Human AC16 cells were obtained from ATCC (Manassas, VA, United States) and cultured in a DMEM medium (Hyclone, SH30243.01, Logan, UT, United States) supplemented with 10% fetal bovine serum (Gibco, 16000e044, Carlsbad, CA, United States) and 1% penicillin–streptomycin (Solarbio, P1400, Beijing, China) and incubated at 37°C with 5% CO₂. AC16 cells were treated with 2 µmol/L DOX (Sigma-Aldrich, 25316-40-9, Shanghai, China) and HucMSC-EVs in different concentrations for 24 h.

HucMSCs were provided by Stem Cell Bank, Chinese Academy of Sciences and cultured in serum-free MSC NutriStem® XF Medium (Biological Industries, Beit HaEmek, Israel). When the cells grew to 80–90% confluence, they were digested with 0.25% trypsin containing 0.01% EDTA. The cells were resuspended in PBS and adjusted to 1 × 10⁶ cells/ml. The mouse antihuman CD34, CD45, CD44, and CD105 were added and incubated at 4°C in dark for 15–30 min. Surface antigens of MSCs were characterized by using a Beckman CytoFLEX flow cytometer (Beckman Coulter Life Sciences, Tokyo, Japan) and the Human MSC Analysis Kit (BD Biosciences, San Jose, CA, United States).

HucMSCs were cultured in EV-free production medium for 24–48 h before conditioned medium was collected. EVs were isolated and purified by ultracentrifugation according to the protocol (Théry et al., 2006). The morphology of isolated HucMSC-EVs was observed by transmission electron microscopy (FEI Tecnai G2 Spirit Twin, Philips, NL). The TSG101 and CD81 protein markers of EVs were detected by Western blot.

PKH67 EV green fluorescent dye (UR52303, Umibio, China) was used to trace EVs being endocytosed by AC16 cells. In brief, EVs were stained with PKH67 dying working solution, in which the PKH liker was mixed with diluent C at a ratio of 1:9 in dark at room temperature. Then they were mixed well and incubated for 10 min in dark. The labeled EVs were incubated on AC16 cells for 24 h at 37°C, and cells were washed with 1 × PBS. Cells were mounted in a mounting medium containing DAPI (4′,6-diamidino-2-phenylindole dihydrochloride) (DAPI-Fluoromount-GTM, Yeasen Biotechnology, Shanghai, China). A laser-scanning confocal microscope was applied to take all of the images (Youn et al., 2019).

The coding region of NOX4 (NM_001143837.2) was cloned into pCDNA3.1 (+) plasmids (Clontech, Mountain View, CA, United States) at Hind III and EcoR1 sites. It was designated to oeNOX4. oeNOX4 was transfected to AC16 cells using Lipofectamine 2000 Kit (Invitrogen, Carlsbad, CA, United States). The negative control included cells with blank vector pCDNA3.1 (+) transfection. The primers used for amplification of the coding sequence of NOX4 were as follows:

NOX4-F:5′-CCCAAGCTTATGAATGTCCTGCTTTTCTGGAAAAC-3' (Hind III)

NOX4-R: 5′-CGG​AAT​TCT​CAG​CTG​AAA​GAC​TCT​TTA​TTG​TAT​TC-3' (EcoR I)

miR-100-5p mimic (5′-AAC​CCG​UAG​AUC​CGA​ACU​UGU​G-3′), miR-100-5p inhibitor (5′-CAC​AAG​UUC​GGA​UCU​ACG​GGU​U-3′), and negative control (NC, 5′-CAG​UAC​UUU​UGU​GUA​GUA​CAA-3′) were synthesized by Beyotime (Beijing, China) and transfected to cells with Lipofectamine 2000 Kit individually.

The NOX4 reporter gene plasmid is constructed by gene synthesis. The NOX4 (NM_001143837.2) sequence was found in NCBI. According to the NOX4 3′-UTR sequence, wild-type and mutant NOX4 3′-UTR sequences with Sac I and Xho I sticky ends were synthesized. The mutation site was based on the binding site of hsa-miR-100-5p and NOX4 3′-UTR sequence. The NOX4 3′-UTR binding site sequence containing mutation was inserted into the vector pGL3-promoter through Sac I and Xho I restriction sites to construct pGL3-Promoter-mutNOX4 3′-UTR. Wild-type NOX4 3′-UTR was inserted into the vector pGL3-Promoter through Sac I and Xho I restriction sites to construct pGL3-Promoter-wtNOX4-3′-UTR. PGL3-Promoter vector had firefly fluorescent gene (luc2) and pRL-TK with Renilla fluorescent gene (hRluc). A map of the predicted binding site of hsa-miR-100-5p to NOX4 and mutant was as following.

