TRIM25 has a dual function in the p53/Mdm2 circuit

Abstract

P53 is an important tumor suppressor that, upon activation, induces growth arrest and cell death. Control of p53 is thus of prime importance for proliferating cells, but also for cancer therapy, where p53 activity contributes to the eradication of tumors. Mdm2 functionally inhibits p53 and targets the tumor suppressor protein for degradation. In a genetic screen, we identified TRIM25 as a novel regulator of p53 and Mdm2. TRIM25 increased p53 and Mdm2 abundance by inhibiting their ubiquitination and degradation in 26 S proteasomes. TRIM25 co-precipitated with p53 and Mdm2 and interfered with the association of p300 and Mdm2, a critical step for p53 polyubiquitination. Despite the increase in p53 levels, p53 activity was inhibited in the presence of TRIM25. Downregulation of TRIM25 resulted in an increased acetylation of p53 and p53-dependent cell death in HCT116 cells. Upon genotoxic insults, TRIM25 dampened the p53-dependent DNA damage response. The downregulation of TRIM25 furthermore resulted in massive apoptosis during early embryogenesis of medaka, which was rescued by the concomitant downregulation of p53, demonstrating the functional relevance of the regulation of p53 by TRIM25 in an organismal context.

Introduction

The p53 protein is a transcription factor that mainly controls the transcription of cell cycle control and cell death genes.1 As an anti-proliferative protein, p53 is tightly controlled. The major regulator for p53 is Mdm2, which binds to the transactivation domain of p53 and mediates its ubiquitination and degradation.2 As mdm2 is also a target gene of p53, both proteins function in a negative feed-back loop. Several other proteins impinge on this negative feed-back loop and affect the abundance and activity of p53 and/or Mdm2.2

TRIM25 belongs to the Tripartite motif protein (TRIM) family, which contains an N-terminal RING (really interesting new gene)-domain, one or two B-boxes and a coiled-coil region.3 TRIM25 is induced by estrogen and is therefore particularly abundant in placenta and uterus. TRIM25 mRNA expression is also high in the thyroid gland, aorta and spleen, but low in most other tissues.4, 5 The absence of TRIM25 leads to underdeveloped uteri in female mice and reduces estrogen-responsiveness,6 whereas its overexpression has been observed in several ovarian and breast tumors and is associated with advanced disease and poor prognosis.7 Similar to other RING-domain proteins, TRIM25 possesses E3 ligase activity and is able to transfer ubiquitin and ISG15 to target proteins.8

We show that TRIM25 enhances p53 and Mdm2 abundance by preventing their proteasomal degradation. At the same time, TRIM25 inhibits p53’s transcriptional activity and dampens the response to DNA damage. TRIM25 is essential for medaka development and this dependence is rescued by silencing of p53.

Results

TRIM25 increases the abundance of p53 and Mdm2 by protein stabilization

We identified TRIM25, a protein belonging to the family of TRIMs, in a cell culture-based expression screen that we performed in order to identify novel regulators of p53 and Mdm2. The overexpression of TRIM25 increased p53 and Mdm2 abundance in a dose-dependent manner (Figure 1a; Supplementary Figure S1A), whereas the downregulation of TRIM25 reduced p53 and Mdm2 levels (Figure 1b). As TRIM25 levels can be increased by β-estradiol,4 we tested the effect of this steroid hormone on the induction of p53 and Mdm2 in order to further investigate the physiologic relevance. When we treated MCF7 cells with β-estradiol, we observed a significant upregulation of both p53 and Mdm2 that accompanied β-estradiol-mediated upregulation of TRIM25 (Figure 1c, Supplementary Figure S1B). The downregulation of TRIM25 by two different small interfering RNAs strongly reduced or blocked β-estradiol- mediated induction of p53 and Mdm2, respectively (Figure 1c, Supplementary Figure S1B). This indicated that the induction of p53 and Mdm2 by β-estradiol is (at least partly) mediated by TRIM25 and that the regulation of p53 and Mdm2 by TRIM25 occurs under physiologically relevant conditions.

Figure 1
figure1

TRIM25 increases the abundance of p53 and Mdm2. (a) H1299 cells were transfected with increasing amounts of V5-tagged human TRIM25 and with p53 and Mdm2. Twenty-four hours after transfection, the amounts of p53, Mdm2 and V5-tagged TRIM25 were monitored by western blotting. Hybridization with an antibody targeted against PCNA (proliferating cell nuclear antigen) was performed for loading control. (b) HCT116 cells were transfected with two different TRIM25 siRNAs (small interfering RNAs) or with control siRNA. Seventy-eight hours after transfection, cells were harvested and the abundance of p53, Mdm2 and TRIM25 was monitored by western blotting. (c) MCF7 cells were transfected with 100 nm of a TRIM25 siRNA or a control siRNA. Forty-eight hours after transfection, cells were treated with 10 nm β-estradiol. Twenty-four hours after β-estradiol addition, the abundance of p53, Mdm2 and TRIM25 was monitored by western blotting.

We next investigated the mechanism leading to increased abundance of p53 and Mdm2 by TRIM25. Since p53 and Mdm2 are connected by a negative feed-back loop,2 we asked whether TRIM25 can increase p53 abundance in the absence of overexpressed Mdm2 (and Mdm2 abundance in the absence of p53). We therefore co-transfected increasing amounts of TRIM25 into H1299 cells together with p53 but without Mdm2 (and together with Mdm2 but without p53). Figure 2a shows that the upregulation of p53 by TRIM25 only occurred when Mdm2 was co-transfected (Figure 2a, lanes 1–4), whereas a TRIM25-mediated increase in Mdm2 levels did not require p53 (Figure 2a, lanes 5–8).

