miR-145 regulates myofibroblast differentiation and lung fibrosis

Abstract

The expression of smooth muscle actin-α (SMA-α) by fibroblasts defines phenotypic transition to myofibroblasts and is a primary contributor to contractile force generation by these differentiated cells. Although the regulation of SMA-α expression has been the focus of many studies, there is presently only limited information concerning miRNA regulation of lung myofibroblast differentiation and the involvement of these miRNAs in pulmonary fibrosis. To determine the role of miR-145 in regulating lung myofibroblast differentiation and pulmonary fibrosis. Wild-type and miR-145^−/− mice were studied. Lung fibrosis models and cell culture systems were employed. miR-145 mimics or inhibitors were transfected into pulmonary fibroblasts. Fibrogenic and contractile activities of lung fibroblasts were determined. We found that miR-145 expression is upregulated in TGF-β1-reated lung fibroblasts. miR-145 expression is also increased in the lungs of patients with idiopathic pulmonary fibrosis as compared to in normal human lungs. Overexpression of miR-145 in lung fibroblasts increased SMA-α expression, enhanced contractility, and promoted formation of focal and fibrillar adhesions. In contrast, miR-145 deficiency diminished TGF-β1 induced SMA-α expression. miR-145 did not affect the activity of TGF-β1, but promoted the activation of latent TGF-β1. miR-145 targets KLF4, a known negative regulator of SMA-α expression. Finally, we found that miR-145^−/− mice are protected from bleomycin-induced pulmonary fibrosis. miR-145 plays an important role in the differentiation of lung myofibroblasts. miR-145 deficiency is protective against bleomycin-induced lung fibrosis, suggesting that miR-145 may be a potential target in the development of novel therapies to treat pathological fibrotic disorders.—Yang, S., Cui, H., Xie, N., Icyuz, M., Banerjee, S., Antony, V. B., Abraham, E., Thannickal, V. J., Liu, G. miR-145 regulates myofibroblast differentiation and lung fibrosis.

Fibroblasts are the primary effectors of tissue fibrosis as they produce collagen and other extracellular matrix (ECM) proteins (1, 2). Fibroblasts often differentiate into myofibroblasts that possess enhanced fibrotic, contractile, and migratory activities (1). Smooth muscle actin-α (SMA-α) is a marker of myofibroblasts and a primary contributor to the contractile force in myofibroblasts (3). The contractility of myofibroblasts is an important regulatory factor in tissue fibrosis, as myofibroblast contraction activates latent transformation growth factor β1 (TGF-β1) via integrins and increases the resistance of these cells to apoptosis (3).

The regulation of SMA-α, particularly at the transcriptional level, has been the focus of numerous recent studies (3). A number of transcriptional factors have been found to regulate SMA-α expression. Among these transcriptional factors, mothers against DPP homolog 2/3 (Smad2/3), myocardin-related transcription factor A/B (MRTF-A/B), and CCAAT/enhancer-binding protein β (C/EBP-β) enhance, whereas Krüppel-like factor 4 (KLF4), peroxisome proliferator-activated receptor γ (PPARγ), and NK2 homeobox 5 (Nkx2.5) suppress SMA-α transcription (3). Epigenetic regulation also plays an important role in the expression of SMA-α during myofibroblast differentiation (3–5).

MicroRNAs (miRNAs) are 21- to 22-nucleotide (nt) noncoding small RNAs (6–10). Aberrant expression of miRNAs is associated with many pathological conditions, including cancer and cardiovascular disease (11–17). Dysregulation of miRNAs has been found to contribute to pulmonary fibrosis. For example, levels of miR-21, a miRNA that is induced by TGF-β1 and promotes TGF-β1 activity, were found to be elevated in the lungs of patients with idiopathic pulmonary fibrosis (IPF) (18). miR-29 and let-7d, which suppress the expression of collagen and inhibit epithelial mesenchymal transition (EMT), respectively, are down-regulated in fibrotic lungs (19–21). However, there is presently only limited information concerning the potential role of miRNA in regulating lung myofibroblast differentiation and generation of contractile activity, as well as the involvement of these miRNAs in pulmonary fibrosis.

MATERIALS AND METHODS

miRNA array

Total RNAs were isolated from human lung fibroblast line MRC-5 that was treated with TGF-β1 for 0, 48, and 96 h using miRNAeasy Mini Kit (Qiagen, Valencia, CA, USA). The miRNA array was performed by Exiqon using miRCURY LNA microRNA Array (Exiqon, Woburn, MA, USA). The data were deposited into Gene Expresion Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) with accession number GSE43992.

Reagents

Rat tail collagen was from Invitrogen (Carlsbad, CA, USA). miRNA mimics and inhibitors were from Ambion (Austin, TX, USA). Human recombinant TGF-β1 was from PeproTech (Rocky Hill, N.J., USA). Sircol collagen assay kit was from Biocolor (Carrickfergus, UK). Bleomycin was from Calbiochem (San Diego, CA, USA).

