Mutant p53-induced Up-regulation of Mitogen-activated Protein Kinase Kinase 3 Contributes to Gain of Function*

Mitogen-activated protein kinase kinase 3 (MAP2K3) is a member of the dual specificity kinase group. Growing evidence links MAP2K3 to invasion and tumor progression. Here, we identify MAP2K3 as a transcriptional target of endogenous gain-of-function p53 mutants R273H, R175H, and R280K. We show that MAP2K3 modulation occurred at the mRNA and protein levels and that endogenous mutant p53 proteins are capable of binding to and activate the MAP2K3 promoter. In addition, we found that the studied p53 mutants regulate MAP2K3 gene expression through the involvement of the transcriptional cofactors NF-Y and NF-κB. Finally, functional studies showed that endogenous MAP2K3 knockdown inhibits proliferation and survival of human tumor cells, whereas the ectopic expression of MAP2K3 can rescue the proliferative defect induced by mutant p53 knockdown. Taken together, our findings define a novel player through which mutant p53 exerts its gain-of-function activity in cancer cells.

Mitogen-activated protein kinase kinase 3 (MAP2K3) is a member of the dual specificity kinase group. Growing evidence links MAP2K3 to invasion and tumor progression. Here, we identify MAP2K3 as a transcriptional target of endogenous gain-of-function p53 mutants R273H, R175H, and R280K. We show that MAP2K3 modulation occurred at the mRNA and protein levels and that endogenous mutant p53 proteins are capable of binding to and activate the MAP2K3 promoter. In addition, we found that the studied p53 mutants regulate MAP2K3 gene expression through the involvement of the transcriptional cofactors NF-Y and NF-B. Finally, functional studies showed that endogenous MAP2K3 knockdown inhibits proliferation and survival of human tumor cells, whereas the ectopic expression of MAP2K3 can rescue the proliferative defect induced by mutant p53 knockdown. Taken together, our findings define a novel player through which mutant p53 exerts its gain-of-function activity in cancer cells.
TP53 gene mutations are the most frequent genetic alterations in human cancers; Ͼ50% of all human cancer cases carry mutations within the TP53 locus (1). Most of these are missense point mutations and are localized in the core DNA-binding domain (2). These alterations disrupt the normal transcriptional capacity of p53 and compromise its tumor suppressor properties by abrogating its transcriptional activity on genes connected with cell cycle arrest, apoptosis, or DNA repair, in response to a variety of stress signals (3,4). Recent studies have shown that mutations of the TP53 gene can confer additional functions (gain of function, GOF) 2 that are exerted in a variety of ways, ranging from enhanced proliferation in culture, increased tumorigenicity in vivo, and enhanced resistance to a variety of commonly used anti-cancer drugs (5,6). The GOF hypothesis has recently been reinforced by studies employing mutant p53 (mutp53) "knock-in" mice, which show a higher frequency of tumor development and increased metastatic potential, compared with p53-deficient mice (7,8). Furthermore, RNA interference (RNAi) studies demonstrated that depletion of mutp53 renders cancer cells more sensitive to DNA-damaging chemotherapeutic agents in vitro (9,10) and reduces tumor malignancy both in vitro and in vivo (10). In agreement with these results, tumor growth delay studies, performed in the HT29 xenograft model, showed that conditional silencing of mutp53 does not only impact on tumor growth but leads to tumor architecture modifications, with consistent reduction in stromal invasion and tumor angiogenesis (11). At the molecular level, these GOF effects were shown to be linked to the ability of mutp53 to modulate the expression of several genes, such as MDR1 (12), c-MYC (13), CD95 (Fas/APO-1) (14), EGR1 (9), MSP/MST-1 (15), GEF-H1 (16), ID2 (17), GRO1 (18), PPARGC1A, FRMD5 (19), and ID4 (20), supporting the hypothesis that mutp53-specific transcriptional activity is required for at least some of the mutp53 GOF effects. However, the molecular mechanisms underlying the GOF of mutp53 proteins are still far from being understood. Two different and not mutually exclusive possibilities are currently considered: (i) mutp53 retains residual transcriptional activity and acts as regulator of transcription (19,(21)(22)(23)(24); and (ii) mutp53 can no longer bind DNA but interacts with other transcription factors and modulates their activities (25)(26)(27).
