Article Figures & Data
Figures
- Figure 1. Snail augments ATP levels via fatty acid metabolism promoting cancer cell survival under glucose starvation.
(A) Clonogenic survival assay of cancer cells following glucose starvation as described in the Materials and Methods section (left). Colonies of more than 50 cells were counted after crystal violet staining (right). Data are expressed as means and SD. The double asterisks denote P < 0.01, one asterisk denoting P < 0.05 (n = 5, means ± SD, t tests). (B) The cancer cells expressing control-shRNA or Snail-shRNA were incubated in the presence (Glc+) or absence (Glc−) of glucose for 4 h, and the relative ATP levels were measured (n = 3, means ± SD, t tests). (C) The breast cancer cells in the absence of glucose (Glc−) were treated with BSA-Palmitate (100 μM) in combination with DMSO control or ETX (100 μM) for 4 h, and the relative ATP levels were measured (n = 3, means ± SD, t tests). (D) The MDA-MB-231 cells expressing control-shRNA or Snail-shRNA were treated with 13C-palmitate (100 μM) in the absence of glucose for 4 h. Mass isotopomer distribution of [U-13C]-palmitate–derived carbon into some TCA metabolites was determined by LC–MS. Filled blue circles represent 13C atoms derived from [U-13C]-palmitate (*P < 0.05, **P < 0.01, ***P < 0.001, t tests). (E) The enzymatic activities of ACADVL, ACADM, and HADHA in breast cancer cells expressing shRNA for control (shControl) or Snail (shSnail) were measured under glucose-starved condition (0.5 mM glucose, n = 3, means ± SD, t tests). (F) The CPT1 activities of breast cancer cells expressing shRNA for control (shControl) or Snail (shSnail) under glucose-starved condition were measured (0.5 mM glucose, n = 3, means ± SD, t tests). (G) Snail was induced by treatment of doxycycline (Dox) for 48 h and ATP (left), NADPH (middle), and cell death (right) in starved condition were measured in combination with CPT1 inhibitor etomoxir (ETX, 100 μM, n = 3, means ± SD, t tests). (H) Mitochondrial oxygen consumption rate (OCR) in Tet-inducible Snail in combination with ETX (n = 3, means ± SD, t tests). (I) Malonyl-CoA abundances in breast cancer cells expressing shRNA for control (shControl) or Snail (shSnail) were measured (n = 3, means ± SD, t tests). (J) A schematic diagram depicting a potential mechanism by which the Snail regulates fatty acid oxidation (FAO).
- Figure S1. Epithelial–mesenchymal transition inducer Snail augments fatty acid oxidation.
(A) Clonogenic capacity under glucose-deprived condition followed by refreshment of normal culture medium. Immunoblot analysis (left) and clonogenic capacity (middle and right) were measured. (B) The cancer cells expressing control-shRNA or independent set of Snail-shRNA-2 were incubated in the presence of glucose (Glc+) or absence (Glc−) for 2 h, and the relative ATP levels were measured (n = 3, means ± SD) (left). Immunoblot analysis shows endogenous Snail protein abundance (right). (C) The cancer cells expressing Tet-inducible Snail (Dox+) were incubated in the presence of glucose (Glc+) or absence (Glc−) for 2 h, and the relative ATP levels were measured (n = 3, means ± SD). (D) The enzymatic activities of ACADVL, ACADM, and HADHA in breast cancer cells expressing Tet-inducible Snail (Dox+) were measured under glucose-starved condition (0.5 mM glucose, n = 3, means ± SD). (E) The CPT1 activities of breast cancer cells expressing Tet-inducible Snail (Dox+) under glucose starved condition were measured (0.5 mM glucose, n = 3, means ± SD). (F) Relative oxygen consumption rate (OCR) normalized to protein abundance over time in control (Dox−, n = 5) and inducible Snail (Dox+, n = 5) in MCF-7 (left) and MDA-MB-231 cells (right). (G) Malonyl-CoA abundances of breast cancer cells expressing Tet-inducible Snail for control (Dox−) or Snail overexpression (Dox+) were measured (n = 3, means ± SD). (H) Lipid (left) and free fatty acid synthesis (right) after inducible Snail for control (Dox−) or Snail overexpression (Dox+). Upper panels represent lipid droplets under Oil Red O staining. Scale bar, 100 μm.
