Human ventricular cardiomyocytes (AC-16) cells were obtained from Beijing Institute for Cancer Research Collection (Beijing, China) and cultured in Dulbecco's modified Eagle's medium/F-12 supplemented (DMEM/F12, Invitrogen, Waltham, MA, USA) supplemented with 12.5 % fetal bovine serum (FBS, Gibco, Carlsbad, CA, USA) and 1 % antibiotic-antimycotic (Gibco) with 5 % CO₂ at 37 °C. Hypoxia was induced by exposing AC-16 cells to 1 % O₂, 94 % N₂, and 5 % CO₂ for 12, 24, or 48 h using a modular incubator. Cells grown under a normoxic atmosphere (incubation with hypoxia condition for 0 h) were used as the control.

When cells were grown to 80 %-90 % confluency, 100 ng of circ_0010729 or TNF Receptor Associated Factor 6 (TRAF6) overexpression vector (circ or TRAF6) or nontarget plasmid (vector or pcDNA) (Promega, Madison, WI, USA), 50 nM of small interfering RNA (siRNA) against circ_0010729 (si-circ) or siRNA negative control (si-NC) (GenePharma, Shanghai, China), 40 nM of miR-370-3p mimic or miR-370-3p inhibitor (miR-370-3p, anti-miR-370-3p) or their negative control (miR-NC, anti-NC) (GenePharma) were transfected into AC-16 cells using Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA).

Following transfection and/or treatment, AC-16 cells (5000/well) were cultivated into a 96-well plate and co-interacted with 10 μL CCK-8 solution (Dojindo Molecular Technologies, Japan) at 37 °C for 4 h. The absorbance of each well was measured at 450 nm using a microplate reader. The results represented as the average of three independent replicates.

After transfection and/or treatment, for cell cycle analysis, AC-16 cells were firstly digested by trypsin to obtain single-cell suspensions. After washing by PBS twice, cells were fixed by 75 % ethanol for 4 h at 4 °C, followed by incubation with 500 uL propidium iodide (PI) staining solution for 15 minutes. The quantitation of cell cycle distribution was analyzed using a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA) with FlowJo software. For cell apoptosis analysis, AC-16 cells were harvested and washed in PBS, then double-stained with 10 μL of Annexin V-FITC and PI (BD Biosciences) for 15 min. Cell apoptosis was analyzed by the flow cytometer. All experiments were repeated three times independently.

The activity of caspase3 was assessed using the commercial caspase-3 activity assay kit (Beyotime, Shanghai, China) following the guidance of producer. The activity was proportional to the absorbance which was detected at optical density (OD) 405 nm using the microplate reader. The results were represented as the average of three independent replicates.

A transwell insert (Cell Biolabs, Inc. Santiago, CA, USA) without Matrigel (BD Biosciences) was employed to detect cell migration. Following transfection and/or treatment, AC-16 cells suspended in 200 μL serum-free medium were placed into the upper chamber of Transwell, then 600 μL medium fixed with FBS was added into the bottom chamber. After incubation for 24 h, migrated cells on the lower face of the chamber were counted by an inverted light microscope in five random fields (100 ×). Experiments were performed three times.

The concentrations of interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) from the supernatants of AC-16 cells following appropriate transfection and/or treatment were determined using commercial IL-6 and TNF-α ELISA kits (R&D Systems, Minneapolis, Minnesota, USA) referring to the instructions of protocol. The results represented as the average of three independent replicates.

After transfection and/or treatment, the supernatants of AC-16 cell culture media were collected, and subjected to the analysis of the consumption or production of glucose and lactate using a Glucose Uptake Assay Kit and L-Lactate Assay Kit (Sigma, St Louis, MO, USA) referring to the producer's guidance using a microplate reader. Experiments were performed three times.

An ATP Assay Kit (Sigma) was applied to detect the level of ATP. Sonicated AC-16 cells were lysed, and the lysate was fixed with ATP reaction mix for 30 min. Finally, the OD570 nm value was examined using a microplate reader. Experiments were performed three times.

Total RNA was extracted using Trizol reagent (Invitrogen) from cells. Then reverse transcription was performed using 1 μg of total RNA with a reverse transcription kit (Takara, Tokyo, Japan) to synthesize cDNA. Subsequently, qRT-PCR was conducted using SYBR Green PCR master mix (Takara) on an ABI 7500 Real-Time PCR system. The 2^−ΔΔCt method was used to calculate the fold changes with U6 or glyceraldehyde 3-phosphate dehydrogenase (GADPH) as an internal control. The same experiment was repeated three times, and the average was taken. The following primers were used: circ_0010729: F, 5'-CAGGCAGAGGTCCGGGCCTGTT-3' and R, 5'-GGACCGTTCTCAATGGCGTATAC-3'; GADPH: F, 5'-GGTGAAGGTCGGAGTCAAC-3' and R, 5'-AGAGTTAAAAGCAGCCCTGGTG-3'; TRAF6: F, 5'-CAGTGGTCGTATCGTGCTTA-3' and R, 5'-CCTTATGGT TTCTTGGAGTC-3'; miR-370-3p: F, 5'-GCCTGCTGGGGTGGAACCTGGT-3' and R, 5'-CTCAACTGGTGTCGTGGA -3'; U6: F, 5'-CTCGCTTCGGCAGCACA-3' and R, 5'-AACGCTTCACGAATTTGCGT-3'.

