Ubc13 increases levels of p53 protein.
In studying the possible role of Ubc13 in responses to cellular stress, we have observed that expression of Ubc13 affects the level and consequently the activity of p53. Expression of wild-type Ubc13, but not the catalytically inactive mutant form, increased expression of exogenous (Fig. ) and also endogenous (Fig. ) p53. The changes elicited by Ubc13 were dose dependent and were observed for different cell lines tested (Fig. ). Ubc13 did not affect levels of p53 transcripts, suggesting that effects on p53 protein are not attributed to increased transcription of the p53 gene (Fig. ). The effect of Ubc13 was not global, as other short-lived proteins, such as c-Myc, were not affected upon elevated expression of Ubc13 (Fig. ). These initial observations, which were confirmed with different cell lines, suggest that elevated expression of Ubc13 increases the steady-state level of p53 protein and that this effect requires Ubc13 conjugating enzyme activity.
FIG. 1. Ubc13 increases p53 levels. (a) Ubc13 stabilizes exogenous p53. HA-Ubc13 (2 μg) was cotransfected with 0.2 μg p53 into H1299 cells, and Western blot analysis was performed on whole-cell extracts (40 μg) with anti-p53 monoclonal (more ...) Ubc13 prolongs p53's half-life while attenuating Hdm2's effect on p53.
To further explore the possible mechanism for increased p53 levels upon overexpression of Ubc13, we monitored the half-life of the p53 protein. Elevated expression of the WT but not the mutant form of Ubc13 doubled the half-life of p53 from 30 to over 60 min (Fig. ).
FIG. 2. Ubc13 increases p53 stability by attenuating Mdm2-mediated degradation. (a) Ubc13 increases p53 half-life. HCT116 p53+/+ cells were transfected with either 2 μg empty plasmid or HA-Ubc13. Cells were treated with cycloheximide (CHX) (more ...)
Since Hdm2 is the primary ligase implicated in the regulation of p53 stability (14
), we next assessed the effect of Ubc13 on Hdm2. Expression of Ubc13 attenuated Hdm2's ability to decrease p53 protein levels, which required the E2-conjugating activity of Ubc13 (Fig. ). Similar results were seen with H1299 cells (data not shown). The ability of Ubc13 to attenuate Hdm2 activity was not due to disruption of the Hdm2-p53 complex, as Hdm2 association with p53 was not affected upon expression of Ubc13 in the presence of proteasome inhibitors (Fig. ). To further assess the possible role of Hdm2 in Ubc13's effects on p53, we next determined whether Ubc13 could affect p53 ubiquitination in the absence of Hdm2. In the presence of proteasome inhibitors, expression of Ubc13 and p53 in p53−/−/mdm2−/−
mouse embryo fibroblasts caused efficient mono- and diubiquitination of p53, as revealed from the nickel pulldown of His-tagged ubiquitinated p53 (Fig. , lane 2). This finding suggests that for its effects on p53, Ubc13 does not require the E3 ligase activity of Hdm2. Since Ubc13 is a ubiquitin-conjugating enzyme rather than a ligase, it is expected that such an effect would engage the activities of an E3 ligase. Of note, however, the expression of Ubc13 led to inhibition of Hdm2-mediated p53 ubiquitination (Fig. , compare lanes 3 and 4), which required Ubc13's own conjugating enzyme activity (Fig. , compare lanes 4 and 5). It is important to note that while Ubc13 causes an efficient decrease of p53 polyubiquitination by Hdm2, it does not affect mono- or diubiquitination of p53 (Fig. , compare lanes 3 and 4). These data suggest that Ubc13 attenuates Hdm2's ability to cause polyubiquitination of p53, thereby reducing its ability to degrade p53 under nonstress conditions. Collectively, the data presented here suggest that Ubc13's effects on p53 are likely to be mediated by an E3 ligase that is different than Hdm2 and that, through its association with or effect on p53, Ubc13 attenuates Hdm2 polyubiquitination of p53.
Ubc13 interacts with p53.
