 4 administration during the 12 weeks of CCl 4 -induced fibrogenesis. At the end of the 12th week, we determined the degree of fibrosis in liver tissue by histology, Masson's trichrome, and Sirius red staining. We observed less liver damage and fewer fibrotic areas in 4 BioMed Research International  These results suggested that HUC-MSC injection could attenuate mouse liver fibrosis in this model.

HUC-MSCs Upregulating the Expression of miR-455-3p of CCl 4 -Induced Liver Fibrosis in Mice by Suppressing PAK2
Expression. To verify whether miR-455-3p is involved in the inhibitory effect of HUC-MSCs on liver fibrosis, we determined miR-455-3p levels in a mouse model of liver fibrosis, which was induced by injecting C57BL/6 mice with CCl 4 for 12 weeks. The increased levels of miR-455-3p in fibrotic liver tissues of mice treated by HUC-MSCs compared with CCl 4 group were confirmed by qRT-PCR analysis ( Figure 5(a)). To examine the efficacy of miR-455-3p at the functional level, we analyzed the mRNA and protein level of PAK2, a target gene of miR-455-3p, which has been confirmed before. As shown in Figure 5(b), a significantly lower expression level of PAK2 was found in CCl 4 -treated mice injected with HUC-MSCs compared with PBS injection; we also found a significantly decreased PAK2 at protein level inspected by Western blotting and immunohistochemical analysis compared with the CCl 4 group (Figures 5(c)-5(f)). To determine whether PAK2 was expressed in activated HSCs, we double-stained the liver tissue slices with α-SMA and PAK2 by immunofluorescence staining, and it could be seen that α-SMA and PAK2 were coexpressed in the space of Disse of the liver (the yellow areas indicated by the arrow) ( Figure 5(g)), and the areas coexpression α-SMA and PAK2 in the HUC-MSC treatment group were significantly smaller than those in the CCl 4 group (Figure 5(h)). It has been reported that PAK2 is highly expressed in activated HSCs [19], the immunohistochemical staining showed that PAK2 was positively expressed in hepatic portal area and sinuses of hepatic stromal cells ( Figure 5(c), combined with immunofluorescence staining, and the results may indicate that PAK2 was positively expressed in activated HSCs in liver fibrosis. In conclusion, our findings revealed that HUC-MSCs alleviated hepatic fibrosis through upregulating miR-455-3p by targeting PAK2.

