miR-21 mediates fibrogenic activation of pulmonary fibroblasts and lung fibrosis

To investigate if miRNAs may participate in the pathogenesis and progression of lung fibrosis, we identified miRNAs whose expression was altered in fibrotic lungs. A miRNA array assay was performed on RNA isolated from the lungs of mice that had been given intratracheal PBS or bleomycin for 1 or 2 wk, a well-characterized model of lung fibrosis (Moore and Hogaboam, 2008). We found that a number of miRNAs had significantly altered expression in bleomycin-exposed lungs (Fig. S1, A and B). Of these miRNAs, miR-21 demonstrated the greatest increase in expression. miR-21 has previously been shown to be induced in TGF-β1–treated human vascular smooth muscle cells and to regulate the expression of genes involved in the contraction of smooth muscle (Davis et al., 2008). Because TGF-β1 is a central pathological mediator of lung fibrosis (Cutroneo et al., 2007; Moore and Hogaboam, 2008), the induction of miR-21 by TGF-β1 suggests that miR-21 may have a potential role in the pathogenesis of lung fibrosis. To validate the miRNA array data, Northern blotting was performed and showed that miR-21 was up-regulated in lungs from bleomycin-exposed mice as early as d 3 after bleomycin administration, reached its highest levels on d 14, and remained at the highest level until at least 24 d after bleomycin administration (Fig. 1 A). Real-time PCR analysis on the same set of samples demonstrated similar results to Northern blotting (Fig. 1 B). Of note, expression of the extracellular matrix protein, fibronectin, was also up-regulated with similar timing as to miR-21 (Fig. 1 C). These data suggest that the enhanced expression of miR-21 is involved in the progression of lung fibrosis.

To demonstrate if TGF-β1 contributed to the enhanced expression of miR-21 in the fibrotic lungs, we examined pulmonary levels of TGF-β1 and found that, similar to miR-21, TGF-β1 expression started to rise at d 3 after bleomycin administration and reached maximal levels on d 7 (Fig. 1 D). TGF-β1 returned to basal levels on d 24 after bleomycin administration, consistent with previous studies (Zhang et al., 1995). To further delineate the role of TGF-β1 in the enhanced expression of miR-21 in fibrotic lungs, we used transgenic mice that inducibly express a dominant-negative form of TGF-β1RII (Ambalavanan et al., 2008), which can inhibit TGF-β1 downstream signaling events by blocking TGF-β1–induced formation of endogenous TGF-β1RI-II complex. As shown in Fig. S1 C, the enhanced expression of miR-21 was significantly attenuated in the mice expressing dominant-negative TGF-β1RII after bleomycin administration. These results suggest that TGF-β1 plays an important role in mediating the increased expression of miR-21, although other mechanisms may also be involved.

In situ hybridization (ISH) of mouse lungs showed only minimal cytoplasmic staining for miR-21, but dramatic increases after bleomycin treatment (Fig. 1 E, i-ii). The expression of miR-21 appeared in sheets in the lung parenchyma, consistent with the pattern of fibroblast/myofibroblast accumulation in bleomycin-treated lungs (Fig. 1 E, ii). In contrast, there was only minimal background staining using control probes with scrambled sequence, supporting the specificity of the miR-21 staining (Fig. 1 E, iii). To determine if miR-21 was indeed expressed in pulmonary myofibroblasts in bleomycin-treated lungs, immunohistochemistry was performed along with ISH and demonstrated that miR-21 expression was primarily colocalized with that for α-SMA (Fig. 1 F), suggesting that myofibroblasts were the main source of the increased miR-21 levels present in bleomycin-treated lungs. Lungs from patients with IPF also showed increased expression of miR-21 (Fig. 1 G), with expression primarily localized to the fibroblastic foci (Fig. 1 H, i). Taken together, these data suggest that miR-21 may participate in the pathogenesis of lung fibrosis by regulating fibroblast/myofibroblast activation.

