2.2. Silencing H19 Inhibits Myofibroblast Activities

The collagen gel contraction assay has been widely used to evaluate the differentiation of myofibroblasts, as a function of their role in wound closure and the fact that α-SMA expression augments their contractile capacity [4,18]. We observed that the size of the collagen gel matrix was increased in two fBMFs with short hairpin RNA (shRNA)-mediated silencing of H19, suggesting that the contraction activity was relieved when H19 was lowered (Figure 2A). Another hallmark of myofibroblasts is their higher migratory capability. As shown in Figure 2B, knockdown of H19 attenuated the migration ability in fBMFs, indicating that the upregulation of H19 promoted the differentiation of myofibroblasts. As such, we aimed to elucidate the potential mechanisms underlying the profibrotic effect of H19.

Open in a separate window

Figure 2

Downregulation of H19 suppresses the phenotypes of myofibroblasts. (A) H19-silenced fBMFs were embedded into collagen gels. After 48 h, the contraction of gels was photographed, and the gel areas were calculated using Image J software v.1.8 (National Institutes of Health, Bethesda, Maryland, USA). (B) A Transwell migration assay was conducted to examine the effect of H19 knockdown on migration ability after 24 h of incubation. Magnification, 100×. * p < 0.05 compared to sh-Luc.

2.3. H19 Promotes Myofibroblast Activation through the Suppression of miR-29b

The interaction between H19 and miR-29b is implicated in bleomycin-induced idiopathic pulmonary fibrosis [16], and miR-29b has been revealed to function as an oncomir in oral cancer [16]. Therefore, we examined whether H19 exerted its effect through targeting miR-29b. Figure 3A illustrates the complementarity between the 3′-UTR regions of H19 and miR-29b to pinpoint the target sequence of miR-29b. Reporter plasmids containing either full-length (Wt-H19) or mutated (mut-H19) forms of the miR-29b-binding region were constructed and cotransfected with miR-29b mimics into fBMFs. We observed that the luciferase activity of Wt-H19 vector was notably reduced when cotransfected with miR-29b mimics, whereas no significant change was present in the mut-H19 vector in fBMFs (Figure 3B). In agreement with this finding, we found that the expression of miR-29b was upregulated in response to H19 inhibition in fBMFs (Figure 3C). To verify if the decreased expression of miR-29b resulted in myofibroblast activation, an miR-29b inhibitor was added to normal BMFs. We demonstrated that both collagen gel contraction (Figure 3D) and migration ability (Figure 3E) were markedly enhanced.

Open in a separate window

Figure 3

Repression of miR-29b by H19 results in upregulation of type I collagen. (A) Schematic of miR-29b and the putative binding sequence, as well as the mutant sequence at the 3’-untranslated region (3′UTR) of H19. (B) Luciferase activity decreased when fBMFs were cotransfected with Wt-H19 and miR-29b mimics. (C) Relative expression of miR-29b in H19-silenced fBMFs. Collagen gel contraction (D) and migration ability (E) of BMFs with miR-29b inhibitor or miR-Scr (scramble). Magnification, 100× (F) Schematic of miR-29b and the putative binding sequence in the 3′UTR sequence of Wt-COL1A1 and mutant sequence. (G) Luciferase activity was decreased when fBMFs were cotransfected with Wt-COL1A1 and miR-29b mimics. (H) Protein expression of type I collagen in BMFs with or without sh-Luc, sh-H19, miR-Scr, and miR-29b inhibitor. * p < 0.05 compared to miR-Scr.

Previously, miR-29b was shown to inhibit the transdifferentiation of C2C12 myoblasts into myofibroblasts via suppression of collagens [19]. Hence, we sought to evaluate the connection between miR-29b and type I collagen. We incorporated the point mutation in the 3′UTR to abolish the putative miR-29 recognition sequence of type I collagen transcripts (COL1A1) (Figure 3F). We showed that cotransfection with miR-29b mimics and Wt-COL1A1 suppressed reporter activity, but luciferase activity was not affected when cotransfected with the mutated form (Figure 3G). This finding indicates that miR-29b directly targeted COL1A1. Next, we demonstrated that the downregulation of miR-29b augmented the expression of type I collagen, whereas the suppression of H19 reduced its expression. Moreover, the decreased expression of type I collagen via knockdown of H19 was reversed when the miR-29b inhibitor was coadministrated in BMFs (Figure 3H and Figures S1A and S3, Supplementary Materials). These results suggest that H19 regulated the protein expression of COL1A1 by counteracting the effect of miR-29b.

