Nucleophosmin Protein Dephosphorylation by DUSP3 Is a Fine-Tuning Regulator of p53 Signaling to Maintain Genomic Stability

Abstract

The dual-specificity phosphatase 3 (DUSP3), an atypical protein tyrosine phosphatase (PTP), regulates cell cycle checkpoints and DNA repair pathways under conditions of genotoxic stress. DUSP3 interacts with the nucleophosmin protein (NPM) in the cell nucleus after UV-radiation, implying a potential role for this interaction in mechanisms of genomic stability. Here, we show a high-affinity binding between DUSP3-NPM and NPM tyrosine phosphorylation after UV stress, which is increased in DUSP3 knockdown cells. Specific antibodies designed to the four phosphorylated NPM's tyrosines revealed that DUSP3 dephosphorylates Y29, Y67, and Y271 after UV-radiation. DUSP3 knockdown causes early nucleolus exit of NPM and ARF proteins allowing them to disrupt the HDM2-p53 interaction in the nucleoplasm after UV-stress. The anticipated p53 release from proteasome degradation increased p53-Ser15 phosphorylation, prolonged p53 half-life, and enhanced p53 transcriptional activity. The regular dephosphorylation of NPM's tyrosines by DUSP3 balances the p53 functioning and favors the repair of UV-promoted DNA lesions needed for the maintenance of genomic stability.

Keywords: DUSP3/VHR; NPM translocation; genomic stability; nucleophosmin (NPM); p53; tyrosine dephosphorylation.

FIGURE 1

Dual specificity phosphatase 3 (DUSP3) binds with high affinity to nucleophosmin (NPM) in vitro and affects its nuclear localization and tyrosine phosphorylation. (A) The bimolecular interaction of purified recombinant proteins was performed by Surface Plasmon Resonance (SPR) and the obtained KD^# is shown in the graph. DUSP3 interacts with higher affinity to NPM full length compared to truncated NPM_9–122, while ERK1 was used as classic control of DUSP3 substrate. (B) MRC-5 and XPA cell lines exposed to UVC radiation show NPM expression was not affected at all indicated times, regardless of the DUSP3 presence. (C) DUSP3 colocalizes with NPM before and after exposure to 18 J/m² UVC. This colocalization occurs even after nucleoplasmic translocation of the NPM (complete kinetics is shown in Supplementary Figure 1). There is no signal of DUSP3 staining in the shDUSP3 cells by immunofluorescence. Representative images are only qualitative and white scale bars are 5 μM length at 63 × magnification. (D) Immunoprecipitation assays using NPM antibody and immunoblotted with anti phospho-Tyr antibody show an increase in the levels of phospho-Tyr-NPM in both MRC-5 and XPA shDUSP3 cells 1 h after UVC exposure. Immunoblottings are representative of experiments performed in triplicates, and the quantification is shown below each band as average ± standard deviation. ^# Note KD = ka/kd (M), where KD = equilibrium dissociation constant, ka = association rate constant, and kd = dissociation rate constant.

FIGURE 2

Conserved tyrosine residues of NPM are dephosphorylated by DUSP3. (A) The complete amino acid sequence of NPM protein shows the four tyrosine residues (in red: Y17, Y29, Y67, and Y271) inserted in decapeptide sequences (in blue) within the primary structure of NPM. The four tyrosine are localized in two regions of the primary and secondary structures of NPM (in green: one at very end N-terminal and the other at very end C-terminal), interspaced by a structurally unknown region (in gray), which have been crystallized and better studied. The four Tyr-containing decapeptide sequences were used as template to synthesize four decaphosphopeptides phosphorylated on each specific Tyr residue, which were used to immunization of rabbits and generation of phospho-specific antibodies. The oligomeric NPM structure (pentameric) proposed from functional studies (in metallic green; PDB 5EHD) was used as model to highlight the spatial position of the four tyrosines on NPM 3D structure (in green; PDB, 5EHD, and 2VXD). (B) Interspecies multiple alignment of the regions containing the four NPM tyrosines show these residues conservation throughout evolution. (C) Cellular lysates from MRC-5 and XPA cell lines (NS or shDUSP3) exposed to 18 J/m² UVC radiation were immunoblotted using the antibodies against the four phospho-tyrosine residues of NPM. DUSP3 knockdown increased the phosphorylation of the 29, 67, and 271 tyrosines. (D) The in vitro dephosphorylation assays confirmed that DUSP3 can specifically dephosphorylate three tyrosine residues (29, 67, and 271), since their phosphorylation levels are decreased by the addition of exogenous DUSP3 to the lysates but are restored in the presence of Na₃VO₄. The immunoblotting images are representative of three independent experiments and the quantification is shown around the bands as mean (red) ± standard deviation (black).