- Figure 2. Transcript abundance of mitochondrial ACC2 is suppressed in human cancer samples as a target of Snail repressor.
(A) ACC2 (ACACB), but not ACC1 (ACACA), transcript abundance in tumor tissue (red) was suppressed compared with adjacent normal tissue (blue) in various cancer types. The Cancer Genome Atlas dataset included breast cancer (BRCA), colorectal adenocarcinoma (COADREAD), head and neck squamous cell carcinoma (HNSC), pan-kidney cohort (KIPAN), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD), stomach adenocarcinoma (STAD), and thyroid carcinoma (THCA). The adjusted P-value of various cancer types were determined by the false discovery rates (FDRs) (Benjamini–Hochberg) method (right panel). (B) Relative transcript (left) and protein (right) abundance of ACC2 after knockdown of Snail (shSnail) in breast cancer cells (n = 3, means ± SD, t tests). (C) Schematic diagram showing positions of potential Snail-binding canonical E-boxes on the ACC2 proximal promoter region and its reporter constructs of wild type or mutant E-boxes. (D) Fold increase of reporter activities in combination with wild type or mutated ACC2 promoter following shRNA-mediated Snail knockdown compared with each control shRNA in breast cancer cells (n = 3, means ± SD, t tests). (E) ChIp-enriched DNA was determined by qRT-PCR using specific primers complementary to the promoter region containing E-box of ACC2 and a positive control PFKP (n = 3, means ± SD, t tests).
- Figure S2. ACC2 and Snail are inversely correlated in clinical samples.
(A) Unsupervised clustering of a 973 breast cancer patient samples in The Cancer Genome Atlas (TCGA) using ACC2 (ACACB) and Snail (SNAI1) to distinguish between subsets of tumors with p53 mutational status and breast cancer subtypes (left panel). Comparison of ACC2 and Snail transcript levels between subsets of p53 mutational status in breast cancer patients. (B) Comparison of Snail (left panel) and ACC2 (right panel) transcript levels according to breast cancer subtypes.
- Figure S3. ACC2 (ACACB) is a target of Snail repressor.
(A) Relative protein abundance of ACC2 after knockdown of Snail (shSnail-2) in breast cancer cells (left). Relative transcript abundance of ACC1 (ACACA) after knockdown of Snail (shSnail) in breast cancer cells (right). (B) Relative transcript (left) and protein (right) abundance of ACC2 after inducible overexpression of Snail (Dox+) in breast cancer cells. (C) Fold suppression of reporter activities in combination with wild-type (WT) or mutated ACC2 promoter after Snail overexpression compared with each control empty vector in breast cancer cells.
- Figure 3. Suppression of ACC2 increases CPT1 activity and fatty acid oxidation, providing pro-survival under metabolic stress.
(A) Relative malonyl-CoA abundance according to ACC2 knockdown (left). ACC2 abundance was determined by immunoblot analysis after transduction of inducible shRNA (right). The double asterisks denote P < 0.01, one asterisk denoting P < 0.05 (n = 3, means ± SD, t tests). (B) Fatty acid oxidation enzymatic activities under glucose-starved condition (0.5 mM) were determined after inducible knockdown (Dox+) of ACC2 compared with control (Dox−). (C) Knockdown of ACC2 increases CPT1 activity in breast cancer cells (n = 3, means ± SD, t tests). (D) ACC2 was knocked down by treatment of doxycycline (Dox) for 48 h, and ATP levels were measured with glucose (Glc+) or glucose-starved (Glc−) condition (n = 3, means ± SD, t tests). (E) ACC2 was knocked down by treatment of doxycycline (Dox) for 48 h, and NADP+/NADPH ratio was determined (n = 3, means ± SD, t tests). (F) Cells were cultured under glucose-starved condition for a 48-h period, and cell death was quantitated by trypan blue exclusion assay (n = 3, means ± SD, t tests). (G) Clonogenic survival assay of cancer cells after glucose starvation for a 72-h period. Colonies of more than 50 cells were counted after crystal violet staining (n = 5, means ± SD, t tests). (H) MDA-MB-231 cells (1 × 106) expressing dsRed (n = 3) or ACC2 (n = 3) were injected orthotopically into the mammary fat pads of nude mice. Tumor initiation and volume were monitored biweekly. Results are shown as means and SEM. Asterisks, P < 0.01 compared with the control by Mann–Whitney test. (I) The MDA-MB-231-luc-D3H2LN cells (5 × 105 cells) either of control (dsRed, n = 3) or of ACC2 overexpression (ACC2, n = 3) were injected intravenously into tail veins of immunodeficient mice. The number of lung metastatic nodules at day 28 was counted under microscopic examination (left). Whole-field images of representative lungs showed the median value for each group. Statistical significance was determined by Mann–Whitney test (right). Arrows indicate metastatic tumor foci in mouse lung. Scale bar, 2 mm. (J) Kaplan–Meier survival graphs for all patients with breast cancer (left) or for those with wild-type p53 (middle) or p53 mutant (right) cancers, on the basis of ACC2 mRNA transcript abundance at an optimal threshold indicated by percentile numbers. Samples with decreased ACC2 expression are represented with blue lines. A log-rank test was used to calculate statistical significances.
