Lentiviral shRNA screen of human kinases identifies PLK1 as a potential therapeutic target for osteosarcoma

Lentiviral shRNA screen of human kinases identifies PLK1 as a potential therapeutic target for osteosarcoma

Cancer Letters 293 (2010) 220–229 Contents lists available at ScienceDirect Cancer Letters journal homepage: www.elsevier.com/locate/canlet Lentivi...

2MB Sizes 0 Downloads 40 Views

Cancer Letters 293 (2010) 220–229

Contents lists available at ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

Lentiviral shRNA screen of human kinases identifies PLK1 as a potential therapeutic target for osteosarcoma q Zhenfeng Duan a,*, Diana Ji b, Edward J. Weinstein b, Xianzhe Liu a, Michiro Susa a, Edwin Choy a, Cao Yang a, Henry Mankin a, Francis J. Hornicek a a b

Center for Sarcoma and Connective Tissue Oncology, Massachusetts General Hospital, Boston, MA 02114, USA Research Biotechnology, Sigma–Aldrich Corporation, St. Louis, MO 63103, USA

a r t i c l e

i n f o

Article history: Received 30 November 2009 Received in revised form 13 January 2010 Accepted 16 January 2010

Keywords: shRNA PLK1 Osteosarcoma

a b s t r a c t We describe an optimized systematic screen of known kinases using osteosarcoma cell lines (KHOS and U-2OS) and a lentiviral-based short hairpin RNA (shRNA) human kinase library. CellTiter 96ÒAQueous One Solution Cell Proliferation Assay was used to measure cell growth and survival. We identified several kinases, including human polo-like kinase (PLK1), which inhibit cell growth and induce apoptosis in osteosarcoma cells when knocked down. cDNA rescue and synthetic siRNA assays confirm that the observed phenotypic changes result from the loss of PLK1 gene expression. Furthermore, a small molecule inhibitor to PLK1 inhibited osteosarcoma cell growth and induced apoptosis. Western blot analysis confirmed that PLK1 is highly expressed and activated in several osteosarcoma cell lines as well as in resected tumor samples. Immunohistochemistry analysis showed that patients with high PLK1 tumor expression levels correlated with significantly shorter survival than patients with lower levels of tumor PLK1 expression. These results demonstrate the capability and feasibility of a highthroughput screen with a large collection of lentiviral kinases and its effectiveness in identifying potential drug targets. The development of more potent inhibitors that target PLK1 may open doors to a new range of anti-cancer strategies in osteosarcoma. Ó 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Osteosarcoma accounts for 60% of primary malignant bone tumors diagnosed in the first two decades of life. Standard treatment for osteosarcoma is surgery and neoadjuvant chemotherapy [1]. Chemotherapy has significantly improved the survival rate from 11% with surgery alone to 60–70% when surgery is combined with chemotherapy [1,2]. Patients with advanced osteosarcoma after front-line chemotherapy usually receive further treatment

q This work was presented, in part, at the annual meeting of the Connective Tissue Oncology Society, London, UK, November 13, 2008 and at the annual meeting of the AACR-NCI-EORTC Molecular Targets and Cancer Therapeutics, Boston, November 15, 2009. * Corresponding author. E-mail address: [email protected] (Z. Duan).

0304-3835/$ - see front matter Ó 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2010.01.014

with additional chemotherapy, which may be considered toxic [3]. Unfortunately, not much progress has been made on improving survival over the past 20 years with regard to the treatment of osteosarcoma. Thus, identification of new targeted therapies is a crucial step forward in the drive towards personalized medicine [1,3,4]. The human kinome contains at least 600 protein kinases that phosphorylate proteins at 250,000 or more sites [5–7]. A functional understanding of the role of the kinome in osteosarcoma is currently incomplete, and a study of these proteins and their functions would contribute to the discovery and development of new therapeutics. Several kinases such as IGF-1R, PI3 K, AKT, PDGFR, and mTOR have been found to be highly expressed in different sarcomas, particularly in advanced stages [8–12]. It has been shown that the suppression of these kinases, such as PI3 K, AKT/mTOR, Src, IGF-R, EGFR, JIK, and JAK, inhibit