As the changes elicited by Ubc13 on p53 require its ubiquitin-conjugating activity, we examined Ubc13 association with and ubiquitination of p53. Association of Ubc13 with p53 was demonstrated in vitro by use of a GST pulldown assay (Fig. ). Interestingly, p53 associated with the intact E2 complex of Ubc13/Mms2 or Ubc13/Uev1A but not with Ubc13 alone. This finding suggests that a functional complex is required for association of Ubc13 with p53. p53 was shown to associate directly with Ubc13/Uev1A and not with Ubc13, MMS2, or Uev1A alone by use of bacterially expressed and purified components (Fig. ). Furthermore, p53's association was specific for the Ubc13/Uev1A E2 complex, because another noncanonical E2, UbcH7, did not associate with p53 (data not shown). The association of Ubc13 and p53 was confirmed in vivo by immunoprecipitating either Ubc13 (Fig. ) or p53 (Fig. ). These data demonstrate that Ubc13 associates with p53.
FIG. 3. Ubc13 interacts with p53. (a) p53 interacts with Ubc13 in vitro. A complex between GST-Ubc13 and either His-MMS or His-Uev1A was preformed by incubating the corresponding proteins for 16 h at 4°C. After a wash, in vitro-transcribed and -translated (more ...) Ubc13 regulates p53 ubiquitination.
Because mutant Ubc13 lacking conjugating activity failed to stabilize p53 levels (Fig. ) or affect Hdm2-mediated ubiquitination of p53 to the degree seen with wild-type Ubc13 (Fig. ), we hypothesized that Ubc13 requires its conjugating activity to facilitate the ubiquitination of p53. Whereas expression of wild-type Ubc13 did not alter the basal level of p53 ubiquitination in vivo, the expression of the Ubc13 mutant reduced p53 ubiquitination (Fig. ). Further, the expression of mutant Ubc13 reduced the levels of mono-, di-, and polyubiquitinated p53 (Fig. ). This was further supported by monitoring the basal level of p53 ubiquitination upon knockdown of Ubc13 expression; a reduction of Ubc13 levels by shRNA led to a decrease in the polyubiquitination and, to a lesser degree, the mono- and diubiquitination of p53 (Fig. ). These results suggest that Ubc13 contributes to the basal level of p53 ubiquitination.
FIG. 4. Ubc13 regulates p53 ubiquitination. (a and b) Ubc13 regulates p53 ubiquitination in vivo. p53−/−/mdm2−/− MEFs were transfected and/or infected with the indicated constructs. Cells were harvested under denaturing conditions, (more ...)
Ubc13 has been shown to facilitate the formation of noncanonical ubiquitin chains by utilizing lysine 63 of ubiquitin. Therefore, we sought to determine if lysine 63 was required for Ubc13's activities towards p53. Performing an in vivo ubiquitination reaction using wild-type ubiquitin or ubiquitin where either lysine 48 or lysine 63 was mutated to arginine (K48R or K63R mutant, respectively) revealed that although neither mutation altered the level of polyubiquitinated p53, both mutations led to a decrease in the level of diubiquitinated p53 by Ubc13, with the K63R mutation having a more pronounced effect (Fig. ). Interestingly, overexpression of the K63R mutant reduced the total level of p53 expression (Fig. ), suggesting that K63-mediated ubiquitination may stabilize p53. Indeed, the effect of K63R was pronounced, as it decreased the level of p53 even below the level seen upon overexpression of Ubc13 (Fig. , compare lanes 3 and 5). An appreciable change in the p53 levels for the K63R mutant (Fig. ) is mostly likely attributed to the different levels of p53 used. These data demonstrate that Ubc13 affects noncanonical p53 ubiquitination.
Ubc13 associates with the C terminus of p53.