Figure 2
figure2

TRIM25 increases p53 and Mdm2 abundance by protein stabilization (a) H1299 cells were transfected with increasing amounts of a plasmid encoding V5-tagged hTRIM25 together with p53 or Mdm2. Twenty-four hours after transfection, the amount of p53, Mdm2 and V5-tagged TRIM25 was monitored by western blotting. (b) H1299 cells were co-transfected with p53, Mdm2 and TRIM25 or with control DNA. Twenty-four hours after transfection, cycloheximid (CHX) was added. Cells were harvested at the indicated times after CHX addition and the amounts of p53, Mdm2 and TRIM25 were monitored by western blotting. Signals for p53, Mdm2 and PCNA (proliferating cell nuclear antigen) were quantified densitometrically and the relative amounts of p53 and Mdm2 were assessed. Mean values and s.d. of three independent experiments were calculated and plotted. Relative abundance of p53 and Mdm2 protein at the time of CHX addition was set to 100%. P-values were calculated by two-tailed t-test (*P<0.05; **P<0.01). (c) H1299 cells were co-transfected with Mdm2, His-tagged ubiquitin (His-ubi) p53 and TRIM25 in the indicated combinations. Twenty-four hours after transfection, the cells were treated with MG132 for 8 h before harvesting. An aliquot of the cells was analyzed for the expression of p53, Mdm2 and TRIM25 (input). The remaining cells were lysed in guanidinium buffer. Ubiquitinated proteins were collected by adsorption to Ni2+-agarose, eluted and analyzed by western blotting for the presence of p53 and Mdm2. (d) H1299 cells were co-transfected with plasmids encoding Mdm2 and p53 or TRIM25 or both. Twenty-four hours after transfection, MG132 was added to a final concentration of 10 μm. Four hours after MG132 addition, the cells were lysed. The cellular lysate (5%) was used to monitor the expression of p53, Mdm2 and TRIM25 by western blotting (input). From the remaining 95% of the cellular lysate p53 was immunoprecipitated. Precipitation with IgG was performed for control. The precipitates were resolved by gel electrophoresis. Immunoprecipitated p53 and associated Mdm2 were detected by western blotting. (e) HCT116 cells were transfected with siRNA targeted against TRIM25. Forty-eight hours after transfection the cells were lysed. An aliquot of the cells was analyzed for the abundance of Mdm2, p53 and TRIM25 (input). From the remaining cells, p300 was immunoprecipitated and associated p53 and Mdm2 were detected by western blotting.

Considering that p53 and Mdm2 are mainly regulated by protein stability,2 we next asked whether TRIM25 increases p53 and Mdm2 abundance by inhibiting their degradation. We therefore used cycloheximide to block de novo protein synthesis after the overexpression of p53, Mdm2 and TRIM25 in H1299 cells and monitored the decay of p53 and Mdm2 by western blotting. Without co-transfection of TRIM25, p53 and Mdm2 showed a half-life of ~20-25 min, however, this was strongly increased after the overexpression of TRIM25 (Figure 2b). TRIM25 did not affect the levels of p53 and Mdm2 mRNA (Supplementary Figure S2A), supporting the idea that the regulation of p53 and Mdm2 abundance by TRIM25 occurs by protecting the proteins from degradation. In order to see whether this principle also applies under more physiologically relevant conditions, we treated MCF7 cells with estrogen, which increases TRIM25 abundance.4 In line with the results obtained after the overexpression of TRIM25, treatment with estrogen increased the half-life of p53 and Mdm2 (Supplementary Figure S2B).

Most proteins that are degraded by 26 S proteasomes require covalent attachment of polyubiquitin chains. As TRIM25 stabilized the p53 and Mdm2 protein, we asked whether TRIM25 affects their polyubiquitination. We therefore transfected H1299 cells with His-tagged ubiquitin together with Mdm2 and/or p53 and TRIM25, purified ubiquitinated proteins by using metal affinity and monitored the abundance of p53 and Mdm2 by western blotting. As shown in Figure 2c, the overexpression of TRIM25 strongly reduced the polyubiquitination of both p53 and Mdm2 (Figure 2c).

P53 can be ubiquitinated by several ubiquitin ligases and these ubiquitin ligases usually need to interact directly with p53.2 As TRIM25 reduced ubiquitination of p53 after the overexpression of Mdm2, we wondered whether this might be caused by a reduction of the interaction of p53 and Mdm2. To test this, we overexpressed p53 and Mdm2 in H1299 cells in the presence and absence of TRIM25, immunoprecipitated p53 and monitored the abundance of Mdm2 by western blotting. In contrast to our expectation, TRIM25 did not reduce the association of p53 and Mdm2. In fact, it slightly enhanced their association (Figure 2d, compare lanes 2 and 3), probably owing to an increase in the abundance of p53 and Mdm2 in the presence of TRIM25.

A further requirement for p53 polyubiquitination is the interaction of Mdm2 with p300.9 We therefore tested whether TRIM25 interferes with the binding of p300 to Mdm2. We downregulated TRIM25 in HCT116 cells, precipitated p300 and monitored the associated Mdm2 by western blotting. As shown in Figure 2e, only a minor amount of Mdm2 was associated with p300 under normal growth conditions. Downregulation of TRIM25, however, resulted in a strong increase in the amount of Mdm2 that was bound to p300, whereas the amount of p53 that was bound to p300 was not affected (Figure 2e). This result suggests that TRIM25 reduces p53 polyubiquitination by hindering the binding of p300 to Mdm2.

