Lipid peroxidation increases hydrogen peroxide permeability leading to cell death in cancer cell lines that lack mtDNA

Abstract

4‐Hydroxynonenal (HNE) is an important product of plasma membrane lipid peroxidation, which is a cause of cell and tissue injury. Mitochondrial DNA (mtDNA)‐depleted ρ⁰ cells were established using human cervical cancer and oral squamous cell carcinoma cell lines. We investigated the effect of reactive oxygen species in ρ⁰ cells, especially the mechanism of hydrogen peroxide (H₂O₂)‐mediated cell death. These cell were subjected to high oxidative stress and, compared with their parental cells, showed greater sensitivity to H₂O₂ and high lipid peroxidation. Upregulation of HNE in the plasma membrane was observed prior to the increase in intracellular H₂O₂. The amount of oxidized lipid present changed H₂O₂ permeability and administration of oxidized lipid led to further cell death after treatment with H₂O₂. Expression levels of lipoxygenase ALOX genes (ie ALOX5, ALOX12, and ALOX15) were upregulated in ρ⁰ cells, as were expression levels of ALOX12 and ALOX15 proteins. ALOX5 protein was mainly distributed in the nucleus, while ALOX12 and ALOX15 proteins were distributed in the nucleus and the cytoplasm. Although expression of COX2 gene was upregulated, its protein expression did not increase. ALOX (especially ALOX15) may be involved in the sensitivity of cancer cells to treatment. These data offer promise for the development of novel anticancer agents by altering the oxidation state of the plasma membrane. Our results showed that lipid peroxidation status is important for H₂O₂ sensitivity and that ALOX15 is involved in lipid peroxidation status.

Abbreviations

•OH

hydroxyl radical

ALOX

lipoxygenase

COX

cyclooxygenase

DOPC

1,2‐dioleoyl‐sn‐glycero‐3‐phosphocholine

H₂O₂

hydrogen peroxide

HNE

4‐Hydroxynonenal

HPF

Hydroxyphenyl fluorescein

MIC

microscope observation

NDGA

nordihydroguaiaretic acid

PINK1

PTEN‐induced kinase 1

POVPC

1‐palmitoyl‐2‐(5′‐oxo‐valeroyl)‐sn‐glycero‐3‐phosphocholine

ROS

Reactive oxygen species

RT

radiation therapy

1. INTRODUCTION

ROS are defined as chemically reactive species containing oxygen, some of which are peroxidases or hydroxyl radicals. ROS production is effective in the chemotherapy and RT of cancer.1, 2, 3 Numerous chemicals can produce ROS than affect cancer cells4, 5, 6, 7, 8 and some of these are under investigation in clinical trials.9 Some ROS, for example H₂O₂, are used as sensitizers of cancer cells by producing reactive oxygen during RT.10, 11 Similarly, ROS are produced by nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in the cytoplasm.12 However, the main source of ROS may be the mitochondrial electron transport chain through oxidative phosphorylation.13, 14 Mitochondrial DNA (mtDNA) encodes 13 proteins that are components of the mitochondrial electron transport chain. Damage to mtDNA results in increased production of ROS, conversely causing neurodegenerative diseases and various types of cancer.15, 16 Therefore, these mtDNA‐damaged cells may be valuable models for studying ROS‐related disease. Mitochondrial DNA devoid ρ⁰ cells were established from a human cervical cancer cell line (HeLa) and from an oral squamous cell carcinoma (SAS). These ρ⁰ cells were sensitive to ROS, in particular H₂O₂.17 Notably, yeast ρ⁰ cells have also shown sensitivity to treatment with H₂O₂.18 Recently, ρ⁰ cells from osteosarcoma and lung carcinoma have shown sensitivity to X‐ray irradiation related to oxidative stress.19 Conversely, there are several studies reporting that ρ⁰ cells are resistant to oxidative stress, especially to radiation. ρ⁰ cells from human fibroblasts have shown resistance to γ‐irradiation by decreasing apoptosis.20 In addition, ρ⁰ cells from the human pancreatic tumor MiaPaCa‐2 cell line showed resistance to X‐ray irradiation via activation of cyclin B1.21 Moreover, ρ⁰ cells from healthy human bronchial epithelial cells have shown resistance to oxidative stress. At present, the sensitivity or resistance of ρ⁰ cells to oxidative stress and the stress response mechanism involved in these processes have not been fully elucidated.

