Cancer Gene Therapy (2014), 1–9 & 2014 Nature America, Inc. All rights reserved 0929-1903/14 www.nature.com/cgt

ORIGINAL ARTICLE

Co-abrogation of Chk1 and Chk2 by potent oncolytic adenovirus potentiates the antitumor efficacy of cisplatin or irradiation F Ye1,2, Z Yang1, Y Liu1, D Gong1, T Ji1, J Wang1, B Xi1, J Zhou1, D Ma1 and Q Gao1 Mammalian checkpoint kinases 1 and 2 (Chk1 and Chk2) are essential kinases that are involved in cell cycle checkpoint control, and the abrogation of either has been proposed as one way to sensitize cancer cells to DNA-damaging agents. However, it remains unclear which kinase is the most therapeutically relevant target, and whether multiple kinases might need to be targeted to achieve the best efficacy because of their overlapping substrate spectra and redundant functions. To clarify this issue, we established asynchronous cell cycle arrest models to investigate the therapeutic outcomes of silencing Chk1 and Chk2 in the presence of irradiation or cisplatin. Our results showed that Chk1- and Chk2-targeting small interference RNAs (siRNAs) demonstrated synergistic effects when both siRNAs were used simultaneously. Interestingly, Chk1 and Chk2 co-expression occurred in B90% of neoplastic tissues examined and showed no difference in neoplastic versus non-neoplastic tissues. Therefore, the selective targeting of Chk1 and Chk2 by oncolytic adenovirus mutants was chosen to treat resistant tumor xenograft mice, and the maximum antitumoral efficacy was achieved with the combined co-abrogation of Chk1 and Chk2 in the presence of low-dose cisplatin. This work deepens our understanding of novel strategies that target checkpoint pathways and contributes to the development of novel, potent and safe checkpoint abrogators. Cancer Gene Therapy advance online publication, 23 May 2014; doi:10.1038/cgt.2014.20

INTRODUCTION DNA-damaging agents are among the most effective anticancer agents in clinical use and constitute the cornerstones of modern cancer treatment. However, their effectiveness largely depends on the function of surveillance systems, which are termed as cell cycle checkpoints.1 Upon genotoxic challenge, these checkpoints work to initiate cell cycle arrest at different phases of the cell cycle to allow sufficient time for DNA repair, thereby promoting cell survival and, as a result, constituting a major mechanism of widespread tumor resistance to chemo- or radiotherapy. Checkpoints that respond to DNA damage have been empirically defined at the G1, S and G2/M phases of the cell cycle.2 Unlike normal cells with a full arsenal of checkpoint responses, tumor cells that are defective in some checkpoints must rely on the remaining checkpoints to survive. Therefore, cell cycle checkpoint abrogation could be an attractive approach to circumvent cancer resistance because of the expected induction of mitotic catastrophe and apoptosis in tumor cells while largely sparing normal cells.3,4 Checkpoint kinases 1 and 2 (Chk1 and Chk2) are two structurally unrelated but functionally similar protein serine/threonine kinases that have emerged as the major mediators of cell cycle modulation in response to DNA-damaging agents. Both kinases are known to relay checkpoint signals from the upstream signal-transducing kinases ataxia telangiectasia mutated and ataxia telangiectasia and RAD3 related to proximal substrates.5 Although Chk1 is known to be activated by phosphorylation at Ser345 by ataxia telangiectasia and RAD3 related, the activation of Chk2 in checkpoint signaling is initiated by phosphorylation at Thr68 in an ataxia telangiectasia mutated-dependent manner.6,7 The inhibition of Chk1 and Chk2 by different approaches, for example, antisense oligodeoxynucleotides,

small interference RNA (siRNA) or small molecule chemicals, was found to abrogate the S and G2/M checkpoints and potentiate the therapeutic efficacy of DNA-damaging agents, and thus has been proposed as adjuvant therapeutics to circumvent tumor resistance to anticancer drugs.8–11 A nonselective inhibitor, UCN-01, has undergone early clinical evaluations either alone or in combination with established cytotoxic agents.12,13 A series of new inhibitors, possibly even more specific than UCN-01, are in preclinical development.14,15 Nevertheless, it remains be fully elucidated which type of inhibitor, that is, against Chk1, Chk2 or both, will be more effective for clinical cancer therapy. Theoretically, the complementary actions of Chk1 and Chk2 that are specifically activated depending on the type and extent of DNA damage and the overlapping substrates Cdc25A/C of Chk1 and Chk2 may provide supportive evidence that the simultaneous inhibition of both Chk1 and Chk2 are necessary to fully abrogate the checkpoint, and hence achieve the best efficacy of potentiation.7,16 To clarify this issue, we investigated the therapeutic outcome of silencing Chk1 and Chk2 both singly and simultaneously, alone or in combination with irradiation or cisplatin treatment. Chk1/Chk2 siRNA significantly knocked down the expression of the Chk1/ Chk2 proteins yet caused little effect on apoptosis and the cell cycle in the absence of DNA-damaging agents. In the presence of DNA-damaging agents, Chk1- or Chk2 siRNA-enhanced irradiationor cisplatin-induced apoptosis and further demonstrated synergistic effects when these two siRNAs were used simultaneously. As the co-expression pattern of Chk1 and Chk2 occurred in B90% of the neoplastic tissues examined and exhibited no difference in cancerous versus non-cancerous tissues, we constructed novel oncolytic adenovirus mutants preferentially targeting Chk1 and Chk2 in cancer rather than in normal cells and demonstrated

1 Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China and 2Department of Neurosurgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Correspondence: Professor Q Gao, Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 430030, China. E-mail: [email protected] Received 19 January 2014; revised 14 April 2014; accepted 14 April 2014

