Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive parenchymal lung disease characterized by chronic tissue damage and aberrant wound healing and is associated with severe pathologic changes in the structure of the alveoli.1 There are limited effective therapeutic options for IPF and patients inevitably succumb to the disease.2, 3 Although pirfenidone and nintedanib reduce the rate at which forced vital capacity declines, their impact on long-term mortality remains unclear.

Another pulmonary disease, lung cancer is the leading cause of cancer-related death in developed countries.4 Approximately 80% of lung cancers are classified histopathologically as non-small cell lung cancers (NSCLC), and the majority of advanced NSCLC patients dying in less than 5 years despite the use of various combination chemotherapy, chemoradiotherapy or molecularly targeted therapy.5 Interestingly, NSCLC and IPF share common risk factors, such as smoking, environmental or professional exposure, viral infections, and chronic tissue injury.6, 7 Indeed, the risk of developing lung cancer is seven times higher in patients with IPF.8 These findings suggest that there are common driver genes and molecular pathways in the two diseases.

The discovery of RNAs that do not code for proteins has provided new directions in the study of human pathogenesis. In this regard, microRNAs (miRNAs) are particularly important. miRNAs are small non-coding RNAs that have pivotal roles in the regulated expression of protein coding/non-coding RNAs. Regulation of gene expression is achieved by repressing the translation of mRNAs or cleaving RNA transcripts in a sequence-specific manner.9, 10 Thus, one miRNA species can regulate the amount of a specific mRNA in human cells.10, 11 Aberrantly expressed miRNAs can disrupt the regulated RNA networks in normal cells, thereby promoting pathologic development and cancer. Many studies have shown that miRNAs are aberrantly expressed in lung cancers and in IPF.12, 13, 14, 15, 16 We hypothesized that the identification of aberrantly expressed miRNAs is an important step toward elucidating miRNA-regulated molecular targets and miRNA-mediated pathogenic RNA networks that contribute to lung cancers and pulmonary fibrosis.

Based on this proposal, we have sequentially identified tumor-suppressive miRNAs and novel miRNA-mediated RNA networks in lung cancer cells by establishing miRNA expression signatures. For example, miR-1 and miR-133a inhibit cancer cell migration and invasion through targeting colonin-1, which is involved in a variety of cellular processes. Restoration of miR-206 suppressed cancer cell aggressiveness via its regulation of EGFR and c-MET oncogenic signaling.17 More recently, we demonstrated that the miR-29 family is significantly downregulated in several types of cancers, including lung squamous cell carcinoma cells. These miRNAs act as antitumor miRNAs through their regulation of several extracellular matrix (ECM) components.18, 19, 20, 21, 22, 23 In cancer cells, aberrantly expressed ECM-mediated oncogenic signals trigger cancer cell migration and invasion and promote metastasis.24, 25 Fibrosis is enhanced by excess stiffness of ECM components, and overexpression of ECM components replaces functional tissue and disrupts organ architecture.26, 27 These observations suggest that the miR-29-family may also have important roles in the pathogenesis of pulmonary fibrosis.

In this study, we focused on miR-29a and investigated its functional significance in lung cancer and IPF. We attempted to identify the genes targeted by miR-29a and how they contributed to pathologic changes. The discovery of miR-29a-regulated genes provides new insight into the potential molecular pathogenesis of these two diseases.

Materials and methods

Clinical specimens and RNA extraction

A total of eight specimens of lung cancer with IPF and 17 normal lung tissues that obtained from lung cancer patients who underwent surgery at Kagoshima University Hospital (Kagoshima, Japan) from 2010 to 2013. IPF was diagnosed according to the official ATS/ERS/JRS/ALAT statement.2 Formalin-fixed paraffin-embedded tissue sections were used for quantitative real-time RT-PCR analysis and immunohistochemical staining.

Samples were staged according to the TNM classification for lung cancer and were histologically graded.28 Our studies were approved by the Institutional Review Board of Kagoshima University (IRB number 25–180). Prior written informed consent and approval were provided by each patient.

