Blocking PLK1 improves radiation therapy by stopping cancer cells during cell division

Accumulating evidence in the research field of radiation biology has demonstrated that the radiosensitivity of cells depends on the phases of the cell cycle; cells are relatively radioresistant in the late S-phase, but are radiosensitive in the mitotic phase (M-phase). This fundamental principle prompted us to hypothesize that the small molecule inhibitor of PLK1, TAK-960, which has been reported to cause the mitotic arrest of cancer cells19, may enhance the cytotoxic effects of radiation. In order to test this possibility, we first determined the optimal conditions of the TAK-960 treatment for mitotic arrest without apparent cytotoxic effects in vitro. We treated human cervical cancer cells, HeLa, with various concentrations of TAK-960 and performed cell cycle analyses using flow cytometry with propidium iodide. Among the various concentrations of TAK-960 examined, 16 nM resulted in the most pronounced cell cycle arrest at the G₂/M phase (Fig. 1a,b), but had a significant cytotoxic effect in the clonogenic survival assay (Supplementary Fig. S1 online). On the other hand, lower concentrations of TAK-960, such as 8 nM, led to G₂/M cell cycle arrest without significant cytotoxicity. We then treated HeLa cells with 8 nM TAK-960 for various periods to optimize the duration of the treatment (Fig. 1c,d). Flow cytometric analyses revealed that the proportion of cells in the G₂/M phase was maximized by the TAK-960 treatment for 12 hours.

In order to monitor the influence of the TAK-960 treatment on the cell cycle status in real-time, we employed HeLa-S FUCCI cells21, which preferentially emit orange (monomeric Kusabira-Orange2: mKO2) and green fluorescence (monomeric Azami-Green1: mAG1) in the G₁-phase and S/G₂/M-phase, respectively. Time-lapse imaging confirmed that cells alternatively exhibited orange and green fluorescence in the absence of the TAK-960 treatment, and the number of round-shaped cells with green fluorescence, which represent mitotic cells, slightly increased, corresponding to proper cell cycle progression (Fig. 1e,f; Supplementary Video S1 online). Upon exposure to 8 nM TAK-960, the number of mitotic cells markedly increased and peaked 12–18 hours after the initiation of the TAK-960 treatment, which was consistent with the results of the flow cytometric analyses (compareFig. 1c,d with 1e–g; Supplementary Video S2 online). Taken together, these results indicated that the optimal condition for the most pronounced cell cycle arrest at the mitotic phase was 8 nM TAK-960 for 12 hours.

We performed clonogenic cell survival assays in vitro to determine whether the optimized TAK-960 treatment actually enhanced the cytotoxic effects of X-irradiation. Not only HeLa cells, but also human non-small cell lung cancer cells, H1299, and human colon cancer cells, HCT116, were subjected to the same TAK-960 treatment (8 nM TAK for 12 hours) before X-irradiation because we confirmed that the TAK-960 treatment clearly induced the mitotic arrest of H1299 and HCT116 cells (Fig. 2a–c, Supplementary Fig. S2 online). Without the radiation treatment, the number of surviving colonies was similar between the control (0 nM) and TAK-960 treatment (8 nM) groups, which is consistent with the observations in Supplementary Fig. S1 online. On the other hand, the TAK-960 treatment significantly decreased clonogenic survival after X-irradiation. The dose of radiation required to reduce the number of surviving colonies by 90% (D₁₀ value) was significantly decreased by the TAK-960 treatment from 4.2 ± 0.2 to 3.4 ± 0.3 Gy in HeLa, from 7.5 ± 0.6 Gy to 5.0 ± 0.6 Gy in H1299, and from 3.7 ± 0.3 Gy to 3.0 ± 0.2 Gy in HCT116 cells (Table 1). The radiation enhancement ratio of the TAK-960 treatment based on the D₁₀ value in HeLa cells was 1.24 ± 0.4, which was similar to that of another PLK1 inhibitor, BI2536 (1.26 ± 0.3; Supplementary Fig. S3 online).

Since TAK-960 was removed from the culture medium just before irradiation in the clonogenic cell survival assays, these results indicated that a cellular response to the TAK-960 pretreatment, possibly mitotic arrest, was required for its radiosensitizing effects. Consistent with this result, the concomitant administration of TAK-960 and X-irradiation, in which cells were exposed to the drug just before irradiation and did not have sufficient time to induce any cellular responses including cell cycle arrest, did not exhibit any radiosensitization (Fig. 2d).

