P120-catenin isoforms 1 and 3 regulate proliferation and cell cycle of lung cancer cells via β-catenin and Kaiso respectively

Abstract

Background: The different mechanisms involved in p120-catenin (p120ctn) isoforms' 1/3 regulation of cell cycle progression are still not elucidated to date.

Methods and findings: We found that both cyclin D1 and cyclin E could be effectively restored by restitution of p120ctn-1A or p120ctn-3A in p120ctn depleted lung cancer cells. When the expression of cyclin D1 was blocked by co-transfection with siRNA-cyclin D1 in p120ctn depleted cells restoring p120ctn-1A or 3A, the expression of cyclin E was slightly decreased, not increased, implying that p120ctn isoforms 1 and 3 cannot up-regulate cyclin E directly but may do so through up-regulation of cyclin D1. Interestingly, overexpression of p120ctn-1A increased β-catenin and cyclin D1 expression, while co-transfection with siRNA targeting β-catenin abolishes the effect of p120ctn-1A on up-regulation of cyclin D1, suggesting a role of β-catenin in mediating p120ctn-1A's regulatory function on cyclin D1 expression. On the other hand, overexpression of p120ctn isoform 3A reduced nuclear Kaiso localization, thus decreasing the binding of Kaiso to KBS on the cyclin D1 promoter and thereby enhancing the expression of cyclin D1 gene by relieving the repressor effect of Kaiso. Because overexpressing NLS-p120ctn-3A (p120ctn-3A nuclear target localization plasmids) or inhibiting nuclear export of p120ctn-3 by Leptomycin B (LMB) caused translocation of Kaiso to the nucleus, it is plausible that the nuclear export of Kaiso is p120ctn-3-dependent.

Conclusions: Our results suggest that p120ctn isoforms 1 and 3 up-regulate cyclin D1, and thereby cyclin E, resulting in the promotion of cell proliferation and cell cycle progression in lung cancer cells probably via different protein mediators, namely, β-catenin for isoform 1 and Kaiso, a negative transcriptional factor of cyclin D1, for isoform 3.

Figure 1. Cyclin D1 and cyclin E expression were significantly decreased in A549 cells with knocked down p120ctn.

(A) The results of western blot analysis showed that proteins cyclin D1 (*, p = 0.001) and cyclin E (*, p = 0.004) were significantly decreased after p120ctn was knocked down in A549 cells. (B and C) The results of MTT and flow cytometry showed suppressed cell proliferative ability, increased G₁ phase cells (*, p = 0.001) and reduced S phase cells (*, p = 0.001) were detected in A549 cells with knocked down p120ctn. All comparisons were made to the groups of A549 cells or the cells transfected with vector alone.

Figure 2. Restitution of p120ctn-1A or p120ctn-3A in p120ctn depleted A549 cells could restore cyclin D1 and cyclin E.

(A) p120ctn-1A or 3A plasmids were transfected in A549 cells depleted of p120ctn (si-p120+1A or si-p120+3A), showing the significantly recovered protein levels of cyclin D1 (*, p<0.001, **, p<0.001) and cyclin E (*, p<0.001, **, p<0.01). (B and C) The results of MTT and flow cytometry showed that cell proliferation was effectively restored (p<0.01), G₁ phase cells were significantly decreased (*, p = 0.001, **, p = 0.004) and S phase cells were significantly increased (*, p = 0.004, **, p = 0.028) after p120ctn depleted A549 cells were transfected with p120ctn-1A or 3A. All the comparisons are made to the groups of A549 cells and the cells transfected with vector alone.

Figure 3. p120ctn promotes cell cycle progression by up-regulating cyclin D1, which subsequently enhances cyclin E expression.

