TRAF6 Restricts p53 Mitochondrial Translocation, Apoptosis, and ...

Download PDF
13 downloads 0 Views 3MB Size Report
Comment
Nov 3, 2016 - Mitochondrial p53 is involved in apoptosis and tumor suppression. However, its regulation is not well studied. Here, we show that TRAF6 E3 ...
Article

TRAF6 Restricts p53 Mitochondrial Translocation, Apoptosis, and Tumor Suppression Graphical Abstract

Authors Xian Zhang, Chien-Feng Li, Ling Zhang, ..., Chang-Hai Tsai, Kounosuke Watabe, Hui-Kuan Lin

Correspondence [email protected]

In Brief Zhang et al. discovered that TRAF6 prevents the mitochondrial translocation of p53 and spontaneous apoptosis by promoting K63-linked ubiquitination of p53 in cytosol. Genotoxic stress overrides this protection mechanism by translocating TRAF6 into nucleus. Deregulation of this mechanism may contribute to cancer development and resistance to chemotherapy and radiotherapy.

Highlights d

TRAF6 restricts p53 mitochondria localization and spontaneous apoptosis

d

TRAF6 promotes K63-linked ubiquitination of p53 at K24

d

K63-linked ubiquitination of p53 in cytosol attenuates tumor suppressive function

d

Genotoxic stress promotes TRAF6 phosphorylation and TRAF6 translocation

Zhang et al., 2016, Molecular Cell 64, 803–814 November 17, 2016 ª 2016 Elsevier Inc. http://dx.doi.org/10.1016/j.molcel.2016.10.002

Accession Numbers GSE69730

Molecular Cell

Article TRAF6 Restricts p53 Mitochondrial Translocation, Apoptosis, and Tumor Suppression Xian Zhang,1,2,3 Chien-Feng Li,9,10 Ling Zhang,2,11 Ching-Yuan Wu,2,12 Lixia Han,2,3 Guoxiang Jin,1,2 Abdol Hossein Rezaeian,2 Fei Han,1,2,3 Chunfang Liu,1,2 Chuan Xu,1,2 Xiaohong Xu,2 Chih-Yang Huang,4,5 Fuu-Jen Tsai,6,7 Chang-Hai Tsai,5,8 Kounosuke Watabe,1 and Hui-Kuan Lin1,2,4,5,13,* 1Department

of Cancer Biology, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA 3The University of Texas Graduate School of Biomedical Sciences at Houston, Houston, TX 77030, USA 4Graduate Institute of Basic Medical Science, China Medical University, Taichung 404, Taiwan 5Department of Biotechnology, Asia University, Taichung 41354, Taiwan 6College of Chinese Medicine, China Medical University, Taichung 40402, Taiwan 7Department of Medical Genetics, Pediatrics, and Medical Research 8Center of Molecular Medicine China Medical University Hospital, Taichung 40402, Taiwan 9National Institute of Cancer Research, National Health Research Institutes, Tainan 704, Taiwan 10Department of Pathology, Chi-Mei Foundational Medical Center, Tainan 710, Taiwan 11Key Laboratory of Laboratory Medical Diagnostics, Ministry of Education, College of Laboratory Medicine, Chongqing Medical University, 1#, Yixueyuan Road, Chongqing, 400016, China 12Department of Chinese Medicine, Chiayi Chang Gung Memorial Hospital, Chiayi 613, Taiwan 13Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.molcel.2016.10.002 2Department

SUMMARY

Mitochondrial p53 is involved in apoptosis and tumor suppression. However, its regulation is not well studied. Here, we show that TRAF6 E3 ligase is a crucial factor to restrict mitochondrial translocation of p53 and spontaneous apoptosis by promoting K63-linked ubiquitination of p53 at K24 in cytosol, and such ubiquitination limits the interaction between p53 and MCL-1/BAK. Genotoxic stress reduces this ubiquitination in cytosol by S13/T330 phosphorylation-dependent translocation of TRAF6 from cytosol to nucleus, where TRAF6 also facilitates the K63-linked ubiquitination of nuclear p53 and its transactivation by recruiting p300 for p53 acetylation. Functionally, K63-linked ubiquitination of p53 compromised p53-mediated apoptosis and tumor suppression. Colorectal cancer samples with WT p53 reveal that TRAF6 overexpression negatively correlates with apoptosis and predicts poor response to chemotherapy and radiotherapy. Together, our study identifies TRAF6 as a critical gatekeeper to restrict p53 mitochondrial translocation, and such mechanism may contribute to tumor development and drug resistance. INTRODUCTION Cell fate determination is a tightly regulated process that is involved in human development and diseases. A key factor

responsible for cell fate determination is p53, which is widely involved in diverse biological processes, including cell-cycle arrest, apoptosis, cellular senescence, and energy homeostasis (Brady et al., 2011; Vousden and Prives, 2009; Vousden and Ryan, 2009). Because p53 is frequently mutated in a wide variety of human cancers, it is considered as the most important tumor suppressor. Therefore, it is critical to understand the molecular mechanism by which p53 activity is regulated. p53 is a short-lived protein whose stability is tightly controlled by Mdm2 E3 ligase (Brooks and Gu, 2006; Haupt et al., 1997; Kubbutat et al., 1997). Upon genotoxic stresses, such as ionizing radiation (IR) and DNA damage agents, p53 protein is stabilized and accumulated in the cells, presumably because of the loss of its interaction with Mdm2 and subsequent reduction in p53 ubiquitination. The majority of p53 resides in the nucleus, where it can bind to the promoter and/or enhancer of its target genes to either induce or repress gene expression. Induction of p21 and Gadd45 is responsible for p53-mediated cell-cycle arrest (elDeiry et al., 1993; Kastan et al., 1992), whereas upregulation of Puma, Bax, and Noxa is thought to mediate p53-dependent apoptosis (Miyashita et al., 1994; Nakano and Vousden, 2001; Oda et al., 2000). Loss of p53 impairs both genotoxic stressinduced cell-cycle arrest and apoptosis, suggesting that p53 is a key mediator in the cell fate decision upon DNA damage. Thus, it is traditionally accepted that p53 exerts its biological functions primarily through its transcriptional activity. In addition to ubiquitination-dependent degradation, p53 activity is also regulated by other posttranslational modifications, such as acetylation (Kruse and Gu, 2009). Acetylation of p53 is generally required for its transcriptional activity (Gu and Roeder, 1997; Tang et al., 2008). p53 acetylation is induced upon exposure to genotoxic stresses, which is the result of enhanced interaction of p53 with HAT domain-containing proteins, such as CBP/p300.