The sequences of circ_0010729 or TRAF6 3'UTR containing the wild-type or mutant potential binding sites of miR-370-3p were cloned into the pmirGLO luciferase vector (Promega), named wild-type/mutant-circ_0010729 or wild-type/mutant-3'UTR TRAF6. Then AC-16 cells placed on the 6-well plates were transfected with these constructed reporter plasmids and miR-370-3p or miR-NC using Lipofectamine 3000 (Invitrogen). Luciferase activities were determined using a dual-luciferase reporter assay kit (Promega). Each group was run in triplicate in 6-well plates.

AC-16 cells were lysed using RIP buffer, and then incubated with RIPA buffer containing magnetic beads conjugated with human Anti-Ago2 antibody (Millipore, Billerica, MA, USA) or normal mouse Anti-IgG (Millipore). After interaction with Proteinase K, the immunoprecipitated RNA was extracted and purified RNA was determined using qRT-PCR. All experiments were repeated three times independently.

Proteins were extracted from cells using RIPA lysis buffer (Beyotime), and approximately 30 μg of extracted protein was subjected to western blot assay as described previously (Park et al., 2018[17]). The following antibodies were used: TRAF6 (1:2000, ab181622), and HRP-conjugated antibody (1:1000, ab9482), which all were obtained from Abcam (Cambridge, MA, USA). β-actin (1: 2000; #ZRB1312, Sigma) served as an internal control, and protein bands were visualized by a Super ECL assay kit (YRBIO, Changsha, Hunan, China). Triplicate individual experiments were performed in this study.

Data from thrice-repeated experiments were exhibited as mean ± standard deviation (SD). All quantitative data were analyzed using the Student's t-test, non-parametric test (Mann-Whitney U tests) (two groups) and one-way analysis of variance (ANOVA) (three or more groups). P values < 0.05 were considered statistically significant.

First, the effects of hypoxia on cardiomyocyte properties were tested. Cardiomyocyte AC-16 cells were exposed to hypoxia for 0, 12, 24, and 48 h, by contrast with the control (0 h) group, hypoxia led to AC-16 cell viability suppression (Figure 1A(Fig. 1)), cell cycle arrest (Figure 1B(Fig. 1)), caspase3 activity enhancement (Figure 1C(Fig. 1)), cell apoptosis promotion (Figure 1D(Fig. 1)), as well as cell migration inhibition (Figure 1E(Fig. 1)). Besides, the levels of inflammatory factor IL-6 and TNF-α were found to be significantly increased under hypoxia at 12, 24, or 48 h (Figure 1F, G(Fig. 1)). Results in Figure 1H-J(Fig. 1) exhibited hypoxia suppressed glycolysis in AC-16 cells, reflected by the decrease of ATP production (Figure 1H(Fig. 1)), glucose uptake (Figure 1I(Fig. 1)) and lactate production (Figure 1J(Fig. 1)) at 12, 24 and 48 h exposure. These results suggested that hypoxia-induced cardiomyocyte injury and suppressed glycolysis.

The molecular mechanism of hypoxia-modulated injury and glycolysis in cardiomyocytes was then investigated. We found circ_0010729 was elevated by hypoxia exposure at 12, 24, and 48 h (Figure 2A(Fig. 2)). To investigate whether the promotion of the hypoxia-induced circ_0010729 could protect cardiomyocytes in hypoxia conditions, AC-16 cells were transfected with circ_0010729 (circ) or specific si-circ_0010729 (si-circ). As expected, circ_0010729 expression was markedly up-regulated in AC-16 cells when transfected with circ_0010729, while circ_0010729 expression was down-regulated by si-circ_0010729 compared with their counterparts, respectively (Figure 2B(Fig. 2)). After treatment with hypoxia for 24 h, circ_0010729 knockdown promoted cell viability (Figure 2C(Fig. 2)) and cell cycle progression (Figure 2D(Fig. 2)), suppressed caspase3 activity (Figure 2E(Fig. 2)), apoptosis (Figure 2F(Fig. 2)), and migration (Figure 2G(Fig. 2)), reduced IL-6 and TNF-α release (Figure 2H, I(Fig. 2)), as well as enhanced ATP production (Figure 2J(Fig. 2)), glucose uptake (Figure 2K(Fig. 2)) and lactate production (Figure 2L(Fig. 2)) in hypoxia-treated AC-16 cells, while the introduction of circ_0010729 in AC-16 cells exhibited inverse effects (Figure 2C-L(Fig. 2)). Taken together, knockdown of circ_0010729 might protect cardiomyocytes through restoration of cardiomyocyte properties and glucose metabolism.