TRIM25 associates with p53 and Mdm2

As TRIM25 interfered with p53 and Mdm2 polyubiquitination, we predicted that TRIM25 would associate with p53 and Mdm2. To investigate this, we precipitated endogenous TRIM25 from MCF7 cells and monitored the associated p53 and Mdm2 by western blotting. In agreement with our prediction, Mdm2 and p53 co-precipitated with TRIM25 (Figure 3a). Furthermore, in reciprocal co-immunoprecipitation experiments, TRIM25 and Mdm2 co-precipitated with p53 (Figure 3b) and TRIM25 and p53 co-precipitated with Mdm2 (Figure 3c). In order to get further evidence that these interactions are indeed physiologic, we performed sucrose density centrifugation. With this approach, it is possible to separate higher-order complexes under native conditions. Although the majority of p53 eluted in fractions that contained protein complexes of a molecular weight >500 kD, whereas TRIM25 eluted in fractions that contained proteins and protein complexes with a molecular weight below <500 kD, both proteins were detected in fractions 20–26. Of note, the majority of Mdm2 that is well known to form a complex with p532 also eluted in fractions with a molecular weight <500 kD. Importantly, the majority of Mdm2 eluted in fraction 20–26, which also contained p53 and TRIM25 (Supplementary Figure S3A). p53 and Mdm2 are primarily nuclear proteins.10 In order to determine whether all the three proteins are in the same cellular compartment we performed cell fractionation with subsequent western blotting. β-Actin (GAPDH in Supplementary Figure S3B.II) and p300 were used as markers for cytoplasmic and nuclear fractions, respectively. As expected, the majority of p53 and Mdm2 were detected in the nuclear compartment. A significant amount (almost 50%) of TRIM25 was also nuclear (Supplementary Figure S3B.I), showing that a significant part of all the three proteins is present in the same cellular compartment. Moreover, as shown in supplementary Figure 3B.II, Mdm2 and p53 co-precipitated with TRIM25 irrespective of whether the cytoplasmic or nuclear lysate was used (Supplementary Figure S3B.II).

Figure 3
figure3

TRIM25 interacts with Mdm2 and p53. (a) MCF7 cells were lysed. The cellular lysate (5%) was used to monitor the expression of p53, Mdm2 and TRIM25 by western blotting (input). From the remaining cellular lysate, TRIM25 was immunoprecipitated and associated p53 and Mdm2 were detected by western blotting. Precipitation with IgG was performed for control. (b) MCF7 cells were lysed. The cellular lysate (5%) was used to monitor the expression of p53, Mdm2 and TRIM25 by western blotting (input). From the remaining cellular lysate, p53 was immunoprecipitated and processed as described in part A. (c) MCF7 cells were lysed. The cellular lysate (5%) was used to monitor the expression of p53, Mdm2 and TRIM25 by western blotting (input). From the remaining cellular lysate, Mdm2 was immunoprecipitated and processed as described in part A.

As TRIM25 associated with p53 and Mdm2, and p53 strongly binds to Mdm22, we next asked whether p53 could influence the binding of Mdm2 to TRIM25 and whether Mdm2 could influence the binding of p53 to TRIM25. We therefore overexpressed TRIM25 and Mdm2 in the presence or absence of p53 and TRIM25 and p53 in the presence and absence of Mdm2. The overexpression of p53 had only a minor effect on the association of Mdm2 with TRIM25 (Supplementary Figure S3C); however, the overexpression of Mdm2 strongly enhanced the association of p53 with TRIM25 (Supplementary Figure S3D).

TRIM25 reduces p53 transcriptional activity

Usually, increased abundance of p53 is accompanied by an increase in p53’s transcriptional activity. To see whether this is also the case for p53 in the presence of TRIM25, we transfected the p53-dependent reporter PG13, which contains 13 repeats of the p53-consensus site fused to a minimal promoter into MCF7 cells, together with increasing amounts of TRIM25 and monitored the reporter activity. Surprisingly, the overexpression of TRIM25 did not increase but actually decreased p53-dependent transcriptional activity (Figure 4a). In agreement with this result, the downregulation of TRIM25 resulted in an increased p53 reporter activity (Figure 4b). Consistently, the downregulation of TRIM25 increased mRNA levels of p21 and 14-3-3σ, whereas it decreased the mRNA abundance of CDC25c, whose expression is repressed by p53 (Supplementary Figure S4A). Interestingly, the abundance of Mdm2 RNA was not significantly changed under these conditions.

Figure 4
figure4

TRIM25 reduces p53’s activity. (a) MCF7 cells were transfected with the p53-dependent reporter PG13 (Firefly luciferase) and Renilla luciferase, together with increasing amounts of TRIM25. 48 h after transfection, Firefly and Renilla luciferase activity was determined. Relative Firefly activity was calculated by normalization with the readings for Renilla luciferase. Relative Firefly activity of cells without TRIM25 transfection was set to 100%. The graph shows mean values and s.d. of three independent experiments. P-values were calculated by two-tailed t-test. *P<0.05; **P<0.01. (b) MCF7 cells were transfected with the p53-dependent reporter PG13 and Renilla luciferase together with a TRIM25 siRNA or a control siRNA. Forty-eight hours after transfection, cells were harvested and processed as described in the legend to part A. (c) HCT116 wild-type (wt) and HCT p53-/- cells were transfected with TRIM25 siRNA or with a control siRNA. Forty-eight hours after transfection, cells were analyzed for expression of p53, acetylated p53 (ac-p53), Mdm2, PARP, cleaved Caspase 3 and TRIM25. (d) Fibroblasts were prepared from murine embryos that possess two, one or no allele of TRIM25. Cells were harvested and analyzed for the expression of p300, p53, acetylated p53, Mdm2 and TRIM25. (e) HCT116 wt cells were transfected with two different siRNAs (small interfering RNA) targeted against TRIM25 or with a control siRNA. Forty-eight hours after transfection, the cells were treated with 25 or 40μM (f.c.) curcumin. Twenty-four hours after curcumin addition, the cells were lysed and analyzed for acetylated p53, Mdm2 and TRIM25.