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

2 potent antitumoral efficacy in the presence of low-dose cisplatin in vivo. Again, simultaneous targeting of Chk1 and Chk2 produced synergistic effects and achieved the best efficacy. MATERIALS AND METHODS Cell lines, viruses and reagents The human cervical carcinoma cell lines HeLa and SiHa, human breast carcinoma cell line MCF-7, human endometrial carcinoma cell line HEC-1-B and human hepatic carcinoma cell line HepG2 were obtained from the American Type Culture Collection and grown in Dulbecco’s modified Eagle medium with 10% fetal calf serum (FCS). The human ovarian cancer cell line A2780, human lung cancer cell line A549, human acute promyelocytic leukemia cell line HL-60 and human chronic myelogenous leukemia cell line K562 were grown in RPMI-1640 with 10% FCS. A human umbilical vein/ vascular endothelium cell line was obtained from American Type Culture Collection and grown in F-12K medium with 10% FCS. Primary leukemia cells were separated from the bone marrow of three cases of newly diagnosed primary acute myeloid leukemia according to standard protocols and cultured in RPMI-1640 with 10% FCS. All Adv5 mutants used in the present study were constructed in-house from the HEK-293 cell line following protocols detailed elsewhere. Ad5/dE1A with a deletion of amino acids 121– 129 in E1A CR2 was previously constructed.16 M2 or M3 was expressed from Adv5/dE1A via the replacement of the 6.7K/gp19K open reading frame in the E3 region by a fragment of reverse Chk1 cDNA (bases 853-250) or Chk2 cDNA (bases 1686-858).17,18 The virus mutant was constructed by homologous recombination in HEK-293 cells.19 Cisplatin (DDP) was obtained from QiLu pharmaceutical company, Shandong, China. Dulbecco’s modified Eagle medium, RPMI-1640, FCS and the Lipofectamine with Plus transfection kit were purchased from Invitrogen (Carlsbad, CA, USA). The annexin V/ propidium iodide (PI) apoptosis detection kit was purchased from BD Biosciences (San Jose, CA, USA). A purified mouse anti-cytochrome c antibody was purchased from BD Pharmingen (San Jose, CA, USA) and a rabbit anti-voltage-dependent anion-selective channel 1 (VDAC1) antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Monoclonal anti-human Chk1 and Chk2 antibodies, cleaved caspase-3, and b-actin were purchased from Cell Signaling Technology (Danvers, MA, USA).

siRNA transfection and drug treatment The sequences of an siRNA fragment against human Chk1 (50 -aactgaagaagcagtcgcagt-30 ), Chk2 (50 -aatgtgtgaatgacaactact-30 ) and a mismatched fragment (Si-con, 50 -aacaagtgaagcagtcgcagt-30 ) were synthesized according to sequences published elsewhere.17,18 For the transfection procedure, siRNA or Si-con was transfected into tumor cells using the Oligofectamine reagent (Invitrogen) according to the manufacturer’s instructions. The DNA-damaging agent was added to the cells at 24 h after transfection.

Western blot analysis The preparation of protein samples and western blotting were performed as described previously.19

Cell cycle and apoptosis analysis Different cell groups were harvested with 0.25% trypsin, washed with phosphate-buffered saline (PBS) and separated into two portions. One portion was freshly prepared by centrifugation, and the cells were resuspended in 300 ml buffer and stained with annexin V and PI; the cells were sorted using a FACScan (BD Biosciences) and analyzed using Cell Fit software (San Jose, CA, USA). The other portion was fixed in 75% ice-cold ethanol overnight at  20 1C. After washing in PBS, the cells were treated for 30 min at 37 1C with 50 mg ml  1 PI and 1 mg ml  1 RNase A. Subsequent analyses of cell cycle distribution and apoptosis were performed using CELLQuest software (BD Biosciences). Each experiment was repeated three times.

Analysis of cell cycle-specific apoptosis (annexin V and propidium iodide method) Cells were harvested with 0.25% trypsin, washed with PBS and resuspended in binding buffer. A 5-ml aliquot of fluorescein isothiocyanateannexin V was added to 100 ml of freshly collected cells suspended in binding buffer at a density of 106 cells per ml and incubated in the dark for 20 min. The sample was then resuspended in 1 ml of 1% methanol-free formaldehyde in binding buffer for 30 min on ice and rinsed twice. The Cancer Gene Therapy (2014), 1 – 9

sample was resuspended in 0.5 ml of PI solution containing 50 mg ml  1 PI, 0.1% RNase A (Sigma, St Louis, MO, USA), 500 mg ml  1 digitonin (Sigma), 10 mM PIPES (Sigma), 2 mM CaCl2 and 0.1 M NaCl, placed at room temperature for 1 h in the dark, and analyzed by flow cytometry for the cell cycle specificity of apoptotic cells. Untreated cells served as the negative control.

Clinical specimens The current study included specimens from 223 patients including 44 patients with primary endometrial carcinoma, 43 patients with primary cervical carcinoma, 39 patients with primary ovarian carcinoma and 33 patients with primary breast carcinoma who underwent surgery. None of the patients with malignant tumors received neoadjuvant chemotherapy or radiotherapy. Non-neoplastic tissues were also collected, including 11 patients with endometria in the proliferative phase, 10 patients with endometria in the secretory phase, 20 patients with non-neoplastic cervical tissues, 10 patients with non-neoplastic ovaries and 13 patients with non-neoplastic breast tissues. The tissue samples were obtained and handled in accordance with a protocol approved by the Institutional Review Board for Human Research of the TongJi Hospital. The slides were evaluated and scored by two pathologists who were blinded to each other to prevent observer bias. Each data point represents the mean of the results from the two pathologists.