Molecular Cell 64, 803–814, November 17, 2016 ª 2016 Elsevier Inc. 803

However, how genotoxic stresses trigger the association of p53 and p300/CBP remains largely unanswered. Beyond its mere role in the nucleus for gene expression, a transcription-independent role of p53 in inducing mitochondrial apoptosis has been recently suggested (Moll et al., 2005; Vaseva and Moll, 2009). Several studies have shown that the p53 mutants with defects in transcription activity are still able to induce apoptosis (Chen et al., 1996; Haupt et al., 1995; Kokontis et al., 2001; Mihara et al., 2003). Later studies unraveled that p53 moves to the mitochondrial outer membrane in response to DNA damage, where it binds to BAK/Bax to promote BAK/Bax oligomerization (Chipuk et al., 2004; Mihara et al., 2003). This event leads to mitochondrial outer membrane permeabilization (MOMP), cytochrome c release, caspase-3 activation, and consequent apoptosis (Chipuk and Green, 2008; David, 2012). Consistent with this notion, recombinant p53 proteins can bind to the purified intact mitochondria and robustly trigger cytochrome c release in the in vitro assay (Chipuk et al., 2004; Mihara et al., 2003). These studies therefore underscore the direct function of p53 in the mitochondria to induce apoptosis. Although there is increasing appreciation of the role of p53 in the mitochondria, it is also critical to elucidate the mechanism by which the complexity of p53 activation and localization in the mitochondria is regulated. Here we provide evidence that TRAF6 is the critical factor to control the p53 mitochondrial translocation. TRAF6 triggers K63-linked ubiquitination of p53 at K24 in the cytoplasm, which reduces the interaction between p53 and MCL-1/BAK, thus keeping p53 away from mitochondria and preventing p53-mediated activation of BAK. Genotoxic stress rapidly removes this K63-linked ubiquitination to trigger the p53-mediated apoptosis pathway in the mitochondria by increasing the free p53-toTRAF6 ratio. In contrast, genotoxic stress triggers the localization of TRAF6 to nucleus, where it also mediates the K63-linked ubiquitination of p53 at K24, which promotes the interaction of p53 with p300, thereby facilitating p53 acetylation and downstream gene expression, including p21 and GADD45, for cell survival under stress conditions. Our results provide great insight into how genotoxic stress coordinates the transcription-dependent and transcription-independent role of p53 in apoptosis by modulating the dynamic status of K63-linked ubiquitination of p53 in distinct cellular compartments. RESULTS p53 translocates from cytosol to mitochondria to trigger cytochrome c release and subsequent apoptosis in response to genotoxic stresses. Although p53 protein is drastically induced by genotoxic stress, both nuclear and cytosolic p53 protein are still detectable in various mice tissues and cancer cell lines without exposing to genotoxic agents (Figures S1A and S1B). Because unmodified recombinant p53 readily binds to and efficiently triggers apoptosis pathway in mitochondria in vitro (Chipuk et al., 2004; Mihara et al., 2003), we hypothesized that there may be a mechanism to restrict p53 in cytosol and prevent its mitochondrial translocation to avoid undesired apoptosis under physiological conditions. Although BCL-2, BCL-xL, and MCL-1 are known direct sequesters for cytosolic p53, it is also reported 804 Molecular Cell 64, 803–814, November 17, 2016

that knockout of liver specific gene IGFBP1 accumulates p53 in mitochondria and induces apoptosis in liver (Leu and George, 2007), suggesting that the whole machinery to antagonize cytoplasmic p53 is more complicated than just the traditional sequesters. Because K63-linked ubiquitination has been linked to protein trafficking, we examined whether p53 is modified through K63-linked ubiquitination in distinct cellular compartments. Surprisingly, we found that cytosolic pool of p53 was ubiquitinated through the K63 linkage, but such modification was not detected for the mitochondrial p53 (Figure S1C). Also, a profound decrease in K63-linked ubiquitination of cytosolic p53 was correlated with the increase in p53 mitochondrial translocation upon genotoxic stress (Figure S1C), indicating that K63linked ubiquitination of p53 could be a potential negative signal to restrict p53 mitochondrial translocation. To study whether K63-linked ubiquitination of p53 plays a suppressive role in the mitochondrial translocation of p53, we aimed to identify E3 ligases that interact with p53 and trigger K63-linked ubiquitination of p53. We screened a panel of E3 ligases and found that ectopic expression of TRAF6 E3 ligase readily induced K63-linked ubiquitination of p53 (Figures 1A, S1D, and S1E). p53 is a short-lived protein whose stability is regulated by K48-linked ubiquitination and proteasomal degradation (Dornan et al., 2004; Haupt et al., 1997; Leng et al., 2003). Our data showed that TRAF6 only promoted K63-linked, not K48linked, ubiquitination of p53 (Figure 1B). We further demonstrated that TRAF6, not TRAF6-C70A E3 ligase dead mutant, induced in vivo and in vitro p53 ubiquitination (Figures 1C and 1D). Reciprocal co-immunoprecipitation assay revealed that TRAF6 interacted with p53 endogenously (Figure 1E). Endogenous ubiquitination assay further demonstrated that K63-linked ubiquitination of p53 in cytosol was reduced in Traf6 / mouse embryonic fibroblasts (MEFs) or upon genotoxic agent treatment (Figure 1F). Together, these data suggest that TRAF6 is a direct E3 ligase for p53. We next determined how genotoxic stress reduced K63-linked ubiquitination of p53 in cytosol, which was also observed in multiple other cell lines with CDDP or etoposide treatment (Figures S1F–S1J). Besides the stabilization of p53 protein in cytosol upon genotoxic stress, we also observed a decreased localization of TRAF6 protein in cytosol (Figures 1G and S1K–S1M), which resulted in an increase in the free p53-to-TRAF6 ratio in cytosol and may lead to the reduction of K63-linked ubiquitination of cytosolic p53 upon genotoxic stress. To study how TRAF6 localization in cytosol is regulated by genotoxic stress, we examined whether genotoxic stress-induced DNA damage signaling is involved. Inhibition of ATM/ATR, which are the major upstream kinases activated by DNA damage, prevented the reduction of TRAF6 level in cytosol (Figure S1N). Immunoprecipitation assay further demonstrated that genotoxic agents induced the phosphorylation of TRAF6 at TQ/SQ motifs (Figures 1H and S1O), which are consensus ATM/ATR substrate sites. Mutations on TQ/SQ sites S13 and T330 to alanine on TRAF6 reduced genotoxic stress-induced phosphorylation of TRAF6 (Figure 1I). Such mutant also displayed resistance to the genotoxic agent-mediated reduction of TRAF6 level in cytosol (Figures 1J and S1P), while its affinity to p53 is similar to TRAF6 WT (Figure S1Q). Together, our data suggest that genotoxic

Figure 1. p53 Is Ubiquitinated by TRAF6 through K63 Linkage (A) U2OS cells were transfected with indicated plasmids for in vivo ubiquitination assay. See also Experimental Procedures. (B) In vivo ubiquitination assay was performed for U2OS cells transfected with indicated plasmids. (C) U2OS cells were transfected with indicated plasmids for in vivo ubiquitination. (D) TRAF6 ubiquitinates p53 in vitro. See also Experimental Procedures. (E) Immunoprecipitation assay was performed for U2OS cells. (F) Immunoprecipitation assay was performed for the cytosol fraction of WT and Traf6 / primary MEFs. (G) Cellular fractionation assay was performed for U2OS cells treated with CDDP for indicated times. (H) Immunoprecipitation assay was performed for U2OS cells treated with CDDP for indicated times. Phosphorylation was detected by the anti-phosphor-TQ/SQ motif, which is the consensus phosphorylation site for ATM and ATR kinase. (I) Immunoprecipitation assay was performed for 293T cells expressing WT and mutant TRAF6 as indicated. CDDP and/or ATM inhibitor was also treated as indicated. (J) Cellular fractionation was performed for 293T cells expressing WT and mutant TRAF6 with or without CDDP treatment.

