Colorectal cancer patients with mutant p53 overexpress β-catenin and the canonical Wnt target gene MYC

TP53 is one of the most frequently altered suppressor genes observed in solid tumors. As the first approach to investigating the possible correlation between altered TP53 gene expression and canonical Wnt signaling in colorectal cancer patients, an in silico analysis was performed using the TCGA public database (Fig. 1A). To accomplish a comparison of colorectal cancer tissues with normal ones, we used the GEPIA platform to make comparisons of mRNA expression levels of TP53 (p53), CTNNB1 (β-catenin), and a typical target gene of p53, CDKN1A (p21). As shown in Fig. 1, TP53 and CTNNB1 were found to be significantly overexpressed in colorectal cancer compared to normal tissue (Fig. 1A, B). This effect was not observed with CDKN1A (p21), consistent with the alteration and nonfunctionality of wtp53 in this type of cancer (Fig. 1C). Comparing the expression of c-Myc, a target gene of the canonical Wnt pathway, in wtp53 versus mut-p53 colorectal samples, it was clearly shown that those patients with altered p53 display higher levels of c-Myc expression (Fig. 1D), which was also observed in the TCGA Pancancer database, in which a higher expression of c-Myc was also found in colorectal cancer patients expressing mut-p53 compared to wt-p53 expression (Fig. 1E). Interestingly, the expression of CDKN1A was higher in those tissues with endogenous wtp53 (Fig. 1E). These results suggest, therefore, that canonical β-catenin Wnt signaling may be more active in mutp53-expressing patients and that it may be related to TP53 status in colon cancer cells.

Fig. 1

Patients with p53 alterations in colon adenocarcinoma samples show high expression of MYC, a canonical Wnt target. A–C Colon adenocarcinoma samples from patients show increased levels of TP53 and CTNNB1 but not CDKN1A in the GEPIA database. D Expression of canonical Wnt target (Myc) in normal and tumoral colon adenocarcinoma samples obtained from the GEPIA database. E. Expression of CDKN1A (p21) and MYC (c-myc) in tumoral samples classified by p53 status (wild-type and mutant) in data obtained from TCGA-Pan Cancer. Graphs represent RNA-Seq by expectation–maximization (RSEM). Nonparametric analysis was used for TCGA data, and the graph represents the median with the 95% confidence interval, **p < 0.01

The TP53-null H1299 cells transfected with wt-p53 or mut-p53 show that only wt-p53 diminishes β-catenin levels

To investigate the specific effects of wt-p53 compared with mut-p53 in Wnt/β-catenin signaling, we first employed as a model the TP53-null H1299 cell line to ectopically express wtp-53 or the mutated versions R175H-p53 and R273H-p53. We transfected increasing concentrations of wt-p53 or empty vector and examined its effect on total β-catenin levels. The results presented in Fig. 2A show that wt-p53 was able to negatively affect β-catenin expression levels in a dose-dependent manner. When we transfected the mutated versions of p53 or wt-p53 to compare their effects on β-catenin levels, we observed again that only wt-p53, but not its mutants, diminished β-catenin levels at 72 h post-transfection (Fig. 2B). We only detected the expression of p21, a typical wt-p53 gene target, in cells expressing wt-p53, corroborating the specific effects of wt-p53 (Fig. 2B). Interestingly, when we compared the protein levels of wtp53 and mutp53, 24 h and 72 h after transfection, our results suggested that the mutant versions of p53 are more stable than wt-p53 (Fig. 2B), but most importantly, after 72 h, β-catenin levels were not increased as a result of mut-p53-R273H expression (the same version of mut-p53 expressed in colon cancer cells) with respect to the levels found in nontransfected or empty controls.

