Human lung carcinoma cell line A549 (cat. # RCB0098, RRID: CVCL_0023) were provided by RIKEN BioResource Research Center through the National Bio-Resource Project of the Ministry of Education, Culture, Sports, Science and Technology (Japan). ACR20 cells, CDDP-resistant cells, were established and cultured as previously described (7). Both cell lines were maintained in Dulbecco's modified Eagle's medium (DMEM, Nacalai Tesque, Japan, cat. 08459-64) with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 µg/ml streptomycin in the presence or absence of 20 µM CDDP (Wako Pure Chemical Industries, Ltd., Osaka, Japan, cat. 033-20091) at 37°C with 5% CO₂ and 95% air. Human cervical cancer HeLa cells (cat. no. CCL-2, RRID: CVCL_0030) were obtained from American Type Culture Collection. ρ° HeLa cells, which are deficient in mtDNA and resistant to 6-thioguanine, were isolated as previously described (14).

Cells (1.2×10⁵ cells) were seeded on a 60-mm dish and incubated in culture media at 37°C in 5% CO₂ and 95% air. After a 24-h incubation, the cells were treated with 80 µM CDDP for 24 h. Annexin V/PI staining (Thermo Fisher Scientific, Inc., cat. no. V13245) was performed following the manufacturer's protocol. The stained cells were analyzed using FACSCalibur (BD Biosciences) and FlowJo software 10.5.3 (Tree Star, Inc.). Apoptotic cells were identified as Annexin V⁺/PI⁻ (early apoptosis) and Annexin V⁺/PI⁺ (late apoptosis) cells.

Cells were seeded at 7×10⁵ cells on a 60-mm dish and incubated in culture media at 37°C in 5% CO₂ and 95% air. In experiments examining cleaved caspase-3 levels, after a 24-h incubation, the cells were treated with 80 µM CDDP for 24 h. In experiments examining IAPs (namely, c-IAP1, c-IAP2, and XIAP) and phosphorylation of IκB-α and NF-кB, after a 1-h incubation, the cells were washed with PBS and treated with 40 mM N-acetyl-L-cysteine (NAC; Sigma-Aldrich; Merck KGaA, cat. no. A7250), an ROS inhibitor, for 24 h. Western blot assays were performed as previously described (15). The antibodies are listed in Table SI. After probing of anti-phospho-IκB-α and anti-phospho-NF-кB antibodies, the blots were stripped and reprobed with anti-IκB-α or anti-NF-кB antibodies. After probing of anti-XIAP, anti-c-IAP1 antibody, antic-IAP2, or anti-NDUFB8 antibodies, the blots were stripped and reprobed with anti-GAPDH or anti-β-actin antibodies.

To evaluate whether ROS affected the sensitivity of cells to CDDP, cells were seeded at 4.8×10⁴ cells/well on a 24-well plate and incubated in culture media at 37°C in 5% CO₂ and 95% air. After a 1-h incubation, the cells were treated with 40 mM NAC for 24 h. The treated cells were washed with PBS and cultured in culture media containing 80 µM CDDP for 48 h. The half maximal inhibitory concentration (IC₅₀) of CDDP was determined following the previous study protocol (7). Cells were fixed with 4% paraformaldehyde, stained with 0.5% crystal violet and lysed with 10% acetic acid. The optical density (O.D.) was measured at 600 nm. The percentage of cell viability and 50% growth inhibitory concentration was calculated as previously described (7).

Cells were incubated for 48 h, washed with HBSS (Nacalai Tesque, cat. no. 09735-75), stained with 10 µM CM-H₂DCFDA (Thermo Fisher Scientific, Inc., cat. no. C6827) or 5 µM MitoSox (Thermo Fisher Scientific, Inc., cat. no. M36008), and incubated for 30 min at 37°C in 5% CO₂ and 95% air. Cells were washed with HBSS and observed with a fluorescence microscope using FITC filters. The number of cells with positive fluorescence was counted and the average number of cells in three fields was taken as the number of ROS-producing cells per field.

Respiration-deficient ρ A549 cells were established by treatment of A549 cells with 50 ng/ml EtBr as previously described (16) for 14 days (Fig. S1A and B).

