LncRNA HOTTIP Knockdown Attenuates Acute Myocardial Infarction via Regulating miR-92a-2/c-Met Axis

Venous blood samples were collected from 42 MI patients (average age: 54.2 ± 4.1, male/female = 22/19) and 42 healthy volunteers (average age: 54.6 ± 3.2, male/female = 23/21) in the First Affiliated Hospital of University of Science and Technology of China. After centrifugation, serum samples were collected and used for quantitative real-time polymerase chain reaction (qRT-PCR) to detect the expression patterns of lncRNA HOTTIP.

C57BL/6 mice (male, approximately 8–20 weeks) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. and kept under standard conditions. Mouse MI model was constructed through left anterior descending coronary artery (LAD) occlusion as previously described [23]. Mice were randomly divided into 5 groups, including sham group, MI group, si-HOTTIP group, miR-92a-2 group (miR-92a-2 mimics), and combined group (si-HOTTIP and miR-92a-2 mimics) with 7 mice in each group. 50 nM of these si-RNAs or mimics were injected into left ventricles through intramyocardial injection each day for 1 week. Mice in all groups except the sham group were subjected to MI operation. At 12 h after MI, blood samples were collected for detection of serum myocardial enzyme activities. After that, mice were euthanized, and myocardial tissues were collected for the subsequent experiments. All procedures were approved by the Ethical Committee of the First Affiliated Hospital, Division of Life Sciences and Medicine, University of Science and Technology of China.

Mouse cardiomyocytes were isolated from 1 to 3-day-old neonatal C57BL/6 mice as previously described [24]. Briefly, the dissected hearts from mice were rinsed with HEPES-buffered saline solution, cut into pieces, and digested with 0.25% trypsin. Subsequently, cells were maintained in Dulbecco’s modified Eagle medium/F-12 (DMEM/F-12; Gibco, Carlsbad, CA, USA) containing 5% fetal bovine serum (Hyclone, Logan, UT, USA), 0.1 mM ascorbate in insulin-transferring-sodium selenite media supplement (Sigma, St. Louis, MO), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.1 mM bromodeoxyuridine (Gibco) in an atmosphere at 37 °C with 5% CO₂.

H/R injury in cardiomyocytes was induced as previously described [25]. In brief, cardiomyocytes were cultured with serum-free medium under normoxic conditions for 24 h. Cells were then cultured under hypoxia (5% CO₂, 95% N₂) for different times (12, 24, and 48 h) and subsequently under re-oxygenation condition (5% CO₂, 95% O₂) for another 4 h to stimulate H/R injury. In addition, control cells were cultured under normoxic conditions all the time. Finally, cells were harvested for the subsequent experiments.

Si-HOTTIP (siRNA against HOTTIP, 5′-AGGCTGAGCTAATACAGTA-3′), si-NC, miR-92a-2 mimics, miR-NC, and si–c-Met (siRNA against c-Met, 5′-ATCTTGAGCCATTCACCGGAA-3′) were purchased from Shanghai GenePharma Co., Ltd. (Shanghai, China) and transfected into mouse cardiomyocytes using Lipofectamine 2000 (Invitrogen). To overexpress HOTTIP and c-Met, the cDNA sequence of HOTTIP and c-Met were amplified with mouse genome as the template and cloned into pcDNA3.1 Expression vector to generate pc-HOTTIP and pc-c-Met. Cells transfected with the empty vector were used as the negative control (pc-NC). Similarly, pc-HOTTIP, pc-c-Met, or pc-NC was transfected into mouse cardiomyocytes using Lipofectamine 2000.

Cell viability was evaluated using a Cell Counting Kit-8 (CCK-8, Dojindo Molecular Technologies, Gaithersburg, MD) as previously described [26]. Briefly, approximately 3 × 10³ transfected cardiomyocytes were seeded into 96-well plates and incubated with additional CCK-8 reagent at 0, 24, 48, 72, and 96 h for another 4 h. Finally, the absorbance at 450 nm was detected with a microplate reader.

To determine the interaction between HOTTIP and miR-92a-2, RIP assay was performed using the EZMagna RIP Kit (Millipore) following the manufacturer’s instructions. Briefly, cells lysates were incubated with magnetic beads conjugated to Ago2 antibody or negative control IgG (Abcam) overnight at 4 °C. On the next day, cell lysates were incubated with 30 μL of magnetic beads for another 2 h at 4 °C, and the immune-precipitated RNAs were purified and used for qRT-PCR analysis.