Immunohistochemistry Immunohistochemistry was carried out on paraffin-embedded sections using standard reagents and protocols (Vector Laboratories, Burlingame, CA, USA). For a semiquantitative evaluation, an immunoreactivity-scoring system was applied. The intensity of staining was designated as either not existent (0), weak (1), moderate (2) or strong (3). The number of cells stained was scored as either 0–9% cells stained (0), 10–39% of cells stained (1), 40–69% of cells stained (2) or 470% of cells stained (3). The immunoreactivity scoring was calculated by the multiplication of these two variables; the scores can range from 0 to 9. We confirmed a 0 score as negative to calculate the positive rate. These data were analyzed as a continuum, and the objective was to use this semiquantitative method to assess differences between the non-neoplastic and neoplastic tissues. The percentage of immunoreactive cells was calculated by counting at least 1000 tumor cells at high magnification (  400).

Mouse tumor model Four-week-old female BALB/c nu/nu mice were purchased from the SLAC Laboratory Animal (Shanghai, China) and maintained in a laminar-flow cabinet under specific pathogen-free conditions. MDA-MB-231 tumor cells (1  106) were implanted into the mammary fat pad of the mice. When the tumors had grown to B4–5 mm, a dose of 2  108 plaque formation unit virus mutant was injected intravenously by the tail vein daily for five consecutive days (12 mice per group). Four days later, cisplatin at a dose of 0.75 mg kg  1 per day was injected into the abdominal cavity for four consecutive days. The tumors were monitored twice weekly until the mice were killed at 70 days after the initiation of treatment or when the tumor volume was 41200 mm3. The lymph nodes (axillary, supraclavicular and paratracheal), lungs and livers were collected and subjected to pathologic evaluation by pathologists for evidence of metastases.

Statistical analysis All the data are presented as the means±s.d. of at least triplicate experiments and were analyzed by a one-way analysis of variance followed by a Student–Newman–Keuls test. Differences at Po0.05 were considered significant. For the in vivo data, the cumulative probability of survival was determined by the Kaplan–Meier method and the significance of differences with the log-rank test. All P-values were two-sided, and values o0.05 were considered significant. SPSS v11.5 software (SPSS, CA, USA) was used for all the statistical procedures.

RESULTS Silencing Chk1 and Chk2 has little effect on apoptosis and cell cycle progression in the absence of DNA-damaging agents Among the intricate network of sensors, transducers and effectors of DNA damage response pathways in eukaryotic cells, the Chk1 & 2014 Nature America, Inc.

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

3

Figure 1. Checkpoint kinases 1/2 (Chk1/Chk2) small interference RNA (siRNA) knocked down Chk1/Chk2 expression without disturbing cell cycle profiles in the absence of DNA damage. HeLa, A2780 and MCF-7 cells were transfected with Chk1 siRNA, Chk2 siRNA or an siRNA control. After 48 h, half of the samples were subjected to a western blot analysis, and the other half were subjected to a FACS analysis. (a) Western blotting results showing the levels of Chk1 and Chk2 and b-actin as the control. (b) Chk1/Chk2-specific siRNA- or Si-con-transfected HeLa cells were checked for levels of Chk1/Chk2 proteins by confocal laser-scanning microscopy analyses. The nuclei are stained red and immunoreactive Chk1/Chk2 proteins are stained green (red arrows). (c) Cell cycle and apoptosis profiles were determined by a FACS analysis in HeLa cells after 48 h of incubation. FACS, fluorescence-activated cell sorting.

and Chk2 kinases have recently attracted much attention because their functions in cell cycle checkpoint mechanisms are capable of blocking the cell cycle and stimulating DNA repair. To determine whether the downregulation of Chk1 and Chk2 expression would affect cell cycle and apoptosis in somatic cells, we transfected Chk1 or Chk2 siRNA into a panel of cancer cell lines, MCF-7, A2780, HeLa, A549, HepG2 and SiHa, for 48 h before the cells were harvested. Expression of the Chk1 and Chk2 proteins were significantly knocked down by Chk1- or Chk2-specific siRNA in the MCF-7, A2780 and HeLa cells compared with the control (Si-con) (Figure 1a), and the nuclear and cytoplasmic expression of Chk1 and Chk2 was also markedly decreased in individual cells, as revealed by confocal laser-scanning microscopy analyses (Figure 1b). A fluorescence-activated cell sorting analysis showed that the downregulation of the Chk1 or Chk2 protein produced the same cell cycle profile as the control siRNA samples, with no increase in apoptotic cells up to 48 h (Figure 1c) or even 72 h (data not shown) in three independent experiments. These observations indicate that Chk1 and Chk2 are dispensable for cell survival and the control of cell cycle progression in somatic cells. DNA damage activates various checkpoints in asynchronous cancer cells Although many studies have reported that the inhibition of Chk1 or Chk2 sensitized tumor cells to DNA-damaging agents, & 2014 Nature America, Inc.

most observations have been performed either in chemically synchronized cells or in cells with an activated G2/M checkpoint. To determine the responsive pattern of asynchronously grown cancer cells to DNA-damaging agents, nine cancer cell lines, such as HeLa, A2780, MCF-7, A2780, HepG2, SiHa, HEC-1-B, HL-60 and K562 cells, were treated by cisplatin or irradiation over a range of varied doses and analyzed for cell cycle profiles at different time points after the initiation of treatment. These cell lines exhibited the most profound cell cycle arrest after the exposure to irradiation or cisplatin because an activated checkpoint was evident by the excessive accumulation of cells at the G1, S and G2/M phases. For the subsequent experiment, we choose three types of cancer cells that varied in their tissues of origin and in their response to DNA damage: cervical cancer cell line HeLa with a mutant Rb and p53 inactivated by HPV E6; lung cancer cell line A2780 with a wild-type p53 and mutant Rb; and breast cancer cell line MCF-7 with a wild-type p53 and wild-type Rb. Activation of the G1 checkpoint became evident at 6 h after the exposure of unsynchronized MCF-7 cells to 6 Gy 60Co irradiation and reached maximal arrest at 24 h (Figure 2a). In contrast, activation of the S checkpoint was observed at the 12-h time point and reached a maximum at 24 h in a time-dependent fashion after the exposure of unsynchronized A2780 to 20 mM cisplatin (Figure 2b). Activation of the G2/M checkpoint after the exposure of unsynchronized HeLa cells to 10 Gy 60Co irradiation became apparent at the 6-h time point and reached a peak at 48 h in a Cancer Gene Therapy (2014), 1 – 9

