MiR-206 promotes extracellular matrix accumulation and relieves infantile hemangioma through targeted inhibition of DNMT3A

ABSTRACT

MiR-206 is abnormally expressed in infant hemangioma endothelial cells (HemECs), but the mechanism is not clear. We explored the intervention of miR-206 in HemECs in relation to extracellular matrix (ECM) metabolism. We selected 48 cases of infantile hemangioma (IH) from volunteer organizations. After the isolated and extracted HemECs were interfered with overexpressed or silenced miR-206, the effects of miR-206 on the proliferation, migration and invasion of HemECs were examined through basic cell function experiments. The expression differences of miR-206, DNA Methyltransferase 3A (DNMT3A) and ECM-related genes were analyzed as needed by qRT-PCR or Western blot. TargetScan and dual-luciferase experiments were applied to predict and confirm the binding relationship between miR-206 and DNMT3A. The correlation between miR-206 and DNMT3A was analyzed in IH tissues by Pearson correlation coefficient, and further confirmed in HemECs by conducting rescue experiments. A nude mouse model of xenograft tumor was constructed to verify the results of in vitro experiments. MiR-206, which was downregulated in proliferative hemangioma, suppressed the malignant development of HemECs by regulating ECM-related genes. As the target gene of miR-206, DNMT3A was high-expressed in IH tissues and was negatively correlated with miR-206. Overexpressed DNMT3A counteracted the inhibitory effect of miR-206 mimic on HemECs and its regulatory effect on ECM. The results of in vivo experiments were consistent with those from cell experiments. Thus, miR-206 could promote ECM accumulation through targeted inhibition of DNMT3A, further inhibiting the malignant development of HemECs and relieving IH.

KEYWORDS:

• Infantile hemangioma
• hemangioma endothelial cells
• miR-206
• DNA methyltransferase 3A
• extracellular matrix

1. Introduction

Infantile hemangioma (IH) is a benign tumor occurring at skin and soft tissues due to the growth of vascular tissues during embryonic period [1]. Excessive proliferation of vascular endothelial cells is the most important biological characteristic of IH that distinguishes it from vascular malformation [2]. The natural course of IH is generally divided into proliferation period, regression period, and regression completion period. The entire growth cycle lasts from several months to several years from the birth of an infant [3]. However, current inadequate understanding of the pathogenesis and regression mechanism of IH greatly limits clinical efficacy of IH treatment [4]. At the same time, only few effective risk assessment indicators for IH have been discovered [5]. Therefore, it is highly necessary to further study the mechanism of the occurrence and development of IH.

Regarding the pathogenesis of IH, it is currently believed that hemangioma can cause abnormal proliferation of vascular endothelial cells, which could alter local microenvironment and abnormal endothelial cell transformation, such as the imbalance of angiogenic factors and inhibitors, compositional or functional abnormalities of immune cells, and elevated estrogen levels [6,7]. However, angiogenic factors alone are not capable enough to induce a rapid growth of hemangioma unless under the joint actions with multiple environmental factors. Study showed that extracellular matrix (ECM) can directly affect the response of endothelial cells to angiogenic factors through its expression of integrins [8]. ECM mainly provides a necessary scaffold for maintaining the vascular tissue structures of vascular endothelial cells through adhesion with integrins on the surface of endothelial cells [9,10]. The interaction of endothelial cell integrins and interstitial matrix proteins (interstitial collagen and fibrin/fibronectin) is a key receptor that stimulates endothelial cell tubular formation and simultaneously activates endothelial cells [11,12]. The main integrins responsible for these interactions are α2β1, α1β1, and αvβ3, α5β1, which are collagen and fibrin and fibronectin receptors, respectively [13–16]. ECM is a dynamic network structure between cells and consists of macromolecular substances such as collagen, proteoglycan and glycoprotein [17]. At present, the most researched ECMs related to the proliferation and degeneration of hemangioma are mainly laminin (LN), fibronectin (FN) and vitronectin (VN) [18,19]. It has been reported that LN was detected in the thickened basement membranes of hemangioma [20]. α6-integrin subunit contribute to tumor angiogenesis and invasive properties of tumor cells [21]. Nonetheless, other related factors involved still require further study.