One function of p53 is to induce apoptosis.2 Indeed, when TRIM25 was downregulated in HCT116 cells by siRNA, a strong increase in cell death was observed (Supplementary Figure S4B) together with the cleavage of PARP and Caspase 3 (Figure 4c; Supplementary Figure S4C). Importantly, cell death did not occur in HCT116 p53-/- cells, demonstrating that TRIM25 regulates cell death in a p53-dependent manner (Figure 4c, Supplementary Figure S4B). It should be noted that in p53-negative HCT116 cells, we also observed some reduction in cell number upon downregulation of TRIM25, however, this decrease was significantly weaker than in HCT116 cells with wild-type p53 (Supplementary Figure S4D).

One condition that would result in increased transcriptional activity of p53 would be increased binding of p53 to promoters of target genes. However, in spite of the increase in p53 transcriptional activity after RNAi-mediated downregulation of TRIM25, we observed a decrease in the binding of p53 to its consensus DNA (Supplementary Figure S4E.I). The reduction in binding most likely mirrors the decrease in the overall abundance of p53 upon reduction of TRIM25 levels (Supplementary Figure S4E.II). This result implies that it is not the absolute amount of p53 bound to promoter sites that determines the activity of p53 but that additional alterations, like post-translational modifications of the bound p53, dictate the transcriptional response.

As p53-dependent induction of cell death requires acetylation of C-terminal lysines,11 we tested whether the downregulation of TRIM25 affected p53 acetylation. We therefore probed the cell lysates with an antibody against acetylated lysine 382 of p53. Indeed, lysates from HCT116 cells treated with TRIM25 siRNA showed a strong increase in p53 acetylation (Figures 4c and e, Supplementary Figure S4C). Also TRIM25-/- mouse embryonic fibroblasts showed enhanced acetylation of p53. This acetylation was accompanied by a strong increase in the abundance of the histone acetyltransferase p300 (Figure 4d), a major acetyltransferase for p53.12, 13 Inhibition of p300 with curcumin14 strongly reduced the acetylation of p53 after the downregulation of TRIM25 (Figure 4e), showing that TRIM25 indeed controls p300-mediated acetylation of p53. Downregulation of TRIM25 and acetylation of p53 furthermore resulted in the induction of the p53 target gene P21 in MCF7 cells. When we concomitantly downregulated p300, acetylation and P21 induction were clearly reduced (Supplementary Figure S4F). Whether p300 is the only acetylase that is involved in this process remains to be determined.

TRIM25 reduces p53 activity in response to DNA damage

P53 activity is most important during the DNA damage response. Here p53 accumulates to high levels and its activity is enhanced.2 As TRIM25 inhibits p53 transcriptional activity, this raises the question whether p53 is released from this TRIM25-mediated repression during the DNA damage response. One possibility by which this could be achieved is via a reduction in the amount of TRIM25 after DNA damage. Treatment of cells with γ-rays or etoposide, however, resulted in a slight but reproducible increase in TRIM25 levels (Figure 5a). Alternatively, TRIM25 could be inhibited and thus no longer be able to repress p53-dependent transcription. To investigate this possibility, we measured p21-dependent reporter activity in response to DNA damage with or without overexpressed TRIM25. In the absence of DNA damage, the overexpression of TRIM25 reduced p21-dependent reporter activity (Figure 5b) as it did with PG13-dependent reporter activity (Figure 4a). As expected, in cells treated with etoposide an increase in p21-dependent reporter activity was detected (Figure 5b). Overexpression of TRIM25 reduced this p21-dependent reporter activity in response to DNA damage (Figure 5b). Likewise, the overexpression of TRIM25 reduced the induction of endogenous P21 protein in response to DNA damage (Supplementary Figure S5A). Conversely, when we downregulated TRIM25, P21 and Bax levels were increased and this increase was raised further in response to DNA damage (Figure 5c, Supplementary Figure S5B). Likewise, mice with a genetic deletion of TRIM25 showed higher levels of the p53 targets P21 and 14-3-3σ and these higher levels were further increased after DNA damage (Figure 5d), showing that TRIM25 reduces p53 activity also in response to DNA damage. Interestingly, transcription of Mdm2 was again rather insensitive to the alterations in TRIM25 abundance (Figure 5d).

Figure 5
figure5

TRIM25 reduces the activation of p53 in response to DNA damage. (a) U2OS cells were irradiated with γ-rays (5 Gy) or treated with etoposide (Eto; 50 μm) for the indicated time. Abundance of TRIM25, p53, Mdm2, P21 and PCNA (proliferating cell nuclear antigen; for loading control) was determined by western blotting. (b) U2OS cells were transfected with the P21-reporter (Firefly luciferase) and Renilla luciferase together with vector DNA or hTRIM25. Forty hours after transfection, Eto or the carrier were added. Sixteen hours after Eto addition, cells were harvested and the Firefly and Renilla luciferase activity was determined. Relative p21-promoter activity was calculated by normalization with the readings for Renilla luciferase. The value of cells treated with carrier was set to 100%. The graph shows mean values and s.d. of three independent experiments. P-values were calculated by two-tailed t-test. *P<0.05; **P<0.01. (c) HCT116 cells were transfected with TRIM25 siRNA or a control siRNA. Forty hours after transfection, cells were irradiated with γ-rays (5 Gy), treated with etoposide (Eto: 50 μm f.c.) or left untreated for control (C). Eight hours after DNA damage, the cells were lysed and the abundance of TRIM25, P21 and Bax was determined by western blotting. (d) Primary fibroblasts from mice with a genetic deletion of TRIM25 or from the corresponding wild-type mice were treated with etoposide (Eto) or left untreated for control (untd). Eight hours after treatment, the cells were harvested and the total RNA was prepared. The relative amount of P21, 14-3-3σ and Mdm2, mRNA was determined by qRT (quantitative reverse transcription)-PCR. Mean values and s.d. were calculated from the 2^dCT values of three independent experiments and blotted. Values of mRNA in cells transfected with control siRNA were set to 1. P-values were calculated by the two-tailed t-test (**P<0.01; *P<0.05; n.s, not significant).

