N6-methyladenosine-induced ERRγ triggers chemoresistance of cancer cells through upregulation of ABCB1 and metabolic reprogramming

Abstract

Background: Drug resistance severely reduces treatment efficiency of chemotherapy and leads to poor prognosis. However, regulatory factors of chemoresistant cancer cells are largely unknown. Methods: The expression of estrogen receptor related receptors (ERRs) in chemoresistant cancer cells are checked. The roles of ERRγ in chemoresistance are confirmed by in vitro and in vivo studies. The mechanisms responsible for ERRγ-regulated expression of ABCB1 and CPT1B are investigated. Results: The expression of ERRγ is upregulated in chemoresistant cancer cells. Targeted inhibition of ERRγ restores the chemosensitivity. ERRγ can directly bind to the promoter of ABCB1 to increase its transcription. An elevated interaction between ERRγ and p65 in chemoresistant cells further strengthens transcription of ABCB1. Further, ERRγ can increase the fatty acid oxidation (FAO) in chemoresistant cells via regulation of CPT1B, the rate-limiting enzyme of FAO. The upregulated ERRγ in chemoresistant cancer cells might be due to increased levels of N6-methyladenosine (m⁶A) can trigger the splicing of precursor ESRRG mRNA. Conclusions: m⁶A induced ERRγ confers chemoresistance of cancer cells through upregulation of ABCB1 and CPT1B.

Keywords: CPT1B; ABCB1; Chemoresistance; ERRγ; FAO.

Figure 1

ERRγ is upregulated in chemoresistant cancer cells. (A&B) Expression of ERRα (ESRRA) and ERRγ (ESRRG) measured in MCF-7/ADR (A), HepG2/ADR (B), and their corresponding parental cells by qRT-PCR; (C) Protein levels of ERRγ in MCF-7/ADR, HepG2/ADR, and their corresponding parental cells measured by Western blot analysis (left) and quantitatively analyzed (right); (D) Subcellular expression of ERRγ in HepG2/ADR and HepG2 cells visualized by confocal imaging; (E) The subcellular expression of ERRγ in HepG2/ADR and HepG2 cells was checked by Western blot analysis (left) and quantitatively analyzed (right); (F) Cells were treated with Dox (2 μM) for the indicated times, then the protein expression of ERRγ was checked by Western blot analysis (left) and quantitatively analyzed (right); (G) Cells were treated with Dox (2 μM) for the indicated times, then the mRNA expression of ERRγ was checked by qRT-PCR. Data were presented as means ± SD from three independent experiments. *p<0.05, **p< 0.01 compared with control.

Figure 2

ERRγ regulates chemoresistance of cancer cells. (A&B) Cell proliferation rate in si-NC- or si-ERRγ-1-transfected HepG2/ADR cells for 24 h and followed by treatment with increasing concentrations of Dox (A) or Tax (B) for 48 h; (C&D) Cell proliferation rate in si-NC- or si-ERRγ-1-transfected MCF-7/ADR cells for 24 h and followed by treatment with increasing concentrations of Dox (C) or Tax (D) for 48 h; (E&F) HepG2/ADR (E) or MCF-7/ADR (F) cells transfected with scrambled shRNA or sh-ERRγ were split and cultured in fresh medium for the next 15 days. The colonies were fixed with methanol/glacial acetic acid (7:1) and stained with 0.1% of crystal violet; (G) Tumor volume measurement in mouse xenografts. HepG2/ADR cells stably transfected with scrambled shRNA or sh-ERRγ were subcutaneously inoculated in nude mice. We randomly divided the mice into Scramble, sh-ERRγ, Dox + Scramble and Dox + sh-ERRγ groups and treated them as described in the Methods. Tumor growth curves were constructed based on the tumor volumes measured in the mice; (H) IHC analysis of mouse xenograft tissues. Expression of ERRγ and proliferation marker Ki-67 was determined in tumor tissue sections from the xenografts using IHC (scale bar, 50 μm) and quantitatively analyzed; Data were presented as means ± SD from three independent experiments. **p< 0.01. NS, no significant.

