Introduction
Lumbar spinal canal stenosis (LSCS) is the most common spinal disease in elderly patients [
41], and hypertrophy of ligamentum flavum (HLF) is considered to be a major cause of LSCS [
27,
42]. There is currently no particularly effective conservative treatment to delay or reverse the HLF-induced LSCS except surgical decompression. Surgical removal of the hypertrophic ligamentum flavum can achieve spinal decompression, but not only the risks and postoperative complications of surgery are particularly prominent but also many patients with underlying diseases (such as diabetes, hypertension and hyperlipidemia) are intolerant to surgery [
21,
37,
43]. Although fibrosis have been proved to be the key pathological feature of HLF, the precise mechanism of the pathology of LF fibrosis has not been fully elucidated [
32]. Therefore, investigating molecular pathways associated with LF fibrosis will provide insight into HLF mechanisms and identify novel targets for prevention and treatment of HLF-induced LSCS.
Transcription factor is a DNA binding proteins that can recognize specific DNA sequences to activate or inhibit gene transcription [
15]. Studies have proved that abnormal expression of transcription factor is closely related to the occurrence and development of many human diseases [
7,
29]. The TCF/LEF transcription factor family can bind to the WRE region within the promoter sequence or enhancer sequence of the target gene, thereby promoting or inhibiting transcription of the target gene [
14]. Transcription factor 7 (TCF7, also known as TCF1) is one of the members of the TCF/LEF transcription factor family and participates in the transcriptional regulation of Wnt/β-catenin signal [
1,
3]. Emerging evidence has shown that TCF7 is involved in tumorigenesis and development [
2,
35,
44], sepsis-induced renal injury [
39], heart development [
40], and cardiac hypertrophy [
26]. However, the roles and underlying mechanism of TCF7 in HLF have not been clarified.
Extensive evidence has shown that TCF7 is negatively regulated by microRNAs (miRNAs) [
12,
28]. MiRNAs is a class of non-coding single-stranded RNA molecules with a length of about 19–25 nucleotides encoded by endogenous genes, which participate in the regulation of post-transcriptional gene expression in plants and animals [
19]. Current studies have been demonstrated that dysregulated miRNAs play an important role in the pathogenesis of HLF [
43]. For example, Li P et al. have demonstrated that miR-10396-3p is significantly decreased in the mechanical stress (MS)-induced HLF and overexpressing miR-10396-3p inhibits MS-induced HLF by targeting the inhibition expression of IL-11 [
17]. Ma et al. reported that delivering the two miRNAs (miR-146a-5p and miR-221-3p) to LF cells markedly suppressed fibrosis and hypertrophy of LF in vitro and vivo [
20]. Our preliminary analysis found that miR-4306 has a potential binding site with TCF7 and negatively regulates TCF7 expression in HLF cells. Previous study have shown that miR-4306 is downregulated in HLF tissues, and miR-4306 expression in HLF tissues were markedly negatively associated with the ratio of LF/spinal canal area [
22]. These studies showed that miR-4306 might be involved in the pathogenesis of HLF, but the specific biological function of miR-4306 in the pathogenesis and development of HLF still remains unclear.
In the current study, we identified the TCF7 is significantly upregulated in HLF tissues and cells by integrating analysis of RNA-sequencing, bioinformatics analysis and validation experiments. Functional experiments demonstrated that TCF7 promoted hyperplasia and fibrosis of the HLF cells in vitro and in vivo. Moreover, our data demonstrated that TCF7 promoted SNAI2 expression by directly activating transcription SNAI2, and further SNAI2 inhibited miR-4306 expression by directly binding the promoter of miR-4306. Finally, our data revealed that miR-4036 negatively regulated by SNAI2 negative feedback regulated TCF7 expression by directly binding to TCF7 mRNA 3′-UTR, thus inhibiting the hyper-proliferation and fibrosis phenotype of HLF cells in vitro. Collectively, our results demonstrated not only an important role of TCF7/SNAI2/miR-4306 signaling in regulating HLF, but also provide a strong theoretical rational for developing new drugs to prevent and treat HLF.
