The Huh7 cell line was obtained from the Shanghai Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) supplemented with 10% fetal bovine serum (FBS) (Gibco; Thermo Fisher Scientific, Inc.) at 37°C in a humidified 5% CO₂ incubator.

Immunoprecipitation (IP) was carried out using the Protein G IP Kit (Roche, Basel, Switzerland) according to the manufacturer's instructions. SDS-PAGE was used for western blotting (WB) in order to separate proteins extracted from tissues or cells, and proteins were transferred onto polyvinylidene fluoride membranes. Membranes were incubated first with primary antibodies against DNM1L (dilution 1:1,000; cat. no. ab184247; Abcam), MID51 (dilution 1:100; cat. no. ab220672; Abcam), FIS1 (dilution 1:1,000; cat. no. ab71498; Abcam), mitochondrial fission factor (MFF) (dilution 1:1,000; cat. no. ab81127; Abcam), hypoxia inducible factor 1 subunit α (HIF1A) (dilution 1:1,000; cat. no. ab187524; Abcam, Cambridge, MA, USA), voltage-dependent anion channel (VDAC) (dilution 1:1,000; cat. no. 4661; Cell Signaling Technology, Danvers, MA, USA), adenine nucleotide translocator (ANT) (dilution 1:1,000; cat. no. 14671; Cell Signaling Technology), cyclophilin D (CypD) (dilution 1:1,000; cat. no. ab110324; Abcam), hexokinase 2 (HK2) (dilution 1:1,000; cat. no. 2867; Cell Signaling Technology), and GAPDH (dilution 1:10,000; cat. no. 60004-1-Ig; Proteintech, Chicago, IL, USA), followed by HRP-conjugated secondary antibody. Enhanced chemiluminescence was used for detection. WB results were quantified using ImageJ software (US National Institutes of Health, Bethesda, MD, USA; imagej.nih.gov/ij) following the WB quantification guidelines.

Total RNA was extracted from cells using TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. RNA was then reverse transcribed and the resulting cDNA was used as a template for quantitative PCR using a Takara RNA PCR kit and SYBR Premix Ex Taq (Takara Bio Inc., Kusatsu, Japan) in accordance with the manufacturer's instructions. GAPDH was used as an internal control. The following primers were used: Human HIF1A, sense, 5′ to 3′-GAACGTCGAAAAGAAAAGTCTCG and antisense, 5′ to 3′-CCTTATCAAGATGCGAACTCACA; human DNM1L, sense, 5′ to 3′-CTGCCTCAAATCGTCGTAGTG and antisense, 5′ to 3′-GAGGTCTCCGGGTGACAATTC; human MID51, sense, 5′ to 3′-CACGGCCATTGACTTTGTGC and antisense, 5′ to 3′-TCGTACATCCGCTTAACTGCC; human GAPDH, sense, 5′ to 3′-GAGTCAACGGATTTGGTCGT and antisense, 5′ to 3′-TTGATTTTGGAGGGATCTCG; human FIS1, sense, 5′ to 3′-GTCCAAGAGCACGCAGTTTG and antisense, 5′ to 3′-ATGCCTTTACGGATGTCATCATT; human MFF, sense, 5′ to 3′-ACTGAAGGCATTAGTCAGCGA and antisense, 5′ to 3′-TCCTGCTACAACAATCCTCTCC. The RT-qPCR protocol was: i) 95°C for 2 min; ii) 95°C for 5 sec, 60°C for 30 sec, 72°C for 30 sec, 40 cycles; iii) dissociation curve. All results were calculated using the 2^−ΔΔCq method (12).

The siRNA sequence (5′-aatatccatctctggcca-3′) for DNM1L was purchased from Sigma-Aldrich; Merck KGaA. A scrambled siRNA precursor was used as the negative control. siRNAs were transfected into cells using Lipofectamine^® 2000 (Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. After 48 h, the transfection efficiency was assessed using qPCR and WB, and the cells were used for further experiments.

The coding sequences of DNM1L were constructed into p3×Flag-CMV7 (Sigma-Aldrich; Merck KGaA). Transfection of siRNA and overexpression plasmids was carried out using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. In the overexpression assays, p3×Flag-CMV7 vectors were used as a control.