(A) Cyclin D1 depletion by siRNA in A549 cells led to reduced cyclin E expression, but conversely, the expression of cyclin D1 was not significantly changed (p = 0.944) in cells with knocked down cyclin E by siRNA, suggesting a unidirectional regulatory relationship between cyclin D1 and cyclin E. (B and C) Cyclin D1 depletion suppressed cell proliferation (p<0.01), elevated G₁ phase cells (p = 0.016) and decreased S phase cells (p = 0.009).(D) After co-transfection of siRNA-cyclin D1 with p120ctn isoform 1 or 3 in the cells depleted of p120ctn for 48 hours, the expression of cyclin E was not increased but was slightly decreased. All these results suggest that p120ctn promotes cell cycle progression by up-regulating cyclin D1, which subsequently enhances expression of cyclin E. The comparison is made to the control group of cells transfected with non-targeted siRNA or vector alone.

Figure 4. p120ctn-1A, but not p120ctn-3A, increases the protein expression of β-catenin.

SPC-K2 cells were transfected with p120ctn-1A or 3A respectively, and the expression of β-catenin and kaiso was measured by western blot (A) and RT-PCR (B). The protein expression of β-catenin was significantly increased (*, p = 0.001) in the cells transfected with p120ctn-1A cDNA, but it showed no significant change (**, p = 0.769) in p120ctn-3A overexpressing SPC-K2. Neither p120ctn-1 nor 3 could affect the transcription of β-catenin (*p = 0.463, **p = 0.401) or the expression of Kaiso. All the comparisons are made to the group of cells transfected with vector alone.

Figure 5. p120ctn isoform 1 could up-regulate cyclin D1 expression through β-catenin.

The expression of active β-catenin and cyclin D1 were significantly reduced in A549 cells (β-catenin, + p<0.001, ++ p<0.001, cyclin D1, + p = 0.003, ++ p = 0.004) when β-catenin or p120ctn was interfered (si-β-cat or si-p120). The expression of active β-catenin (* p = 0.001) and cyclin D1 (* p<0.001) rebounded when p120ctn-1A expression was reconstituted by p120ctn-1A cDNA transfection (si-p120+1A). Co-transfection of p120ctn-1A and the siRNA targeting β-catenin (si-p120+1A+si-β-cat) showed no change of cyclin D1 (^Δ p = 0.248) expression, suggesting that p120ctn-1A up-regulates cyclin D1 via β-catenin. Restoration of p120ctn-3A (si-p120+3A) did not increase the expression of active β-catenin (** p = 0.891) but could still restore the expression of cyclin D1 (** p = 0.001), implying that p120ctn-3 up-regulates expression of cyclin D1 independent of β-catenin. Co-transfection of p120ctn-3A and the siRNAs targeting p120ctn/β-catenin (si-p120+3A+si-β-cat) could significantly increase cyclin D1 expression in A549 cell line in comparison with the group of p120ctn/β-catenin siRNA transfection alone (si-p120+si-β-cat, ^ΔΔ p<0.01), whereas, co-transfection of p120ctn-3A seems to have no impact on the levels of active β-catenin. Similar results were obtained in SPC cells. Note, although cyclin D1 expression is restored by restitution of either p120ctn 1A or p120ctn 3A, the rescue effect is more prominent by p120ctn 1A than by p120ctn 3A. The difference between the group si-p120+3A and the group si-p120+3A+si-β-cat may be explained by an effect of residual endogenous β-catenin in the former. This effect of residual endogenous β-catenin can also be seen in the difference between si-p120 and si-p120+si-β-cat in both cell lines. A synergistic effect resulting in additional decrease in both β-catenin and cyclin D1 can be seen in the latter group.

Figure 6. N-56-101 amino acid residues of p120ctn-1 are essential for up-regulating β-catenin.

Two deletion mutants of p120ctn-1, M1 and M2 were transfected in p120ctn depleted A549 cells. The results showed that M1 could up-regulate β-catenin (*** p<0.001), while M2 could not (**** p = 0.175). M1 could up-regulate the expression of cyclin D1 (*** p<0.001) but M2 could not (**** p = 0.906). Therefore, it may be due to deletion of N-56-101 amino acid residues for p120ctn-3 to fail to regulate β-catenin. All the comparisons are made to the group of cells co-transfected with vector alone.