- Figure S4. ACC2 downstream of Snail plays a critical role in fatty acid oxidation.
(A) Relative malonyl-CoA abundances of breast cancer cells expressing vector control (dsRed) or ACC2 (n = 3, means ± SD). (B) The protein abundance (left) and relative CPT1 activities (right) of breast cancer cells transfecting control or ACC2 siRNA (siACC2). (C) The cancer cells expressing control or ACC2-siRNA (siACC-2) were incubated in the presence of glucose (Glc+) or absence (Glc−) for 2 h and the relative ATP levels were measured (n = 3, means ± SD). (D) The cancer cells expressing control (ACC2−) or ACC2 were incubated in the presence of glucose (Glc+) or absence (Glc−) for 2 h and the relative ATP levels were measured (n = 3, means ± SD). (E) Clonogenic survival assay of breast cancer cells expressing control vector (ACC2−) or ACC2 after glucose starvation as described in the Materials and Methods section (left). Colonies of more than 50 cells were counted after crystal violet staining (right). (F) Clonogenic survival of breast cancer cells against paclitaxel treatment as indicated by concentration. The ACC2 was inducible knocked down with doxycycline (Dox+) for 48 h before paclitaxel treatment (left). The colony number was determined by stereomicroscopic examination under high power field (HPF) as described in the Materials and Methods section (right) (n = 3, means ± SD).
- Figure 4. Snail-ACC2 axis controls fatty acid oxidation and NADPH homeostasis.
(A) Relative ATP level in breast cancer cells transduced with shRNA control (shControl) or with Snail shRNA (shSnail). ACC2 was knocked down by treatment with doxycycline (Dox+) for 48 h in Snail shRNA cells under either nourished condition (Glc+) or glucose-starved condition (Glc−). The double asterisks denote P < 0.01, one asterisk denoting P < 0.05 (n = 3, means ± SD, t tests). (B, C, D) Inducible knockdown of ACC2 rescued metabolic reprogramming by lack of Snail. The fatty acid oxidation activity (B), clonogenic capacity (C), and NADP+/NADPH ratio (D) after glucose starvation were measured. Data are means ± SD from n = 3 for (B, C), from n = 5 (D).
- Figure S5. Metabolic reprogramming by Snail-ACC2 axis.
(A) Relative ATP levels in breast cancer cells with inducible expression of Snail (Dox+) for 48 in combination with ACC2 induction under either nourished condition (Glc+) or glucose starved condition (Glc−). (B, C, D) Overexpression of ACC2 rescued Snail-induced metabolic rewiring of fatty acid oxidation activities (B), clonogenic capacity (C), and NADPH (D) (n = 3, means ± SD).
- Figure 5. Snail augments catabolic metabolism via activation of pentose phosphate pathway and fatty acid oxidation (FAO).
(A) A schematic diagram depicting a potential catabolic mechanism by which the Snail regulates pentose phosphate pathway flux and FAO activity. Open arrows denote metabolic outcomes of catabolic metabolism regulated by increased Snail abundance in cancer cells. (B) Relative ATP levels (left) and immunoblot (right) in breast cancer cells according to ACC2 expression in combination with inducible glucose-6-phosphate dehydrogenase (G6PD) knockdown (Dox) under physiologic glucose concentration (5.5 mM, n = 3, means ± SD, t tests).