Total RNA (including miRNA) was extracted from tissues as previously described.18, 29 The RNA integrity was assessed using the RNA 6000 Nano Assay Kit and a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

Cell culture

The EBC-1 and A549 human lung cancer cell lines and MRC-5 fibroblast cell line were obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). The cell lines were maintained in RPMI 1640 medium or Eagle’s minimum essential medium supplemented with 10% fetal bovine serum.

Quantitative real-time PCR analysis

Quantitative Taqman real-time PCR analysis was used to evaluate lysyl oxidase-like 2 (LOXL2) and serpin peptidase inhibitor clade H, member 1 (SERPINH1) mRNA expression levels. Primers and probes were assay-on-demand gene expression products: LOXL2 (P/N: Hs00158757_m1, Applied Biosystems, Foster City, CA, USA), SERPINH1 (P/N: Hs01060397_m1, Applied Biosystems). Stem-loop RT-PCR for miR-29a (P/N: 002112, Applied Biosystems) was used to quantify the expression levels of miRNAs according to the manufacturer’s protocol. Quantitative Taqman real-time PCR relative to human GAPDH (P/N: Hs02758991_m1; Applied Biosystems) and RNU48 (P/N: 440888; Applied Biosystems) as the internal controls were performed to assess the expression levels of mRNA and miRNAs in the sample, using the delta–delta Ct method.

Restoration and inhibition of miR-29a

Transfection was performed as described previously.17, 29 We used Pre-miR-29a miRNA precursors (P/N: AM 17100; Applied Biosystems), miR-29a inhibitor (P/N: AM 4464084; Applied Biosystems) and negative-control miRNA (P/N: AM 17111; Applied Biosystems).

Cell proliferation, migration and invasion assays

Cell proliferation, migration and invasion assays were performed as described before.18 All experiments were performed in triplicate.

Identification of putative target genes of miR-29a

TargetScan database (http://www.targetscan.org) was used to identify miR-29a target genes. In addition, we performed gene expression analysis using miR-29a-transfected cells (EBC-1, A549 and MRC-5). Oligo-microarray human 60Kv (Agilent Technologies) was used for gene expression studies. To investigate the expression status of candidate miR-29a target genes in NSCLC clinical specimens, we analyzed gene expression profiles in the gene expression omnibus database (accession number: GSE 19188). Microarray procedures and data mining methods were described previously.18, 29

Western analysis

Expression of LOXL2 and SEPINH1 protein in vitro was assessed using western analysis as described previously.17, 29 Immunoblotting was done with anti-LOXL2 antibody (1:1000; ab96233; Abcam, Cambridge, UK) or with anti-SERPINH1 antibody (1:500; sc-5293; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-GAPDH antibody (1:10000; MAB374; Chemicon, Temecula, CA, USA).

Plasmid construction and dual-luciferase reporter assays

Partial wild-type sequences of the 3′-untranslated region (UTR) of LOXL2 and SERPINH1 or those with a deleted miR-29a target site (positions 555–561 or 757–763 of LOXL2 3′-UTR and 57–63 of SERPINH1 3′-UTR) were inserted between the XhoI and PmeI restriction sites in the 3′-UTR of the hRluc gene in the psiCHECK-2 vector (C8021; Promega, Madison, WI, USA). Dual-luciferase reporter assay was performed as described previously.17, 29

Immunohistochemistry

Formalin-fixed paraffin-embedded tissue sections were treated as previously described.17 Anti-LOXL2 rabbit polyclonal antibody (ab96233; Abcam) and anti-SERPINH1 mouse monoclonal antibody (sc-5293; Santa Cruz Biotechnology) were used.

Statistical analysis

All data are shown as mean±s.e. Statistical comparison was made by analysis of variance, followed by Bonferroni’s post hoc analysis. Expert StatView software, version 4 (SAS Institute, Cary, NC, USA), was used in these analyses.

Results

Expression levels of miR-29a in lung cancer, IPF and normal lung tissues

The expression level of miR-29a was evaluated by quantitative real-time PCR methods by using formalin-fixed paraffin-embedded specimens of lung cancer lesions, fibrotic lesions and normal lung tissues. Representative formalin-fixed paraffin-embedded specimens are shown in Figures 1a and b. The patient backgrounds and clinicopathological characteristics are summarized in Tables 1A and 1B. The expression level of miR-29a was significantly reduced in lung cancer and fibrotic lesions compared with normal lung tissues (Figure 1c).