MicroRNAs (miRNAs), which are a member of the small RNA family with about 20 bases, can partially or completely recognize specific targeted mRNAs, and degrade the mRNAs at the transcription level or inhibit its expression, thereby completely silencing or inhibiting the formation of a target protein [22]. Up to now, a large number of studies have clarified the important role of miRNA in tumors [23,24]. Changes in miRNA activity are also found to be of great significance in the course of IH [25,26]. For example, miR-424 inhibits the bFGF/FGFR1 pathway and prevents the phosphorylation of ERK1/2, thereby inhibiting the development of IH [27]; miR-187-3p and miR-382 could improve the condition of IH [28,29]. The intervention effect of miR-206 in tumors has been extensively studied. For example, in hepatocellular carcinoma in which tumor cells sustain intense metabolic rewiring [30–33], miR-206 has been shown to repress genes that were involved in carcinogenesis and lipid metabolism [34]. Furthermore, it has been reported that VEGFA is a downstream target gene of miR-206, and the up-regulation of miR-206 promotes the apoptosis of human lung microvascular endothelial cells [35,36]. Research also found that the expression of miR-206 in IH tissues is abnormally reduced [37]. These results suggest that miR-206 may be a specific molecular marker in the development of IH.

This study aimed to explore the regulatory effect and mechanism of miR-206 in IH, and to clarify the effects of ECM metabolism on the proliferation and metastasis of infant human hemangioma endothelial cells (HemECs).

2. Materials and methods

2.1. Ethics statement and sample collection

From 2015 to 2019, the research team collected IH tissue samples and normal subcutaneous tissue samples adjacent to hemangioma from 48 pairs of IH patients at Nanfang Hospital. According to histopathological diagnosis, the 48 IH cases consisted of 27 cases of proliferative hemangioma and 21 cases of involuting hemangioma. In compliance with the ethical requirements of clinical research, an application was submitted to the Nanfang Hospital Ethics Committee and approved (No.2015003-XGWK) before the initiation of this study. Before sample collection, the guardians of the infant patients signed a written informed consent and were informed of the use of the sample tissues for research purpose. All the samples were removed and cleaned by surgical resection, and then frozen in liquid nitrogen for subsequent research.

The animal research was approved by the Nanfang Hospital Ethics Committee (No. 2,019,027-XGWK). Thirty male balb/c nude mice (6 weeks old, SPF grade) ordered from GDMLAC (China) were kept in a breeding center at a temperature of 22℃ and a humidity of 55% and were provided with free access to eating and drinking.

2.2. Separation and processing of HemECs

The hemangioma tissue of one child with proliferative hemangioma was sectioned into 3 tissue blocks of about 1 mm². A petri dish added with complete medium (90% DMEM medium + 10% fetal bovine serum + 100 U/ml penicillin + 100 µg/ml streptomycin) in advance was placed in a 37°C incubator (D180-P, RWD, China) and incubated with the tissue blocks for 1 hour (h). The tissues were then inoculated on the bottom of the petri dish at a space of about 1 mm from each other, when completely attached, the tissues were cultured in the incubator for 4–6 h. After culture, whether the small tissue pieces were firmly attached was confirmed again before turning over the petri dish for continuing the culture. The medium was changed every 2 days (d). After 10–14 d, the HemECs around the tissues were digested using trypsin-EDTA digestion solution (T1300, Solarbio, China) and collected. Next, HemECs at a ratio of 1:3 were transferred to culture flasks for re-cultivation.

To study the effects of miRNA and target genes on HemECs, we transfected miR-206 mimic, miR-206 inhibitor, overexpressed DNA Methyltransferase 3A (DNMT3A) and their respective negative controls into HemECs by liposome transfection techniques. MiR-206 mimic (miR10,000462-1-5), miR-206 inhibitor (miR20000462-1-5) and their negative controls mimic control (miR1N0000002-1-5), inhibitor control (miR2N0000002-1-5) were all synthesized and provided by Guangzhou RIBOBIO Biological Company (China). The overexpressed plasmid of DNMT3A was constructed by inserting the amplified sequence of DNMT3A into the pcDNA3.1(+) vector (V790-20, Invitrogen, USA), and its negative control was constructed by inserting the scramble sequence into the pcDNA3.1(+) vector. A Lipofectamine 3000 kit (L3000008, Invitrogen, USA) was used for transfection. The synthetic plasmids were sequentially transfected into HemECs under the instructions of the kit. 48 h after the transfection, the transfection efficiency of the cells was detected by RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR).