- Figure 6. Combined pharmacological inhibition of pentose phosphate pathway and CPT1 synergistically suppresses cancer progression.
(A) A glucose-6-phosphate dehydrogenase inhibitor DHEA (20 μg/ml) in combination with CPT1 inhibitor etomoxir (ETX, 200 μM) was treated for a 16-h period. The NADP+/NADPH ratio (left) and ATP levels (right) were determined from breast cancer cells under physiologic glucose concentration (5.5 mM). The double asterisks denote P < 0.01, one asterisk denoting P < 0.05 (n = 3, means ± SD, t tests). (B) Live cell density of breast cancer cells after treatment of DHEA (20 μg/ml) in combination with etomoxir (ETX, 200 μM) for a 48-h period under physiologic glucose concentration (n = 5, means ± SD, t tests). (C) Clonogenic survival of breast cancer cells under glucose-deprived condition (0.5 mM) in combination with DHEA (6.25 μg/ml) and/or etomoxir (ETX, 100 μM) followed by refreshment of normal culture medium (left). The colony number (right) was determined by stereomicroscopic examination as described in the Materials and Methods section (n = 5, means ± SD). (D) MDA-MB-231 cells (1 × 106) were orthotopically injected into the mammary fat pads of either vehicle (n = 10), or oral administration of DHEA (100 mg/kg, n = 10), or intraperitoneal administration of etomoxir (ETX, 50 mg/kg, n = 10), or a combination of DHEA and ETX (n = 9). The drugs were given five times a week and tumor growth was measured twice a week (means ± SEM). Two asterisks denote P < 0.01 by Mann–Whitney test. (E) Lung metastasis by tail vein injection of MDA-MB-231-D3H2LN cells (5 × 105). The mice were administrated by either vehicle (n = 8), or oral administration of DHEA (100 mg/kg, n = 8), or intraperitoneal administration of etomoxir (ETX, 50 mg/kg, n = 8), or a combination of DHEA and ETX (n = 8). The number of lung metastatic nodules at day 28 was counted under microscopic examination (left), and statistical significance was determined by Mann–Whitney test. Whole-field images of representative lungs showing median value for each group (right). Arrows indicate metastatic tumor foci in mouse lung. Scale bar, 1 mm.
- Figure S6. Combined inhibition of pentose phosphate pathway and fatty acid oxidation intervenes catabolic rewiring of cancer cells.
(A) Chemical structure of CB83 as a specific inhibitor of glucose-6-phosphate dehydrogenase (G6PD) (left). Molecular docking studies of CB83 with the hG6PD (middle and right). Surface representation of the substrate-binding pocket of G6PD with docked conformation of CB83 ligand (middle), and key residues involved in the ligand–protein interactions (right) are shown as cartoon model. Protein surface is colored in light blue, residues and CB83 are shown as yellow and green sticks, respectively. The red-dashed lines represent hydrogen bonds. Figures were drawn using PyMol (Delano Scientific LLC). (B) A specific G6PD inhibitor CB83 (2.5 μM) in combination with a CPT-1 inhibitor etomoxir (ETX, 200 μM) was treated for a 16-h period. The NADP+/NADPH ratio (left) and ATP levels (right) were determined from breast cancer cells under physiologic glucose concentration (5.5 mM). (C) Live cell density of breast cancer cells according to treatment of CB83 (2.5 μM) in combination with etomoxir (ETX, 200 μM) for a 48-h period under physiologic glucose concentration. (D) Clonogenic survival of breast cancer cells under glucose-deprived condition (0.5 mM) in combination with CB83 (1.25 μM) and/or etomoxir (ETX, 100 μM) followed by refreshment of normal culture medium (left). The colony number (right) was determined by stereomicroscopic examination as described in the Materials and Methods section (n = 5, means ± SD). (E) Gross pictures of tumor growth described in Fig 6D (left) and body weight change (right) during DHEA (100 mpk) and/or ETX (50 mpk) administration.
Source data are available for this figure.
Source Data for Figure S6[LSA-2020-00683_SdataFS6.xls]
Supplementary Materials
Supplemental Data 1.
Supplemental Data 2.