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

tumor cell growth and proliferation, suggesting the potential utility of these kinases as drug targets [13,14]. Certain kinases have been reported to be highly expressed and activated in osteosarcoma [1,15]. Identification of novel kinases whose inhibition induces osteosarcoma cell lethality would be of high clinical value. The ability to use small interfering RNA (siRNA) as a tool for functional gene silencing in tumor cells has enabled us to perform unbiased genetic loss-of-function analyses in tissue culture systems [16,17]. While siRNA has been shown to be effective for short-term gene inhibition in mammalian cell lines, its use for stable transcriptional knockdown have been problematic [18]. Recently, short hairpin RNA (shRNA) libraries in lentiviral vectors have been described and can be used to generate stable cell lines in a wide variety of cells [19,20]. In the present study, the roles of protein kinases in supporting osteosarcoma cell growth were examined using MISSIONÒ LentiExpress™ Human Kinases shRNA library. We show that inhibition of polo-like kinase (PLK1), as well as several other kinases can decrease growth and induce apoptosis in osteosarcoma cells. Additionally, we observe that PLK1 has high endogenous expression levels in osteosarcoma cell lines and osteosarcoma tissues. Overexpression of PLK1 in tumors closely correlates with a poor prognosis in patients. Furthermore, a small molecule inhibitor for PLK1 significantly inhibits osteosarcoma cell growth and induces apoptosis. These data suggest that inhibition of PLK1 can slow proliferation and induce apoptosis in osteosarcoma. 2. Materials and methods 2.1. Cell lines and cell culture Dr. Efstathios Gonos (Institute of Biological Research and Biotechnology, Athens, Greece) provided the osteosarcoma KHOS cell line [21]. The human osteosarcoma cell lines, U-2OS and Saos, were obtained from the ATCC (Rockville, MD). Osteosarcoma cell line OSA344 was established from primary osteosarcoma tissue. Human osteoblast cells HOB-c were obtained from PromoCell GmbH (Heidelberg, Germany), osteoblast cells NHOst were obtained from Lonza Wallkersville Inc. (Walkersville, MD), and osteoblast cells hFOB were obtained from ATCC. Osteoblast cells were cultured in osteoblast growth medium (PomoCell) with 10% fetal bovine serum (FBS). All other cell lines were cultured in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% FBS, 100-units/ml penicillin and 100 lg/ml streptomycin (Invitrogen). 2.2. PLK1 inhibitor and human PLK1 cDNA clone PLK1 inhibitor, scytonemin, a natural marine product, was purchased from EMD Chemicals, Inc. (Gibbstown, NJ). Scytonemin is a cell-permeable, dimeric indolo-phenol PLK1 kinase inhibitor that exhibits anti-proliferative and anti-inflammatory properties [22]. Scytonemin was dissolved in DMSO at a concentration of 25 mg/ml and stored at 20 °C. Human PLK1 cDNA clone (Genbank accession no.: NM_005030.3) was purchased from OriGene Technol-

221

ogies, Inc. (Rockville, MD). PLK1 cDNA was cloned into pCMV6-XL5 vector as transfection-ready DNA. 2.3. MISSIONÒ LentiExpress™ human kinases shRNA library The role of protein kinases in maintaining or supporting osteosarcoma cell growth were examined using MISSIONÒ LentiExpress™ Human Kinases shRNA library (Sigma, Saint Louis, MO). This library contains 3109 lentiviruses carrying shRNA sequences targeting 673 human kinase genes allowing for quick, high-throughput loss-of-function screens. The library was cloned into a lentiviral-based pLKO.1-puro shRNA expression vector. Each gene is represented by a target set that consists of 3–5 individual constructs, targeting different regions of the gene sequence. The LentiExpress™ human kinases shRNA library consists of 41 plates. The particles are pre-diluted to approximately 5000 viral particles per well in a single reaction volume of 30 ll. Each well contains lentiviral particles encoding a single shRNA that is designed to target a single kinase mRNA. In addition to the virus generated for the human kinases, each plate includes negative controls (empty vector and non-target shRNA controls) to monitor transduction efficiency and well-to-well comparison of results. 2.4. Determination of the optimal cell density (cells/ml) for transduction Transduction conditions in KHOS was optimized using MISSIONÒ non-target shRNA Control Transduction particles (SHC002 V) (Sigma) in determining its efficiency as a negative control, optimal puromycin concentrations for selection and assay times prior to the kinase panel screen. Optimization of cell density was determined in KHOS by MISSIONÒ LentiExpress™ Optimization Plate (SHXC01) as a positive control for experiment. This plate allows rapid determination of the optimal cell density for transduction. The plate contains pre-arrayed aliquots of TurboGFP™ Control Transduction particles, so that various seeding densities of the cell type of interest may be added with ease. Efficiency of transduction is subsequently monitored by the assessment of GFP fluorescence under fluorescence microscopy. 2.5. Lentiviral human kinase shRNA library screen After optimization of cell seeding density, concentration of puromycin and assay times, the effects of kinase knockdown by lentiviral shRNA were carried out with the LentiExpress™ human kinases shRNA library. In brief, on day 1, the LentiExpress™ kinase plate was thawed at room temperature and KHOS cells were diluted to 40,000 cells/ml in complete medium. Polybrene was added to a final concentration of 11.3 lg/ml and then 70 ll of cell suspension was added to each well of a LentiExpress™ kinase plate. The plate was then covered and transferred to the tissue culture incubator and incubated overnight. This allows time for the shRNAs to be expressed, to bind to their target kinase mRNAs, and cause a reduction in expression of the encoded protein. On day 2, the plates were removed from the incubator and the media was gently aspirated and dis-

222

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

posed of in 10% bleach solution. Hundred microliters of complete media with puromycin (1 lg/ml) was replaced in each well. Plates were covered and returned to the tissue culture incubator. From day 3 to day 6, fresh medium was replaced as necessary, as described previously, and wells were evaluated for cell proliferation under the microscope. On day 6, the number of viable cells was determined via CellTiter 96ÒAQueous One Solution Cell Cytotoxicity Assay (Promega, Madison, WI). 2.6. Determination of the screen results and analysis The general format of the screen was designed to evaluate each kinase target through shRNA lentiviral-mediated gene knockdown. The experiments were carried out using the LentiExpress™ human kinases shRNA panel which also

included empty vector controls, non-target shRNA controls, and medium-only controls on each plate to permit well-to-well, as well as plate-to-plate comparisons (Supplementary Fig. 1). The control wells (empty vector and non-target shRNA controls) were used to evaluate transduction efficiency. shRNA targeted kinase genes that are associated with cell death in each well were identified as positive ‘‘hits” and selected for further study. Each of the kinase genes is represented by three to five different shRNA lentiviral particle constructs targeting different sites of each gene. Because these shRNA are complimentary to different regions of the mRNA, it reduces the risk of off-target effects. To further minimize the possibility of off-target hits, we focused on genes identified as being functionally essential in osteosarcoma survival and growth; where only genes targeted by two or more shRNA

