INTRODUCTION
Primary bronchogenic carcinoma mainly occurs in bronchial mucosal epithelial cells, with high malignancy and mortality rate1. Non-small cell lung cancer (NSCLC) is the most common histological type of lung cancer, including squamous cell carcinoma (LUSC), adenocarcinoma (LUAD), and large cell carcinoma2. Although the pathogenesis of lung cancer is still unclear, smoking, environmental pollution, and diet are all related to the occurrence of lung cancer3. Smoking is the most important risk factor for lung cancer, especially with a direct impact on the occurrence of NSCLC3. Although previous studies have shown that ferroptosis-related genes were upregulated in chronic obstructive pulmonary disease patients who smoke compared to non-smoking healthy individuals4, to our knowledge, no study has shown the roles and mechanisms of smocking in ferroptosis in patients with lung cancer.
Nicotine, as the main alkaloid in cigarettes, is not carcinogenic but can promote the development of lung cancer5. Benzo(a)pyrene (BaP) is the main carcinogen in smoke and is closely related to the occurrence of lung cancer6. Both nicotine and BaP promote tumor cell proliferation, invasion, epithelial-mesenchymal transition (EMT), and angiogenesis5-7. Previous studies have demonstrated that benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide induces ferroptosis in neuroblastoma cells8. However, no study has shown the roles of nicotine in ferroptosis. Ferroptosis is an iron-dependent new type of programmed cell death, which differs from apoptosis, necrosis, and autophagy9. The main mechanism of ferroptosis is that divalent iron or ester oxygenase catalyzes the peroxidation of unsaturated fat acid, leading to cell death9. The depletion of intracellular glutathione leads to a decrease in glutathione peroxidase 4 (GPX4) activity, and lipid peroxides cannot be metabolized through the reduction reaction catalyzed by GPX4, leading to the accumulation of a large amount of lipid peroxidation and inducing ferroptosis10. Ferroptosis has great potential in cancer treatment by inhibiting tumor growth and killing tumor cells10.
In this study, we identified the expression profiles of ferroptosis-related genes in the smoking and non-smoking patients with LUSC and LUAD, based on The Cancer Genome Atlas (TCGA) database. In addition, we analyzed the influence of nicotine and BaP on ferroptosis in NSCLC cells. According to these hints, the distinct molecular mechanisms of nicotine and BaP in ferroptosis of LUSC and LUAD cells were found.
METHODS
Study design
This study was a secondary bioinformatics analysis followed by an experimental cell culture analysis.
Bioinformatics analysis
Lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD) mRNA data (HTSeq FPKM) and corresponding clinical data, were downloaded from the TCGA database (http://cancergenome.nih.gov/). R programming ggplot2 and pheatmap packages were used to show the different expression levels of ferroptosis-related genes between the patients with LUSC and LUAD. R programming Survival package was used to evaluate the association of ferroptosis-related genes with the prognosis of the patients.
Cell culture
Lung adenocarcinoma cells, A549, and lung squamous cell carcinoma cells, H1869, were purchased from the American Type Culture Collection (ATCC, Bethesda, MD) and grown in DMEM medium (Hyclone, Logan, UT) supplemented with 10% fetal bovine serum and antibiotics (100 U/mL penicillin and 100 μL/mL streptomycin).
shRNA transfection and nicotine/BaP treatment
A549 cells and H1869 cells were transfected with shFDFT1 plasmid (sc-61610-SH, Santa Cruz Biotechnology, Shanghai, China) or shSLC1A5 plasmid (sc-60210-SH, Santa Cruz) using Lipofectamine 2000 (Invitrogen, Shanghai, China). Cells with or without shRNA transfection were treated with 0.5 μM nicotine (613223, Sigma-Aldrich, Shanghai, China) or BaP (B1760, Sigma-Aldrich) to keep a constant concentration for 48 h before experiments.
Real-time reverse transcription PCR
Total RNA was isolated from cells using an RNeasy Mini Kit (Beyotime Biotechnology, Shanghai, China). FDFT1 primers were: 5’-TGTGACCTCTGAACAGGAGTGG-3’ (sense) and 5’-GCCCATAGAGTTGGCACGTTCT-3’ (antisense).
SLC1A5 primers were: 5’-TCCTCTTCACCCGCAAAAACCC-3’ (sense) and 5’-CCACGCCATTATTCTCCTCCAC-3’ (antisense).
GAPDH was used as an internal control for normalization of the results. GAPDH primers were: 5’-AATGGACAACTGGTCGTGGAC-3’ (sense) and 5’-CCCTCCAGGGGATCTGTTTG-3’ (antisense).
MTT assay
Cytotoxicity assay kit was purchased from Beyotime. After cells (1×103 cells/well) were attached on 96-well plates for 48 h, 20 μL of MTT solution was added to each well. Absorbance density values were checked at 570 nm to determine the cell viability after 4 h using a TECAN microplate reader (Tecan Trading AG, Switzerland).