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

4

Figure 2. DNA-damaging agents induced G1, S and G2/M cell cycle arrest models by activating various checkpoints in asynchronous tumor cells. Unsynchronized tumor cells (MCF-7, A2780 and HeLa) were treated by cisplatin or irradiation at the indicated doses. At various time points, the cells were pelleted for cell cycle and apoptosis analyses. The arrested cells gave rise to a net increase in G1-phase in MCF-7 cells after irradiation (a), a net increase in S-phase in A2780 cells after cisplatin treatment (b), and a net increase in G2/M-phase in HeLa cells after irradiation (c); the respective apoptosis net increase is also shown. The right panel represents a typical result of G1-, S- or G2/M-phase arrest, as determined by flow cytometry in the three cell lines after DNA-damaging agent treatment. IR, irradiation.

time-dependent fashion (Figure 2c). Very interestingly, the most profound cell cycle accumulation was always accompanied with a minor apoptotic response under this experimental condition, and the apparent apoptotic cells always appeared later than the maximum cell cycle arrest time point in all the cell lines tested. This finding indicates that the G1, S and G2/M checkpoints provide tumor cells with sufficient time to repair before cell cycle progression is recovered, and that the damaged cells eventually move toward apoptosis because of the expected induction of mitotic catastrophe (Figure 2). Chk1 and Chk2 siRNAs abrogate checkpoints, enhance apoptosis induced by DNA-damaging agents, and show synergistic effects when used in combination To determine whether Chk1 or Chk2 siRNAs, which have no apparent toxicity in the absence of DNA-damaging agents, can abrogate cell cycle arrest and sensitize cancer cells to apoptosis in the presence of DNA-damaging agents, we transfected cells with either 50 nM Chk1- or Chk2-specific siRNA Cancer Gene Therapy (2014), 1 – 9

and incubated for 24 h before subjecting the cells to irradiation or cisplatin. The floating and adherent cells were then harvested at the maximum cell cycle arrest time point mentioned above for each cell model for cytometry analyses. Although the Si-coninfected HeLa cells exhibited a typical G2/M arrest, the exposure to Chk1 or Chk2 siRNA but not Si-con completely abolished this G2/M arrest and enhanced apoptosis by 200% compared with the corresponding control siRNA (Po0.01) (Figures 3a-f). This enhanced apoptosis could be detected by cleaved caspase-3 and the release of cytochrome c from the mitochondria (Figure 3c and d). Similar results were found with regard to induced S-phase arrest in A2780 cells treated with cisplatin and induced G1-phase arrest in MCF-7 cells treated with irradiation (data not shown). More interestingly, the combined use of Chk1 and Chk2 siRNAs produced synergistic effects compared with the treatment with Chk1- or Chk2-specific siRNA alone in all of the cell models tested (Po0.05). To examine the basis for Chk1- or Chk2-specific siRNAenhanced apoptosis, the API method was used to analyze cell cycle-specific apoptosis in Chk1 or Chk2 siRNA-transfected MCF-7 & 2014 Nature America, Inc.

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

5

Figure 3. Checkpoint kinases 1 and 2 (Chk1 and Chk2) small interference RNA (siRNAs) abrogated checkpoints, enhanced the apoptosis induced by DNA-damaging agents, and showed synergistic effects when used in combination. Tumor cells (HeLa or primary acute leukemia cells) were transfected with Chk1/Chk2 siRNA or Si-con and incubated for 24 h. The cells were then irradiated at a dose of 10 Gy (for HeLa cells) or 2 Gy (for leukemia cells) and maintained in complete medium for another 48 h for HeLa or 24 h for primary leukemia cells before analysis. The results represent data from three independent experiments. Control, no treatment; Si-con, siRNA control at a concentration of 50 nM compared with Chk1/Chk2 siRNA treatment alone or 100 nM compared with Chk1 and Chk2 siRNA combined treatment; Si-Chk1 þ 2, 50 nM of Chk1 siRNA plus 50 nM of Chk2 siRNA. (a) Combined Chk1- and Chk2 siRNAs showed synergistic effects in inducing apoptosis in tumor cells (HeLa). (b) The representative cell morphology is shown. (c) Cleaved caspase-3 was clearly found in the combined Chk1 and Chk2 siRNAtriggered apoptosis. GAPDH was used as a control for equal protein loading. (d) HeLa cells were harvested after treatment, and the mitochondria was purified by subcellular fractionation. Cytochrome c in the supernatant was detected with an anti-cytochrome c monoclonal antibody. Voltage-dependent anion-selective channel 1 (VDAC1) from total mitochondria was used as a control for equal protein loading. Combined Chk1 and Chk2 siRNAs showed synergistic effects by significantly enhanced apoptosis (*Po0.05) (e) and abrogated cell cycle arrest (f ) in HeLa cells. **Po0.01. (g) Although Si-con-transfected MCF-7 cells exposed to irradiation showed preferentially triggered G1-phase apoptosis, Chk1/2 siRNA-transfected MCF-7 cells showed either G1-, S- or G2/M-phase apoptosis. (h) Primary acute leukemia cells (PLCs) were isolated and tested for the effects of Chk1- or Chk2-specific siRNA. Combined Chk1- and Chk2 siRNAs showed synergistic effects in PLCs (Po0.05). IR, irradiation.

cells. Interestingly, the MCF-7 cells treated by irradiation alone underwent apoptosis preferentially in G1-phase, whereas those treated with irradiation plus Chk1- or Chk2-specific siRNA but not Si-con showed apoptosis in either G1, S or G2/M-phase, suggesting that the inhibition of endogenous Chk1 or Chk2 sensitized the MCF-7 cells by altering the executive machinery of apoptosis (Figure 3g). To confirm our finding in primary cancer cells, we extended our analysis to three clinical specimens of & 2014 Nature America, Inc.