Developmental defects in medaka embryos induced by knocking down TRIM25 are rescued by the silencing of p53

A TRIM25-like gene has been reported to be essential for early zebra-fish development.15 To investigate whether this is also the case for medaka, we performed a functional analysis of these proteins in medaka embryos during early development. Owing to teleost-specific gene duplications, medaka possesses two homologs of TRIM25 (Supplementary Figure S6A). Both homologs contain the RING-domain, the two B-boxes and the coiled-coil domain (Supplementary Figure S6B) and both homologs of medaka TRIM25 are capable of increasing p53 and Mdm2 abundance (Supplementary Figure S6C). Although homolog-2 was more efficient in regulating p53 and Mdm2 abundance than homolog 1, we cannot exclude that this difference is caused by different expression levels, due to the lack of antibodies.

A prerequisite for the regulation of p53 by TRIM25 during development is that both proteins are temporally and spatially co-expressed. By whole-mount in situ hybridization we found that p53, mdm2 and trim25 were all ubiquitously expressed in medaka (Supplementary Figure S6D). Expression of p53 and trim25 was also constant during the first 6 days of development, whereas the expression of mdm2 was reduced from day 1 to day 2 post fertilization but stayed constant during the following days (Supplementary Figure S6E).

In order to investigate whether TRIM25 is required for medaka development, we injected antisense-morpholino-oligonucleotides (MO) targeting both homologs of medaka TRIM25 at the one-cell stage. This led to a dose-dependent shortening of the embryonic axis and retarded development (Figure 6a, Supplementary Figure S6F). The phenotype induced by TRIM25-MOs was rescued to a significant part by the co-injection of TRIM25 mRNA (Supplementary Figure S6G), demonstrating that the developmental defect upon injection of TRIM25-MOs is specific to the absence of TRIM25 and not due to off-target effects.

Figure 6
figure6

Malformations caused by TRIM25 knockdown in medaka embryos are rescued by the silencing of p53. (a) The indicated MOs were injected into fertilized medaka eggs. One day after injection, the morphology of the embryos was monitored by microscopy. The graph shows the mean values and s.d. of three independent experiments with 28 to 46 analyzed embryos in each experiment. The total number of analyzed embryos was set to 100%. The P-value was calculated from the embryos injected with TRIM25-MOs and the embryos injected with TRIM25- and p53-MOs by two-tailed t-test. **P<0.01. (b) Fertilized medaka eggs were injected with the indicated MOs. One day after injection, the embryos were fixed and apoptotic cells were labelled using the TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling) technique and detected by microscopy.

Importantly, co-injection of p53-MOs together with TRIM25-MOs reduced the number of embryos with an abnormal phenotype from 70 to 20%, whereas the injection of p53-MOs alone had no effect (Figure 6a). As the phenotype of TRIM25-MO-injected embryos was reminiscent of apoptosis, we performed TUNEL staining and found that injection of TRIM25-MOs indeed induced apoptosis and that this induction of apoptosis was also reversed by p53-MOs (Figure 6b). These data show that TRIM25 also controls p53 function in vivo. Likewise, a reduction in body weight that was observed in TRIM25-/- mice was rescued when p53 was also deleted (Supplementary Figure S7A). Although the phenotype in mice was not as strong as in medaka, and the cohort was rather small, these data indicate that TRIM25 also controls p53 activity in mammals.

TRIM25 is frequently overexpressed in human ovarian tumors

The p53/Mdm2 loop is an important hub for tumor development, which is displayed by the fact that ~50% of human tumors have mutated p53. We investigated the expression level of TRIM25 in a cohort of ovarian tumors with known p53 status.16 In agreement with previous reports, we found that TRIM25 was frequently overexpressed in human tumors.17 In fact, we found a high expression of TRIM25 in 11 out of 24 tumors. Three out of seven tumors with wild-type p53 and five out of eleven tumors with a missense mutation in the p53 gene showed high expression of TRIM25. Strikingly, in tumors with a nonsense mutation resulting in a shortened p53 protein, TRIM25 was overexpressed in three out of four cases (Table 1, Supplementary Figure S7B).

Table 1 Relative expression of TRIM25 and p53 mutational status of 24 tumor samples from patients with ovarian carcinoma

Discussion

We identified the Tripartitite-Motif-protein TRIM25 as a novel regulator for p53 and Mdm2 by a cell culture-based overexpression screen using a complementary DNA library.18 Increased abundance of TRIM25 either by its direct overexpression or via estrogen-dependent TRIM25-induction resulted in a higher abundance of p53 and Mdm2.

Increased abundance of p53 and Mdm2 after estrogen treatment has been reported before,19, 20, 21 but the dependence on TRIM25 has not been investigated. Downregulation of TRIM25 completely blunted the induction of Mdm2 by estrogen, demonstrating that the induction of TRIM25 is responsible for the increase. Induction of p53 was, however, not completely inhibited under these conditions. Earlier reports showed that estrogen increases p53 mRNA levels 21 and protein stability, 19 and it is most likely that these mechanisms also contribute to the induction of p53 by estrogen.