Mice

C57BL/6 mice were from the National Cancer Institute (Frederick, MD, USA). miR-145^−/− mice were generous gifts from Dr. Eric Olson (University of Texas Southwestern Medical Center, Dallas, TX, USA; ref. 22).

Cell lines

Human pulmonary fibroblast line MRC-5 was purchased from American Type Culture Collection (Manassas, VA, USA). Transformed mink lung cells (TMLCs), reporter cells that stably express a firefly luciferase reporter gene under the control of the TGF-β1-responsive elements within the plasminogen activator inhibitor 1 (PAI-1) promoter, was a generous gift from Dr. Daniel Rifkin (New York Medical Center, New York, NY, USA).

Human lung tissue

IPF and normal lung tissue samples were obtained from the University of Alabama at Birmingham (UAB) Tissue Procurement and Cell Culture Core. The protocol was approved by the UAB Institutional Review Board (IRB).

Isolation of primary mouse lung fibroblasts

Primary mouse lung fibroblasts were isolated as described previously (23). In brief, lung tissues were minced and digested with 0.1% collagenase, 0.005% trypsin, and 0.04% DNase. Selection for lymphocytes/macrophages was performed by incubation on CD16/32- and CD45-coated Petri dishes for 30 min at 37°C. Negative selection for fibroblasts was performed by adherence of the suspension for additional 45 min on cell culture dishes. The suspended cells were used as alveolar epithelial cells (AECs). The adherent lung fibroblasts were cultured in MEM containing 10% FBS. The fibroblasts at passage 2 were trypsinized, and the same numbers of cells were plated for experiments.

Experimental pulmonary fibrosis model

Pulmonary fibrosis was induced as described previously (24).

Immunohistochemistry

Immunohistochemistry was performed as described in our previous studies (24). The area of SMA-α expression was measured by color image analysis software (Image-Pro Plus, Media Cybernetics, Silver Spring, MD, USA).

Real-time PCR

Taqman probes for small nucleolar RNAs sno135 and U48, miR-143, and miR-145 were purchased from Applied Biosystems (Foster City, CA, USA). RNA levels of SMA-α, fibronectin (Fn), and GAPDH and human pri-mir-143/145 transcripts were determined by real-time PCR using SYBR Green master mix kit (Roche, Indianapolis, IN, USA). Primer sequences were as follows: human SMA-α: sense, 5′-CATCACCAACTGGGACGACATGGAA-3′, antisense, 5′-GCATAGCCCTCATAGATGGGGACATTG-3′; mouse SMA-α: sense, 5′-GACGCTGAAGTATCCGATAGAACACG-3′, antisense, 5′-CACCATCTCCAGAGTCCAGCACAAT-3′; human pri-miR-143/145: sense, 5′- GACCGAGGAGCAGGAGGAGAACA-3′, antisense, 5′-GAAGGGGGTAGGAGGAAGGAAGC-3′; human GAPDH: sense, 5′-GCTGGCGCTGAGTACGTCGTGGAGT-3′, antisense, 5′-CACAGTCTTCTGGGTGGCAGTGATGG-3′; mouse GAPDH: sense, 5′-CGACTTCAACAGCAACTCCCACTCTTCC-3′, antisense, 5′-TGGGTGGTCCAGGGTTTCTTACTCCTT-3; human Fn: 5′-GTGTTGGGAATGGTCGTGGGGAATG-3′, antisense, 5′-CCAATGCCACGGCCATAGCAGTAGC-3′; human SM22-α: sense, 5′-GGGGTCATCAAGACTGACATGTTCC-3′, antisense, 5′-CGCTTTCTTCATAAACCAGTTGGGA-3′; mouse fibroblast-specific protein 1 (FSP1): sense, 5′-TCCACAAATACTCAGGCAAAGAGGG-3′, antisense, 5′-TGTTGCTGTCCAAGTTGCTCATCAC-3′; mouse E-Cadherin: sense, 5′-GTGTGCTCACCTCTGGGCTGGAC-3′, antisense, 5′-GAGTGTTGGGGGCATCATCATCG-3′; and human KLF4: sense, 5′-GTTGACTTTGGGGTTCAGGTGCC-3′, antisense, 5′-CGAACGTGGAGAAAGATGGGAGC-3′.

Fibroblast contraction assay

Lung fibroblasts were mixed with 1.5 mg/ml rat tail collagen I diluted in medium 199 and then seeded into a 24-well plate at 1.5-2.5 × 10⁵ cells/well. The collagen gels were freed from attachment to the walls after they were incubated at 37°C for 30 min. Diameters of the gels were measured at the indicated time points after gel release.