Fig. 1. Optimization of cell concentrations for LentiExpress™ kinase screening of osteosarcoma KHOS cells (A) and effect of lentiviral shRNA transduction on KHOS cells (B and C). (A) Optimization of cell concentration. Cells were seeded into the LentiExpress™ optimization plate at various initial cell densities. (A1) 1  104 cells/ml; (A2) 2  104 cells/ml; (A3) 4  104 cells/ml; (A4) 8  104 cells/ml; (A5) 16  104 cells/ml. After 24 h, media was replaced with fresh media supplemented with 1 lg/ml of puromycin. After five additional days of growth with regular media changes, cells were imaged by fluorescence microscopy to determine optimal plating density. 4  104 cells/ml was selected as the optimal plating density for further experimentation. (B) Effect of lentiviral shRNA transduction on KHOS cells. KHOS cells 7 days post-transduction and 5 days under puromycin (1 lg/ml) selection. Puromycin (1 lg/ml) causes complete cell death of KHOS in the media-only wells containing untransduced cells in 7 days (B3) while cells grow unaffected in wells containing pLKO.1 empty vector control particles (B1) and non-target shRNA control particles (B2). (C) Representational plate that shows screen results of the MISSIONÒ LentiExpress™ Kinase shRNA library. KHOS cells were plated into the LentiExpress™ plate and selected with puromycin. To find kinases required for proliferation, cells were then subjected to a cell viability assay. While quantification of this data is possible by CellTiter96ÒAQuesos One Solution, simple visual inspection was sufficient to determine potential ‘‘hits”. C: empty vector control; M: media control; N: non-target shRNA control.

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

clones were selected as a positive hit. Once hits were identified, we validated knockdown at the protein levels by Western blot analysis. 2.7. Proliferation assay Proliferation was assessed using the CellTiter 96ÒAQueous One Solution Cell Assay as previously described [23].

223

containing 10% fetal bovine serum. Twenty-four hour post-transfection; the PLK1 transfected cells were collected and transduced with PLK1 lentiviral shRNA targeting either 30 UTR or open reading frame (ORF) in a 96well plate. At 72 h post-transduction, proliferation of cell growth was determined by CellTiter 96ÒAQueous One Solution Reagent as described above.

2.9. Synthetic PLK1 siRNA and transfection 2.8. PLK1 cDNA rescue assay In order to verify the observed PLK1 knockdown phenotype in osteosarcoma, PLK1 lentiviral shRNA clones against the 30 UTR was utilized in a cDNA rescue assay. With the introduction of PLK1 cDNA (without the 30 UTR), the clone could rescue the knockdown phenotype induced by lentiviral shRNA targeting the 30 UTR of PLK1. Transcripts from an exogenous cDNA clone lacking a 30 UTR would not be susceptible to silencing compared to the endogenous copy. Lentiviral shRNA targeting the 30 UTR of PLK1 was purchased from Sigma (TRCN0000011006). First, KHOS cells were transfected with PLK1 cDNA clone. Transfections were performed using LipofectAmine Plus reagents (Invitrogen) as follows: 5  105 KHOS cells were plated into 90 mm tissue culture dishes and cultured overnight. Prior to transfection, the growth medium was replaced with serum free RPMI 1640 and cultured for 3 h. LipofectAmine reagent containing 5 lg of PLK1 cDNA was combined with LipofectAmine Plus reagent and applied to the cells. After 4 h of culture, the media was replaced with RPMI 1640

Further confirmation of PLK1 knockdown phenotype in osteosarcoma was carried out with synthetic human PLK1 siRNA purchased from Ambion at Applied Biosystems (Foster City, CA). The siRNA sequence targeting PLK1 (Genbank accession no. NM_005030) corresponded to coding regions(sense 50 -CCAUUAACGAGCUGCUUAAtt-30 , antisense 50 -UUAAGCAGCUCGUUAAUGGtt-30 ) of the PLK1 gene. The siRNA oligonucleotides were dissolved in nuclease-free water at a concentration of 100 lM and kept at 20 °C until the following transfection experiment. The non-specific siRNA oligonucleotides (Dharmacon, Chicago, IL) were used as negative controls. KHOS cells were either plated on 96-well plates for cell proliferation assays or plated on dishes for Western blot protein isolation. Transfections were performed with TransMessenger transfection reagent (Qiagen) as directed by the manufacturer. For PLK1, each 96-well plate received 0.1 lg siRNA per well in a volume of 200 ll in triplicate, and each 60 mm dish received 5 lg siRNA per dish in a volume of 10 mL. Medium was replaced with RPMI 1640 supplemented with 10% fetal bovine ser-

Fig. 2. Effects of PLK1 knockdown in osteosarcoma cell lines. (A) Results of lentiviral shRNA directed against PLK1 in KHOS cells from initial screen. The data is representing from a single well of 96-well plate of MISSIONÒ LentiExpress™ Human kinases shRNA library as described in Section 2. (B) PLK1 cDNA rescue assay. Pre-transfection of full length of PLK1 cDNA constructs which lack the 30 UTR could reverse the cell death phenotype when induced by a specific PLK1 lentiviral shRNA which targets the 30 UTR region.

