Plk1 overexpression induces chromosomal instability and suppresses tumor development

Abstract

Polo-like kinase 1 (Plk1) is overexpressed in a wide spectrum of human tumors, being frequently considered as an oncogene and an attractive cancer target. However, its contribution to tumor development is unclear. Using a new inducible knock-in mouse model we report here that Plk1 overexpression results in abnormal chromosome segregation and cytokinesis, generating polyploid cells with reduced proliferative potential. Mechanistically, these cytokinesis defects correlate with defective loading of Cep55 and ESCRT complexes to the abscission bridge, in a Plk1 kinase-dependent manner. In vivo, Plk1 overexpression prevents the development of Kras-induced and Her2-induced mammary gland tumors, in the presence of increased rates of chromosome instability. In patients, Plk1 overexpression correlates with improved survival in specific breast cancer subtypes. Therefore, despite the therapeutic benefits of inhibiting Plk1 due to its essential role in tumor cell cycles, Plk1 overexpression has tumor-suppressive properties by perturbing mitotic progression and cytokinesis.

Fig. 1

Generation of Plk1-inducible mice. a Schematic representation of the alleles used in this work. A cassette containing the human FLAG-Plk1 cDNA downstream of the tetO sequences is inserted in the endogenous ColA1 locus after homologous recombination in KH2 ES cells. This allele [ColA1(Plk1) or (Plk1) in short] is later combined with the Rosa26-rtTA allele expressing the tetracycline transactivator. b (+/Plk1);(+/rtTA) ES cells were treated with Dox and FLAG and Plk1 signal was detected using specific antibodies at the indicated time points. Vinculin was used as a loading control. c Immunofluorescence of (+/Plk1);(+/rtTA) ES cells treated with Dox for 12 h. FLAG (green) is concentrated at the spindle poles with some signal in the spindle microtubules and additional diffuse signal as expected for Plk1. α-tubulin is in red and DAPI in blue. Scale bar 5 μm. d Immunodetection of Flag-Plk1 in the indicated tissues from (+/+);(rtTA/rtTA); (+/Plk1)(+/rtTA) and (+/Plk1)(rtTA/rtTA) mice treated with Dox for 8 weeks. Scale bar 100 μm. e Tumor-free survival of (+/+);(+/rtTA) and (+/Plk1)(+/rtTA) mice fed with Dox since birth during 85 weeks. (+/+)(+/rtTA), 19 mice; (+/Plk1)(+/rtTA), 24 mice. p = 0.2836; Log-rank (Mantel–Cox) test. f Sections of Dox-treated (+/+);(+/rtTA) and (+/Plk1)(+/rtTA) mice after staining with hematoxylin and eosin (H&E) or immunodetection of Plk1 (right panel). Cells with abnormally large nuclei are indicated with arrows. Black scale bar: 100 μm, blue scale bar: 50 μm

Fig. 2

Plk1 overexpression results in proliferative defects in cultured MEFs. a Expression of Plk1 in MEFs with the indicated genotypes, after 24 h treatment with doxycycline. Cell lysates were immunoblotted with anti-Plk1, anti-Flag, anti-pH3, and anti-vinculin as a loading control. b Quantification of confluence in cultures with the indicated genotypes in the absence (–Dox) or presence (+Dox) of doxycycline for 7 days. ****p < 0.0001; two-way ANOVA. c Percentage of EdU-positive cells in (+/Plk1);(rtTA/rtTA) cultures untreated or treated with Dox for 3 days. Cells were exposed to EdU for 1 h before the analysis. ***p < 0.001; Student’s t-test. d Percentage of mitotic cells in (+/Plk1);(rtTA/rtTA) cultures untreated or treated with Dox for 2 days as detected by phospho-Ser10 Histone H3 immunofluorescence. **p < 0.05; one-way ANOVA. e DNA content in (+/Plk1);(rtTA/rtTA) cells untreated or treated with Dox for 1 or 2 days. The percentage of 2N, 4N, or 8N cells is indicated in the histograms. f Metaphase spreads of (+/Plk1);(rtTA/rtTA) cells untreated or treated with Dox for the indicated time. The number of chromosomes per cell is shown in the plot (n = 25 (–Dox), n = 41 (24 h), n = 25 (48 h), n = 16 (72 h) cells per condition). ns not significant; *p < 0.05, ***p < 0.001; one-way ANOVA. g Focus formation assays of (+/Plk1);(rtTA/rtTA) MEFs transfected with oncogenic HrasV12 or control GFP-expressing vectors in the absence (–Dox) or presence (+Dox) of doxycycline. The number of foci is indicated in the histogram. ****p < 0.0001 (n = 3 replicates); one-way ANOVA. h Soft agar colony formation of immortal Hras transformed MEFs. Colonies are grown in the absence of Dox for 20 days (gray upper panels), or in the presence of Dox since plating (purple mid panels). Additionally, colonies grown in the absence of Dox were supplemented with Dox at day 8 (lower panels). Colony diameter (in microns) is quantified in the right histogram. Each dot represents one colony (over 50 colonies are quantified in each set). *p < 0.01; ****p < 0.0001, one-way ANOVA Bonferroni test