ROS react with polyunsaturated fatty acids in the lipid membranes, inducing lipid peroxidation that leads to cell death. HNE, the end product of lipid peroxidation reacts with low‐molecular‐weight compounds (eg, glutathione, proteins, and DNA), and is considered a second messenger of oxidative stress.22, 23 Notably, HNE production occurs through the ROS generation and an enzymatic process.24 The key enzyme in the production of HNE is ALOX15. The relationship between the level of HNE and ALOX15 has been previously reported.25 It has also been reported HNE induces expression of the COX2 gene.26 However, so far, the relationship between ρ⁰ cells and ALOXs and/or COX2 has not been investigated.

Moreover, the mechanism of sensitivity to H₂O₂ in ρ⁰ cells via oxidation of the plasma membrane has not been elucidated. In this study, we investigated the oxidation state of the plasma membrane in ρ⁰ cells, and identified factors that control this process.

2. MATERIALS AND METHODS

2.1. Cell culture

Human cancer cell lines HeLa and SAS were obtained from the Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University (Sendai, Japan). ρ⁰ cells were established through culture in RPMI1640 (Wako Pure Chemical Industries Ltd., Osaka, Japan) containing 5% FBS (Gibco Invitrogen Corp., Carlsbad, CA, USA), 50 ng/mL ethidium bromide (Nacalai Tesque Inc., Kyoto, Japan), 50 μg/mL uridine (Sigma‐Aldrich, St Louis, MO, USA), and 110 μg/mL sodium pyruvate (Sigma‐Aldrich) for 3‐4 wk.17 Cells were maintained in RPMI 1640 supplemented with 10% FBS, 50 μg/mL uridine, and 110 μg/mL sodium pyruvate in a humidified atmosphere at 37°C with 5% carbon dioxide. Exponentially growing cells were used in all experiments.

2.2. Relative levels of internal H₂O₂

Internal H₂O₂ was visualized using HYDROP™ (Goryo Chemical Inc., Hokkaido, Japan) as previously described.17 Briefly, cells in glass‐bottomed dishes (Matsunami Glass Ind., Ltd., Osaka, Japan) were cultured in RPMI 1640 with or without 50 μmol/L H₂O₂ for 10 min, 30 min, 1 h, and 2 h. The cultured cells were washed with noncontaining H₂O₂ RPMI 1640 twice to remove the H₂O₂ from the medium, and subjected to treatment with 2.5 μmol/L HYDROP™ in RPMI 1640 at 37ºC for 20 min. Subsequently, the cells were washed with RPMI 1640 twice, and fluorescence images were obtained using a BZ‐8000 fluorescence microscope (Keyence Corporation, Osaka, Japan). The ImageJ software (Rasband, W.S., ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997‐2012) was used to measure fluorescence intensity.

2.3. Intracellular intake of H₂O₂ using a stable isotope

A H₂O₂ ¹⁸O₂ solution (Sigma‐Aldrich) was used to measure the intracellular intake of H₂O₂. Cells were subjected to treatment with 50 μmol/L H₂O₂ ¹⁸O₂ solution for 1 h. After treatment, the cells were washed with phosphate‐buffered saline (PBS) three times. Following the wash, the cells were collected and dried using an FDU‐2200 freeze dryer (Tokyo Rikakikai Co. Ltd., Tokyo, Japan) for 3 h. Dried samples were sent for analysis to Taiyo Nippon Sanso Corporation (Tokyo, Japan). The levels of ¹⁸O₂ in treated cells were measured using stable isotope‐ratio mass spectrometry. The results are shown as delta ‰ compared with the standard mean ocean water (^18Oδ SMOW ‰).

2.4. Immunofluorescence

Cells were cultured in glass‐bottomed dishes using RPMI 1640 with or without 50 μmol/L H₂O₂ for 1 and 2 h. Subsequently, the cells were fixed with 4% formaldehyde in PBS for 30 min, and rinsed three times with PBS. Plasma membranes were permeabilized by incubation in 95% ethanol with 5% acetic acid for 10 min. After washing five times with PBS, cells were incubated for 30 min in blocking solution (5% skimmed milk in PBS‐T; PBS with 0.05% Tween 20). The primary antibodies (incubated at 4°C overnight) were mouse anti‐HNE (1:200 dilution; Japan institute for the control of aging, Shizuoka, Japan), rabbit anti‐PINK1, anti‐aquaporin11, anti‐ALOX5, anti‐ALOX12, and anti‐ALOX15 antibodies (Novus Biologicals, Centennial, CO, USA: BC100‐494; ABGENT San Diego, CA, USA: AP58056; Abcam, Cambridge, UK: ab169755, ab211506, and ab80221; 1:1000 dilution), mouse anti‐COX1 (Abcam: ab695; 1:500 dilution), and rabbit anti‐COX2 (Abcam: ab15191; 1:100 dilution). The secondary antibodies (1:200 dilution, incubated at room temperature for 1 h) were the Alexa Fluor 488 goat anti‐mouse IgG and Alexa Fluor 568 goat anti‐rabbit IgG (Thermo Fisher Scientific, Waltham, MA, USA: A11001, and A11011). For staining the nuclei, cells were incubated with DAPI (0.5 μg/mL) at room temperature for 10 min. A BZ‐8000 fluorescence microscope was used to obtain fluorescence images, and the ImageJ software was used to measure fluorescence intensity.