Materials and methods
Human LF sample collection
This study was approved by the Institutional Research Ethics Committee of the Zhujiang Hospital of Southern Medical University. A total of 30 LF tissues samples, including hypertrophied and non-hypertrophied LF tissues, were collected from patient under-going lumbar spine surgery in Zhujiang Hospital of Southern Medical University. The hypertrophied LF tissues (> 4 mm thickness) were obtained from patient with LSCS due to LF hypertrophy, and non-hypertrophied LF tissues (≤ 4 mm thickness) were obtained from age- and gender-matched lumbar disc herniation (LDH) patient without LF hypertrophy as control. All enrolled patients were excluded from diseases such as cancer, heart disease, kidney disease, rheumatism and autoimmune diseases. All patients underwent magnetic resonance imaging scan to confirm the thickness of the LF before surgery, and all LF samples were obtained from the anatomical region (L4/5). Fibrosis score was assessed by Masson's trichrome staining according to previously report [
38]. Informed consent was obtained from each patient prior to this study. Patient information in this study is summarized in Additional file
5: Table S1.
mRNA sequencing and data analysis
Total RNAs were extracted from three independent samples of hypertrophied or non- hypertrophied LF using TRIzol reagent (Invitrogen) according to the manufacturer’s recommended protocol, and the RNA quantity was assessed with a NanoDrop ND-2000 spectrophotometer (NanoDrop Technologies). After purifying the mRNA using RiboZero Magnetic Gold Kit, the cDNA libraries were constructed for the KAPA Stranded RNA-Seq Library Prep kit (Illumina, Inc.) according to the manufacturer’s instructions. Subsequently, we used Agilent 2100 and qPCR to assess the quality and quantification of the cDNA library. Finally, the RNA-sequencing was performed by Next-Generation Sequencing with an Illumina HiSeq Xten platform. Clean data were obtained from the raw data by removing reads containing adapters, reads containing over 10% poly N, and low-quality reads, which were aligned to the specified reference genome (Homo sapiens. GRCh38, NBCI) to obtain the mapped data. The differentially expressed mRNAs between hypertrophied LF tissues and non- hypertrophied LF tissues were performed using EBseq R package. The fold changes (FCs) ≥ 2 or − 2 and false discovery rates (FDRs) < 0.05 served as the screening criteria to get differentially expressed mRNA.
Primary LF cell culture
According to our previously described method [
4], LF cells were isolated and cultured from the LF tissues of patients with LSCS or LDH. Briefly, the obtained LF samples were cut into small pieces and digested for 2 h at 37 °C using Dulbecco’s modified Eagle’s medium (DMEM, Gibco) with 0.2% type I collagenase (Gibco), then seeded on to cell culture dish and incubated with DMEM containing 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Invitrogen). The isolated cells were observed for fibroblast morphology and identified the expression of specific markers [collagen I (1:100, #72026, Cell Signaling Technology) and Vimentin (1:500, ab16700, Abcam)] by immunofluorescence staining. Subsequent experiments were conducted using cells from the third passage to forth passage of LF cells.
Adenovirus construction and infection
The adenovirus of TCF7 overexpression/knockdown or SNAI2 overexpression/knockdown were constructed by Genechem (Nanjing, China). The adenovirusl particles of miR-4306 mimic/inhibitor and negative control (NC) mimics/inhibitor were obtained from Ribobio Inc. (Guangzhou, Guangdong, China). All cells were infected with the adenovirus according the manufacturer’s instructions, and the infection efficiency was verified by real-time quantitative PCR (RT-qPCR) or Western blotting analyses. All shRNA and miR-4306 mimics/inhibitor sequences are listed in Additional file
6:Table S2.
RT-qPCR
Total RNA from the cells or tissues sample was extracted using Trizol (Invitrogen, Carlsbad, CA). Reverse transcription and RT-qPCR for miR-4306 were carried out using miRNA 1st Strand cDNA Synthesis Kit (MR101-01, Vazyme) and miRNA Universal SYBR qPCR Master Mix (MQ101-01, Vazyme) on a LightCycler
®96 (Roche). Reverse transcription and RT-qPCR for the mRNAs were performed as described in our previous studies [
4].U6 small nuclear RNA and GAPDH were used as an internal control for miR-4306 and mRNAs, respectively. The relative RNA expression of genes was calculated according to the Ct (2
−ΔΔCt) method. All experiments were performed in triplicate. All specific primers sequence in study are listed in Additional file
7:Table S3.
Western blotting
The nuclear protein extraction was performed using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, Shanghai, China) according to the protocol of manufacturer, and then subjected to western blotting analysis. Western blotting was conducted as described in our previous studies [
4]. Primary antibodies antibody information is as follows: cleaved Caspase-3 (1:500, ab32042, Abcam), Bax (1:1000, ab182733, Abcam), Bcl-2 (1:500, ab182858, Abcam), TCF7 (1:1000, #2203, Cell Signaling Technology), SNAI2 (1:1000, #9585, Cell Signaling Technology), collagen I (1:1000, ab260043, Abcam), collagen III (1:1000, ab7778, Abcam), MMP2 (1:500, ab181286, Abcam), MMP13 (1:1000, ab39012, Abcam), TGF-1β (1:1000, ab215715, Abcam), and GAPDH (1:1000, ab8245, Abcam).
