Phosphorylation of the proapoptotic BH3-only protein bid primes mitochondria for apoptosis during mitotic arrest

Abstract

Mitosis is a moment of exquisite vulnerability for a metazoan cell. Failure to complete mitosis accurately can lead to aneuploidy and cancer initiation. Therefore, if the exit from mitosis is delayed, normal cells are usually removed by apoptosis. However, how failure to complete mitosis activates apoptosis is still unclear. Here, we demonstrate that a phosphorylated form of the BH3-only protein Bid regulates apoptosis if mitotic exit is delayed. Bid is phosphorylated on serine 66 as cells enter mitosis, and this phosphorylation is lost during the metaphase-to-anaphase transition. Cells expressing a nonphosphorylatable version of Bid or a BH3-domain mutant were resistant to mitotic-arrest-induced apoptosis. Thus, we show that Bid phosphorylation primes cells to undergo mitochondrial apoptosis if mitotic exit is delayed. Avoidance of this mechanism may explain the selective pressure for cancer cells to undergo mitotic slippage.

Graphical abstract

Figure 1

Bid Is Required for Apoptosis following Delayed Mitotic Exit (A) Knockdown and re-expression of Bid in human carcinoma cells. RKO cells stably expressing control pVenus, pVenus-shBid, or pVenus-shBid-mBidYFP were immunoblotted for human Bid (hBid) and BidYFP. Vinculin was immunoblotted as a loading control. IB, immunoblot. (B) RKO and DLD1 cells expressing pVenus, pVenus-shBid, or pVenus-shBid-mBidYFP were left untreated or treated with paclitaxel for 18 hr. Cells were collected and apoptosis quantified by immunostaining for active caspase 3. The error bars represent SEM. Data represent the mean of three independent experiments. Data were analyzed by ANOVA. n/s, not significant. (C) In the left panel, RKO cells stained with Hoechst. RKO cells remained in mitosis when knockdown of Bid prevented them undergoing apoptosis following 18 hr in paclitaxel. In the right panel, RKO cells treated with paclitaxel immunostained for cytochrome c and active caspase 3, as well as Hoechst. The cell indicated by the arrow shows that active caspase 3 corresponds with loss of mitochondrial cytochrome c and pyknotic nuclei. (D) Bid^−/− mouse embryonic fibroblasts (MEF) were stably infected with lentivirus expressing either BidYFP-WT or BidYFP-G94E, before being treated with combinations of paclitaxel and ABT-737 for 18 hr. Apoptosis was quantified as in (B). The error bars represent SEM. Data represent the mean of three independent experiments.

Figure 2

Endogenous Bid Undergoes Reversible Posttranslational Modification during Mitosis (A) Schematic representation of mouse Bid, showing known sites of posttranslational modification within the loop between α helices 2 and 3. (B) Endogenous Bid in MEFs migrates as multiple forms. MEFs were treated with etoposide for 1 hr and lysates analyzed by immunoblotting for Bid, with or without alkaline phosphatase (AP) treatment. The higher-migrating form of Bid is indicated (pBid). The slower-migrating form of Bid phosphorylated in response to etoposide is also indicated (Bid p61/p78). (C) Lysates from MEFs arrested in either G1 or mitosis were analyzed by immunoblotting for Bid or phosphorylated histone H3 (pSer10-H3). (D) MEFs arrested in mitosis were collected by overnight treatment with nocodazole and shaking from the dish. These cells were replated into media with or without MG132. Lysates were collected at the indicated times and immunoblotted for Bid, cyclin B, phospho-histone H3, and calnexin. (E) MEFs enriched in G1 were released for 8 hr into the indicated combinations of nocodazole and RO-3306. In the right lane, cells were released into nocodazole for 8 hr and RO-3306 added for 1 hr prior to lysis. Lysates were immunoblotted for Bid and pSer10-H3. (F) Mitotic MEFs were collected by overnight treatment in nocodazole, followed by shake off, and replated in the indicated combinations of nocodazole, RO-3306, and MG132 for 2 hr. Lysates were immunoblotted for Bid, cyclin B, phospho-histone H3, and calnexin.