The mitogen-activated protein kinase-kinase 3 (MAP2K3) belongs to a dual specificity kinase group (MKK Ϫ ) and is activated by MKKK proteins (MEKK1-4) through Ser-189 and Thr-193 phosphorylation (28). MAP2K3 is a specific upstream activator of the p38 MAPK protein (28). Recent studies found that MAP2K3 up-regulation was involved in invasion and progression of gliomas and breast tumors (29). Our recent quantitative PCR validation of microarray data, from cells expressing endogenous mutp53 (HT29 and SKBR3) and xenograft tumors, indicated that MAP2K3 (also known as MKK3) is a mutp53 target gene (11). To unveil a possible role of MAP2K3 up-regulation in mutp53 GOF activity, we studied the molecular mechanisms through which different mutp53 proteins up-regulate MAP2K3 expression and the biologic effects of its up-regulation in different cell lines.
Here, we find that different mutp53 proteins bind the MAP2K3 promoter, enhancing its transcriptional activity. Moreover, by dissecting the promoter region, we identify co-activators required by mutp53 to up-regulate MAP2K3 expression. Finally, biological studies indicate that MAP2K3 up-regulation plays an important role in cell proliferation and survival. Overall, our study defines MAP2K3 as a novel target of mutp53 and provides new insights for the understanding of mutp53 GOF activity.
MAP2K3 Promoter Constructs-A fragment of 1.0 kb of the MAP2K3 gene 5Ј-regulatory region (Ϫ989 to Ϫ2 with respect to the translational start site) was PCR-amplified from HT29 genomic DNA by using information obtained from the NCBI database. Primers were designed as follows. The forward primer includes a Sac1 restriction enzyme sequence plus a short tail (MAP2K3-For, 5Ј-tatagactat-GAGCTCACCACCGACCC-3Ј). The reverse primers include a BglII restriction enzyme sequence plus a short tail (MAP2K3-Rev, 5Ј-tataagatct-TGC-AAGTGGGTCCTGGAC-3Ј). PCR reaction was performed with HOT-MASTER Taq (Eppendorf). Amplified fragment was digested with SacI/BglII restriction enzymes (Invitrogen) and cloned into digested pGL3-Luc promoter-less vector generating pMAP2K3-Luc. The new vector was then sequence analyzed. Deletion derivatives of the MAP2K3 cloned promoter were generated as follows. pdel1-Luc (deletion Ϫ989 to Ϫ716) was generated by cutting cloned promoter (pMAP2K3-Luc) with SacI restriction enzyme, T4 DNA polymerase blunt-ended (Biolabs), PvuII-digested (Invitrogen), and self-ligated (Takara DNA ligation kit). pdel2-Luc (deletion Ϫ782 to Ϫ499) was generated by cutting with SmaI (Invitrogen) dual restriction enzyme sites and self-ligated. pdel3-Luc (deletion Ϫ711 to Ϫ167) was generated by cutting with PstI restriction enzyme (Roche Applied Science) and self-ligated.
Transactivation Assay-Cells were plated in 6-well plates (5 ϫ 10 4 cells/well), and on the following day, cells were transiently transfected with vector reporter (0.9 g/well) and pCMV␤-gal (0.1 g/well) (transfection efficiency) vector by following Lipofectamine Plus guidelines (Invitrogen). Luciferase and ␤-galactosidase assays were performed 24 h later on whole-cell extract, as described (30). Luciferase values were normalized to ␤-galactosidase activity and protein contents. H1299 cells were plated in 6-well plates (5 ϫ 104 cells/well), and 24 h later, cells were transiently co-transfected with either the pMAP2K3-Luc or the partial deleted promoters (0.8 g/well) along with the pCMV␤-gal (0.1 g/well) and p53-expressing (0.1 g/well) vectors. Luciferase and ␤-galactosidase assays were performed 24 h later as reported.
