Mitochondrial Uncoupling Induces Epigenome Remodeling and Promotes Differentiation in Neuroblastoma

Abstract

The Warburg effect is the major metabolic hallmark of cancer. According to Warburg himself, the consequence of the Warburg effect is cell dedifferentiation. Therefore, reversing the Warburg effect might be an approach to restore cell differentiation in cancer. In this study, we used a mitochondrial uncoupler, niclosamide ethanolamine (NEN), to activate mitochondrial respiration, which induced neural differentiation in neuroblastoma cells. NEN treatment increased the NAD+/NADH and pyruvate/lactate ratios and also the α-ketoglutarate/2-hydroxyglutarate (2-HG) ratio. Consequently, NEN treatment induced promoter CpG island demethylation and epigenetic landscape remodeling, activating the neural differentiation program. In addition, NEN treatment upregulated p53 but downregulated N-Myc and β-catenin signaling in neuroblastoma cells. Importantly, even under hypoxia, NEN treatment remained effective in inhibiting 2-HG generation, promoting DNA demethylation, and suppressing hypoxia-inducible factor signaling. Dietary NEN intervention reduced tumor growth rate, 2-HG levels, and expression of N-Myc and β-catenin in tumors in an orthotopic neuroblastoma mouse model. Integrative analysis indicated that NEN treatment upregulated favorable prognosis genes and downregulated unfavorable prognosis genes, which were defined using multiple neuroblastoma patient datasets. Altogether, these results suggest that mitochondrial uncoupling is an effective metabolic and epigenetic therapy for reversing the Warburg effect and inducing differentiation in neuroblastoma.

Significance: Targeting cancer metabolism using the mitochondrial uncoupler niclosamide ethanolamine leads to methylome reprogramming and differentiation in neuroblastoma, providing a therapeutic opportunity to reverse the Warburg effect and suppress tumor growth. See related commentary by Byrne and Bell, p.167.

Figure 1.. NEN treatment promotes neuron differentiation and reduces MYCN expression.

(A) Left: morphological features of NB cells treated with DMSO or 1 μM NEN for 96 h (scale bar: 50 μM). Right: Quantification of neurite outgrowth with NeuronJ. (B) Immunofluorescence staining of β-tubulin III (red) and DAPI (blue) in cells treated with DMSO or 1 μM NEN for 96 h (scale bar: 25μm). (C) Relative cell proliferation with NEN treatment for 3 d. (D) Volcano plot of gene expression distribution from the RNA-seq data (n=3) of SK-N-BE(2) cells treated with 1 μM NEN for 16 h. A horizontal dashed line denotes an adjusted p-value of 0.05. Vertical dashed lines denote an absolute log₂ fold change of In(2). (E) The top 10 GO pathways enriched for NEN-upregulated genes in the DAVID analysis. (F) Left panel: MYCN mRNA levels in cells treated with 1 μM NEN for 24 h measured using RT-qPCR. Right panel: N-Myc protein levels in cells treated with NEN for the indicated time measured using immunoblots. (G) GSEA of N-Myc pathway genes from RNA-seq data in (D). Key: *, p < 0.05; **, p <0.01; ***, p < 0.001, ****, p < 0.0001 based on a Student’s two-tailed t-test. Please check the experimental replicates information in the Methods section.

Figure 2.. NEN treatment increases the NAD⁺/NADH ratio to accelerate glutaminolysis.

(A) Relative metabolite levels in SK-N-BE(2) cells treated with DMSO or 1 μM NEN measured using LC-MS. (B) ¹³C-labeling patterns of TCA cycle metabolites derived from U-¹³C-glutamine (Blue circle: ¹³C; hollow circle: ¹²C. (C) Relative L-glutamine and L-glutamate levels in SK-N-BE(2) cells treated with DMSO or 1 μM NEN for 5 h. (D) SK-N-BE(2) cells were pretreated with DMSO or 1 μM NEN for 3 h, then labeled with U-¹³C-glutamine for 2 h. (E) Left: morphological feature of cells treated with/without 3.5 mM DMKG for 96 h (scale bar: 50 μM). Right: Quantification of neurite outgrowth with NeuronJ. (F) Immunofluorescence staining for β-tubulin III (red) and DAPI (blue) in cells treated as in (E; scale bar: 25μm). (G) Cells treated with/without 3.5 mM DMKG for 3 d. All cell numbers were normalized to the control group. Key, *, p < 0.05, **, p < 0.01, ***, p < 0.001 based on a Student’s two-tailed t-test.

Figure 3.. NEN treatment remodels the epigenetic landscape.

(A, B)The regional distribution of differentially methylated probes on the Illumina MethylationEPIC array in SK-N-BE(2) cells treated with DMSO or 1 μM NEN for 24 h (n=3). (C, D) GO pathway enrichment of differentially methylated probes in promoter or gene body regions. (E) Morphological feature of cells treated with DMSO or 2 μM 5-AZA for 96 h (scale bar: 50 μM). Neurite outgrowth was quantified using NeuronJ. Key: *, p < 0.05; **, p < 0.01 based on a Student’s two-tailed t-test.

Figure 4.. NEN treatment inhibits 2-HG generation, promotes DNA demethylation, and inhibits HIF signaling under hypoxia.

(A) LC-MS-based relative metabolite levels or ratios in SK-N-BE(2) cells treated with DMSO or 1 μM NEN under hypoxia (0.5% oxygen) for 3 h or 6 h. (B) The regional distribution of differentially methylated CpG sites on the Illumina MethylationEPIC array in SK-N-BE(2) cells treated with DMSO or 1 μM NEN under hypoxia for 24 h (n=3). (C) Left: Morphological features of SK-N-BE(2) cells treated with DMSO or 1 μM NEN under normoxia or hypoxia (0.5% oxygen) for 4 d (scale bar: 50 μM). Right: Quantification of neurite outgrowth with NeuronJ. (D) The cells in (C) were trypsinized, plated in 12-well plates without treatment, and counted after 5 d. (E) Protein levels of HIFs in SK-N-BE(2) and NB16 cells treated with DMSO, 2 μM NEN, 4 μM NEN, or 3.5 mM α-KG for the indicated time under normoxia or hypoxia (0.5% oxygen). Key: ns, p > 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001 based on a one-way ANOVA test.

Figure 5.. NEN supplementation reduces NB growth in vivo.

(A) Schematic of in vivo orthotopic NB xenograft experiment. (B) LC-MS-based relative metabolite levels in tumor samples (n=5). (C, D) The mice’s tumor growth, survival, and body weight curves. (E) H&E-stained tumors in control and NEN groups stained for N-Myc and β-catenin (scale bar: 50 μm); arrows indicate prominent enlarged nucleolar formations. Key: ns, p > 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001 based on a Student’s two-tailed t-test.

Figure 6.. NEN treatment induces gene expression profile changes that indicate a favorable prognosis in NB patients.

GO enrichment of (A) overlapping favorable prognosis genes (p < 0.05; >1000 genes) and (B) overlapping unfavorable prognosis genes (p < 0.05; >1000 genes) from seven available NB datasets submitted for DAVID analysis. (C) GSEA enrichment in RNA-seq gene expression data (n=3). The gene sets (favorable or unfavorable prognosis; p < 0.05) were created from 10 available NB datasets in the R2 database. (D) Representative GSEA analysis plot from (C).