2.5. Liposome assay

Liposomes containing luminol and horseradish peroxidase (HRP) were prepared as follows: DOPC (Avanti Polar Lipids Inc., Albaster, AL, USA) and POVPC (Avanti Polar Lipids Inc.) were dissolved in chloroform (1 mg/mL) and mixed together. The dissolved lipids were dried to remove the chloroform. Subsequently, the mixture was added to 50 mmol/L Tris‐HCl (pH 8.6) buffer containing 10 mmol/L luminol (Nacalai Tesque Inc.: 20751‐34) and 100 μmol/L HRP (Wako Pure Chemical Industries Ltd.: 169‐10791) and was then sonicated The diameter of the prepared liposomes was calculated via the Dynamic Light Scattering method using ELS‐Z2M (Otsuka Electronics Co. Ltd., Osaka, Japan). Subsequently, 8.8 mmol/L H₂O₂ were administered to the liposomes, and luminescence was observed using a luminometer (JASCO, Tokyo, Japan: FP‐6300). Passage of H₂O₂ through the liposome membrane resulted in luminol reaction, producing blue luminescence.

2.6. Administration of oxidized lipid

POVPC (12.5‐50 μmol/L) was administered to the cultured medium using the ethanol injection method.27, 28 This lipid has been shown to localize to the plasma membrane.27, 28 After 10 min (SAS) or 30 min (HeLa) of treatment, POVPC was removed, and the cells were treated with 25 μmol/L H₂O₂ for 48 h. After administration of POVPC and H₂O₂, cell survival was analyzed using the CCK‐8 assay kit (Dojindo Molecular Technologies, Inc., Kumamoto, Japan), as previously described.17

2.7. Quantitative PCR

Total RNA was extracted using ISOGEN (Nippon Gene, Toyama, Japan), and reverse transcription was performed as previously described.17 cDNA equivalent to 1 ng of total RNA was used for the Quantitative PCR (qPCR). The reactions were performed with Step One Plus (Applied Biosystems; Foster City, CA, USA) using THUNDERBIRD^® qPCR Mix (TOYOBO, Osaka, Japan). β‐actin was used as the loading control. For the amplification of genes, one cycle of denaturation (95°C for 10 min) was performed, followed by 40 cycles of amplification (95°C for 10 s and 60°C for 60 s). Each experiment was performed in triplicate. Table 1 lists the primer sequences used in this experiment.

Table 1.

Primer sequences

Primer name	Primer sequence
ALOX5 F	5′‐CTGGGCATGTACCCAGAAGAGCATTTTAT‐3′
ALOX5 R	5′‐ACAAGTAGTAATATGGCAGCTGCTTCTTCT‐3′
ALOX12 F	5′‐TTCAAATGGCCATCTCATGGCATCTGAGT‐3′
ALOX12 R	5′‐ATCTGTTCGGAATTGGTTTAGCACAGCTTT‐3′
ALOX15 F	5′‐ATCTATCGGTATGTGGAAGGAATCGTGAGT‐3′
ALOX15 R	5′‐TAAAGAGACAGGAAACCCTCGGTCCT‐3′
COX1 F	5′‐AATCCCATCTGTTCTCCGGAGTACTG‐3′
COX1 R	5′‐GAAACGTAGGGACAGGTCTTGGTGTT‐3′
COX2 F	5′‐TGGAGCACCATTCTCCTTGAAAGGACTTAT‐3′
COX2 R	5′‐GACTGTTTTAATGAGCTCTGGATCTGGAAC‐3′
β‐Actin F	5′‐AGAGCTACGAGCTGCCTGAC‐3′
β‐Actin R	5′‐AGCACTGTGTTGGCGTACAG‐3′