Cells were seeded onto coverslips coated with poly-L-lysine, and fixed with 4% paraformaldehyde before staining. Cells were permeabilized with 0.5% Triton X-100 in phosphate-buffered saline (PBS) at room temperature for 5 min, and blocked with 1% bovine serum albumin. Then, cells were incubated with primary antibodies (anti-DNML1 and/or anti-HK2), secondary antibodies, and then stained with DAPI and photographed with a TCS SP5 imaging system (Leica Microsystems, Wetzlar, Germany). For mitochondrial morphology, cells were incubated with 100 nM MitoTracker Deep Red (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C for 10 min.

Mitochondrial transmembrane potential was determined using JC-1 staining (Sigma-Aldrich; Merck KGaA) according to the manufacturer's instructions. Briefly, cells were incubated with JC-1 for 20 min at 37°C. JC-1 monomers were stained green and JC-1 aggregate was stained red. Cell images were obtained and analyzed by IMF.

2,7-Dichlorofluorescein diacetate (DCFH-DA; Sigma-Aldrich; Merck KGaA) was used to determine intracellular ROS levels according to the manufacturer's instructions. Briefly, cells were incubated with 20 µM DCFH-DA for 30 min at 37°C. After washing the cells with PBS three times, images were obtained and analyzed by IMF.

Briefly, treated cells were incubated with 2 µM calcein (Sigma-Aldrich; Merck KGaA) and 100 nM MitoTracker Deep Red for 30 min in the absence of light. Cells were subsequently exposed to hypoxia (1% O₂, 5% CO₂, 94% N₂) for 24 h. After washing the cells with PBS three times, images were obtained and analyzed using IMF.

Mitochondrial proteins were isolated from tumor cells using Mitochondria Isolation Kit (MP-007; Invent Biotechnologies, Plymouth, MN, USA) according to the manufacturer's instructions. Briefly, the cell lysate was centrifuged at 20,000 × g for 1 min and the precipitate was suspended briefly. The floccule was centrifuged for 1 min at 1,500 × g after overhanging and then the supernatant was mixed by vortexing with 400 µl buffer. After centrifugation at 20,000 × g for 10 min, the supernatant was removed completely and the pellet was suspended in 200 µl buffer. After a further 30 min of centrifugation at 20,000 × g, the supernatant was discarded and the pellet comprised isolated mitochondria.

Cell viability was quantified with a Cell Counting Kit-8 (CCK-8) (Dojindo, Tokyo, Japan), according to the manufacturer's instructions. The cells were plated at a density of 3,000 cells per well in 96-well plates after transfection with siRNA in a 10 cm dish for 48 h. CCK-8 assays were assessed at 12, 24 and 48 h after transfection by measuring the absorbance at 450 nm.

All data are presented with error bars denoting standard deviation. All statistical data are based on experiments conducted in triplicate. Statistical analysis was performed with GraphPad Prism 5 software (GraphPad Software, Inc., La Jolla, CA, USA). Differences between two groups were examined by Student's two-tailed t-test; two-way analysis of variance was used when comparing more than two groups. For comparisons among multiple groups, one-way analysis of variance (ANOVA) followed by Scheffe post hoc test was used. Correlations between two groups were analyzed using nonparametric Spearman's r-test. Two-sided P-values <0.05 were considered statistically significant.

Mitochondrial function is affected by hypoxic exposure in different types of cells. To determine the effect of hypoxia on HCC cells, we measured two indicators of mitochondrial function, mitochondrial membrane potential and ROS generation, in hypoxia-exposed Huh7 cells. JC-1 fluorescence, which was used to determine the mitochondrial membrane potential, was significantly decreased by hypoxia, whereas ROS generation was significantly elevated (Fig. 1A and B). Meanwhile, a significant increase in mitochondrial fragments was observed in hypoxia-exposed Huh7 cells (Fig. 1C), suggesting that mitochondrial dynamic equilibrium of fusion and fission in Huh7 cells is disrupted by hypoxia. Next we determined whether regulators of mitochondrial fission, such as DNM1L, MID51, FIS1 and MFF, are affected in Huh7 cells under hypoxia. HIF1A was used as positive control under hypoxia. No significant changes were observed for any of these regulators at the mRNA or protein levels (Fig. 1D and E). Since the effect of DNMlL depends on their translocation into mitochondria, we also determined their levels in mitochondrial and cytoplasmic fractions. Among these proteins, only DNM1L was decreased in the cytoplasm and increased in mitochondria under hypoxic conditions (Fig. 1F). In addition, confocal images verified that hypoxia increased the recruitment of DNM1L into the mitochondria (Fig. 1G). Taken together, these data indicate that hypoxia induces mitochondrial translocation of DNM1L in HCC cells.