Latent TGF-β1 activation assay

Latent TGF-β1 activation was determined as described previously (25). Lung fibroblasts were seeded into a 24-well plate and cultured for 24 h. Cells were transfected with control mimics or miR-145 mimics. At 6 d after transfection, cells were topped with TMLCs at 2.5 × 10⁵ cells/well and then cocultured in MEM with 0.1% FBS for 24 h. Cocultured cells were harvested for luciferase assay. Luciferase activity was measured using a luciferase activity assay kit (Promega, Madison, WI, USA).

Immunofluorescence

Immunofluorescence was performed as described previously (26). Fibroblasts growing on cover slides were fixed in 4% formaldehyde for 20 min. After being permeabilized with 0.5% Triton X-100 for 2 min, the cells were blocked in PBS containing 5% BSA for 1 h. Cells were then incubated with a mixture of anti SMA-α antibody and FITC-conjugated phalloidin, anti-Fn, or anti-vinculin overnight at 4°C. Cells were washed 3 times and then incubated with Alexa Fluor 594-conjugated secondary antibody (Invitrogen) for 1 h. Cells were then mounted with DAPI-containing mounting solution. Fluorescent images were taken with a Leica confocal microscope (Leica Microsystems, Wetzlar, Germany).

Western blot analysis

Western blot analysis was performed as described previously (24). Mouse anti-α-SMA antibody was obtained from Sigma (St. Louis, MO, USA). Mouse anti-fibronectin and anti-GAPDH antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA. Mouse anti-vimentin antibody was from Abcam (Cambridge, MA, USA). Mouse anti-E-cadherin antibody was from BD Biosciences (San Jose, CA, USA). Rabbit antiKLF4, anti-p-Smad2, and anti-Smad2 antibodies were from Cell Signaling (Beverly, MA, USA.

RESULTS

TGF-β1 up-regulates miR-145 and miR-143 in lung fibroblasts

To identify TGF-β1 regulated miRNAs in lung fibroblasts, we performed miRNA profiling on RNA samples isolated from control or TGF-β1-treated normal human lung fibroblasts. We found that the expression of miR-145 and miR-143 is increased in TGF-β1-treated cells. The miR-145 and miR-143 genes are clustered on the same genomic locus. It has also been shown that miR-145 and miR-143 are vascular smooth muscle cell (VSMC)-specific miRNAs and are essential to the contractile phenotype of VSMCs (22, 27).

We next performed real-time PCR assays and confirmed that both miR-145 and miR-143 are induced in the TGF-β1-treated lung fibroblasts (Fig. 1A). We examined the levels of the primary transcripts of the miR-145/143 cluster genes (pri-miR-143/145) and found that they are up-regulated in TGF-β1-treated lung fibroblasts, suggesting that the up-regulation of mature miR-145 and miR-143 results from enhanced transcription of the miR-145/143 gene cluster (Fig. 1B). The mRNA levels of Fn and SMA-α were also increased in the TGF-β1-treated lung fibroblasts, consistent with myofibroblast differentiation (Fig. 1B).

Figure 1.

TGF-β1 up-regulates miR-145 and miR-143 in lung fibroblasts. A) Human lung fibroblast line MRC-5 was treated with TGF-β1 for 0, 48, or 96 h. RNA was isolated, and levels of miR-145 and miR-143 were determined by real-time PCR. ***P < 0.001 vs. time 0. B) Human lung fibroblasts were treated with TGF-β1 for 0, 8, or 24 h. RNA was isolated, and levels of pri-miR-143/145, Fn, and SMA-α were determined by real-time PCR. Values are presented as means ± sd; n = 3. Experiments were performed 3 times with similar results. *P < 0.05, **P < 0.01, ***P < 0.001 vs. time 0. C) Levels of miR-145 in normal human lungs and lungs of patients with IPF were determined by real-time PCR; n = 8 for normal lungs, n = 17 for IPF lungs. Values are presented as means ± sem. *P < 0.05 vs. normal lungs.

As TGF-β1 is a critical cytokine in the pathogenesis of human lung fibrosis (28), we hypothesized that miR-145 participates in this process. Therefore, we examined the level of miR-145 in normal human lungs and in lungs from patients with IPF. As shown in Fig. 1C, miR-145 was expressed at higher levels in IPF lungs than that in normal lungs. These data suggest that the profibrotic effects of TGF-β1 in IPF lungs may be partially mediated by enhanced expression of miR-145.