2.8. Western blotting

Each cell lysate (30 μg per lane) was subjected to SDS‐PAGE under reduced conditions using a 10% or 15% polyacrylamide gel. The proteins were subsequently blotted on a PVDF membrane. After blocking with blocking buffer (3% skimmed milk in TBS‐T; TBS with 0.05% Tween 20, for ALOX12 antibody and 5% skimmed milk in PBS‐T for the other primary antibodies), the blotted membranes were incubated with primary antibodies (rabbit anti‐ALOX5, anti‐ALOX12, anti‐ALOX15, mouse anti‐COX1, rabbit anti‐COX2; Abcam: ab169755, ab211506, ab80221, ab695, and ab15191, rabbit anti‐Akt, pAkt^T308, pAkt^S473, FOXO1, pGSK3β^S9: Cell Signaling Technology, Danvers, MA, USA: #4691, #13038, #4060, #2880, #5558, mouse anti‐BCL2, Bax; Santa Cruz Biotechnology, San Diego, CA, USA: SC‐7382, SC‐7480) in blocking buffer at 4ºC overnight. After five washes with TBS‐T (ALOX12 antibody) or PBS‐T (other antibodies), the membranes were incubated with peroxidase‐conjugated anti‐rabbit IgG antibody or anti‐mouse IgG antibodies (GE Healthcare UK Ltd., Amersham Place, England) at room temperature for 2 h. Immunoreactive proteins were visualized with ECL prime (GE Healthcare) using ChemiDoc XRS Plus (BIO‐RAD Laboratories, Inc., Hercules, CA, USA). An anti‐β‐actin antibody (Novus Biologicals LLC, Centennial, CO, USA; NB100‐56874) was used as loading control. The dilution factor for all antibodies was 1:1000. Table 2 represents the mean and standard error of the mean (SEM) of three independent ρ⁰/parent intensity ratios after normalization against β‐actin.

Table 2.

Relative levels of protein expression in HeLa and SAS ρ⁰/parent cells (n = 3)

	ρ0/parent	SEM
ALOX5
HeLa	0.61	.16
SAS	0.61	.01
ALOX12
HeLa	1.51	.14
SAS	1.24	.05
ALOX15
HeLa	1.42	.04
SAS	1.24	.05
COX1
HeLa	1.14	.01
SAS	1.08	.05
COX2
HeLa	0.53	.14
SAS	0.36	.05
Akt
HeLa	0.87	.07
SAS	1.00	.01
pAktT308
HeLa	0.64	.13
SAS	0.49	.20
pAktS473
HeLa	0.52	.17
SAS	0.63	.14
FOXO1
HeLa	1.44	.17
SAS	1.28	.07
GSK3βS9
HeLa	0.71	.11
SAS	0.72	.16
BCL2
HeLa	0.78	.00
SAS	0.67	.11
Bax
HeLa	1.88	.04
SAS	1.31	.01

2.9. ALOX inhibitor assay

Cells were treated with 10 μmol/L caffeic acid (Nacalai Tesque Inc.), the ALOX5 inhibitor, or 10 μmol/L NDGA (Sigma‐Aldrich), the nonspecific ALOXs inhibitor, at 37ºC for 10 min. After removal of the inhibitor‐containing medium, the cells were treated with 50 μmol/L H₂O₂ for 1 h. HYDROP™ and HNE antibody were used as described above to detect internal H₂O₂ or HNE.

2.10. Detection of OH using HPF

Internal levels of ･OH were detected using HPF (Goryo Chemical Inc.) according to the protocol provided by the manufacturer with slight modification. HPF (10 μmol/L) with MIC buffer (130 mmol/L sodium chloride, 5.3 mmol/L potassium chloride, 0.8 mmol/L magnesium sulfate, 1 mmol/L disodium phosphate, 2 mmol/L glucose, 20 mmol/L HEPES, 1 mmol/L sodium pyruvate, 2.5 mmol/L sodium bicarbonate, 1 mmol/L ascorbic acid, 1.5 mmol/L calcium chloride, and 1.5 mg/mL bovine serum albumin (BSA)) was added to the cells in glass‐bottomed dishes. After incubation at 37°C for 15 min, the HPF with MIC buffer was replaced with new MIC buffer. Images were obtained using a BZ‐8000 fluorescence microscope. Nuclei were counterstained with DAPI as described above.

2.11. Statistical analysis

One‐way analysis of variance with Scheffe's F test was performed for the water‐soluble tetrazolium assay and ALOX inhibitor experiment. All other statistical analyses were performed using Student's t test. A P < .05 denoted statistical significance. The results were expressed as means ± SEM.