224

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

um 24 h after transfection. Total protein was isolated 48 h post PLK1 siRNA transfection. 2.10. Apoptosis assay Quantification of apoptosis was evaluated using the M30-Apoptosense ELISA assay kit, as per manufacturer’s instructions (Peviva AB, Bromma, Sweden). KHOS, U-2OS, Saos or normal osteoblast cell lines transduced with PLK1 lentiviral shRNA or transfected with synthetic PLK1 siRNA were seeded at 8000 cells/well in a 96-well plate for 24 h before addition of PLK1 inhibitor, scytonemin (EMD Chemicals, Inc., Gibbstown, NJ). After incubation, the cells were then lysed by adding 10 ll of 10% NP-40 per well, and the manufacturer’s instructions for the apoptosis assay were then followed.

2.12. Osteosarcoma tissue microarray and immunohistochemistry The osteosarcoma tissue microarray was purchased from Imgenex Corp. (San Diego, CA) which contains 57 samples from 57 osteosarcoma patients. Immunohistochemistry was performed by following the manufacturer’s instructions. PLK1 positive samples were defined as those showing nuclear and cytoplasmic staining pattern of tumor tissue. PLK1 staining intensity was graded into four groups: no staining (0), weak staining (1+), moderate staining (2+) and intense staining (3+). The correlation between PLK1 expression levels and prognosis was analyzed by Kaplan–Meier survival analysis (GraphPad PRISMÒ 4 software, GraphPad Software, San Diego, CA). A two-sided Student’s t test (GraphPad PRISMÒ 4 software, GraphPad Software, San Diego, CA) was used to compare the PLK1 intensity scores among survivors and non-survivors.

2.11. Western blotting The mouse monoclonal antibody to human PLK1 and actin was purchased from Sigma–Aldrich (St. Louis, MO). The phosphor-cdc25c (Ser216) antibody (Cat#9528) was purchased from the Cell Signaling Technology, Inc. (Cambridge, MA). Western blot was performed as previously reported [23].

2.13. Effects of PLK1 inhibitor, scytonemin, on the proliferation of osteosarcoma cells KHOS or U-2OS cells were treated with scytonemin at various concentrations. Cells were either plated on 96-well plates for sensitivity assays or plated on dishes for Western bloting of PLK1 protein detection. The relative sensitivity of

Fig. 3. Expression of PLK1 in osteosarcoma cell lines and validation of PLK1 knockdown induces apoptosis in multiple osteosarcoma cell lines. (A) Expression of PLK1 in osteosarcoma cell lines as compared with normal human osteoblast cell lines determined by Western blot. Anti-actin monoclonal antibody was used to assess relative protein levels in the sample lanes. (B) Expression of PLK1 in osteosarcoma tissues determined by Western blot. OST1– OST6 represent six different patients’ tissue samples. (C) Knockdown PLK1 induces apoptosis in multiple osteosarcoma cell lines. Apoptosis was evaluated using the M30-Apoptosense ELISA assay kit on cell lines: U-2OS, Saos, OSA344 and normal osteoblast cell HOB-c, after transduction with PLK1 shRNA (TRCN0000006247). PLK1 knockdown induces apoptosis in osteosarcoma cells, but not in normal osteoblast cells.

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

each cell line to scytonemin was determined by CellTiter 96ÒAQueous One Solution Reagent 72 h post-treatment as described above. Total protein was isolated after 72 h of treatment with scytonemin.

3. Results 3.1. Establishment of cell based high-throughput screening of kinase targets in osteosarcoma To investigate the potential function of kinases in the proliferation and survival of osteosarcoma cells, we used a lentiviral shRNA kinase library to screen an osteosarcoma cell line, KHOS. Initial experiments to optimize cell-selection conditions included treatment with 1 lg/ml of puromycin, which was effective for generating complete cell death against uninfected cells after 5 days of incubation. KHOS cells transduced with MISSONÒ TurboGFP™ control lentivirus and subsequently selected with puromycin, confirmed a high transduction efficiency when using 2  104–8  104 ml/cell concentration, as demonstrated by GFP expression percentage (Fig. 1A2–4). Likewise, transduction with MISSIONÒ empty vector control particles or with MISSONÒ non-target lentivirus containing a ‘non-target shRNA’ control that fails to target any known human or mice genes was non-toxic to KHOS (Fig. 1B1 and B2). After optimizing the transduction efficiency and puromycin selection concentration, we transduced KHOS cells with the lentiviral shRNA kinase library that targeted each of the 673 kinase genes. After 6 days, the specific gene knockdown associated with cell death was identified. The results confirmed that exposure of untransduced KHOS to 1 lg/ml of puromycin induced completely cell death (Fig. 1B3). A as a comparison, cells exposed to 1 lg/ml of puromycin infected with either an empty vector control or a non-target shRNA control showed strong cell growth (Fig. 1B1, B2 and C). We found that knockdown of most kinases have no effect on KHOS cell growth and survival; several kinases only demonstrated inhibited cell growth in 1 out of 4 or 1 out of 5 shRNA target sites (Fig. 1C). To further reduce the risk of off-target effects and decrease the chance of analyzing false positives, we only focused on the kinases that had two or more shRNAs targeting the same gene showing similar effects. The absorbance values from each well on the plate were determined by CellTiter 96ÒAQueous One Solution Cell Cytotoxicity Assay. For example, the OD values from the media control wells were always around 0.1–0.2 while the OD values for empty vector and non-target shRNA controls were always around 0.8–1 in KHOS cells. Any well showing the value of absorbance to be below 0.5 after a specific kinase knockdown was treated as significant.

225

3.4. PLK1 is highly expressed in sarcoma cell lines and in osteosarcoma tissues High levels of PLK1 expression were visible in all four osteosarcoma cell lines, including KHOS, U-2OS, Saos and OSA344. However, normal human osteoblast cell lines HOB-c, NHOst and hFOB expressed very low levels of PLK1 (Fig. 3A). To preclude the possibility that PLK1 expression is an artifact induced by in vitro propagation, we also examined six freshly isolated primary osteosarcoma specimens. We observed expression of PLK1 in all six of the osteosarcoma patient samples (Fig. 3B).

