|Home | About | Journals | Submit | Contact Us | Français|
As a key cellular regulatory protein p53 is subject to tight regulation by several E3 ligases. Here we demonstrate the role of HECT domain E3 ligase, WWP1, in regulating p53 localization and activity. WWP1 associates with p53 and induces p53 ubiquitination. Unlike other E3 ligases, WWP1 increases p53 stability; inhibition of WWP1 expression or expression of a ligase-mutant form results in decreased p53 expression. WWP1-mediated stabilization of p53 is associated with increased accumulation of p53 in cytoplasm with a concomitant decrease in its transcriptional activities. WWP1 effects are independent of Mdm2 as they are seen in cells lacking Mdm2 expression. Whereas WWP1 limits p53 activity, p53 reduces expression of WWP1, pointing to a possible feedback loop mechanism. Taken together, these findings identify the first instance of a ubiquitin ligase that causes stabilization of p53 while inactivating its transcriptional activities.
The tumor suppressor protein p53 is a primary coordinator of cellular responses to a wide range of stresses (Oren 2003). After genotoxic stress, p53 is rapidly activated and either promotes cell growth arrest or apoptosis, depending on the type, strength and duration of the stimulus (reviewed in Oren 2003; Senegupta & Harris 2005; Poyurovsky & Prives 2006). Under normal growth conditions, however, the level of p53 expression and activity is kept under tight control, preventing its activity under non-warranted conditions. Such tight regulation is for the most part mediated by limiting p53 stability by one of several ubiquitin ligases, including Mdm2, ARF-BP1, Pirh1 and the component of the signalosome, Cop9, which efficiently targets p53 to ubiquitination and proteasome-dependent degradation (Haupt et al., 1997; Kubbutat et al., 1997; Doran et al., 2004; Leng et al., 2003; Shmueli & Oren 2005; Harris & Levine 2005; Brooks & Gu 2003; Aschroft & Vousden 1999). Common to these ligases is their role in limiting p53 stability and activity. In searching for protein ligases with different effects on p53 activity, we identified the HECT domain E3 ligase WWP1, whose interaction with p53 increases its stability while reducing its transcriptional activity.
WW domain-containing protein 1 (WWP1) was first identified as a novel protein based on its WW modules - a 35-40 amino acid region exhibiting high affinity towards the PY motif, a proline-rich sequence followed by a tyrosine residue (Verdecia et al., 2003). WWP1 shares a characteristic domain organization with the E3 ligases Nedd4 and Smurfs, which consists of a C2 domain, 2-4 WW domains, and a HECT domain (Kasanov et al., 2001). Although WWP1 has been shown to function as an E3 ubiquitin ligase, only few substrates have been identified. Among these are Krüppel-like factors (Conkright et al., 2001; Zhang et al., 2004; Chen et al., 2005) and Smad7 (Moren et al., 2005), which have been shown to be regulated by WWP1. Additionally, it has been reported that WWP1 is essential for embryonic development in Caenorhabditis elegans (Huang et al., 2000).
WWP1 is a member of the Nedd4 family of E3 ligases, which includes Nedd4, Itch and WWP2 (Sudol & Hunter 2000; Ingham et al., 2004). Members of the Nedd4 family have been shown to be involved in regulating cellular signaling and protein sorting (Galinier et al., 2002). Nedd4 was initially found to regulate cell surface stability of the epithelial sodium channel (ENaC) (Staub et al., 1996). Itch has been shown to be regulated by JNK and to play a role in TNF signaling (Chang et al., 2006). Additionally, Itch has been shown to interact with, ubiquitinate, and degrade p73 (Rossi et al., 2005). WWP2 has been shown to regulate the stability of the Oct-4 transcription factor, a master regulator affecting the fate of pluripotent embryonic stem cells (Xu et al., 2004). WWP1, WWP2 and Itch have been implicated in vascular protein sorting, which is exploited by enveloped viruses (Martin-Serrano et al., 2005).