Fig. 2

Mutant p53 gain-of-function acts as a positive regulator of canonical Wnt signaling, while wild-type p53 version represses it in the H1299 cell line. A Graph and representative immunoblot of β-catenin corresponding to the H1299 cell line (null to TP53) transfected with different doses of wtp53 or empty vector; GAPDH was used as a loading control. B Graph and representative immunoblots of β-catenin, p53, and p21 obtained from overexpression assays of cells transfected with empty vector, wtp53, R175H p53, and R273H p53 for 72 h. C Immunoblot of p53 and comparation of the sizes of spheres obtained from the H1299 cancer cell line transfected with empty vector, wtp53, R175H p53, and R273H p53. Representative spheres were grown for 72 h. D Representative immunoblots corresponding to p-GSK3β Ser9 and p-AKT Ser473 from transfected H1299 spheres and treated with either wortmannin inhibitor (150 nM) or DMSO control; GAPDH and total AKT were used as loading controls. Densitometric analysis from at least three independent experiments is shown in graphs, (means ± SEM). *p < 0.05, **p < 0.01, ***p < 0.001

Then, we explored the effects of p53 and its mutants on sphere-forming ability in H1299 spheroid cultures. In this context, TP53-null H1299 cells were transfected with wt-p53 or mut-p53 and cultured in ultra-low-adherence plates for 72 h as described in the Methods section. Nontransfected cells were considered as controls. The expression of wt-p53 and mut-p53 at 72 h after transfection were validated by western blotting, as can be seen in Fig. 2C. We observed that spheroids expressing wt-p53 were significantly smaller than those formed in cells transfected with empty vector or with mut-p53 versions (Fig. 2C), indicating that wild-type p53 negatively affects cell proliferation. It has been reported that mut-p53 may stimulate proliferation in other cancer types (breast and prostate cancer) by activating the PI3K axis (Muller et al. 2009; Valentino et al. 2017). In turn, PI3K/Akt can stimulate canonical Wnt signaling by inducing the inhibition of GSK-3β via its phosphorylation at Ser 9. To explore the mechanism by which wt-p53 versus mut-p53 affect H1299 spheroid proliferation, H1299-transfected spheroids were treated in the absence or the presence of the PI3K inhibitor wortmannin, and the effect on GSK-3β phosphorylation status at Ser 9 (inactivation), or the Akt phosphorylation status at Ser 473 (activation), was examined by western blotting. The results shown in Fig. 2D indicate that in the absence of wortmannin, mut-p53 favored a GOF that increased p-GSK-3β Ser 9 levels, and therefore may induce activation of canonical Wnt by stabilizing β-catenin levels, while wt-p53 or empty vector did not show this effect. Interestingly, we observed that wortmannin treatment not only greatly reduced GSK-3β inactivation but also induced a significant reduction in p-AKT Ser 473, which resulted in greater inhibition under conditions where mutant versions of p53, particularly the mut-p53-R273H version, were expressed, indicating that mut-p53 GOF activates the PI3K/AKT axis (Fig. 2D).

Mutant p53 stimulates the Wnt/β-catenin pathway in colon cancer cells

Once we knew in a null-TP53 cell context model that mut-p53 induces β-catenin stabilization levels and confirmed that wt-p53 induces the opposite, negatively regulating the canonical Wnt pathway, we focused on investigating the role played by mut-p53 on the canonical Wnt pathway in colon cancer cells. As described in detail in the Materials and Methods section, we used RKO cells expressing wild-type p53 and wild-type APC (normal Wnt activation) as well as SW480 or SW620 cells expressing mut-p53 (TP53 R273H) with constitutively active canonical Wnt signaling because they express a truncated version of APC. Consistent with this, the analysis of the active (non-phosphorylated) β-catenin levels found in SW480 and SW620 by western blotting were much higher in these cells compared with those found in RKO cells, as can be observed in Fig. 3A. Interestingly, it can also be observed in this Figure that endogenous mut-p53 expression levels are much higher in SW480 and SW620 cells than the wt-p53 levels expressed in RKO cells. We then blocked the mut-p53 expression by transient knockdown with siRNA and evaluated the knockdown efficiency using western blotting. Figure 3B shows that expression of mut-p53 was decreased considerably in SW480 cells transfected with the siRNA plasmid with respect to those transfected with the scramble control plasmid (Fig. 3B). Then, we analyzed the effect of mut-p53 knockdown on transcriptional activity mediated by β-catenin in SW480 cells. SW480 cells were co-transfected with siRNA p53 and luciferase reporter plasmid pTOP or with the reporter control plasmid pFOP. The β-catenin-dependent transcriptional activity was measured 72 h after co-transfection. The results in Fig. 3C clearly show that mut-p53 knockdown decreased β-catenin transcriptional activity. Consistent with this, increasing the concentration of siRNA mut-p53 blocked the expression of c-myc, a typical canonical Wnt gene target, in a dose-dependent manner (Fig. 3D). To study the effect of mut-p53 knockdown in a functional role, we performed a colony formation assay in scramble- or shp53-transduced SW480 cancer cells. In line with findings that silencing mutant p53 diminishes canonical Wnt pathway activation, we observed a significantly decreased capacity of mut-p53-silenced cells to form colonies compared to control (scramble) cells (Fig. 3E). Altogether, these results indicated that mut-p53 stimulates canonical Wnt signaling in colon cancer cells.

