EDD enhances cell survival and cisplatin resistance and is a therapeutic target for epithelial ovarian cancer

Abstract

The E3 ubiquitin ligase EDD is overexpressed in recurrent, platinum-resistant ovarian cancers, suggesting a role in tumor survival and/or platinum resistance. EDD knockdown by small interfering RNA (siRNA) induced apoptosis in A2780ip2, OVCAR5 and ES-2 ovarian cancer cells, correlating with loss of the prosurvival protein myeloid cell leukemia sequence 1 (Mcl-1) through a glycogen synthase kinase 3 beta-independent mechanism. SiRNA to EDD or Mcl-1 induced comparable levels of apoptosis in A2780ip2 and ES-2 cells. Stable overexpression of Mcl-1 protected cells from apoptosis following EDD knockdown, accompanied by a loss of endogenous, but not exogenous, Mcl-1 protein, suggesting that EDD regulated Mcl-1 synthesis. Indeed, EDD knockdown induced a 1.87-fold decrease in Mcl-1 messenger RNA and EDD transfection enhanced murine Mcl-1 promoter-driven luciferase expression 5-fold. To separate EDD survival and potential cisplatin resistance functions, we generated EDD shRNA stable cell lines that could survive initial EDD knockdown and showed that these cells were 4- to 21-fold more sensitive to cisplatin. Moreover, transient EDD overexpression in COS-7 cells was sufficient to promote cisplatin resistance 2.4-fold, dependent upon its E3 ligase activity. In vivo, mouse intraperitoneal ES-2 and A2780ip2 xenograft experiments showed that mice treated with EDD siRNA by nanoliposomal delivery [1,2-dioleoyl-sn-glycero-3-phophatidylcholine (DOPC)] and cisplatin had significantly less tumor burden than those treated with control siRNA/DOPC alone (ES-2, 77.9% reduction, P = 0.004; A2780ip2, 75.9% reduction, P = 0.042) or control siRNA/DOPC with cisplatin in ES-2 (64.4% reduction, P = 0.035), with a trend in A2780ip2 (60.3% reduction, P = 0.168). These results identify EDD as a dual regulator of cell survival and cisplatin resistance and suggest that EDD is a therapeutic target for ovarian cancer.

Fig. 1.

EDD is overexpressed in ovarian cancer cell lines and EDD knockdown induces apoptosis. (A) EDD expression was determined by immunoblotting lysates from ovarian cell lines. (B) A2780ip2, (C) OVCAR5 and (D) ES-2 cells were transfected with control siRNA or one of two siRNAs to EDD. After transfection for the indicated time, floating and adherent cells were harvested and cell lysates were immunoblotted for EDD expression and PARP. Uncleaved (Un) and cleaved (Clv) PARP are indicated with arrows. (E) Cells were transfected with control siRNA or EDD siRNA1 for 48h and floating and adherent cells were stained with propidium iodide. Flow cytometry was used to determine the percentage of cells with sub-2n DNA content, an indicator of apoptosis. The results are from three independent experiments. (F) A2780ip2 cells were transfected with control siRNA or EDD siRNA1 for 12 or 24h and lysates from floating and adherent cells were immunoblotted for EDD expression and PARP cleavage. (G) ES-2 cells were transfected with control siRNA or EDD siRNA1 for 24 or 48h and lysates from floating and adherent cells were immunoblotted for EDD expression and PARP cleavage. (H) A2780ip2 cells were treated with 50 μg/ml of cycloheximide for the indicated time. Cell lysates were immunoblotted for EDD and actin. The number under each lane indicates the relative intensity of the EDD band compared with actin, with the amount in time zero set at 1.

Fig. 2.

EDD downregulation decreases Mcl-1 protein levels, whereas Mcl-1 overexpression inhibits apoptosis upon EDD knockdown. (A) Cells were either untreated (none) or transfected with control or EDD siRNA1 for 24h. Lysates from floating and adherent cells were immunoblotted for EDD, PARP and Bcl2 family members as indicated. (B) A2780ip2 and ES-2 cells were untreated (none) or transfected with control siRNA or EDD siRNA1 and simultaneously treated with either Q-VD-OPH pan caspase inhibitor (+) or the negative control Z-FA-FMK (−). After 24h, floating and adherent cells were collected, lysed, run on sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotted for EDD, PARP, Mcl-1, Bcl-xL, p53 upregulated modulator of apoptosis and actin. (C) A2780ip2 and (D) ES-2 cells were either untreated (none) or were transfected with the control siRNA, EDD siRNA1 or siRNA to Mcl-1 or Bcl-xL for 24h. Floating and adherent cells were fixed, stained with propidium iodide and the percentage of sub-2n cells determined by flow cytometry. P values represent significance compared with the control siRNA-transfected cells. (E) Cells were transfected with siRNA as in (D) for 24h. Lysates from floating and adherent cells were immunoblotted to confirm knockdown and to determine PARP cleavage. (F) Stable populations of ES-2 cells and (G) stable clones of A2780ip2 cells expressing either pBabe vector (Vec) or pBabe-Flag-Mcl-1 (Mcl-1) were generated by retroviral transduction. Cell lysates were immunoblotted as indicated. Arrows indicate endogenous Mcl-1 and the slower-migrating Flag-Mcl-1. (H) Stable ES-2 or (I) A2780ip2 cells were transfected with control or EDD siRNA1 for 24h and cell lysates immunoblotted as indicated. Arrows indicate endogenous Mcl-1 and Flag-Mcl-1.