Cell proliferation
Cell proliferation ability was detected using a Cell Proliferation Reagent Kit I (MTT, Sigma-Aldrich) according to the manufacturer’s recommended protocol. Absorbance values were measured at the wavelength of 490 nm. All experiments were performed in three to five biological duplicates.
Cell apoptosis
Cells apoptosis was measured by flow cytometry using Annexin V-FITC/propidine iodide (PI) double-staining kit (Abcam) according to the manufacturer’s instructions. All experiments were performed in three biological duplicates.
Chromatin immunoprecipitation
A chromatin immunoprecipitation (ChIP) assay was carried out using the ChIP assay kit (ab500, Abcam) according to the manufacturer’s instructions. Briefly, the following antibodies were used to immunoprecipitate cross-linked protein-DNA complexes: rabbit anti-TCF7 (1:50, #2203, Cell Signaling Technology, USA), rabbit anti-SNAI2 (1:50, #9585, Cell Signaling Technology, USA), and normal rabbit IgG (1:50, #2729, Cell Signaling Technology, USA). After cross-linked protein-DNA complexes to free DNA, the immunoprecipitated DNA was purified for RT-qPCR analyses with specific primers.
Dual-luciferase reporter assay
For validation of transcription factor interactions with target genes, the TCF7-binding motif in the promoter region of SNAI2 and the SNAI2-binding motif in the promoter region of miR-4306 were predicted through JASPAR (
http://jaspar.genereg.net/). According to the predicted results, different truncated plasmid of SNAI2 or miR-4306 promoter was co-transfected with corresponding transcription factors plasmid into 293 T cells. For the verification of interaction between miR-4306 and TCF7, the wild type or mutant type of TCF7 3′-UTR (containing the binding site of miR-4306) was cloned into the luciferase vector, and then transfected into 293 T cells together with miR-4306 mimics or the negative control (NC mimics), respectively. Transfer after 24 h, luciferase activities were assessed using a Dual Luciferase Assay Kit (Promega, Madison, WI, USA) in accordance with the manufacturer’s instructions. The relative luciferase activities were calculated by normalizing the signal value of Renilla luciferase to Firefly luciferase, and then compared with the negative control.
RNA immunoprecipitation
RNA immunoprecipitation (RIP) experiments were performed with a Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore, Billerica, MA) according to the manufacturer’s instructions. AGO2 antibody ((Millipore, Billerica, MA, USA)) or negative control IgG antibody (Millipore, Billerica, MA, USA) was used for RIP. Co-precipitated RNAs were finally extracted with RNeasy Mini Kit (QIAGEN, China) for RT-qPCR to demonstrate the presence of the binding targets.
Animal experiments
All animal experimental protocols were approved by the Ethics Committee for Animal Research of the Zhujiang Hospital of Southern Medical University (Guangzhou, China). All mice purchased from the Experimental Animal Center of Southern Medical University were 8-week-old C57BL/6 male mice. The mice model with HLF was constructed by the hydrophobic characteristics of mice according to the previously described [
45,
46]. In brief, the mice were placed in a beaker with 5 mm of animal water at the bottom to induce a bipedal standing posture. The mice were maintained in a bipedal standing posture for 6 h a day with an interval of 2 h of free activity, and the above operations were continued for 8 weeks to construct the HLF mice model. The control mice were placed in a same condition as the mice with HLF model except for the bottom without animal water. To verify the therapeutic effect of silencing TCF7 on the mouse HLF model, 20 mice were randomly assigned to the control group, HLF model group, HLF + AAV-shNC group, and HLF + AAV-shTCF7 group. After six weeks of bipedal induction, we made a longitudinal skin incision in the mouse lumbar spine, and removed the dorsal paravertebral muscle from the spinous processes and laminae to expose the L5/6 LF under a surgical microscope, thereby AAV-shNC or AAV-shTCF7 (1 × 10
12 vg/ml, 3 µl) from Hanbio Biotechnology (Shanghai, China) was injected into L5/6 LF in anesthetized mice with a microinjector (NF36BV 36GA, NanoFil, United States). After 4 weeks of AVV injection, mice were euthanized and L5/6 vertebrae were collected to obtain LF tissue for later histological analyses and molecular biological analyses. Histological analyses were performed to measure the area of the LF.