In order to directly examine whether the radiosensitizing effects of the TAK-960 treatment were dependent on its PLK1-inhibiting activity and resultant mitotic arrest, we employed mutant forms of PLK1, which were previously reported to be less sensitive to PLK1 inhibitors, and thus, promoted M-phase progression even in the presence of the inhibitor. We constructed expression vectors for the wild-type (wt) and three mutant forms of PLK1: a phosphorylation-mimicking mutant, PLK1-T210D22,23, a mutant less sensitive to PLK1 inhibitors, PLK1-R136G24, and a mutant with both mutations, PLK1-R136G&T210D (Fig. 3a). Flow cytometry-based cell cycle analyses using propidium iodide revealed that the forced expression of these mutants had no effect on the cell cycle distribution of HeLa cells in the absence of the TAK-960 treatment (Fig. 3b,c). In contrast, the TAK-960-mediated increase in the proportion of cells in the G₂/M phase was significantly suppressed by the overexpression of PLK1-R136G&T210D, but not that of PLK1-T210D or PLK1-R136G in our experimental setting (Fig. 3b,c). Based on these results, we employed the expression vector for PLK1-R136G&T210D and directly tested the importance of mitotic arrest in the radiosensitizing effects of the PLK1 inhibitor, TAK-960 (Fig. 3d–f). Clonogenic cell survival assays demonstrated that the radiosensitizing effects of TAK-960 were almost completely abolished by the introduction of PLK1-R136G&T210D (compareFig. 3f with Fig. 3d,e).

We employed paclitaxel (PTX), which is known to induce cell cycle arrest at the mitotic phase as an inhibitor of microtubular depolymerization25 to further test the importance of mitotic arrest in TAK-960-mediated radiosensitization. The clonogenic cell survival assay demonstrated that the relative radiosensitizing effects of TAK-960 were significantly decreased after the cell cycle was synchronized at the mitotic phase by paclitaxel (Fig. 3g,h). These results clearly showed that TAK-960 enhanced the cytotoxic effects of radiation through its own PLK1-inhibiting activity and consequent cell cycle arrest at the radiosensitive mitotic phase.

Prior to evaluating the radiosensitizing effects of TAK-960 in vivo, we determined the optimal TAK-960 treatment to maximize the amount of cells in mitotic arrest without influencing xenografted tumor growth in vivo. We first analyzed changes in the number of mitotic cells after the administration of TAK-960 by immunohistochemistry. Mice bearing a subcutaneous HeLa tumor xenograft were administered with various doses of TAK-960 and subjected to immunohistochemical analyses using an antibody against a marker of mitotic cells, phosphorylated histone H3 (Ser10) (pHH3(S10)). The maximal number of pHH3(S10)-positive cells was observed 24 hours after the administration of 10 or 20 mg/kg TAK-960 (Fig. 4a,b). In order to confirm this result, we conducted an optical in vivo imaging experiment using the HeLa-S FUCCI tumor xenograft model. The use of green fluorescence as an indicator of mitotic cells was justified by immunocytochemical data in which the TAK-960 treatment significantly increased the proportion of cells that were round in shape, emitted green fluorescence, and were stained with the marker of mitotic cells, pHH3(S10) (Supplementary Fig. S4 online; Supplementary Table S1 online). Subcutaneous HeLa-S FUCCI tumor xenografts were injected with the indicated dose of TAK-960, and green and orange fluorescence from mAG1 and mKO2, respectively, was detected by the IVIS-SPECTRUM imaging system in real-time (Fig. 4c). In order to exclude the possibility that the intensity of fluorescence was influenced by tumor growth (increases in tumor volume over time), we calculated the ratio of the intensity of green fluorescence (mAG1) to that of red fluorescence (mKO2) as an index of mitotic cells (Fig. 4d). The mitotic index was significantly increased and peaked 24–36 hours after the administration of 10 or 20 mg/kg TAK-960, which was consistent with the immunohistochemical data for pHH3(S10) (Fig. 4d). We then assessed the influence of the various doses of the TAK-960 treatment on the growth of HeLa tumor xenografts. The tumor growth delay assay revealed that 10 mg/kg was the maximum dose that did not influence tumor growth and did not reduce the body weights of mice (Supplementary Fig. S5 online; Supplementary Table S2 online). Collectively, the optimal conditions that caused significant cell cycle arrest at the M-phase with minimum effects on tumor growth in vivo for subsequent in vivo experiments were determined to be 24 hours after the administration of 10 mg/kg TAK-960.