Cybrids were established as previously described (17). In brief, A549 cells or ACR20 cells were seeded at 1.5×10⁶ cells or 3.0×10⁶ cells, respectively, on a 100-mm dish for 48 h. To prepare enucleated A549 cells or ACR20 cells (the mtDNA donor), cells were pretreated with 10 µg/ml cytochalasin B (Sigma-Aldrich; Merck KGaA, cat. no. 14930-96-2) for 15 min and centrifuged at 17,300 × g for 20 min. The resultant cytoplasts were fused with ρ° HeLa cells, using polyethylene glycol. After 2 days, selective isolation of the cybrids was performed by culture in selection medium with 6-thioguanine, to exclude unfused A549 cells or ACR20 cells, and without uridine and pyruvate, to exclude parental ρ° HeLa cells. The resultant cybrids contained nuclear DNA from ρ° HeLa cells and mtDNA from enucleated A549 cells or ACR20 cells.

mtDNA extraction (Wako Pure Chemical Industries, Ltd., cat. no. 291-55301), polymerase chain reaction (PCR) (Takara Shuzo, Otsu, Japan, cat. no. RR001C), and purification of the PCR products (Takara Shuzo, cat. no. 740609) were performed following the manufacturers' protocols. Primer sequences are shown in Table SII. DNA sequencing was conducted by Eurofins Genomics (Tokyo, Japan).

The percentage levels of mtDNA mutations in A549 cells and ACR20 cells were examined by PCR-RFLP. In addition, to confirm that the mtDNA derived from A549 and ACR20 cells was incorporated into A549^cyb and ACR20^cyb cells, respectively, we performed PCR-RFLP. In brief, mtDNA was extracted from the A549 and ACR20 cells. Total DNA was extracted from A549^cyb and ACR20^cyb cells. PCR amplification was performed according to the manufacturer's protocol. Purified PCR products were digested with the restriction enzyme. The restriction fragments were analyzed by gel electrophoresis. PCR primer sequences, restriction enzyme, and annealing temperature are shown in Table SIII.

A XF24 Extracellular Flux Analyzer (Seahorse Biosciences, USA) was used to determine the intracellular bioenergetic profiles. A549 or ACR20 cells were seeded at 4×10⁴ cells/well on the XF24 V7 Cell Culture microplate. Extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) were measured as previously described (18). Mitochondria from cultured cells were isolated using the Mitochondria Isolation Kit (BioVision, Inc., cat. no. K288-50). Protein concentrations of isolated mitochondrial samples were estimated using Bradford assays. Isolated mitochondria were stored at −80°C until used. Mitochondrial complex activities were measured using MitoCheck Complex I Activity Assay Kit (Cayman Chemical Company, cat. no. 700930). Complex I activity was determined as activity without inhibitor of complex I subtracted from the activity with complex I inhibitor.

RT-qPCR was established as previously described (7). RT-qPCR was performed using PrimeScript™ RT Master Mix (Takara Shuzo, cat. no. RR036) and TB Green^® Premix Ex Taq™ II (Takara Shuzo, cat. no. RR820) according to the manufacturer's protocol. Primer sequences are shown in Table SIV. Relative expression was calculated using the ΔΔCt method (19) with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as the reference gene.

The structures of the mutant ND proteins were analyzed at Altif Laboratories Inc. (Tokyo, Japan).

Data are expressed as the mean ± standard error (SEM) of independent determinations for each experiment. Normality was assessed by Kolmogorov-Smirnov normality test. Unpaired t-test was used to evaluate the differences between two groups. Comparisons of more than two groups were evaluated by one-way analysis of variance (ANOVA), followed by the Student-Newman-Keuls post hoc test. P-value <0.05 is indicative of a statistically significant difference.

We previously established CDDP-resistant cells derived from A549 cells, which we named ACR20 cells (7). To determine whether ACR20 cells are resistant to CDDP-induced apoptosis, we evaluated CDDP-induced apoptosis using Annexin V/PI staining assays. In the CDDP-treated A549 cells, a high percentage of cells were positively stained for Annexin V and cleaved caspase-3 expression was detected by western blotting, thus indicating that CDDP induced apoptosis in the A549 cells (Fig. 1A-C). In contrast, the percentage of Annexin V-stained cells was low (11%) and cleaved caspase-3 was not detected in the CDDP-treated ACR20 cells. These results indicate that CDDP-induced apoptosis was impaired in the ACR20 cells.