Viral Vectors-Lentiviral vectors were produced in HEK293T cells by transient transfection as described previously (11). Lentiviruses were harvested 48 h later, centrifuged for 5 min at 3,000 rpm, aliquoted, and stored at Ϫ80°C. Lentiviral stocks were titered following standard protocols (32). Routinely, a viral titer of 106 transducing units per ml was achieved.
Cellular Transduction-For constitutive RNA interference, cells were seeded in 24-well plates (3 ϫ 10 4 cells/well). 24 h later, recombinant lentiviruses supplemented with 8.0 g/ml of polybrene reagent (Sigma-Aldrich, H-9268) were added to the cells. After 16 h, cells were washed and replenished with fresh medium and, 96 h later, processed for Western blot analysis and/or proliferation assays. For conditional RNA interference, HT29 cells were plated in 24-well plates (3 ϫ 10 4 cells/well). After 24 h, medium containing LV-THsh/MAP2K3 and LV-tTR-KRAB lentiviruses supplemented with 8.0 g/ml of polybrene was added to the cells. Following 16 h of incubation, cells were washed and replenished with fresh medium. To monitor shRNA expression upon doxycycline hydrochloride treatment (ϩDox) (D9891, Sigma-Aldrich), new generated cell lines were plated, and after 24 h, Dox (1.0 g/ml) was added to the medium. Dox was replenished every 3 days. Then, 5 days later, cells were analyzed by Western blot.
For stable ectopic expression of MAP2K3, cells were transfected with either vector pCDNA3 (empty) or pRc/RSV-FLAG-MAP2K3 (Addgene) following Lipofectamine Plus guidelines (Invitrogen). Cells were selected by Geneticin (G418 sulfate, Invitrogen), and stable transfected cells were analyzed by Western blot to monitor ectopic MAP2K3 expression.
Flow Cytometry Analysis-Infected HT29 cells were maintained in either ϮDox conditions (1.0 g/ml). At different time points, cells were harvested, washed once in phosphate-buffered saline/NaN3 1ϫ, and fixed in MetOH:acetic acid solution (4:1) for 60 min at 4°C. Then, cells were centrifuged, resuspended in phosphate-buffered saline/NaN3 1ϫ supplemented with 2 g/ml RNase (150 units/ml), and incubated, during the last 30 min, in the dark with 0.1 g/ml of propidium iodide. Subsequently, cells were analyzed by flow cytometry.
Quantitative PCR-Quantitative PCR was performed using SYBR Green (Applied Biosystems) as a marker for DNA amplification on an ABI Prism 7500 apparatus (Applied Biosystems), with 50 cycles of two-step amplification. Samples were quantified in triplicate from two independent immunoprecipitations. The relative proportions of immunoprecipitated promoter fragments were determined based on the comparative threshold (DCt) method (31). Primers sequences of the hMAP2K3 promoter used in the quantitative PCR reactions were: forward, 5Ј-TTAACCCCCGCCCACTTC-3Ј and reverse, 5Ј-TGC-GTCGTCTGGAAAAAACC-3Ј.
Statistical Analysis-All experiments were performed in triplicate. Numerical data were reported as means ϩ S.D. Significance was assessed by Student's t test analysis.

