LC3A-mediated autophagy regulates lung cancer cell plasticity

Abstract

ATG14: autophagy related 14; CDH2: cadherin 2; ChIP-qPCR: chromatin immunoprecipitation quantitative polymerase chain reaction; CQ: chloroquine; ECAR: extracellular acidification rate; EMT: epithelial-mesenchymal transition; EPCAM: epithelial cell adhesion molecule; MAP1LC3A/LC3A: microtubule associated protein 1 light chain 3 alpha; MAP1LC3B/LC3B: microtubule associated protein 1 light chain 3 beta; MAP1LC3C/LC3C: microtubule associated protein 1 light chain 3 gamma; NDUFV2: NADH:ubiquinone oxidoreductase core subunit V2; OCR: oxygen consumption rate; ROS: reactive oxygen species; RT-qPCR: reverse-transcriptase quantitative polymerase chain reaction; SC: scrambled control; shRNA: short hairpin RNA; SNAI2: snail family transcriptional repressor 2; SOX2: SRY-box transcription factor 2; SQSTM1/p62: sequestosome 1; TGFB/TGF-β: transforming growth factor beta; TOMM20: translocase of outer mitochondrial membrane 20; ZEB1: zinc finger E-box binding homeobox 1.

Keywords: Autophagy; LC3A; SOX2; cancer cell plasticity; lung cancer; mitochondria dynamics.

Figure 1.

Differential LC3A expression and autophagic activity in lung cancer cells. (A) One-way ANOVA and Tukey’s multiple comparison analysis of LC3A expression with histological grades in non-small cell lung cancer (N = 109) from GSE43580 database. *p < 0.05, **p < 0.01, ***p < 0.001. (B) Chi-square correlation analysis of LC3A expression with distant metastasis in lung adenocarcinoma (N = 369) from TCGA-LUAD cohort. *p < 0.05. (C) Represented IHC images (upper) of heterogeneous LC3A expression in lung adenocarcinoma (case #1) and squamous cell carcinoma (case #2). Quantitative analysis (lower) of LC3A expression and heterogeneity of non-small cell lung cancer (N = 44). Scale bar: 200 μm. (D) Clonogenic (left) and cell-tracking migration (middle) assays of CL1-0 versus CL1-5 cells. RT-qPCR (right) of LC3A expression in CL1-0 or CL1-5 cells. ***p < 0.001. (E) Clonogenic (left) and cell-tracking migration (middle) assays of H1650 versus H1650-T cells. RT-qPCR (right) of LC3A expression in H1650 and H1650-T cells. **p < 0.01, ***p < 0.001. (F) Immunoblotting analysis of LC3A and GAPDH expression in CL1-0 and CL1-5 cells treated with or without chloroquine (CQ, 12.5 μM) for 24 h. (G) Immunofluorescence analysis to assess the expression of LC3A (green) in CL1-0 and CL1-5 cells treated with or without chloroquine (CQ, 25 μM) for 24 h. Nuclei and F-actin were stained with DAPI (blue) and phalloidin (red), respectively. Scale bar: 20 μm.

Figure 2.

LC3A expression is associated with lung cancer proliferation. (A) RT-qPCR (left upper) and immunoblotting analysis (left lower) of LC3A and GAPDH expression and clonogenic analysis (right) of CL1-0 cells transduced with the lentiviral vector encoding shLC3A (#1 and #2) or scrambled control (SC) for 14 days. shLC3A#1 and shLC3A#2 target different regions in LC3A mRNA. **p < 0.01, ***p < 0.001. (B) RT-qPCR (left upper) and immunoblotting analysis (left lower) of LC3A and GAPDH expression and clonogenic analysis (right) of H1650 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. *p < 0.05, **p < 0.01, ***p < 0.001. (C) Cell cycle analysis of CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. Proportions of cell cycle phases were quantified in the table.

Figure 3.

Autophagy regulates mitochondrial dynamics and degradation as well as ROS production. (A) Representative confocal fluorescence images of mitochondrial morphology in CL1-0 (fragmented, left) and CL1-5 (hyperfused, right) cells stained with mitochondrial marker TOMM20. Scale bar: 10 μm. Proportions of mitochondrial phenotypes (fragmented and hyperfused) were quantified. **p < 0.01. (B) Immunoblotting analysis to assess NDUFV2 expression from mitochondria fraction (Mito) and SQSTM1 and GAPDH expression from total cell lysate (TCL) of CL1-0 and CL1-5 cells treated with or without chloroquine (CQ, 30 μM) for 24 h. (C) OCR analysis of CL1-0 and CL1-5 cells. **p < 0.01, ***p < 0.001. (D) MitoSOX flow cytometry analysis to detect ROS levels in CL1-0 and CL1-5 cells treated with or without chloroquine (CQ, 12.5 μM) for 24 h. *p < 0.05. (E) Luminol chemiluminescence analysis to assess ROS levels in CL1-0 and CL1-5 cells treated with or without chloroquine (CQ, 12.5 μM) for 24 h. (F) Quantitative analysis of mitochondrial morphologies in CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. **p < 0.01. (G) OCR analysis of CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. **p < 0.01, ***p < 0.001. (H) Luminol chemiluminescence analysis of CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days.