primary acute myeloid leukemia cells and tested the effects of Chk1- or Chk2-specific siRNA knockdown. The irradiated leukemic cells exhibited a typical G1 accumulation. Treatment with either Chk1- or Chk2-specific siRNA for 24 h before 2 Gy irradiation led to at least a 200% increase in apoptosis. Again, combined Chk1 and Chk2 knockdown resulted in synergistic effects compared with treatment with Chk1- or Chk2- specific siRNA alone (Po0.05; Figure 3h). Cancer Gene Therapy (2014), 1 – 9

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

6 Table 1. Chk1 and Chk2 expression as defined by immunostaining score in endometrial carcinoma, cervical cancer, ovarian cancer and breast cancer, and corresponding non-neoplastic tissues Groups

Immunostaining score (mean±s.d.) Chk1

Endometrium Cervix Ovary Breast

Chk2

Nonneoplastic

Neoplastic

Nonneoplastic

Neoplastic

3.86±2.46 4.45±1.88 2.40±2.21 6.08±3.35

2.98±3.46 4.95±3.10 2.87±3.11 3.52±2.84

3.76±2.72 4.40±1.90 2.30±1.25 4.92±4.29

5.36±3.76 4.00±3.72 3.38±3.34 4.70±3.40

Abbreviations: Chk1, checkpoint kinase 1; Chk2, checkpoint kinase 2.

tumor specimens with evidence of metastasis and B90% of the neoplastic tissues examined. Figure 4. Expression of checkpoint kinases 1 and 2 (Chk1 and Chk2) proteins in neoplastic and non-neoplastic tissues. Immunohistochemical staining of Chk1 and Chk2 proteins was performed in a total 223 tissue samples, including neoplastic tissues (endometrial carcinoma ¼ 44, cervical carcinoma ¼ 43, ovarian carcinoma ¼ 39 and breast carcinoma ¼ 33) and their corresponding non-neoplastic tissues (endometria ¼ 21, cervices ¼ 20, ovaries ¼ 10 and breasts ¼ 13). (a) Immunoreactive Chk1 and Chk2 were readily detectable in both cancerous and non-cancerous tissues. The brown staining indicates immunoreactive Chk1 or Chk2. (b) Different distributions of Chk1 and Chk2 in muscle, venous vessels, arterial vessels and cervical epithelial tissues. The representative images depict the similar data observed in all samples examined.

Expression of Chk1 and Chk2 proteins in primary cancer tissues versus non-neoplastic tissues of respective organs To be an ideal target for cancer therapy, a molecule must be preferentially expressed in cancer cells rather than the normal cells of the respective organs. There is, however, only very limited information available on the expression of the Chk1 and Chk2 proteins in cancers versus non-neoplastic tissues of the respective organs. To provide a rationale for the potential combination of Chk1- and Chk2-targeting therapy, we further determined the expression of the Chk1 and Chk2 proteins in primary ovarian carcinoma, endometrial carcinoma, cervical cancer and breast cancer specimens and compared with the corresponding nonneoplastic tissue specimens of respective organs. The Chk1 protein was ubiquitously expressed in all specimens examined. Chk2-positive staining of the cytoplasm and nucleus was found in B90% of the neoplastic tissues examined (Figure 4a). A semiquantitative immunoreactivity H-scoring system was used to compare the expression of the Chk1 and Chk2 proteins in primary cancer specimens versus corresponding non-neoplastic tissues of origin; the results are illustrated in Table 1. Although Chk1 and Chk2 protein expression was found to be at similar levels in the cancer specimens in comparison with the non-neoplastic tissue specimens of respective organs, the distribution in normal tissues was slightly different (Figure 4b). The positive staining of blood vessels and muscle was observed in sections of the Chk1 proteinexpressing tissues but not in the Chk2 protein-expressing tissues. In normal cervical epithelium, the expression of the Chk2 protein was found in the total epithelium, but the positive staining of Chk1 was determined to become progressively weaker from the basement to the surface of the cervical epithelium. The coexpression pattern of Chk1 and Chk2 occurred in almost all of the Cancer Gene Therapy (2014), 1 – 9

Selectively targeting Chk1 and Chk2 shows synergistic effects in vitro when used in combination The universal expression pattern of the Chk1 and Chk2 proteins in neoplastic and non-neoplastic tissues at least indicated that targeting Chk1 and Chk2 made sense in targeted therapy design, and also hinted that a novel approach was urgently needed to preferentially target Chk1 and Chk2 in cancer rather than in normal cells to achieve therapeutic windows. As described previously, M2 and M3 were generated from Adv5/dE1A, an adenoviral mutant with a deletion of amino acids 121–129 in E1A CR2, through the replacement of the E3 6.7K/gp19K by a fragment of reverse Chk1 cDNA (M2) or reverse Chk2 cDNA (M3).19,20 Both M2 and M3 have been demonstrated in our previous reports to exhibit potent efficacy in vitro and in vivo by virtue of combining oncolysis with efficient Chk1/Chk2 silencing.19,20 In the present study, we sought to determine whether the combined silencing of Chk1 and Chk2 could achieve synergistic effects. As shown in Figure 5a, although the apoptosis induced by DNA-damaging agent cisplatin in A2780 cells was apparently potentiated by M2 or M3 (Po0.01), the killing efficacy was still strikingly enhanced by the combination of M2 and M3 (Po0.05 compared with M2 or M3 alone). These data were further confirmed in MCF-7 and HeLa cells (Figure 5b). Conversely, M2 or M3 did not significantly sensitize human umbilical vein endothelial cells to DNA-damaging agents featuring a good therapeutic window (Figure 5b). Combined silencing of Chk1 and Chk2 demonstrates superior antitumoral efficacy in vivo To further assess the chemosensitized effects of the combined silencing Chk1 and Chk2 in vivo, we compared the antitumoral potency of M2 and M3 in combination with cisplatin in MDA-MB231, an orthotopic breast cancer model, which has been well characterized as having a high frequency of metastasis and resistance to treatment. The MDA-MB-231 breast cancer-bearing mouse model was established previously.21 When the tumors had grown to B4–5 mm in size, the mice were injected intravenously with various adenoviral mutants followed by treatment with cisplatin. Although all the mice (12/12) treated with cisplatin plus PBS died before day 35 after treatment, M2 or M3 injection alone or in combination plus cisplatin improved survival. Either M2 or M3 was significantly superior to improving survival compared with Adv5/dE1A (P ¼ 0.045, 0.017; Figure 6a). In accordance with the in vitro results, the combination of M2 and M3 further improved survival compared with M2 or M3 injection alone (P ¼ 0.0394, 0.0393; Figure 6a). A higher number of CR was achieved following & 2014 Nature America, Inc.