miR-145, but not miR-143, regulates SMA-α expression in lung fibroblasts

TGF-β1 promotes pulmonary fibrosis, in part, by inducing differentiation of fibroblasts into myofibroblasts, characterized by SMA-α expression and enhanced contractile activity (29, 30). To determine whether miR-145 and miR-143 regulate SMA-α expression, we transfected human lung fibroblasts with control mimics, mimics for miR-143, or miR-145 and evaluated SMA-α induction in these cells. As shown in Fig. 2A and Supplemental Fig. S1, overexpression of miR-143 had no effect on SMA-α expression in untreated or TGF-β1-treated lung fibroblasts. However, overexpression of miR-145 markedly increased the baseline levels of the SMA-α transcripts and SMA-α protein in lung fibroblasts (Fig. 2A and the SMA-α panel in Fig. 2C). miR-145 also enhanced TGF-β1-induced SMA-α expression (Fig. 2B, C). Given that TGF-β1 up-regulates miR-145 in lung fibroblasts, these data suggest that miR-145 may mediate TGF-β1-induced SMA-α expression. Interestingly, miR-145 did not affect the baseline levels or TGF-β1-induced fibronectin expression (Fig. 2C), suggesting that miR-145 preferentially regulates SMA-α expression. Of note, untreated lung fibroblasts had minimum expression of SM22-α, a specific SMC marker, compared to pulmonary smooth muscle cells (Supplemental Fig. S2B). Both miR-145 and TGF-β1 enhanced expression of SMA-α and SM22-α (Supplemental Fig. S2), suggesting that they promote differentiation of lung fibroblasts into myofibroblasts that characteristically express SMC markers.

Figure 2.

miR-145, but not miR-143, regulates SMA-α expression in lung fibroblasts. A) Human lung fibroblasts were transfected with 10 nM control mimics or mimics for miR-143 or miR-145. At 3 d after transfection, RNA was isolated, and levels of SMA-α were determined by real-time PCR; n = 3. ***P < 0.001 vs. control miR group. B) Human lung fibroblasts were transfected with 10 nM control mimics or mimics for miR-145. At 2 d after transfection, cells were starved in MEM containing 0.1% FBS for 1 d. Cells were then treated with TGF-β1 for 24 h. RNA was isolated, and levels of SMA-α were determined by real-time PCR; n = 3. ***P < 0.001 vs. untreated control group; ^###P < 0.001 vs. TGF-β1-treated control group. C) Human lung fibroblasts were transfected with 10 nM control mimics or mimics for miR-145. Cells were treated with 2 ng/ml TGF-β1 for 0, 24, and 48 h. Levels of SMA-α, Fn, and GAPDH were determined by Western blot analysis. D) Human pulmonary fibroblasts were transfected with 20 nM control inhibitors or inhibitors against miR-145. At 2 d after transfection, cells were starved in MEM containing 0.1% FBS for 1 d. Cells were then treated with TGF-β1 for 6 h. Levels of SMA-α were determined by real-time PCR. ***P < 0.001 vs. con⁻ group; **P < 0.01 vs. con⁺ group. E) Human lung fibroblasts were transfected with 20 nM control inhibitors or inhibitors against miR-145. Cells were then treated with TGF-β1 for 24 h. Levels of SMA-α were determined by Western blot analysis (E). F) Densitometrical and statistical analyses of TGF-β1-induced expression of SMA-α shown in E. ***P < 0.001 vs.anti-con group. Similar experiments were performed ≥3 times. G) Human lung fibroblasts were transfected with 20 nM control inhibitors or inhibitors against miR-145 for 3 d. Levels of miR-145 were determined by real-time PCR; n = 3. ***P < 0.001 vs.untreated control group. Values are presented as means ± sd.

To determine whether miR-145 is required for TGF-β1-induced SMA-α expression, we transfected human lung fibroblasts with control inhibitors or specific inhibitors against miR-145 and then evaluated SMA-α levels in the cells. As shown in Fig. 2D–F, knockdown of miR-145 attenuated TGF-β1 induced SMA-α expression at both mRNA and protein levels in lung fibroblasts. These data indicate that TGF-β1-mediated SMA-α expression requires, at least in part, the induction of miR-145. As expected, miR-145 is markedly down-regulated by specific antimiR-145 inhibitors in these cells (Fig. 2G).

TGF-β1-induced SMA-α expression is attenuated in miR-145-deficient (miR-145^−/−) lung fibroblasts

To gain further insight into the role of miR-145 in regulating TGF-β1-induced SMA-α expression, we treated lung fibroblasts from wild-type and miR-145^−/− mice with TGF-β1. We first confirmed that miR-145 was undetectable in miR-145^−/− mouse lung fibroblasts (Fig. 3A). Next, we assessed TGF-β1 responses in wild-type mouse and miR-145^−/− mouse lung fibroblasts. While TGF-β1 increased SMA-α expression in wild-type lung fibroblasts, there was minimum induction of SMA-α in miR-145^−/− mouse lung fibroblasts (Fig. 3B). These data support an essential role for miR-145 in TGF-β1-induced myofibroblast differentiation. Of note, the mouse lung fibroblasts had much greater levels of FSP1, whereas they demonstrated minimum expression of E-cadherin, as compared to mouse lung epithelial cells (Supplemental Fig. S3). These data indicate the homogenous identity of the mouse primary lung fibroblasts, although the possibility of contamination of the lines with smooth muscle cells that may dedifferentiate during culture cannot be ruled out.