Figure 3

Bid Is Phosphorylated on Serine 66 during Mitosis (A) BidYFP was transiently expressed in HEK293T cells, which were untreated, or treated with nocodazole. BidYFP was isolated on GFP-trap agarose, separated by SDS-PAGE, and stained with Coomassie blue. pBid is indicated as the sample isolated from nocodazole-treated cells. (B) Fragmentation spectrum of the AspN-generated Bid peptide corresponding to amino acids 59–74. The data include diagnostic ions that identify the peptide and also a fragment ion that demonstrates that the peptide is phosphorylated at Ser66 (y₉). No fragment ions were observed supporting phosphorylation at any other serine residues present in the peptide. Fragment ions b₃ and b₅ discount S61 phosphorylation. (C) Fragmentation spectrum of synthetic peptide representing mouse Bid amino acids 59–74, phosphorylated on S66. This peptide shows the same diagnostic fragment ions identified in (B). (D) Bid^−/−MEFs expressing either BidYFP-WT or BidYFP-S66A were untreated or treated with nocodazole and precipitated on GFP-Trap agarose. Precipitates and WCLs were separated by SDS-PAGE and immunoblotted for total Bid or Bid phospho-S66. A phospho-S66-positive band was only detected in the BidYFP-WT cells treated with nocodazole. (E) MEFs transiently expressing the indicated BidYFP constructs were arrested in either G1 or mitosis. Mitotic cells were collected by shaking (M) and the remaining cells attached to the dish were also collected (A). Lysates were analyzed by immunoblotting for Bid. (F) Comparison of the amino acids sequences from mouse (top) and human (bottom) Bid in the regulatory loop. The regions of divergence around mouse S66 are indicated. (G) Human Bid does not show a shift in mobility by SDS-PAGE. DLD1 and RKO cells were treated with paclitaxel for 18 hr and lysates analyzed by immunoblotting for endogenous Bid. (H) Fragmentation spectrum of the AspN-generated peptide corresponding to amino acids 55–74 from hBid-YFP, isolated on GFP-Trap agarose from nocodazole-treated HEK293T cells. The data include diagnostic ions showing hBid-YFP is phosphorylated uniquely on S67.

Figure 4

Bid Phosphorylation on Serine 66 Sensitizes Cells to Apoptosis during Mitotic Arrest (A) RKO cells stably infected with pVenus, pVenus-shBid, or pVenus-shBid coexpressing the indicated mouse (left panel) and human (right panel) BidYFP variants under the ubiquitin promoter were analyzed by immunoblotting with an antibody that recognizes both human and mouse Bid. Immunoblotting for Erk was used as a loading control. Endogenous human Bid is only present in the control cells. (B) The control RKO lines and those expressing mouse BidYFP variants were untreated or treated with 1 μM paclitaxel for 18 hr. Lysates were analyzed by immunoblotting for Bid and active caspase 3. Erk was a loading control. Note the shift in mobility of BidYFP-WT and BidYFP-G94E in paclitaxel-treated RKO cells. (C) The RKO lines from (A), untreated or treated with 1 μM paclitaxel for 18 hr, were immunostained for active caspase 3 and apoptosis quantified. The data represent the mean of three independent experiments. The error bars represent SEM. Data were analyzed by ANOVA. (D) Images of the paclitaxel-treated RKO cell lines from (C), immunostained for active caspase 3. Nuclei were stained with Hoechst. (E) Bid^−/− MEFs, infected with the indicated pVenus lentiviruses, were left untreated or treated with 1 μM paclitaxel. Apoptosis was quantified as above. The data represent the mean of three independent experiments. The error bars represent SEM. Data were analyzed by ANOVA. (F) RKO cells infected with the indicated lentiviruses expressing human Bid or human BidS67A were treated with paclitaxel as in (C). Cells showed similar responses to those expressing the mouse BidYFP. The data represent the mean of three independent experiments. The error bars represent SEM. Data were analyzed by ANOVA. (G) The indicated RKO lines, untreated or treated with monastrol for 18 hr, were immunostained for active caspase 3 and apoptosis quantified. The data represent the mean of three independent experiments. The error bars represent SEM. Data were analyzed by ANOVA.

Figure 5

Bid Phosphorylation on Serine 66 Increases Mitochondrial Priming in Mitotic Cells (A) Wild-type and Bid^−/−MEFs were treated for 18 hr with the indicated combinations of ABT-737, paclitaxel, and etoposide. Apoptosis was quantified by immunostaining for active caspase 3. Data represent the mean of three independent experiments. The error bars represent SEM. (B) RKO cells infected with control pVenus, pVenus-shBid, pVenus-shBid-BidYFP-WT, or pVenus-shBid-BidYFP-S66A were untreated or treated for 18 hr with 5 μM ABT-737, 1 μM paclitaxel, or both ABT-737 and paclitaxel. Apoptosis was quantified by immunostaining for active caspase 3. Data represent the mean of three independent experiments. The error bars represent SEM. (C) DLD1 cells infected with pVenus, pVenus-shBid, pVenus-shBid-BidYFP-WT, or pVenus-shBid-BidYFP-S66A were left untreated or treated with nocodazole overnight. Lysates were immunoblotted for Bid. (D) DLD1 and RKO cells, control or expressing mouse BidYFP-WT, were arrested overnight in nocodazole and mitotic cells collected by shaking. Lysates were immunoblotted for Bid. Quantification of the pBid/Bid ratios using Odyssey-based imaging showed no difference between the cell types. (E) DLD1 cells expressing endogenous Bid were treated with the indicated combinations of drugs for 18 hr. Cells were immunostained for active caspase 3 and apoptosis quantified. Data represent the mean of three independent experiments and analyzed by ANOVA. The error bars represent SEM. (F). Representative images from (E).