2.4. Overexpression of miR-29b Suppresses Myofibroblast Phenotypes and Expression of Fibrosis Markers

To further evaluate the functional role of miR-29b in myofibroblast activation, fBMFs with ectopic expression of miR-29b were subjected to various assays to examine the myofibroblast features and the expression of fibrosis markers. As expected, collagen gel contraction (Figure 4A) and migration capacity (Figure 4B) were diminished in miR-29b-overexpressing fBMFs. Furthermore, the protein expression levels of α-SMA, type I collagen, and fibronectin were all reduced (Figure 4C, Figures S1B and S4, Supplementary Materials). The scratch wound healing assay revealed that a higher closure rate was observed in fBMFs compared to normal BMFs, whereas overexpression of miR-29b in fBMFs delayed the repopulation of cells in the artificial gap (Figure 4D).

Open in a separate window

Figure 4

Overexpression of miR-29b mitigates the features of myofibroblasts. (A) Collagen gel contraction and (B) migration ability of fBMFs with miR-Scr or miR-29b mimics. Magnification, 100× (C) Relative expression of SMA, type I collagen, and fibronectin (FN1) in fBMFs with miR-Scr or miR-29b mimics. (D) Wound healing capacity of BMFs, fBMFs with miR-Scr, or fBMFs with miR-29b mimics. Magnification, 100×, * p < 0.05 compared to miR-Scr.

2.5. The Expression of miR-29b Is Downregulated in OSF Specimens and Negatively Correlated with Fibrosis Markers

Subsequently, we verified the aberrantly reduced expression of miR-29b in OSF tissues (Figure 5A) and fBMFs derived from these specimens (Figure 5B). Moreover, we found that the expression of miR-29b was inversely related to the expression of myofibroblast marker α-SMA (Figure 5C). Similarly, the expression of miR-29b was inversely correlated to fibronectin, which was required for collagen matrix assembly (Figure 5D). Altogether, these results support the antifibrotic property of miR-29b in OSF and the association of the inadequate expression of miR-29b with oral fibrogenesis.

Open in a separate window

Figure 5

The expression of miR-29b is diminished in OSF and inversely associated with fibrosis markers. (A) Gene expression of miR-29b in OSF (red dots) and normal tissues (n = 20; blue dots). ** p < 0.01 (B) Relative expression of miR-29b in BMFs and fBMFs derived from OSF specimens (n = 5). ** p < 0.01 The expression of miR-29b was negatively correlated with (C) α-SMA and (D) FN1 expression in OSF tissues (blue dots) ** p < 0.01.

4.2. Tissue Acquisition

As for tissue collection, all procedures followed the tenets of the Declaration of Helsinki and were reviewed by the Institutional Review Board of Human Subjects Research Ethics Committee at Chung Shan Medical University, Taichung, Taiwan (approval number: CSMUH No. CS2-16142 (approval number: CSMUH No. CS2-16142; approval date: 10/05/2017). A total of 20 histological fibrotic mucosa or normal tissues were obtained from OSF patients or normal subjects recruited at the Department of Dentistry, Chung Shan Medical University Hospital.

4.3. Cell Culture

Upon approval by the Institutional Review Board of the Chung Shan Medical 364 University Hospital (approval number: CSMUH No: CS2-16142), two normal human BMFs isolated from normal controls, two fBMFs isolated from OSF tissues, and 293T cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% v/v foetal bovine serum (FBS) (Thermo Fisher Scientific, Inc., Carlsbad, CA, USA) and 1% streptomycin/penicillin (Hyclone, Logan, UT, USA) as previously described [18].

4.4. RNA Sequencing

TRIZOL reagent (Invitrogen) was conducted to collect total RNAs for identification of the differential expression of the transcriptomes between normal and OSF tissues. The quality control of the RNA isolate from each sample was assured by the manufacturer of Genomics inc. before library construction. RNA-Seq libraries were sequenced paired-end on a HiSeq2500 (Illumina) according to the manufacturer’s instructions. Raw reads were processed using the Illumina CAS v.1.8.2 (Illumina, San Diego, CA, USA) to filter out the low-quality reads. The library preparations were sequenced on an Illumina platform and the paired-end reads. For the bioinformatics analysis, the sequence reads were aligned and calculated expression values according to FPKM (fragments per Kb of transcript per million mapped reads) for detection of discrepancies in the transcripts [5].

4.5. Quantitative Real-Time PCR

The TRIzol reagent was used for the isolation of total RNAs from cells, which were reverse-transcribed into complementary DNA (cDNA) using the SuperScript^® III First-Strand Synthesis System (Invitrogen Life Technologies, Carlsbad, CA, USA) following the manufacturer’s instructions for RT-PCR. The resulting first-strand cDNAs were used as templates for qPCR using an ABI StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and SYBR Green reagent with specific primers to anneal to the target sequences. GAPDH was used as the reference gene. [6]. The primer sequences used in this study were as follows: H19, 5′–CTCACCTTCCAGAGCCGATT–3′ and 5′–TAAGTTTCCGGGTCCGAACT–3′; GAPDH, 5′–CTCATGACCACAGTCCATGC–3′ and 5′–TTCAGCTCTGGGATGACCTT–3′.