Total RNAs were extracted from myocardial tissues of MI mice or cultured cells using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). After reversely transcribing into cDNA using the PrimeScript RT reagent Kit (TaKaRa), qRT-PCR analysis was carried out using SYBR Green PCR Kit (Thermo) on an ABI 7500 Fast Real-Time PCR system (Applied Biosystems, Carlsbad, CA, USA). The expression of target genes was normalized to GAPDH or U6 using the 2^−ΔΔCt method [27] with GAPDHA and U6 as the internal references. The primers used in this study were HOTTIP forward 5′-TCTGGTATTGCCTGGAACGCCAA-3′ and reverse 5′-CAGTGGTAATCAAGTGGAGAATC-3′, miR-92a-2 forward 5′-GAGCTAGCGAATGGCACCCT-3′ and reverse 5′-GCAGGAACGAAGTCGACTTA-3′, c-Met forward 5′-GAGCCACTTAGAATCGAGGA-3′ and reverse 5′-CTGAGGCTATAGATTCGTGCC-3′, GAPDH forward 5′-CCAAGGTCATCCATGACAAC-3′ and reverse 5′-GCTTCACCACCTTCTTGATG-3′, and U6 forward 5′-AGAGAAGATTAGCATGGCCCCTG-3′ and reverse 5′-AGTGCAGGGTCCGAGGTATT-3′.

Total proteins from cultured cells were extracted using RIPA Lysis Buffer (Beyotime, Shanghai, China). Approximately equal amounts of proteins were separated by 10% SDS-PAGE and transferred onto PVDF membranes. After blocking with 5% non‐fat milk, the membranes were incubated with primary antibodies against c-Met (1:1000, Abcam), Bax (1:1000, Abcam), Bcl-2 (1:1000, Abcam) and GAPDH (1:10,000, Cell Signaling Technology) overnight at 4 °C. The membranes were subsequently exposed to horseradish peroxidase (HRP)-conjugated secondary antibodies for 2 h at room temperature. The protein bands of interest were detected using the ECL kit (Millipore), and protein levels were quantified using ImageJ software.

Targetscan v7.2 and Starbase v.8 web tools were used to predict the binding site between HOTTIP and miR-92a-2 and between miR-92a-2 and c-Met. The wild type (WT) or mutant type (MUT) 3′-UTR of HOTTIP and c-Met containing the putative miR-92a-2 binding site was inserted into the pmirGLO vectors (Promega, Madison, WI, USA). Then mouse cardiomyocytes were co-transfected with luciferase reporter vectors and miR-92a-2 or miR-NC using Lipofectamine 2000 (Invitrogen). At 48 h post-transfection, relative luciferase activity was evaluated by the dual-luciferase reporter system.

Cardiomyocyte apoptosis was evaluated with Annexin V-FITC Apoptosis Kit (BD Biosciences). Briefly, cardiomyocytes under normoxia or hypoxia conditions were stained with Annexin V-fluorescein isothiocyanate (FITC) and propidium iodide (PI) and then analyzed by flow cytometry without re-oxygenation.

The day before euthanization, mice in different groups were subjected to M-mode echocardiography to analyze left ventricular ejection fraction (LVEF), left ventricular fractional shortening (LVFS), left ventricular end-systolic diameter (LVESd) and left ventricular end-diastolic diameter (LVEDd) with ATL HDI 5000 ultrasound system with parameters of 15 MHz linear and 12-MHz sectorial scan heads.

After euthanization, the myocardial tissues of mice were collected, cut into 5-μm-thick sections, embedded with paraffin, and stained with TTC solution for 15 min. After fixation with formalin, samples were photographed using a digital camera, and infarct size was analyzed using Image J software (Olympus).

IHC and IF staining were performed as previously described [28]. Briefly, 5-µm-thick paraffin-embedded sections of myocardial tissues were incubated with primary antibodies against c-Met (1:200, Abcam) and Cas-3 (1:200, Abcam) overnight at 4 °C respectively. After washed with PBS, the protein signals were visualized by incubating with Alexa FluorVR 647-labeled goat anti-rabbit IgG (ab150083, 1:1500, Abcam) and biotin-labeled goat anti-mouse lgG (ZSGB-Bio, Beijing, China) as secondary antibodies for 60 min. The samples were covered using anti-fluorescence quenching tablets containing 4,6-diamidino-2-phenylindole (DAPI). The immunohistochemistry staining was stopped with a DAB chromogen kit (ZSGB-BIO, Beijing, China) before hematoxylin staining. The positive staining signals were observed under a fluorescence microscope (Nikon, Japan).