3.5. PLK1 knockdown decreases cell proliferation and induces apoptosis in multiple osteosarcoma cells To determine whether PLK1 shRNA knockdown induced growth inhibition and apoptosis in KHOS is a general phenomenon in osteosarcoma cells, we extended our studies to additional osteosarcoma cell lines. The additional cell lines included U-2OS, Saos and a primary osteosarcoma in culture, OSA344. To confirm that PLK1 shRNA is specific to malignant osteoblasts, we also examined the effect of PLK1 siRNA on a benign human osteoblast cell line. We found that inhibition with PLK1 shRNA sig-

3.2. Identification of PLK1 as a new regulator of osteosarcoma cell survival and apoptosis We performed a screen that targeted 673 kinase genes by lentiviral shRNA and determined its effects on osteosarcoma cell survival and apoptosis. After completing the KHOS cell screen, eight kinases, when knocked down, displayed inhibitory growth effects. We further validated these shRNA clones targeting the kinase hits in a secondary osteosarcoma cell line, U-2OS. Finally, four kinases were validated when it was shown that their individual knockdowns could significantly decrease osteosarcoma cell proliferation and induce apoptosis. These kinases included PLK1, DYRK1b, ROCK1 and PITSLRE. Because PLK1 has been reported plays a critical role in several types of human cancers and the importance of PLK1 in the osteosarcoma has not been studied, we decided to further investigate the function role of PLK1 in the osteosarcoma. Lentiviral shRNA directed against PLK1 in KHOS cells significantly reduced the cell proliferation and eventually led to cell death (Fig. 2A).

3.3. PLK1 cDNA rescue results To confirm the above observation of PLK1, PLK1 lentiviral shRNA clones against the 30 UTR was utilized in a cDNA rescue assay. The results showed that pre-transfection of full length PLK1 cDNA constructs lacking the 30 UTR could reverse the cell death phenotype when transduced with the PLK1 shRNA which targets the 30 UTR region (Fig. 2B).

Fig. 4. Synthetic siRNA targeting PLK1 decreases cell proliferation and induces apoptosis in KHOS cells. (A) KHOS cells were transfected with PLK1 siRNA or non-specific siRNA, and cell proliferation post-transfection was determined by CellTiter 96ÒAQueous One Solution Reagent. (B) Apoptosis was evaluated using the M30-Apoptosense ELISA assay kit. (C) Confirmation of PLK1 protein knockdown by Western blot.

226

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

nificantly decreased cell proliferation in several osteosarcoma cell lines including KHOS, U-2OS, Saos, and OSA344, but not in the benign osteoblast cell line, HOB-c (Fig. 3C). 3.6. Confirmation that PLK1 knockdown induces apoptosis by synthetic siRNA To further confirm the shRNA knockdown induced phenotype, we decreased PLK1 expression using synthetic siRNA. This siRNA oligonucleotide has been validated in other cell lines for inhibiting PLK1 expression. After the synthetic PLK1 siRNA or non-specific siRNA oligonucleotides were transfected into KHOS, we measured proliferation of osteosarcoma cells. Although transfection with non-specific siRNA did not affect the growth rate of the cells, transfection with PLK1 siRNA significantly inhibited cell proliferation (Fig. 4A). We also examined apoptosis induced by PLK1 inhibition. It was observed that transfection with nonspecific siRNA did not show any effect on apoptosis, whereas PLK1 siRNA induced strong apoptosis (Fig. 4B). Western blot confirmed that after transfection with PLK1 siRNA, PLK1 protein was efficiently depleted (Fig. 4C). These results are thus consistent with the PLK1 shRNA knockdown results. 3.7. PLK1 gene expression and the rate of survival in patients with osteosarcoma We analyzed PLK1 expression by immunohistochemistry using an osteosarcoma tissue microarray. PLK1 was detectable in all osteosarcoma tissues with its expression ranging from weak (1+) to moderate (2+) to strong (3+). Of the 57 patients with osteosarcoma, 21 (36.8%) were classified as weak, 25 (43.8%) were classified as moderate, and 11 samples

(19.3%) contained strong expression levels of PLK1, according to the criteria described under Section 2 (Fig. 5B). Kaplan–Meier analysis of 57 cases showed that the overall survival rates of patients with weak PLK1 expression was significantly higher than those with moderate to strong PLK1 staining (P = 0.008) (Fig. 5A).

3.8. Reduction of PLK1 expression by the PLK1 inhibitor, scytonemin, induces apoptosis in osteosarcoma cell lines The high frequency of PLK1 overexpression in osteosarcoma strongly suggests that PLK1 may play an important role in maintaining the tumorogenic phenotype of osteosarcoma cells. The shRNA data also suggests that drugs targeting PLK1 may induce growth arrest and apoptosis with some degree of cancer cell selectivity. To test this hypothesis, we examined the effect of scytonemin, a small molecule inhibitor of PLK1, to furthermore assess the consequences of PLK1 inhibition in osteosarcoma cells. After exposing the cell lines to scytonemin for 72 h, the relative number of viable cells was determined by CellTiter 96ÒAQueous One Solution Cell Cytotoxicity Assay. The assay showed that osteosarcoma cell lines, KHOS and U-2OS, exhibited a decline in growth rate when treated with scytonemin (Fig. 6A and B). Incubation of these osteosarcoma cell lines with increasing concentrations of scytonemin was found to decrease proliferation in a dose-dependent manner. Effects of scytonemin on the expression of PLK1 and PLK1 specific downstream target phosphorcdc25c (pCdc25c) in KHOS and U-2OS were analyzed by Western blot. The KHOS cells were incubated with a range of concentrations of scytonemin for 24 h. We observed that scytonemin reduced PLK1 as well as pCdc25c proteins expression in a dose-dependent manner (Fig. 6C), even with concentration levels as low as 0.1 lM, which was consistent with