Mdm2 and p53 are short-lived proteins that are degraded in 26 S proteasomes upon prior ubiquitination. TRIM25 reduced p53 and Mdm2 ubiquitination, resulting in increased stability of p53 and Mdm2. Polyubiquitination and degradation of p53 requires the association of Mdm2 with p300.9 This association was strongly reduced in the presence of TRIM25 resulting in reduced ubiquitination and degradation of p53. Whether the interaction of p300 with Mdm2 is also required for ubiquitination of Mdm2 remains to be determined.

Most surprisingly, although p53 levels were elevated, p53’s transcriptional activity was reduced by TRIM25. This result suggests that TRIM25 has two separate activities with regard to p53, one is regulation of p53 abundance and the other is control of its transcriptional activity. The increase in p53’s transcriptional activity was accompanied by increased acetylation of p53, a modification that is required for transcription particularly of growth-arresting and pro-apoptotic target genes.22 The histone acetyltransferase P300 contributes significantly to p53 acetylation.2 Interestingly, fibroblasts with a genetic deletion of TRIM25 showed much stronger expression of p300. The mechanism for this upregulation of p300 in the absence of TRIM25 is unclear. Surprisingly, despite the increase in target gene activation, the amount of p53 bound to DNA was reduced upon the downregulation of TRIM25. This reduction in DNA-bound p53 was probably due to an overall reduction in p53 abundance after the downregulation of TRIM25. The increase in the amount of acetylated p53, however, could most likely compensate for this reduction, resulting in increased transcriptional activation despite a reduction in the overall p53 levels.

Increased activity of p53 after downregulation of TRIM25 resulted in the activation of Caspase 3, cleavage of PARP and death of wt HCT116 cells, whereas HCT116 cells with a genetic deletion of both alleles of p53 (HCT p53-/-) showed no signs of cell death, demonstrating the p53-dependence of this process. Nevertheless, HCT p53-/- cells also showed some, although much weaker, reduction in proliferation upon downregulation of TRIM25. This reduction is most likely due to increased abundance of 14-3-3σ, a p53 target gene that is targeted for degradation by TRIM257, 23 and this activity of TRIM25 does not depend on p53.

The observation that TRIM25 increases p53 abundance while at the same time inhibiting its transcriptional activity appears counter-intuitive. One possible explanation is that TRIM25 keeps p53 levels high to allow rapid p53-dependent responses but inactive enough to ensure cell survival. When p53 activity is required, the activity-suppressing function of TRIM25 could be rapidly switched off, resulting in the immediate presence of high levels of active p53. How the p53-inhibiting activity of TRIM25 is released remains to be determined.

One major function of p53 is to inhibit cell proliferation in response to DNA damage.24 As TRIM25 inhibits p53 activity, the question arose whether TRIM25 is inactivated in response to DNA damage. Surprisingly, TRIM25 levels were even enhanced after DNA damage and also the activity of TRIM25 appeared not to be restricted as we always observed reduced transcriptional activation of p53 in the presence of TRIM25, irrespectively if DNA lesions were present or not. As the supervision of genomic integrity is one of the main routes by which p53 suppresses carcinogenesis, abundance and activity of TRIM25 during the DNA damage response is an important issue. Indeed, we found that TRIM25 is over-expressed in ovarian tumors and its over-expression in other tumors has also been reported17, 25 and this study. Moreover, downregulation of TRIM25 strongly reduced tumor growth in an experimental setting.7 Whether these observations are directly linked to the regulation of p53 will, however, require a larger cohort of tumors with known p53 status. Importantly, as DNA damage is not only an important trigger for the onset of carcinogenesis but also frequently used for cancer treatment, which often involves activation of p53, the status of TRIM25 might be an important determinant for successful tumor therapy.

TRIM25 is essential for zebrafish development15 and as we show also for the development of medaka. The downregulation of TRIM25 resulted in severe malformations and the embryos eventually die by p53-induced apoptosis. These results show in a whole animal that the control of p53 activity by TRIM25 is of vital importance. In contrast to fish, mice with a genetic deletion of TRIM25 are viable. However the uterus of these mice is underdeveloped and the proliferation of fibroblasts derived from these mice is strongly reduced.6, 7 Furthermore, mice with a deletion of TRIM25 frequently showed reduced body weight, which was not observed when we compared it with sex-matched siblings from the same litter. All these properties are consistent with an overly active p53 protein. The strong dependence of teleost development on TRIM25-mediated control of p53 raises the question why this dependence is not seen so robustly in the mouse. The most likely explanation is that in higher eukaryotes, the abundance of TRIM25 might be too low in fetal tissues and therefore the influence of TRIM25 is limited. In teleosts, the impact of TRIM25 might be higher as the protein might be more abundant. However, if TRIM25 expression is low in most tissues, why then is p53 not overly active in these settings? Most likely, other proteins can replace TRIM25 in controlling p53 activity in most tissues and this could have restricted the requirement for TRIM25 to certain tissues such as the uterus where TRIM25 is present in higher amounts.6

Materials and methods

Cell lines and their treatments

H1299, U2OS, MCF7, HCT116 and HCT116 p53-/- were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 1% penicillin/streptomycin according to standard conditions. H1299 cells were transiently transfected by calcium-phosphate DNA co-precipitation26 or with PromoFectin (PromoKine, Heidelberg, Germany). U2OS and HCT116 cells were transfected with GeneJuice (Novagene, Darmstadt, Germany). For transfection of siRNA, Ribojuice (Novagen) or Viromer (Lipocalyx, Halle, Germany) was used. Sequences of small interfering RNAs are available on request.