stress promotes TRAF6 phosphorylation, which reduces the level of TRAF6 in cytosol and may contribute to the reduction of K63-linked ubiquitination of p53 in cytosol. If K63-linked ubiquitination of cytosolic p53 acts as a suppressive signal for p53 mitochondrial localization, TRAF6 would serve as a suppressor for p53 mitochondrial localization and apoptosis, whereas its loss will favor p53 mitochondria accumulation and spontaneous apoptosis. Using Traf6 / genetic mouse model, we found that Traf6 / thymus, spleen, and lung displayed a higher spontaneous apoptosis rate than their wildtype (WT) counterparts (Figures 2A, 2B, and S2A). The spontaneous apoptosis level in thymus cells from Traf6 / mice was also comparable with that in thymus cells from WT mice upon IR treatment (Figure 2A). Notably, treating Traf6 / mice with pifithrin-m, a potent inhibitor that specifically inhibits p53 mitochondrial translocation, reversed the spontaneous apoptosis in

Traf6 / thymus, while pifithrin-a, the specific inhibitor for p53 transcriptional activity, failed to do so (Figures 2C and 2D), suggesting that mitochondrial p53 contributes to spontaneous apoptosis upon TRAF6 deficiency. It should be noted that a recent study suggests that in addition to preventing p53 mitochondria accumulation, pifithrin-m can also disrupt the interaction between Hsp70 and its chaperones, thus leading to apoptosis (Leu et al., 2009). Therefore, it is highly unlikely that pifithrin-m acts through disrupting Hsp70 activity to reverse spontaneous apoptosis in Traf6 / thymus. Consistently, knockdown of TRAF6 in U2OS cells also triggered spontaneous apoptosis (Figures 2E and S2B). As in mice thymus, such spontaneous apoptosis upon TRAF6 deficiency acts through p53, as knockdown of p53 reversed the elevated spontaneous apoptosis in TRAF6-deficient U2OS cells (Figures 2E and S2B). By isolating mitochondria, we found a spontaneous accumulation of p53 in mitochondria in Traf6 / thymus, MEFs, and TRAF6-knockdown U2OS cells, reaching a level similar to that observed in control cells treated with CDDP (Figures 2F, 2G, and S2C). However, CDDP treatment did not further enhance the level of p53 in mitochondria in these TRAF6-deficient cells (Figures 2G and S2C). Consistent with the mitochondrial translocation of p53 and apoptosis, cytochrome c, the effector of the mitochondrial apoptosis pathway, was released from Molecular Cell 64, 803–814, November 17, 2016 805

Figure 2. TRAF6 Inhibits p53 Translocation to Mitochondria and Apoptosis (A) WT and Traf6 / total thymus cells were analyzed by TUNEL assay. See also Experimental Procedures. (B) WT and Traf6 / total thymus cells were analyzed by western blot. (C) WT and Traf6 / mice were treated with pifithrin-m or pifithrin-a and apoptosis of total thymus cells was analyzed by TUNEL assay. (D) Mitochondrial fractionation was performed for WT and Traf6 / thymus treated with pifithrin-m or pifithrin-a. See also Experimental Procedures. (E) TUNEL assay was performed for U2OS cells with indicated knockdown and examined by fluorescence microscopy. (F and G) Mitochondria were isolated from total thymus cells (F) or U2OS cells (G). U2OS cells were treated with or without CDDP for 6 hr. (H) Mitochondria fractionation was performed for the WT and Traf6 / mice thymus treated with or without IR. (I) TRAF6 deficiency induced mitochondrial fragmentation was p53 dependent. Immunofluorescence assay was performed for U2OS cells with indicated knockdown. All data are represented as mean ± SEM. *p < 0.01, Student’s t test.

mitochondria to cytosol in Traf6 / thymus before IR, while cytochrome c was only released upon IR in WT thymus (Figure 2H). Because mitochondrial fragmentation and reduced mitochondrial potential are also important readouts for the activation of mitochondrial apoptosis pathways (Arnoult, 2007; Brooks et al., 2007; Ly et al., 2003), we examined whether TRAF6-deficient cells also exhibited such phenotypes. Remarkably, TRAF6 deficiency readily induced mitochondrial fragmentation and reduced mitochondrial potential even before genotoxic stress (Figures S2D–S2F). Knockdown of p53 partially reversed mitochondrial fragmentation phenotype in TRAF6-knockdown cells (Figure 2I). Accordingly, these results suggest that TRAF6 serves as a key player to restrict p53 mitochondrial translocation and spontaneous apoptosis in unstressed cells both in vivo and in vitro. To directly support the functional role of K63-linked ubiquitination of p53 in these processes, we then determined TRAF6dependent ubiquitination site(s) on p53. We mutated every conserved lysine (K) residue on p53 and found that TRAF6 preferentially ubiquitinates p53 at K24 in the in vivo ubiquitination assay (Figures 3A and S3A–S3C). Mutation of K24R on p53 also reduced in vitro ubiquitination of p53 by TRAF6 (Figure S3D). We then determined whether K63-linked ubiquitination of p53 affects p53 mitochondrial localization in a manner similar to TRAF6 does. Ectopic expression of p53 K24R mutant showed increased localization to mitochondria (Figure 3B), phenocopying TRAF6 deficiency. Hence, TRAF6-mediated K63-linked ubiquitination acts as a suppressive signal for p53 mitochondrial localization. 806 Molecular Cell 64, 803–814, November 17, 2016

As a previous report and our data showed that BAK but not Bax is required to recruit p53 to mitochondria upon genotoxic stress, and BAK is the principal target for p53 in mitochondria (Figure S3E) (Leu et al., 2004; Pietsch et al., 2007), we then rationalized that TRAF6-mediated K63-linked ubiquitination of p53 may inhibit p53 mitochondrial translocation by disrupting the interaction between p53 and BAK. In support of this notion, TRAF6 knockdown increased the binding between p53 and BAK without genotoxic agent treatment (Figure 3C). As a consequence, TRAF6 deficiency readily promoted BAK oligomerization in mouse thymus and multiple cell lines (Figures 3D, 3E, and S3F), while p53 deficiency inhibited such effect (Figure S3G). Upon CDDP treatment, TRAF6 overexpression also attenuated genotoxic stress-induced BAK oligomerization (Figure 3F). Restoration of TRAF6 but not TRAF6 E3 ligase dead mutant (C70A) inhibited BAK oligomerization in TRAF6-deficient cells (Figure 3G). Similarly, p53 K24R mutant displayed higher binding affinity to BAK and stronger activity to induce BAK oligomerization in comparison with p53 WT (Figures 3H and 3I). Of note, TRAF6 knockdown or p53 K24R mutation also increased the binding between p53 and MCL-1 (Figures 3C and 3H), which is the major sequester of BAK in mitochondria. Because p53 is also shown to interact with MCL-1 and antagonize the sequestering effect of MCL-1 on BAK (Leu et al., 2004), the increased interaction between p53 and MCL-1 triggered by the deficiency in K63-linked ubiquitination of p53 may also contribute to the release of BAK from inhibition, BAK oligomerization and subsequent apoptosis. To validate our notion further, we incubated recombinant p53 or recombinant p53 that was ubiquitinated by TRAF6 in vitro with

