Silencing MIAT reduces the myofibroblastic properties of fBMFs

To test our presumption, we used a lentiviral vector-mediated short hairpin (sh) RNA targeting MIAT to knock down the expression of MIAT in fBMFs (Figure 2A) and examined the effect of MIAT on myofibroblast phenotypes. First, a collagen gel contraction assay was conducted to assess the contractile forces generated by myofibroblasts, which propagated throughout the collagen matrix and resulted in decreased matrix size. As shown in Figure 2B, the downregulation of MIAT markedly attenuated the collagen gel contractility of fBMFs. Another feature of myofibroblasts is that they proliferate and migrate to the damaged site during wound healing. By using Transwell migration and scratch (wound healing) assays, we demonstrated that suppression of MIAT significantly inhibited the cell migration (Figure 2C) and wound-healing ability (Figure 2D) of fBMFs.

Open in a separate window

Figure 2

Knockdown of MIAT suppresses the myofibroblastic properties. (A–D) Primary fBMFs (obtained from two patients with OSF; fBMFs−1 and −2) were transfected with lentiviruses expressing non-targeting ShRNA (Sh-Luc.) and Sh-MIAT (Sh-MIAT−1 and −2). The MIAT knockdown efficiency was assessed using qRT-PCR analysis (A). The cells (fBMFs−1 and −2) were then cultured in collagen gel for additional 48 hours, and the gel area after cell contraction was measured (B). Cells (fBMFs−1 and −2) were cultured in Transwell system for an additional 24 hours, and their migration ability was quantified (C). Data are presented as mean ± SD (n=3); *p < 0.05 vs. Sh-Luc. (A–C). Confluent monolayers of fBMFs−2 were scratched and cultured for 48 hours, and the wound closure was assessed. Scale bar, 50 μm (D).

Loss of MIAT inhibits the arecoline-induced myofibroblast transformation of nBMFs

We previously demonstrated that arecoline treatment can effectively induce nBMFs to acquire myofibroblastic properties [15, 25–27]. Thus, to understand whether MIAT is involved in the transformation of fibroblasts to myofibroblasts, we assessed the expression of MIAT in nBMFs after treatment with 0, 5, 10, and 20 μg/mL arecoline for 24 hours. RNA sequencing analysis showed that MIAT levels were up-regulated in nBMFs treated with 10 and 20 μg/mL arecoline (Figure 3A). Also, results of qPCR analysis demonstrated that arecoline gradually increased MIAT expression in a dose-dependent manner (Figure 3B). Based on our previous findings [15, 25–27], in this study, nBMFs were cultured in medium containing 20 μg/mL arecoline for 24 hours. This treatment induced an increase in α-SMA expression (Figure 3D and Supplementary Figure 1), contractility (Figure 3E), and cell migration (Figure 3F), collectively confirming the acquisition of myofibroblast-like properties. As expected, the silencing of MIAT in the arecoline-treated nBMFs (Figure 3C) successfully attenuated the expression of α-SMA (Figure 3D), and significantly reduced both collagen gel contractility (Figure 3E) and Transwell migration capacity (Figure 3F). These results implied that downregulation of MIAT may prevent myofibroblast transformation and potentially mitigate the progression of OSF.