To determine if miR-21 is involved in the development of experimental lung fibrosis, we used locked nucleic acid (LNA)–modified miR-21 antisense probes to modulate miR-21 expression. LNA-modified miRNA antisense probes were previously shown to be capable of effectively sequestering miR-122 in vivo by forming duplexes with this miRNA (Elmén et al., 2008). Intratracheal instillation of miR-21 antisense probes before bleomycin administration completely sequestered miR-21, as demonstrated by the formation of miR-21:anti-miR-21 duplexes (bands with retarded migration) in lungs harvested 2 wk after bleomycin administration (Fig. 2 A). Real-time PCR demonstrated no mature miR-21 expression in miR-21 antisense probe-treated lungs, whereas the expression of an unrelated miRNA, miR-155, was not affected by administration of the miR-21 antisense probes, consistent with specific targeting of miR-21 by the antisense probes (Fig. 2 B). Treatment with miR-21 antisense probes prevented the enhanced collagen deposition (Fig. 2 C) and elevated expression of ECM proteins, such as Fn and collagen, at both RNA and protein levels that occur in the lungs after bleomycin administration (Fig. 2, D and E). The control probes had no effects on bleomycin-induced lung fibrosis (Fig. S2 E). These data indicate that sequestration of miR-21 with antisense probes prevents experimental pulmonary fibrosis. Of note, α-SMA expression was not increased in bleomycin-treated lungs from mice given miR-21 antisense probes, suggesting a role for miR-21 in the induction of myofibroblast differentiation (Fig. 2 E). H&E staining demonstrated dramatic attenuation of bleomycin-induced lung fibrosis in mice pretreated with miR-21 antisense probes; this finding was confirmed by Masson’s trichrome staining, which highlights collagen deposition (Fig. 2 F). Likewise, pretreatment with miR-21 antisense probes prevented the accumulation of myofibroblasts in bleomycin-treated lungs, as demonstrated by immunohistochemistry staining with anti α-SMA antibody (Fig. 2 G).

To explore if inhibition of miR-21 has therapeutic potential in the treatment of lung fibrosis, we administered miR-21 antisense probes i.p. once a day from d 5 to 7 after intratracheal instillation of bleomycin, a time when inflammatory responses start to subside and active fibrogenesis occurs (Vittal et al., 2005; Moore and Hogaboam, 2008), and then analyzed the extent of lung fibrosis 14 d after administration of bleomycin. Under these conditions, the majority of mature miR-21 appeared to be sequestered (Fig. 3 A), and miR-21 antisense probes attenuated collagen deposition (Fig. 3 B) as well as Fn and collagen expression at both RNA and protein levels in bleomycin-treated lungs (Fig. 3, C and D). To further define the therapeutic potential of miR-21 sequestration, we administered miR-21 antisense probes intratracheally on d 7 after intratracheal instillation of bleomycin and then determined the extent of fibrosis in the lungs 3 wk after bleomycin injection. As shown in Fig. S2 A, intratracheal instillation of miR-21 antisense probes after bleomycin administration sequestered miR-21 in mouse lungs. Collagen deposition in the lungs was significantly attenuated 3 wk after bleomycin administration in anti-miR-21–treated mice (Fig. S2 B). Administration of miR-21 antisense probes intratracheally on d 14 after intratracheal instillation of bleomycin was still able to attenuate lung fibrosis, although the effects were less significant than those in mice that were treated with anti–miR-21 probes at earlier time points (Fig. S2, C and D).

Pulmonary fibroblasts are primary effectors of lung fibrotic diseases (Hinz et al., 2007). TGF-β1 induces fibroblast differentiation into more fibrogenic myofibroblasts (Hinz et al., 2007). We showed above that enhanced miR-21 expression in bleomycin-treated mice was primarily localized to α-SMA–expressing myofibroblasts. To determine if the mechanisms leading to the anti-fibrotic effects of miR-21 antisense probes in vivo involve potential regulation of fibrogenic activities of pulmonary fibroblasts by miR-21, we studied the role of miR-21 in the activation of pulmonary fibroblasts by TGF-β1. TGF-β1 up-regulated miR-21 expression in a dose- and time-dependent manner in pulmonary fibroblasts (Fig. 4, A and B; Fig. S3 A), suggesting the potential involvement of miR-21 in TGF-β1–related signaling events. Increasing miR-21 levels by transfection of miR-21 precursors enhanced TGF-β1–induced transcription of Fn and α-SMA in pulmonary fibroblasts (Fig. 4 C). Conversely, knocking down miR-21 attenuated TGF-β1–dependent transcription of Fn and α-SMA (Fig. 4 D). Similarly, increasing miR-21 levels enhanced, whereas knocking down miR-21 attenuated, Fn and α-SMA protein levels in pulmonary fibroblasts (Fig. 4, E and F). These data suggest that the anti-fibrotic effects of miR-21 antisense probes may be mediated through regulation of TGF-β1–related signaling events. Consistent with this hypothesis, increasing miR-21 levels enhanced, whereas knocking down miR-21 attenuated, Smad2 phosphorylation in response to TGF-β1 stimulation (Fig. 4, G and H).