β-estradiol was used at a final concentration of 10 nm in RPMI (Roswell Park Memorial Institute) medium without phenol red, supplemented with 10% charcoal-stripped fetal bovine serum and 1% penicillin/streptomycin. Cells were transferred to RPMI medium 2 days before estrogen treatment. MG132 was used at 10 μm (f.c.), cycloheximide at 60 μg/ml (f.c.) and etoposide at 50 μm (f.c.). γ-irradiation was performed in culture medium with 5 Gy at a dose rate of 0.5 Gy per minute using a 60cobalt-γ-source.

Fish strains and housing

The medaka inbred line Cab was maintained at 26° C. Embryos were raised at 28° C.

Plasmids

The plasmids encoding p53, Mdm2 and His-ubiquitin, have been described previously.26 TRIM25 was amplified by PCR and cloned into pcDNA3 vector. For cloning of V5-tagged TRIM25 the same strategy was used with inclusion of the V5-tag sequence in the reverse primer. The plasmids for mdTRIM25 homolog 1 and 2 were from the cDNA (complementary DNA) library. Sequences of cloning primers are available on request.

Antibodies

The antibodies DO-1 (p53), HRP (horseradish peroxidase)-coupled DO-1, PC10 (anti-proliferating nuclear cell antigen; PCNA), 9E10 (Myc) and antibodies directed against TRIM25 (E-4 and E-12), PARP (H-250) and P21 (C-19) were from Santa Cruz (Dallas, TX, USA). The antibodies against acetylated p53 (K382), cleaved Caspase 3 (Asp175) and Bax were from Cell Signalling (Danvers, MA, USA), the antibodies against p300 and p53 (AB-2) were from Millipore (Darmstadt, Germany) and the anti-V5 antibody was from Serotec (Puchheim, Germany). HRP-coupled secondary antibodies were from Dako (Darmstadt, Germany).

Small interfering RNAs

Sequences of small interfering RNAs are available on request

Antisense morpholino oligonucleotide and injection

MOs (Gene Tools, Philomath, OR, USA) were dissolved in RNAse/DNAse-free water, diluted to the desired concentration and injected into medaka embryos at the one- or two-cell stage.

SDS–PAGE and western blotting

SDS–PAGE and western blotting was performed as described in Kulikov et al.26

Ubiquitination assay

Ubiquitination assay was performed as described in Kulikov et al.26

Immunoprecipitation

Cells were lysed in lysis buffer (20 mm Tris pH 7.5, 1 mm EDTA, 100 mm NaCl, 0.5% NP40, 10% glycerol and EDTA-free Protease Inhibitor Cocktail (Roche, Mannheim, Germany)) for 15 min on ice. The lysate was cleared by centrifugation at 13 200 r.p.m. for 10 min at 4 °C. The protein lysate (5%) was analyzed by western blotting (input). The remaining lysate was added to protein A and G sepharose precoupled with antibody and incubated for 4 h at 4 °C. The precipitates were washed with lysis buffer, suspended in sample buffer and separated by SDS–PAGE.

Quantitative reverse transcription PCR

Quantitative reverse transcription PCR. was performed as described in Solozobova et al.27

Luciferase assay

For luciferase assays, 1x104 cells per well were plated in 96-well plates. For TRIM25 overexpression studies, each well was transfected with 50 ng of the p53-dependent reporter PG13 or p21 and 5 ng of a plasmid encoding Renilla luciferase together with a plasmid encoding TRIM25 or with vector DNA. For TRIM25 downregulation studies, each well was transfected with 30 ng of PG13 or p21 and 3 ng of a plasmid encoding Renilla luciferase together with TRIM25 or control siRNA. Forty-eight hours after transfection, the cells were lysed and the luciferase activity was determined.

Whole-mount in situ hybridization

Whole-mount in situ hybridization was performed as described in Loosli et al.28

Ethical issues

Patient samples were used in an anonymized form. The local ethical committee has approved the study.

Figure 7
figure7

Regulation of p53 by TRIM25. Mdm2, as an E3 ligase, can ubiquitinate p53 and target it for proteasomal degradation. Efficient polyubiquitination and degradation of p53, however, requires p300. P300 furthermore acetylates p53 and this modification is essential for the transcriptional activation of most target genes of p53. TRIM25 inhibits these activities of p300 resulting in higher levels of p53, which is, however, usually transcriptionally inactive p53.

References

  1. 1

    Riley T, Sontag E, Chen P, Levine A . Transcriptional control of human p53-regulated genes. Nat Rev Mol Cell Biol 2008; 9: 402–412.

  2. 2

    Boehme KA, Blattner C . Regulation of p53—insights into a complex process. Crit Rev Biochem Mol Biol 2009; 44: 367–392.

  3. 3

    Nisole S, Stoye JP, Saib A . TRIM family proteins: retroviral restriction and antiviral defence. Nat Rev Microbiol 2005; 3: 799–808.

  4. 4

    Inoue S, Orimo A, Hosoi T, Kondo S, Toyoshima H, Kondo T et al. Genomic binding-site cloning reveals an estrogen-responsive gene that encodes a RING finger protein. Proc Natl Acad Sci USA 1993; 90: 11117–11121.

  5. 5

    Shimada N, Suzuki T, Inoue S, Kato K, Imatani A, Sekine H et al. Systemic distribution of estrogen-responsive finger protein (Efp) in human tissues. Mol Cell Endocrinol 2004; 218: 147–153.

  6. 6

    Orimo A, Inoue S, Minowa O, Tominaga N, Tomioka Y, Sato M et al. Underdeveloped uterus and reduced estrogen responsiveness in mice with disruption of the estrogen-responsive finger protein gene, which is a direct target of estrogen receptor alpha. Proc Natl Acad Sci USA 1999; 96: 12027–12032.

  7. 7

    Urano T, Saito T, Tsukui T, Fujita M, Hosoi T, Muramatsu M, Ouchi Y et al. Efp targets 14-3-3 sigma for proteolysis and promotes breast tumour growth. Nature 2002; 417: 871–875.