Open in a separate window

Figure 3

Knockdown of MIAT impairs the arecoline-induced myofibroblastic transformation in BMFs. (A) Primary nBMFs (obtained from two healthy individuals; nBMFs−1 and −2) were cultured with arecoline (0, 10, and 20 μg/mL) for 24 hours, followed by RNA-sequencing analysis to determine the levels of MIAT (p < 0.01). (B) Normal BMFs (−1 and −2) were cultured with arecoline (0, 5, 10, and 20 μg/mL) for 24 hours, followed by qRT-PCR analysis to determine the MIAT expression. (C–F) Normal BMFs (−1 and −2) were transfected with lentiviruses expressing non-targeting ShRNA (Sh-Luc.) and Sh-MIAT (Sh-MIAT−1 and −2). After 48 hours, the cells were cultured with or without arecoline (20 μg/mL) for 24 hours for the induction of myofibroblasts transdifferentiation. The expression of MIAT in each group was assessed using qRT-PCR analysis. Data are presented as mean ± SD (n=3); *p < 0.05 vs. Sh-Luc.; #p < 0.05 vs. Sh-Luc. with arecoline treatment (C). The protein expression of α-SMA in each group was determined using Western blotting analysis (D). Cells (nBMFs−1 and −2) were cultured in collagen gel for an additional 48 hours, and the resulting gel area after cell contraction was measured (E). Cells (nBMFs−1 and −2) were cultured in Transwell system for an additional 24 hours, and their migration ability was quantified (F). Data are presented as mean ± SD (n=3); *p < 0.05 vs. Sh-Luc.; #p < 0.05 vs. Sh-Luc. with arecoline treatment; Scale bar, 50 μm (E, F).

MIAT promotes myofibroblastic properties by acting as a sponge of miR-342-3p

Recent studies focused on the biological roles of lncRNAs as miRNA sponges. Due to the miRNA-mediated post-transcriptional suppression of targeted mRNA occurs in the cytoplasm, determining the subcellular localization of MIAT is precedence. Herein, we showed that MIAT was preferentially located in the cytoplasm of fBMFs (Figure 4A). By using bioinformatics prediction tools, several putative miRNA binding sites contained in MIAT sequences were identified, including miR-342-3p. Since a recent study has demonstrated the interplay between MIAT and miR-342-3p in retinal pericytes [28] and miR-342-3p was downregulated in OSF tissues using RNA sequencing analysis (Figure 4B), we then investigated whether miR-342-3p mediated the fibrosis effect of MIAT. Bioinformatics analysis predicted that MIAT 3’UTR sequence had a potential binding region for miR-342-3p (Figure 4C). To confirm the significance of miR-342-3p in OSF, the qRT-PCR analysis was conducted to evaluate its expression in clinical specimens. As expected, miR-342-3p was significantly decreased in OSF tissues (Figure 4D). Notably, we also found a significantly negative correlation between MIAT expression and miR-342-3p levels in OSF tissues (r = − 0.931, p<0.05, Figure 4E). Results demonstrated that the miR-342-3p mimics evidently lowered the activity in wt-MIAT group but had no effect in mut-MIAT group (Figure 4F). Furthermore, we showed fBMFs with miR-342-3p mimics exhibited lower myofibroblast features, including collagen gel contraction, Transwell migration and wound healing abilities (Figure 4G–4I).

Open in a separate window

Figure 4

MiR-342-3p negatively correlates to MIAT expression in OSF tissues and acts as an anti-fibrotic miRNA in OSF. (A) RNA of cytoplasmic and nuclear fractions from primary fBMFs (−1 and −2) were analyzed by qRT-PCR to determine the subcellular localization of MIAT. Data are presented as mean ± SD (n=3); *p < 0.05 vs. Nucleus. (B) RNA-sequencing analysis showed that miR-342-3p was a down-regulated differentially expressed gene (fold change ≤ -2.0; p < 0.01) in samples of fibrotic tissues (OSF; n=2) compared to normal tissues (N; n=2). (C) An illustration of the predicted pairing region between miR-342-3p and MIAT 3’UTR, discovered using the miRanda database, and the 3’ UTR regions of full-length (wt-MIAT) and mutated MIAT (mut-MIAT) complementarity to the seed site of miR-342-3p, predicted by TargetScan in silico browser. (D) The relative expression of miR-342-3p in normal (N; n=25) and fibrotic (OSF; n=25) tissues was assessed by qRT-PCR analysis. Data are presented as mean ± SD. (E) A significant negative correlation was observed between MIAT and miR-342-3p expression fibrotic tissue samples (OSF; n=45). (F) Fibrotic BMFs (−1) were co-transfected with either miR-Scramble (miR-Src.) or miR-342-3p mimics, along with the indicated pmirGLO-based constructs shown in (C). Luciferase reporter activity was measured 24 hours post-transfection. Data are presented as mean ± SD (n = 3); *p < 0.05 vs. wt-MIAT with miR-Src. (G–I) Fibrotic BMFs (−1 and −2) expressing Sh-Luc or Sh-MIAT were transfected with either miR-Src or miR-342-3p inhibitor for 24 hours. The cells (fBMFs−1 and −2) were cultured in collagen gel for an additional 48 hours, and the resulting gel area after cell contraction was measured (G). Cells (fBMFs−1 and −2) were cultured in Transwell system for an additional 24 hours, and their migration ability was quantified (H). Data are presented as mean ± SD (n=3); *p < 0.05 vs. miR-Scr. (G, H). Confluent monolayers of fBMF−2 were scratched and cultured for an additional 48 hours, and the wound closure was assessed (I). Scale bar, 50 μm.