293T cells were then co-transfected with miR-100-5p inhibitor, miR-100-5p mimic and pGL3-NOX4-WT or miR-100-5p inhibitor, miR-100-5p mimic, and pGL3-NOX4-MUT plasmids. After transfection, the cells were treated with a Dual-Luciferase Reporter Gene Detection System Test Kit. The firefly luciferase activity and Renilla luciferase activity were detected by a microplate reader. Luciferase activity ratio in this study was the ratio of the firefly luciferase activity to Renilla luciferase activity.

A dichlorodihydrofluorescein diacetate (DCFH-DA) fluorescent probe (Sigma-Aldrich, D6883, Shanghai, China) combined with the flow cytometric analysis was used to detect the changes of reactive oxygen species (ROS) levels. The reactive oxygen species in the cell can oxidize nonfluorescent DCFH to produce fluorescent DCF. By detecting the fluorescence of DCF, the level of reactive oxygen species in the cell can be known. According to the production of red fluorescence in living cells, the amount and change of the cell ROS content can be judged. Briefly, AC16 cells were resuspended in 1x PBS, and the density was adjusted to 5 × 10⁵ cells/ml. AC16 cells were then incubated with 10 μM DCFH-DA for 20 min in dark at 37°C and subsequently subjected to the flow cytometric analysis (Shi et al., 2016).

AC16 cells were seeded in 6-well plates with 1 × 10⁵ per well and cultured for 12–24 h before use. AC16 cells were harvested 24 h after being treated with DOX combined with HucMSC-EVs or EV miR-100-5-p inhibitor or oeNOX4. Cells were prepared with the Annexin V-FITC Apoptosis Detection Kit (Beyotime, C1062s, Beijing, China) according to the manufacturer’s recommendations. Briefly, AC16 cells were centrifuged at 1,000 g for 5 min and resuspended to a concentration of 1 × 10⁶ cells/ml. 1 × 10⁶ resuspended cells were centrifuged at 1,000 g for 5 min. The supernatant was discarded, and 195 μL of Annexin V-FITC binding solution was added to gently resuspend the cells. And 10 μL of Annexin V-FITC was added and mixed gently. Then 5 μL of propidium iodide staining solution was added to the mix gently. The cell suspension was gently vortexed and incubated in dark at room temperature for 15 min, and then placed in an ice bath. At the same time, a tube without Annexin V-FITC and PI was used as a negative control. Flow cytometry was performed within 1 h. The following method was used: The Annexin V-negative/PI-negative part represented viable cells. The Annexin V-positive/PI-negative part represented early apoptotic cells, and the Annexin V-positive/PI-positive part represented late apoptotic and dead cells.

The levels of lactate dehydrogenase (LDH), superoxide dismutase (SOD), and malondialdehyde (MDA) in cells were measured, respectively, using the LDH (A020-2), SOD (A001-3), and MDA (A003-1) kits (Jiancheng Biotechnology Research Institute, Nanjing, Jiangsu, China) according to the manufacturer's recommendations. Assays were performed in triplicate, and the mean values of each sample were calculated manually.

AC16 cells were lysed by the addition of RIPA lysis buffer supplemented with a protease and phosphatase inhibitor cocktail (p8340 and p8250, Sigma, St Louis, MO, United States). 25 µg of total protein was separated by SDS-PAGE and transferred onto a nitrocellulose membrane. Membranes were further blocked with 5% skim milk and immersed into antibody solutions against TSG101 (1:2000, ab120511, Abcam, Cambridge, MA, United States), CD81 (1:2000, ab109201, Abcam, Cambridge, MA, United States), NOX2 (1:5000, ab129068, Abcam, Cambridge, MA, United States), cleaved-caspase-3 (1:500, ab13847, Abcam, Cambridge, MA, United States), β-actin (1:2000, ab8226, Abcam, Cambridge, MA, United States), and NOX4 (1:2000, 14347-1-AP, Proteintech, Rosemont, IL, United States). Then the membranes were immersed into the secondary antibody solution linked to horseradish peroxidase (A0208 and A0216, Beyotime, Shanghai, China). Signals were captured by a chemiluminescence system.

All data were expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was applied to assess the statistically significant differences between more than two groups. Statistical analysis was performed by Prism 8.0.2 software (GraphPad, San Diego, United States). p value < 0.05 was considered significant.

