Fibrotic liver tissues (n=24) were obtained from patients clinically and pathologically diagnosed with liver fibrosis. Normal liver tissues (n=20) were obtained from the surgical resection of liver metastases. All samples were collected at the Affiliated Hospital of Shandong Medical College. The characteristics of the patients and healthy donors are listed in Table I. The protocol was approved by the Investigation and Ethics Committee of the Affiliated Hospital of Shandong Medical College. Each patient provided written informed consent. The collected tissue samples were divided into two portions; one was preserved in liquid nitrogen for the preparation of total RNA and total protein lysates, and the other was used for the isolation and culture of HSCs.

Primary human HSCs (pHSCs) were isolated from fragments of normal liver tissue obtained from the surgical resection of liver metastases, as described in detail elsewhere (21). pHSCs were cultured for three days, and then displayed a quiescent phenotype (qHSCs). After being cultured for 10 days, pHSCs transformed into activated cell types (aHSCs) (22). The obtained HSCs were cultured in RPMI medium supplemented with 20% fetal bovine serum (FBS), nonessential amino acids, 1.0 mM sodium pyruvate and antibiotic-antimycotic (all from Life Technologies, Carlsbad, CA, USA), as referred to a previous study (23).

LX-2 cells (human activated HSCs) were cultured with Dulbeccos modified Eagles medium (DMEM) (HyClone, Logan, UT, USA) containing 10% newborn calf serum, 100 U/ml penicillin and 100 mg/ml streptomycin at 37°C in a humidified atmosphere containing 5% CO₂.

Total RNA was extracted from normal liver and liver cirrhosis tissues as well as qHSCs and aHSCs with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). For mRNA quantification, 500 ng of RNA was used for the synthesis of cDNA with reverse transcriptase using the M-MLV First Strand kit (Taraka, Dalian, China) according to the manufacturer's instructions. cDNA (1 µl) was used for real-time PCR with GoTaq qPCR Master Mix (Promega, Madison, WI, USA). For each sample, the relative mRNA level was normalized to β-actin expression. The primers for qPCR are listed in Table II.

For miRNA quantification, the GoScript Reverse Transcription System kit (Promega) was used with the stem loop primer. For each sample, the relative mRNA level was normalized to U6. The specific forward primer for miR-9 was 5′-TCTTTGGTTATCTAGCTGTATGA-3′. Relative expression levels of miRNA or mRNA were analyzed using the Bio-Rad C1000 Thermal Cycler (Bio-Rad, Hercules, CA, USA).

Western blotting was performed as previously described (24). The primary antibodies used were rat anti-mouse MRP1, α-SMA, Col-1, Timp-1, Smo, Ptc, Gli2 and Gli3 (Abcam, Cambridge, UK). Secondary antibodies included horseradish peroxidase-conjugated goat anti-rat IgG (H+L) (Pierce, Rockford, IL, USA). The Western-Light chemiluminescent detection system (Image Station 4000 MM PRO, XLS180; Kodak, Rochester, NY, USA) was used to visualize the signals. The gray values were analyzed with BandScan5.0 software.

LX-2 cells were seeded 24 h prior to transfection into 24-well or 6-well plates or 6-cm dishes. LX-2 cells were transfected with an miRNA mimic control, miR-9 mimic, miRNA inhibitor control or miR-9 inhibitor (Applied Biosystems, Foster City, CA, USA) using Lipofectamine 2000 (Invitrogen) at a final concentration of 80 nM. The cells were harvested at 24 h (for RNA extraction) and at 48 h (for protein extraction).

For the expression plasmid construct, the wild-type MRP1 cDNA sequence without the 3′-UTR was selected and cloned into the pcDNA™3.2-DEST vector (Invitrogen).

Cell proliferation was measured using the CCK-8 kit (Solarbio, Beijing, China). In brief, 48 h after transfection with the miR-9 mimic or miR-9 inhibitor, LX-2 cells were seeded at 5×10³ cells/well in 96-well plates in triplicate. At 0, 24, 48 and 72 h, 10 µl of CCK-8 solution mixed with 90 µl of DMEM was added to each well. After 2 h of incubation, absorbance was measured at 450 nm.

The human MRP1 3′-UTR was amplified and cloned into the XbaI site of the pGL3-control vector (Promega), downstream of the luciferase gene, to generate the pGL3-MRP1-3′-UTR plasmid. As a control, the pGL3-MRP1-3′-UTR-mut plasmids were also constructed using cDNA fragments containing corresponding mutated nucleotides for miR-9. For the luciferase reporter assay, LX-2 cells were co-transfected with the luciferase reporter vectors and control and miR-9 mimics, control or miR-9 inhibitors, using Lipofectamine 2000. A β-actin promoter Renilla luciferase reporter was used for normalization. After 48 h, luciferase activity was analyzed using the Dual-Luciferase assay system (Promega), according to the manufacturer's protocols.

Male C57BL/6 mice (6-weeks-old) were purchased from the Experimental Animal Center of Shandong Province, housed with a 12-h light/dark cycle and allowed free access to normal food and water. To induce liver fibrosis, mice (n=10) received 0.6 ml/kg body weight CCl₄ (Sigma-Aldrich, St. Louis, MO, USA) by i.p. injection, which was dissolved in corn oil, twice a week for eight weeks (25). Control mice (n=10) were injected i.p. with the same amount of saline. All mice were sacrificed to obtain serum and liver samples at 48 h post the last injection of CCl₄ or corn oil. All experimental procedures were approved by the Institutional and Local Committee on the Care and Use of Animals of Shandong Medical College, and all animals received humane care according to the National Institutes of Health (Bethesda, MD, USA) guidelines.