RESULTS
Mutp53 Up-regulates MAP2K3 Protein Expression-We previously showed (11) that exogenous expression of mutp53R175H, in p53-null H1299 cells, significantly increases MAP2K3 mRNA. Conversely, depletion of endogenous mutp53R175H or mutp53R273H proteins by RNAi induced a sizable reduction of MAP2K3 mRNA levels in human cancer cell lines (11). We wished to verify whether the observed upregulation of MAP2K3 gene expression by mutp53 translates into increased MAP2K3 protein levels. To this aim, we first explored its modulation in p53-null H1299 cells upon ectopic expression of the wild-type (wt) p53, mutp53R175H, or FIGURE 1. Mutant proteins but not wild-type p53 contribute in MAP2K3 protein up-regulation. A, p53null H1299 cells were transiently transfected with empty (pcDNA3), wild-type p53-, mutp53R175H-, or mutp53R273H-expressing vectors. Then, 24 h later, cells were processed, and MAP2K3 protein level was established by Western blot analysis. Protein lysate (30 g/lane) were resolved and probed with specific antibodies: anti-p53 (DO1), anti-MAP2K3, and anti-actin (loading control). B, human cancer cell lines harboring p53 mutations or wild-type p53 proteins were infected with either the sh/p53 or sh/scr lentivirus, and then 96 h later, cells were processed, and Western blot analysis was performed as reported. Intensities of MAP2K3 bands were quantified by Scion Image software and normalized to actin protein bands. mutp53R273H proteins. As shown in Fig. 1A, ectopic expression of the p53 mutants induced a significant increase in MAP2K3 proteins levels, compared with empty-vector (pcDNA3) transfection. On the contrary, no significant effects were observed upon ectopic expression of wt p53 (Fig. 1A). To confirm these results in a more relevant cellular context, similar experiments were performed with a panel of human cancer cell lines naturally harboring mutations in the TP53 locus. MAP2K3 modulation was assessed upon RNAi depletion of endogenous mutp53 in SKBR3 (mutp53R175H), MDA-MB468 (mutpR273H), MDA-MB231 (mutpR280K), and HT29 (mutpR273H) cell lines. The cells were transduced with either lentiviral vectors carrying shRNAs specific to p53 (sh/p53) or control scrambled shRNA (sh/scr) (11). mutp53 depletion induced a significant reduction in MAP2K3 protein levels in all tested cell lines (Fig. 1B,  lanes 2, 4, 6, and 8). In contrast, similar experiments performed with wt 53-expressing cells (HCT116) showed no modulation of MAP2K3 protein levels (Fig. 1B, lane 10). Overall, these data indicate that, in our cancer cell line panel, MAP2K3 is likely to be a common target of mutp53R273H, mutp53R175H, and mutp53R280K proteins.
p53 Mutants Transactivate the MAP2K3 Promoter-To determine whether mutp53 up-regulates MAP2K3 expression through transactivation of its promoter, we cloned the MAP2K3 gene 5Ј regulatory region (Ϫ989 to Ϫ2) upstream of a luciferase gene, generating the pMAP2K3-Luc vector (see "Experimental Procedures"). The effect of mutp53 on MAP2K3 promoter activity was evaluated by transient transfection assays in our cohort of human cancer cells. Cells infected with either the sh/p53-or the sh/scr-bearing lentiviral vector were transiently transfected 96 h later with either the pBasic-Luc (promoter-less) or the pMAP2K3-Luc vector. Promoter activity was evaluated after 24 h by luciferase assay. Fig. 2A shows that mutp53 actively contributes to MAP2K3 gene expression, though to different extents, in all tested cell lines. Consistently, depletion of endogenous mutp53 induces a significant reduction in luciferase activity. Similar experiments were then performed with p53-null H1299 cells, in which MAP2K3 promoter activity was assessed upon transient expression of either wt or mutp53 proteins (R175H, R273H). The results showed that the baseline MAP2K3 promoter activity, present in control cells, was not affected by wt p53 protein, whereas the ectopic expression of the R175H or R273H mutant induced significant increases in MAP2K3 promoter activity (Fig. 2B). These data indicate that the analyzed cancer-associated p53 mutants can up-regulate MAP2K3 expression via transactivation of the MAP2K3 promoter.