Some 5 µm-thick paraffin-embedded sections of myocardial tissues were incubated with recombinant terminal deoxynucleotidyl transferase (rTdT) solution at 37 °C for 1 h in the dark and stained with DAPI solution. TUNEL stained cells were observed under a fluorescence microscope. The apoptotic index was evaluated as the percentage of TUNEL-positive cells to the total cells. In addition, some sections were stained with hematoxylin–eosin solution and observed under an optical microscope as previously described [29].

Serum creatine kinase MB (CK-MB) and lactate dehydrogenase (LDH) levels in mice from different groups were detected using a Creatine Kinase Activity Assay kit and a LDH Colorimetric Assay kit, respectively.

Data were presented as mean ± SD derived from at least three independent experiments and statistically analyzed using GraphPad Prism software. Differences were determined by Student’s t test (two groups) or one-way ANOVA (among three or more groups). P < 0.05 was considered statistically significant.

HOTTIP expression in the serum samples of MI patients was detected. The results showed that HOTTIP expression was upregulated in MI patients compared with the healthy controls (Fig. s1). To investigate the potential role of HOTTIP in MI, HOTTIP expression in hypoxia-treated cardiomyocytes and myocardial tissues of MI mice was evaluated. The results indicated that HOTTIP was significantly upregulated in hypoxia-treated cardiomyocytes in a time-dependent manner (p < 0.05) (Fig. 1A) and in myocardial tissues of MI mice compared with the sham mice (p < 0.01) (Fig. 1B). In addition, miR-92a-2 was downregulated in mouse cardiomyocytes exposed to hypoxia in a time-dependent manner (p < 0.05) (Fig. 1C). Consistent with the findings in vitro, miR-92a-2 was also obviously downregulated in myocardial tissues of MI mice compared with the sham mice (p < 0.01) (Fig. 1D). These results suggested that HOTTIP and miR-92a-2 might play crucial roles in ischemic cardiac injury.

The potential targets of HOTTIP were predicted using Starbase. The results revealed a putative binding site between HOTTIP and miR-92a-2 (Fig. 2A), suggesting that miR-92a-2 might be a target of HOTTIP. Luciferase reporter assay was performed and indicated that miR-92a-2 mimics significantly decreased the relative luciferase activity of WT HOTTIP vector (p < 0.01), while exhibited no impact on the MUT vector (Fig. 2B). Meanwhile, RIP assay using anti-Ago2 was carried out and found that both HOTTIP and miR-92a-2 were significantly enriched in immune-precipitates containing anti-Ago2 compared with that containing negative control anti-IgG (p < 0.01) (Fig. 2C). In addition, mouse cardiomyocytes were transfected with si-HOTTIP, si-NC, pc-HOTTIP or pc-NC, and the expression of HOTTIP and miR-92a-2 was evaluated by qRT-PCR. The results showed that si-HOTTIP significantly decreased HOTTIP expression (p < 0.01) and increased miR-92a-2 expression compared with si-NC (p < 0.001), while pc-HOTTIP markedly increased HOTTIP expression (p < 0.001) and decreased miR-92a-2 expression compared with pc-NC (p < 0.01) (Fig. 2D). Furthermore, the expression level of MiR-17-92a-1 Cluster Host Gene (MIR17HG) did not decrease in the myocardial tissues of MI mice (Fig. s2). These results suggested that HOTTIP plays its role in MI via directly sponging miR-92a-2 to decrease miR-92a-2 level.