3. RESULTS

3.1. Relative levels of internal H₂O₂ and intake of H₂O₂

After treatment for 1 h, the internal levels of H₂O₂ were increased only in ρ⁰ vs parental cells (Figure 1A). The results obtained from the treatment with ¹⁸O₂‐labeled H₂O₂ showed that the intake of H₂O₂ in ρ⁰ cells was significantly higher than that observed in parental cells after treatment with H₂O₂ for 1 h (Figure 1B).

Figure 1.

Internal levels of H₂O₂ and intracellular intake of H₂O₂ in ρ⁰ cells. A, Temporal change of H₂O₂ after treatment with H₂O₂ in ρ⁰ and parental cells using HYDROP™. In ρ⁰ cells, the relative H₂O₂ intensity was increased after treatment with H₂O₂ for 1 h. B, The intracellular intake of H₂O₂ was detected using a stable isotope_. Cells were subjected to 50 μmol/L of H₂O₂ ¹⁸O₂ solution, and the content of ¹⁸O₂ was measured using stable isotope‐ratio mass spectrometry. In ρ⁰ cells, the intake of H₂O₂ was significantly higher than that observed in parental cells after treatment. The results are expressed as the mean ± SEM. **P < .01 using Student's t test

3.2. Timing of increase in lipid peroxidation and internal levels of H₂O₂

Significantly higher intensity was observed in ρ⁰ cells (Figure 2A, B). Co‐staining with PINK1 (localized in mitochondria) and aquaporin 11 (localized in the plasma membrane) revealed that HNE was mainly localized in the plasma membrane (Figure 2C, D). Moreover, temporal change of lipid peroxidation was observed to investigate the timing of this process after treatment with H₂O₂ (Figure 2E). Lipid peroxidation was increased 1 h after treatment (Figure 2E, F), whereas the internal levels of H₂O₂ were increased in parental cells 2 h after treatment (Figure 2F).

Figure 2.

Timing of increase in lipid peroxidation and internal H₂O₂ levels. A, Immunofluorescence images of 4‐hydroxynonenal (HNE). B, Relative HNE intensity. (C, D) HNE staining with PINK1 (localized in the mitochondria, C) and AQP11 (localized in the plasma membrane, D). HNE was mainly co‐localized with AQP11 that is present in the plasma membrane. E, Immunofluorescence images of HNE in HeLa and SAS parental cells after treatment with H₂O₂. F, Timing of lipid peroxidation and internal H₂O₂ levels. The results are expressed as the mean ± SEM. **P < .01 using Student's t test

3.3. H₂O₂ permeability using liposome and administration of oxidized lipid

We prepared liposomes encapsulating luminol and HRP to quantify the permeation of H₂O₂. Representative phospholipids of the plasma membrane (ie, DOPC and oxidized lipid POVPC) were used in this model. Internal levels of H₂O₂ can be detected as luminescence of luminol catalyzed by HRP (Figure 3A). The intensity of luminescence was increased following the administration of H₂O₂. Differential DOPC/POVPC ratio liposomes were prepared to investigate changes in H₂O₂ permeability. Maximum H₂O₂ transport appeared in the presence of several (2‐4) mol% of POVPC (Figure 3B). The POVPC administration experiment was conducted to confirm whether administration of this oxidized lipid led to cell death. The results showed that the administration of 12.5 μmol/L POVPC significantly increased the sensitivity of HeLa and SAS parental cells to H₂O₂. However, a higher concentration of POVPC did not induce a similar increase in sensitivity to H₂O₂ (Figure 3C). Furthermore, administration of different concentrations of POVPC revealed that the relationship between cell viability and POVPC concentration did not show a linear concentration‐dependent pattern. However, it showed a reverse bell shape, with the lowest viability in HeLa and SAS parental cells observed after the administration of 12.5 μmol/L POVPC (Figure 3D). In contrast, administration of 12.5 μmol/L POVPC did not increase sensitivity to H₂O₂. The relationship between cell viability and the concentration of POVPC showed a linear concentration‐dependent pattern in HeLa and SAS ρ⁰ cells that contained more oxidized lipid before treatment (Figure 3E, F).

Figure 3.