Because miR-21 regulates TGF-β1 signaling events under in vitro conditions in pulmonary fibroblasts, we expected that it also might have such actions in vivo. In bleomycin-treated mice, administration of miR-21 antisense probes attenuated Smad2 phosphorylation in the lungs (Fig. 4 I), an event mediated by TGF-β1 activation (Wang et al., 2006). The computational algorithm TargetScan predicted that Smad7, an inhibitory Smad, is a miR-21 target. Consistent with this prediction, up-regulating miR-21 levels decreased, whereas blocking miR-21 increased, Smad7 expression in pulmonary fibroblasts (Fig. 4, J and K). Smad7 expression was also decreased in bleomycin-treated lungs, concomitant with enhanced miR-21 expression under these conditions. Treatment with miR-21 antisense probes prevented the decrease in Smad7 expression in bleomycin-treated lungs (Fig. 4 L). These data suggest that miR-21 targets Smad7 to enhance signaling events downstream of TGF-β1. To determine if miR-21 directly regulates Smad7 expression, the 3′ UTR of the Smad7 gene was cloned into a luciferase reporter, pMIR-reporter. A 3′ UTR mutant that contains mutations in the predicted miR-21 seeding sequence was also cloned into the pMIR-reporter (Fig. S4 A). Transfection of miR-21 precursors significantly down-regulated luciferase activity for the reporter containing wild-type, but not mutant, Smad7 3′ UTR (Fig. 4 M), consistent with direct regulation of Smad7 by miR-21. Furthermore, the Smad7-expressing vector containing wild-type, but not mutant Smad7 3′ UTR, was subject to regulation by miR-21 (Fig. 4 N).

In this study, we showed that expression of miR-21 is increased in the lungs of bleomycin-treated mice and in the lungs of patients with IPF. miR-21 expression was enhanced in locations with fibroblast/myofibroblast accumulation and was localized to myofibroblasts. TGF-β1–induced miR-21 expression and miR-21 in turn promoted TGF-β1–induced fibrogenic activation of pulmonary fibroblasts by targeting the inhibitory Smad, Smad7. Thus, miR-21 appears to function in an amplifying circuit to enhance TGF-β1 signaling events and to promote fibrotic lung diseases in which TGF-β plays a contributory role, including IPF.

miRNAs normally have multiple targets (Bushati and Cohen, 2007; Lodish et al., 2008). Though Smad7 was previously shown to negatively regulate lung fibrosis (Nakao et al., 1999; Shukla et al., 2009), inhibition of Smad7 may not be the sole mechanism by which miR-21 exerts pro-fibrogenic effects. For example, Spry1, a negative regulator of Erk activation, was previously shown to be a miR-21 target (Thum et al., 2008). Erk activation has been demonstrated to promote the fibrogenic activities of TGF-β1 (Ding et al., 2008). In our experiments, Spry1 was decreased in bleomycin-treated lungs, whereas Erk phosphorylation was increased (Fig. S4 B). The decrease in Spry1 expression and increase in Erk phosphorylation was prevented in the lungs of mice treated with miR-21 antisense probes (Fig. S4 B). In experimental models of myocardial infarction, phosphatase and tensin homologue (PTEN) was down-regulated by miR-21 in cardiofibroblasts from the infracted areas (Roy et al., 2009). Furthermore, PTEN has been shown to negatively regulate experimental lung fibrosis (White et al., 2006). Such studies suggest that PTEN could also be involved in the profibrotic effects of miR-21. Thus, miR-21 occupies an important role in integrating functionally connected pathways involved in pulmonary fibrotic disease through its ability to regulate multiple important signaling events involved in fibrogenesis. Targeting miR-21 in IPF may represent a better therapeutic strategy than approaches aimed at a single pathway. Our data highlight the ability of miRNAs to fine tune various cellular and developmental events, rather than abolishing the expression of a single protein (Bushati and Cohen, 2007; Lodish et al., 2008).