Detection of reactive oxygen species (ROS)
Concentrations of ROS in cells were measured using the ROS detection kit (Beyotime) according to the manufacturer’s instructions. Cells were washed with PBS and incubated with DCFH-DA (1:1000) at 37oC. After incubation, cells were analyzed by flow cytometry (Becton, Dickinson and Company, Shanghai, China).
Detection of malondialdehyde (MDA)
Concentrations of MDA were determined using the lipid peroxidation MDA assay kit (Beyotime) according to the manufacturer’s instructions. Absorbance was measured at 532 nm using a TECAN microplate reader (Tecan Trading AG).
Glutamine uptake assay
Cells were incubated with [3H]-L-Gln (200 nmol/L) in Gln-free medium at 37oC for 20 min and then were harvested for Gln measurements using a liquid scintillation counter (PerkinElmer, Shanghai, China).
Western blot
Cellular protein (30 μg) was separated with 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and transferred to transferred to nitrocellulose (NC) filter membranes (Beyotime). The membranes were then incubated in 5% milk for 2 h at room temperature and with the first antibody overnight at 4°C. Primary antibodies were FDFT1 (sc-271602, Santa Cruz), SLC1A5 (8057, Cell Signaling Technology, Shanghai, China), GPX4 (sc-166570, Santa Cruz) and GAPDH (sc-74512, Santa Cruz). After 24 h, the membranes were incubated with secondary antibodies for 2 h at room temperature. The signals were detected using an enhanced chemiluminescence kit (Beyotime).
RESULTS
Identification of the different genes related to ferroptosis in LUAD and LUSC tissues
In order to identify the influence of smocking on ferroptosis, we compared the genes related to ferroptosis in the LUAD and LUSC patients with different smoking habits based on the TCGA database. There was a statistical difference in FDFT1 expression in the smoking and non-smoking LUSC patients and SLC1A5, ALT1, LPCAT3, DPP4, and RPL8 in the smoking and non-smoking LUAD patients (Figures 1A and 2A, p<0.05). FDFT1 was expressed lower in the non-smoking LUSC patients than in the smoking LUSC patients (Supplementary file Figure 1A, p<0.05). Lower SLC1A5 and RPL8 expression were observed in the non-smoking LUAD patients than in the smoking LUAD patients, while ALT1, LPCAT3, and DPP4 were opposite (Supplementary file Figure 1B, p<0.05). In addition, we checked the correlation among these genes in each tissue. A positive correlation was found between FDFT1 and GPX4 in smoking LUSC patients but not in non-smoking LUSC patients (Figures 1B and 1C). SLC1A5 was positively correlated with GPX4 in the smoking LUAD patients, while no correlation was found between them in the non-smoking LUAD patients (Figures 2B and 2C).
Figure 1
Differentially ferroptosis-related genes in the LUSC patients with different smocking status: A) Heatmap of differentially ferroptosis-related genes in smocking and non-smocking LUSC patients; B) G1: non-smocking patients; G2: smocking patients. Circles represent the ferroptosis-related mRNA, line represents the relationship between genes in non-smocking LUSC patients; and C) Smocking LUSC patients
Figure 2
Differentially ferroptosis-related genes in the LUAD patients with different smocking status: A) Heatmap of differentially ferroptosis-related genes in smocking and non-smocking LUAD patients; B) G1: non-smocking patients; G2: smocking patients. Circles represent the ferroptosis-related mRNA, and lines represent the relationship between genes in non-smocking LUAD patients; and C) Smocking LUAD patients
The function and molecular mechanisms of smocking in ferroptosis of LUAD and LUSC cells
Based on the bioinformatic results, we checked the influence of two main chemical components of smoking, nicotine and benzo(a)pyrene (BaP), on ferroptosis in LUAD and LUSC cells. The proliferation of A549 and H1869 cells was inhibited after BaP treatment, compared with the parental cells (Figure 3A, p<0.05). After FDFT1 knockdown, the proliferation of A549 and H1869 cells was higher than the parental cells (Figure 3A, p<0.05). Compared with matched cells, ROS and MDA levels were upregulated in the cells with BaP treatment (Figures 3B and 3C, p<0.05). Lower glutamine (Gln) consumption was observed in the cells with BaP treatment compared with the parental cells (Figure 3D, p<0.05). However, BaP-induced ferroptosis in LUAD and LUSC cells was inhibited by FDFT1 knockdown (Figure 3, p<0.05). Nicotine induced the proliferation of A549 and H1869 cells (Figure 4A, p<0.05) and downregulated the ROS level and the MDA levels in these two cells (Figures 4B and 4C, p<0.05). Nicotine upregulated Gln consumption in A549 and H1869 cells (Figure 4D, p<0.05). Interestingly, SLC1A5 knockdown can counteract the effect of nicotine on ferroptosis of A549 and H1869 cells (Figure 4, p<0.05).