2.3. RNA extraction and qRT-PCR

Before qRT-PCR detection, the RNA isolation reagent Trizol (15,596–026, Invitrogen, USA) was use to extract total RNA from the sample tissues and HemECs, and the RNA concentration was measured with the ALLSHENG Nano-100 spectrophotometer. Then cDNA was synthesized with the extracted RNA using a RevertAid First Strand cDNA Synthesis Kit (K1621, Thermo Scientific™, USA). With reference to a previous report [38], we mixed cDNA, gene primers (Sangon, China) and SYBR® Green I nucleic acid gel stain (S9430, Sigma-Aldrich, USA) in a qRT-PCR reaction system. After adding the system to the PCR detection instrument (T100, BIO-RAD, USA), qRT-PCR detection was performed with a 96-well plate under the following condition: pre-denaturation at 95℃ for 10 minutes (min), denaturation at 95℃ for 15 seconds (s), annealing at 58℃ for 1 min, for a total of 40 cycles. As previously demonstrated [38], the final gene mRNA level was calculated according to the 2^−ΔΔCT calculation method. β-actin and U6 were internal references. The primer sequences used for gene detection are listed in Table 1.

Table 1. Primers for qRT-PCR

Gene	Forward primer (5ʹ-3ʹ)	Reverse primer (5ʹ-3ʹ)
miR-206	GTGGGTCGTATCCAGTGCAA	GTCGTATCCAGTGCGTGTCG
DNMT3A	CGCAAAGGGTTTCACCAAGG	GATGGTGGGGGCCTCATAAA
MMP-2	ACATACAAAGGGATTGCCAGGA	CAAAGGGGTATCCATCGCCA
MMP-9	TCTATGGTCCTCGCCCTGAA	TTGTATCCGGCAAACTGGCT
COL2A1	GTAGAGACCCGGACCCGC	ACTCTCCGAAGGGGATCTCA
Aggrecan	AAGGGCGAGTGGAATGATGT	CGTTTGTAGGTGGTGGCTGTG
U6	CTCGCTTCGGCAGCACA	AACGCTTCACGAATTTGCGT
β-actin	CATGTACGTTGCTATCCAGGC	CTCCTTAATGTCACGCACGAT

2.4. Cell counting cell-8 (CCK-8)

The transfected HemECs were made into a single cell suspension at a concentration of 1 × 10⁵ cells/mL, and then seeded into a 96-well plate. Under conventional incubation conditions, HemECs were allowed to incubate for 24 h, 48 h and 72 h. 10 μL of CCK-8 solution (Cat. No.: HY-K0301, MedChemExpress, USA) was added to each well. To allow CCK-8 to fully bind to the mitochondria in HemECs, the mixture was incubated for another 4 h. Finally, a SynergyTM LX multi-function microplate detector (BioTek, USA) was applied to perform function detection. The effects of different treatment conditions on the viability of HemECs were examined by analyzing the absorbance (OD) of HemECs at 450 nm.

2.5. Colony formation assay

Two hundred HemECs in each group were transferred to petri dishes containing 10 mL of complete medium. The dishes were slowly shaken to evenly distribute the HemECs. The HemECs in the petri dishes were collected after 14 d of routine propagation. During the cultivation, the culture medium was changed every two days. The collected HemECs were stained with Giemsa for 20 min. Finally, a NIKON microscope (E100, Japan) was used to count the number of HemECs in the field of view and calculate relative colony formation rate.