Figure 3. K63-Linked Ubiquitination of p53 Inhibited the Mitochondrial Translocation of p53 (A) In vivo ubiquitination assay was performed for U2OS cells transfected with indicated plasmids. (B) Mitochondria were isolated from H1299 cells transfected with p53 WT or K24R. (C) Immunoprecipitation assay was performed for control and TRAF6-knockdown U2OS. (D and E) Mitochondria were isolated from total thymus cells (D) or U2OS cells (E) and crosslinked as described in Experimental Procedures. (F) U2OS cells were treated with or without CDDP for 6 hr before BAK oligomerization assay. (G) Mitochondria were isolated from TRAF6-knockdown U2OS cells restored with TRAF6 WT or C70A and crosslinked by BMH. (H) Immunoprecipitation assay was performed for H1299 cells transfected with p53 WT and K24R. (I) Mitochondria were isolated from H1299 cells transfected with p53 WT and K24R and crosslinked by BMH. (J and K) Ubiquitinated p53 failed to interact with mitochondria and trigger BAK oligomerization in vitro. See also Figure S3H and Experimental Procedures. (L) In vitro binding assay was performed for recombinant MCL-1, BAK, and GST-p53 derived from in vitro ubiquitination assay.

the mitochondria isolated from p53-deficient H1299 cells (Figures S3D and S3H). Strikingly, p53 that was ubiquitinated by TRAF6 in vitro not only failed to interact with mitochondria but also lost its ability to trigger BAK oligomerization and cytochrome c release in vitro compared with non-ubiquitinated p53 and p53 K24R mutant (Figures 3J and 3K), providing strong evidence that TRAF6-mediated K63-linked ubiquitination of p53 prevents p53 from binding to mitochondria and inducing BAK oligomerization. We also performed in vitro assay to test whether K63-linked ubiquitination of p53 affects its capability to antagonize MCL-1. Although non-ubiquitinated p53 WT or p53 K24R mutant could readily disrupt the binding between BAK and MCL-1 in vitro, ubiquitinated p53 failed to do so (Figure 3L). This was also supported by the in vitro binding assay showing that ubiquitinated p53 failed to interact with MCL-1 (Figure S3I). In addition, ubiquitinated p53 also showed reduced affinity to

BAK (Figure S3J), suggesting that K63linked ubiquitination of p53 may inhibit its binding to both BAK and MCL-1, and these two mechanisms may contribute to the suppression of BAK oligomerization and activation at the same time. Our finding that TRAF6 protect cells from spontaneous apoptosis by inhibiting mitochondrial localization of p53 in multiple cell types prompted us to examine whether TRAF6-mediated p53 ubiquitination at K24 regulates p53-mediated tumor suppression. We found that introducing p53 WT or p53 K24R to H1299 cells could inhibit in vitro colony formation and in vivo tumor growth in xenograft models (Figures 4A–4C). Notably, although TRAF6 overexpression could block p53-mediated cell killing and tumor suppression, it failed to do so in cancer cells expressing p53-K24R mutant (Figures 4A–4C), suggesting that TRAF6-mediated p53 ubiquitination at K24 generally protects cancer cells from p53-mediated tumor suppression. As p53 plays a critical role in chemotherapy- and radiotherapy-induced apoptosis of cancer cells, we examined whether TRAF6 affect cancer cell survival under the treatment of genotoxic agents. We found that TRAF6 knockdown or C70A mutation sensitize the cells to genotoxic stress (Figures S4A–S4D), confirming the protective role of TRAF6 against genotoxic agents. We further studied the clinical relevance of TRAF6 and chemotherapy- or IR-induced apoptosis. Strikingly, higher TRAF6 protein levels negatively correlated with the level of cleaved caspase-3 and Molecular Cell 64, 803–814, November 17, 2016 807

Figure 4. TRAF6 Inhibited p53 WT but Not p53 K24R-Induced Tumor Suppression (A–C) Colony-forming assay (A), xenograft tumor growth assay (B), and western blot assay (C) were performed for H1299 cells expressing indicated exogenous proteins. (D) Representative rectal cancers after concurrent CCRT showing low (left) and high (right) TRAF6 immunoreactivity were significantly associated with a higher tumor regression grade (TRG) and increased TUNEL and acspase-3 expression (left) and vice versa (right). (E and F) In post-CCRT rectal cancers, TRAF6 expression is significantly and negatively related to cell death as determined by TUNEL and active caspase-3 staining. (G and H) Survival analysis plotted by using Kaplan-Meier methods discloses TRAF6 expression in post-CCRT biopsy specimens is significantly predictive for disease-free and metastasis-free survival. All data are represented as mean ± SEM. *p < 0.01, Student’s t test.

TUNEL-positive staining, two readouts for apoptosis, and predicted the poor survival outcome of colorectal cancer in response to chemotherapy and IR treatment (Figures 4D–4H and S4E; Table S1). Thus, clinical data indicate that TRAF6 overexpression is correlated with less apoptosis and predicts poor 808 Molecular Cell 64, 803–814, November 17, 2016

response of colorectal cancer patients with WT p53 to chemotherapy and IR treatment. Our finding may provide a mechanism to explain how TRAF6 regulates apoptosis in the p53 WT context. Because another important role of p53 is the stress-sensing transcription factor, which modulates cell survival upon stress,

Figure 5. TRAF6 Is Required for Genotoxic Stress-Induced p53 Target Gene Expression (A) WT and Traf6 / mice aged 2.5 weeks were treated with IR (15 Gy) for 10 hr before thymus was isolated and total RNA or total protein was extracted. Genes with greater than 1.5-fold change in mRNA expression were used to generate the heatmap. (B and C) The expression of p53 target genes, p21 and GADD45, was validated by qPCR and western blot in the similar sample from Figure 5D. (D and E) Representative rectal cancers showing low (left) and high (right) TRAF6 immunoreactivity in pre-treatment (Pre-Tx) biopsy specimens were linked to low p21 (left) and high p21 (right) expression, respectively. In the pre-treatment (Pre-Tx) biopsy specimens, TRAF6 immunoexpression is significantly and positively associated with p21 expression.

we tested whether TRAF6 also regulates the transcriptional activity of p53. Hence, we applied gene expression microarray using the mouse thymus isolated from WT and Traf6 / mice treated with or without IR. Interestingly, the IR-induced expression pattern of p53 target genes was deregulated in Traf6 / thymus (Figures 5A and S5A). Importantly, p21 and Gadd45, which are important downstream mediators of p53 for the cell survival upon genotoxic stresses (Figure S5B) (Gartel and Tyner, 2002; Lu et al., 2008), are not induced in Traf6 / thymus, as confirmed by both real-time PCR and western blot analysis (Figures 5B and 5C). The impairment of IR-induced p21 and Gadd45 expression upon Traf6 deficiency could be extended to MEFs and various human cell lines (Figures S5C–S5E). Again, by analyzing human colorectal cancer samples with WT p53, we found that TRAF6 protein expression was significantly correlated with the expression of p21 (Figures 5D and 5E; Table S1). Collectively, our data suggest that TRAF6 regulates the genotoxic stress-induced expression of p53 downstream targets, p21 and Gadd45, in diverse cell types both in mouse and human, which may account for the role of TRAF6 in cell survival upon genotoxic stress and poor response to chemoradiotherapy (CCRT) in colorectal cancer patients with WT p53 (Figures 4D–4H). Meta-analysis of microarray data sets further revealed that E1A binding protein p300 (EP300 or p300), a critical transcriptional co-activator for p53, may be responsible for the TRAF6mediated gene expression pattern (Figure 6A). Because p300