Figure 3.

TGF-β1-induced SMA-α is attenuated in miR-145 deficient (miR-145^−/−) pulmonary fibroblasts. A) Levels of miR-145 in wild-type (WT) and miR-145^−/− lung fibroblasts were determined by real-time PCR. ***P < 0.001 vs. WT. B) WT and miR-145^−/− lung fibroblasts were starved in MEM containing 0.1% FBS for 1 d. Cells were then treated with 2 ng/ml TGF-β1 for 24 h. RNA was isolated, and levels of SMA-α were determined by real-time PCR; n = 3. ***P < 0.001 vs. WT untreated group; **P < 0.01 vs.WT TGF-β1-treated group. Experiments were performed twice with similar results. Values are presented as means ± sd.

miR-145 enhances the contractility of lung fibroblasts

As shown above, miR-145 enhances baseline levels and TGF-β1-induced SMA-α expression in pulmonary fibroblasts. SMA-α is critical for the generation of contractile force (1, 31). These findings, therefore, suggest that miR-145 may regulate the contractility of lung fibroblasts. To test this hypothesis, we transfected human lung fibroblasts with control mimics or mimics for miR-145 and then evaluated contractility using collagen gel contraction assays. As shown in Fig. 4A, lung fibroblasts transfected with miR-145 demonstrated greater contractility, as compared to those transfected with control mimics. These data are concordant with increased SMA-α expression in miR-145-transfected lung fibroblasts. TGF-β1 increased contractility of lung fibroblasts, and this effect was further enhanced by miR-145 (Fig. 4A).

Figure 4.

miR-145 enhances the contractility of lung fibroblasts. A) Human lung fibroblasts were transfected with 10 nM control mimics or 10 nM miR-145 mimics. At 2 d after transfection, cells were trypsinized and mixed with 1.5 mg/ml rat tail collagen. Gels were freed from the wall of the wells 30 min after gel formation, and cells were cultured in medium containing 0.1% FBS for 1 d. Cells were treated without or with 4 ng/ml TGF-β1 for 2 d. Quantification of collagen gel surface areas was performed and plotted; n = 4. *P < 0.05 vs. untreated con-miR group; ***P < 0.001 vs. untreated con-miR group; ^#P < 0.05 vs. TGF-β1 treated con-miR group. B) Wild-type (WT) and miR-145^−/− lung fibroblasts were trypsinized and mixed with 1.5 mg/ml rat tail collagen. Gels were freed from the wall of the wells 30 min after gel formation, and cells were cultured in medium containing 0.1% FBS for 1 d. Cells were treated without or with 4 ng/ml TGF-β1 for 2 d. Quantification of collagen gel surface areas was performed and plotted; n = 4. **P < 0.01 vs. untreated WT group; ^#P < 0.05 vs. TGF-β1-treated WT group. Experiments were performed twice with similar results. Values are presented as means ± sd.

Because we have found that the TGF-β1-induced SMA-α is suppressed in miR-145^−/− mouse lung fibroblasts, we next asked whether TGF-β1-enhanced contractility is defective in these cells. As shown in Fig. 4B, TGF-β1-induced contraction of miR-145^−/− mouse lung fibroblasts was diminished, as compared to that found with wild-type fibroblasts.

miR-145 enhances the formation of focal and fibrillar adhesions in lung fibroblasts

SMA-α contributes to the contractility and adhesion capacity of myofibroblasts by promoting the formation of filamentous F actin and mature focal and fibrillar adhesions (3, 31). Since we found that miR-145 enhances SMA-α expression in lung fibroblasts, we next determined whether miR-145 regulates the formation of filamentous F actin and mature focal and fibrillar adhesions. As shown in Fig. 5A, overexpression of miR-145 increased SMA-α⁺ stress fiber, which generally overlapped with strengthened long filamentous F-actin. Furthermore, overexpression of miR-145 enhanced focal adhesion formation in lung fibroblasts, demonstrated by an increase in both the number and size of the focal adhesion protein vinculin (Fig. 5B). Overexpression of miR-145 also increased fibrillar adhesion and fibronectin organization, as reflected by enhanced fibronectin staining on the cell perimeter and strengthened organization of fibronectin network (Fig. 5B). These data suggest that miR-145 contributes to the contractility and adhesion capacity of myofibroblasts by up-regulating SMA-α, increasing the formation of filamentous F actin along with mature focal and fibrillar adhesions.