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

7

Figure 5. M2 and M3 sensitized tumor cells to DNA-damaging agents and produced synergistic effects when used in combination. Cancer cells (MCF-7, A2780 and HeLa) and human umbilical vein endothelial cells (HUVECs) were infected with virus at an MOI of 1 and then cultured for 24 h. The cells were then treated with either 20 mM cisplatin (for A2780) or 6 Gy irradiation (for MCF-7, HeLa or HUVECs) and cultured for another 48 h for HeLa, 24 h for MCF-7, 24 h for A2780 and 24 h for HUVECs before being subjected to apoptosis analysis. (a) A representative graph is shown for A2780 cells. (b) The results are the means of three independent experiments. IR/DDP treated with 6 Gy irradiation for MCF7 cells, 10 Gy irradiation for HeLa cells or HUVECs or 20 mM cisplatin DDP for A2780 cells. *Po0.05; M2 and M3 plus IR/DDP compared with M2 or M3 alone plus IR/DDP.

M2- (4/12) or M3- (3/12) treatment alone or in combination (6/12) when compared with Adv5/dE1A (1/12) (Figure 6b). All the mice treated with cisplatin plus PBS were found to have metastases within the lymph nodes at the time of killing, whereas only 1/12 mice treated with M2, 2/12 with M3 and 0/12 with the M2 and M3 combination showed this metastasis (Figure 6c). Efficient infection of the primary tumor was detected by in situ hybridization staining for fiber mRNA after five doses of the systemic administration of M2 or M3 (data not shown). A consequent reduction in Chk1 or Chk2 protein was evident in the M2-, M3- or M2 plus M3-treated mice model (Figure 6d). DISCUSSION Various reports using Chk1 pathway interference methods have already suggested the dominant role of Chk1 in the control of checkpoints in somatic cells. This hypothesis was substantially reaffirmed by our findings that using Chk1-specific siRNA inhibition was sufficient to abrogate DNA damage checkpoints. & 2014 Nature America, Inc.

However, to date, the role of Chk2 in DNA-damaging agent-induced apoptosis is controversial. Chk2 is conflictingly recognized as both a promoter and an inhibitor of apoptosis. Chk2 has been regarded as a tumor suppressor that enhances apoptosis by stabilizing p53.22–24 However, Chk2 has been proven to provide a survival signal in tumor cells because the inhibition of Chk2 facilitates CPT- or doxorubicin-induced apoptosis.25,26 According to our results, Chk2 siRNA markedly enhanced irradiation- or cisplatin-induced apoptosis in various types of cancer cells, suggesting that Chk2 is likely a universal apoptosis inhibitor. It appears that the role of Chk2 varies with the cellular genetic background, form of damage and potency. We propose that Chk2 may circumvent its own tumor-suppressive functions by promoting survival in the current DNA damage models. Given the evidence that Chk1 and Chk2 are both central mediators in the DNA damage response and have been shown to share some overlapping substrates and possess redundant functions such as controlling cell cycle progression,27,28 regulating DNA repair29,30 and coordinating cell survival and death;31,32 Cancer Gene Therapy (2014), 1 – 9

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

8

Figure 6. Antitumoral effects of i.v. administration of M2 and M3 plus cisplatin in the metastatic MDA-MB-231 human breast cancer model. MDA-MB-231 human breast cancer model mice were injected intravenously with various adenoviral mutants at a dose of 2  108 p.f.u. daily for five consecutive days. Four days later, cisplatin at a dose of 0.75 mg kg  1 per day was injected into the abdominal cavity for four consecutive days. Tumors were monitored twice weekly until the mice were killed at 70 days after the initiation of treatment or when the tumor volume was larger than 1200 mm3. (a) Kaplan–Meier survival curves following the injection of M2, M3, M2 þ M3, Adv5/dE1A or PBS in combination with cisplatin in the MDA-MB-231 breast cancer-bearing mouse model. Each group included 12 animals. Po0.05, M2 and M3 plus cisplatin compared with M2 plus cisplatin or M3 plus cisplatin. (b) The tumor CR in each group treated by the various virus mutants plus cisplatin was evaluated at the time of study termination. Phosphate-buffered saline (PBS) plus cisplatin was included as a negative control. Columns, % mice with CR. (c) The ratio of mice with metastases at the time of study termination. (d) The virus mutants were tested for the ability to inhibit Chk1 or Chk2 protein expression in vivo. After five doses of i.v. injection of virus mutants, 30 mg of total protein from the tumor tissues were isolated and subjected to a western blot analysis. i.v., intravenous.