HucMSCs were cultured and collected. HucMSCs were identified by a flow cytometer to detect their surface markers. Flow cytometry exhibited that HucMSC surface markers, such as CD44 and CD105, were highly expressed, while hematopoietic stem marker CD34 and leukocyte surface antigen CD45 exhibited low expression (Figure 1A). Next, HucMSC-EVs were extracted and identified. Transmission electron microscope showed that HucMSC-EVs were small round or elliptical membranous bi-lipid membrane vesicles. Their diameters ranged in size from 30 to 100 nm. There were low electron densities in the vesicles (Figure 1B). The specific markers, namely, TSG101 and CD81, of HucMSC-EVs were detected by Western blot. While there were no expression for the specific markers TSG101 and CD81 of HucMSC-EVs in the HucMSC medium (Figure 1C). HucMSC-EVs were labeled by PKH67 and incubated with AC16 cells for 24 h. Under a laser scanning microscope, the green EVs were located in the cytoplasm (Figure 1D). These results suggest that HucMSC-EVs were successfully isolated.

AC16 cells were treated with 2 µmol/L DOX and HucMSC at concentrations of 0 μg/ml, 50 μg/ml, 100 μg/ml, and 200 μg/ml for 24 h. Flow cytometry showed that 2 µmol/L DOX obviously induced ROS levels to increase, while HucMSC-EV treatment decreased ROS levels that were increased by DOX. HucMSC-EVs significantly reduced ROS levels at the concentrations of 50 μg/ml, 100 μg/ml, and 200 μg/ml. The effect of HucMSC-EV treatment was concentration dependent (Figure 2A). Meanwhile, 2 µmol/L DOX increased LDH release and MDA levels, and decreased SOD levels. However, HucMSC-EV treatment inhibited the increases of LDH release and MDA levels, and the decreases of SOD levels which were induced by DOX. The functions of HucMSC-EVs on LDH release, SOD levels, and MDA levels were in a concentration-dependent manner (Figures 2C–E). Furthermore, 2 µmol/L DOX induced AC16 cell apoptosis. Flow cytometry displayed that HucMSC-EVs reduced apoptotic cells that were increased by DOX with the concentrations of 50, 100, and 200 μg/ml. The action of HucMSC-EVs on AC16 cell apoptosis was in a concentration-dependent manner (Figure 2B). Western blot exhibited that HucMSC-EVs markably repressed the cleaved-caspase-3 expression with the different concentrations of 50, 100, and 200 μg/ml, which were highly expressed by DOX (Figure 2F). These findings suggest that HucMSC-EVs obviously inhibit cardiomyocyte oxidative stress and apoptosis that are induced by DOX in AC16 cells.

AC16 cells were treated with 2 µmol/L DOX and HucMSC-EVs with the concentrations of 0, 50, 100, and 200 μg/ml for 24 h. qRT-PCR indicated that DOX extremely increased the NOX2 and NOX4 mRNA expression. HucMSC-EVs could repress the mRNA expression of NOX2 and NOX4 at concentrations of 50, 100, and 200 μg/ml, which were increased by DOX. Compared to the 2 μmol/L_DOX+50 μg/ml_HucMSC-EV group, HucMSC-EV treatment at 100 μg/ml apparently suppressed NOX2 and NOX4 mRNA expression, which was induced by DOX (Figure 3A). Western blot displayed that HucMSC-EVs suppressed protein expression of NOX2 and NOX4 with the concentrations of 50, 100, and 200 μg/ml, which was increased by DOX (Figure 3B). AC16 cells were treated with 2 µmol/L DOX and 100 μg/ml HucMSC-EVs for 0, 6, 12, 24, and 48 h. qRT-PCR showed that 2 µmol/L DOX remarkably induced NOX4 mRNA expression increase. HucMSC-EVs attenuated NOX4 mRNA expression with a concentration of 100 μg/ml at 6, 12, 24, and 48 h, which were increased by DOX. Compared to 12 h, HucMSC-EV treatment markedly suppressed NOX4 mRNA expression with 100 μg/ml at 24 h, which was increased by DOX (Figure 3C). Similarly, Western blot displayed that the NOX4 protein expression was ameliorated at a concentration of 100 μg/ml for 6, 12, 24, and 48 h, which was increased by DOX (Figure 3D). Collectively, the data suggest that HucMSC-EVs markedly inhibit NOX4 expression in a time- and concentration-dependent manner, which is induced by DOX.