To determine the mechanism underlying the impairment of CDDP-induced apoptosis in ACR20 cells, we analyzed the production of ROS using the ROS indicator CM-H₂DCFDA. Green fluorescence was not detected in A549 cells not treated with CDDP but was observed after CDDP treatment (Figs. S2 and 2A). This result indicates that CDDP induced ROS generation in the cytoplasm of the A549 cells. Unexpectedly, green fluorescence was detected in ACR20 cells not treated with CDDP, and the number of green fluorescence-positive cells was increased compared to that of the A549 cells (Fig. 2A). These results indicate that IRLs were elevated in ACR20 cells. NAC, an inhibitor of ROS, pretreatment improved the sensitivity of ACR20 cells to CDDP (Fig. 2B). Together, these findings suggest that the elevation of IRLs was involved in the acquisition of CDDP resistance.

NF-κB regulates the gene expression of IAPs (20). ROS-stimulated IκB kinase leads to phosphorylation of IκB and NF-κB, which promotes IκB degradation and modulates NF-κB activation (21). Thus, we hypothesized that the elevation of IRLs in ACR20 cells would induce the expression of IAPs and impair CDDP-induced apoptosis. The expression levels of phosphorylated IκB-α and NF-κB in ACR20 cells were significantly higher than those in A549 cells (P<0.01), and these levels were suppressed in ACR20 cells by NAC pretreatment (Fig. 2C and D). Similarly, the levels of IAPs in ACR20 cells were significantly elevated compared to those in A549 cells (P<0.01), and the expression of IAPs in ACR20 cells was suppressed by NAC pretreatment (Fig. 2E-G). The expression of phosphorylated IκB-α, phosphorylated NF-κB, and IAPs in A549 cells was not affected by NAC pretreatment (data not shown). There results suggest that elevated IRLs decreased the sensitivity of ACR20 cells to CDDP by activating NF-κB signaling and inducing IAP expression.

Intracellular levels of ROS are dependent on the balance between ROS generation and ROS elimination. Increased IRLs in cancer cells may be caused by malfunction of the mitochondrial respiratory chain (22,23). Extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) values are important indicators of mitochondrial respiration and glycolysis. We found that ECAR was not changed and that OCR was decreased in the ACR20 cells when compared to the A549 cells (Fig. 3A). Mitochondrial superoxide is generated as a byproduct of ATP production (24,25). In A549 cells loaded with MitoSox, an indicator of mitochondrial superoxide, red fluorescence was observed; however, in contrast, red fluorescence was not detected in the ACR20 cells (Fig. 3B). Together, these results indicate a loss of mitochondrial function in ACR20 cells. To investigate whether mitochondrial dysfunction causes CDDP resistance in A549 cells, we analyzed CDDP sensitivity in mtDNA-depleted ρ A549 cells, which showed loss of mitochondrial function. The IC₅₀ values of CDDP in A549 and ρ A549 cells were 4.64±1.04 and 11.56±2.26 µM, respectively (Fig. S1C). Together, these results indicate that mitochondrial dysfunction causes CDDP-resistance in the A549 cells.

The mitochondrial genome encodes mitochondrial respiratory chain components. Thus, decreased quality and quantity of mtDNA lead to mitochondrial dysfunction (26–28). To examine whether a change in the number of mtDNA copies in ACR20 cells is involved in the loss of mitochondrial function, we evaluated the expression of mitochondria-encoded COX I (MT-COI) levels normalized to α-tubulin levels to determine the number of mtDNA copies (29,30). Contrary to our expectations, MT-COI expression was increased in the ACR20 cells when compared to the expression level in the A549 cells (Fig. 4A). The number of mtDNA copies is regulated by transcription factor A, mitochondrial (TFAM) and RNA polymerase mitochondrial (POLRMT) (31,32). We found that the expression of TFAM and POLRMT mRNAs was increased in the ACR20 cells when compared to the A549 cells (Fig. 4B and C). These results indicate that the number of mtDNA copies was increased in ACR20 cells, and this increase might cause the upregulation of TFAM and POLRMT mRNA expression.