Fig. 3

Mutant p53 knockdown in the SW480 cell line decreases canonical Wnt pathway. A Comparation of non-phosphorylated β-catenin (active) and p53 protein levels measured in different colon cancer cell lines with different TP53 status, RKO (wild-type p53), SW480 (R273H p53), and SW620 (R273H p53). Densitometric analysis of immunoblots of p53 and active β-catenin was performed using GAPDH as a loading control. B Representative immunoblots and graph of p53 protein in SW480 under different doses of siRNA p53 (25–100 nM); the scramble condition was used as a control (100 nM). C Luciferase activity measured with the TOP/FOP system in both scramble and siRNA p53 (100 nM) conditions. The relative units were normalized to Renilla activity. D Representative immunoblot and graph of c-myc normalized to GAPDH. E. Representative colonies and relative area quantification corresponding to scramble and shp53 conditions of SW480-transduced cells. Graphs represent densitometric analysis from at least three independent experiments (means ± SEM). *p < 0.05, **p < 0.01, ***p < 0.001

Mut-p53 stimulates canonical Wnt signaling through Akt-mediated β-catenin Ser 552 phosphorylation

Because colon cancer cells co-expressing mut-p53 and truncated version of APC, such as SW480 cells, do not possess a functional β-catenin degradation complex, we reasoned that the inhibition of GSK-3β by its phosphorylation at Ser 9 observed in a null-p53 H1299 cell line expressing mut-p53 cannot affect the β-catenin levels in colon cancer SW480 cells, and thus, another mechanism must exist in these cells to explain how mutant p53 can stimulate β-catenin-mediated transcriptional activity. In this regard, it has been demonstrated in HCT116 colon cancer cells and in intestinal organoid cultures that phosphorylation of β-catenin at Ser 552 by AKT contributes to β-catenin stability and enhances its transcriptional activity (Behrouj et al. 2021; Wang et al. 2020; Fang et al. 2007).

To explore if mut-p53 GOF in colon SW480 cells activate Akt to induce the phosphorylation of β-catenin at Ser 552 (depicted in Fig. 4A), we first corroborated that the knockdown of mut-p53 in SW480 cells does not affect the active β-catenin levels in these cells. Indeed, as shown in Fig. 4B, the levels of active β-catenin were not significantly modified by mut-p53 knockdown. However, mut-p53 silencing decreased p-β-catenin Ser552 and Akt activation, visualized as a decrease in p-Akt Ser 473, in a dose-dependent manner of siRNA (Fig. 4C). To confirm that Akt produces β-catenin Ser 552 phosphorylation, we made use of the Akt-specific inhibitor AZD-5363. The results presented in Fig. 4D clearly show that β-catenin Ser 552 phosphorylation was diminished in a dose-dependent manner by the Akt-specific inhibitor.

Fig. 4

Mutant p53 knockdown decrease p-β-catenin Ser552 in a specific manner of AKT axis. A Schematic representation of the mechanisms influencing β-catenin activity in colon cancer cells. B Representative immunoblot and graph of non-phosho β-Catenin (active); GAPDH was used as a loading control. C Representative immunoblots and graphs of p-β-catenin Ser552 normalized to total β-catenin and p-AKT Ser473 normalized to total AKT. D Representative immunoblot and graph of p-β-catenin Ser552 normalized to total β-catenin measured in SW480 cancer cells treated with different doses of AKT inhibitor (AZD-5363) in the range from 0.1 to 2.0 μM; DMSO was used as control. Densitometric analysis from at least three independent experiments is shown in graphs, (means ± SEM). *p < 0.05, **p < 0.01, ***p < 0.001

Altogether, these results suggest that the mechanism by which mutant p53 stimulates canonical Wnt activation in colon cancer cells with a nonfunctional β-catenin degradation complex is by Akt-mediated phosphorylation of β-catenin at Ser 552.