Figure 5.

miR-145 enhances the formation of focal and fibrillar adhesions in lung fibroblasts. A) Human lung fibroblasts were transfected with 10 nM control mimics or miR-145 mimics. At 2 d after the transfection, cells were starved in medium containing 0.1% FBS for 1 d, followed by treatment without or with 2 ng/ml TGF-β1 for 1 d. Cells were then fixed and incubated with a mixture of anti-SMA-α antibody and FITC-conjugated phalloidin for staining of F-actin overnight. After washing 3 times, cells were incubated with Alexa Fluor-594-conjugated secondary antibody for 1 h. Confocal microscopy was performed. Representative experiments are shown. B) Human lung fibroblasts were treated as in A. Cells were then fixed and incubated with anti-vinculin or anti-Fn antibody overnight. After washing 3 times, cells were incubated with Alexa Fluor-594-conjugated secondary antibody for 1 h. Confocal microscopy was performed. Representative experiments are shown.

miR-145 targets KLF4

As shown in our initial experiments, miR-145 enhanced baseline levels and TGF-β1 induced SMA-α in lung fibroblasts. However, we found that neither overexpression nor knockdown of miR-145 affects phosphorylation of Smad2/3 in lung fibroblasts treated with or without TGF-β1 (Fig. 6A, B). These data suggest that miR-145 enhances SMA-α expression independently of TGF-β receptor signaling. As found in previous studies and suggested by TargetScan, a computational program that predicts miRNA targets, KLF4, appears to be a major target of miR-145 (22, 27, 32, 33). KLF4 was previously shown to suppress SMA-α expression by interfering with the binding of Smad2/3 to the SMA-α promoter and association with activators of the TGF-β control element (TCE; refs. 34, 35). To determine whether miR-145 regulates KLF4 in lung fibroblasts, we transfected human lung fibroblasts with control mimics or mimics for miR-145 and then evaluated KLF4 levels in the cells. As shown in Fig. 6D, miR-145 reduced KLF4 protein levels in lung fibroblasts. miR-145 had no effect on the levels of the KLF4 transcripts (Fig. 6C), suggesting that miR-145 affects the translation, but not the stability, of the KLF4 transcripts in lung fibroblasts. Because TGF-β1 up-regulates miR-145, which targets KLF4, we next asked whether TGF-β1 inhibits KLF4 expression in lung fibroblasts. As shown in Fig. 6E, TGF-β1 decreased KLF4 expression. Furthermore, we found that the reduced expression of KLF4 in TGF-β1-treated cells is attenuated by anti-miR-145 inhibitors (Fig. 6E). We found that there are higher levels of KLF4 in miR-145^−/− lung fibroblasts than in the wild-type controls (Fig. 6F), concordant with our findings that the induction of SMA-α by TGF-β1 is defective in miR-145-deficient lung fibroblasts. Together, these data suggest that miR-145 enhances SMA-α expression by suppressing the inhibitory transcriptional factor KLF4.

Figure 6.

miR-145 targets KLF4. A, B) Human lung fibroblasts were transfected with 10 nM control mimics (A), miR-145 mimics (A), 20 nM control inhibitors (B), 20 nM inhibitors against miR-145 (B). At 2 d after the transfection, cells were starved in medium containing 0.1% FBS for 1 d, followed by treatment with 2 ng/ml TGF-β1 for 0, 0.5, 1, or 24 h. Levels of p-Smad2, Smad2, and GAPDH were determined by Western blot analysis. C, D) Human lung fibroblasts were transfected with 10 nM control mimics or miR-145 mimics. At 3 d after transfection, KLF4 and GAPDH were determined by real-time PCR (C) and Western blot analysis (D). E) Human lung fibroblasts were transfected with 20 nM control inhibitors or inhibitors against miR-145. At 1 d after transfection, cells were starved in MEM containing 0.1% FBS for 1 d. Cells were then treated with TGF-β1 for 24 h. Levels of KLF4 and GAPDH were determined by Western blot analysis. F) Levels of KLF4 and GAPDH in wild-type and miR-145^−/− lung fibroblasts were determined by Western blot analysis. Experiments were performed twice with similar results.

miR-145 activates latent TGF-β1

miR-145 enhances SMA-α expression and increases the contractility of lung fibroblasts. It is known that the mechanical stimuli derived from cellular contraction can activate latent TGF-β1 in the pericellular matrix (36, 37). We tested whether miR-145 promotes the activation of latent TGF-β1. Of note, although miR-145 enhanced SMA-α expression, it had no effect on the baseline levels of Smad2/3 phosphorylation or on TGF-β1-dependent fibronectin expression (Figs. 2C and 6A). However, these experiments were performed in cells that were in culture for only 3 d. We reasoned that there may be little endogenous latent TGF-β1 produced by the fibroblasts and deposited into ECM within this short period of time. To overcome this limitation and to determine whether miR-145 promotes latent TGF-β1 activation, lung fibroblasts transfected with control mimics or mimics for miR-145 were cultured for 6 d to allow accumulation of ECM and deposition of latent TGF-β1 produced by the fibroblasts into ECM. TMLC reporter cells were then plated on the top of the transfected cells, and the coculture was kept in MEM containing 0.1% FBS for 24 h. TMLC reporter cells contain a luciferase reporter that is under the control of the TGF-β1-responsive elements from the PAI-1 gene (38). These cells demonstrate luciferase activity that correlates with the availability of active TGF-β1 on the cell surface and receptor signaling. As shown in Fig. 7A, greater luciferase activity was observed in fibroblasts transfected with miR-145 mimics, suggesting that there is an increased level of latent TGF-β1 activation under such conditions. As expected, levels of fibronectin and SMA-α in the miR-145-transfected fibroblasts at this time point were greater than in the cells transfected with the control mimics (Fig. 7B).