In order to examine whether the TAK-960 treatment in the optimized treatment regimen actually enhanced the therapeutic effects of radiation in vivo, we performed tumor growth delay assays using the subcutaneous HeLa tumor xenograft model (Fig. 5a; Table 2). Tumor xenografts were locally irradiated at a dose of 0 or 10 Gy 24 hours after the oral administration of 0 or 10 mg/kg TAK-960. We intentionally selected a relatively low dose of TAK-960 (10 mg/kg), which was confirmed to be sufficient for mitotic arrest with minimal effects on the tumor growth rate (Fig. 4; Supplementary Fig. S5 online); therefore, tumor growth after the TAK-960 treatment alone was not significantly delayed more that with the sham treatment. Tumor growth doubling time (TGDT) after the TAK-960 treatment alone was 8.3 ± 1.6 days and was not significantly longer than that after the sham treatment (9.1 ± 1.7 days; P = 0.33). Conversely, the TAK-960 treatment significantly suppressed tumor growth when combined with IR. TGDT after the combination of radiation therapy with the TAK-960 treatment was 27.4 ± 10.0 days, and was approximately 2.0-fold longer than that after radiation therapy (14.0 ± 5.1 days; P < 0.05). The same radiosensitizing effects of TAK-960 were confirmed in another tumor xenograft model with human non-small cell lung cancer cells, H1299 (Fig. 5c; Table 2). This combination did not cause any obvious side effects (Fig. 5b,d). Taken all together, these results demonstrated that the TAK-960 treatment exhibited radiosensitizing effects in the optimized schedule. In order to further examine the importance of mitotic arrest in the radiosensitizing effects of TAK-960 in vivo, we intentionally irradiated subcutaneous HeLa tumor xenografts just after the administration of TAK-960, and TAK-960 was not allowed to induce mitotic arrest. A delay in tumor growth was not observed even after the combination of TAK-960 with radiation (Fig. 5e,f; Table 3), which supported our conclusion that the radiosensitizing effects of TAK-960 in vivo may be attributed to mitotic arrest.

The human cervical epithelial adenocarcinoma cell line (HeLa) and human lung carcinoma cell line (H1299) were purchased from the American Type Culture Collection (Manassas, VA, USA). The human colon cancer cell line (HCT116) and HeLa-S FUCCI cell line were kindly provided by Prof. Bert Vogelstein (The Johns Hopkins University, Baltimore, MD) and RIKEN BRC, respectively. Cells were cultured with 10% FBS-containing Dulbecco’s modified Eagle’s medium (DMEM) in a well-humidified incubator with 5% CO₂ and 95% air at 37 °C. TAK-960 was dissolved in DMSO and 0.5% (w/v) hydroxylpropyl methylcellulose solution for in vitro and in vivo experiments, respectively.

In order to construct the plasmid pcDNA3.1/PLK1, the coding sequence of the PLK1 gene was amplified by PCR from the HeLa cDNA library using the following primers: (5′-AAAAGCTTACCATGAGTGCTGCAGTGACTG-3′ and 5′-AGCGGCCGCTTAGGAGGCCTTGAGACGG-3′), and inserted between the HindIII-NotI sites of pcDNA3.1/myc-His C (Invitrogen Corp., Carlsbad, CA, USA). Regarding construction of the plasmids pcDNA3.1/PLK1-R136G, pcDNA3.1/PLK1-T210D, and pcDNADNA3.1-PLK1-R136G&T210D, PCR-based site-directed mutagenesis was performed using the following primer sets: (5′-TGCCGCCGGGGGTCTCTCCTGGAGCTGCAC-3′ and 5′-CAGGAGAGACCCCCGGCGGCAGAGCTCC-3′ for R136G, and 5′-AGGAAGAAGGACCTGTGTGGGACTCCTAAT-3′ and 5′-CCCACACAGGTCCTTCTTCCTCTCCCCGTC-3′ for T210D), and the resultant DNA fragments encoding PLK1-R136G, PLK1-T210D, and PLK1-R136G&T210D were inserted between the HindIII-BamHI sites of pcDNA3.1/PLK1.

HeLa cells were seeded in 6-cm dishes for flow cytometric analyses (2 × 10⁵ cells/dish) or in 6-well plates for western blotting (1 × 10⁵ cells/well). In the cell cycle analysis using flow cytometry, cells were fixed with ice-cold 70% ethanol, re-suspended in PBS containing both RNase A (1 mg/ml) and propidium iodide (1 mg/ml), incubated at 37 °C for 30 min, and subjected to flow cytometry (BD Bioscience, Franklin Lakes, NJ, USA), as described previously36. The results obtained were analyzed with DIVA software (BD Bioscience). Cell lysates prepared with CelLytic M cell lysis reagent (Sigma-Aldrich, St Louis, MO, USA) were subjected to Western blotting for the detection of PLK1 and β-actin with an anti-PLK1 mouse monoclonal antibody (1:1000; Abcam, Cambridge, UK) and anti-β-actin mouse monoclonal antibody (1:500; BioVision Research Products, Mountain View, CA, USA), respectively. In time-lapse imaging, HeLa-S FUCCI cells were seeded in glass-coated 3.5-cm dishes (5 × 10⁴ cells/dish) and cultured for 12 h. Just after the addition of the indicated concentrations of TAK-960, orange (monomeric Kusabira-Orange2: mKO2) and green fluorescence (monomeric Azami-Green1: mAG1) emitted from HeLa-S FUCCI cells was observed with FLUOVIEW FV10i (Olympus, Tokyo, Japan) every 20 min for 72 h.