4.6. Inhibition and Overexpression of H19

Lentivirus-mediated short hairpin RNA (shRNA) targeting H19 was generated as follows: the pLV-RNAi vector was purchased from Biosettia Inc. (Biosettia, San Diego, CA, USA) for efficient delivery of RNA interference (RNAi) into cells via lentiviral transduction. The oligonucleotide sequence of lentivirus expressing shRNA against H19 was synthesized and cloned into the pLV-RNAi vector, which was then cotransfected with a plasmid DNA mixture consisting of helper plasmids (VSVG and Gag-Pol) into 293T cells (American Type Culture Collection, Manassas, VA, USA) using Lipofectamine 2000 (LF2000, Invitrogen, Carlsbad, CA, USA). The target sequences for H19 were listed as follows: Sh-H19-1: 5′- AAAAGCTTTCCTGTCTTTCCTTTATGGATCCATAAAGGAAAGACAGGAAAGC-3′; Sh-H19-2: 5′-AAAAGCTTTCCTGTCTTTCCTTTATGGATCCATAAAGGAAAGACAGGAAAGC-3.

4.7. Collagen Gel Contraction

A mixture of collagen gel and 2 × 10⁵ BMFs/fBMFs cells (with sh-H19, miR-29b inhibitor, or miR-29b mimic) was added to a six-well plate for polymerization. Then, the gels were detached from the well using a sterile pipette tip for measurement of gel contraction. The size of the gel was calculated using ImageJ software v.1.8 (National Institutes of Health, Bethesda, MD, USA).

4.8. Transwell Migration Assay

A two-chamber Transwell cell culture system with an 8 μm pore size polycarbonate membrane was used (Corning, Cambridge, MA, USA). 2 × 10⁵ BMFs/fBMFs were added to the upper chamber along with a serum-free medium. In the lower compartment, a serum-containing medium was used as the chemoattractant. After 48 h of incubation, cells on the other side of the insert were stained with crystal violet (Sigma-Aldrich, Sigma-Aldrich, St. Louis, MO, USA) prior to fixation. Cell numbers were counted from five different visual areas of 100-fold magnification under a microscope.

4.9. Luciferase Reporter Assay

The full length of H19 was cloned and inserted into a pmirGLO plasmid (Promega, Madison, WI, USA) to generate the pmirGLO-H19-Wt reporter according to the manufacturer’s instructions, and the pmirGLO-H19-mut reporter was constructed via mutagenesis. For different experiments, fBMFs were cotransfected with pmirGLO-H19-Wt reporter, pmirGLO-H19-mut reporter, miR-29b mimic, or miR-Scr (scramble) using Lipofectamine 2000 reagent, followed by analysis of luciferase activity.

4.10. Western Blot Analysis

Protein samples were extracted from the BMFs or fBMFs. After denaturation, equal amounts of extracted proteins (30 μg) were separated by 10% SDS-PAGE with Tris-Glycine SDS running buffer and transferred with transfer buffer onto a nitrocellulose membrane (GE 446 Healthcare, Little Chalfont, Buckinghamshire, UK). Then, transferred membranes were incubated with BlockPRO blocking buffer (visual protein, energenesis biomedical co. ltd). After washing with TBST buffer (20 mM Tris, 150 mM NaCl and 0.1% Tween 20; pH 7.4) for 5 min 3 times, the transferred membranes were incubated with the primary anti-bodies against α-SMA (1:1000 dilution; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), COL1A1(1:1000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), fibronectin (1:1000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), or GADPH (1:5000 dilution; Thermo Fisher Scientific, Waltham, MA, USA). Then, transferred membranes were washed with TBST buffer for 5 min 3 times followed by incubation with the corresponding secondary antibodies. Developed chemiluminescence signals from catalyzed ECL substrate (Pierce Biotechnology) were detected using the LAS-1000 plus Luminescent Image Analyzer (GE Healthcare, Piscataway, NJ, USA). After the intensity of each band was measured by densitometry, the relative intensities were calculated by normalizing to GAPDH.

4.11. Wound Healing Assay

Confluent fBMFs were mechanically wounded by creating a straight line in the center of the well using a sterile 200 µL plastic pipette tip. The rate of the cells toward the denuded area was monitored and photographed at 0 and 48 h under a microscope.