Although TGF-β1 is an important mediator of fibrotic diseases, it may not be the only factor causing miR-21 up-regulation during fibrosis, as indicated by our data that miR-21 was still up-regulated in the lungs of mice expressing dominant-negative TGF-β1RII after intratracheal administration of bleomycin, although to a lesser extent than in wild-type mice (Fig. S1 C). Other pro-fibrotic growth factors, such as EGF signaling, have been shown to regulate miR-21 expression in cancer cells (Seike et al., 2009). We found that bFGF, another important profibrotic growth factor involved in the pathogenesis of IPF (Inoue et al., 2002), enhances miR-21 expression in human primary fibroblasts (Fig. S3 B). Thus, dysregulation of miR-21 could arise from aberrations in multiple critical signaling events involved in pulmonary fibrosis.

Our studies provide proof of concept and suggest a novel approach using miRNA therapeutics, specifically directed to miR-21, in treating clinically important fibrotic diseases such as IPF, for which cures have long been elusive.

C57BL/6 mice were purchased from NCI-Frederick. Transgenic mice that inducibly express a dominant-negative form of TGF-β1RII were a gift from Drs. Rosa Serra and Namasivayam Ambalavanan (University of Alabama at Birmingham, Birmingham, AL). To induce the expression of dominant-negative TGF-β1RII, mice were given water containing 10 mM ZnSO₄ for 2 wk before intratracheal administration of bleomycin and kept on ZnSO₄-containing water for the entire period of the experiments. As controls, wild-type mice were provided with the same ZnSO₄-containing water. For bleomycin instillation, 8-wk-old mice were anesthetized with isoflurane. After the tongues of the anesthetized mice were gently pulled forward with forceps, bleomycin (from EMD; 1–1.5 U/kg body weight in 50 µl PBS) was delivered into the oropharyngeal cavity. The tongue was kept extended until all of the liquid was inhaled into the lungs. The mice were sacrificed at the indicated time points. The animal protocol was approved by the UAB Institutional Animal Care and Use Committee (IACUC).

Human recombinant TGF-β1 was purchased from PeproTech. LNA-modified control knockdown and miR-21 knockdown probes for in vivo applications were synthesized by Exiqon.

Human primary pulmonary cell lines, MRC-5 and IMR-90, as well as HEK-293 cells, were purchased from American Type Culture Collection (Manassas, VA). Cells were cultured at 37°C with 5% CO₂ in DME containing 10% FBS.

8 IPF lung tissue samples were obtained from surgical remnants of biopsies or lungs explanted from patients with IPF that underwent pulmonary transplant and 8 controls were obtained through the University of Pittsburgh Health Sciences Tissue Bank from samples resected from patients with lung cancer. The protocol was approved by the Institutional Review Board (IRB) of the University of Pittsburgh.

Total RNAs were isolated from mouse lungs harvested at d 0, 7, and 14 after bleomycin instillation (three mice per group) with the smiRNAeasy Mini kit (QIAGEN). The miRNA array was performed by Exiqon using miRCURY LNA microRNA Array (Exiqon). Array Express accession no. E-MEXP-2749.

The assay was performed as described previously (Liu et al., 2009). In brief, total RNA (10 µg) was resolved on a 12% denatured polyacrylamide gel containing 8M urea. The RNA was then transferred to a Hybond nylon membrane (GE Life Sciences). After UV cross-linking, the membrane was incubated in prehybridization buffer (50% formamide [USB], 0.5% SDS, 5xSSC [USB], 5×Denhardt’s solution [USB], and 20 µg/ml sheared, denatured, salmon sperm DNA [Invitrogen]) at 55°C for 30 min and then hybridized with specific γ[P³²]-labeled human LNA miR-21 probes (Exiqon) at 55°C for 24 h. The membrane was washed for 10 min three times with buffer (0.5% SDS, 2xSSC) and exposed to film. After hybridization with miR-21 probes, the membrane was stripped and reblotted with specific γ[P³²]-labeled mouse U6 probes (Exiqon) as loading controls.