H₂O₂ permeability using liposome and administration of oxidized lipid. A, Structure of the liposome. The representative phospholipids of the plasma membrane (ie, 1,2‐dioleoyl‐sn‐glycero‐3‐phosphocholine (DOPC) and oxidized lipid 1‐palmitoyl‐2‐(5′‐oxo‐valeroyl)‐sn‐glycero‐3‐phosphocholine (POVPC)) were used. B, Effect of the oxidized lipid ratio to the H₂O₂ permeability of the liposome. Up to several (2‐4) mol%, the administration of oxidized lipid increased H₂O₂ permeability. C, D, Administration of oxidized lipid in HeLa and SAS parental (C) or ρ⁰ (D) cells before H₂O₂ treatment. The numbers in parentheses represent the concentration of POVPC (μmol/L). E, F, Effect of the concentration of POVPC on the viability of HeLa and SAS parental (E) or ρ⁰ (F) cells before H₂O₂ treatment (25 μmol/L). The results are expressed as the mean ± SEM. *P < .05, **P < .01 using Scheffe's F test (vs control), ^# P < .05, ^## P < .01 using Scheffe's F test (vs treatment with H₂O₂). C: control, P: treatment with POVPC (12.5 μmol/L), H: treatment with H₂O₂ (25 μmol/L for 48 h), P+H: treatment with H₂O₂ after treatment with POVPC

3.4. Gene expression and protein expression of ALOX

Gene expression of ALOX5, ALOX12, and ALOX15 – involved in lipid peroxidation of the plasma membrane – was upregulated in ρ⁰ cells (Figure 4A: ALOX5, 4B: ALOX12, and 4C: ALOX15). In addition, protein expression of ALOX12 and ALOX15, unlike that of ALOX5, was upregulated in both HeLa and SAS ρ⁰ cells (Figure 4D and Table 2). Investigation of spatial distribution showed that the ALOX12 and ALOX15 proteins were distributed in the nucleus and cytoplasm. Conversely, the ALOX5 protein was distributed only in the nucleus (Figure S1). A semiquantitative analysis of immunofluorescence revealed that the relative expression of the ALOX12 and ALOX15 proteins was upregulated in ρ⁰ cells (Figure 5A, C: ALOX12, Figure 5B, D: ALOX15).

Figure 4.

Expression of the ALOX gene and protein in ρ⁰ cells. A, Expression of the ALOX5 gene. B, Expression of the ALOX12 gene. C, Expression of the ALOX15 gene. The results are expressed as the mean ± SEM. **P < .01 using Student's t test. D, Western blotting of ALOX5, ALOX12, and ALOX15. Unlike ALOX5, the expression of ALOX12 and ALOX15 was upregulated in HeLa and SAS ρ⁰ cells

Figure 5.

Immunofluorescence of ALOX12 and ALOX15. A, B, Spatial distribution of ALOX12 (A) and ALOX15 (B). C, D, Relative expression of ALOX12 (C) and ALOX15 (D). The results are expressed as the mean ± SEM. *P < .05, **P < .01 using Student's t test

3.5. Inhibition of ALOX

Treatment with 10 μmol/L caffeic acid (ALOX‐5 inhibitor) or 10 μmol/L NDGA (universal ALOX inhibitor) was performed before treatment with H₂O₂ to investigate the effect of ALOX on the levels of internal H₂O₂ or HNE. The results showed that, unlike caffeic acid, NDGA significantly inhibited the levels of internal H₂O₂ and lipid peroxidation (Figure 6).

Figure 6.

Effect of ALOX inhibitor on internal H₂O₂ and lipid peroxidation. Treatment with 10 μmol/L caffeic acid (ALOX‐5 inhibitor) or 10 μmol/L nordihydroguaiaretic acid (NDGA) (universal ALOX inhibitor) was performed to investigate their effect on internal H₂O₂ and lipid peroxidation. A, Immunofluorescence images of internal H₂O₂. B, C, Relative intensity of internal H₂O₂ in HeLa (B) and SAS cells (C). D, Immunofluorescence images of 4‐hydroxynonenal (HNE). E, F, Relative intensity of internal H₂O₂ in HeLa (E) and SAS (F) cells. Control: no treatment. H₂O₂: treatment with 50 μmol/L H₂O₂. +CA: treatment with 10 μmol/L caffeic acid before treatment with 50 μmol/L H₂O₂. +NDGA: treatment with 10 μmol/L NDGA before treatment with 50 μmol/L H₂O₂. The results are expressed as the mean ± SEM.*P < .05, **P < .01 using Scheffe's F test (vs control), ^## P < .01 using Scheffe's F test (vs treatment with H₂O₂)

3.6. Gene and protein expressions of COX

Expression of the COX1 gene was downregulated in SAS ρ⁰ cells (Figure 7A). Conversely, expression of the COX2 gene was significantly upregulated in HeLa and SAS ρ⁰ cells (Figure 7B). Protein expression of COX1 did not change significantly between parental and ρ⁰ cells, whereas that of COX2 was downregulated in ρ⁰ cells (Figure 7C and Table 2). Immunofluorescence of COX revealed that COX1 was mainly localized in the nucleus, whereas COX2 was mainly localized in the cytoplasm (Figure 7D).