To examine the effect of miR-9 in vivo, mice received 0.4 ml/kg body weight CCl₄ dissolved in corn oil by i.p. injection, three times a week for three weeks (26). Next, the mice were randomly divided into four experimental groups: treatment with 0.4 ml/kg body weight of CCl₄ twice a week in parallel with one i.p. injection of PBS (n=10, CCl₄ group); miR-9 mimic (n=10, miR-9 group), Ad-con (n=10, Ad-emp group) or Ad-MRP1 (n=10, miR-9 + MRP1 group)/week for three weeks.

The liver histological sections from the control, CCl₄ and miR-9-treated mice were examined by light microscopy after staining with hematoxylin and eosin (H&E).

All data are expressed as the mean ± SD of results derived from three independent experiments performed in triplicate. Statistical analysis was performed using Student's t-test and ANOVA. A difference was accepted as significant at p<0.05.

qRT-PCR and western blotting were used to determine the expression of miR-9 and MRP1 in fibrotic liver tissues and cultured human HSCs. The results showed that the mRNA and protein levels of MRP1 were markedly increased in the human fibrotic liver tissues (Fig. 1B and C). A microRNA database was used to screen miRNA candidates targeted to MRP1. As a candidate target miRNA of MRP1, miR-9 showed significantly decreased expression in the human fibrotic liver tissues as compared with the healthy controls (Fig. 1A). Furthermore, qRT-PCR and western blotting results showed that MRP1 was upregulated in aHSCs compared with the qHSCs (Fig. 1E and F). The decreased level of miR-9 was also demonstrated in aHSCs (Fig. 1D). Thus, a decrease in miR-9 or an increase in MRP1 in fibrotic liver tissues plays an important role in the development of liver fibrosis.

In light of the decreased expression of miR-9 in activated HSCs, we next investigated the effect of miR-9 on the proliferation and activation of HSC cell line LX-2. miR-9 expression was markedly upregulated by miR-9 mimics (Fig. 2A). As determined by CCK-8 and BrdU assays, the transfection of miR-9 led to an inhibition of cell proliferation in the LX-2 cells as compared to that noted in the controls (Fig. 2B and C). In addition, we transfected miR-9 mimics into HSCs to clarify the role of miR-9 overexpression in HSC activation and collagen deposition. As shown in Fig. 2D and E, overexpression of miR-9 significantly suppressed mRNA and protein expression of α-SMA, Col-1 and Timp-1, which are characterized as genes involved in the activation of HSCs. In contrast, the mRNA expression and protein level of these genes were reversed by the miR-9 inhibitor, suggesting that miR-9 may suppress HSC activation as well as ECM deposition.

Examination of the homology showed that 12 nucleotides in the seed region of miR-9 were complementary to MRP1 bases (Fig. 3A). To determine whether miR-9 binds directly to the predicated sites of the MRP1 3′-UTR, we performed luciferase reporter assays. miR-9 significantly reduced MRP1 3′-UTR-dependent luciferase activity, but did not affect mutant reporter luciferase activity, whereas the miR-9 inhibitor had no effect on wild-type or mutant reporter luciferase activity (Fig. 3B). Western blot results showed that the protein expression of MRP1 was decreased in the miR-9-transfected HSCs, but increased in the miR-9 inhibitor-transfected HSCs (Fig. 3C). These results suggested the interaction between miR-9 and the 3′-UTR of MRP1.

To investigate whether the downregulation of MRP1 by miR-9 affects the expression of Hh ligands and Hh-target genes, we analyzed Smo, Ptc, Gli2 and Gli3 mRNA levels in HSCs by qRT-PCR. As compared to the control, mRNA expression of Smo, Ptc, Gli2 and Gli3 was significantly decreased after miR-9 transfection. However, these mRNA levels were increased after transfection with the miR-9 inhibitor (Fig. 4A). In addition, exogenous MRP1 was transfected into HSCs, and BrdU assay indicated that cell proliferation was decreased after miR-9 transfection, but increased after MRP1 transfection. However, the suppression of miR-9 on HSC proliferation was restored by the overexpression of MRP1 (Fig. 4B). Furthermore, miR-9 decreased the protein levels of Smo, Gli2 and α-SMA, but the overexpression of MRP1 significantly counteracted this inhibitory effect (Fig. 4C). Taken together, these results indicate that MRP1 overexpression rescues the miR-9-mediated suppressive effect on the activation and proliferation of HSCs by activating the Hh signaling pathway.

To better investigate the effect of miR-9 on liver fibrosis in vivo, we first examined the expression of miR-9 and MRP1 in fibrotic mouse liver tissues. qRT-PCR and western blotting demonstrated decreased miR-9 and increased MRP1 levels in the CCl₄-induced liver fibrosis mouse model (Fig. 5A). Western blotting suggested that the protein expression of Timp-1, α-SMA and Col-1 was increased in the CCl₄ group. miR-9 reduced the expression of Timp-1, α-SMA and Col-1 in the CCl₄-induced mouse model, whereas these effects were restored by overexpression of MRP1 (Fig. 5B). H&E staining showed that liver tissues in the control group exhibited normal structure with no abnormal morphological changes. In contrast, the number of fibrotic septa was increased in the livers of animals from the CCl₄ group, and the normal lobular architecture was lost. The group transfected with miR-9 showed less pathological damage than that observed in the CCl₄ group. Furthermore, the results showed that the overexpression of MRP1 suppressed the effects of miR-9 on CCl₄-induced liver fibrosis (Fig. 5C). These results indicate that miR-9 treatment slowed the progression of liver fibrosis in CCl₄-induced mice, resulting in a reduction in the deposition of ECM in the liver by reducing MRP1.