  8. 8

    Zou W, Zhang DE . The interferon-inducible ubiquitin-protein isopeptide ligase (E3) EFP also functions as an ISG15 E3 ligase. J Biol Chem 2006; 281: 3989–3994.

  9. 9

    Grossman SR, Perez M, Kung AL, Joseph M, Mansur C, Xiao ZX et al. p300/MDM2 complexes participate in MDM2-mediated p53 degradation. Mol Cell 1998; 2: 405–415.

  10. 10

    Yuan J, Luo K, Zhang L, Cheville JC, Lou Z . USP10 regulates p53 localization and stability by deubiquitinating p53. Cell 2010; 140: 384–396.

  11. 11

    Tang Y, Zhao W, Chen Y, Zhao Y, Gu W . Acetylation is indispensable for p53 activation. Cell 2008; 133: 612–626.

  12. 12

    Sakaguchi K, Herrera JE, Saito S, Miki T, Bustin M, Vassilev A et al. DNA damage activates p53 through a phosphorylation-acetylation cascade. Genes Dev 1998; 12: 2831–2841.

  13. 13

    Kobet E, Zeng X, Zhu Y, Keller D, Lu H . MDM2 inhibits p300-mediated p53 acetylation and activation by forming a ternary complex with the two proteins. Proc Natl Acad Sci USA 2000; 97: 12547–12552.

  14. 14

    Balasubramanyam K, Varier RA, Altaf M, Swaminathan V, Siddappa NB, Ranga U et al. Curcumin, a novel p300/CREB-binding protein-specific inhibitor of acetyltransferase, represses the acetylation of histone/nonhistone proteins and histone acetyltransferase-dependent chromatin transcription. J Biol Chem 2004; 279: 51163–51171.

  15. 15

    Amsterdam A, Nissen RM, Sun Z, Swindell EC, Farrington S . Hopkins. Identification of 315 genes essential for early zebrafish development. Proc Natl Acad Sci USA 2004; 101: 12792–12797.

  16. 16

    Bartel F, Jung J, Böhnke A, Gradhand E, Zeug K, Thomssen C, Hauptmann S . Both germ line and somatic genetics of the p53 pathway affect ovarian cancer incidence and survival. Clin Cancer Res 2008; 14: 89–96.

  17. 17

    Watanabe T, Imoto I, Kosugi Y, Ishiwata I, Inoue S, Takayama M et al. A novel amplification at 17q21-23 in ovarian cancer cell lines detected by comparative genomic hybridization. Gynecol Oncol 2001; 81: 172–177.

  18. 18

    Souren M, Martinez-Morales JR, Makri P, Wittbrodt B, Wittbrodt J . A global survey identifies novel upstream components of the Ath5 neurogenic network. Genome Biol 2009; 10: R92.

  19. 19

    Okumura N, Saji S, Eguchi H, Hayashi S, Nakashima S . Estradiol stabilizes p53 protein in breast cancer cell line, MCF-7. Jpn J Cancer Res 2002; 93: 867–873.

  20. 20

    Kinyamu HK, Archer TK . Estrogen receptor-dependent proteasomal degradation of the glucocorticoid receptor is coupled to an increase in mdm2 protein expression. Mol Cell Biol 2003; 23: 5867–5881.

  21. 21

    Hurd C, Dinda S, Khattree N, Moudgil VK . Estrogen-dependent and independent activation of the P1 promoter of the p53 gene in transiently transfected breast cancer cells. Oncogene 1999; 18: 1067–1072.

  22. 22

    Brooks CL, Gu W . The impact of acetylation and deacetylation on the p53 pathway. Protein Cell 2011; 2: 456–462.

  23. 23

    Nakayama H, Sano T, Motegi A, Oyama T, Nakajima T . Increasing 14-3-3 sigma expression with declining estrogen receptor alpha and estrogen-responsive finger protein expression defines malignant progression of endometrial carcinoma. Pathol Int 2005; 55: 707–715.

  24. 24

    Lane DP . Cancer. p53, guardian of the genome. Nature 1992; 358: 15–16.

  25. 25

    Ikeda K, Orimo A, Higashi Y, Muramatsu M, Inoue S . Efp as a primary estrogen-responsive gene in human breast cancer. FEBS Lett 2000; 472: 9–13.

  26. 26

    Kulikov R, Letienne J, Kaur M, Grossman SR, Arts J, Blattner C . Mdm2 facilitates the association of p53 with the proteasome. Proc Natl Acad Sci USA 2010; 107: 10038–10043.

  27. 27

    Solozobova V, Blattner C . Regulation of p53 in embryonic stem cells. Exp Cell Research 2010; 316: 2434–2446.

  28. 28

    Loosli F, Koster RW, Carl M, Krone A, Wittbrodt J . Six3 a medaka homologue of the Drosophila homeobox gene sine oculis is expressed in the anterior embryonic shield and the developing eye. Mech Dev 1998; 74: 159–164.

Download references

Acknowledgements

We thank Germana Meroni (CBM S.c.r.l., Trieste) for the TRIM25 plasmid, Yi Su for help with the screening and Christina Bauer, Tanja Kuhn, Beate Heydel and Cathrin Herder for technical assistance. PZ was a CSC fellow. GD acknowledges the funding from the DFG (FOR 1036) for the screening experiments. This work is supported by COST Action BM1307.

Author information

Correspondence to G Davidson or C Blattner.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Zhang, P., Elabd, S., Hammer, S. et al. TRIM25 has a dual function in the p53/Mdm2 circuit. Oncogene 34, 5729–5738 (2015). https://doi.org/10.1038/onc.2015.21

Download citation

Further reading