The finding of p53 association with Ubc13 led us to determine the region on p53 that is required for Ubc13 binding and/or activity. To map the region on p53, we utilized different deletion or mutant constructs of p53 and monitored their association with Ubc13 in coimmunoprecipitation experiments. p53 lacking the C terminus (the Δ363-393 mutant) or mutated on the seven C-terminal lysines (the 7KR mutant) associated to a lesser degree with Ubc13 than did WT p53 (Fig. ). Surprisingly, a mutant of p53 which lacks the oligomerization domain (the Δ325-354 mutant) showed increased association with Ubc13 (Fig. , lane 3). Furthermore, while WT Ubc13 did not increase basal ubiquitination of WT p53, it caused an increase in the di- and triubiquitination of the Δ325-354 mutant in the absence of proteasome inhibitors (data not shown). These findings suggest that, through its higher affinity towards a monomeric form of p53, Ubc13 elicits a more pronounced effect on di- and triubiquitination of p53.
FIG. 5. Ubc13 highly associates with monomeric p53. The C-terminal region of p53 is required for Ubc13 interaction. H1299 cells were cotransfected with 2 μg wild-type (HA-Ubc13wt) or mutant (HA-Ubc13ca) Ubc13 and WT p53, p53 whose oligomerization domain (more ...) Ubc13 facilitates the formation of monomeric p53.
The greater association of Ubc13 with the tetramerization mutant of p53 raised the possibility that Ubc13's association with the monomeric p53 may inhibit p53 tetramerization. To test this possibility, we monitored the sizes of p53-containing protein complexes that were prepared from Ubc13- and p53-expressing H1299 cells and analyzed by gel sieve chromatography. Whereas p53 was found to localize within high-molecular-mass fractions of 150 to 600 kDa, coexpression of p53 with Ubc13 resulted in formation of a new p53 pool within the lower-molecular-mass fractions with an estimated mass of 50 to 60 kDa (i.e., monomers) (Fig. ). Coexpression of mutant Ubc13 was unable to facilitate the shift of p53; rather, it caused a shift in the p53 pool to even higher molecular mass fractions (Fig. ). These observations suggest that the catalytic activity of Ubc13 is required for its ability to promote the formation of monomeric p53 forms. The ability of Ubc13 to promote the formation of a monomeric pool of p53 was further assessed in vitro. To this end, lysates containing exogenously expressed p53 were incubated with recombinant Ubc13/Uev1A. Whereas mock treatment did not alter p53 distribution, addition of recombinant Ubc13/Uev1A promoted the formation of a pool of monomeric p53 (data not shown). These observations were further confirmed by monitoring the oligomerization status of endogenous p53 upon expression of wild-type or mutant Ubc13; wild-type but not mutant Ubc13 was able to induce the formation of monomeric p53 (Fig. ). These results collectively demonstrate that Ubc13 acts to shift p53 from oligomeric complexes to a more monomeric form and requires Ubc13 catalytic activity.
FIG. 6. Ubc13 affects p53 oligomerization status. (a) Expression of Ubc13 increases monomeric levels of p53. H1299 cells were transfected with 1.5 μg p53 or p53 and 2 μg WT Ubc13, and then equal amounts of cellular lysate (400 μg) were (more ...) Ubc13 affects the localization and transcriptional activity of p53.
Former studies have revealed that as a monomer, p53 is subjected to efficient export from the nucleus to the cytoplasm, which coincides with reduced transcriptional activity (6
). Thus, to further assess the functional implications of the Ubc13-dependent monomerization of p53, we monitored p53 localization, transcriptional activity, and ability to induce apoptosis. Coexpression of p53 and Ubc13 in p53−/−/mdm2−/−
MEFs revealed a marked effect on p53 localization. Expression of wild-type but not mutant Ubc13 substantially increased the percentage of cells exhibiting cytoplasmic p53 localization (Fig. ). Treatment with leptomycin B, a nuclear export inhibitor, blocked Ubc13's ability to cause cytoplasmic accumulation of p53 (Fig. ), suggesting that Ubc13 promotes nuclear export of p53. These observations suggest that Ubc13 expression increases cytoplasmic accumulation of p53.
FIG. 7. Ubc13 affects p53 activity and localization. (a) Ubc13 promotes nuclear export of p53. p53−/−/mdm2−/− MEFs were transfected with the indicated constructs (100 ng p53 or 500 ng Ubc13). Some cells were treated with leptomycin (more ...)