To determine the quality of mtDNA, mtDNA was extracted from the A549 and ACR20 cells for sequencing and identification of mtDNA mutations. Moreover, the percentage levels of mtDNA mutations were analyzed by PCR-RFLP. 4587 T>C and 11384 C>A were included in 100% of mtDNAs in the ACR20 cells (Fig. S3A and B and Table I). 13148 C>A and 14162 G>T were specifically detected in ACR20 cells and were included in 44.9 and 53.3% of the mtDNAs, respectively (Fig. S3C and D and Table I). Thus, four mtDNA mutations with varying percentage levels were identified in ACR20 cells and these mutations were located in genes encoding mitochondrial complex I. We further found that the activity of mitochondrial complex I was decreased in ACR20 cells compared to the A549 cells (Fig. 4D). NDUFB8 is a component of the supernumerary subunits in mammalian complex I, which are central to the structure, stability, and assembly (33). mRNA and proteins levels of NDUFB8 were comparable between the A549 and ACR20 cells (Fig. 4E-G). Thus, these results suggest that the identified mtDNA mutations caused the decrease of mitochondrial complex I activity in ACR20 cells.

To clarify whether the identified mtDNA mutations in ACR20 cells contributed to the elevation of IRLs and the decreased activity of complex I, cybrids (cells with the same physiological functions such as nuclear DNA, but different only in mtDNA) were established (17) (Fig. 5A). mtDNA derived from A549 and ACR20 cells was incorporated into A549^cyb and ACR20^cyb cells, respectively (Fig. S4). The number of ROS-producing cells was significantly increased and the activity of complex I was decreased in the ACR20^cyb cells compared to the A549^cyb cells (P<0.01) (Fig. 5B and C). The expression of IAPs was significantly increased in the ACR20^cyb cells compared with the A549^cyb cells (XIAP and c-IAP1, P<0.05 and P<0.01; c-IAP2; P<0.05) (Fig. 5D-F). We further examined whether the mtDNA mutations impacted the sensitivity to CDDP. The IC₅₀ value of CDDP in the A549^cyb and ACR20^cyb cells were 0.27±0.02 and 0.69±0.02 nM, respectively (Fig. 5G). Together, these results indicate that the four identified mtDNA mutations with varying percentage levels in ACR20 cells caused CDDP resistance through the increased expression of IAPs via elevation of IRLs and inactivation of complex I.

We next determined which of the four identified mtDNA mutations may be associated with complex I inactivation using the three-dimensional structure data of human mitochondrial complex I. The four identified mutations (ND2 F40L, ND4 L209I, ND5 P271Q, and ND6 S21Y) in the mitochondrial complex I were marked on the structure of human respiratory complex I (PDB ID: 5XTB) (Fig. 6A). The mutations were located in the transmembrane. The electric charge and polar amino acid, which exist in proton channels affect proton translocation (34,35). Therefore, the protein structure of the four mutations was modeled and the effect of the mutations on polar amino acids in proton translocation was investigated. Four mutations and key residues were marked to three-dimensional protein structures of complex I and the pathway of proton translocation was marked with reference to Fiedorczuk et al (34) and Fiedorczuk and Sazanov (35) (Fig. 6B). ND2 F40L is located near the charged side chains (Glu34, Lys105, and Lys135) and hydrophilic side chains (Gln134) (Fig. 6B and C). The side chain of Phe residue contacts TMH5 and is packed by hydrophobic interactions with the surrounding hydrophobic residues (Fig. 6C). Although Leu is hydrophobic, the side chain of Leu is smaller than that of Phe. Thus, we speculated that the F40L mutation eliminated these hydrophobic interactions and the packing between TMH2 and TMH5 (Fig. 6C). Furthermore, the conformation of Glu134 may be changed by displacement from Phe40 to Ile40, which may cause the structure change of the proton translocation pathway. In contrast, the other mutations were present at different positions of the proton translocation pathway (Fig. 6B). These results show the possibility that ND2 F40L affects the proton pathway, leading to the decrease of complex I activity.