Figure 1
figure 1

Comparison of expression levels of miR-29a in clinical specimens and cell lines. (a) Typical specimen of lung cancer with idiopathic pulmonary fibrosis (IPF). (b) Normal lung tissue. (c) Expression levels of miR-29a in lung cancer, IPF, normal lung tissues and cell lines. *P<0.05, **P<0.005. A full color version of this figure is available at the Journal of Human Genetics journal online.

Full size image
Table 1a Characteristics of the patients
Full size table
Table 1b Normal lung tissue
Full size table

Effects of miR-29a restoration on the proliferation, migration and invasion of NSCLC cell lines and a fibroblast cell line

To evaluate the functional roles of miR-29a, we carried out gain-of-function studies using mature miRNA transfection into two NSCLC cell lines (EBC-1 and A549) and a fibroblast cell line (MRC-5).

The XTT assays revealed inhibition of cell proliferation in EBC-1, A549 and MRC-5 cells transfected with miR-29a in comparison with mock- or control-transfected cells (Figure 2a). The wound-healing assays showed significant inhibition of cell migration activities in miR-29a transfectants (EBC-1, A549 and MRC-5) compared with that in mock- or miR-control-transfected cells (Figure 2b). Moreover, Matrigel invasion assays demonstrated that miR-29a transfection significantly inhibited cell invasion. The number of invading cells was significantly reduced in miR-29a transfected EBC-1 and A549 cells (Figure 2c). In this assay, we omitted examination of MRC-5 cells because the ability for invasion of this cell line is extremely low.

Figure 2
figure 2

Effects of miR-29a transfection on EBC-1, A549 and MRC-5 cells. (a) Cell proliferation was determined with XTT assays 96 h after transfection with 10 nM miR-29a, miR-control or mock. (b) Cell migration activity was determined by migration assay 48 h after transfection. (c) Cell invasion activity was determined by Matrigel invasion assay 72 h after transfection. *P<0.05.

Full size image

In contrast to restoration study, inhibition of miR-29a enhanced EBC-1 cell proliferation, migration and invasion (Supplementary Figure 1).

Identification of putative miR-29a regulated genes in lung cancer and fibroblast cells

To identify miR-29a-regulated genes, we carried out a combination of in silico analysis and genome-wide gene expression analysis of miR-29a transfectant cells (EBC-1, A549 and MRC-5). By using TargetScan database analysis, a total of 3538 genes were identified as putative miR-29a-regulated genes, that is, those that had putative target sites for miR-29a in their 3′-UTRs. We undertook genome-wide gene expression analysis using EBC-1, A549 and MRC-5 cells. Among the 3538 genes, we selected a total of 34 genes that were commonly downregulated in EBC-1, A549 and MRC-5 cells (average log2 ratio <−1.0) following miR-29a transfection compared with the expression levels in mock cells. Finally, we investigated the expression status of those 34 genes in lung cancer specimens by using the gene expression omnibus database (accession number: GSE 19188). From this selection, 24 candidate genes were identified as targets of miR-29a (Table 2).

Table 2 Downregulated genes in miR-29a transfectant
Full size table

Extracellular matrix has a key role in the progression of cancer and fibrosis (24–27). Therefore, we focused on the ECM-related genes.

Among these candidate genes, we chose LOXL2 and SERPINH1 genes that code for collagen cross-linking enzymes and those genes were subjected to further analyses. Our method for selecting miR-29a target genes is shown in Figure 3.

Figure 3
figure 3

Flow chart illustrating the strategy for identification of miR-29a target genes. In total, 3538 genes were identified as putative targets of miR-29a according to the TargetScan database. We merged the expression analysis data of downregulated genes in miR-29a-transfected EBC-1, A549 and MRC-5 cells (Log2 ratio <−1.0). Upregulated genes were determined according to the gene expression data set of lung NSCLC clinical specimens according to the GEO database (accession no. GSE 19188). From this selection, 24 candidate genes were identified as putative targets of the miR-29a.