Figure 7. Cyclin D1 is one of downstream target genes of Kaiso.

(A) Western blot analyses show the increased expression of Kaiso in A549 cells transfected with Kaiso cDNA. Kaiso overexpression remarkably down-regulated the expression of cyclin D1 (p = 0.000) and cyclin E (p = 0.003). (B) Western blot analyses show reduced expression of Kaiso in SPC cells transfected with Kaiso siRNA. Kaiso interference significantly increased the expression of cyclin D1 (p = 0.001) and cyclin E (p = 0.012). Overall findings suggest that Kaiso could regulate the expression of cyclin D1 and cyclin E. (C) Chromatin immunoprecipitation (ChIP) assay confirmed the Kaiso monoclonal antibody could precipitate the cyclin D1 gene promoter fragment containing KBS. Kaiso could bind to KBS of cyclin D1 promoter. All the comparisons are made to the group of cells transfected with vector alone.

Figure 8. Only p120ctn-3 could bind to Kaiso in vivo.

(A and B) Co-immunoprecipitation assays showed that p120ctn monoclonal antibody (the lower panel of left column in Figure 8A), but not specific p120ctn-1,2 monoclonal antibody (6H11) (the lower panel of right column in Figure 8A), can effectively precipitate Kaiso protein both in the nucleus and cytoplasm (Figure 8B). However, Kaiso monoclonal antibody cannot effectively precipitate p120ctn (the upper panel of middle column in figure 8A). (C) Co-immunoprecipitation with sufficient p120ctn mAb after overexpression of p120ctn-1 or 3 isoform showed that overexpression of p120ctn-3 led to more Kaiso protein precipitated in A549 cells. No change was detected when p120ctn isoform 1 was overexpressed, compared with the control. Specific p120ctn-1, 2 monoclonal antibodies (6H11) could not precipitate Kaiso when p120ctn isoform 1 was significantly increased. Note the presence of p120ctn after coprecipitation with 6H11 but absence of kaiso in the precipitate, suggesting there is no significant interaction between p120 isoform 1, 2 and kaiso. Since there are mainly p120ctn isoforms 1 and 3 in both of the lung cancer cell lines, we thought that Kaiso might bind to p120ctn isoform 3 but not to p120ctn isoform 1 in vivo.

Figure 9. p120ctn-3 inhibit Kaiso from binding to cyclin D1 promoter by regulating the nuclear/cytoplasmic shuttle of Kaiso.

(A) Transfection of p120ctn-3A in A549 cells significantly increased Kaiso in the cytoplasm (CYTO, *, p = 0.000) and reduced Kaiso in the nucleus (NE, *, p = 0.000). However, no significant difference of subcellular localization of Kaiso was found between p120ctn-1A overexpression cells and control cells (CYTO, **, p = 0.135, NE, **, p = 0.774). These results suggest that a high level of p120ctn-3A may be able to promote the nuclear export of Kaiso, but p120ctn-1A does not appear to have this function. Incubating cells transfected with p120ctn-3A with LMB or transfecting NLS-p120ctn-3A in A549 cells increased nuclear p120ctn-3A (*, p = 0.000, or +, p = 0.000) and nuclear Kaiso (NE, +, p = 0.000 or ++, p<0.001), but reduced cytoplasmic Kaiso (CYTO, +, p = 0.000 or ++, p = 0.002) compared to the cells with only overexpressed p120ctn-3A, implying that the nuclear export of Kaiso is likely to be p120ctn-3-dependent. (B) p120ctn-1A overexpression did not change the binding of Kaiso to KBS on the cyclin D1 promoter, whereas p120ctn-3A overexpression significantly reduced the binding of Kaiso to KBS on the cyclin D1 promoter.