Next, Targetscan was applied to predict the potential targets of mIR-92a-2, and the results indicated that c-Met might be a target of miR-92a-2 (Fig. 3A). Luciferase reporter assay showed that miR-92a-2 mimics significantly decreased the relative luciferase activity of WT c-Met vector compared with miR-NC (p < 0.01) but had no obvious effect on that of MUT c-Met vector (Fig. 3B). Mouse cardiomyocytes were transfected with miR-92a-2 mimics, miR-NC, or co-transfected with miR-NC and pc-HOTTIP or miR-92a-2 mimics and pc-HOTTIP. QRT-PCR indicated that miR-92a-2 mimics significantly increased miR-92a-2 expression compared with miR-NC (p < 0.001), and pc-HOTTIP decreased miR-92a-2 expression compared with miR-NC (p < 0.01), while co-transfection of miR-92a-2 mimics and pc-HOTTIP obviously reversed the effect of miR-92a-2 mimics (p < 0.01) (Fig. 3C). Meanwhile, miR-92a-2 mimics markedly decreased c-Met expression at both mRNA (p < 0.01) (Fig. 3C) and protein (p < 0.01) (Fig. 3D) levels compared with miR-NC. HOTTIP overexpression significantly increased c-Met expression at both mRNA (p < 0.01) (Fig. 3C) and protein (p < 0.01) (Fig. 3D) levels compared with miR-NC. By contrast, co-transfection of miR-92a-2 mimics and pc-HOTTIP obviously reversed the effect of miR-92a-2 mimics on c-Met expression at mRNA level (p < 0.01) (Fig. 3C) and protein level (p < 0.05) (Fig. 3D). In addition, IHC assay showed that c-Met expression was obviously increased in myocardial tissues of MI mice compared with that of the sham mice (Fig. 3E). Similarly, c-Met expression was increased in hypoxia-treated cardiomyocytes in a time-dependent manner (p < 0.05) (Fig. 3F). These results suggested that HOTTIP might elevate c-Met expression through sponging miR-92a-2 in MI.

Next, cardiomyocytes were transfected with si-HOTTIP, si-NC, miR-92a-2 mimics, or co-transfected with miR-92a-2 mimics and si-HOTTIP, and cell viability and apoptosis were evaluated. As shown in Fig. 4A and Fig. 4B, hypoxia significantly decreased viability and induced apoptosis of cardiomyocytes compared with normoxia condition (p < 0.001), and si-HOTTIP and miR-92a-2 mimics significantly promoted viability and inhibited apoptosis of hypoxia-treated cardiomyocytes compared with si-NC (p < 0.01), and co-transfection of miR-92a-2 mimics and si-HOTTIP further enhanced cell viability and inhibited apoptosis of hypoxia-treated cardiomyocytes compared with si-HOTTIP group (p < 0.05). We further detected the expression of Bcl-2 and BIM. The results showed that hypoxia significantly decreased bcl-2 expression and induced BIM expression compared with normoxia condition (p < 0.001), and si-HOTTIP and miR-92a-2 mimics significantly promoted bcl-2 expression and inhibited BIM expression compared with si-NC in hypoxia-treated cardiomyocytes (p < 0.01). Furthermore, co-transfection of miR-92a-2 mimics and si-HOTTIP further enhanced the effect of si-HOTTIP (Fig. 4C, p < 0.05). Similarly, the results of CCK-8 assay (Fig. 4D) and flow cytometry assay (Fig. 4E) showed that both si–c-Met and si-HOTTIP significantly promoted viability and inhibited apoptosis of hypoxia-treated cardiomyocytes compared with si-NC (p < 0.01), and co-transfection of si–c-Met and si-HOTTIP further enhanced cell viability and inhibited apoptosis of hypoxia-treated cardiomyocytes compared with si-HOTTIP group (p < 0.05). We also found that both si–c-Met and si-HOTTIP significantly promoted bcl-2 expression and inhibited BIM expression in hypoxia-treated cardiomyocytes compared with si-NC (Fig. 4F, p < 0.01). We further demonstrated that c-Met overexpression reversed the protective effect of si-HOTTIP or miR-92a-2 mimics on AMI progression, including cardiomyocyte apoptosis (Fig. s3). These results indicated that HOTTIP knockdown promoted growth and inhibited apoptosis of hypoxia-treated cardiomyocytes by modulating the miR-92a-2/c-Met axis in vitro.