The assay was performed as described previously (Liu et al., 2009). Taqman probes for hsa-miR-21, hsa-miR-155, and mouse Sno135 were purchased from Applied Biosystems. The expression of α-SMA, fibronectin, Col1A1, Col1A2, and TGF-β1 was determined using the SYBR Green Master Mix kit (Roche). GAPDH or HPRT was used as an internal control. The sequences of the primers: mouse HPRT: sense, 5′-GGGACATAAAAGTTATTGGTGGAGATG-3′; antisense, 5′-CAACAACAAACTTGTCTGGAATTTCAA-3′. Human fibronectin: sense, 5′-GTGTTGGGAATGGTCGTGGGGAATG-3′; antisense, 5′-CCAATGCCACGGCCATAGCAGTAGC-3′. Mouse fibronectin: sense, 5′-TCTGGGAAATGGAAAAGGGGAATGG-3′; antisense, 5′-CACTGAAGCAGGTTTCCTCGGTTGT-3′. Human α-SMA: sense, 5′-CATCACCAACTGGGACGACATGGAA-3′; antisense, 5′-GCATAGCCCTCATAGATGGGGACATTG-3′. Mouse α-SMA: sense, 5′-GACGCTGAAGTATCCGATAGAACACG-3′; antisense 5′-CACCATCTCCAGAGTCCAGCACAAT-3′. Mouse TGF-β1: sense, 5′-AGCGGACTACTATGCTAAAGAGGTCACCC-3′; antisense, 5′-CCAAGGTAACGCCAGGAATTGTTGCTATA-3′. Mouse Col1A1: sense, 5′-GGAGGGCGAGTGCTGTGCTTT-3′; antisense, 5′-GGGACCAGGAGGACCAGGAAGT-3′. Mouse Col1A2: sense, 5′-TGGTCTTACTGGGAACTTTGCTGC-3′; antisense, 5′-ACCCTGTGGTCCAACGACTCCTCTC-3′. Mouse GAPDH: sense, 5′-CGACTTCAACAGCAACTCCCACTCTTCC-3′; antisense, 5′-TGGGTGGTCCAGGGTTTCTTACTCCTT-3′. Human GAPDH: sense, 5′-GCTGGCGCTGAGTACGTCGTGGAGT-3′; antisense, 5′-CACAGTCTTCTGGGTGGCAGTGATGG-3′.

The assay was performed as described previously (Obernosterer et al., 2007). In brief, mice were sacrificed and the lungs inflated with intratracheal injection of 1 ml OCT (Thermo Fisher Scientific). The lungs were then embedded with OCT, and 10-µm-thick frozen sections prepared. The sections were dried at room temperature for 30 min, followed by fixation in 4% paraformaldehyde (Thermo Fisher Scientific) for 30 min. The sections were treated in acetylation solution for 10 min and then in PBS containing 10 µg/ml proteinase K (Sigma-Aldrich) for 5 min. Sections were then blocked with hybridization solution for 4 h at room temperature and incubated with digoxigenin (Dig)-conjugated miR-21 probes (Exiqon) or Dig-conjugated control probes with scrambled sequence (Exiqon) overnight. The sections were washed with 0.2× SSC followed by incubation with HRP-conjugated anti-Dig antibody (Roche) overnight at 4°C. After washing three times with buffer B1, the sections were developed with NBT/BCIP (Roche) for 24 h, with light blue cytoplasmic staining being positive. To determine the colocalization of miR-21 and α-SMA, FITC-conjugated miR-21 probes (Exiqon) in combination with Cy3-conjugated anti–α-SMA antibody (Sigma-Aldrich) were used in the hybridization assays. The sections were examined by confocal microscopy. miR-21 was shown as green fluorescence and α-SMA as red fluorescence.