We next monitored p53 transcriptional activity by using promoter sequences of p53 target genes p21, Bax, and PUMA linked to a luciferase reporter gene in p53−/−/mdm2−/− MEFs. Although p53 expression led to activation of these target genes, Ubc13 expression inhibited (30 to 50%) luciferase activity driven by all three target genes (Fig. ; also data not shown). Furthermore, treatment with leptomycin B was able to attenuate Ubc13-mediated inhibition of p53 transactivation activity (Fig. ). Ubc13 was not able to affect the transcriptional activities or localization of the C-terminal mutant p53 forms with which it weakly associated (see Fig. S1a in the supplemental material; also data not shown). Expression of Ubc13 did not inhibit the transactivation of β-catenin, which was used as a representative of non-p53 target genes (see Fig. S1b in the supplemental material), thereby pointing to the specificity of Ubc13's effect. Of interest, mutant Ubc13 was also able to decrease p53-mediated transcription (data not shown), although mutant Ubc13 reduces p53 ubiquitination and does not cause its nuclear export, as seen with the WT form of p53 (Fig. ). These findings suggest that association with p53 (which is seen for both WT and mutant forms of Ubc13) suffices to interfere with its transcriptional activities, whereas changes in tetramer formation, nuclear exclusion, and stability require Ubc13's conjugating enzyme activity. Thus, although WT Ubc13 increases p53 stability, it limits transcriptional activity of p53, pointing to a yet to be identified function of p53 upon its nuclear exclusion by Ubc13.
Additionally, the effect of Ubc13 on p53-mediated apoptosis was determined. Saos-2 cells were transfected with either p53 alone or p53 with wild-type or mutant Ubc13. Expression of p53 led to a marked increase in the number of cells undergoing apoptosis, which was enhanced upon coexpression with mutant Ubc13 and attenuated upon coexpression with wild-type Ubc13 (Fig. ). Irradiation (IR) of cells that were inhibited for Ubc13 expression caused their accumulation within the G2/M phase of the cell cycle (data not shown). Since mutant Ubc13 attenuates p53 transcriptional activity, its ability to promote apoptosis may be independent of p53 transcriptional activities. To further assess the role of Ubc13 in regulating the activity of endogenous p53, we have monitored the levels of p53 and its target genes in cells that were inhibited for Ubc13 expression. Knockdown of Ubc13 by shRNA against Ubc13 that was infected into U2OS cells decreased the levels of p53, whereas the levels of p21 and Bax were increased (Fig. ). Similar results were seen with HCT116 cells (data not shown). These results are in agreement with the luciferase studies in which Ubc13 expression leads to an inhibition of p53 transcriptional activities; thus, knockdown of Ubc13 relieved such suppression, resulting in increased p53 transcriptional activities. Together, these findings suggest that Ubc13 reduces p53 tetramerization, resulting in its nuclear export and consequently reducing its transcriptional activity and ability to induce apoptosis.
FIG. 8. Ubc13 affects p53 apoptotic activity. (a) Ubc13 attenuates p53-mediated apoptosis. Saos-2 cells were transfected with 1 μg p53 alone or with p53 and 2 μg wild-type (Ubc13wt) or mutant (Ubc13ca) Ubc13. Thirty-six hours after transfection, (more ...) Ubc13 associates with polysomes, and its overexpression affects the ribosomal distribution of p53.