To further determine the cardio-protective effects of HOTTIP in vivo, si-HOTTIP, miR-92a-2 mimics, or si-HOTTIP plus miR-92a-2 mimics were injected into mice through intramyocardial injection. The results of qRT-PCR showed that si-HOTTOP significantly decreased HOTTIP expression (p < 0.01), increased miR-92a-2 expression (p < 0.01), and decreased c-Met expression (p < 0.01); MiR-92a-2 mimics markedly increased miR-92a-2 expression (p < 0.01) and decreased c-Met expression (p < 0.01); Co-injection of si-HOTTIP and miR-92a-2 mimics obviously further elevated the effect of si-HOTTIP on the expression of miR-92a-2 (p < 0.05) and c-Met (p < 0.05) (Fig. 5A). Meanwhile, si-HOTTIP and miR-92a-2 mimics all significantly decreased c-Met protein level (p < 0.05), and si-HOTTIP plus miR-92a-2 mimics further decreased c-Met protein level compared with si-HOTTIP group (p < 0.05) (Fig. 5B). In addition, we found that the infarct size of myocardial tissues was significantly increased in the MI group and HOTTIP knockdown and miR-92a-2 mimics obviously decreased infarct size. Meanwhile, the inhibitory effect of si-HOTTIP was significantly enhanced by miR-92a-2 mimics (Fig. 5C). In addition, cardiac function indexes including LVEF (Fig. 5D), LVFS (Fig. 5E), LVESd (Fig. 5F), and LVEDd (Fig. 5G) of different groups were evaluated by echocardiographic measurement. The results indicated that LVEF and LVFS of myocardial tissues in MI group were significantly decreased, while LVESd and LVEDd of myocardial tissues in MI group were increased compared with the sham group (p < 0.01). HOTTIP knockdown led to a significant increase in LVEF and LVFS and a marked decrease in LVESd and LVEDd (p < 0.05). Meanwhile, the protective effect of si-HOTTIP was significantly enhanced by miR-92a-2 overexpression (p < 0.05). These results revealed that HOTTIP downregulation could effectively alleviate myocardial dysfunction through upregulating miR-92a-2 in vivo.

We also evaluated the degree of myocardial cell injury by measuring serum CK-MB and LDH levels. The results (Fig. 6A and B) showed that serum CK-MB and LDH levels were significantly increased in the MI group than in the sham group (p < 0.001), and both si-HOTTIP and miR-92a-2 mimics markedly decreased serum CK-MB and LDH levels (p < 0.01). Meanwhile, miR-92a-2 mimics significantly enhanced the inhibitory effect of si-HOTTIP on serum CK-MB and LDH levels (p < 0.05). The myocardial infarct size was increased in all groups except for the sham group (p < 0.05). Si-HOTTIP and miR-92a-2 mimics significantly decreased the myocardial infarct size (p < 0.05), while the inhibitory effect of si-HOTTIP was enhanced by miR-92a-2 mimics (p < 0.05) (Fig. 6C). TUNEL staining assay was used to evaluate cardiomyocyte apoptosis rate in mouse heart tissues, and the results indicated that apoptosis rate was significantly increased in MI group than in the sham group (p < 0.05). Both si-HOTTIP and miR-92a-2 mimics markedly decreased apoptosis in MI mice (p < 0.05). Meanwhile, the inhibitory effect of si-HOTTIP was obviously enhanced by miR-92a-2 mimics (p < 0.05) (Fig. 6D and E). H&E staining assay (Fig. 6E) also showed a similar trend with TUNEL staining assay. In other words, cardiomyocyte apoptosis was significantly increased in MI group than in the sham group, and both si-HOTTIP and miR-92a-2 mimics obviously inhibited cardiomyocyte apoptosis. Furthermore, the protective effect of si-HOTTIP was obviously enhanced by miR-92a-2 mimics. IF staining also showed that Cas-3 expression in MI was inhibited when treated with si-HOTTIP or miR-92a-2 mimics (Fig. s4). In addition, the apoptosis-related makers, including pro-apoptotic factor Bax and anti-apoptotic factor Bcl-2, in mouse heart tissues was evaluated by Western blot (Fig. 6F). The results indicated that Bax was significantly upregulated (p < 0.001), while Bcl-2 (p < 0.001) was downregulated in MI mice compared with the sham mice. Si-HOTTIP and miR-92a-2 mimics markedly decreased Bax expression (p < 0.05) and increased Bcl-2 expression (p < 0.05), and si-HOTTIP plus miR-92a-2 mimics further enhanced Bcl-2 expression (p < 0.05) and decreased Bax expression (p < 0.05). These results demonstrated that HOTTIP downregulation effectively improved myocardial injury and inhibited cardiomyocyte apoptosis by targeting miR-92a-2 in vivo.