Antioxidants stimulate BACH1-dependent tumor angiogenesis

Abstract

Lung cancer progression relies on angiogenesis, which is a response to hypoxia typically coordinated by hypoxia-inducible transcription factors (HIFs), but growing evidence indicates that transcriptional programs beyond HIFs control tumor angiogenesis. Here, we show that the redox-sensitive transcription factor BTB and CNC homology 1 (BACH1) controls the transcription of a broad range of angiogenesis genes. BACH1 is stabilized by lowering ROS levels; consequently, angiogenesis gene expression in lung cancer cells, tumor organoids, and xenograft tumors increased substantially following administration of vitamins C and E and N-acetylcysteine in a BACH1-dependent fashion under normoxia. Moreover, angiogenesis gene expression increased in endogenous BACH1-overexpressing cells and decreased in BACH1-knockout cells in the absence of antioxidants. BACH1 levels also increased upon hypoxia and following administration of prolyl hydroxylase inhibitors in both HIF1A-knockout and WT cells. BACH1 was found to be a transcriptional target of HIF1α, but BACH1's ability to stimulate angiogenesis gene expression was HIF1α independent. Antioxidants increased tumor vascularity in vivo in a BACH1-dependent fashion, and overexpressing BACH1 rendered tumors sensitive to antiangiogenesis therapy. BACH1 expression in tumor sections from patients with lung cancer correlated with angiogenesis gene and protein expression. We conclude that BACH1 is an oxygen- and redox-sensitive angiogenesis transcription factor.

Keywords: Angiogenesis; Hypoxia; Lung cancer.

Figure 1. Antioxidants stabilize BACH1 and induce angiogenesis gene expression in NSCLC organoids and tumors by upregulating BACH1 expression.

(A) Experimental design. (B) BACH1 protein levels in spheroids incubated for 7 days with 25 μM VitC and BACH1 levels by densitometry (n = 3 experiments). (C) Reverse transcription quantitative PCR (RT-qPCR) of BACH1 in A549 and H838 spheroids , lung tumor organoids incubated for 7 days with 25 μM VitC, and A549 xenograft tumors from mice administered VitC (3.47 g/L) in the drinking water for 7 weeks (n = 3 experiments). (D) RT-qPCR of angiogenesis genes in spheroids incubated with 25 μM VitC for 7 days (n = 3 experiments). (E) VEGFR2 protein levels in spheroids incubated with 25 μM VitC for 7 days and VEGFR2 levels determined by densitometry (n = 6 experiments). Ctrl, control. Data indicate the mean ± SEM. Statistical significance was determined by 2-tailed, unpaired Student’s t test (C–E) and 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (B).

Figure 2. BACH1 controls the expression of angiogenesis genes under normoxia.

(A) RT-qPCR of angiogenesis genes in BACH1^OE and BACH1^WT spheroids under normoxia (n = 3 experiments). (B) VEGFR2 protein levels in BACH1^OE and BACH1^WT spheroids and VEGFR2 levels by densitometry (n = 4 experiments). (C) RT-qPCR of angiogenesis genes in BACH1^+/+ and BACH1^–/– spheroids under normoxia (n = 4 experiments). (D) VEGFR2 protein levels in BACH1^+/+ and BACH1^–/– spheroids and VEGFR2 levels by densitometry (n = 3 experiments). (E) RT-qPCR of angiogenesis genes in BACH1^+/+ and BACH1^–/– spheroids incubated for 7 days with 25 μM VitC or vehicle (Ctrl) (n = 3 experiments). (F) Top, VEGFR2 and BACH1 protein levels in BACH1^+/+ and BACH1^–/– spheroids incubated for 7 days with 25 μM VitC and VEGFR2 protein levels by densitometry (n = 4 experiments). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001, by 2-tailed, unpaired Student’s t test (A–D) and 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (E and F).

Figure 3. BACH1-mediated expression of angiogenesis and glycolysis genes correlates with BACH1-dependent epigenetic changes at promoter regions.

(A) Genome-wide profiling of BACH1 chromatin enrichment in A549 spheroids using CUT&Tag. (B) Transcription factor DNA-binding motif analysis of BACH1 CUT&Tag peaks. (C) Genome-wide plot of H3K27ac peak density in BACH1^+/+ and BACH1^–/– A549 spheroids; note that the 2 lines for each genotype replicate (rp1/rp2) overlap. (D and E) Integrative Genomics Viewer (IGV) tracks showing H3K27ac levels at the indicated angiogenesis (D) and glycolysis (E) gene loci in BACH1^+/+ and BACH1^–/– A549 spheroids. BACH1 peaks are shown at the bottom to indicate overlap with H3K27ac-marked regions. Regions with significant H3K27ac changes in BACH1^–/– compared with BACH1^+/+ A549 spheroids are highlighted in blue; the percentage of change is indicated in red. (F) RT-qPCR of the expression of a broader set of angiogenesis-related genes in tumors from mice engrafted with BACH1^+/+ and BACH1^–/– A549 lung cancer cells. The mice were given VitC (3.47 g/L) or normal drinking water for 7 weeks (n = 3 experiments). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001, by 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (F).