Figure 3
The roles of FDFT1 and BaP in ferroptosis of NCSLC cells: A) Cell viability was assayed by MTT assays of A549 cells and H1869 cells after FDFT1 knockdown and/or BaP treatment and their parental cells; B) Intracellular ROS; C) MDA formation; and D) Gln uptake activities of A549 cells and H1869 cells after FDFT1 knockdown and/or BaP treatment and their parental cells
Figure 4
The roles of SLC1A5 and nicotine in ferroptosis of NCSLC cells: A) Cell viability was assayed by MTT assays of A549 cells and H1869 cells after SLC1A5 knockdown and/or nicotine treatment and their parental cells; B) Intracellular ROS; C) MDA formation; and D) Gln uptake activities of A549 cells and H1869 cells after SLC1A5 knockdown and/or nicotine treatment and their parental cells
Western blot results showed that nicotine-induced EGFR, SLC1A5, and GPX4 expression in A549 and H1869 cells (Supplementary file Figure 3A, p<0.05). In SLC1A5 knockdown cells, nicotine can still induce GPX4 expression slightly (Supplementary file Figure 3B, p<0.05). BaP induced FDFT1 expression and inhibited GPX4 expression in A549 and H1869 cells (Supplementary file Figure 3C, p<0.05). FDFT1 knockdown can restore GPX4 protein levels in A549 and H1869 cells with BaP treatment (Supplementary file Figure 3D, p<0.05).
DISCUSSION
Smoking is the most important cause of chronic obstructive pulmonary disease (COPD) and an important trigger for lung cancer, especially lung squamous cell carcinoma3. Cigarette smoke induces lipid peroxidation, ROS accumulation, and reduced mitochondrial volume in human bronchial epithelial cells4. Iron metabolism disorders are closely related to the occurrence and development of lung cancer11. High expression of GPX4 indicates poor prognosis in lung cancer patients12. In this study, we screened the distinct ferroptosis-related genes in LUAD and LUSC patients with different smoking habits based on the TCGA database for the first time. We confirmed high FDFT1 expression in smoking LUSC patients and high SLC1A5 expression in smoking LUSC patients. Their expressions were correlated with GPX4 expression. FDFT1 and SLC1A5 have been proven to be regulatory factors of ferroptosis in physiological and pathological conditions13-15. Consistent with previous studies13-15, we also found that FDFT1 knockdown will inhibit ROS accumulation and MDA elevation in both LUAD and LUSC cells, while SLC1A5 knockdown was the opposite. All the evidence showed us that smoking will influence the ferroptosis in LUAD and LUSC patients by changing the expression of FDFT1 and SLC1A5.
Cigarette smoke contains over 4000 known chemicals, of which over 400 are toxic to the human body5. Nicotine is the highest content of alkaloids in smoke and is the main chemical substance that makes people dependent on cigarettes5. A previous study showed nicotine-induced blood testis barrier (BTB) damage by causing ferroptosis in the testis16. Nicotine-induced epidermal growth factor receptor (EGFR) expression promotes the progression of non-small-cell lung cancer17. Downregulation of the EGFR-SLC1A5 complex sensitizes human head and neck squamous cell carcinoma cells to ROS-induced apoptosis18. This study found nicotine-induced EGFR expression, followed by increased SLC1A5, in LUAD and LUSC cells. SLC1A5 takes up glutamine, which is metabolized into glutamate through mitochondrial glutaminase GLS2, to protect cells from ferroptosis19. BaP is a representative carcinogen of polycyclic aromatic hydrocarbons, mainly produced by high-temperature decomposition and incomplete combustion of carbon-containing substances5. The function of BaP for pulmonary diseases has been studied widely and comprehensively, including COPD, pulmonary fibrosis, and lung cancer20. It can induce lung toxicity and inflammation and convert COPD into lung cancer via epithelial-mesenchymal transition (EMT)21. The roles and mechanisms of BaP in ferroptosis are relatively complex. BaP has been used to construct lung cancer animal models22. Reduced GSH and MDA levels and increased oxidative stress were observed in BaP-induced lung cancer in rat models22. Benzo(a)pyrene diol epoxide (BPDE), an active metabolite of BaP, downregulated GPX4 and SLC7A11 and induced ferroptosis in neuroblastoma cells8. It seems that BaP has a dual effect on lung cancer.
On the one hand, it induces tumorigenesis; on the other hand, it can induce ferroptosis. This study found BaP-induced ferroptosis in LUAD and LUSC cells through upregulation of FDFT1. After FDFT1 knockdown, BaP cannot induce GPX4 expression and ferroptosis in these cells. FDFT1 has been identified as a ferroptosis-associated gene in colorectal cancer and renal cell carcinoma13,23. To our knowledge, this is the first study to show the roles of FDFT1 in the ferroptosis of lung cancer cells with BaP treatment.
CONCLUSIONS
In this study, the ferroptosis-associated gene signature was identified in LUAD and LUSC patients with different smoking status. We confirmed nicotine-protected LUAD and LUSC cells from ferroptosis by upregulating EGFR and SLC1A5 expression. BaP-induced ferroptosis in these cells depends on FDFT1 expression. Interestingly, our research shows that cigarette smoke also contains chemical components that can induce tumor cell death.