Given that miR-342-3p is negatively correlated with MIAT and that adding this miRNA to fBMFs inhibited myofibroblast differentiation, it is crucial to investigate whether MIAT directly sponges miR-342-3p in its exertion of fibrotic trait. Since the evasion of apoptosis by myofibroblasts is a hallmark of fibrotic disorders, understanding this interaction is essential [10]. Thus, we measured the percentage of apoptotic cells in fBMFs infected with MIAT shRNA along with miR-342-3p inhibitor. Our results showed that transfection of miR-342-3p inhibitor into fBMFs successfully avoided cell apoptosis induced by silencing of MIAT (Figure 5A). Similarly, miR-342-3p inhibitor counteracted the effect of silencing of MIAT on collagen gel contraction (Figure 5B), protein expression of α-SMA and COL1A1 (Figure 5C and Supplementary Figure 2), and cell migration (Figure 5D). Taken together, these data indicate that MIAT functions as a miR-342-3p sponge to promote myofibroblast properties in fBMFs.

Open in a separate window

Figure 5

The silencing of MIAT induces apoptosis and inhibits myofibroblastic properties by targeting miR-342-3p. (A–D) Fibrotic BMFs (fBMFs−1 and −2) expressing the Sh-Luc. or Sh-MIAT were transfected with either miR-Scramble (miR-Src.) or miR-342-3p inhibitor for 24 hours. Cell apoptosis (annexin V⁺ or annexin V⁺/PI⁺) was assessed using flow cytometry (A). Cells (fBMFs−1 and −2) were cultured in collagen gel for an additional 48 hours, followed by the measurement of the gel area after cell contraction (B). Data are presented as mean ± SD (n=3); *p < 0.05 vs. Sh-Luc.; #p < 0.05 vs. Sh-MIAT with miR-Src. (A, B). The protein expression of α-SMA and COL1A1 in fBMFs−1 was analyzed using Western blotting (C). Cells (fBMFs−1 and −2) were cultured in Transwell system for an additional 24 hours, and their migration ability was quantified. Data are presented as mean ± SD (n=3); *p < 0.05 vs. Sh-Luc.; #p < 0.05 vs. Sh-MIAT with miR-Src.; Scale bar, 50 μm (D).

MiR-342-3p mitigates myofibroblast activation by suppressing SOX6

Several studies have shown that SRY-box transcription factor 6 (SOX6) may be a target of miR-342-3p in renal or cardiac injuries [29, 30] and implicate in renal fibrosis [29, 31], so we sought to examine whether SOX6 plays a role for the anti-fibrosis effect of miR-342-3p during oral fibrogenesis. By using RNA sequencing analysis, we showed that SOX6 was upregulated in OSF tissues compared to normal specimens (Figure 6A). We also predicted a potential miR-342-3p binding region on the SOX6 3’UTR sequence via miRDB bioinformatic analysis (Figure 6B), and the results from a luciferase assay showed that the luciferase activity was significantly reduced in wild-type SOX6 in nBMFs (wt-SOX6), while there was no change in mutated SOX6 (mut-SOX6, Figure 6C). Besides, the expression of SOX6 in fBMFs was downregulated after transfection of miR-342-3p mimics (Figure 6D and Supplementary Figure 3). Additionally, we showed that overexpression of SOX6 in nBMFs (Figure 6E and Supplementary Figure 4) elicited multiple myofibroblast phenotypes, including higher collagen gel contractility (Figure 6F), Transwell migration (Figure 6G) and wound healing (Figure 6H) capacities. These results supported that miR-342-3p suppresses myofibroblastic properties by targeting SOX6.