Cells (100, 200, 700, and 2,000 HeLa cells/60-mm dish, 200, 300, 500, and 1,000 H1299 cells/60-mm dish, and 100, 200, 1,000, and 10,000 HCT116 cells/60-mm dish for 0, 2, 4, and 6 Gy, respectively) were precultured in medium containing TAK-960 for 0 (Fig. 2d) or 12 h (Figs 2a–c and 3d–f). These cells were then exposed to 0, 2, 4, or 6 Gy X-radiation (ACROBIO Co., Tokyo, Japan) just after replacing the culture medium to a TAK-960-free one. The cells were further cultured for 2 weeks. Surviving colonies were fixed with 70% ethanol and stained with Giemsa solution. Colonies consisting of more than 50 cells were counted as surviving colonies. The plating efficiency and surviving fraction were calculated as described previously37.

Cells were transiently transfected with pcDNA3.1/myc-His (as an empty vector (EV)), pcDNA3.1/PLK1, pcDNA3.1/PLK1-R136G, pcDNA3.1/PLK1-T210D, or pcDNA3.1/PLK1-R136G&T210D using the Polyfection transfection reagent (Qiagen Inc., Valencia, CA, USA) according to the manufacturer’s instructions. These cells were then cultured for 48 h and subjected to each in vitro experiment.

The suspension of HeLa cells (2 × 10⁶ in 100 μL PBS) or H1299 cells (3 × 10⁶ in 100 μL PBS) was subcutaneously inoculated into the right hind legs of 8-week-old nude mice (BALB/c nu/nu mice; SHIMIZU Laboratory Supplies Co., Ltd; Kyoto, Japan). The indicated dose of TAK-960 was orally administered to tumor-bearing mice. In the radiation treatment, tumor xenografts were locally irradiated with the indicated dose of ¹³⁷Cs γ-rays using a Gammacell 40 Exactor (MDS Nordion International Inc., Ottawa, Ontario, Canada).

HeLa tumor xenografts were surgically excised 24 h after the administration of TAK-960. The tumor xenografts were embedded in OCT compound and frozen at −80 °C. Frozen sections were rinsed quickly in PBS, fixed in 4% paraformaldehyde for 10 min, blocked with the serum-free protein block solution (Dako, Glostrup, Denmark) for 20 min, and permeabilized with 0.2% Triton X-100 in PBS for 10 min. The sections were then treated with an anti-pHH3(S10) mouse monoclonal antibody (1:1000; Abcam, Cambridge, MA, USA). pHH3(S10) was detected with Alexa Fluor 594 goat anti-mouse IgG (1:1000; Invitrogen Corp.). The reproducibility of each staining was confirmed in at least three independent tumors and representative results are shown.

HeLa-S FUCCI cells on glass coverslips were fixed with ice-cold acetone for 2 min and permeabilized with 0.2% Triton X-100 in PBS for 10 min. After being incubated for 20 min in the serum-free protein block solution, cells were treated with an anti-pHH3(S10)-antibody (1:1000). pHH3(S10) was detected with Alexa Fluor 594 donkey anti-mouse IgG (1:1000; Invitrogen Corp.). The reproducibility of each staining was confirmed in at least three independent tumors and representative results are shown.

Optical in vivo imaging was carried out with an IVIS-SPECTRUM (Caliper). Fluorescence was acquired with the filter sets EX/535 nm and EM/580 nm for mKO2 and with EX/465 nm and EM/520 nm for mAG1. During imaging, mice were anesthetized with 2.5% isoflurane gas in the oxygen flow (1.5 L per min). Signal intensities were quantified and analyzed using Living Image 2.50-Igor Pro 4.90 software (Caliper), as described previously38,39,40.

Tumor volumes were calculated as 0.5 × length × width², and compared with the initial value to calculate relative tumor volume.

The significance of differences was determined using the Student’s t-test. A P value < 0.05 was considered to be significant.

All animal experiments were approved by the Animal Research Committee of Kyoto University, and performed according to guidelines governing animal care in Japan.