the simultaneous modulation of their activity should be necessary to fully abrogate the checkpoint, and hence confer maximal efficacy of potentiation. The data presented here clearly show that simultaneously silencing Chk1 and Chk2 synergistically enhanced irradiation- or cisplatin-induced apoptosis and achieved maximal efficacy in vitro and in vivo. This overall strategy clearly benefits the development of novel small molecule inhibitors such as AZD7762, a highly specific ATP-competitive Chk1/2 inhibitor, which was found to avoid many of the problems associated with the relatively nonspecific Chk1 inhibitor UCN-01 in initial clinical trials.33 Moreover, Chk1/Chk2 inhibitors may indeed have a place as a substantial therapeutic option, holding particularly great promise for cancers that are refractory to conventional treatment modalities. It was noteworthy that asynchronized cells were chosen for examining the effects of Chk1/Chk2 abrogators. The present cell models are more clinically relevant compared with the checkpoints initiated by synchronizing agents. Synchronizing agents per se can affect the expression pattern of cell cycle-related proteins involved in apoptosis and hence give rise to artificial results. Furthermore, synchronized cell models cover only part of the cell cycle-responsive patterns to DNA-damaging agents and are not sufficient to mimic the real situations of tumor cells undergoing chemo- or radiotherapy in a clinical context. In clinical cancer treatment, DNA-damaging agents would activate heterogeneous patterns of checkpoint activation rather than a Cancer Gene Therapy (2014), 1 – 9

homogeneous G2/M checkpoint activation. Our present data show that tumor cells maintained the G1, S and G2/M checkpoints to some extent, and that silencing Chk1 or Chk2 abrogated G1phase arrest in MCF-7 cells and primary acute myeloid leukemia cells as well as S-phase and G2/M-phase arrest, thus sensitizing tumor cells to DNA-damaging agents in all asynchronized checkpoint models. The present study presents interesting data on the expression levels of the Chk1 and Chk2 proteins in tissues. Although the distribution of the Chk1 and Chk2 proteins in normal tissues varies somewhat, the co-expression pattern of Chk1 and Chk2 occurred in B90% of the neoplastic tissues examined and showed no difference in cancerous versus non-cancerous tissues. To be an ideal target for cancer therapy, a molecule must be preferentially expressed in cancer cells versus normal cells of respective organs. Niida et al. also proved that the combined loss of Chk1 and Chk2 caused the accumulation of cells with spontaneous DNA damage under unperturbed conditions, leading to genomic instability and then tumor development, and suggesting that the inhibition of Chk1 or Chk2 as an approach to cancer therapy should be undertaken with careful consideration.34 Novel adenoviral mutants capable of silencing tumor-associated Chk1 or Chk2 enabled us to overcome the obstacles mentioned above, and further demonstrate that maximum antitumoral efficacy was achieved with the combined coabrogation of Chk1 and Chk2 in the presence of low-dose cisplatin in a refractory to conventional treatment cancer model. & 2014 Nature America, Inc.

Abrogation of Chk1 and Chk2 produces synergistic effects F Ye et al

9 Our findings emphasize the importance of the Chk1 and Chk2 kinases functioning in concert to maintain checkpoints when the cells are under DNA-damaging conditions. Having demonstrated a valuable therapeutic strategy involving the co-abrogation of Chk1 and Chk2 in treatment-resistant cancer, this work deepens our understanding of novel strategies targeting checkpoint pathways and contributes to the development of novel, potent and safe checkpoint abrogators. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS This work was supported by grants from the National Natural Science Foundation of China (81072135, 81372801, 30901749 and 81272426); and the ‘973’ Program of China (no. 2009CB521808).

REFERENCES 1 Zhou BB, Elledge SJ. The DNA damage response: putting checkpoints in perspective. Nature 2000; 408: 433–439. 2 Wong CF, Guminski A, Saunders NA, Burgess AJ. Exploiting novel cell cycle targets in the development of anticancer agents. Curr Cancer Drug Targets 2005; 5: 85–102. 3 Dixon H, Norbury CJ. Therapeutic exploitation of checkpoint defects in cancer cells lacking p53 function. Cell Cycle 2002; 1: 362–368. 4 Zhou BB, Bartek J. Targeting the checkpoint kinases: chemosensitization versus chemoprotection. Nat Rev Cancer 2004; 4: 216–225. 5 Abraham RT. Cell cycle checkpoint signaling through the ATM and ATR kinases. Genes Dev 2001; 15: 2177–2196. 6 Mailand N, Falck J, Lukas C, Syljuasen RG, Welcker M, Bartek J et al. Rapid destruction of human Cdc25A in response to DNA damage. Science 2000; 288: 1425–1429. 7 Bartek J, Lukas J. Chk1 and Chk2 kinases in checkpoint control and cancer. Cancer Cell 2003; 3: 421–429. 8 Zhao H, Watkins JL, Piwnica-Worms H. Disruption of the checkpoint kinase 1/cell division cycle 25A pathway abrogates ionizing radiation-induced S and G2 checkpoints. Proc Natl Acad Sci USA 2002; 99: 14795–14800. 9 Sorensen CS, Syljuasen RG, Falck J, Schroeder T, Ronnstrand L, Khanna KK et al. Chk1 regulates the S phase checkpoint by coupling the physiological turnover and ionizing radiation-induced accelerated proteolysis of Cdc25A. Cancer Cell 2003; 3: 247–258. 10 Falck J, Mailand N, Syljuasen RG, Bartek J, Lukas J. The ATM-Chk2-Cdc25A checkpoint pathway guards against radioresistant DNA synthesis. Nature 2001; 410: 842–847. 11 Falck J, Petrini JH, Williams BR, Lukas J, Bartek J. The DNA damage-dependent intra-S phase checkpoint is regulated by parallel pathways. Nat Genet 2002; 30: 290–294. 12 Shao RG, Cao CX, Shimizu T, O’Connor PM, Kohn KW, Pommier Y. Abrogation of an S-phase checkpoint and potentiation of camptothecin cytotoxicity by 7-hydroxystaurosporine (UCN-01) in human cancer cell lines, possibly influenced by p53 function. Cancer Res 1997; 57: 4029–4035. 13 Lara Jr. PN, Mack PC, Synold T, Frankel P, Longmate J, Gumerlock PH et al. The cyclin-dependent kinase inhibitor UCN-01 plus cisplatin in advanced solid tumors: a California cancer consortium phase I pharmacokinetic and molecular correlative trial. Clin Cancer Res 2005; 11: 4444–4450. 14 Kohn EA, Yoo CJ, Eastman A. The protein kinase C inhibitor Go6976 is a potent inhibitor of DNA damage-induced S and G2 cell cycle checkpoints. Cancer Res 2003; 63: 31–35.