Open in a separate window

Figure 6

SOX6 is a target of miR-342-3p. (A) RNA-sequencing analysis showed that SOX6 was an up-regulated differentially expressed gene (fold change ≥ 2.0; p < 0.01) in fibrotic tissues samples (OSF; n=2) compared to normal tissues samples (N; n=2). (B) An illustration of the predicted pairing region between miR-342-3p and SOX6 3’UTR were discovered using the miRDB database, and the 3’ UTR regions of full-length (wt-SOX6) and mutated SOX6 (mut-SOX6) complementarity to the seed site of miR-342-3p, predicted by the TargetScan in silico browser. (C) Fibrotic BMFs (−1) were co-transfected with either miR-Scramble (miR-Src.) or miR-342-3p mimics, along with the indicated pmirGLO-based constructs shown in (B). Luciferase reporter activity was measured 24 hours post-transfection. Data are presented as mean ± SD (n = 3); *p < 0.05 vs. wt-SOX6 with miR-Src. (D) Fibrotic BMFs (−1 and −2) were transfected with either miR-Src or miR-342-3p inhibitor for 24 hours, followed by Western blotting analysis to determine the protein expression of SOX6. (E–H) Normal BMFs (−1 and −2) were transfected with lentiviruses expressing control vector (Ctrl.) or SOX6 (ov-SOX6). The overexpression efficiency of SOX6 was assessed by Western blotting analysis (E). Cells (nBMFs−1 and −2) were cultured in collagen gel for an additional 48 hours, and the resulting gel area after cell contraction was measured (F). Cells (nBMFs−1 and −2) were cultured in Transwell system for an additional 24 hours, and their migration ability was quantified (G). Confluent monolayers of cells (nBMFs−1 and −2) were scratched and cultured for an additional 48 hours, and the wound closure was assessed (H). Data are presented as mean ± SD (n=3); *p < 0.05 vs. Ctrl. (F–H). Scale bar, 50 μm.

Cell culture

Fibrotic buccal mucosa fibroblasts (fBMFs) and normal BMFs (nBMFs) were extracted from buccal mucosa tissues of OSF patients and healthy individuals, respectively. The details of primary cell isolation and culture were previously described [16]. In brief, buccal mucosa tissues obtained from surgery were incubated in Hanks’ Balanced Salt Solution (HBSS) at 4° C and transferred to the laboratory immediately for further processing. After trypsinization, the tissues were cultured in DMEM medium (with 10% fetal bovine serum [FBS], and 1% penicillin-streptomycin cocktail) and plated into 25-T flasks for 14 days. The cells which are spindle-shaped and migrated out of the tissues were defined as buccal fibroblasts. All cells were continuously passaged and used for subsequent experiments between the 3rd and 8th passages and were negative for mycoplasma contamination verified using short tandem repeat (STR) DNA profiling. Arecoline was purchased from Sigma (St. Louis, MO, USA). Arecoline was dissolved in phosphate-buffered saline (PBS) as a stock solution stored at -20° C before use. When nBMFs reached a designated density, cells were treated with arecoline at a series concentration (0-20 μg/mL) for 24 hours.

RNA-sequencing analysis

Total RNA was extracted from the collected buccal tissue samples and primary BMF cells using Trizol reagent as directed by the manufacturer (Invitrogen, Carlsbad, CA, USA). RNA quality and quantity were evaluated using NanoDrop (Thermo Fisher Scientific, Waltham, MA, USA). All procedures of RNA library preparation, sequencing, and transcriptome discrepancies analysis were performed at Genomics Inc. The detail was described in our previous studies [57]. To verify the RNA-sequencing results, the selected differentially expressed gene expressions were further confirmed by qRT-PCR analysis.