is known to critically regulate p53 activity by acetylating p53 protein on the C-terminal lysine (K) residues (Grossman, 2001), we investigated whether p53 acetylation is regulated by TRAF6. Remarkably, Traf6 deficiency in thymus, MEFs, and human cancer cell lines impaired p53 acetylation upon IR treatment (Figures 6B, S5D, and S6A–S6C). Mechanistically, we found that the endogenous interaction between p53 and p300 upon IR stimulation was markedly reduced in Traf6 / MEFs (Figure 6C), while p300 protein level remains unaffected by TRAF6. Hence, TRAF6 serves as an essential regulator for p53 transactivation by recruiting p300 to p53 and subsequent acetylation of p53. Because p53 is ubiquitinated by TRAF6 in the cytoplasm, we found that p53 in nucleus is also ubiquitinated through K63 linkage (Figure 6D). This ubiquitination modification is induced by genotoxic stress and is TRAF6 dependent (Figures 6D and S1G–S1J), which is correlated with the genotoxic stress-induced phosphorylation of TRAF6 at S13/T330 sites and subsequent S13/T330 phosphorylation-dependent localization of TRAF6 in nucleus (Figures 1G–1J and S1G–S1P). p53 K24R mutant that is deficient in K63-linked ubiquitination also did not interact with p300 in the co-immunoprecipitation assay and was insufficient to induce p21 and Gadd45 expression in the p53-deficient H1299 cells (Figures 6E, 6F, and S6D). However, either ubiquitinated p53 or non-ubiquitinated p53 readily binds to p300 in vitro (Figure 6E), suggesting that the mechanism of how ubiquitinated p53 gains advantage in recruiting p300 in Molecular Cell 64, 803–814, November 17, 2016 809

Figure 6. TRAF6 Is Required for p300 Recruitment and p53 Acetylation (A) Predicted regulators for TRAF6-dependent transcription regulation. (B) Thymus prepared as described in Figure 5D was analyzed by western blot. (C) Immunoprecipitation assay was performed for WT and Traf6 / MEFs treated with or without IR (10 Gy) for 4 hr. (D) Immunoprecipitation assay was performed for the nuclear fraction of WT and Traf6 / primary MEFs to access p53 ubiquitination. (E) Immunoprecipitation assay was performed for the H1299 cells transfected with p53 WT or K24R to access its binding with p300. (F) Western blot was performed for H1299 cells transfected with p53 WT or K24R to detect p21 and GADD45.

the cellular context is complicated, which may involve other regulatory factors to be studied in the future. In addition, both p53 WT and K24R mutant bind to p300 similarly (Figure S6F), suggesting that the effect of p53 K24R mutant when expressing in the cells is probably due to the loss of K63-linked ubiquitination rather than change in its protein structure caused by K24 mutation. Together, these data suggest that genotoxic stressinduced K63-linked p53 ubiquitination at K24 by TRAF6 in nucleus is needed for p300 recruitment, p53 acetylation, and the expression of p53 target genes such as p21 and Gadd45. DISCUSSION p53 displays potent tumor suppressive effect by eliciting apoptosis, senescence, cell-cycle arrest, and metabolism regulation (Bieging et al., 2014; Levine and Oren, 2009). Among these, p53 executes apoptosis through its transcriptional activation and mitochondrial translocation. Notably, p53 readily translocates from cytosol to mitochondria to trigger cytochrome c release and subsequent apoptosis in response to genotoxic stresses. However, how genotoxic stress induces mitochondrial translocation of p53 remains largely unclear. Although one earlier study proposed that Mdm2 may be involved in genotoxic stress-induced p53 mitochondrial translocation and apoptosis (Marchenko et al., 2007), such a concept may conflict with well-established genetic evidence that loss of Mdm2 in mice 810 Molecular Cell 64, 803–814, November 17, 2016

causes mouse embryonic lethality through eliciting p53-dependent apoptosis (Jones et al., 1995; Montes de Oca Luna et al., 1995). Thus, a fundamental mechanism accounting for such phenomenon is still lacking and yet to be discovered. Because cytosolic p53 is still detectable in the unstressed cells (Figures S1A and S1B) and unmodified recombinant p53 readily binds to and triggers apoptosis pathway in mitochondria in vitro (Chipuk et al., 2004; Mihara et al., 2003), we hypothesized that there may be a cytosolic gatekeeper keeping p53 away from mitochondria under the unstressed condition. Our study offers the great insight into how such regulatory mechanism is operated in the cells. We identify TRAF6 E3 ligase as a critical factor to elicit such inhibitory mechanism for cytosolic p53. Mechanistically, we show that TRAF6 interacts with cytosolic p53 and triggers K63-linked ubiquitination of p53 at K24, which prevents p53 from binding to BAK/MCL-1 in the mitochondria, thereby keeping p53 away from the mitochondria and protect cells from spontaneous apoptosis. Importantly, we uncover that such inhibitory mechanism by TRAF6 is shut off upon genotoxic stress, empowering p53 to move to the mitochondria to initiate apoptosis, by genotoxic stress-induced TRAF6 phosphorylation and subsequent translocation. Interestingly, our result is consistent with a study showing that the amount of ubiquitinated p53 in mitochondria is too low to be detected by western blot, while ubiquitinated p53 in the cytoplasm is readily detectable (Marchenko et al., 2007). Remarkably, loss of TRAF6 in various mouse

tissues and human cancer cells is sufficient to drive p53 mitochondria localization and spontaneous apoptosis even without any genotoxic stress. Thus, our study defines the critical mechanism by which genotoxic stress drives cytosolic p53 mitochondrial translocation and subsequent apoptosis by inhibiting TRAF6-mediated K63-linked ubiquitination of p53 at K24 in cytosol. Also, this inhibitory mechanism may be potentially heightened in colorectal cancer patients with WT p53, because TRAF6 overexpression in such setting indicates less apoptosis and could predict poor patient survival. In addition, p53 K24R mutant displays heightened cell-killing effect under TRAF6 overexpression condition compared with p53 WT, while p53 K24R displays deficiency in p53 target gene expression. Thus, mitochondrial action may be the major tumor suppressive pathway of p53 when it is overexpressed in the in vitro cell growth assay and in vivo xenograft tumor growth assay using H1299 cell line without the treatment of genotoxic agents. Besides our finding showing a direct role of K63-linked ubiquitination in inhibiting the binding between p53 and BAK, mitochondrial BAK is also suppressed by MCL-1 (Leu et al., 2004). Our data show that K63-linked ubiquitination also inhibited the binding between p53 and MCL-1, resulting in the BAK suppression by MCL-1. In vitro binding assay further indicated that p53 ubiquitination could directly inhibit the binding of p53 to both BAK and MCL-1, suggesting that p53 ubiquitination inhibits p53-BAK binding both directly and indirectly through MCL-1. Because BCL-2/BCL-xL is also reported to bind and sequester cytosolic p53 and/or BAK, our study does not rule out the possibility that TRAF6-mediated cytosolic p53 ubiquitination may also regulate the crosstalk between BCL-2/BCL-xL and p53 and/or BAK. Although our study reveals that TRAF6 is a negative regulator for p53-mediated apoptosis in the mitochondrial pathway, we also find that TRAF6 is required for p53 acetylation, transcriptional activation, and the expression of its target genes, including p21 and Gadd45, upon genotoxic stress. It should be noted that the basal expression of p53 target genes before genotoxic stress and the expression of non-IR-inducible p53 target genes are not affected by TRAF6 (Figure S5A), consistent with the fact that K63-linked ubiquitination of p53 by TRAF6 in nucleus is induced only upon genotoxic stress. These data also demonstrate that the spontaneous apoptosis in TRAF6-deficient cells before genotoxic stress is not related to the transcriptional function of p53, because TRAF6 does not regulate p53 target expression at this time. The impairment in the induction of p21 and Gadd45 expression upon genotoxic stress in TRAF6-deficient cells sensitizes the cells to genotoxic stress, which is also consistent with the clinical data showing that TRAF6 is upregulated, is positively correlated with the expression of p21, and predicts poor response of colorectal cancer patients to chemotherapy and IR treatment (Figures 4D–4H). We have also noticed that the TRAF6 knockout affected many other p53 target genes involved in multiple biological processes, including apoptosis pathway (Puma and Bax), in the microarray data sets upon genotoxic stress (Figure S5A). First, this observation should not be confused with the role of TRAF6 in protecting cells from apoptosis before genotoxic stress, because the expression of those apoptosis genes is not affected by TRAF6 before genotoxic stress. More important, TRAF6 deficiency does not