Figure 7.

miR-145 activates latent TGF-β1. A) Human lung fibroblasts were transfected with 10 nM control mimics or miR-145 mimics. At 6 d after transfection, cells were topped with TMLC cells and then cocultured for 24 h. Luciferase activity in the cells were determined. *P < 0.05 vs. con miR group. B) Human lung fibroblasts were transfected with 10 nM control mimics or miR-145 mimics. At 7 d after transfection, levels of SMA-α, Fn, and GAPDH were determined by Western blot analysis. Experiments were performed twice with similar results.

miR-145 deficiency attenuates bleomycin-induced pulmonary fibrosis

Our in vitro experiments demonstrated that miR-145 enhances SMA-α expression and contractility of lung fibroblasts. We also found that miR-145 promotes the activation of latent TGF-β1. All of these observations are important events that contribute to the pathogenesis of pulmonary fibrosis. Thus, we asked whether miR-145 participates in the development of pulmonary fibrosis. To examine this issue, wild-type mice and miR-145^−/− mice were treated intratracheally with bleomycin, and the severity of lung fibrosis was evaluated after 14 d. As shown in Fig. 8A, there was a significant increase in collagen deposition in the lungs of bleomycin treated wild-type mice. However, the amount of bleomycin-induced collagen deposition was diminished in the lungs of miR-145^−/− mice. These data suggest that miR-145 participates in the development of lung fibrosis. The attenuation of bleomycin-induced lung fibrosis in miR-145^−/− mice was confirmed by histological examination of the lungs (Fig. 8B).

Figure 8.

miR-145 deficiency attenuates bleomycin-induced pulmonary fibrosis. A) Wild-type (WT) and miR-145^−/− mice were intratracheally instilled with saline or bleomycin (BLM). At 14 d after the treatment, collagen contents in the right lungs were determined by Sircoll assays. Values are presented as means ± sem; n = 3–9. ***P < 0.001 vs. WT saline; **P < 0.01 vs. WT BLM. B–D) Experiments were performed as in A. Lung tissue sections were prepared, and hematoxyln and eosin (H&E) staining (B) and immunohistochemistry staining for KLF4 (C) and SMA-α (D) were performed. Original view: ×40. E) Quantification of SMA-α expression in the lungs of saline- or bleomycin-treated WT and miR-145^−/− mice was performed by color image analysis software. Values are presented as means ± sd. ***P < 0.001 vs. WT saline; **P < 0.01 vs. WT BLM.

As we have shown that miR-145 targets KLF4 in vitro in lung fibroblasts, we next asked whether miR-145 regulates KLF4 expression in the lungs. As shown in Fig. 8C, KLF4 was primarily located in the nuclei of AECs in saline treated wild-type mice. The levels of KLF4 expression in AECs of saline-treated miR-145^−/− mice appeared comparable to those in AECs of saline-treated wild-type mice. We found many more KLF4⁺ cells appeared in the thickened parenchyma, likely fibrotic areas, in the lungs of bleomycin-treated wild-type mice. However, cells in the same areas of the lungs of bleomycin treated miR-145^−/− mice had more intense KLF4 staining (Fig. 8C). In contrast to the pattern of KLF4 expression in wild-type and miR-145^−/− mouse lungs, there was a greater expression of SMA-α in the lungs of bleomycin treated wild-type mice than in the lungs of bleomycin treated miR-145^−/− mice (Fig. 8C, D). Taken together, these data suggest that miR-145 suppresses KLF4, thereby promoting SMA-α expression and enhancing bleomycin-induced pulmonary fibrosis.

DISCUSSION

Myofibroblasts are the key effector cells in the pathogenesis of lung fibrosis (2, 3). SMA-α is a major molecular marker for myofibroblasts and a primary contributor to the generation of contractile force in myofibroblasts (29, 30). The regulation of SMA-α expression has been the focus of numerous studies, particularly at the transcriptional level (3, 31). The Smad2/3-dependent regulation of SMA-α by TGF-β1 is one of the best-studied mechanisms (35). In the past 3 yr, miRNAs have emerged as important regulators of lung fibrosis through modulation of TGF-β1 activity and EMT of lung epithelial cells (18, 19, 24). However, it has not been previously determined whether miRNAs participate in the differentiation of pulmonary myofibroblasts. To study the role of miRNAs in regulating SMA-α expression in pulmonary fibroblasts, we performed miRNA profiling in these cells treated with or without TGF-β1 and found that miR-145 and miR-143 are upregulated by TGF-β1 in lung fibroblasts. We found that levels of miR-145 in IPF lungs are greater than those in normal controls. Such an increase could be caused by elevated expression of miR-145 in IPF fibroblasts. However, it is also possible that it is due to alterations of cellular composition in fibrotic lungs.