QRT-PCR analysis

The methods of RNA extraction, quality control and quantitation are described above. For MIAT subcellular localization detection, the nuclear and cytoplasmic RNA from fBMFs were isolated using a PARIS™ Kit (Thermo Fisher Scientific). cDNA was prepared using The Superscript III first-strand synthesis system (Invitrogen) to reverse-transcribe RNA. ABI StepOne™ Real-Time PCR Systems (Applied Biosystems, Waltham, MA, USA) were used for PCR experiments with the resultant cDNAs. The specific primer sequences are listed as follows (5’-3’): MIAT, TATTTGCAGGGGGTGCTCTG (forward), GGGCAGGGGGTCTAACTCTA (reverse); α-SMA, AGCACATGGAAAAGATCTGGCACC (forward), TTTTCTCCCGGTTGGCCTTG (reverse); COL1A1, GATTCCCTGGACCTAAAGGTGC (forward), AGCCTCTCCATCTTTGCCAGCA (reverse); FN1, ACTGCGAGAGTAAACCTGAAGC (forward), GCGGTTTGCGATGGTACAGCT (reverse); miR-342-3p, GTGCTATCTGTGATTGAGGGA (forward), CGGGTGCGATTTCTGTG (reverse); SOX6, GTCGCTTAATGTGTGGCTCG (forward), TGTTCTTCCTGCCCTGACATT (reverse); GAPDH, GTGGCTGGCTCAGAAAAAGG (forward), GGGGAGATTCAGTGTGGTGG (reverse); U6, CTCGCTTCGGCAGCACA (forward), AACGCTTCACGAATTTGCGT (reverse). Relative expression levels were normalized using GAPDH as the internal control for total and cytoplasmic RNA, while U6 served as the internal control for nuclear RNA and miRNA normalization, and the 2-ΔCt method was applied.

Knockdown of MIAT

A lentiviral pLKO-shRNA-expressing vector was purchased from the National RNAi Core Facility (Academia Sinica, Taipei, Taiwan). Lentivirus particle production was performed in 293T cells by co-transfecting the pLKO-shRNA-expressing vectors with packaging and envelope plasmids with a ratio of 10:10:1 using Lipofectamine 2000 reagent (Invitrogen). The virus-containing supernatant was collected after 48 post-transfection and filtered through a 0.45-μm PVDF filter. Fibrotic BMFs were cultured with the filtered virus-containing supernatant supplemented with 8 μg/mL polybrene (Merck, Darmstadt, Germany) for 48 hours. The lentivirus-infected cells were selected by puromycin. Quantitative RT-PCR analysis was performed to verify the knockdown efficiency. The sequences for Sh-MIAT are listed below: Sh-MIAT-1, AAAAGCAGTCCAGGGTCTATTTATTGGATCCAATAAATAGACCCTGGACTGC; Sh-MIAT-2, AAAAGGGTTTGAACCTTTAGGATTTGGATCCAAATCCTAAAGGTTCAAACCC. A pLKO-luciferase-expressing vector was used as the non-targeting control (Sh-Luc.).

Collagen gel contraction assay

A total of 0.5 mL of the collagen solution (2 mg collagen/mL) was mixed with 1 × 10⁵ cells and added into each well of a 24−well plate. After gel polymerization by 30 minutes of incubation at 37° C, 0.5 mL DMEM medium was added to cover the polymerized gels entirely. Then the gels were scraped out and plated into another 24−well plate for 48 hours of incubation at 37° C. Images of the contracted collagen gels were captured and quantified using ImageJ software (NIH). The area of the contracted gel was measured and compared to the initial area to calculate the percentage of contraction. A decrease in relative gel area represents an increase in the gel contraction ability of cells.