Mutp53 Proteins Are Recruited onto the MAP2K3 Promoter and Modulate Its Activity through the Ϫ499 to Ϫ167 Regulatory Region-To identify the promoter region required for mutp53 to up-regulate MAP2K3 gene expression, we generated three FIGURE 2. mutp53 but not wild-type protein transactivates MAP2K3 promoter. The MAP2K3 gene 5Ј regulatory region was amplified by PCR from genomic DNA and cloned in pGL3-Luc vector (see under "Experimental Procedures" for details). A, transcriptional activity of MAP2K3 promoter was assessed by luciferase assays upon depletion of endogenous mutp53-expressing cells. Cells were transduced with either the sh/p53 or sh/scr lentivirus, and 96 h later, cells were transiently co-transfected with either the pGL3-Luc (promoterless) or pMAP2K3-Luc vector along with the pCMV␤-gal vector (internal control). Luciferase and ␤-galactosidase assays were performed 24 h post-transfection (for details, see under "Experimental Procedures"). Transcriptional activity (Luc) was normalized to ␤-galactosidase activity (transfection efficiency) and protein concentration. Gray bars, sh/scr infected cells; white bars: sh/p53 infected cells. B, p53-null H1299 cells were co-transfected with the pMAP2K3-Luc and pCMV␤-gal (internal control) vectors along with the mutp53R175H, mutp53R273H, or wild-type p53-expressing vectors. 24 h later, luciferase and ␤-galactosidase assays were performed as described. Black bar, control cells; white bar, wt p53-transfected cells; gray bar, mutp53R175H-transfected cells; dark gray bar, mutp53R273H-transfected cells. Means and S.D. of three independent experiments are reported. partial deletions in the cloned promoter, namely pdel1-Luc (Ϫ989 to Ϫ716 deletion), pdel2-Luc (Ϫ782 to Ϫ499 deletion), and pdel3-Luc (Ϫ711 to Ϫ167 deletion) ( Fig. 3A and see "Experimental Procedures"). To assess transcriptional activity, we initially performed experiments with HT29 cells harboring the p53R273H mutation. Cells were transiently transfected with the partially deleted, the full-length (pMAP2K3-Luc), or the promoter-less (pBasic-Luc) reporter vectors, and transcriptional activities were measured 24 h later. The pMAP2K3-Luc vector showed an activity ϳ10-fold greater than the promoterless construct. Similar activities were displayed by the pdel1-Luc and pdel2-Luc constructs. However, the Ϫ711 to Ϫ167 deletion completely abrogated promoter activity (pdel3-Luc, Fig. 3B). For confirmation, experiments were then performed with p53-null H1299 cells where the transcriptional activity of mutp53 proteins on the MAP2K3 promoter was explored by exogenously expressing different p53 mutants. The Luc reporter vectors already described were co-transfected along with either the mutp53R175H-or the mutp53R273H-expressing vector, and transcriptional activity was evaluated 24 h later. Full-length MAP2K3 promoter activity increased significantly upon expression of either p53 mutant protein (Fig. 3C). These two mutants showed similar activities on the pdel1-Luc and pdel2-Luc reporter constructs, whereas the pdel3-Luc construct showed transcriptional activity independent of the presence of p53 mutant proteins (Fig. 3C). In conclusion, our results show that the response to mutp53 proteins maps to the Ϫ499 to Ϫ167 promoter region.
Next, we asked whether mutp53 proteins contribute directly to MAP2K3 up-regulation via physical recruitment onto the MAP2K3 promoter. To this end, chromatin immunoprecipitation assays were performed on chromatin isolated from our human cancer cell lines infected with either the sh/p53 or the sh/scr lentivirus. Cyclin B2 was included in our experiments as a positive control, because it has been reported that the p53R175H mutant protein is recruited onto cell cycle-related gene promoters (25). All endogenous mutants investigated (R273H, R175H, and R280K) were found to be recruited onto the MAP2K3 promoter in all tested cell lines (Fig. 3D, lanes 1, 3,  5, and 7). The specificity of these results was confirmed by the absence of specific amplifications from mutp53-depleted cells. B, HT29 cells were transiently transfected with promoter-less (pBasic-Luc), full-length (pMAP2K3-Luc), or partially deleted mutant constructs along with pCMV␤-gal vector (internal control). C, p53-null H1299 cells were transiently co-transfected with full-length (pMAP2K3-Luc) or partial deleted mutants and pCMV␤-gal (internal control) constructs along with either the mutp53R175H or mutp53R273H expressing vectors. All luciferase and ␤-galactosidase assays were performed 24 h later, and transcriptional activities (Luc) were normalized as reported. White bars, vector reporter alone; black bars, vector reporter and mutp53R175H-expressing vector; gray bars, vector reporter and mutp53R273H-expressing vector. Means and S.D. of three independent experiments are reported. *, p Ͼ 0.05 and **, p Ͼ 0.1. D, multiple mutp53 proteins are physically recruited on the MAP2K3 regulatory region. Chromatins derived from our panel of cancer cell lines either transduced with the sh/p53 or sh/scr lentivirus were immunoprecipitated with anti-p53 (p53, Ab7) or no antibody (No-ab) as negative control. PCR analyses were performed on immunoprecipitated DNA samples, by using a specific set of primers for MAP2K3 and cyclin B2 promoters (see under "Experimental Procedures"). Similar PCR analyses were performed in an adjacent region (Ϫ3200 bp, negative control probe) to MAP2K3 regulatory region chosen to verify the specificity of achieved results. A representative panel of several analyses is reported.