Earlier studies have implicated the role of the proline-rich domain of p53 in its activities (Walker & Levine 1996; Zhu et al., 2000). For example, the phosphorylation of Thr81 within this domain has been shown to be important in recruitment of the prolyl isomerase Pin1 (Zacchi et al., 2002), which contributes to p53 stability and activity following stress. In searching for E3 ligases that may associate with the proline-rich domains (Bedford et al., 2000; Verdecia et al., 2000), we have compared different members of the 4 WW-domains for their association with p53. Among those, WWP1 exhibited the strongest association (see data below). To map the region on p53 required for interaction with WWP1, we performed a GST-pull down assay. Surprisingly, WWP1 associated, not with the proline-rich region (1-100) but rather with the DNA-binding region of p53 (100-295) (Figure 1a). However, the proline-rich region was required for efficient interaction in vivo, because a mutant of p53 that lacks this region (aa 68-91), p53Δ6, showed a weaker association with WWP1 in a co-immunoprecipitation assay (Figure 1b). Another mutant which lacks a region downstream of the proline domain (aa 97-116), p53Δp7, still efficiently bound to WWP1. This observation suggests that the conformation of p53, generated in the presence of the proline-rich domain, is important for efficient association with WWP1. Since WWP1 is part of a family of HECT-E3 ligases that contain a C-terminal HECT domain and N-terminal WW domains, we next determined whether other members of the Nedd4 family can also associate with p53. GST-pull down assay using full-length p53 and the four members of the Nedd4 family revealed that WWP1 displayed the strongest association. WWP2 showed weaker association whereas Nedd4 and Itch showed no association (Figure 1c). These data suggest that WWP1 is the primary Nedd4 family member that associates with p53 and that both the DNA-binding and the proline-rich domains of p53 are required for this association.
As WWP1 is an E3 ligase, we next determined the possible role of WWP1 in p53 stability. To exclude the possible role of Mdm2, p53-/-/Mdm2-/- mouse embryo fibroblasts (DKO MEFs) were used. Co-expression of WWP1 with p53 in the DKO MEFs increased p53 steady-state levels (Figure 2a). A ligase-dead mutant of WWP1 (C883A; where the catalytic cysteine residue is mutated to arginine) failed to stabilize p53. In fact, p53 stability was remarkably reduced upon expression of mutant WWP1, suggesting that it serves as an efficient dominant negative (Figure 2a). Consistent with this conclusion is the observation that the effect of the WWP1 mutant could be attenuated upon addition of the proteasome inhibitor MG132, suggesting that the WWP1 mutant increases p53 degradation (Figure 2a). These data suggest that stabilization of p53 by WWP1 requires its E3 ligase activity. WWP1 stabilized p53 in a dose-dependent manner in the DKO MEFs (Figure 2b), suggesting an MDM2-independent effect on p53 stability. Furthermore, WWP1 also affected endogenous p53 steady-state levels; in U2OS cells, wild-type WWP1 led to an increase in p53 levels, whereas the ligase mutant WWP1 decreased p53 levels (Figure 2c). Again, proteasome inhibitors could attenuate the effect of mutant WWP1 on p53 and the levels of hdm2 were unaffected by expression of WWP1. To determine whether WWP1 mediates increase in p53 levels post-transcriptionally, we performed a cycloheximide chase analysis in DKO MEFs. p53 alone had a half-life of approximately 1.5h. Upon co-expression with WWP1, the half-life of p53 increased to more than 3h (Figure 2d). These data suggest that WWP1 increases p53 levels by prolonging its half-life. We next sought to determine whether WWP1 is required for p53 stability. Using shRNAi against WWP1, we observed that knocking down exogenously expressed WWP1 led to a corresponding decrease in p53 levels (Figure 2e). Furthermore, similar results were obtained using endogenous proteins; knockdown of endogenous WWP1 lead to a decrease in p53 levels (Figure 2f). These results are consistent with the effects observed with mutant WWP1: in both cases, attenuating WWP1 activity, either by mutation or knockdown, decreased p53 stability.
To address the possible mechanism of WWP1’s effect on p53 stability, we determined WWP1’s effect on p53 ubiquitination. Expression of WWP1 in DKO MEFs revealed that wild-type WWP1 facilitated p53 ubiquitination, whereas the ligase mutant could not ubiquitinate p53 (Figure 3a). The level of ubiquitination of p53 by WWP1 was modest compared to that of p53 by Mdm2. Of note, however, whereas expression of WWP1 increased steady state levels of p53, Mdm2 decreased them (compare lanes 2 and 5 in Figure 3a). WWP1’s ability to facilitate ubiquitination of p53 was confirmed by performing an in vitro ubiquitination assay (Figure 3b). In this assay, WWP1 promoted ubiquitination of p53 to a degree similar to that seen with Mdm2; in both cases, in addition to polyubiquitination, a clear monoubiquitinated form of p53 was seen.