completely wipe out their expression but rather reduces their expression to the level comparable with the WT cells without genotoxic stress, which is still sufficient to mediate cellular apoptosis process. Thus, the biological effect of TRAF6 knockout should not be simply considered as equal to the knockout of those apoptotic genes, which are demonstrated as essential in p53-mediated apoptosis (Jeffers et al., 2003). For example, the protein level of Bax is not much induced by genotoxic stress or affected by TRAF6 deficiency upon genotoxic stress (data not shown), and such basal level of Bax in TRAF6 deficiency may be sufficient to mediate apoptosis. In this regard, it is not surprising to see the increased cell death upon genotoxic stress in TRAF6-deficient cells, because it mainly reflects the loss of protective effect from p21 and Gadd45 against apoptosis. Thus, it is possible that the net outcome of TRAF6 regulated p53 transactivation upon genotoxic stress favors cell survival and could also contribute to the poor response to the CCRT in the colorectal cancer patients. TRAF6 is shown to be involved in the innate immune response and tumorigenesis by regulating TLR4/IL-1R and growth factorinduced AKT signaling (Chen, 2005; Yang et al., 2009, 2010a, 2010b). In this study, we further extend its role in regulating both transcription-dependent and transcription-independent function of p53 in apoptosis. Interestingly, TRAF6 is responsible for K63-linked ubiquitination of p53 at K24 residue in the nucleus upon genotoxic stresses. This event facilitates the interaction between p53 and p300 and is required for cell survival upon genotoxic stress through p21 and Gadd45 expression. This is in sharp contrast to Mdm2, which triggers K48-linked ubiquitination and degradation of p53. Consistent with the important role of TRAF6 in NF-kB signaling pathway (Dickson et al., 2004), NF-kB is predicted as one of the top potential transcription factors regulated by TRAF6 in our microarray data sets (Figure 6A). However, NF-kB and p53 are shown to antagonize each other in their target gene expression (Webster and Perkins, 1999). Thus it is unlikely that TRAF6 activates p53 target genes indirectly through its well-established role in NF-kB pathway. It is also unlikely that AKT signaling pathway promotes the p53 transactivation, because AKT is shown to enhance MDM2-mediated degradation of p53 (Ogawara et al., 2002). To exclude any potential effect of AKT and NF-kB pathway on the mitochondrial translocation of p53, we overexpressed constitutive active AKT (myristoylated AKT) or p65 (S276D) in TRAF6-knockdown cells and found that they cannot rescue the accumulation of p53 in the mitochondria in TRAF6-deficient cells (Figure S7A), further supporting that TRAF6 may directly regulate p53 mitochondrial translocation and such regulation is independent of NF-kB and AKT pathway. However, we are not excluding the possibility that TRAF6 protects cells from genotoxic stress, likely through the AKT and NF-kB pathway, because NF-kB or AKT activation upon genotoxic stress may also contribute to cell survival. We also demonstrated that TRAF6 is phosphorylated upon genotoxic stress at the ATM and ATR substrate consensus sites, as shown in the immunoprecipitation assay (Figures 1H, 1I, and S1O). However, without further experimental evidence, we cannot pinpoint which kinases are directly involved in genotoxic stress-induced TRAF6 phosphorylation and translocation. Thus, it would be of interest and significance to further explore this Molecular Cell 64, 803–814, November 17, 2016 811

Figure 7. Schematic Model Before genotoxic stress, p53 in the cytosol is bond by TRAF6 and K63-linked ubiquitinated by TRAF6. This ubiquitination inhibited the binding between p53 and BAK/MCL-1, thus keeping p53 away from mitochondria. In the nucleus, the level of TRAF6 and K63-linked ubiquitination of p53 is low, which keeps p53 activity at basal level. Upon genotoxic stress, p53 level is drastically increased, while TRAF6 level is reduced in cytosol, which results in the more nonubiquitinated p53 and allows p53 translocation to mitochondria for BAK oligomerization and subsequent apoptosis. Meanwhile, genotoxic stress enhances the level of TRAF6 in the nucleus and drives the K63-linked ubiquitination of p53, which facilitates the recruitment of p300 for p53 transactivation. Thus, p53 target genes, including p21 and GADD45 were actively transcribed and exert their functions in cell-cycle control and cell survival. In the TRAF6deficient condition, p53 accumulated in the mitochondria triggers Bak oligomerization and apoptosis under the unstressed condition. Upon genotoxic stress, TRAF6-deficient cells failed to activate p53 and express p21 and GADD45 and are more sensitive to genotoxic stress.

Total Thymus Cell Isolation Total thymus cells were isolated by mechanical disruption of thymus obtained from mice using two cover slides. Cells were then filtered and used in various assays.

aspect in the future to discover novel pathways in DNA damage response signaling. In summary, our study reveals important insights into how genomic and non-genomic p53 activity is regulated under genotoxic stress (Figure 7). On the basis of our findings, we propose a working model to delineate how genomic and non-genomic p53 activity is regulated in response to genotoxic stress (Figure S7B). Our findings unravel distinct patterns of K63-linked ubiquitination of p53 in various cellular compartments, which serves as an important molecular switch that enables p53 transcriptional activation in the nucleus and p53 transcription-independent function in the mitochondrial for apoptosis under stress conditions. This study also identifies TRAF6 as a cytosolic gatekeeper, which restricts p53 mitochondrial translocation and subsequent apoptosis. Thus, therapeutically targeting TRAF6 is a promising strategy to elicit p53-mediated apoptosis for cancer therapy and overcoming drug resistance.

In Vivo and In Vitro Ubiquitination Assays In vivo ubiquitination assay was performed as previously described (Yang et al., 2009). In brief, cells were transfected with the indicated plasmids for 48 hr and harvested by denatured buffer (6 M guanidine-HCl, 0.1 M Na2HPO4/NaH2PO4, 10 mM imidazole). The cell extracts were then incubated with nickel beads (NiNTA) for 3 hr, washed, and subjected to western blot analysis. In vitro ubiquitination assay was also performed as described (Yang et al., 2009). Briefly, recombinant TRAF6 (GST tag was removed by thrombin digestion) proteins and GST-p53 protein purified from bacteria were incubated with 0.5ug E1 (Boston Biochem), 1.5 mg ubiquitin (Boston Biochem), 1 mg UBC13/UEV1A (Boston Biochem), 1 mg UBCH5c (Boston Biochem), and 2.5 mM ATP in a final 30 mL reaction buffer (1.5 mM MgCl2, 5 mM KCl, 1 mM DTT, 20 mM HEPES [pH 7.4]) at 30 C for 2 hr. GST-p53 protein was then isolated from reaction by glutathione beads, and its ubiquitination was detected by western blot analysis.