2.6. Wound healing assay

A wound healing-culture insert (80,206, Ibidi, Germany) was applied to create a wound in this study. According to the instructions of the insert, the treated HemECs were inoculated into sterile insert holes. After 24 h of standing, the insert was removed from the petri dish, and 2 mL of HemECs complete medium was added to the petri dish. The location of cell scratches by taking pictures under a microscope (magnification 100 ×) and the time point as 0 h. HemECs were incubated again for 24 h under normal conditions and photographed accordingly.

2.7. Cell invasion assay

Cell invasion detection was performed using Transwell (Corning, USA) and Matrigel in an 8.0 µm Pore Polyester membrane insert. In brief, 3 × 10⁴ HemECs were inoculated on the upper layer of the Transwell chamber covered with homogenized Matrigel glue (356,234), while the lower layer of the Transwell chamber was supplemented with HemEC complete medium. After conventional culture for 48 h, the HemECs cells in the lower layer were collected, washed and fixed. For an easier observation, HemECs were stained with Giemsa. Finally, the number of invading cells was counted with a microscope at a magnification of 100 times.

2.8. Western blot

For total protein extraction, both the separated HemECs and the removed transplanted tumor tissues were fully processed with RIPA Lysis and Extraction Buffer (89,900, Thermo Scientific™, USA). The transplanted tumor tissues were processed into a homogenate prior to the HemECs. The protein concentration was determined using a BCA kit (23,227, Thermo Scientific™, USA) and a spectrophotometer. The protein was then firmly fixed on the 0.22 μm nitrocellulose filter membrane (20 mA, 90 min; N8645, Millipore, USA) after high-temperature denaturation, SDS-PAGE (the electrophoresis voltage of the upper concentrated gel was 80 V, and the electrophoresis voltage of the separation gel was 100V), and blocking treatment at room temperature. According to our experimental design, primary antibodies were used to cover the protein on the membrane overnight at 4℃. Secondary antibody (room temperature, 1.5 h) was probed after washing odd the primary antibody. The formed HRP-labeled antigen-antibody complex was reacted with Immobilon ECL Ultra Western HRP Substrate (WBULS0100, Millipore, USA) for developing fluorescent signal, which was collected by a Gel Doc™ XR+ instrument (BIO-RAD, USA) and processed into a visualized strip chart. The antibodies used in this experiment were DNMT3A (ab232391, 102 kDa, 1:1000), matrix metalloprotein-2 (MMP-2, ab92536, 74 kDa, 1:2000), MMP-9 (ab38898, 92 kDa, 1:1000), collagen 2A1 (COL2A1, ab188570, 142 kDa, 1:3000), Aggrecan (ab3778, 250 kDa, 1:1000), β-actin (1:3000, ab8226, 42kDa, 1:5000), Goat Anti-Rabbit antibody (1:10,000, ab6721), Goat Anti-Mouse antibody (1:10000, ab205719). All the antibodies were produced by Abcam (USA).

2.9. Bioinformatics analysis and verification

The binding sequence of miR-206 and DNMT3A was determined on the TargetScan database (http://www.targetscan.org). The wild-type (WT) and mutant (MUT) reporter gene plasmids of DNMT3A were designed and synthesized by Nanjing COBIOER based on the targeted binding sequence. According to the grouping, DNMT3A-WT and DNMT3A-MUT were co-transfected with miR-206 mimic and mimic control into HemECs for 48 h, respectively. Next, the transfected HemECs were processed with reagents in the TransDetect® Double-Luciferase Reporter Assay Kit (FR201-01, TransGen Biotech, China). After the reaction, the fluorescence intensity of HemECs was analyzed on a GloMax 20/20 dual-luciferase detector (Promega, USA).

2.10. Xenograft experiment

After allowing to adapt to the environment for 3 d, all the nude mice were randomly divided into 3 groups (n = 10), HemECs group (subcutaneous injection of HemECs), HemECs+agomiR-NC group (subcutaneous injection of HemECs transfected with agomiR negative control), and HemECs+agomiR-206 group (subcutaneous injection of HemECs transfected with agomiR miR-206). AgomiR miR-206 (miR40017004-4-5) and its negative control (agomiR-NC, miR4N0000001-4-5) were also provided by RIBOBIO. The treated HemECs (1 × 10⁷ cells/mL) were injected into the armpits (100 μL) of the nude mice in different groups once every 5 days.