& 2014 Nature America, Inc.

15 Jobson AG, Cardellina 2nd JH, Scudiero D, Kondapaka S, Zhang H, Kim H et al. Identification of a Bis-guanylhydrazone [4,4’-Diacetyldiphenylurea-bis(guanylhydrazone); NSC 109555] as a novel chemotype for inhibition of Chk2 kinase. Mol Pharmacol 2007; 72: 876–884. 16 Collins I, Garrett MD. Targeting the cell division cycle in cancer: CDK and cell cycle checkpoint kinase inhibitors. Curr Opin Pharmacol 2005; 5: 366–373. 17 Chen Z, Xiao Z, Chen J, Ng SC, Sowin T, Sham H et al. Human Chk1 expression is dispensable for somatic cell death and critical for sustaining G2 DNA damage checkpoint. Mol Cancer Ther 2003; 2: 543–548. 18 Ou YH, Chung PH, Sun TP, Shieh SY. p53 C-terminal phosphorylation by CHK1 and CHK2 participates in the regulation of DNA-damage-induced C-terminal acetylation. Mol Biol Cell 2005; 16: 1684–1695. 19 Gao Q, Huang X, Tang D, Cao Y, Chen G, Lu Y et al. Influence of chk1 and plk1 silencing on radiation- or cisplatin-induced cytotoxicity in human malignant cells. Apoptosis 2006; 11: 1789–1800. 20 Gao Q, Zhou J, Huang X, Chen G, Ye F, Lu Y et al. Selective targeting of checkpoint kinase 1 in tumor cells with a novel potent oncolytic adenovirus. Mol Ther 2006; 13: 928–937. 21 Zhou J, Gao Q, Chen G, Huang X, Lu Y, Li K et al. Novel oncolytic adenovirus selectively targets tumor-associated polo-like kinase 1 and tumor cell viability. Clin Cancer Res 2005; 11: 8431–8440. 22 Hirao A, Kong YY, Matsuoka S, Wakeham A, Ruland J, Yoshida H et al. DNA damage-induced activation of p53 by the checkpoint kinase Chk2. Science 2000; 287: 1824–1827. 23 Hirao A, Cheung A, Duncan G, Girard PM, Elia AJ, Wakeham A et al. Chk2 is a tumor suppressor that regulates apoptosis in both an ataxia telangiectasia mutated (ATM)-dependent and an ATM-independent manner. Mol Cell Biol 2002; 22: 6521–6532. 24 Jack MT, Woo RA, Hirao A, Cheung A, Mak TW, Lee PW. Chk2 is dispensable for p53-mediated G1 arrest but is required for a latent p53-mediated apoptotic response. Proc Natl Acad Sci USA 2002; 99: 9825–9829. 25 Castedo M, Perfettini JL, Roumier T, Yakushijin K, Horne D, Medema R et al. The cell cycle checkpoint kinase Chk2 is a negative regulator of mitotic catastrophe. Oncogene 2004; 23: 4353–4361. 26 Huang M, Miao ZH, Zhu H, Cai YJ, Lu W, Ding J. Chk1 and Chk2 are differentially involved in homologous recombination repair and cell cycle arrest in response to DNA double-strand breaks induced by camptothecins. Mol Cancer Ther 2008; 7: 1440–1449. 27 Yu Q, Rose JH, Zhang H, Pommier Y. Antisense inhibition of Chk2/hCds1 expression attenuates DNA damage-induced S and G2 checkpoints and enhances apoptotic activity in HEK-293 cells. FEBS Lett 2001; 505: 7–12. 28 Xiao Z, Chen Z, Gunasekera AH, Sowin TJ, Rosenberg SH, Fesik S et al. Chk1 mediates S and G2 arrests through Cdc25A degradation in response to DNA-damaging agents. J Biol Chem 2003; 278: 21767–21773. 29 Sorensen CS, Hansen LT, Dziegielewski J, Syljuasen RG, Lundin C, Bartek J et al. The cell-cycle checkpoint kinase Chk1 is required for mammalian homologous recombination repair. Nat Cell Biol 2005; 7: 195–201. 30 Wang HC, Chou WC, Shieh SY, Shen CY. Ataxia telangiectasia mutated and checkpoint kinase 2 regulate BRCA1 to promote the fidelity of DNA end-joining. Cancer Res 2006; 66: 1391–1400. 31 Ghosh JC, Dohi T, Raskett CM, Kowalik TF, Altieri DC. Activated checkpoint kinase 2 provides a survival signal for tumor cells. Cancer Res 2006; 66: 11576–11579. 32 Zaugg K, Su YW, Reilly PT, Moolani Y, Cheung CC, Hakem R et al. Cross-talk between Chk1 and Chk2 in double-mutant thymocytes. Proc Natl Acad Sci USA 2007; 104: 3805–3810. 33 Ma CX, Cai S, Li S, Ryan CE, Guo Z, Schaiff WT et al. Targeting Chk1 in p53-deficient triple-negative breast cancer is therapeutically beneficial in human-in-mouse tumor models. J Clin Invest 2012; 122: 1541–1552. 34 Niida H, Murata K, Shimada M, Ogawa K, Ohta K, Suzuki K et al. Cooperative functions of Chk1 and Chk2 reduce tumour susceptibility in in vivo. EMBO J 2010; 29: 3558–3570.

Cancer Gene Therapy (2014), 1 – 9