A complete description of experimental and analysis methods can be found in Supplemental Experimental Procedures.

Mitochondrial Fractionation and Crosslinking Cells were cultured in DMEM high glucose and harvested for mitochondrial fractionation by mitochondrial isolation kit for mammalian cells (Pierce, Thermo Scientific) according to the glass homogenizer-based fractionation protocol. Purified mitochondria were then re-suspended in MRM-S buffer (250 mM sucrose, 10 mM HEPES, 1 mM ATP, 5 mM succinate, 0.08 mM ADP, 2 mM K2HPO4 at [pH 7.4]) for further analysis, as previously described (Wei et al., 2000). For crosslinking, mitochondria were incubated in PBS containing 1 mM bis(maleimido)hexane (BMH) (Invitrogen) under room temperature for 30 min and boiled in 13 SDS sample buffer for subsequent western blot analysis.

Cell Culture Cells were cultured in DMEM high glucose (Hyclone) containing 10% fetal bovine serum (FBS).

Gene Expression Microarray Total RNA was extracted and purified using RNeasy mini kit (Qiagen) according to the manufacturer’s manual. Microarray analysis was performed for total

EXPERIMENTAL PROCEDURES

812 Molecular Cell 64, 803–814, November 17, 2016

RNA on Illumina HumanHT12v4 platform or Illumina MouseWG-6v2 following Illumina’s standard procedure.

Bieging, K.T., Mello, S.S., and Attardi, L.D. (2014). Unravelling mechanisms of p53-mediated tumour suppression. Nat. Rev. Cancer 14, 359–370.

Cell Apoptosis Assay Briefly, cell were labeled by TUNEL assay kit (Roche) according to the manufacturer’s protocol.

Brady, C.A., Jiang, D., Mello, S.S., Johnson, T.M., Jarvis, L.A., Kozak, M.M., Kenzelmann Broz, D., Basak, S., Park, E.J., McLaughlin, M.E., et al. (2011). Distinct p53 transcriptional programs dictate acute DNA-damage responses and tumor suppression. Cell 145, 571–583.

Colony-Forming Assay Cell colonies stained with crystal violet were dissolved in DMSO and quantified by the A600 absorbance. Cellular Fractionation Cells were fractionated by homogenize in hypertonic buffer and followed by centrifuge. Drug Treatment and Xenograft Mouse Model Pifithrin-m (inhibitor of p53 mitochondrial translocation, Cayman Chemical, #10748) or pifithrin-a (inhibitor of p53-dependent transcriptional activation, Cayman Chemical, #13326) dissolved in 1% EtOH/30% PEG/1% Tween-80 was injected intraperitoneally into 1-week-old WT and TRAF6 / mice (10 mg/kg, every other day). After treatment for 1 week, thymus was then isolated for subsequent analysis. For the xenograft model, 0.5 million H1299 cells were injected into each nude mouse, and tumor size was monitored for 4 weeks (n = 5). All animal experiments were performed under Institutional Animal Care and Use Committee approval protocol. ACCESSION NUMBERS The accession number for the microarray data reported in this paper is GEO: GSE69730. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, seven figures, and one table and can be found with this article online at http://dx.doi.org/10.1016/j.molcel.2016.10.002. AUTHOR CONTRIBUTIONS X.Z. and H.-K.L. designed all the experiments and wrote the manuscript. X.Z., C.-F.L., L.Z., C.-Y.W., L.H., G.J., A.H.R., F.H., C.L., C.X., and X.X. performed the experiments. C.-Y.H., F.-J.T., C.-H.T., and K.W. provided scientific input and suggestions. All authors have read and discussed the manuscript. ACKNOWLEDGMENTS We thank Drs. Wei Gu, Susan Taylor, Tom Hamilton, Ed Harhaj, and Yi-Chieh Nancy for reagents. We also thank the members of Dr. Hui-Kuan Lin’s laboratory for their support and critical comments on this study. This work was supported by start-up funds from Wake Forest School of Medicine, the Endowed Professorship Fund from the Anderson Family, and National Institutes of Health grants (NIH R01CA182424 and NIH R01CA193813) to H.-K.L. This study was also partly supported by Ministry of Health and Welfare (DOH102TD-M-111-102001 and MOHW103-TD-B-111-05 to C.-F.L.). We are grateful to the BioBank of Chi Mei Medical Center for providing the tumor samples. Received: March 16, 2016 Revised: August 2, 2016 Accepted: September 30, 2016 Published: November 3, 2016 REFERENCES Arnoult, D. (2007). Mitochondrial fragmentation in apoptosis. Trends Cell Biol. 17, 6–12.

Brooks, C.L., and Gu, W. (2006). p53 ubiquitination: Mdm2 and beyond. Mol. Cell 21, 307–315. Brooks, C., Wei, Q., Feng, L., Dong, G., Tao, Y., Mei, L., Xie, Z.J., and Dong, Z. (2007). Bak regulates mitochondrial morphology and pathology during apoptosis by interacting with mitofusins. Proc. Natl. Acad. Sci. U S A 104, 11649–11654. Chen, Z.J. (2005). Ubiquitin signalling in the NF-kappaB pathway. Nat. Cell Biol. 7, 758–765. Chen, X., Ko, L.J., Jayaraman, L., and Prives, C. (1996). p53 levels, functional domains, and DNA damage determine the extent of the apoptotic response of tumor cells. Genes Dev. 10, 2438–2451. Chipuk, J.E., and Green, D.R. (2008). How do BCL-2 proteins induce mitochondrial outer membrane permeabilization? Trends Cell Biol. 18, 157–164. Chipuk, J.E., Kuwana, T., Bouchier-Hayes, L., Droin, N.M., Newmeyer, D.D., Schuler, M., and Green, D.R. (2004). Direct activation of Bax by p53 mediates mitochondrial membrane permeabilization and apoptosis. Science 303, 1010– 1014. David, R. (2012). Apoptosis: a lipid trigger of MOMP. Nat. Rev. Mol. Cell Biol. 13, 208–209. Dickson, K.M., Bhakar, A.L., and Barker, P.A. (2004). TRAF6-dependent NF-kB transcriptional activity during mouse development. Dev. Dyn. 231, 122–127. Dornan, D., Wertz, I., Shimizu, H., Arnott, D., Frantz, G.D., Dowd, P., O’Rourke, K., Koeppen, H., and Dixit, V.M. (2004). The ubiquitin ligase COP1 is a critical negative regulator of p53. Nature 429, 86–92. el-Deiry, W.S., Tokino, T., Velculescu, V.E., Levy, D.B., Parsons, R., Trent, J.M., Lin, D., Mercer, W.E., Kinzler, K.W., and Vogelstein, B. (1993). WAF1, a potential mediator of p53 tumor suppression. Cell 75, 817–825. Gartel, A.L., and Tyner, A.L. (2002). The role of the cyclin-dependent kinase inhibitor p21 in apoptosis. Mol. Cancer Ther. 1, 639–649. Grossman, S.R. (2001). p300/CBP/p53 interaction and regulation of the p53 response. Eur. J. Biochem. 268, 2773–2778. Gu, W., and Roeder, R.G. (1997). Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain. Cell 90, 595–606. Haupt, Y., Rowan, S., Shaulian, E., Vousden, K.H., and Oren, M. (1995). Induction of apoptosis in HeLa cells by trans-activation-deficient p53. Genes Dev. 9, 2170–2183. Haupt, Y., Maya, R., Kazaz, A., and Oren, M. (1997). Mdm2 promotes the rapid degradation of p53. Nature 387, 296–299. Jeffers, J.R., Parganas, E., Lee, Y., Yang, C., Wang, J., Brennan, J., MacLean, K.H., Han, J., Chittenden, T., Ihle, J.N., et al. (2003). Puma is an essential mediator of p53-dependent and -independent apoptotic pathways. Cancer Cell 4, 321–328. Jones, S.N., Roe, A.E., Donehower, L.A., and Bradley, A. (1995). Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53. Nature 378, 206–208. Kastan, M.B., Zhan, Q., el-Deiry, W.S., Carrier, F., Jacks, T., Walsh, W.V., Plunkett, B.S., Vogelstein, B., and Fornace, A.J., Jr. (1992). A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell 71, 587–597. Kokontis, J.M., Wagner, A.J., O’Leary, M., Liao, S., and Hay, N. (2001). A transcriptional activation function of p53 is dispensable for and inhibitory of its apoptotic function. Oncogene 20, 659–668. Kruse, J.P., and Gu, W. (2009). Modes of p53 regulation. Cell 137, 609–622. Kubbutat, M.H., Jones, S.N., and Vousden, K.H. (1997). Regulation of p53 stability by Mdm2. Nature 387, 299–303.