During the experiment, we recorded the changes in tumor volume every week from the first week to the fifth week. After 5 weeks of tumorigenesis, all the nude mice were quickly sacrificed by cervical dislocation. The transplanted tumor was removed from the armpit of the nude mice with no vital signs and weighed.

2.11. Statistical analysis

The experimental data were all statistically analyzed by GraphPad Prism 8.0 (GraphPad Software, USA) to obtain the final results. Paired t-test was used to analyze the expression differences of miR-206 and DNMT3A between adjacent, proliferative hemangioma and involuting hemangioma. Differences between multiple groups were compared by one-way ANOVA followed by Tukey’s test. The correlation between miR-206 and DNMT3A was analyzed by Pearson Correlation Coefficient. Statistical significance was determined when p < 0.05.

3. Results

3.1. MiR-206 was down-regulated proliferative hemangioma, but its up-regulation attenuated the malignant development of HemECs by regulating ECM-related genes.

In the clinical samples collected, the expression of miR-206 in proliferative hemangioma tissues was significantly lower than that in adjacent and involuting hemangioma (p < 0.001, Figure 1a). To clarify the effect of miR-206 on IH, we transfected the miR-206 mimic, mimic control, miR-206 inhibitor and inhibitor control into the extracted HemECs. The results showed that miR-206 mimic significantly up-regulated the mRNA expression of miR-206, while the silencing effect of miR-206 inhibitor on the cells was not significant (p < 0.001, Figure 1b). In the following cell function experiments, miR-206 mimic effectively reduced the viability of HemECs and inhibited their proliferation, migration and invasion, whereas miR-206 inhibitor promoted the viability, proliferation, migration and invasion of HemECs (p < 0.05, Figure 1c-E and Figure 2a -D). Further study showed that miR-206 mimic increased the protein and mRNA expression of COL2A1 and Aggrecan, while decreased the protein and mRNA expressions of DNMT3A, MMP-2 and MMP-9 (p < 0.01, Figure 3a-c). In addition, compared with the IC group, the expressions of DNMT3A, MMP-2 and MMP-9 were increased, while the expression of COL2A1 and Aggrecan was decreased (p < 0.05, Figure 3a-c).

Figure 1. MiR-206 was down-regulated in proliferative hemangioma, but its up-regulation reduced the viability and proliferation of HemECs

(A) The mRNA level of miR-206 in adjacent, proliferative hemangioma and involuting hemangioma was detected by qRT-PCR. p < 0.001 vs. proliferative hemangioma. (B) The transfection efficiency of miR-206 mimic and inhibitor was determined by qRT-PCR. U6 was an internal reference. (C) The effects of miR-206 mimic and inhibitor on the viability of HemECs were analyzed by CCK-8 experiment. (D-E) The effects of miR-206 mimic and inhibitor on the proliferation of HemECs were analyzed through colony formation experiments. All experiments were repeated three times to obtain average values. *p < 0.05, ***p < 0.001 vs. MC; ^{^}p < 0.05, ^{^^}p < 0.01 vs. IC. HemECs, human hemangioma endothelial cells; qRT-PCR, quantitative real-time polymerase chain reaction; CCK-8, cell counting kit-8; MC, mimic control; M, miR-206 mimic; IC, inhibitor control; I, miR-206 inhibitor.

Figure 2. MiR-206 reduced the migration and invasion of HemECs

(A-B) The effects of miR-206 mimic and inhibitor on the migration of HemECs were analyzed through wound-healing experiments. The magnification was 100 × . (C-D) The effects of miR-206 mimic and inhibitor on HemEC invasion were analyzed by Transwell experiment. The magnification was 100 × . All experiments were repeated three times to obtain average values. ***p < 0.001 vs. MC; ^{^^^}p < 0.001 vs. IC. MC, mimic control; M, miR-206 mimic; IC, inhibitor control; I, miR-206 inhibitor.