Mut-p53 participates in the induction of chemoresistance to 5-FU treatment in colon cancer cells

We recently reported that canonical Wnt activation plays an essential role in inducing chemoresistance to 5-Fluorouracil (5-FU) in colon cancer cells (Moreno-Londoño et al. 2023). Given that in this study, we found that mut-p53 activates β-catenin/Wnt signaling, we then investigated whether mut-p53 could also participate in this resistance induction. First, we measured the effect of increasing concentrations of 5-FU for 72 h on cell viability using RKO cells expressing wt-p53 and SW480 or SW620 cells expressing mut-p53. In agreement with our previous reports, we observed that RKO cells are more sensitive to 5-FU (IC-50 value of 10.79 μM) than SW480 or SW620 cells, which are highly resistant to 5-FU (149.6 and 817.1 μM respectively) (Fig. 5A). Interestingly, when we blocked wt-p53 expression in RKO cells, 5-FU induced cell death, visualized as cleaved-PARP (c-PARP) and cleaved Caspase 3 apoptosis markers, was significantly diminished, as can be observed in Fig. 5A.

Fig. 5

The functionality of p53 can determine the cytotoxic effect of 5-FU. A Dose–response curve of the RKO, SW480, and SW620 cell lines treated with 5-FU for 72 h; IC50 values were calculated on the basis of cell viability determined by MTT assay. Estimated IC50 values were RKO = 10.79 μM, SW480 = 149.6 μM, and SW620 = 817.1 uM. Representative immunoblot and graphs of cleaved caspase 3 and cleaved PARP in the RKO cell line transfected with scramble or p53 siRNA for 24 h and then treated with 5-FU at the IC50 value. Densitometric analysis was performed using total protein with ponceau staining. B Immunoprecipitation assay (IP) for MDM2 or p53 were measured in the cellular lysates of SW480 cell line treated for 72 h with DMSO or RITA at 1 μM. Western blot analysis for each antibody was performed and input lysate 10% was considered as control; arbitrary units were determined considering the values of IgG light chain. C Cell viability assay of the SW480 cell line co-treated with 5-FU in curve response doses and RITA at 1 μM for 72 h. Graphs represent the analysis from at least three independent experiments, (means ± SEM). *p < 0.05, **p < 0.01, ***p < 0.001

When we tried to investigate the effect of transient knockdown of mut-p53 on 5-FU exposure of SW480 cells, we noticed that it is very difficult to efficiently obtain a decrease in the amount of mut-p53 protein levels in these cells (data not shown). Thus, we decided to measure the effect of reestablishing canonical functions of wt-p53 in SW480 cells employing the small-molecule RITA (Reactivation of p53 and Induction of Tumor cell Apoptosis) (Grinkevich et al. 2009; Zhao et al. 2010), as depicted in Fig. 5B. It has been described that RITA can reactivate mut-p53 to induce its canonical tumor suppressor functions. Although RITA mechanism action is not clearly defined, it has been reported that the main mechanisms by which RITA acts is by disrupting the interaction of the MDM2/p53 complex. We validated this effect of RITA in mut-p53-expressing SW480 cells employing a reciprocal co-immunoprecipitation assay. Figure 5B shows how RITA treatment produced a significant decrease in the MDM2/p53 interaction (Fig. 5B) and importantly, that RITA treatment re-established the negative effect on active β-catenin levels in a dose-dependent manner, mimicking the action of wild-type p53 on β-catenin levels (Fig. 5C). Finally, we analyzed the effect of both RITA and 5-FU treatment on cellular viability of SW480 cancer cells. As can be seen in Fig. 5C, RITA treatment alone decreased the cellular viability; however, the combination of 5-FU and RITA produced an additive effect, making SW480 more sensitive to 5-FU treatment (Fig. 5C). Taken together, our results indicate that mut-p53 expression in colon cancer cells favors the induction of chemoresistance to 5-FU and suggest that it may be by stimulating the canonical Wnt pathway in colon malignant cells.