Furthermore, no amplification of an adjacent region (Ϫ3200 bp) on the MAP2K3 promoter was detected. These findings indicate that up-regulation of MAP2K3 occurs through in vivo recruitment of mutp53 proteins onto specific MAP2K3 regulatory regions.

Mutp53 Regulates MAP2K3 Gene Expression through the NF-Y and NF-B Transcriptional Cofactors-
Studies published in the past few years indicate that mutp53 may exert its GOF activity via transcriptional regulation of target genes through the formation of large transcriptional complexes (reviewed in Ref. 34). To identify transcriptional cofactors through which mutp53 might regulate MAP2K3 expression, we searched for consensus binding sites using the MatInspector software tool. An analysis of the mutp53-responsive promoter region (Ϫ499 to Ϫ167) showed that it includes consensus sequences for several transcription factors, but no TATA box. Therefore, among the transcription factors potentially binding this region, we focused on those known to functionally interact or form complexes with mutp53, such as NF-Y (25) and NF-B (20,26,27) (Fig. 4A). To ascertain their involvement in mutp53-mediated MAP2K3 up-regulation, we first analyzed the in vivo recruitment of NF-Y and NF-B on MAP2K3 regulatory regions. Chromatin immunoprecipitation analyses were performed on chromatin from HT29 cells transduced with either the sh/p53 or sh/scr lentivirus. Chromatin was immunoprecipitated with antibodies to NF-YB, NF-B (p50), or acetylated histone H4 (H4ac). Acetylated histone H4 was chosen to assess the accessibility of the chromatin regions analyzed. The results showed that both transcriptional cofactors are recruited to the MAP2K3 regulatory region. In particular, NF-Y occupancy did not vary between the analyzed conditions (sh/p53 and sh/scr cells) (Fig. 4B, left panel, lane 1), in agreement with published data (24). In contrast, a reduced recruitment of NF-B (p50) was observed in sh/p53, compared with sh/scr cells (Fig. 4B, left panel, lane 3). Acetylated histone 4 behaved similarly to NF-B (Fig. 4B, left  panel, lane 2), further confirming that the MAP2K3 promoter activity was significantly reduced by mutp53 depletion. Quantitative PCR measurements confirmed quantitatively the observed effect (Fig. 4B, right panel). The absence of signal in an adjacent region (Ϫ3200 bp) demonstrated the specificity of the results. To rule out the possibility that mutp53 affects the steady state levels of NK-B, thereby altering MAP2K3 expression, we monitored NF-B protein in H1299 cells upon exogenous expression of either the mutp53R175H or mutp53R273H protein. Supplemental Fig. 1 shows that ectopic expression of the mutants does not modify NF-B protein levels with respect to control cells.