Figure 7.

Gene and protein expression of COX. A, Relative gene expression of COX1. B, Relative gene expression of COX2. C, Western blotting of COX1 and COX2. D, Spatial distribution of COX1 (green) and COX2 (red) in SAS parent cells. The results are expressed as the mean ± SEM. *P < .05, **P < .01 using Student's t test

3.7. Phosphorylation of Akt and protein expressions involved in apoptosis

Phosphorylation of Akt, which inhibits apoptosis and regulates cell proliferation, was investigated using western blotting. This analysis revealed a marked decrease in Akt phosphorylation in ρ⁰ cells (Figure 8 and Table 2). Expression of Akt downstream proteins was also investigated. Expression of FOXO1, which is downregulated by Akt phosphorylation and inhibited apoptosis, was increased. Moreover, pGSK3β^S9, which is the inactivated form of GSK3β, was decreased in ρ⁰ cells. Expression of the anti‐apoptotic protein BCL2 was decreased, whereas expression of the apoptosis‐inducing protein Bax was increased in ρ⁰ cells (Figure 8 and Table 2).

Figure 8.

Downregulation of Akt phosphorylation and promotion of apoptosis signaling in ρ⁰ cells. Western blots of Akt, pAkt^T308, pAkt^S473, FOXO1, pGSK3β^S9, BCL2, Bax, and β‐actin are shown. Downregulation of Akt phosphorylation was observed in both HeLa and SAS ρ⁰ cells. The expression of apoptosis‐promoting proteins (FOXO1 and Bax) was upregulated, whereas that of apoptosis‐inhibiting proteins (pGSK3β^S9 and BCL2) was downregulated

4. DISCUSSION

The present results indicated that the susceptibility of ρ⁰ cells to H₂O₂ may be attributed to an increase in the levels of intracellular ROS due to peroxidation of the plasma membrane. ρ⁰ cells from osteosarcomas have previously shown sensitivity to X‐ray irradiation and produced more ROS from mitochondria than their parental cells.14, 19 Similarly, lung carcinoma ρ⁰ cells have shown sensitivity to X‐ray irradiation by decreasing the expression of CuZn‐SOD.19 Conversely, ρ⁰ cells may be resistant to oxidative stress; and ρ⁰ cells from GM701 and BEAS‐2B have shown resistance to irradiation.19, 20 Additionally, GM701 ρ⁰ and BEAS‐2B ρ⁰ cells have shown resistance through inhibition of apoptosis and reduction of DNA damage, respectively. However, in these cell lines, intracellular levels of ROS were not determined. Therefore, it remained unknown whether ρ⁰ cells exhibited resistance to ROS, such as H₂O₂. Collectively, the increased production of internal ROS via oxidative stress may lead to increased cell death in ρ⁰ cells.

In our study, liposome membrane experiments showed that the content of oxidized lipid in liposomes increased H₂O₂ permeability by at least several (2‐4) mol% (Figure 3A, B). Administration of oxidized lipid to HeLa and SAS parent cells produced a decrease in cell viability following treatment with H₂O₂ (Figure 3C, D). The reduced cell viability induced by 12.5 μmol/L POVPC may reflect the results of the liposome experiments. As shown in Figure 3B, the highest uptake of H₂O₂ was caused at a POVPC concentration of several (2‐4) mol%, and administration of 12.5 μmol/L POVPC (Figure 3D) induced marked cell damage. In contrast, the administration of oxidized lipid did not affect cell viability following treatment with H₂O₂ in ρ⁰ cells, in which lipid peroxidation was enhanced (Figure 3E, F). From these results, the concentration of the oxidized lipids was thought to influence H₂O₂ permeability and viability. However, administration of oxidized lipids to parental cells whose plasma membrane was not highly oxidized, is effective in inducing cell death after treatment with H₂O₂. Recently, it was reported that the density of phospholipids in the plasma membrane is involved in the efficiency of membrane permeability (ie, uptake of ions or low‐molecular‐weight substances, such as H₂O₂).29 The results of our liposome experiment are in accordance with those findings reported by a recent study,29 suggesting that different densities of POVPC influence H₂O₂ membrane permeability. Uptake of H₂O₂ may be largely related to changes in the plasma membrane itself, rather than the presence of functional proteins on the membrane (ie, aquaporins) that are involved in the permeation of H₂O₂ into the cell.30, 31, 32, 33