Figure 3

P-gp is involved in ERRγ-regulated chemoresistance of cancer cells. (A&B) mRNA expression of ABC transporters measured in HepG2/ADR (A) or MCF-7/ADR (B) cells 24 h post transfection with si-NC or si-ERRγ-1; (C) Expression of P-gp protein measured by Western blot analysis (left) and quantitively analyzed (right) in HepG2/ADR or MCF-7/ADR cells 24 h post transfection with si-NC or si-ERRγ-1/2; (D) Expression of P-gp protein measured by Western blot analysis (left) and quantitively analyzed (right) in HepG2 or MCF-7 cells 24 h post transfection with vector control or pcDNA/ERRγ; (E) P-gp function analyzed by flow cytometric measurement of the intracellular accumulation of Rh123 in HepG2/ADR or MCF-7/ADR cells 24 h post transfection with scrambled siRNA or si-ERRγ-1; (F) IC₅₀ values of Dox in HepG2/ADR or MCF-7/ADR cells co-transfected with si-ERRγ and P-gp construct. Data were presented as means ± SD from three independent experiments. *p< 0.05, **p< 0.01. NS, no significant.

Figure 4

ERRγ interacts with p65 to regulate ABCB1 transcription. (A) Schematic representation of ERREs in the promoter region of ABCB1 with changes of nucleotides in ERRE1 and ERRE2 shown as indicated; (B) ChIP-PCR assay showing ERRγ binding to ERRE1 and ERRE2 in ABCB1 promoter. The input (5%), binding between ERRγ and the promoter of ABCB1 at the potential binding site ERRE1/2, was amplified by qPCR (right) and confirmed by 2% agarose gel electrophoresis (left); (C) Schematic representation of mutated ERRE positions in pGL-ABCB1 vector; (D) Reporter gene assay performed in HepG2 cells 24 h post transfection with pGL-ABCB1-WT or pGL-ABCB1-Mut1/2/3 by dual-luciferase analysis; (E) Examination of ERRγ interaction with different transcription factors in HepG2/ADR and MCF-7/ADR cells following immunoprecipitation with ERRγ or control antibody and analyzed by Western blot analysis; (F) Interaction between ERRγ and p65 in HepG2 and HepG2/ADR cells monitored by immunoprecipitation using anti-ERRγ antibody; After ERRγ was immunoprecipitated, the binding between ERRγ and p65 was examined by Western blot analysis. An equal amount of ERRγ was loaded for normalization according to a pre-Western blot; (G) Expression and localization of p65 (green) and ERRγ (red) in HepG2 and HepG2/ADR cells visualized by confocal imaging; (H) Interaction between ERRγ and p65 in HepG2/ADR cells treated with or without BAY 11-7082 for 12 h and then analyzed by immunoprecipitation using an antibody against ERRγ; (I) Dual-luciferase reporter gene assay performed in HepG2 cells transfected with pGL-ABCB1-WT or pGL-ABCB1-Mut1/2, with or without pcDNA/ERRγ, for 12 h and then further treated with or without BAY 11-7082 for 12 h; (J) HepG2/ADR cells were treated with si-RNA or si-ERRγ combined with or without BAY 11-7082 for 12 h and then further treated with 5 μM Dox for 48 h; (K) Model for ERRγ/p65-promoted transcription of ABCB1 in chemoresistant cancer cells. Data were presented as means ± SD from three independent experiments. **p< 0.01. NS, no significant.

Figure 5

ERRγ dictates the metabolic reprogramming in chemoresistant cancer cells. (A~D) Relative glucose consumption (A), lactate production (B), PDH activity (C), and ATP levels (D) measured in HepG2/ADR or MCF-7/ADR cells following comparison with those measured in their corresponding parental cells; (E) OCR in HepG2/ADR and HepG2 cells measured by Seahorse XF24 analyzer; (F) Relative glucose consumption, lactate production, and mitochondrial mass in HepG2/ADR cells transfected with sh-Con or sh-ERRγ; (G) Relative ATP levels in HepG2/ADR cells transfected with sh-Con or sh-ERRγ; (H) OCR in HepG2/ADR cells transfected with sh-Con or sh-ERRγ measured by Seahorse XF24 analyzer; (I~J) Basal (I) and maximal (J) OCR measured by Seahorse XF24 analyzer in HepG2/ADR or HepG2 cells transfected with sh-Con or sh-ERRγ. Data were presented as means ± SD from three independent experiments. *p< 0.05, **p< 0.01.