Transwell migration assay

A total of 0.4 mL of a serum-free DMEM containing 1 × 10⁵ cells was added into the upper 8-μm pore insert, then the lower chamber was filled with 0.6 mL DMEM containing 10% FBS. After incubation at 37° C for 24 hours, non-migrated cells on the upper side of the insert membrane were removed, and the migrated cells on the lower side were fixed with cold-100% methanol for 30 minutes at room temperature. The fixed cells were stained with 0.1% crystal violet solution for 20 minutes. Migrated cells were visualized by capturing images using a microscope. For the evaluation of migrated cells, 0.4 mL of 30% acetic acid was added to each insert and incubated for 15 minutes to dissolve the crystal violet stain. The absorbance value at 570 nm of crystal violet stain from each insert was measured using a microplate reader (Infinite M200 Pro, Tecan Group Ltd., Männedorf, Germany) and quantified according to the manufacturer’s protocols.

Wound healing assay

After cell confluence reached 90%, a straight wound was introduced across the center of each well using a 200 μL sterile pipette tip. Each well was washed with PBS to remove any detached cells and debris, and cells were cultured with a serum−free DMEM medium for 48 hours. Images of the wound area were captured at the time points of 0 and 48 hours by microscope.

Western blot analysis

In brief, whole-cell lysate containing 20 μg of protein was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then transferred onto a polyvinylidene fluoride membrane. After blocking for non-specific sites with 5% bovine serum albumin (BSA), the membrane was incubated with primary antibodies for 16 hours at 4° C, and subsequently incubated with the corresponding HRP-conjugated secondary antibody for 1 hour at room temperature. The signals of antibody binding were developed by ECL substrate (Merck) and then captured using the LAS−1000 plus analyzer (GE Healthcare, Piscataway, NJ, USA). GAPDH was used as an internal reference. All the antibodies were purchased from Cell Signaling Inc. (Danvers, MA, USA).

MiR-342-3p knockdown and overexpression

MiR-342-3p mimics, miR-342-3p inhibitor, and miR-scramble negative control (miR-Scr.) were purchased from Applied Biosystems (Waltham, MA, USA), and a total of 100 nM oligonucleotide sequences were transfected into cells using Lipofectamine 2000 (Invitrogen). After 48 hours of transfection, cells were collected for subsequent experiments.

Luciferase reporter activity assay

The binding of miR-342-3p to MIAT or SOX6 was examined using a luciferase reporter activity assay (Promega, Madison, WI, USA). MIAT and SOX6 3’-UTR with full-length (wild-type, wt) and mutant-type (mut) were synthesized and cloned into pmirGLO vectors according to the manufacturer’s protocols. Luciferase vectors were mixed with miR-342-3p mimics and miR-scramble (miR-Src.), respectively, and the mixtures were then co-transfected into cells using Lipofectamine 2000 reagent. The activity of luciferase reporter was detected using the Luciferase Assay System according to the manufacturer’s instructions. The relative activity of each group was normalized with Renilla luciferase internal control (VA).

Apoptosis assay

Briefly, 2 × 10⁵ cells were incubated with stained with 1.5 μl Annexin V-FITC reagent and 1.5 μl propidium iodide (PI) for 5 minutes at room temperature in the dark. The stained cells were then directly measured using a Calibur flow cytometer (Becton Dickinson, San Jose, CA, USA). Data were analyzed using FlowJo software v10 (FlowJo LLC, Ashland, OR, USA). The Annexin single-positive and Annexin/PI double-positive populations were considered apoptotic cells.

Overexpression of SOX6

The SOX6 cDNA was cloned into the pCDHI-MCS1-EF1-CopGFP plasmid (System Biosciences, Mountain View, CA, USA). The pCDH and two helper plasmids (packaging and envelope plasmids) were co-transfected into 293T cells using Lipofectamine 2000 reagent to produce lentiviral particles. The conditions for lentivirus production and infection were as described in the Knockdown of MIAT section. Following lentivirus infection, GFP-positive cells were sorted using flow cytometry. Overexpression of SOX6 was confirmed using qRT-PCR and Western blot analysis. An empty pCDHI-MCSI-EF1-COpGFP vector was used as the control group (Ctrl.).