Fig. 5. Overexpression of PLK1 is associated with poor osteosarcoma survival. (A) Kaplan–Meier survival curve of osteosarcoma patients are subgrouped as either PLK1 low expression group (PLK1 staining = 1+) or PLK1 high expression group (PLK1 staining P 2+). The prognosis of PLK1 high expression group was significantly shorter than that of PLK1 low expression group (p = 0.008). (B) Representative immunohistochemical staining of PLK1 expression levels in osteosarcoma tissues. Upper panels: PLK1 expression in osteosarcoma. PLK1 staining intensity was graded into four groups: no staining (0), weak staining (1+), moderate staining (2+) and intense staining (3+). Lower panels: HE staining.

227

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

B

Absorbance (490nM)

Absorbance (490nM)

A

Scytonemin (µM)

C Scytonemin (µM)

Scytonemin (µM)

D Scytonemin (µM)

PLK1

PLK1

pCdc25c

pCdc25c

Actin

Actin

Fig. 6. Effects of PLK1 inhibitor scytonemin in osteosarcoma cell lines. (A and B) PLK1 inhibitor, scytonemin, inhibits osteosarcoma cell proliferation and induces apoptosis. Cells were treated with the scytonemin at the indicated concentrations. The relative sensitivity of each line to scytonemin was determined by CellTiter 96ÒAQueous One Solution Reagent 72 h post-treatment. (C and D) Dose dependent concentrations of scytonemin down-regulated PLK1 protein in osteosarcoma cells. KHOS (C) or U-2OS (D) was treated with scytonemin in for 24 h. Total cellular proteins were subjected to immunoblotting with specific antibody to PLK1 and b-actin as described in Section 2.

the proliferation data. Similar results were also found in U-2OS cell lines (Fig. 6D). Furthermore, we tested the effects of scytonemin on osteoblast cell lines and found that a higher resistance to scytonemin was observed in normal osteoblast cell lines as compared to osteosarcoma cells (Supplementary Fig. 2).

4. Discussion We performed a kinase shRNA screen and identified essential kinases based on their role in the growth and survival of osteosarcoma cells. This study provides a genomic view on the role of kinases in the regulation of osteosarcoma cell survival and death. Our study identifies several kinases whose loss-of-function resulted in marked cell death, and therefore may indicate an unrecognized role of kinases in osteosarcoma. By targeting PLK1 with shRNA and siRNA, osteosarcoma cell growth was inhibited, inducing apoptosis among various cell lines. In addition, a small molecular inhibitor for PLK1 allowed us to further characterize PLK1 roles in osteosarcoma cells. PLK1 is a serine/threonine kinase that functions to regulate many stages of mitosis and to maintain genomic stability [24]. Other studies have observed overexpression of PLK1 in several human cancers [24]. Many of these studies demonstrate that PLK1 overexpression correlates with tumor progression and patient survival in a variety of cancers [24–29]. Overexpression of PLK1 in NIH3T3 fibroblasts transformed cells into an oncogenic foci in soft agar and more importantly, and leads to tumor formation when injected into nude mice [30]. Since PLK1 is considered as a

‘‘proto-oncogene,” inhibition of PLK1 could be an effective treatment for cancers. Several strategies for inhibition or depletion of PLK1 activity have been tested in cancer therapeutic trials [24,31–34]. For example, intravesical administration of PLK1 siRNA suppressed bladder cancer growth in an orthotropic bladder cancer mouse model [35]. Normal hTERT-RPE1 and MCF10A cells, but not cancer Hela cells survive siRNA based PLK1 depletion [36]. In another model, plasmid-based U6 promoter-driven shRNA for PLK1 has also been effective in suppressing the growth of HeLa S3 xenografts [37]. These data suggest that inhibition of PLK1 activity is crucial for the observed growth inhibition and apoptosis. Here, we also show that knockdown of PLK1 in benign osteoblast cells was not effective at inducing growth arrest, implying that PLK1 has a unique function in promoting the integrity and proliferation of malignant osteoblast cells. The mechanisms of PLK1 signaling still remain incompletely described, but evidence support an important role of PLK1 in tumor cells, suggesting that PLK1 could be pursued as a novel strategy in the treatment of osteosarcoma. The results generated herein extend our understanding of PLK1 importance to osteosarcomas, making PLK1 a good target for further study. First, we show that osteosarcoma cells have a high level of PLK1 expression when compared to normal osteoblast cells. Secondly, our data is consistent with previous characterizations of PLK1, where expression of PLK1 was found to be below the limit of detection in most adult normal tissues [36,38,39]. Thirdly, we show that expression of PLK1 is strongly correlated with clinical