Fig. 3

Mitotic defects in Plk1-overexpressing MEFs. a Time-lapse microscopy of Plk1 MEFs untreated (upper panel) and after 8 h on Dox (lower panels), indicating mitotic cells; H2B-GFP (green), classified based on the three major phenotypes resulting in a tetraploid progeny in the case of Plk1 overexpression. Scale bar 20 μm. b Percentage of occurrence of each major phenotype in –Dox (n = 141) and +Dox (n = 164) MEFs. c Percentage of mitotic errors per MEF (–Dox: 141 cells; +Dox: 164 cells); points represent individual MEF line. d Duration of mitosis in the MEF cultures (–Dox: 141 cells; +Dox: 164 cells). In c, d, ****p < 0.0001, Student’s t-test. e Immunofluorescence against α-tubulin and pericentrin in primary MEFs after 24 h on Dox followed by quantification of each mitotic phase and percentage of mitotic aberrancies (–Dox: n = 149 mitotic cells; +Dox: n = 166 mitotic cells). ****p < 0.0001; two-way ANOVA. Scale bar, 10 μm. f Immunofluorescence against α-tubulin and γ-tubulin in (+/Plk1)(rtTA/rtTA) MEFs untreated (–Dox) or treated with Dox for 48 h (+Dox). Binucleated cells (arrowheads) from the total population (>100 cells from at least four random microscope fields are quantified in each replicate). ****p < 0.0001 (n = 2 replicates); one-way ANOVA. Scale bar, 10 μm

Fig. 4

Plk1 overexpression induces defects in centrosome and chromosome segregation dynamics. a Immunofluorescence against p-Plk1-T210 and pericentrin in MEFs untreated or after 24 h with doxycycline (+Dox). The histogram shows the quantification of mean fluorescence intensity (MFI) of p-Plk1-T210 staining at the centrosomes during late G2 phase (separated; –Dox: n = 54 centrosomes; +Dox: n = 34 centrosomes) or earlier interphase (joint; –Dox: n = 88 centrosomes; +Dox: n = 58 centrosomes; n = 2 replicates). Scale bar, 10 μm. b Immunofluorescence against p-Plk1-T210 and pericentrin in MEFs untreated or after 24 h on Dox. The histogram shows the quantification of mean fluorescence intensity (MFI) of p-Plk1-T210 staining at each mitotic phase [Prometaphase (PM): –Dox n = 22 cells; +Dox: n = 26 cells; Metaphase (M): –Dox n = 17 cells; +Dox: n = 21 cells; Anaphase (A): –Dox n = 15 cells; +Dox: n = 12 cells; Cytokinesis (C): –Dox n = 13 cells; +Dox: n = 15 cells; n = 3 replicates]. c Immunofluorescence against p-Plk1-T210 in primary MEFs untreated or after 24 h on Dox. The histogram shows the quantification of mean fluorescence intensity (MFI) of p-Plk1-T210 staining at individual kinetochores of cells in prometaphase (–Dox: n = 51 kinetochores; +Dox: n = 51 kinetochores; n = 2 replicates) **p < 0.01; Mann–Whitney test. Scale bar, 10 μm. d Immunofluorescence against Sgo1 in primary MEFs untreated or after 24 and 72 h on Dox. The first histogram shows the quantification of mean fluorescence intensity (MFI) of Sgo1 staining in the entire cell at prometaphase (–Dox n = 33 cells; +Dox 24 h: n = 35 cells; +Dox 72 h: n = 52 cells; n = 3 replicates). The second histogram shows the percentage of prometaphase cells with diffused Sgo1 staining (n = 3 replicates). Scale bar, 10 μm. In a, b, d, n.s. not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; one-way ANOVA. e Chromosome spreads (DAPI stained) from (+/Plk1);(rtTA/rtTA) MEFs untreated or treated with Dox at the indicated times. Chromosome cohesion was classified in three different status: “parallel arms” as readout of full chromatid cohesion (dark blue box), “separated arms” as readout of chromatid arm separation (mid blue), and “separated sisters” as readout of fully separated chromatids (light blue). (–Dox, n = 24; +Dox 24 h, n = 40; +Dox 48 h, n = 25; +Dox 72 h, n = 22 cells). Scale bars, 20 μm