Figure 4.

LC3A-mediated autophagy is engaged in lung cancer cell plasticity. (A) RT-qPCR of LC3A expression in CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. ***p < 0.001. (B) Immunoblotting analysis to assess LC3A, SQSTM1, and GAPDH expression in CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 3 days. (C) Clonogenic assay of CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. ***p < 0.001. (D) Cell-tracking migration (left) and matrigel invasion (right) assays of CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC) for 14 days. ***p < 0.001.

Figure 5.

Association of LC3A levels with SOX2 and EMT markers expression in lung cancer. (A) RT-qPCR analysis to assess SOX2 and LC3A expression in CL1-0, CL1-5 and CL1-5-S3 cells. *p < 0.05, ***p < 0.001. (B) RT-qPCR analysis to assess EPCAM and SNAI2 expression in CL1-0, CL1-5 and CL1-5-S3 cells. **p < 0.01, ***p < 0.001. (C) Clonogenic (left), matrigel invasion (middle), and cell-tracking migration (right) analysis of CL1-0, CL1-5 and CL1-5-S3 cells. *p < 0.05, **p < 0.01, ***p < 0.001. (D) RT-qPCR of LC3A expression in CL1-5-S3 cells transduced with the lentiviral vector encoding shLC3A (+) or scrambled control (-) for 14 days. ***p < 0.001. (E) Clonogenic (left), matrigel invasion (middle), and cell-tracking migration (right) analysis of CL1-5-S3 cells transduced with the lentiviral vector encoding shLC3A (+) or scrambled control (-) for 14 days. ***p < 0.001. (F) Hierarchical clustering analysis of LC3A, SOX2, NBR1, epithelial marker (CDH1, EPCAM, and TJP3), and mesenchymal marker (CDH2, SNAI1, SNAI2, TWIST1, TWIST2, VIM, ZEB1, and ZEB2) expression in primary lung adenocarcinoma from TCGA-LUAD (N = 514). The heatmap was generated using Morpheus software.

Figure 6.

SOX2 regulates LC3A in lung cancer cells and predicts patient prognosis in lung adenocarcinoma. (A) RT-qPCR (left) and immunoblotting (right) assays to assess SOX2 and GAPDH expression in CL1-0 cells transduced with the lentiviral vector encoding shLC3A or scrambled control (SC). ***p < 0.001. (B) RT-qPCR to assess SOX2 expression in H1650 versus H1650-T cells. ***p < 0.001. (C) RT-qPCR (left) and immunoblotting (right) assays to assess SOX2 and LC3A expression in CL1-5 cells transduced with the lentiviral vector encoding SOX2 cDNA (SOX2) or empty control (Ctrl). ***p < 0.001. (D) ChIP-qPCR analysis to access the occupancy of SOX2 at the indicated regions (1–6) along the LC3A enhancer (Chr20: 34,514,477–34,518,001) in CL1-0 cells. (E) Scatter plots of correlation between SOX2 and LC3A expression in primary lung adenocarcinoma from GSE27262 (left) and TCGA-LUAD (right), displaying positive correlations between SOX2 and LC3A. (F) Hierarchical clustering analysis of SOX2, LC3A, LC3B, and LC3C expression in primary lung adenocarcinoma from TCGA-LUAD (N = 514). (G) Kaplan–Meier analysis to assess the correlation of SOX2 (left) and LC3A (middle) expression with the overall survival of lung adenocarcinoma patients (N = 500) from TCGA-LUAD cohort. The overall survival analysis was further stratified by SOX2-high/LC3A-high and SOX2-low/LC3A-low signatures (right) for Kaplan–Meier analysis (N = 325). Different groups were compared using log-rank test. *p < 0.05.

Figure 7.

Model of the crosstalk between LC3A-mediated autophagy and SOX2 proliferation signaling in lung cancer. SOX2 signaling promotes proliferation but inhibits SNAI2 EMT signaling in lung cancer. SOX2-positive high-proliferative lung cancer cells exhibit higher oxygen consumption rate (OCR) and generate more reactive oxygen species (ROS), which cause mitochondrial damage and dysfunction. LC3A-mediated autophagy works with mitochondrial fission to remove damaged mitochondria (left). Deficient LC3A expression limits SOX2 proliferation signaling and tilts the balance toward the enrichment of low-proliferative/high-invasive cancer cells, which exhibit lower OCR and decreased ROS, concomitant with increased mitochondrial fusion for mitochondria restoration (right).