Full size image

Direct regulation of LOXL2 and SERPINH1 by miR-29a in lung cancer cells and fibroblasts

We conducted qRT-PCR and western blotting to confirm that restoration of miR-29a resulted in downregulation of LOXL2 and SERPINH1 in EBC-1, A549 and MRC-5 cells. The mRNA and protein expression levels of LOXL2 and SERPINH1 were significantly repressed in miR-29a transfectants compared with mock or miR-control transfectants in EBC-1, A549 and MRC-5 (Figures 4a and b). In contrast to restoration study, inhibition of miR-29a resulted in upregulation of LOXL2 and SERPINH1 in these cells (Supplementary Figure 2).

Figure 4
figure 4

Effects of miR-29a transfection on LOXL2 and SERPINH1 expression in EBC-1, A549 and MRC-5 cells. (a, bar graph) LOXL2 mRNA expression 72 h after transfection with miR-29a. GAPDH was used as an internal control. (a, blot) LOXL2 protein expression 72 h after transfection with miR-29a. GAPDH was used as a loading control. (b, bar graph) SERPINH1 mRNA expression 72 h after transfection with miR-29a. (b, blot) SERPINH1 protein expression 72 h after transfection with miR-29a. *P<0.0001.

Full size image

Next, we performed dual-luciferase reporter assays in A549 cells to confirm whether LOXL2 and SERPINH1 mRNAs were directly regulated by miR-29a. The TargetScan database predicted that putative miR-29a-binding sites existed in the 3′-UTRs of LOXL2 (positions 555–561 and 757–763; Figure 5a) and SERPINH1 (position 57–63; Figure 5b). We used vectors encoding either the partial wild-type sequence of the 3′-UTR of LOXL2 or SERPINH1 mRNA, including the predicted miR-29a target sites, or deletion vectors lacking the miR-29a target sites. We found that the luminescence intensities were significantly reduced by transfection with miR-29a and vectors carrying the wild-type 3′-UTR of LOXL2 and SERPINH1, whereas transfection with deletion vectors blocked the decrease in luminescence (Figure 5). These data suggested that miR-29a bound directly to specific sites in the 3′-UTRs of LOXL2 and SERPINH1 mRNA.

Figure 5
figure 5

Direct regulation of LOXL2 and SERPINH1 by miR-29a in A549 cells. (a) The miR-29a-binding site in the 3′-UTR of LOXL2 mRNA. Luciferase reporter assays used vectors that included (WT) or lacked (DEL) wild-type sequences of the putative miR-29a target sites at position 555–561 and 757–763. Renilla luciferase values were normalized to firefly luciferase values. (b) The miR-29a-binding site in the 3′-UTR of SERPINH1 mRNA. Luciferase reporter assays used vectors that included (WT) or lacked (DEL) wild-type sequences of the putative miR-29a target sites at position 57–63. *P<0.001.

Full size image

Expression of LOXL2 and SERPINH1 in clinical specimens of lung cancer with IPF

We used immunohistochemical staining to examine the expression status of LOXL2 and SERPINH1 in clinical specimens of lung cancer with associated IPF (Figures 6a and b). LOXL2 was strongly expressed in cancer cells and fibroblasts (Figures 6c and d). Similar overexpression of SERPINH1 was observed in cancer cells and fibroblasts (Figures 6e and f).

Figure 6
figure 6

Immunohistochemical staining of LOXL2 and SERPINH1 in clinical specimens of lung cancer with IPF. (a) Hematoxylin-eosin staining. (b) Masson’s trichrome staining. Cancer lesions (thin arrow) and fibrotic lesions (thick arrow) are noted. (c, d) LOXL2 staining. Cancer cells (open arrowheads) and fibroblasts (arrowheads) are noted. (e, f) SERPINH1 staining. Cancer cells (open arrowheads) and fibroblasts (arrowheads) are noted. A full color version of this figure is available at the Journal of Human Genetics journal online.