Histological analyses
hematoxylin and eosin (H&E) staining and Elastica van Gieson (EVG) staining were performed to measure the area of the LF and degree of LF fibrosis, respectively. The LF samples from mice or humans were fixed overnight with 4% paraformaldehyde (Beyotime, shanghai, China), paraffin-embedded, and cut into slices (4 μm). After dewaxing and dehydration, the slices were performed by H&E kit (Beyotime, shanghai, China) or EVG kit (JianChen, Nanjing, China) according to the manufacturer’s instructions. Quantitative analyses of the LF areas and degree of LF fibrosis (the ratio of elastic fibers to collagen fibers) were obtained using ImageJ software (NIH, United States). Each sample was detected three times, and the average value was taken.
Immunohistochemistry
Immunohistochemical staining was conducted using Histostain®-SP Kits according to the manufacturer’s instructions. In brief, the LF samples from humans were fixed overnight with 4% paraformaldehyde (Beyotime, shanghai, China), paraffin-embedded, and then cut into slices with 4 μm thickness. After dewaxing and antigen retrieval, the slices were organization background closed using serum blocking reagent. Then these slices were incubated with TCF7 primary antibodies (1:200, #2203, Cell Signaling Technology) and β-catenin (1:100, ab16051, Abcam) overnight at 4 °C. Subsequently, these sections were incubated with the respective secondary antibodies (Proteintech) at room temperature. Immunohistochemical results were visualized using a Zeiss microscope (Carl Zeiss Meditec AG, Jena, Germany) and then analyzed by ImageJ software (NIH, United States).
Statistical analysis
Results are presented as the means ± SD. All the statistical analyses were performed using the GraphPad Prism 6.0 (GraphPad Software, La Jolla, USA). Statistical significance for comparisons between two groups were analyzed using Student’s t-test, and statistical significance for comparisons among more than two group was analyzed using one-way ANOVA. The p < 0.05 was considered as significant difference.
Discussion
Although fibrosis is considered as the key pathological feature of HLF, but internal molecular mechanisms has not been fully elucidated. Herein, we confirmed for the first time that TCF7 significantly upregulated in HLF, and the TCF7 expression was remarkably positively correlated with the thickness and fibrosis degree of LF. Moreover, we demonstrated that silencing TCF7 inhibited LF hypertrophy and fibrosis by inhibiting LF cell proliferation, promoting apoptosis, and aggravating ECM degradation in vitro and vivo. Mechanically, our data revealed that augmented TCF7 led to transcriptional activation of SNAI2, and SNAI2 inhibited the transcription of miR-4306 by binding to the promoter region of miR-4306, which in turn promoted the TCF7 expression. Most importantly, we found that inhibition of TCF7/SNAI2/miR-4306 signaling suppressed LF hypertrophy and fibrosis, indicating that targeting TCF7/SNAI2/miR-4306 signaling may be a novel strategy for the prevention and treatment of HLF.
Accumulating evidence indicates that TCF7 is a critical function downstream of the typical WNT/β-catenin signaling pathway, which is involved in biological processes of many diseases [
2,
18,
44]. A recent study reported that TCF7 expression was elevated in mouse heart tissue after TAC and in cardiomyocytes treated with Ang-II, and inhibition of TCF7 suppressed the occurrence of cardiac hypertrophy [
18]. Base on integrating analysis of RNA-sequencing, bioinformatics analysis and validation experiments, we identified for first time that the mRNA and protein expression of TCF7 was significantly increased in HLF tissues and cells. Correlation analysis showed that TCF7 expression had significant positive correlation with LF thickness and fibrosis score, suggesting that TCF7 involved in the HLF development. Then our data demonstrated that TCF7 overexpression significantly induced proliferation and fibrosis of HLF cells. On the contrary, the silencing TCF7 significantly suppressed proliferation and fibrosis of HLF cells in vitro and suppressed the LF fibrosis and hypertrophy in vivo. In addition to proliferation and fibrosis, apoptosis is also involved in the pathological process of HLF, but the role of apoptosis in HLF remains controversial. Several investigators have shown that activation of apoptotic pathways induce cell apoptosis in HLF tissue [
6,
24]. However, some researchers have shown that apoptosis is inhibited in HLF tissue. For example, Zhou et al. revealed that inhibition of apoptosis in HLF tissue compared to non-HLF tissue, which manifested as inhibition of apoptosis marker (Bax and cleaved caspase3) expression and reduction of TUNEL-positive cell percentage in human HLF tissue, and reduction cleaved caspase3 expression in the rat model with HLF [
47]. At the same time, they showed that LPAR1 significantly upregulated in HLF tissue compared with that in non-HLF tissue, overexpression of LPAR1 improved inhibited apoptosis in LF cells, whereas knockdown of LPAR1 has the opposite effect [
47]. Sun et al. demonstrated that upregulated WISP1 in HLF decreased apoptosis of LF cells, which manifested as inhibition of Bax and activation of Bcl-2 [
30]. Furthermore, DNMT1-mediated ACSM5 significantly downregulated in LFH tissue, and ACSM5 knockdown inhibited apoptosis of HLF cells in vitro [
4]. Like these studies, our data showed that upregulated TCF7 could promoted apoptosis of HLF cells, whereas silencing TCF7 significantly promoted apoptosis of HLF cells in vitro. Thus, these results implied that TCF7 inhibition might be considered as a novel potential therapeutic strategy for HLF.