Given WWP1’s effect on ubiquitination-dependent stabilization of p53, we next explored its effect on p53 transcriptional activity using a luciferase construct driven by the p21-promoter. Expression of wild-type WWP1 led to a dose-dependent decrease in p21 transactivation by p53, whereas the mutant WWP1 had no effect (Figure 3c). This result was surprising in light of our finding that WWP1 increases p53 stability. To determine a possible mechanism that may explain these observations, we monitored localization of p53 after WWP1 expression. The primary localization of p53 within the nucleus turned out to be cytosolic upon expression of wild-type WWP1 (Figure 3d). Mutant WWP1 was unable to facilitate cytosolic localization of p53. These findings suggest that WWP1 promotes nuclear export of p53. This conclusion is supported by the observation that treatment with the nuclear-export inhibitor Leptomycin B attenuated WWP1’s ability to cause redistribution of p53. Consistent with these findings is the observation that WWP1 elicited a similar effect on Smad7. Wild-type, but not mutant WWP1, causes export of Smad7 from the nucleus to the cytosol (Moren et al., 2005). Taken together, these findings suggest that increase in p53 levels after WWP1 expression is associated with its export to the cytosol, which results in decreased p53 transcriptional activities. This is further supported by the finding that knockdown of WWP1 leads to an increase in the levels of p21 expression, a marker for increased p53 activity (Figure 2f).
Interestingly, increased expression of p53 in DKO MEFs decreased WWP1 levels (Figure 4a). This effect could not be blocked by addition of MG132, suggesting that the p53-mediated decrease in WWP1 is not proteasome-dependent. This effect was substantiated in cells that express p53 or lack such expression. p53-/- MEFs exhibited higher WWP1 levels that were reduced upon reconstitution with p53 (Figure 4b). Since MG132 could not block p53’s effect on WWP1, we tested the possibility that p53 affects WWP1 transcriptionally. To this end, levels of WWP1 transcripts were followed in p53+/+ and p53-/- MEFs under normal growth conditions and after exposure to UV-irradiation or γ-irradiation. UV-irradiation decreased transcript levels in p53+/+ cells. This effect was time dependent and not seen within the first 4h in p53-/- cells (Figure 4c). This time frame corresponded to activation of p53, as was assessed by monitoring p53 target genes p21 and survivin. In fact, WWP1 showed similar changes to those seen with survivin, which is known to be transcriptionally repressed by p53 (Hoffman et al., 2002; Wang et al., 2004). Similar results were obtained when treating p53+/+ or p53-/- MEFs with γ-irradation (Figure 4d). These results suggest a possible feedback mechanism between p53 and WWP1.
Our findings point to the role of WWP1 in inactivation of p53 transcriptional activities under non-stressed conditions. At this point we cannot exclude the possibility that localization of p53 to the cytosol by WWP1 does not serve an as yet undisclosed function of p53 within the cytosol, which is expected in light of its increased stability. Given that WWP1 (along with other members of this ligase family) has been implicated in vacuolar protein sorting (Martin-Serrano et al., 2005), it would be of interest to explore possible role of WWP1-modified p53 in this process. WWP1-modified p53 is expected to carry a non-canonical ubiquitin chain, given that such ubiquitination results in its stabilization rather than degradation. Upon stress, inactivation of WWP1 is expected to enable a complete activation of p53. The inverse correlation between WWP1 and p53 is expected to persist in cells that harbor functional p53 but not in those where p53 is mutated or deleted. Consistent with this hypothesis is the observation that tumor cells with inactive p53 exhibit elevated levels of WWP1 expression (http://symatlas.gnf.org).
Overall, our current findings highlight a previously undescribed mechanism for regulation of p53 transcriptional activities. The role of WWP1 in mediating increased expression but decreased activity of the tumor suppressor p53 represents a previously unrecognized regulatory mode. The effect of WWP1 is independent of Mdm2 but nevertheless resembles changes after mono-ubiquitination of p53 by low levels of Mdm2 (Li et al., 2002).
One would expect that WWP1 activity occurs primarily under conditions in which Mdm2 is not active or available; it is also plausible that elevated WWP1 expression as seen in human tumors may override the effect of Mdm2, thereby attenuating p53 activities. Further studies will be required to address such relationships as well as the mechanism underlying p53 nuclear export by WWP1 and the function of p53 in the cytosol under nonstressed conditions.
We thank Marius Sudol and Tony Hunter for providing us with WWP1 reagents, Stephen Jones for p53/Mdm2 DKO cells, and ZQ Pan, J Manfredi and M O’Connell for advice. A. Laine performed these studies as part of his work for the M.D./Ph.D. program at Mount Sinai School of Medicine, New York, NY. Support by NCI grant CA78419 to ZR is gratefully acknowledged.