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

228

prognosis in osteosarcoma patients. It has been reported that the level of PLK1 transcripts in tumor samples could be linked to the prognosis of patients with non-small-cell lung cancer [40]. Several follow-up studies showed that quantification of PLK1 expression levels has a prognostic value for a variety of cancers, including breast cancer, colon cancer, melanoma and hepatoblastoma [41–44]. PLK1 expression has been reported is a reliable marker for identifying high risk of metastatic patients in melanoma, breast cancer and thyroid cancer [45–47]. The correlation between PLK1 expression with clinical stage and histological grade of a tumor has the potential to aid clinicians in their search for improving treatment decisions for different cancer patients. PLK1 inhibition by shRNA, siRNA, and small molecule inhibitor, scytonemin, can significantly constrain osteosarcoma cell growth and induce apoptosis. Interestingly, while this manuscript was in the publication process, a siRNA library screening in osteosarcoma cell line Saos has been reported [48]. This study also showed siRNA targeting PLK1 significantly induced cell death in osteosarcoma cell lines [48]. In conclusion, we propose that PLK1 can be a potential target for the treatment of osteosarcoma. As PLK1 kinase inhibitors displayed the same phenotype as shRNA knockdown, efforts to develop kinase inhibitors of PLK1 with enhanced potency and selectivity would be valuable. Conflicts of interest None declared. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.canlet. 2010.01.014. References [1] A.J. Chou, D.S. Geller, R. Gorlick, Therapy for osteosarcoma: where do we go from here?, Paediatr Drugs 10 (2008) 315–327. [2] J.C. Trent, Rapid evolution of the biology and treatment of sarcoma, Curr. Opin. Oncol. 20 (2008) 393–394. [3] H.J. Siegel, J.G. Pressey, Current concepts on the surgical and medical management of osteosarcoma, Expert Rev. Anticancer Ther. 8 (2008) 1257–1269. [4] A.J. Chou, R. Gorlick, Chemotherapy resistance in osteosarcoma: current challenges and future directions, Expert Rev. Anticancer Ther. 6 (2006) 1075–1085. [5] J.P. MacKeigan, L.O. Murphy, J. Blenis, Sensitized RNAi screen of human kinases and phosphatases identifies new regulators of apoptosis and chemoresistance, Nat. Cell Biol. 7 (2005) 591–600. [6] J.S. Sebolt-Leopold, J.M. English, Mechanisms of drug inhibition of signalling molecules, Nature 441 (2006) 457–462. [7] J. Du, P. Bernasconi, K.R. Clauser, D.R. Mani, S.P. Finn, R. Beroukhim, M. Burns, B. Julian, X.P. Peng, H. Hieronymus, R.L. Maglathlin, T.A. Lewis, L.M. Liau, P. Nghiemphu, I.K. Mellinghoff, D.N. Louis, M. Loda, S.A. Carr, A.L. Kung, T.R. Golub, Bead-based profiling of tyrosine kinase phosphorylation identifies SRC as a potential target for glioblastoma therapy, Nat. Biotechnol. (2008). [8] K. Scotlandi, P. Picci, Targeting insulin-like growth factor 1 receptor in sarcomas, Curr. Opin. Oncol. 20 (2008) 419–427. [9] L. Cao, Y. Yu, I. Darko, D. Currier, L.H. Mayeenuddin, X. Wan, C. Khanna, L.J. Helman, Addiction to elevated insulin-like growth factor I receptor and initial modulation of the AKT pathway define the responsiveness of rhabdomyosarcoma to the targeting antibody, Cancer Res. 68 (2008) 8039–8048.