Emerging evidence has demonstrated that dysregulated miRNAs is involved in the occurrence and progression of HLF [
22,
43]. Yu et al. identified that miR-221 was down-regulated in HLF tissues, and overexpressing miR-221 inhibited the expression of collagens I and collagens III by sponging TIMP-2, thus inhibiting HLF formation [
36]. Sun et al. revealed that miR-21 may play an important role in HLF, upregulation of miR-21 might contribute to the HLF by promoting inflammation and fibrosis via the induction of IL-6 expression [
36]. The above studies suggested that miRNAs played pivotal roles in the occurrence and development of HLF, which has aroused great attention of scholars. Notably, previous study have identified that the down-regulation of miR-4306 in HLF tissues is significantly negatively correlated with increased LF/spinal canal area ratio (LSAR) [
22], but the role and underlying mechanism of miR-4306 in HLF cells has not been investigated. Herein, we found that overexpression of miR-4306 in HLF cells suppressed cell hyper-proliferation and pro-fibrosis, whereas inhibition of miR-4306 displayed the opposite effect. These results showed that miR-4306 might play an anti-hypertrophy role in HLF. Furthermore, a growing body of research has revealed that miRNA-mediated target gene expression exerted an appreciable promoting or inhibiting HLF progression [
5,
17,
31,
36,
43]. We conducted bioinformatics analysis using the RAID and miRDB software and found that TCF7 contained potential binding sequences for miR-4306. The RIP assay and dual luciferase activity assay confirmed that TCF7 was a direct target for miR-4306.
Remarkably, a host of studies have shown that miRNAs expressions is regulated by transcription factors (TFs) in gene regulatory networks, and the interaction between TFs and miRNAs can precisely regulate gene expressions to maintain cell homeostasis [
25]. The SNAI2 encoded by the SNAI2 gene is an evolutionarily conserved C2H2 zinc finger protein that orchestrates biological processes critical to tissue development and tumorigenesis, and its main role is to facilitate the epigenetic regulation of transcriptional programs [
48]. And previous studies have demonstrated that SNAI2 interacts with miRNAs promoters to inhibit its expression and then regulates cell biological functions [
8,
9,
34]. Consistent with previous studies, we identified that SNAI2 inhibited the transcription of miR-4306 by directly binding to the promoter region of miR-4306. Moreover, extensive studies have revealed that SNAI2 plays an important role in various cell proliferation and fibrosis processes [
10,
11,
13,
16]. However, no studies have investigated the potential biological function of SNAI2 in HLF cells. Our results revealed that SNAI2 overexpression promoted proliferation and fibrosis of HLF cells, and the inhibition of SNAI2 suppressed proliferation and fibrosis of HLF cells, indicating SNAI2 may play pro-hypertrophy roles in LF. Further rescue experiments disclosed that silencing SNAI2 could eliminate the inhibitory effect of overexpression of miR-4306 on the proliferation and fibrosis of HLF cells. Thus, our results revealed that miR-4306 was directly inhibited at the transcriptional level by SNAI2, and thereby promoted the proliferation and fibrosis of HLF cells. In addition, we found that TCF7 directly targets SNAI2 promoter and activates its transcription expression, and activation of SNAI2 contributes to TCF7-mediated LF hypertrophic and fibrotic phenotype in vitro. Therefore, our data revealed that TCF7 promoted HLF formation through mediating the TCF7/SNAI2/miR-4306 feedback loop (Fig.
7).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.