Figure 3. MiR-206 significantly regulated the expressions of DNMT3A and ECM-related genes

(A-B) The effects of miR-206 mimic and inhibitor on the expressions of DNMT3A and ECM-related genes in HemECs were analyzed by Western blot. β-actin was an internal control. (C) The effects of miR-206 mimic and inhibitor on the mRNA expressions of DNMT3A and ECM-related genes in HemECs were analyzed by qRT-PCR. β-actin was an internal control. All experiments were repeated three times to obtain average values. **p < 0.01, ***p < 0.001 vs. MC; ^{^}p < 0.05, ^{^^}p < 0.01, ^{^^^}p < 0.001 vs. IC. DNMT3A, DNA Methyltransferase 3A; ECM, extracellular matrix; MC, mimic control; M, miR-206 mimic; IC, inhibitor control; I, miR-206 inhibitor.

3.2. Up-regulated DNMT3A in proliferative hemangioma could bind to miR-206 and was negatively correlated with miR-206.

To verify the binding sequence between DNMT3A and miR-206 (Figure 4a) acquired from TargetScan, dual-luciferase experiment was conducted. As shown in Figure 4b, co-transfection of DNMT3A-WT and miR-206 mimic reduced the luciferase activity (p < 0.001, Figure 4b). Subsequent experiments confirmed that the expression of DNMT3A in proliferative hemangioma was significantly up-regulated (p < 0.001, Figure 4c), and DNMT3A expression was negatively correlated with the mRNA level of miR-206 (r = −0.488, p = 0.010, Figure 4d).

Figure 4. Up-regulated DNMT3A in proliferative hemangioma could bind to miR-206 and was negatively correlated with miR-206

(A) The TargetScan database (http://www.targetscan.org) was used to determine the binding sequence of miR-206 and DNMT3A. (B) The dual-luciferase experiment was used to verify the binding of miR-206 and DNMT3A. ***p < 0.001 vs. MC. (C) The mRNA level of DNMT3A in adjacent, proliferative hemangioma and involuting hemangioma was detected by qRT-PCR. p < 0.001 vs. proliferative hemangioma. (D) The correlation between miR-206 and DNMT3A expressions was analyzed by Pearson correlation coefficient. r = −0.488, p = 0.010. All experiments were repeated three times to obtain average values. MC, mimic control; M, miR-206 mimic.

3.3. Overexpressed DNMT3A counteracted the inhibitory effect of miR-206 mimic on HemECs

Overexpressed DNMT3A plasmid was transfected into HemECs by lipofection to overexpress DNMT3A (p < 0.05, Figure 5a-c), and we found that the expression of DNMT3A was higher in DNMT3A group than that in NC group, and that DNMT3A overexpression partially neutralized the inhibitory effect of miR-206 mimic on DNMT3A expression (p < 0.05, Figure 5a-c). This finding was further confirmed by the results of a series of rescue experiments. Overexpressed DNMT3A significantly promoted the viability, malignant proliferation, migration and invasion of HemECs (p < 0.001, Figure 5d-F and Figure 6a-D), and it counteracted the inhibitory effect of miR-206 mimic on the malignant development of HemECs (p < 0.05, Figure 5d-F and Figure 6a-D).

Figure 5. Overexpressed DNMT3A counteracted the inhibitory effect of miR-206 mimic on the viability and proliferation of HemECs

(A-C) The effects of overexpressed DNMT3A and miR-206 mimic on DNMT3A expression were analyzed by Western blot (A-B) and qRT-PCR (C). β-actin was an internal reference. (D) The effects of overexpressed DNMT3A and miR-206 mimic on the viability of HemECs were analyzed by CCK-8 experiment. (E-F) The effects of overexpressed DNMT3A and miR-206 mimic on the proliferation of HemECs were analyzed by colony formation experiment. All experiments were repeated three times to obtain average values. *p < 0.05, ***p < 0.001 vs. NC; ^{^}p < 0.05, ^{^^^}p < 0.001 vs. Control; ^###p < 0.001 vs. DNMT3A; ^†p < 0.05, ^††p < 0.01, ^†††p < 0.001 vs. M + NC. NC, negative control; M, miR-206 mimic.