Figure 6

ERRγ regulates the FAO via Cpt1b in chemoresistant cancer cells. (A~B) Relative FA uptake (A) and FA β oxidation rate (B) between HepG2 and HepG2/ADR cells; (C~D) Relative FA uptake (C) and FA β oxidation rate (D) in cells transfected with sh-Con or sh-ERRγ; (E) Cell proliferation measured by CCK-8 kit in HepG2 cells pre-transfected with vector control or pcDNA/ERRγ for 6 h and then treated with or without Dox (1 μM) combined with or without ETO for 24 h; (F) mRNA levels of FAO-related genes measured by qRT-PCR in HepG2 and HepG2/ADR cells; (G) mRNA levels of FAO-related genes measured by qRT-PCR in HepG2/ADR cells transfected with sh-Con or sh-ERRγ; (H) Cell proliferation measured by CCK-8 kit in HepG2/ADR cells pre-transfected with sh-Con or sh-ERRγ and then transfected with vector or a Cpt1b expression construct, followed by further treatment with Dox (10 μM) for 24 h; (I~J) ATP (I) and FA β oxidation rate (J) measured in HepG2/ADR cells transfected with sh-Con or sh-ERRγ with further transfection with vector or a Cpt1b expression construct for 24 h; (K) Nucleotide sequences of ERREs in CPT1B and the mutated (GACCTTG to AGAACCG) nucleotides in pGL3-CPT1B-Mut-Luc vector; (L) ChIP assay measuring ERRγ binding to CPT1B promoter in both HepG2 and HepG2/ADR cells; (M) Dual-luciferase reporter gene assay performed in HepG2 and HepG2/ADR cells transfected with pGL3-CPT1B-WT-Luc or pGL3-CPT1B-Mut-Luc; (N) Dual-luciferase reporter gene assay performed in HepG2 cells transfected with pGL-ABCB1-WT or pGL3-CPT1B-Mut reporter with or without pcDNA/ERRγ for 12 h and then further treated with or without BAY 11-7082 for 12 h; (O) Model for ERRγ-regulated FAO via Cpt1b in chemoresistant cancer cells. Data were presented as means ± SD from three independent experiments. *p< 0.05.

Figure 7

The m⁶A-facilitated splicing is responsible for the upregulation of ERRγ. (A) The m⁶A/A ratio of total mRNA in HepG2/ADR and MCF-7/ADR cells were determined by LC-MS/MS and compared with that in their parental cells; (B) The expression of Mettl3 and ALKBH5 in HepG2/ADR and HepG2 cells was checked by western blot analysis and quantitatively analyzed; (C) m⁶A RIP-qPCR analysis of ERRγ mRNA in HepG2/ADR and HepG2 cells; (D) The expression of ERRγ in HepG2/ADR and MCF-7/ADR cells transfected with sh-Con or sh-Mettl3 was checked by western blot analysis and quantitatively analyzed; (E) HepG2/ADR cells transfected with sh-Con or sh-Mettl3 were further treated with increasing concentrations of Dox, the cell proliferation was tested by CCK-8 kit; (F~G) The mature (E) and precursor (F) mRNA of ERRγ in HepG2/ADR cells transfected with sh-Con or sh-Mettl3 were checked by qRT-PCR; (H) HepG2/ADR cells transfected with sh-Con or sh-Mettl3 were pre-treated with Act-D for 90 min, then the precursor mRNA of ERRγ was checked by qRT-PCR; (I) The promoter activity of pGL-ESRRG-Basic in HepG2/ADR cells transfected with sh-Con or sh-Mettl3 was checked by dual luciferase assay; (J) HepG2 cells were transfected with sh-Con, sh-Mettl3, vector control or ERRγ construct alone or together for 24 h, the expression of targets was measured and quantitatively analyzed. Data were presented as means ± SD from three independent experiments. **p< 0.01. NS, no significant.

Figure 8

The m⁶A/ ERRγ axis and in vivo cancer progression. (A) IHC (ERRγ and Mettl3)-stained paraffin-embedded sections obtained from HepG2 and HepG2/ADR xenografts when the tumor volumes were about 100 mm³ for each group; (B) The sh-control and sh-Mettl3 HepG2/ADR cells were subcutaneously inoculated in nude mice. IHC (P-gp and Cpt1b)-stained paraffin-embedded sections obtained at the end of experiment; (C) Tumor volume measurement in mouse xenografts. HepG2/ADR cells stably transfected with sh-ERRγ were subcutaneously inoculated in nude mice. We randomly divided the mice into sh-ERRγ, sh-ERRγ + Elacridar, sh-ERRγ + Etomoxir, and sh-ERRγ + Elacridar + Etomoxir and then treated with Dox as described in the Methods. Tumor growth curves were constructed based on the tumor volumes measured in the mice; (D) Expression of ESRRG in HCC tumor tissues and normal liver tissues from Oncomine database (Guichard and TCGA liver cancers); (E) ESRRG expression in liver cancers of T1 (n=153), T2 (n=77), and T3 (n=65) stages from TCGA database; (F) Expression of Mettl3 in HCC tumor tissues and normal liver tissues from Oncomine database (Roessler liver); (G) Correlation between ESRRG and ABCB1 in liver cancer patients (n=371) from TCGA database; Data were presented as means ± SD from three independent experiments. *p<0.05, **p< 0.01.