Targeting Mitochondrial Metabolism in Clear Cell Carcinoma of the Ovaries

Abstract

Ovarian clear cell carcinoma (OCCC) is a rare but chemorefractory tumor. About 50% of all OCCC patients have inactivating mutations of ARID1A, a member of the SWI/SNF chromatin-remodeling complex. Members of the SWI/SNF remodeling have emerged as regulators of the energetic metabolism of mammalian cells; however, the role of ARID1A as a modulator of the mitochondrial metabolism in OCCCs is yet to be defined. Here, we show that ARID1A loss results in increased mitochondrial metabolism and renders ARID1A-mutated cells increasingly and selectively dependent on it. The increase in mitochondrial activity following ARID1A loss is associated with increase in c-Myc expression and increased mitochondrial number and reduction of their size consistent with a higher mitochondrial cristae/outer membrane ratio. Significantly, preclinical testing of the complex I mitochondrial inhibitor IACS-010759 showed it extends overall survival in a preclinical model of ARID1A-mutated OCCC. These findings provide for the targeting mitochondrial activity in ARID1A-mutated OCCCs.

Keywords: ARID1A; OCCC; mitochondria; ovarian cancer.

Figure 1

ARID1A-deficient cells upregulate OXPHOX pathways and increase mitochondrial respiration. (A) Western blot analysis for levels of ARID1A in scramble (scr.) versus ARID1A knockdown (KD) in ARID1A WT ES-2 cells. Tubulin was used as loading control. (B) GSEA data analysis of microarray transcriptomic profiling in ARID1A KD versus scramble ES-2 cells revealed that the OXPHOS pathway is among the top-most upregulated pathways following ARID1A loss. RNA-sequencing data were obtained using 3 independent samples for each cohort (scramble and KD). (C) Representative graph for enriched mitochondrial OXPHOS pathway (p = 0.000, q = 0.001). (D) Left, Top ten gene ontology (GO) biological process (p < 0.001) in ARID1A KD versus scramble ES-2 cells. (E) Genes coding for mitochondrial electron transport chain (ETC) components are upregulated in ARID1A KD versus scramble ES-2 cells. Results are from three independent experiments and are expressed as mean ± SD. (F) Amido black staining of mitochondrial (Mit.) and cytoplasmic (Cyt.) fractions isolated from scramble (scr.) and ARID1A KD (KD) ES-2 cells. (G) Western blot analysis for levels of Tom20 in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT ES-2 isolated mitochondria. Tubulin was used as a loading control. (H) Isolated mitochondrial pellets were processed in native gels following the protocol described in [47]. Shown are complexes and supercomplexes of the mitochondrial ETC.

Figure 2

ARID1A controls mitochondria metabolism in OCCC cells. (A) Western blot analysis for levels of ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT ES-2 cells. Actin was used as a loading control. (B) Real-time oxygen consumption rate (OCR) in scramble versus ARID1A KD ES-2 cells as measured by Seahorse. The injection order: a. glucose, b. oligomycin, c. FCCP, and d. rotenone and antimycin. (C) ATP-linked respiration in ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT ES-2 cells. (D) Maximal respiration in scramble versus ARID1A KD ES-2 cells. (E) Fold change of basal ATP total amount in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT ES-2 cells. (F) Western blot analysis for levels of ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1 cells. Actin was used as a loading control. (G) Real-time oxygen consumption rate (OCR) in scramble versus ARID1A KD RMG1 cells as measured by Seahorse. The injection order: a. glucose, b. oligomycin, c. FCCP, and d. rotenone and antimycin (H) ATP-linked respiration in ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1 cells. (I) Maximal respiration in scramble versus ARID1A KD RMG1 cells. (J) Fold change of basal ATP total amount in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1 cells. (K) Western blot analysis for levels of ARID1A empty vector (EV) versus ARID1A overexpressing (OE) ARID1A-mutated TOV-21G cells. Tubulin was used as a loading control. (L) Real-time oxygen consumption rate (OCR) in ARID1A empty vector (EV) versus ARID1A overexpressing (OE) ARID1A-mutated TOV-21G cells. (M) ATP-linked respiration in ARID1A empty vector (EV) versus ARID1A overexpressing (OE) ARID1A-mutated TOV-21G cells. (N) Maximal respiration in ARID1A empty vector (EV) versus ARID1A overexpressing (OE) ARID1A-mutated TOV-21G cells. The injection order: a. oligomycin, b. FCCP, and c. rotenone and antimycin. Results are from three independent experiments and are expressed as mean ± SD.