Discussion
Liver fibrosis is the end result of the most kinds of chronic liver damage and eventually will develop into liver cirrhosis, liver failure, and even liver cancer [20]. Activation of HSCs was believed to be the crucial component of this process [21]. However, there have been no effective therapeutic strategies for liver fibrosis up to now. Recently, MSCs have been shown to inhibit the activation of HSCs in vitro and improve liver fibrosis in animal models and in clinical trials [22][23][24]. However, the underlying mechanisms have not yet been fully elucidated.
Recent studies have focused on miRNA mechanisms in the pathophysiology of hepatic fibrosis and the activation of HSCs. miR-455-3p is generally regarded as an oncogene or tumor suppressor in different types of tumors [25,26]. Recently, miR-455-3p has been demonstrated participating in fibrosis. For instance, upregulation of miR-455-3p could suppress idiopathic pulmonary fibrosis through repression of Bax expression [27]. miR-455-3p overexpression suppressed renal fibrosis through repressing Rho-associated protein kinase 2 (ROCK2) expression [28]. And miR-455-3p can inhibit liver fibrosis by alleviating HSC activation through suppressing heat shock transcription factor 1 (HSF1) expression [14]. A recent study showed that miR-455-3p derived from HUC-MSC exosomes has a promising treatment for acute liver injury [29]. However, it remains unclear whether miR-455-3p was involved in the mechanism of MSCs regulating liver fibrosis.
In the current study, we found that miR-455-3p was significantly upregulated in activated HSCs co-cultured with HUC-MSCs in vitro. In vivo, we found that miR-455-3p was upregulated in CCl 4 -induced hepatic fibrosis models treated by HUC-MSCs, which was in accordance with the results in vitro. Our findings suggested that miR-455-3p might play an important role in the MSC treatment of liver fibrogenesis, and its upregulation might be associated with the alleviation of hepatic fibrosis. However, how HUC-MSCs inhibit the activation of HSCs by regulating the expression of miR-455-3p is a question we need for a further study. According to the research by Shao et al. [29], IL-6 stimulation of HUC-MSCs can significantly upregulate the expression of miR-455-3p in its exosomes, while the exosomes with high expression of miR-455-3p can significantly reduce lipopolysaccharide-(LPS-) mediated liver injury in mice. We speculate that HUC-MSCs may inhibit the activation of HSCs and liver fibrosis by secreting exosomes rich in miR-455-3p.
To elucidate the target gene of miR-455-3p, we predicted it by bioinformatics analysis and confirmed that PAK2 was the target gene of miR-455-3p by luciferase reporter assay. PAK2 is one of the P21-activated kinases (PAKs) of the serine/threonine kinases family [30], which is a critical effector of the Rho family of small GTPases linked to cytoskeleton reorganization [31]. PAK2 is widely distributed in the whole body and regulates a variety of biological behaviors, including promoting tumorigenesis, cellular senescence, and organismal aging [32,33]. TGFβ has long emerged as a prominent master regulator of fibrogenesis [34]. The cellular responses evoked by TGFβ were mediated by canonical SMAD signaling and noncanonical (non-SMAD) signaling, including mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K)/protein-serine-threonine kinase (Akt) signaling [35,36]. As a downstream component of Akt, PAK2 has been reported as an activated receptor of TGFβ in mesenchymal cells; PAK2 activation is essential for the  10 BioMed Research International proliferative/profibrotic action of TGFβ [37]. PAK2 was also reported as a non-SMAD effector of TGFβ in renal interstitial fibrogenesis; overexpression of PAK2 promotes renal fibrogenesis [38]. PAK2 was strongly expressed in activated HSCs [19], and it was highly likely that PAK2 positively modulates TGFβ/Akt signaling in HSCs. In our study, we found that HUC-MSC treatment could inhibit the expression of profibrotic markers, upregulate miR-455-3p, and downregulate PAK2 in activated HSCs, and the effect could be enhanced by miR-455-3p mimics and attenuated by miR-455-3p inhibitors. In vivo, we also found that HUC-MSCs could upregulate miR-455-3p and inhibit PAK2 in a mouse hepatic fibrosis model. However, it has been reported that miR-455-3p can promote TGFβ/SMAD signaling in chondrocytes and inhibit the development of osteoarthritis by directly targeting PAK2 [17]. Therefore, it might be that PAK2 positively modulates TGFβ/Akt (non-SMAD) signaling in mesenchymal cells, while negatively regulating TGFβ/SMAD signaling by attenuating the receptor-SMAD interaction in epithelial cells [39].

Conclusion
Our study demonstrated that HUC-MSCs significantly alleviate HSC activation and CCl 4 -induced liver fibrosis in a mouse model; overexpression of miR-455-3p by targeting PAK2 might be the underlying mechanism for the therapeutic effects of HUC-MSCs in hepatic fibrosis.

Data Availability
The datasets generated and/or analyzed during the present study are available from the corresponding author upon reasonable request.

Conflicts of Interest
The authors declare that they have no conflicts of interest.

Authors' Contributions
ZQ performed the study and wrote the manuscript. GTF established the animal models and carried out the histopathological studies. ZY performed the study, carried out the data analysis, and prepared the figures. LHD and ZSXM carried out cell culture and did the molecular studies. KYT and XSR conceived the ideas, supervised the study, and revised the manuscript. Qing Zhou, Tengfei Gu, and Yong Zhang contributed equally to this work and were co-first authors.