In HeLa and SAS ρ⁰ cells, lipid peroxidation in the plasma membrane was higher than that observed in parental cells. The following three factors may be involved in lipid peroxidation and cell death: (a) downregulation of the antioxidant enzymes; (b) upregulation of the endogenous levels of ROS; and (c) upregulation of intracellular oxidase. It has been reported that catalase activity and expression of the Mn‐SOD gene were upregulated in ρ⁰ cells. In addition, expression of the GPx1 gene was downregulated in SAS ρ⁰ cells, whereas it was upregulated in HeLa ρ⁰ cells.17 In osteosarcoma, rhabdomyosarcoma, and lung ρ⁰ cells, glutathione peroxidase activity and the expression levels of the Mn‐SOD and CuZn‐SOD genes were upregulated. However, expression of the CuZn‐SOD protein in osteosarcoma ρ⁰ cells was downregulated.34 These results suggested that antioxidant enzymes are not the main factors involved in lipid peroxidation in ρ⁰ cells. It has been reported that the levels of superoxide from the mitochondria are decreased in ρ⁰ cells.35

Moreover, it has been reported that the administration of HNE leads to increased levels of internal ROS using 2′,7′‐dichlorodihydrofluorescein diacetate in parental cells. However, the levels of internal ROS in ρ⁰ cells did not increase significantly.36 Our data showed that the levels of HNE and ･OH were increased in ρ⁰ cells (Figures 2 and S2). Currently, the involvement of mitochondria‐derived ROS in lipid peroxidation of the plasma membrane in ρ⁰ cells remains unknown. It has been reported that an increase in ･OH from the mitochondria was detected in osteosarcoma ρ⁰ cells.14 Additionally, it has been reported that ･OH in cells was involved in the initiation of lipid peroxidation.37 Moreover, H₂O₂ reportedly produced ･OH in the presence of ferrous ion, leading to lipid peroxidation and cell death.38 These results indicated a positive correlation between the levels of ･OH and lipid peroxidation of the plasma membrane in ρ⁰ cells.

The expression of ALOX12 and ALOX15 (mRNA and protein levels) and COX2 was upregulated in ρ⁰ cells (Figure 4, 5, 6, 7). ALOX15 produces 13‐hydroxyoctadecadienoic acid, which is a precursor of HNE from linoleic acid.39 Inhibition of ALOX 12 and 15, unlike ALOX5, resulted in the downregulation of HNE and H₂O₂ permeability in ρ⁰ cells (Figure 6). Moreover, overexpression of ALOX15 enhanced erastin and RAS‐selective lethal compound 3‐induced ferroptosis, an iron‐dependent cell death via activation of the HNE.40 Furthermore, LOX15 reportedly plays a central role in the initiation and execution of ferroptosis.41 These results suggested that the expression of internal oxidative enzymes is more important than that of antioxidative enzymes for lipid oxidation in ρ⁰ cells. Additionally, ALOX15 plays an important role in the peroxidation of the plasma membrane in ρ⁰ cells. To date, there is no evidence showing that COX2 directly peroxidizes lipids and produces HNE. However, it has been reported that HNE induces the expression of the COX2 gene, and that this expression is stabilized by the p38 mitogen‐activated protein kinase.42 In addition, COX2 has been reported to be involved in PI3K/Akt signaling.43 Recently, it was reported that administration of HNE decreased Akt phosphorylation and induced apoptosis in an osteosarcoma cell line.44 These findings indicated a strong relationship between lipid peroxidation and intracellular signal pathways (eg, Akt signaling). It has been shown that phosphorylation of Akt was downregulated in HeLa ρ⁰ cells.45 Our results showed that Akt phosphorylation was downregulated in both types of ρ⁰ cells (Figure 8). Moreover, an apoptosis‐promoting signal was activated in ρ⁰ cells (Figure 8 and Table 2). These results indicated that the downregulation of Akt phosphorylation promoted cell death (ie, apoptosis and/or ferroptosis) in ρ⁰ cells via upregulation of COX2.

In conclusion, the sensitivity of ρ⁰ cells to treatment with H₂O₂ may be due to peroxidation of the plasma membrane, leading to increases in H₂O₂ permeability and cell death. Moreover, alterations in the state of the membrane may change the susceptibility of cells to drugs. These findings offer promise for the development of novel anticancer drugs, altering the oxidation state of the plasma membrane. Such drugs may be helpful in overcoming resistance to RT and drug therapy in cells.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

Supporting information

 

Tomita K, Takashi Y, Ouchi Y, et al. Lipid peroxidation increases hydrogen peroxide permeability leading to cell death in cancer cell lines that lack mtDNA. Cancer Sci. 2019;110:2856–2866. 10.1111/cas.14132