Figure 3

Loss of ARID1A results in increased mitochondrial membrane potential and increased mitochondrial content and fragmentation. (A) Representative images (original magnification 60×) of scramble (scr.) and ARID1A KD (KD) ES-2 (top four panels) and RMG1 (bottom four panels) ARID1A WT cells stained with DAPI (blue) and MitoTracker green. Merge image is also shown. (B) Quantification of MitoTracker fluorescence intensity in ES-2 (top panel) or RMG1 (bottom panel) cells under the abovementioned conditions expressed as arbitrary units (a.u). (C) Representative image of mitochondrial content in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1 cells evaluated via TOM20 staining. (D) Quantification of mitochondrial content expressed per each condition. (E) Mitochondrial DNA content in scr. versus ARID1A knockdown (KD) ARID1A WT RMG1 cells. (F) Representation of parameters used to evaluate mitochondrial morphology. (G) Representative images of TOM20 and DAPI staining in scr. versus ARID1A knockdown (KD) ARID1A WT RMG1 cells followed by 3D reconstruction of mitochondria. (H) Average mitochondrial perimeter in scr. vs. KD cells (top left), average mitochondrial area in scr. vs. KD cells (top right), average mitochondrial area/perimeter in scr. vs. KD cells (bottom left). Minor axis in scr. vs. KD cells (bottom right). Results are from three independent experiments and are expressed as mean ± SD. (I) Left, Western blot analysis for levels of ARID1A and c-Myc in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT ES-2 cells. Tubulin was used as a loading control. Right, Western blot analysis for levels of ARID1A and c-Myc in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1 cells. Tubulin was used as a loading control.

Figure 4

Loss of ARID1A results in selective sensitivity to mitochondrial inhibition in 2D. (A) Table of compounds with mitochondrial inhibitions properties in clinical trials. Compound names and conditions for which they are used are indicated [57]. (B) Dose-dependent sensitivity of scramble (scr.) and ARID1A KD (KD) ES-2 WT cells to the panel of mitochondrial inhibitors. Results are expressed as residual colony formation as compared to control. Drug treatment was conducted over a period of 7 days. (C) Western blot analysis for expression levels of ARID1A in scramble and ARID1A KD RMG1 cells obtained by lentiviral-mediated delivery of shRNA-ARID1A. Actin was used as a loading control. (D) Residual colony formation in scramble and ARID1A KD RMG1 cells exposed to 10 nM of the mitochondrial inhibitor IACS-010759 over a period of 10 days. (E) Real-time oxygen consumption rate (OCR) in scramble versus ARID1A KD RMG1 cells exposed to increasing concentrations of the mitochondrial inhibitor IACS-010759. The injection order: a. IACS-010759, b. oligomycin, c. FCCP, and d. rotenone and antimycin. Inset shows the OCR alterations after IACS-010759 injection during the first 100 min. (F) ATP-linked respiration in ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1-2 cells exposed to increasing concentrations on the mitochondrial inhibitor IACS-010759. (G) Maximal respiration in ARID1A in scramble (scr.) versus ARID1A knockdown (KD) ARID1A WT RMG1-2 cells exposed to increasing concentrations on the mitochondrial inhibitor IACS-010759.

Figure 5

Loss of ARID1A results in selective sensitivity of spheroids to mitochondrial inhibition. (A) ES-2 derived spheroids exposed to the indicated concentrations of either IACS-010759 (left) or dapagliflozin (right) over a period of 48 h. (B) Quantification of residual spheroid volume in cells exposed to IACS-010759 (top) or dapagliflozin (bottom). (C) Real-time oxygen consumption rate (OCR) in scramble versus ARID1A KD spheroids derived from ES-2 cells exposed to the indicated concentrations of IACS-010759. The injection order: a. oligomycin, b. FCCP, and c. rotenone and antimycin.

Figure 6

Preclinical testing of IACS-010759 in a xenograft model of OCCC-mutated tumor. (A) The percentage survival of mice. (B) Days the mice are alive. (C) Percentage body weight overtime. (D). Ascitic fluid volume after intraperitoneal injection of TOV21G cells followed by oral dosing of either vehicle control or indicated dose of IACS-010759 (N = 9 mice/group).