Fig. 5

Overexpression of Plk1 impairs Cep55 and ESCRT loading into the cytokinesis midbody. a (+/Plk1);(rtTA/rtTA) MEFs were untreated (–Dox) or treated for 24 h with doxycycline (+Dox), fixed and stained for α-tubulin (red) and DAPI (DNA, green). Cells in cytokinesis (n > 100 per condition; n = 3 replicates) were evaluated for aberrant cytokinesis, considering the midbody formation and shape, and correct distribution of DNA into the two daughter cells. Scale bars, 10 μm. b The length of the cytokinesis bridge was evaluated by measuring the distance in between the two daughter nuclei in MEFs untreated or after 24 h of Dox treatment (–Dox, n = 17 cells; +Dox, 23 cells). Scale bars, 10 μm. In a, b *, p < 0.05; Student’s t-test. c MEFs expressing a CEP55-EGFP fusion (green) were treated for 24 h with Dox (+Dox), in the absence or presence of 1 μM of the Plk1 inhibitor BI2536 for 1 h at the end of the Dox time (Dox + BI), or left untreated as control (–Dox). α-tubulin is in red. Data represent the percentage of cells with positive CEP55-EGFP signal at the midbody (–Dox, n = 134; +Dox, n = 94; Dox + BI, n = 47 cells; n = 3 replicates (–Dox, +Dox) or 2 replicates (Dox+BI)). Scale bar, 5 μm. d MEFs expressing a Tsg101-mCherry fusion (red) were treated for 24 h with Dox (+Dox), in the absence or presence of 1 μM of BI2536 for 1 h at the end of the Dox time (Dox + BI), or left untreated as control (–Dox). α-tubulin is in green. Data represent Tsg101-mCherry mean intensity at the midbody (–Dox, n = 81; +Dox, n = 60; Dox + BI, n = 37 cells; n = 2 replicates). Scale bars, 5 μm. e MEFs were treated for 24 h with Dox (+Dox), in the absence or presence of 1 μM of BI2536 for 1 h at the end of the Dox time (Dox + BI), or left untreated (–Dox). Cells were stained for α-tubulin (red) and DAPI (DNA, green) and binucleation index was quantified from more than 600 cells in each sample (n = 5 replicates). Scale bars, 50 μm. In c–e, n.s. not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; one-way ANOVA

Fig. 6

Plk1 overexpression reduces tumor development induced by Kras or Her2 oncogenes. a Percentage of tumor-free mice after doxycycline administration (control, n = 33; Plk1/MMTV-rtTA n = 35; Kras^G12D n = 87; Kras^G12D/Plk1 n = 40 mice). b Percentage of tumor-free survival after doxycycline administration (control, n = 33; Her2, n = 51; Her2/Plk1, n = 36 mice). In a, b, ****p < 0.0001, Mantel–Cox test. c Number of tumors per animal in the indicated genotypes; **p < 0.01; ****p < 0.0001; Mann–Whitney test. d Interphase-Fluorescent in situ hybridization (I-FISH) on paraffin sections of mammary tumors using two centromeric probes against chromosomes 16 and 17. e Quantification of the frequency of aneuploidy (left) in samples from the indicated genotypes (control, n = 4; Kras, n = 4; Kras/Plk1, n = 3; Her2, n = 4; Her2/Plk1, n = 4 mice). *p < 0.05; ***p < 0.001; one-way ANOVA. f Percentage of cells with 2, 3, or 4 or more chromosomes in the indicated genotypes. g Representative micrographs of Her2 and Her2/Plk1 tumor cells in vitro (H2B-GFP green). Top: mitotic cell with a lagging chromosome. Bottom: cytokinesis failure resulting in binucleation. h Percentage of cells in Her2 tumors (H) and Her2/Plk1 (HP) with the indicated mitotic errors

Fig. 7

Mitotic aberrations in cultured Plk1-overexpressing mammary organoids. a Time-lapse microscopy of Plk1/MMTV-rtTA mammary organoids expressing H2B-GFP either untreated (upper panel) or after 24 h on Dox (lower panel). Yellow circles indicate mitotic cells; H2B-GFP (green). Scale bar, 18 μm. b Duration of mitosis in the organoid cultures (–Dox, 57 cells; +Dox, 43 cells). c Percentage of mitotic errors per mouse (–Dox, 57 cells from 5 mice; +Dox, 47 cells from 5 mice). In b, c, ****p < 0.0001, Student’s t-test. d Classification of mitotic phenotypes in organoid cells overexpressing Plk1 after 36 h on Dox

Fig. 8

Plk1 expression in human breast cancers. a Plk1 expression relative to TBP (log2) in genome-doubled (GD) and non-genome-doubled (nGD) breast cancers from the TCGA. p = 4.62e-09, t-test. b Plk1 expression in breast cancers with and without TP53 mutations. The association between GD and Plk1 expression is significant in TP53 wt tumors. p = 3.37e-06, t-test. c Survival analysis of breast cancer (BRCA) patients with low Plk1 expression (bottom quartile) and those with higher Plk1 expression (remaining quartile) p = 0.0099; hazard ratio (HR), 0.59; 95% confidence interval (CI), 0.40–0.89