Molecular Cell 64, 803–814, November 17, 2016 813

Leng, R.P., Lin, Y., Ma, W., Wu, H., Lemmers, B., Chung, S., Parant, J.M., Lozano, G., Hakem, R., and Benchimol, S. (2003). Pirh2, a p53-induced ubiquitin-protein ligase, promotes p53 degradation. Cell 112, 779–791.

Nakano, K., and Vousden, K.H. (2001). PUMA, a novel proapoptotic gene, is induced by p53. Mol. Cell 7, 683–694.

Leu, J.I., and George, D.L. (2007). Hepatic IGFBP1 is a prosurvival factor that binds to BAK, protects the liver from apoptosis, and antagonizes the proapoptotic actions of p53 at mitochondria. Genes Dev. 21, 3095–3109.

Oda, E., Ohki, R., Murasawa, H., Nemoto, J., Shibue, T., Yamashita, T., Tokino, T., Taniguchi, T., and Tanaka, N. (2000). Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis. Science 288, 1053– 1058.

Leu, J.I., Dumont, P., Hafey, M., Murphy, M.E., and George, D.L. (2004). Mitochondrial p53 activates Bak and causes disruption of a Bak-Mcl1 complex. Nat. Cell Biol. 6, 443–450.

Ogawara, Y., Kishishita, S., Obata, T., Isazawa, Y., Suzuki, T., Tanaka, K., Masuyama, N., and Gotoh, Y. (2002). Akt enhances Mdm2-mediated ubiquitination and degradation of p53. J. Biol. Chem. 277, 21843–21850.

Leu, J.I., Pimkina, J., Frank, A., Murphy, M.E., and George, D.L. (2009). A small molecule inhibitor of inducible heat shock protein 70. Mol. Cell 36, 15–27.

Pietsch, E.C., Leu, J.I., Frank, A., Dumont, P., George, D.L., and Murphy, M.E. (2007). The tetramerization domain of p53 is required for efficient BAK oligomerization. Cancer Biol. Ther. 6, 1576–1583.

Levine, A.J., and Oren, M. (2009). The first 30 years of p53: growing ever more complex. Nat. Rev. Cancer 9, 749–758. Lu, X., Yang, C., Hill, R., Yin, C., Hollander, M.C., Fornace, A.J., Jr., and Van Dyke, T. (2008). Inactivation of gadd45a sensitizes epithelial cancer cells to ionizing radiation in vivo resulting in prolonged survival. Cancer Res. 68, 3579–3583. Ly, J.D., Grubb, D.R., and Lawen, A. (2003). The mitochondrial membrane potential (deltapsi(m)) in apoptosis; an update. Apoptosis 8, 115–128. Marchenko, N.D., Wolff, S., Erster, S., Becker, K., and Moll, U.M. (2007). Monoubiquitylation promotes mitochondrial p53 translocation. EMBO J. 26, 923–934.

Tang, Y., Zhao, W., Chen, Y., Zhao, Y., and Gu, W. (2008). Acetylation is indispensable for p53 activation. Cell 133, 612–626. Vaseva, A.V., and Moll, U.M. (2009). The mitochondrial p53 pathway. Biochim. Biophys. Acta 1787, 414–420. Vousden, K.H., and Prives, C. (2009). Blinded by the light: the growing complexity of p53. Cell 137, 413–431. Vousden, K.H., and Ryan, K.M. (2009). p53 and metabolism. Nat. Rev. Cancer 9, 691–700. Webster, G.A., and Perkins, N.D. (1999). Transcriptional cross talk between NF-kappaB and p53. Mol. Cell. Biol. 19, 3485–3495.

Mihara, M., Erster, S., Zaika, A., Petrenko, O., Chittenden, T., Pancoska, P., and Moll, U.M. (2003). p53 has a direct apoptogenic role at the mitochondria. Mol. Cell 11, 577–590.

Wei, M.C., Lindsten, T., Mootha, V.K., Weiler, S., Gross, A., Ashiya, M., Thompson, C.B., and Korsmeyer, S.J. (2000). tBID, a membrane-targeted death ligand, oligomerizes BAK to release cytochrome c. Genes Dev. 14, 2060–2071.

Miyashita, T., Krajewski, S., Krajewska, M., Wang, H.G., Lin, H.K., Liebermann, D.A., Hoffman, B., and Reed, J.C. (1994). Tumor suppressor p53 is a regulator of bcl-2 and bax gene expression in vitro and in vivo. Oncogene 9, 1799–1805.

Yang, W.L., Wang, J., Chan, C.H., Lee, S.W., Campos, A.D., Lamothe, B., Hur, L., Grabiner, B.C., Lin, X., Darnay, B.G., and Lin, H.K. (2009). The E3 ligase TRAF6 regulates Akt ubiquitination and activation. Science 325, 1134–1138.

Moll, U.M., Wolff, S., Speidel, D., and Deppert, W. (2005). Transcription-independent pro-apoptotic functions of p53. Curr. Opin. Cell Biol. 17, 631–636.

Yang, W.L., Wu, C.Y., Wu, J., and Lin, H.K. (2010a). Regulation of Akt signaling activation by ubiquitination. Cell Cycle 9, 487–497.

Montes de Oca Luna, R., Wagner, D.S., and Lozano, G. (1995). Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53. Nature 378, 203–206.

Yang, W.L., Zhang, X., and Lin, H.K. (2010b). Emerging role of Lys-63 ubiquitination in protein kinase and phosphatase activation and cancer development. Oncogene 29, 4493–4503.

814 Molecular Cell 64, 803–814, November 17, 2016