Mouse lungs were inflated with 10% formalin solution and embedded in paraffin. 10-µm-thick sections were prepared, deparaffinized with xylene, and then rehydrated in water through graded ethanol. Antigen retrieval was performed in a pressure cooker in Tris-EDTA solution, pH 9.0, for 5 min. After incubation in TBST buffer for 10 min and 3% H₂O₂ for 10 min, the sections were washed with TBST and blocked with avidin/biotin blocker and affinity-purified goat anti–mouse IgG (H+L). The sections were then incubated with mouse anti–α-SMA (Sigma-Aldrich) overnight at 4°C, secondary antibody for 45 min, avidin-HRP for 45 min, and DAB chromagen, sequentially. Finally, the sections were counterstained with hematoxylin.

The right lungs from mice were collected and homogenized in 5 ml 0.5 M acetic acid in PBS containing 0.6% pepsin. The extracts were rotated at 4°C overnight and cleared by centrifugation at 13,200 rpm for 15 min. Collagen content was measured using the Sircol Collagen Assay kit (Biocolor Ltd.) according to the manufacturer’s instructions. Collagen content is presented as μg acid-soluble collagen/right lung.

Western blotting was performed as described previously (Liu et al., 2008). Mouse anti-Fn antibody, goat anti-Col1A1 antibodies, rabbit anti-GAPDH antibodies, and rabbit anti-Spry1 antibodies were from Santa Cruz Biotechnology, Inc. Mouse anti–α-SMA was from Sigma-Aldrich. Rabbit anti–p-Smad2, Smad-2, P-Erk, and Erk were from Cell Signaling Technology. Rabbit anti-Smad7 antibodies were from Thermo Fisher Scientific.

The full-length 3′ UTR of mouse Smad7 gene was amplified by PCR using mouse genomic DNA as a template. The primers were: sense 5′-ACTAGTACCGTTCAAACTACTTGCTGCTAACC-3′ and antisense 5′-AAGCTTTGCTAAATTTCCAATGAGAATGCTTC-3′. The PCR fragment was cloned downstream of the luciferase gene between the SpeI and HindIII sites in pMIR-Report (Applied Biosystems). To generate Smad7 3′ UTR mutants containing mutations in the conserved miR-21 binding site, site-directed mutagenesis was performed using the wild-type 3′ UTR as the template. In the 3′ UTR mutant, the nucleotide sequence complementary to nt 2–5 of miR-21 was mutated to the same sequence as that in miR-21 (from AGCT to TCGA). 0.5 µg of the luciferase reporters containing the wild-type or mutant 3′ UTR were cotransfected with control miRNA precursors or miR-21 precursors into HEK-293 cells. As an internal control, renilla luciferase reporters were also included. 2 d after transfection, the cells were collected and dual luciferase activities were measured according to the manufacturer’s instructions.

The assay was performed by the UAB Tissue Histology Core facility as described previously (Hecker et al., 2009).

The assay was performed by the UAB Tissue Histology Core facility as described previously (Hecker et al., 2009).

One-way ANOVA followed by the Holm-Sidak or Tukey-Kramer test was performed for multiple group comparisons. The Student’s t test was used for comparison between two groups. P < 0.05 was considered significant.

Fig. S1 shows enhanced expression of miR-21 in the lungs of mice with experimental lung fibrosis. Fig. S2 shows that sequestering miR-21 attenuates experimental lung fibrosis in mice. Fig. S3 demonstrates that miR-21 is induced by TGF-β1 and bFGF in human primary fibroblasts. Fig. S4 shows that miR-21 regulates Smad7 and Spry1.

G. Liu conceived of, designed, conducted, and supervised the experiments. A. Friggeri and Y. Yang contributed to mouse studies. J. Milosevic, Q. Ding, V.J. Thannickal, and N. Kaminski contributed to the data on human samples. E. Abraham designed and supervised the study. G. Liu and E. Abraham wrote the manuscript.

The authors thank Drs. Rosa Serra and Namasivayam Ambalavanan at the University of Alabama at Birmingham for supplying transgenic mice expressing dominant-negative TGF-β1RII. The authors also thank Drs. Dezhi Wang and Lingling Guo of UAB for technical assistance in mouse studies.

This work was supported by National Institutes of Health grants R21HL097218 (to G. Liu), R01HL076206 (to E. Abraham), R01GM87748 (to E. Abraham), R01LM009657 (to. N. Kaminski), and R01HL095397 (to. N. Kaminski). N. Kaminski is the Simmons Endowed Chair for Pulmonary Research.

The authors have no conflicting financial interests.