Figure 6. Overexpressed DNMT3A counteracted the inhibitory effect of miR-206 mimic on the migration and invasion of HemECs

(A-B) The effects of overexpressed DNMT3A and miR-206 mimic on the migration of HemECs were analyzed through wound-healing experiments. The magnification was 100 × . (C-D) The effects of overexpressed DNMT3A and miR-206 mimic on the invasion of HemECs were analyzed by Transwell experiment. The magnification was 100 × . All experiments were repeated three times to obtain average values. ***p < 0.001 vs. NC; ^{^^^}p < 0.001 vs. Control; ^#p < 0.05, ^###p < 0.001 vs. DNMT3A; ^††p < 0.01, ^†††p < 0.001 vs. M + NC. NC, negative control; M, miR-206 mimic.

The expressions of ECM-related molecules also showed consistent changes. Overexpressed DNMT3A up-regulated the expression of MMP-2 and MMP-9 yet down-regulated COL2A1 and Aggrecan, and it neutralized the inhibitory effect of miR-206 mimic on MMP-2 and MMP-9 expressions as well as the promotive effect of miR-206 mimic on COL2A1 and Aggrecan expressions (p < 0.01, Figure 7a-c).

Figure 7. Overexpressed DNMT3A counteracted the regulatory effect of miR-206 mimic on ECM-related genes

(A-B) The effects of overexpressed DNMT3A and miR-206 mimic on the expressions of ECM-related genes in HemECs were analyzed by Western blot. β-actin was an internal reference. (C) The effects of overexpressed DNMT3A and miR-206 mimic on the mRNA expressions of ECM-related genes in HemECs were analyzed by qRT-PCR. β-actin was an internal reference. All experiments were repeated three times to obtain average values. **p < 0.01, ***p < 0.001 vs. NC; ^{^^^}p < 0.001 vs. Control; ^#p < 0.05, ^###p < 0.001 vs. DNMT3A; ^††p < 0.01, ^†††p < 0.001 vs. M + NC. NC, negative control; M, miR-206 mimic.

3.4. The up-regulation of miR-206 inhibited the growth of transplanted tumors by down-regulating DNMT3A and promoting ECM accumulation.

The transplanted tumor inoculation experiment verified the results of the cell experiment in vivo. AgomiR-206 significantly increased the expression of miR-206 in transplanted tumors, as compared with HemECs+agomiR-NC group (p < 0.001, Figure 8a). The up-regulation of miR-206 greatly inhibited the increase of tumor volume and weight (p < 0.001, Figure 8b-d). Consistently, agomiR-206 up-regulated the expressions of COL2A1 and Aggrecan genes, but down-regulated the expressions of DNMT3A, MMP-2 and MMP-9 (p < 0.001, Figure 8e-f).

Figure 8. Up-regulation of miR-206 inhibited the growth of xenografts by down-regulating DNMT3A and promoting ECM accumulation

(A) The transfection efficiency of agomiR-206 was evaluated by qRT-PCR. U6 was an internal reference. (B-D) AgomiR-206 suppressed the increase in the size and weight of transplanted tumors. (E-F) The regulatory effect of agomiR-206 on DNMT3A and ECM related proteins in transplanted tumors was analyzed by Western blot. β-actin was an internal reference. All experiments were repeated three times to obtain average values. ^{^^^}p < 0.001 vs. HemECs + agomiR-NC. AgomiR-206, agomiR miR-206; agomiR-NC, agomiR negative control.

4. Discussion

IH is the most common benign tumor among infants and young children, with an incidence of about 4%-10% [39]. Despite the self-limiting property, IH can still affect the vision, appearance and psychology of patients. In severe cases, it can cause disfigurement, airway obstruction, intracranial hemorrhage, and even death [40]. Therefore, active treatment and effective strategies for IH management should be developed.