Several reports indicate that cytoplasmic p53 protein associates with the ribosomes through covalent interaction with 5.8S rRNA (11
). Since WT Ubc13 increased the levels and stability of cytoplasmic p53, we tested the possibility that Ubc13 may perturb ribosomal distribution of p53. To this end, we transfected U2OS cells with WT Ubc13, the C83A mutant, or empty vector and fractionated ribosomal pellets on a continuous (10 to 40%) sucrose gradient. Expression of wild-type Ubc13 lead to an increase in p53 levels (see Fig. S1c in the supplemental material). Surprisingly, the distributions of WT and mutant Ubc13 protein across the gradient were markedly different. Wild-type Ubc13 associated with heavier polysomes (Fig. , right panel, fractions 8, 9, and 10), while the mutant form of Ubc13 was found only in submonosomal fractions (Fig. , middle panel, fractions 1 to 4). Significantly, cells overexpressing the wild-type form of Ubc13 exhibited a substantial increase in the amount of p53 associated with polysomes (Fig. , right panel, fractions 7 and 8) compared to the distribution of p53 in vector-transfected cells (Fig. , left panel, fractions 1 to 5). In contrast, the mutant form of Ubc13, which lacks ubiquitin-conjugating activity, failed to induce a comparable shift of p53 protein to the heavier polysomes (Fig. ). In the latter case, p53 protein was found mainly in the submonosomal and monosomal fractions (Fig. , middle panel, fractions 1 to 5) and yielded a profile identical to that of vector-transfected cells (Fig. , left panel). Levels of expression for both endogenous and exogenous Ubc13 are shown in Fig. S1d in the supplemental material.
Importantly, expression of the aforementioned constructs did not affect global ribosomal profiles as monitored by absorbance at 254 nm or the distribution of rRNAs or ribosomal protein S6 across the gradient (Fig. ). The latter data indicate that the association of p53 with the heavier polysomes, upon forced expression of the wild-type form of Ubc13, is not due to the major perturbations in polysome assembly.
Taken together, these data indicate that Ubc13 associates with the polysomes. This association requires the ubiquitin-conjugating activity of Ubc13. Furthermore, overexpression of the wild-type form of Ubc13 leads to increased association of p53 protein with the polysomes. The latter finding suggests that through its ubiquitin-conjugating activity Ubc13 recruits (and possibly modifies) p53 to polysomes or affects p53 organization (monomeric rather tetrameric form) within the polysomes. The presence of p53 within the polysomes is expected to impact a translational program(s) in a Ubc13-dependent manner.
IR attenuates Ubc13's effect on p53.
Given the role of p53 in DNA damage response, we next monitored the effect of Ubc13 on the formation of monomeric p53 following treatment with γ-irradiation. To this end, cells were subjected to 10 Gy of γ-irradiation before proteins were prepared and subjected to fractionation. While Ubc13 caused a shift in the p53 pool in the nontreated sample, this shift was no longer apparent after γ-irradiation (Fig. ). In fact, p53 was found in higher-molecular-mass complexes following IR (Fig. , compare fractions 11 to 13 prior to and after IR), resembling the changes seen upon expression of the Ubc13 mutant (Fig. ).
FIG. 9. p53 regulates Ubc13 stability. (a) IR attenuates Ubc13's activity on p53. H1299 cells were transfected with the indicated constructs. Cells were treated with γ-irradation (10 Gy) and then harvested 4 h later. Lysates were subjected to fractionation (more ...) Regulation of Ubc13 expression by p53.
The notion that DNA damage induced by IR attenuates Ubc13's ability to promote p53 monomerization (Fig. ) led us to explore possible changes in Ubc13 following DNA damage. Treatment of cells with IR led to a decrease in Ubc13 levels, both exogenous and endogenous (Fig. ). A decrease in Ubc13 expression levels was also seen after UV irradiation (Fig. ). IR-induced decreases in exogenous as well as endogenous Ubc13 expression were observed only with cells that express p53 (Fig. ). Expression of wild-type p53 caused greater suppression of endogenous Ubc13 expression (Fig. ) than did expression of a transcriptionally inactive mutant (p5322/23/175) (Fig. , compare lanes 5 and 6), suggesting that transcriptional activity of p53 contributes to the regulation of Ubc13 levels. Hdm2 was not required for p53-mediated down-regulation of Ubc13, shown by experiments using p53−/−/mdm2−/− MEFs (data not shown). Changes in Ubc13 protein expression elicited by p53 occur after a delay of 12 to 24 h, suggesting that this is part of a secondary response.