The Trim39 ubiquitin ligase inhibits APC/CCdh1-mediated degradation of the Bax activator MOAP-1

Abstract

Proapoptotic Bcl-2 family members, such as Bax, promote release of cytochrome c from mitochondria, leading to caspase activation and cell death. It was previously reported that modulator of apoptosis protein 1 (MOAP-1), an enhancer of Bax activation induced by DNA damage, is stabilized by Trim39, a protein of unknown function. In this paper, we show that MOAP-1 is a novel substrate of the anaphase-promoting complex (APC/C(Cdh1)) ubiquitin ligase. The influence of Trim39 on MOAP-1 levels stems from the ability of Trim39 (a RING domain E3 ligase) to directly inhibit APC/C(Cdh1)-mediated protein ubiquitylation. Accordingly, small interfering ribonucleic acid-mediated knockdown of Cdh1 stabilized MOAP-1, thereby enhancing etoposide-induced Bax activation and apoptosis. These data identify Trim39 as a novel APC/C regulator and provide an unexpected link between the APC/C and apoptotic regulation via MOAP-1.

Figure 1.

Trim39 is an Xnf7-related E3 ubiquitin ligase, and ligase activity is required for APC/C inhibition. (A) Schematic representation of the domains found within Xnf7 and Trim39 proteins. BBC represents the coiled-coil region C terminal to the B-Box domains. (B) In vitro ubiquitylation using recombinant Trim39 WT, RING domain mutant, C44A, and C52A in the absence or presence of ubiquitin and E1/E2 (UbcH5a). (C) Cell lysates were prepared from HeLa cells synchronized after a 1-h release from nocodazole (allowing further release in vitro) and incubated with 1 µM recombinant proteins, as indicated. Samples were collected and immunoblotted with cyclin B1 antibody. (D) Radiolabeled in vitro translated cyclin B1 fragment (1–106) was incubated with ubiquitin, E1 and E2 (UbcH5a, UbcH5c, or UbcH10, as indicated), ATP, His-Cdh1, and APC/C immunoprecipitated from HeLa cell G1 lysates using Cdc27 antibody. Reactions were supplemented with control buffer, MBP, MBP-Trim39 WT, or C44A recombinant protein for 45 min. Samples were detected by SDS-PAGE and autoradiography.

Figure 2.

MOAP-1 protein is degraded in a cell cycle–specific manner, and Trim39 attenuates the degradation. (A) HeLa cells were synchronized at prometaphase with nocodazole and collected at the indicated times after release. Radiolabeled in vitro translated cyclin B1 and MOAP-1 proteins were incubated in each lysate. Samples were taken at each time point, and reactions were stopped by addition of SDS loading buffer. (B) HeLa cells were synchronized with nocodazole and collected after release at the indicated times. Lysates were analyzed by immunoblotting with the indicated antibodies. (C) 1 µM MBP-Trim39 WT and C44A recombinant proteins were incubated in nocodazole-synchronized cell lysates for 3 h and then supplemented with in vitro translated, radiolabeled MOAP-1 or Geminin. Samples were taken at each time point, and reactions were stopped by addition of SDS loading buffer. The line graphs below represent quantitation of three independent experiments. Error bars are SD.

Figure 3.

Cdh1 regulates MOAP-1 stability. (A, top) HeLa cells were treated with control or Cdh1 siRNA, arrested in nocodazole, released for 4 h, and lysed, and lysates were supplemented with radiolabeled in vitro translated MOAP-1 protein. Samples were withdrawn into SDS loading buffer at the indicated times and resolved by SDS-PAGE and autoradiography. (bottom) Immunoblot of Cdh1 from the Cdh1 siRNA–treated (siCdh1) cells above. siCtr, siRNA control. (B) 293T cells were transfected with 0.5 µg Myc–MOAP-1 plasmid and increasing amounts of HA-Cdh1, as indicated. Cell lysates were collected and probed with anti-Myc, -HA, or -actin antibody. Relative Myc-MOAP1 levels (normalized to levels seen in the absence of exogenous Cdh1 in lane 1) are quantitated below based on three repetitions of the experiment. Error bars are SD. (C) 293T cells were transfected with 0.5 µg HA-Cdh1 and 1.5 µg Myc–MOAP-1. Cell lysates were prepared and immunoprecipitated (IP) by normal IgG or anti-HA antibody and probed with anti-HA or -Myc antibody. (D) Anti-Cdh1 or control IgG immunoprecipitates from 20 µM MG132-treated 293T cells were immunoblotted with anti–MOAP-1 or anti-Cdh1 antibody.

Figure 4.

MOAP-1 is an APC/C^Cdh1 substrate with D-box motifs. (A) Sequences of putative D-boxes on MOAP-1. D-box consensus amino acids mutated to Ala and transfected MOAP-1 WT or D-box mutant constructs into 293T cells are underlined. (B) 293T cells were transfected with HA-Cdh1 and Myc–MOAP-1 WT or MOAP-1 MT. Cells were collected and immunoprecipitated (IP) by anti-HA antibody. Samples resolved by SDS-PAGE were immunoblotted using anti-Myc or -HA antibodies. (C) MBP or MBP–MOAP-1 WT or MBP–MOAP-1 MT was incubated with His-Cdh1 for 2 h at 4°C, and proteins were retrieved from the mixture using amylose beads. After SDS-PAGE, samples were immunoblotted with anti-MBP or -Cdh1 antibodies. (D) 293T cells were transfected with Myc–MOAP-1 WT or Myc–MOAP-1 MT and treated with 100 µM cycloheximide (CHX). Cell lysates were prepared at the indicated times after cycloheximide treatment and immunoblotted with anti-Myc or -actin antibodies. (E) 293T cells were transfected with Myc–MOAP-1 WT or MT in the presence or absence of HA-Cdh1. After 48 h, cells were treated with 20 µM MG132 for 16 h, and lysates were collected for immunoblotting with anti-Myc, -HA, or -actin antibodies. (F) Recombinant MBP–MOAP-1 WT or MT proteins were incubated with ubiquitin, E1 and E2 (UbcH10 and Ube2S), ATP, and APC/C precipitated from HeLa G1 cell lysates using anti-Cdc27 antibody. Samples were immunoblotted with anti–MOAP-1 antibody.

Figure 5.

Higher Bax activation in Cdh1 knockdown cells as a result of MOAP-1. (A) 293T cells treated with Cdh1 (siCdh1), MOAP-1 (siMOAP-1), or control siRNA (siCtr) were treated with control or 100 µM etoposide for 48 h. Bax activation was monitored by IP with 6a7 antibody from 293T cells followed by immunoblotting of total Bax (N20). Input lysates were immunoblotted with anti-actin and -Bax (N20) antibodies. (B) Cell lysates from A were immunoblotted with anti-Cdh1 or anti–MOAP-1 antibodies. (C) Control, Cdh1, and MOAP-1 knockdown PC3 cells were treated with 100 µM etoposide for 48 h. The percentage of cells with sub-G1 DNA content was measured by PI staining and flow cytometry. (D) PC3 cells were processed as in C but treated with 20 µM cisplatin. (C and D) The asterisks indicate that the difference between the two experiments is significant (P < 0.05), using an unpaired t test. The data shown represent three independent experiments. Error bars are SEM. (E) Cdh1 or MOAP-1 was knocked down using siRNA in PC3 cells, and samples were immunoblotted for Cdh1, MOAP-1, or actin. (F) Cells treated as in D were lysed, and lysates were immunoblotted to detect caspase 3 processing.