Fig. 3.

EDD regulates Mcl-1 levels through transcriptional regulation. (A) EDD knockdown inhibits Mcl-1 mRNA expression. ES-2 and A2780ip2 cells were transfected with EDD siRNA1 for 24 or 12h, respectively, and RNA was harvested. Quantitative real-time PCR was performed using Mcl-1-specific primers. The y-axis represents the fold change in Mcl-1 mRNA in EDD siRNA1-transfected cells compared with that in control siRNA-transfected cells. The results are a combination of three independent experiments. (B) EDD activates the Mcl-1 promoter. HeLa cells were transfected with p(−2389/+10)mcl-luc, an Mcl-1 promoter-driven firefly luciferase plasmid (16), TK Renilla luciferase and either Flag-EDD, Flag-EDD-C2768A or empty vector. Cells were harvested at 48h and firefly luciferase activity was normalized to Renilla luciferase activity in each sample. P values indicate significance (P < 0.05) within a group between Flag-EDD- and vector-transfected cells. These results are a combination of four independent experiments. (C) Western blot of Flag-EDD from (B).

Fig. 4.

Stable EDD knockdown increases cisplatin sensitivity. (A) ES-2, (B) A2780ip2 and (C) OVCAR5 cells were retrovirally transduced with control or one of three EDD shRNAs and clones (ES-2 and A2780ip2) or populations (OVCAR5) were selected. Cell lysates were immunoblotted for EDD expression. Multiple clones from the same shRNA are designated as A or B. (D) ES-2 control shRNA or EDD shRNA cells were treated with cisplatin for 72h and cell viability measured by MTS assay. Percent survival was plotted against the log of the cisplatin concentration. The results are from three independent experiments performed in quadruplicate. (E) A2780ip2 and (F) OVCAR5 shRNA cells were treated with cisplatin for 24h and cell lysates from floating and adherent cells were immunoblotted for EDD and PARP cleavage. The numbers underneath the blot represent the relative intensity of cleaved PARP in each lane compared with the first lane of each blot.

Fig. 5.

EDD overexpression is sufficient to induce cisplatin resistance, dependent upon its ubiquitin ligase activity. (A) COS-7 cells on coverslips were transfected with Flag-EDD or GFP for 24h and then treated with cisplatin for an additional 24h. Fixed cells were stained for transfected and apoptotic cells and 4′,6-diamidino-2-phenylindole stained as shown in Supplementary Figure 5, available at Carcinogenesis Online. The percentage of apoptotic transfected cells was determined by cell counting. At least 500 cells were counted per condition in each of four experiments. (B) Same as in (A), but cells were transfected with GFP, Flag-EDD or Flag-EDD-C2768A, a ubiquitin ligase-deficient mutant. Cells were treated with 15 μM cisplatin for 24h on the day following transfection and the percentage of apoptotic transfected cells was determined by cell counting 24h later.

Fig. 6.

DOPC nanoparticle delivery of EDD siRNA in vivo reduces tumor burden. (A) ES-2 or A2780ip2 cells were injected intraperitoneally into 40 female athymic nude mice per cell line. Mice were either treated with control siRNA in DOPC (lane 1), EDD siRNA1 in DOPC (lane 2), control siRNA in DOPC plus cisplatin (lane 3) or EDD siRNA1 in DOPC plus cisplatin (lane 4). Mice were treated for 4 weeks before killing and tumor collection. Tumors were excised and total tumor weight determined. The number above each lane represents the mean tumor weight in grams. Immunohistochemistry demonstrates EDD knockdown in vivo. A2780ip2 tumors from mice treated with (B) control siRNA in DOPC, (C) control siRNA in DOPC plus cisplatin, (D) EDD siRNA1 in DOPC and (E) EDD siRNA1 in DOPC plus cisplatin were immunostained with EDD antibody followed by horseradish peroxidase secondary antibody.