IH shows different histopathological manifestations at each stage of its natural course [41,42]. A solid mass structure is formed by massive proliferation of endothelial cells and an immature vascular cavity during proliferation stage. Later in the regression period, the vascular cavity continues to expand, the proliferation of endothelial cells slows down, and deposits of fibrous and adipose tissues appear in the intercellular substance. In the regression completion period, the endothelial cells in the tumor begin to degenerate, and the vascular cavity is replaced by a large amount of fibrous fat tissue, collagen and mesh organization. It is not difficult to see that the regression period and the regression completion period play an important role in the formation and accumulation of ECM. In our basic cell function experiments, the up-regulation of miR-206 significantly reduced the biological function and malignant development of HemECs. This was because miR-206 promoted ECM accumulation through silencing MMP-2/MMP-9 and up-regulating COL2A1/Aggrecan. Both MMP-2 and MMP-9 are important members of the MMP family, and their silencing is indicative of increased accumulation of ECM [43]. COL2A1 encodes type II collagen α1 chain and further synthesizes type II collagen. Type II collagen and Aggrecan are the main components of ECM [44]. Up-regulated COL2A1 and Aggrecan also suggest increased accumulation of ECM. Our findings are also supported by the experimental evidence of other researchers. For instance, Anita Rotter showed that in the proliferation phase of IH, vascular endothelial growth factor (VEGF) is upregulated, and the positive expression rates of MMP-2 and MMP-9 were significantly increased [45]; Katarzyna Taran found that propranolol treatment could significantly reduce the high expressions of MMP-2 and MMP-9 in the tissues of IH patients, and believed that the mRNA expressions of MMP-2 and MMP-9 in tissues and serum could serve as a diagnostic criterion for IH [46]. Thus, to investigate whether miR-206 can activate ECM accumulation has become the direction for our further research.

Studies have reported that miR-206 inhibits cancer cell proliferation and promotes apoptosis of a variety of cancer cells, such as lung cancer, colon cancer, and prostate cancer [47–49]. MiRNAs can be combined with gene sequences in a complementary paired manner to inhibit gene translation and expression, thus affecting protein synthesis and eventually affecting disease occurrence and development [50,51]. Therefore, finding the target genes of miRNAs and clarifying their regulatory effects have important significance in understanding the mechanism of miRNA expression regulation. This study found that DNMT3A was the target gene of miR-206. Consistently, Wang et al [52] also reported that DNMT3A is as a direct target of miR-206, and there is a negative feedback interaction between miR-206 and DNMT3A in laryngeal squamous cell carcinoma. DNMT3A, which is a member of the DNMT family and a synthetic DNA methyltransferase, functions to produce specific DNA methylation patterns. Abnormal DNA methylation can lead to developmental abnormalities and pathological changes [53]. Increasing evidence has showed that cancer is related to DNA methylation. DNMT3A could induce aberrant CpG site methylation in human cancer cells [54]. Methylation-mediated silencing of miR-133a-3p promotes breast cancer cell migration and stemness via miR-133a-3p/MAML1/DNMT3A positive feedback loop [55]. MiR-708-5p suppressed NSCLC initiation, development, and stemness through interfering DNMT3A-dependent DNA methylation [56]. These studies indicated that DNMT3A played a key role in the human cancer. Researchers have observed that a lower DNA methylation level in a variety of cancers is related to tumor occurrence and invasion [57]. Our results showed that the expression of DNMT3A in hemangioma tissues of children with IH was significantly higher than that in normal tissues adjacent to the tumor. In in vitro experiments, we found for the first time that DNMT3A improved the activity and malignant metastasis ability of HemECs. Moreover, consistent with the results of Ma et al. and Liu et al [58,59], our exploration results on the molecular mechanism revealed that DNMT3A reduced the accumulation of ECM by regulating related genes.

Based on the above experimental results, we confirmed that the combination of miR-206 and DNMT3A inhibits the activity of DNMT3A, thereby activating ECM accumulation, and ultimately inhibiting the malignant development of HemECs to alleviate the symptoms of IH. The differential expression and important intervention effects of miR-206 and DNMT3A in IH suggest that miR-206 and DNMT3A may be potential molecular markers for the diagnosis and treatment of IH.

Availability of Data and Materials

The analyzed data sets generated during the study are available from the corresponding author on reasonable request.

Acknowledgments

Not applicable

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [81701916].