To evaluate the functional involvement of NF-Y and NF-B in mutp53-mediated MAP2K3 upregulation, we exploited dominant-negative mutants of NF-Y (YA13m29) (35) or NF-B (IB␣ super-repressor) (36). p53-null H1299 cells were transiently cotransfected with the full-length promoter (pMAP2K3-Luc) along with the mutp53R175H or mutp53R273H or empty (pcDNA3) vector and increasing amounts of either YA13m29-or IB super-repressorexpressing vector. Transcriptional activity was measured 24 h later. These experiments showed that the increased MAP2K3 promoter activity induced by p53 mutants was significantly repressed in the presence of either dominant-negative mutant (Fig. 4C). Taken together, our findings indicate that mutp53 up-regulates MAP2K3 expression through the involvement of NF-Y and NF-B transcriptional cofactors.
Knockdown of Endogenous MAP2K3 Protein Interferes with mutp53 GOF Effects-To evaluate whether MAP2K3 contributes to mutp53 GOF activity, we explored the biologic effects of depleting the endogenous MAP2K3 protein by RNAi. To this end, we constructed a lentiviral vector bearing a MAP2K3 shRNA (see "Experimental Procedures"). HT29 cells conditionally expressing the MAP2K3 or the scr shRNA were generated as described previously (11). The engineered cell lines were then challenged with doxycycline (ϩDox), and MAP2K3 depletion efficiency was monitored. Maximal depletion of endogenous MAP2K3 was achieved after 144 h of ϩDox treatment (Fig. 5A,  left panel). In addition, because MAP2K3 is a specific upstream activator of p38 MAPK (28), we evaluated whether silencing MAP2K3 compromises its signaling cascade by monitoring the phosphorylation status of ATF2, a downstream factor of p38 MAPK. Indeed, depletion of MAP2K3 causes a significant reduction in phospho-ATF2 at 144 h of ϩDox treatment (Fig. 5A, left panel). Accordingly, depletion of the endogenous mutp53R273H protein sharply reduced ATF2 phosphorylation (Fig. 5A, right panel). Together, these results suggest that the knockdown of mutp53 and the ensuing reduc- tion of MAP2K3 levels impair p38 MAPK signaling. We previously reported that the knockdown of endogenous mutp53 impacts on the proliferation and survival of HT29 and SKBR3 cells (10). To explore whether MAP2K3 depletion affects mutp53 GOF activities on cell proliferation and cell survival, engineered HT29 cells were plated in the presence (ϩDox) or absence (ϪDox) of doxycycline and total cell numbers, cell viability, and cell cycle profiles were analyzed in time. MAP2K3 depletion strongly impaired cell proliferation (Fig. 5B, left panel) and cell survival (Fig. 5B, right panel) and induced significant cell accumulation in G 2 /M phase (Fig. 5C).
Similar experiments were performed with our other cancer cell lines. We first established that mutp53 enhances proliferation and survival of MDA-MB231 (R280K) and MDA-MB468 (R273H) cell lines. Similarly to HT29 and SKBR3 cells (10), depletion of mutp53R280K and mutp53R273H, respectively, in MDA-MB231 and MDA-MB468 cells compromises cell proliferation and survival (supplemental Fig. 2). Afterward, we explored whether depletion of MAP2K3 compromises proliferation and survival of our other cancer cell lines. Efficient knockdown of MAP2K3 (supplemental Fig. 3, A-C) affects prolifera-tion of all studied cell lines (Fig. 6, A-C, left) and survival of MDA-MB468 and SKBR3 cells (Fig. 6, A-C, right). One possible reason of the different sensitivity to survival of MDA-MB231 may be due to additional survival signals provided by expression of activated mutant Ras protein in these cells (37). Overall, our findings suggest that, in our panel of cancer cell lines, MAP2K3 up-regulation contributes to mutp53 GOF activities. We therefore wished to determine whether MAP2K3 expression might be relevant only for mutp53 expressing cells or might represent a more general required factor. To this aim, we performed RNAi experiments with p53-null H1299 cells. We found that efficient MAP2K3 depletion (supplemental Fig. 4A) significantly compromises H1299 cell proliferation (supplemental Fig. 4B). In conclusion, our data suggest MAP2K3 that MAP2K3 is a more generally required factor for cell proliferation.