To explore molecular mechanism underlying the action of circ_0010729 in hypoxia-modulated cardiomyocyte properties, the online database CircInteractome was applied to predict the potential microRNA (miRNA) that could be interacted with circ_0010729. Then miR-370-3p was identified to have the potential binding sites of circ_0010729 (Figure 3A(Fig. 3)). Afterwards, the transfection efficiency of miR-370-3p or miR-NC was validated, as expected, miR-370-3p expression was greatly overexpressed in AC-16 cells after miR-370-3p transfection (Figure 3B(Fig. 3)). Immediately, the dual-luciferase reporter assay showed miR-370-3p overexpression significantly reduced the luciferase activity in AC-16 cells transfected with wild type-circ_0010729 (Figure 3C(Fig. 3)). Meanwhile, data from RIP assay revealed that circ_0010729 and miR-370-3p were highly enriched in the complex precipitated by Anti-Ago2 compared with nonspecific Anti-IgG (Figure 3D(Fig. 3)). Importantly, the effect of circ_0010729 on miR-370-3p expression was investigated, qRT-PCR analysis indicated miR-370-3p expression in AC-16 cells was decreased by circ_0010729 overexpression, but increased by circ_0010729 down-regulation (Figure 3E(Fig. 3)). Altogether, circ_0010729 directly bound to miR-370-3p and negatively regulated its expression.

Based on the relation between circ_0010729 and miR-370-3p, we then investigated whether circ_0010729 regulated cardiomyocytes properties under hypoxia was through binding to miR-370-3p. First of all, AC-16 cells were transfected with anti-miR-370-3p or anti-NC, as expected, anti-miR-370-3p introduction caused significant reduction of miR-370-3p expression relative to anti-NC (Figure 4A(Fig. 4)). Next, we transfected anti-miR-370-3p into circ_0010729-decreased AC-16 cells, and found the introduction of anti-miR-370-3p attenuated circ_0010729 knockdown-induced miR-370-3p overexpression in hypoxia condition (Figure 4B(Fig. 4)). Then under hypoxia for 24 h, we found miR-370-3p inhibitor reversed the regulatory effects of si-circ_0010729 on AC-16 cell viability (Figure 4C(Fig. 4)), cell cycle (Figure 4D(Fig. 4)), caspase3 activity (Figure 4E(Fig. 4)), apoptosis (Figure 4F(Fig. 4)), migration (Figure 4G(Fig. 4)), IL-6 and TNF-α release (Figure 4H, I(Fig. 4)) as well as glucose metabolism (Figure 4J-L(Fig. 4)). Altogether, knockdown of circ_0010729 might protect cardiomyocytes against hypoxia-induced injury through the restoration of cardiomyocytes' properties and glycolysis via miR-370-3p.

The downstream target genes of miR-370-3p were then explored. Through searching online database Targetscan, TRAF6 was identified as a potential target of miR-370-3p (Figure 5A(Fig. 5)). Then the significant reduction of luciferase activity in AC-16 cells co-transfected with wild-type-3'UTR TRAF6 and miR-370-3p confirmed their direct interaction (Figure 5B(Fig. 5)). After that, the effect of miR-370-3p on TRAF6 expression was detected, and we found TRAF6 expression both at mRNA and protein levels was decreased by miR-370-3p up-regulation, but increased by miR-370-3p down-regulation in AC-16 cells (Figure 5C, D(Fig. 5)). Thus, we confirmed miR-370-3p targetedly suppressed TRAF6.

Given the direct interaction between miR-370-3p and TRAF6, we then studied the functions miR-370-3p/TRAF6 axis on cardiomyocytes. First, AC-16 cells were transfected with pcDNA or TRAF6, and TRAF6 expression was markedly elevated in cells after TRAF6 transfection compared to pcDNA (Figure 6A, B(Fig. 6)). Next, AC-16 cells were co-transfected with miR-NC, miR-370-3p, miR-370-3p + pcDNA, or miR-370-3p + TRAF6, and we found that introduction of TRAF6 markedly rescued miR-370-3p-induced decrease of TRAF6 level in AC-16 cells (Figure 6C, D(Fig. 6)). Thereafter, transfected cells were exposed to hypoxia condition for 24 h and rescue assay was then performed. Results showed miR-370-3p re-expression attenuated hypoxia-induced AC-16 cell viability suppression (Figure 6E(Fig. 6)), cell cycle arrest (Figure 6F(Fig. 6)), caspase3 activity enhancement (Figure 6G(Fig. 6)), apoptosis promotion (Figure 6H(Fig. 6)), migration inhibition (Figure I(Fig. 6)), IL-6 and TNF-α levels increase (Figure J, K(Fig. 6)), and glycolysis suppression (Figure 6L-N(Fig. 6)), while these conditions were reversed by following TRAF6 overexpression (Figure 6E-N(Fig. 6)). Overall, miR-370-3p might protect cardiomyocytes through restoration of cardiomyocyte properties and glucose metabolism via TRAF6 in hypoxia condition.

Whether specific crosstalk existed between circ_0010729 and TRAF6 through competition for miR-370-3p binding was further investigated. As shown in Figure 7A, B(Fig. 7), we found miR-370-3p inhibition rescued circ_0010729 decrease-induced TRAF6 down-regulation under hypoxia condition. Thus, we confirmed that circ_0010729 could indirectly regulate TRAF6 via miR-370-3p.

See also Supplementary data.