Figure 4. BACH1 expression under normoxia and hypoxia is HIF1α dependent.

(A) RT-qPCR of HIF1A and HIF2A in spheroids incubated for 7 days with antioxidants under normoxia (n = 3 experiments). (B) Top: HIF1α levels in spheroids incubated with VitC by Western blotting. Middle: HIF1α levels by densitometry (n = 3 experiments). Bottom: HIF2α protein levels by Western blotting. (C) RT-qPCR of BACH1 expression in spheroids under normoxia (21% O₂) and hypoxia (1% O₂) (n = 3 experiments), BACH1 protein levels by Western blotting, and BACH1 levels by densitometry (n = 3 experiments. (D) Left top: BACH1 protein levels by Western blotting in spheroids incubated for 16 hours with prolyl hydroxylase inhibitors. Right top: BACH1 levels by densitometry (n = 3 experiments). Left bottom: HIF1α protein levels by Western blotting. Right bottom: HIF1α levels by densitometry (n = 3 experiments). (E) Top: RT-qPCR of BACH1 expression in HIF1α-overexpressing (HIF1A^OE) and control (HIF1A^WT) spheroids under normoxia (n = 3 experiments). Middle: BACH1 protein levels by Western blotting. Bottom: BACH1 levels by densitometry (n = 4 experiments). (F) Experiments similar to those in E using HIF2α-overexpressing (HIF2A^OE) and control (HIF2A^WT) spheroids (n = 6 experiments). Data indicate the mean ± SEM. P values were determined by 2-tailed, unpaired Student’s t test (A, C, E, and F) and 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (B and D).

Figure 5. BACH1 increases angiogenesis gene expression in HIF1A-deficient lung cancer cells.

(A) HIF1A-knockout validation with RT-qPCR and Western blotting. (B) BACH1 protein levels by Western blotting in HIF1A^–/– and HIF1A^+/+ spheroids under normoxia and hypoxia and BACH1 levels by densitometry (n = 3 experiments). (C) BACH1 protein levels by Western blotting in HIF1A^–/– spheroids incubated for 16 hours with prolyl hydroxylase inhibitors and BACH1 levels by densitometry (n = 4 experiments). (D) Genome-wide BACH1 CUT&Tag peak density plot of HIF1A^+/+ and HIF1A^–/– spheroids. (E) RT-qPCR of BACH1 and angiogenesis genes in HIF1A^–/– spheroids with lentiviral BACH1 overexpression (BACH1^L-OE) and controls (BACH1^L-CTR) (n = 3 experiments). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001, by 2-tailed, unpaired Student’s t test (A and E) and 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (B and C).

Figure 6. BACH1 expression correlates with angiogenesis gene and protein expression in human NSCLC samples.

(A) Heatmap showing TCGA lung cancer cases with low (left) and high (right) BACH1 expression. Angiogenic genes whose expression differed significantly between the 2 groups are listed on the right along with the P value for the correlation with BACH1 expression. (B) Representative immunohistochemical staining for BACH1, VEGFA, and VEGFR2 in consecutive sections of tumors from patients with KRAS-mutant NSCLC. Tumor sections with low BACH1 expression (left); tumor section with high BACH1 expression (right). Original magnification, ×20. Scale bars: 100 μm. (C) Comparisons of VEGFA and VEGFR2 expression with BACH1 protein expression in human NSCLC tumor sections (n = 20). Data were analyzed using Pearson’s correlation test (C).

Figure 7. BACH1 increases tumor vascularity and the response to anti-VEGF therapy in xenografts.

(A) Tumor vascularity (peak enhancement) in NSG mice injected s.c. with 5 × 10⁵ BACH1^+/+ or BACH1^–/– A549 cells and administrated water (n = 9 and 6 for +/+ and –/–, respectively), 1 g/L NAC (n = 9 and 5), or 3.47 g/L VitC (n = 9 and 7) for 7 weeks. (B) Representative images of tumor vascularity from ultrasound imaging analyses. (C–E) Tumor growth in NSG mice injected s.c. with 5 × 10⁵ BACH1^OE (C and D) and BACH1^–/– (C and E) A549 cells. When tumors were palpable, the mice were injected i.p. with PBS (n = 6 in D and E) and 40 mg/kg DC101 (D, n = 7; E, n = 6) 3 times per week for 5 weeks. Tumors were measured 3–5 times per week. (F) Curves from D and E are shown in the same graph. Data indicate the mean ± SEM. Statistical significance was determined by 1-way ANOVA with Tukey’s post hoc test for multiple comparisons (A) and 2-way ANOVA with Šidák’s post hoc test for multiple comparisons (F).