br Chronic Antioxidant Administration Increases Lung Cancer
Chronic Antioxidant Administration Increases Lung Cancer Cell Invasiveness, Reduces ROS, and Increases Reducing Equivalents
To define mechanisms underlying the pro-metastatic effect of chronic antioxidant administration, we established cell lines from tumors of control and NAC-treated K mice (mTC and mTN, respectively) and cultured them in the absence of antioxidants (Figure 1G). mTC and mTN DAPT proliferated at similar rates and retained functional p53 (Figures S2A–S2C), but mTN cells had higher invasive and migratory capacities (Figures 1H, 1I, and S2D). The migration of mTC cells increased to mTN levels after in-cubation with NAC for 7 days (Figure S2D). After intravenous (i.v.) injection into syngeneic recipient mice, mTN cells produced far more lung metastases than mTC cells and formed occasional metastatic tumors in the liver, kidney, heart, and rib cage
(I) Transwell invasion assay of mTC and mTN cells (n = 2 biological replicates/condition).
(J) Left, lung metastases in syngeneic mice 3 weeks after i.v. injection of mTC and mTN cells (0.5 3 105 cells/mouse; n = 10 mice/cell type). Red dots indicate mice with metastases to organs other than the lung. Right, representative lung sections.
(K) Left, percentage of NSG mice with lymph node metastasis 3 weeks after s.c. transplantation of mTC and mTN cells (2.5 3 105 cells/mouse; n = 6 mice/condition). Right, weight of primary s.c. tumors 3 weeks after transplantation.
Figure 2. Antioxidants Stabilize BACH1 by Reducing Free Heme Levels
(A) Venn diagram of RNA-seq data showing the number of genes differentially expressed in mTN versus mTC cells (blue). Small circles show genes with NRF2 and BACH1 binding motifs in their proximal promoter.
(B) Real-time qPCR analyses of BACH1 and NRF2 target genes in mTC and mTN cells (n = 3). Data are first normalized to Rplp0 expression and then to mTC and are mean of two experiments.
(C) Kaplan-Meier graph of TCGA data showing survival of lung adenocarcinoma (LUAD) patients with high and low BACH1 expression (Z score = 2).
(D) Schematic of ROS- and hemin-induced BACH1 degradation.
(E) Left, western blots showing steady-state amounts of BACH1 in mTC and mTN cell lines (n = 3). Right, amounts of BACH1 determined by densitometry of protein bands from two experiments. ACTIN was the loading control.
(F) Left, western blot showing amounts of BACH1 in mTC cells incubated with NAC or Trolox (>7 days). Right, amounts of BACH1 determined by densitometry of protein bands from three experiments.
(legend continued on next page)
(Figure 1J). Moreover, after subcutaneous (s.c.) implantation, only mTN cells metastasized to lymph nodes (Figure 1K).
Incubation with NAC and Trolox (a water-soluble vitamin E analog) for 7 days increased migration and invasion (Figures S2E–S2N) in 5 of 7 human non-small cell lung cancer (NSCLC) cell lines. These effects did not depend on the mutational profile and remained when antioxidants were withdrawn 48 h before the assay (Figures S2O–S2R).
ROS levels were lower in mTN cells and in NSCLC cells pre-incubated for 7 days with NAC than in controls (Figures S2S and S2T). Analyses with genetically encoded cytosolic H2O2 and glutathione disulfide (GSSG)/glutathione (GSH)-biosensors revealed that baseline levels of oxidation were lower in NAC-treated than control A549 cells and returned faster to redox homeostasis after addition of the pro-oxidant menadione (Fig-ure S2U). Moreover, the total amounts of GSH, NADH, and NADPH were higher in mTN cells and in NAC- and Trolox-treated NSCLC cells than in controls (Figures S2V–S2AA). NAC is a pre-cursor of GSH, but the increased migration of mTN cells did not depend on new GSH synthesis, as the migration of mTN and also mTC cells was unaffected by the GSH inhibitor buthionine sul-foximine and GSH (Figures S2AB–S2AD).
Antioxidants Stabilize BACH1 by Reducing Levels of ROS and Free Heme
We next did a transcriptomic analysis (RNA sequencing [RNA-seq]), which identified 1,025 genes whose mRNA levels differed significantly between mTN and mTC cell lines (Figure 2A). Gene ontology analyses revealed that mTN cells were enriched in mRNAs that help regulate cell motility and migration, cell-cell adhesion, and stem cell differentiation, consistent with activation of a pro-metastatic transcriptional program (Figure S3A). Conversely, genes involved in antioxidant mechanisms were depleted in mTN cells, reflecting the low levels of oxidative stress (Figure S3B). NRF2 target genes were expressed at lower levels in mTN than mTC cells (Figures S3C and S3D). Moreover, NRF2 protein levels were lower in lung tumors from antioxidant-treated mice than controls (Figure S3E). In line with these findings, we identified an NRF2-binding motif and also a BACH1-binding motif in the proximal promoters of 267 and 310 genes, respec-tively (Figure 2A). BACH1 displaces NRF2 from promoters and represses the expression of antioxidant genes (Figure S3F). Indeed, 207 of 267 genes with a NRF2 motif also contained a consensus motif recognized by BACH1 (Figure 2A).