Ectopic Expression of MAP2K3 Can Rescue the Proliferative Defect Induced by Knockdown of mutp53-To assess whether MAP2K3 is enable to mediate mutp53 GOF activities, we stably expressed exogenous MAP2K3 in our cohort of human cancer cell lines. Results show MDA-MB468 cells as more suitable cell hosts for stable and sustained expression of ectopic MAP2K3 (Fig. 7A, left panel). To measure the effect of MAP2K3 on mutp53 GOF activities, stable transfected cells were infected with either the sh/p53 or sh/scr lentivirus, and cell proliferation was assessed. Depletion of mutp53R273H markedly reduces proliferation of empty vector-transfected cells (pcDNA3) (Fig. 7A, right panel). In contrast, upon stable ectopic expression of MAP2K3, depletion of mutp53 had little or any effect on cell proliferation of MDA-MB468 cells (Fig. 7A,  right panel). These findings suggest that mutp53 is required for cell proliferation of MDA-MB468 cells and ectopic expression of MAP2K3, as a target of mutp53, is sufficient to compensate for mutp53 knockdown.
Finally, to highlight the relevance of our studies, we evaluated the MAP2K3 expression of primary tumors by querying public gene expression data repositories (Oncomine 4 research edition) (38). Analysis of data sets obtained from specimens of primary tumors revealed that the MAP2K3 transcription is significantly higher in breast (39) and colon (40) cancers carrying p53 mutations in comparison with those carrying wt p53 protein (Fig. 7B). In conclusion, our results confirmed MAP2K3 as a mutp53 target gene and provided evidence that one of the mechanisms by which mutp53 exerts its GOF activity is through the up-regulation of MAP2K3 expression.

DISCUSSION
Recently reported data indicate that the MAP2K3 up-regulation is involved in invasion and progression of glioma and breast tumors. In this study, we show the first evidence that different endogenous mutp53 proteins transcriptionally upregulate MAP2K3 expression in diverse human tumor cell lines. Ample data indicate that mutp53 proteins do not lose only their tumor suppressive functions, but they do gain new abilities that promote tumorigenesis by influencing cancer cell transcriptome and phenotype. In particular, we found that knockdown of mutp53 inhibits, whereas ectopic expression of mutp53 increases MAP2K3 transcription. Contrarily, no significant effects were observed upon ectopic expression of wt p53 in p53-null H1299 cells or upon depletion of endogenous wt p53 in the HCT116 cell line. These data showed that, in our panel of cancer cell lines, MAP2K3 is likely to be a common target of mutp53 proteins.
By analysis for deletions, we identified a MAP2K3 regulatory region (Ϫ499 to Ϫ167) required to mutp53 proteins to modulate MAP2K3 expression. We found that MAP2K3 up-regulation occurs by physical recruitment of p53 mutant proteins onto the MAP2K3 regulatory region. Moreover, we found that NF-Y and NF-B transcriptional cofactors are recruited on MAP2K3 promoter and thus required for MAP2K3 up-regulation. Consistently with data found in the literature, our findings are indicative that both transcriptional cofactors are relevant players in the mutp53 transcriptome. Their involvement in biologic processes such as cell proliferation (NF-Y) and inflammation (NF-B) highlight the relevance of their contribution in mutp53 GOF activity.
To better understand the involvement of MAP2K3 up-regulation in the mutp53 GOF activity, we explored biologic effects linked to modulation of the MAP2K3 expression in our panel of cancer cell lines. We found that knockdown of endogenous MAP2K3 inhibits cell proliferation and survival. In contrast, the exogenous expression of MAP2K3 is sufficient to compensate alterations of cell proliferation upon silencing of mutp53.
Moreover, by querying public gene expression data repositories, we showed a higher expression of MAP2K3 mRNA in specimens of human primary tumors harboring p53 mutations in respect to those expressing wt p53 protein. In conclusion, our present study suggests that one possible mechanism through which mutants of p53 may acquire their GOF activities is via up-regulation of MAP2K3.