FIGURE 3

The NPM translocation and oligomerization, the global RNA transcription, and the nuclear and nucleolar morphology are all affected by DUSP3 knockdown. (A–D) To verify the location of Tyr-phosphorylated NPM after 18 J/m² UVC exposure, confocal microscopy was performed in MRC5 and XPA cell lines (NS or shDUSP3) and compared with the staining of total NPM. The phosphorylation of Y29, Y67, and Y271 residues of NPM is observed in the nucleolus at basal conditions colocalizing with total NPM and remain phosphorylated after its translocation to the nucleoplasm (the complete kinetics is in the Supplementary Figure 3). In shDUSP3 cells p- Y29-, p- Y67-, and p-Y271-NPM reached the nucleoplasm 3 h after UVC, while in non-silencing (NS) cells they remain in nucleolus. Representative images are only qualitative and white scale bars are 5 μM length at 63 × magnification. (E) The NPM translocation was measured by ImageJ software as percentage of NPM present in nucleolus of at least 100 individual nuclei. shDUSP3 cells show an early nucleolus-nucleoplasm translocation of NPM. The same collected confocal images were used to count the number of nucleoli per nucleus (F) and the nuclear area (G). In MRC-5 cells, the DUSP3 knockdown implied in greater number of nucleoli and larger nuclei compared to XPA cells. (H) General assay for RNA transcription using ethynyl uridine (EU) shows that MRC-5 shDUSP3 cells present greater transcriptional activity, size, and number of nucleoli per nucleus (Supplementary Figure 4). (I) Immunoblotting for NPM performed in gradient semi-native gels of total lysates from MRC-5 and XPA cells submitted or not to UV radiation. Representative blottings from three independent assays show greater levels of monomeric NPM under DUSP3 knockdown and after UVC exposure. Note: “–” indicates DUSP3 knockdown (shDUSP3 cells) and “+” indicates DUSP3 presence (NS cells). Anova: ****: p < 0.0001.

FIGURE 4

DUSP3 knockdown increases p53(Ser15) phosphorylation and p53 activity. MRC-5 and XPA cells were exposure to UVC radiation (18 J/m²) and submitted to confocal microscopy as indicated. (A) The p53 phosphorylation on Ser15 (p-p53) accompanies the nucleolus-nucleoplasm translocation of NPM: it peaks at 3 h after UVC exposure in shDUSP3 and it occurs at 6 h in NS cells. (B) The spatiotemporal colocalization of ARF and p-p53 occurs the same way as NPM and is also earlier in DUSP3 silenced cells. (C) Immunofluorescence images show p-p53 levels in DUSP3 knockdown cells before exposure to UV and peaking 3 h after stress, but only 6 h after in NS cells (complete kinetics are shown in Supplementary Figure 6). Representative images are only qualitative and white scale bars are 5 μM length at 63 × magnification. (D) The levels of p53(Ser15) phosphorylated are elevated in both DUSP3 knockdown cells compared to NS cells from 0 to 24 h after UV radiation. (E) The differences in p53(Ser15) phosphorylation caused by DUSP3 silencing were quantified and plotted from three independent experiments. (F) Firefly Luciferase gene reporter with a promoter containing p53 responsive element was transfected in cells and used to measure transcriptional activity of p53. Both MRC5 and XPA shDUSP3 cells exhibit greater p53 activity compared to NS cells, which is still higher in XPA cells. After exposure to UV radiation, p53 activity is increased in the control group and much more evidenced in DUSP3 silenced cells.

FIGURE 5

DUSP3 silencing relocates ARF and HDM2 in the nucleus and enhances p53 stability. MRC-5 and XPA cells were exposure to 18 J/m²UVC radiation and confocal microscopy was performed as indicated. (A) NPM colocalizes with HDM2 earlier (0 to 3 h) in shDUSP3 cells compared to NS (only in 6 h) after UV exposure. (B) Likewise, ARF colocalizes strongly with HDM2 as early as 3 h after exposure to UV in both DUSP3 silenced cells, while this colocalization is seen only in 6 h in NS cells. (C) The reduction of p53-HDM2 colocalization in the nucleoplasm is observed earlier in the shDUSP3 cells compared with NS controls. Representative images are only qualitative and white scale bars are 5 μM length at 63 × magnification. (D) Percentage of ARF protein present in nucleolus of at least 100 nuclei per condition measured using ImageJ and showing that ARF translocated earlier from nucleolus-to-nucleoplasm in DUSP3 knockdown cells. (E) The presence of HDM2 in nucleoli was measured by its colocalization with NAT10, a specific constitutive nucleolar marker, and it was expressed as percentage of nucleoli containing HDM2. Both MRC-5 and XPA shDUSP3 cells show high presence of HDM2 at nucleolus regardless UV radiation stress. Besides that, HDM2 is more retained in the nucleoli of XPA cells in basal conditions. (F) p53, NPM, and HDM2 proteins stability was verified in non-stressed MRC-5 cells by CHX treatments and followed by immunoblotting assays. Bands were quantified assuming the control condition (C = without CHX) of each cell line as 100% and normalized accordingly by the Actin loading control. An apparent increase in p53 protein stability is observed in MRC-5 shDUSP3 cells compared to NS cells. Blots are representative of three independent experiments. Anova: ***p < 0.001; ****p < 0.0001.

FIGURE 6

Schematic model on the contribution of DUSP3-NPM axis to p53 actions in the maintenance of genomic stability. Under conditions of cellular homeostasis, the NPM present in the nucleoli is constantly interacting with and being dephosphorylated by DUSP3 on residues Y29, Y67, and Y271. In the absence of DUSP3, these three residues remain phosphorylated and favor the dissociation equilibrium of NPM homo-oligomerization and/or its association with ARF, therefore promoting an early nucleoplasmic translocation of monomeric NPM and ARF. Once in the nucleoplasm, these two proteins can induce the dissociation of the HDM2-p53 interaction through the individual binding to one or the other protein. This mitigates the process of p53 degradation (via proteasome), increasing its half-life and, therefore, allowing its phosphorylation in Ser15 (through kinases of the PIKK family) that subsequently increase its transcriptional activity. Therefore, as previously reported in a DUSP3 deficiency scenario, the greater p53 activation modulates the downstream pathway to regulate cellular responses to genotoxic stress, causing cell cycle arrest associated with the absence or insufficient DNA repair, followed by senescence and reduced cell proliferation/survival (Torres et al., 2017; Monteiro et al., 2019; Russo et al., 2020).