Our confocal images revealed that hypoxia triggered mitochondrial fission and aggravated mitochondrial fragmentation (Fig. 1). To determine whether DNM1L is critical for this process, DNM1L was knocked down in Huh7 cells. Knockdown efficiency was determined by qPCR and WB analysis (Fig. 2A). DNM1L knockdown significantly elevated the mitochondrial membrane potential and significantly reduced ROS levels in both hypoxia-exposed and unexposed Huh7 cells, and significantly alleviated hypoxia-induced elevation of ROS generation and reduction of mitochondrial membrane potential (Fig. 2B and C). The alleviation was also observed with treatment of Huh7 cells with the DNM1L translocation inhibitor Mdivi (Fig. 2B and C). These data demonstrated the important role of DNM1L in hypoxia-induced excessive mitochondrial fission and the subsequent mitochondrial dysfunction. Since mitochondrial dynamics have been reported to regulate apoptosis and proliferation of HCC cells, we also determined the effects of DNM1L knockdown on the viability of Huh7 cells under hypoxia. DNM1L knockdown decreased the viability of Huh7 cells without hypoxic treatment, whereas hypoxia reversed the effects of DNM1L knockdown on viability, indicating that DNM1L alleviated the suppressive effects of hypoxia on tumor cell viability, and the oncogenic effects of DNM1L depends on hypoxic condition (Fig. 2D).

To note, compared to Mdivi treatment, DNM1L knockdown conferred stronger effects on mitochondrial function (Fig. 2B and C), suggesting that DNM1L regulates mitochondrial function in ways other than mitochondrial fission. Since DNM1L has been reported to regulate calcium signaling in HCC (6), and calcium signaling is closely associated with mPTP opening (13), we speculated that DNM1L may also be able to regulate mPTP opening in HCC cells. We analyzed the mPTP opening levels in calcein-CoCl₂ tests, and found that DNM1L knockdown significantly inhibited mPTP opening in hypoxia-exposed Huh7 cells (Fig. 3A). Further, we overexpressed DNM1L in Huh7 cells (Fig. 3B) and treated these cells with CsA (an mPTP opening antagonist). CsA treatment alleviated the effects of DNM1L on mitochondrial membrane potential, ROS generation and cell viability (Fig. 3C-E).

To further investigate the regulatory mechanism of DNM1L on mPTP opening, the expression and mitochondrial translocation of several mPTP structural proteins, including VDAC, ANT, CypD and HK2 were assessed. WB analysis showed that the total levels of these proteins were not altered after DNM1L knockdown in hypoxic Huh7 cells (Fig. 4A). Meanwhile, as control, no changes in mitochondrial translocation were observed for mitochondrial proteins CypD, VDAC, and ANT, whereas mitochondrial HK2 was increased after DNM1L knockdown (Fig. 4A), indicating that instead of regulating the expression of mPTP-related proteins, DNM1L regulates mitochondrial detachment of HK2 in hypoxia-exposed Huh7 cells. This effect of DNM1L on HK2 was demonstrated in our confocal images (Fig. 4B). To determine how DNM1L regulates HK2 translocation, we performed IP analysis and found a slight interaction between DNM1L and HK2 in Huh7 cells (Fig. 4C). Of note, hypoxia enhanced their binding affinity (Fig. 4C). Moreover, hypoxia increased the phosphorylation of HK2, which was attenuated by DNM1L knockdown (Fig. 4C). These data indicate that DNM1L interacts with HK2 and mediates the activation of HK2, which induces mitochondrial detachment, thus regulating the mPTP (Fig. 4D).