HucMSCs were transfected with miR-100-5p inhibitor (Inhibitor) or miR-100-5p mimic (Mimic). Q-PCR displayed that miR-100-5p expression was abolished by miR-100-5p inhibitor, while the miR-100-5p expression was aggravated by miR-100-5p mimic (Figure 3A). After HucMSCs were transfected with miR-100-5p inhibitor or miR-100-5p mimic, HucMSC-EVs were extracted, and Q-PCR was performed. Q-PCR showed that miR-100-5p inhibitor abolished miR-100-5p expression in EVs compared to the negative control, whereas miR-100-5p mimic exceedingly exacerbated miR-100-5p expression in EVs (Figure 3B). AC16 cells were treated with 2 µmol/L DOX and 100 μg/ml miR-100-5p negative control in HucMSC-EVs (NC-EVs) or miR-100-5p inhibitor in HucMSC-EVs (Inhibitor-EVs) or miR-100-5p mimic in HucMSC-EVs (Mimic-EVs) for 24 h. Flow cytometry demonstrated that DOX remarkably increased ROS levels, which were decreased with supplement of NC-EVs or Mimic-EVs. Compared to NC-EVs, Inhibitor-EVs reversed those effects in which NC-EVs decreased ROS production which were increased by DOX, whereas Mimic-EVs had similar inhibitory effects to NC-EVs for ROS production (Figure 4C). Biochemical assay exhibited that 2 µmol/L DOX aggravated LDH release and MDA level increases as well as SOD level decreases, which were inhibited with supplement of NC-EVs or Mimic-EVs. In comparison to the NC-EV group, Inhibitor-EVs reversed those effects in which NC-EVs inhibited DOX induced the increases of LDH and MDA levels as well as the decreases of SOD levels, while Mimic-EVs had the similar effects to NC-EVs (Figures 4E–G). Furthermore, apoptosis was examined with flow cytometry and Western blot. NC-EVs and Mimic-EVs markedly reduced the percentages of apoptotic cells, which were induced to increase y DOX. Compared to the NC-EV group, Mimic-EVs had a similar effect to NC-EVs for reducing the percentages of apoptotic cells, while inhibitor-EVs reversed those effects in which NC-EVs inhibited DOX-induced apoptotic cell increases (Figure 4D). NOX4 and cleaved-caspase-3 protein expression were apparently attenuated by NC-EVs or Mimic-EVs, which was induced to increase by DOX. Compared to the NC-EV group, Inhibitor-EVs reversed those effects which NC-EVs decreased, NOX4 and cleaved-caspase-3 protein expression, which was increased by DOX, while Mimic-EVs had a similar effect to NC-EVs (Figure 4H). Taken together, these findings suggest that downregulation of EV miR-100-5-p reverses those effects in which HucMSC-EVs inhibit DOX-induced oxidative stress and apoptosis.

To study whether miR-100-5p targets the NOX4 protein, 293T cells were co-transfected with pGL3-NOX4-WT, miR-100-5p inhibitor (Inhibitor), and miR-100-5p mimic (Mimic) or pGL3-NOX4-MUT, miR-100-5p inhibitor (Inhibitor), and miR-100-5p mimic (Mimic). The miR-100-5p inhibitor markedly increased the luciferase activity of pGL3-NOX4-WT, while the miR-100-5p mimic obviously reduced the luciferase activity of pGL3-NOX4-WT. There was no change in the luciferase activity of pGL3-NOX4-MUT (Figure 5A). qRT-PCR exhibited that the miR-100-5p inhibitor increased the NOX4 mRNA expression, while the miR-100-5p mimic decreased the NOX4 mRNA expression (Figure 5B). Western blot illustrated that the NOX4 protein expression was increased by the miR-100-5p inhibitor, whereas the NOX4 protein expression was reduced by the miR-100-5p mimic (Figure 5C). These results demonstrate that NOX4 is the targeting protein of miR-100-5p. NOX4 expression is negatively regulated by miR-100-5p. Next, NOX4 overexpression plasmid (oe-NOX4) was constructed and transfected to AC16 cells. oe-NOX4 markedly increased NOX4 mRNA expression (Figure 5D) and NOX4 protein expression (Figure 5E). Furthermore, AC16 cells were transfected with oe-NOX4 for 24 h, and then treated with 2 µmol/L DOX combined with 100 μg/ml HucMSC-EVs for another 24 h. Flow cytometry revealed that HucMSC-EVs reduced ROS levels which were induced to increase by DOX, and this was reversed by the overexpression of NOX4 (Figure 5F). At the same time, HucMSC-EVs inhibited increased LDH and MDA levels as well as decreased SOD levels which were induced by DOX; these effects were reversed by the overexpression of NOX4 (Figures 5H–J). Moreover, DOX induced the percentages of apoptotic cell increase. HucMSC-EVs ameliorated the percentages of apoptotic cells which were increased by DOX, and this was reversed by NOX4 overexpression (Figure 5G). Western blot showed that HucMSC-EVs ameliorated the expression of NOX4 and cleaved-caspase-3 which was increased by DOX, and this was reversed by NOX4 overexpression (Figure 5K). Collectively, these results indicate that overexpression of NOX4 abolishes those effects in which HucMSC-EVs inhibit DOX-induced oxidative stress and apoptosis.