Full size image

Discussion

Our present study showed that expression of miR-29a was significantly reduced in lung cancer tissues and in fibrotic lesions of IPF and that restoration of miR-29a significantly suppressed proliferation, migration and invasion in cancer cells and fibroblasts. Those data suggested that miR-29a and miR-29a-regulated genes were deeply involved in these diseases. Our recent studies of cancer cells showed that all members of the miR-29-family were downregulated in several types of cancer tissues and that restoration of these miRNAs inhibited cancer cell aggressiveness through targeting collagen biosynthesis genes, laminins and integrins.30 Most types of cells require ECM-integrin-mediated signaling pathways for proliferation, migration, invasion and survival.31, 32 Aberrant expression of ECM-related genes disturbs tightly regulated integrin-meditated signaling and these phenomena promote cancer cell aggressiveness.30 Thus, elucidation of tumor-suppressive miRNAs that regulate novel ECM-related pathways should point to new therapies for cancer.

Recent studies have indicated that dysregulated miRNAs contribute to the disruption of normal cellular RNA networks, promoting the development of cancer cells and associated pathology. We have identified novel miRNA-regulated cancer pathways based on the aberrant expression of miRNAs.17, 18, 33 In this study, we identified miR-29a-regulated genes that participate in the pathogenesis of lung cancer and IPF.

Basically, the function of LOXL2 is covalent cross-linking of collagen and/or elastin in the ECM.34, 35 Recent studies of cancer cells showed that aberrant expression of LOXL2 was correlated with disease progression and aggressiveness in several types of cancers.35 Our present data indicated that overexpression of LOXL2 enhanced cancer cell migration and invasion in several cancers.18, 36, 37 SERPINH1 (also known as HSP47) is a 47 kDa collagen-binding glycoprotein localized in the endoplasmic reticulum (ER) and as a collagen-specific molecular chaperon.38 Many studies have reported that SERPINH1 is overexpressed in cancers and fibrotic diseases.39, 40, 41 To our surprise, our past studies demonstrated that LOXL2 was involved in cancer cell metastasis and directly regulated by the miR-29-family in renal cell carcinoma, head and neck cancer and lung squamous cell carcinoma.18, 36, 37 Moreover, SERPINH1 was regulated by miR-29a in cervical cancer.23 Our present data showed that both LOXL2 and SERPINH1 were directly regulated by miR-29a in lung cancer cells and fibroblasts. These findings indicated that direct regulation of LOXL2 and SERPINH1 by miR-29a is widespread and indispensable for the normal function of the cells.

Fibrosis is caused by excess accumulation of ECM components, such as collagens, fibronectins, elastin and alpha-smooth muscle actin. ECM-related genes also contribute to the development of fibrosis.26, 27, 42 Recent studies reported that members of the miR-29-family have potential roles in the inhibition of ECM synthesis, indicating anti-fibrotic activity.43, 44 Interestingly, the search for miR-29 family targets by in silico analysis revealed that many collagen genes and several ECM component genes were putative targets of miR-29. Downregulation of the miR-29 family was demonstrated in several types of human fibrotic tissues, including IPF,13, 15, 45, 46 supporting our present findings. It would be important to confirm the downregulation of miR-29a in fibroblasts and cancer cells in this study. The in situ hybridization to detect the reduction of miR-29a signals in these cells will be examined in future studies.

Regulation of miR-29a expression was investigated by chromatin immunoprecipitation assay and in silico analysis.47 The promoter region of miR-29a contained two putative E-box sites, a Gli-binding site and four nuclear factor-kB-binding sites. These transcription factors bound to these regions and negatively regulated miR-29a expression.47 Moreover, TGF-β inhibited the expression of the miR-29 family and induced the expression of ECM components.46, 48 We know that TGF-β has a pivotal role in the pathogenesis of IPF by promoting the proliferation, activation and differentiation of epithelial cells and myofibroblasts.1, 49 Therefore, suppression of the miR-29 family by TGF-β is a possible critical event in the progression of IPF, and inhibition of these pathways might be a therapeutic target of the disease.

In conclusion, downregulation of miR-29a was observed in clinical specimens of lung cancer and IPF, and miR-29a might contribute to these diseases by targeting LOXL2 and SERPINH1. Overexpression of LOXL2 and SERPINH1 was confirmed in clinical specimens of lung cancer with IPF and they enhance the aggressiveness of these diseases. The identification of novel pathways regulated by the downregulation of miRNA may lead to a better understanding of molecular pathogenesis in lung cancer and pulmonary fibrosis.