In the present studies, we demonstrated that miR-145, but not miR-143, increases SMA-α expression. Interestingly, miR-145 increased SMA-α expression in lung fibroblasts is not caused by an activation of TGF-β1 signaling, as miR-145 does not affect Smad2/3 phosphorylation, at baseline, or in TGF-β1-treated lung fibroblasts. Such actions of miR-145 in lung fibroblasts are different from those found with miR-21 in that miR-21 enhances SMA-α expression by promoting TGF-β1 signaling (18).

We found that miR-145 enhances the contractility of lung fibroblasts. This effect appears to be primarily due to increased SMA-α expression in the cells transfected with miR-145. Consistent with these results, we found that miR-145 promotes the formation of stress fibers. miR-145 also increased focal and fibrillar adhesions in lung fibroblasts, suggesting that miR-145 is capable of enhancing the activated cellular phenotype of myofibroblasts.

We found that miR-145 promotes latent TGF-β1 activation in a delayed manner, as suggested by the elevated expression of fibronectin in lung fibroblasts 7 d after they were transfected with miR-145. The increased activation of latent TGF-β1 by miR-145 is most likely due to its activity in enhancing SMA-α expression and the resulting increase in myofibroblast contractility. It has been previously shown that mechanical forces generated by cell contraction activates latent TGF-β1 (36). The ability of miR-145 to enhance latent TGF-β1 activation likely contributes to its role in promoting pulmonary fibrosis.

Consistent with previous observations in VSMCs, we found that miR-145 targets KLF4 in lung fibroblasts. KLF4 is known to inhibit the activation of the CArG box [CC(A/T)₆GG] by serum response factor and its coactivators, myocardin (Myocd), or MRTFs (33, 39). Therefore, the enhanced expression of SMA-α in miR-145 overexpressing lung fibroblasts could be a result of derepression of the CArG box present in the SMA-α promoter.

There is evidence showing that KLF4 is involved in pathological fibrogenesis. Mice with cardiomyocyte-specific deletion of KLF4 develop increased myocardial fibrosis in response to transverse aortic constriction (TAC; ref. 40). Such a protective role of KLF4 in cardiac fibrosis is supported by our findings that miR-145^−/− mice are resistant to bleomycin-induced lung fibrosis. Indeed, there is a marked increase in KLF4 expression in the lungs of bleomycin-treated miR-145^−/− mice, compared to that in the lungs of bleomycin-treated wild-type mice. It is likely that the enhanced expression of KLF4 confers protection against injury-induced lung fibrosis in miR-145^−/− mice. Although KLF4 is expressed in AECs of wild-type mice, we did not observe a significant increase in KLF4 expression in these cells of uninjured miR-145^−/− mice. These results could be explained by low basal levels of miR-145 present in AECs. Therefore, although we showed that miR-145 promotes EMT in vitro, the role of miR-145 in AECs in lung fibrosis remains to be defined.

miR-145 and miR-143 have tumor suppressor activity (41, 42). Reduced expression of miR-145 and miR-143 has been shown to be associated with cancer progression, metastasis, and angiogenesis (41, 42). miR-145 was also found to be dysregulated in mouse models of pulmonary arterial hypertension (PAH), and enhanced expression of miR-145 contributes to pulmonary vascular remodeling (43). Down-regulation of miR-145 protects against the development of PAH (43). These previous studies and our findings demonstrate the prevalence of miR-145 dysregulation in various pulmonary disorders.

Finally, we found that miR-145^−/− mice develop less severe lung fibrosis than wild-type mice after intratracheal administration of bleomycin. This adds to a growing list of miRNAs, including miR-21, miR-155, miR-200, let-7d, and miR-29, that are involved in the pathogenesis of lung fibrosis (18–21, 24, 44). However, we need to be aware that there was no direct correlation between the reduced fibrosis in the lungs of bleomycin-treated miR-145^−/− mice and the diminished contractile activity and SMA-α expression in miR-145-deficient lung fibroblasts in vitro. Furthermore, it is also important to recognize that targeting a single miRNA may not be as effective as targeting 2 or more miRNAs in the complex pathogenesis of pulmonary fibrosis. However, it is currently unknown which combination of miRNAs will be most efficacious in treating lung fibrosis. The answer to this question may hinge on specific miRNAs that are altered in individual patients. Conceivably, a personalized miRNA targeting approach may represent the best strategy.