[10] X. Wan, L.J. Helman, The biology behind mTOR inhibition in sarcoma, Oncologist 12 (2007) 1007–1018. [11] M.M. Mita, A.W. Tolcher, The role of mTOR inhibitors for treatment of sarcomas, Curr. Oncol. Rep. 9 (2007) 316–322. [12] E. Taniguchi, K. Nishijo, A.T. McCleish, J.E. Michalek, M.H. Grayson, A.J. Infante, H.E. Abboud, R.D. Legallo, S.J. Qualman, B.P. Rubin, C. Keller, PDGFR-A is a therapeutic target in alveolar rhabdomyosarcoma, Oncogene 27 (2008) 6550–6560. [13] J.A. George, T. Chen, C.C. Taylor, SRC tyrosine kinase and multidrug resistance protein-1 inhibitions act independently but cooperatively to restore paclitaxel sensitivity to paclitaxel resistant ovarian cancer cells, Cancer Res. 65 (2005) 10381–10388. [14] D. LeRoith, L. Helman, The new kid on the block(ade) of the IGF-1 receptor, Cancer Cell 5 (2004) 201–202. [15] L.J. Helman, P. Meltzer, Mechanisms of sarcoma development, Nat. Rev. Cancer 3 (2003) 685–694. [16] J.M. Bosher, M. Labouesse, RNA interference: genetic wand and genetic watchdog, Nat. Cell Biol. 2 (2000) E31–E36. [17] Z. Duan, K.A. Brakora, M.V. Seiden, Inhibition of ABCB1 (MDR1) and ABCB4 (MDR3) expression by small interfering RNA and reversal of paclitaxel resistance in human ovarian cancer cells, Mol. Cancer Ther. 3 (2004) 833–838. [18] B.R. Cullen, Induction of stable RNA interference in mammalian cells, Gene. Ther. 13 (2006) 503–508. [19] J. Moffat, D.A. Grueneberg, X. Yang, S.Y. Kim, A.M. Kloepfer, G. Hinkle, B. Piqani, T.M. Eisenhaure, B. Luo, J.K. Grenier, A.E. Carpenter, S.Y. Foo, S.A. Stewart, B.R. Stockwell, N. Hacohen, W.C. Hahn, E.S. Lander, D.M. Sabatini, D.E. Root, A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen, Cell 124 (2006) 1283–1298. [20] C.T. Dann, New technology for an old favorite: lentiviral transgenesis and RNAi in rats, Transgenic Res. 16 (2007) 571–580. [21] M. Lourda, I.P. Trougakos, E.S. Gonos, Development of resistance to chemotherapeutic drugs in human osteosarcoma cell lines largely depends on up-regulation of clusterin/apolipoprotein J, Int. J. Cancer 120 (2007) 611–622. [22] C.S. Stevenson, E.A. Capper, A.K. Roshak, B. Marquez, C. Eichman, J.R. Jackson, M. Mattern, W.H. Gerwick, R.S. Jacobs, L.A. Marshall, The identification and characterization of the marine natural product scytonemin as a novel antiproliferative pharmacophore, J. Pharmacol. Exp. Ther. 303 (2002) 858–866. [23] Z. Duan, E.J. Weinstein, D. Ji, R.Y. Ames, E. Choy, H. Mankin, F.J. Hornicek, Lentiviral short hairpin RNA screen of genes associated with multidrug resistance identifies PRP-4 as a new regulator of chemoresistance in human ovarian cancer, Mol. Cancer Ther. 7 (2008) 2377–2385. [24] K. Strebhardt, A. Ullrich, Targeting polo-like kinase 1 for cancer therapy, Nat. Rev. Cancer 6 (2006) 321–330. [25] T.C. Nappi, P. Salerno, H. Zitzelsberger, F. Carlomagno, G. Salvatore, M. Santoro, Identification of polo-like kinase 1 as a potential therapeutic target in anaplastic thyroid carcinoma, Cancer Res. 69 (2009) 1916–1923. [26] W.S. Yang, B.R. Stockwell, Inhibition of casein kinase 1-epsilon induces cancer-cell-selective, PERIOD2-dependent growth arrest, Genome Biol. 9 (2008) R92. [27] R. Guan, P. Tapang, J.D. Leverson, D. Albert, V.L. Giranda, Y. Luo, Small interfering RNA mediated polo-like kinase 1 depletion preferentially reduces the survival of p53-defective, oncogenic transformed cells and inhibits tumor growth in animals, Cancer Res. 65 (2005) 2698– 2704. [28] A. Rizki, J.D. Mott, M.J. Bissell, Polo-like kinase 1 is involved in invasion through extracellular matrix, Cancer Res. 67 (2007) 11106– 11110. [29] B. Spankuch, S. Heim, E. Kurunci-Csacsko, C. Lindenau, J. Yuan, M. Kaufmann, K. Strebhardt, Down-regulation of polo-like kinase 1 elevates drug sensitivity of breast cancer cells in vitro and in vivo, Cancer Res. 66 (2006) 5836–5846. [30] M.R. Smith, M.L. Wilson, R. Hamanaka, D. Chase, H. Kung, D.L. Longo, D.K. Ferris, Malignant transformation of mammalian cells initiated by constitutive expression of the polo-like kinase, Biochem. Biophys. Res. Commun. 234 (1997) 397–405. [31] H.A. Lane, E.A. Nigg, Antibody microinjection reveals an essential role for human polo-like kinase 1 (Plk1) in the functional maturation of mitotic centrosomes, J. Cell Biol. 135 (1996) 1701–1713. [32] J.P. Cogswell, C.E. Brown, J.E. Bisi, S.D. Neill, Dominant-negative pololike kinase 1 induces mitotic catastrophe independent of cdc25C function, Cell Growth Differ. 11 (2000) 615–623. [33] X.Y. Liu, T. Zhou, K. Chang, L. Wang, M.X. Zheng, Y.W. Luo, Study on intermolecular interaction between 4-aminopyridine and

Z. Duan et al. / Cancer Letters 293 (2010) 220–229

[34]

[35]

[36] [37]

[38]

[39]

[40]

[41]

methacrylic acid using two dimensional FTIR spectroscopy, Guang Pu Xue Yu Guang Pu Fen Xi 28 (2008) 2073–2076. X. Liu, R.L. Erikson, Polo-like kinase (Plk)1 depletion induces apoptosis in cancer cells, Proc. Natl. Acad. Sci. USA 100 (2003) 5789–5794. M. Nogawa, T. Yuasa, S. Kimura, M. Tanaka, J. Kuroda, K. Sato, A. Yokota, H. Segawa, Y. Toda, S. Kageyama, T. Yoshiki, Y. Okada, T. Maekawa, Intravesical administration of small interfering RNA targeting PLK-1 successfully prevents the growth of bladder cancer, J. Clin. Invest. 115 (2005) 978–985. X. Liu, M. Lei, R.L. Erikson, Normal cells, but not cancer cells, survive severe Plk1 depletion, Mol. Cell Biol. 26 (2006) 2093–2108. B. Spankuch, Y. Matthess, R. Knecht, B. Zimmer, M. Kaufmann, K. Strebhardt, Cancer inhibition in nude mice after systemic application of U6 promoter-driven short hairpin RNAs against PLK1, J. Natl. Cancer Inst. 96 (2004) 862–872. U. Holtrich, G. Wolf, A. Brauninger, T. Karn, B. Bohme, H. RubsamenWaigmann, K. Strebhardt, Induction and down-regulation of PLK, a human serine/threonine kinase expressed in proliferating cells and tumors, Proc. Natl. Acad. Sci. USA 91 (1994) 1736–1740. R. Hamanaka, S. Maloid, M.R. Smith, C.D. O’Connell, D.L. Longo, D.K. Ferris, Cloning and characterization of human and murine homologues of the Drosophila polo serine-threonine kinase, Cell Growth Differ. 5 (1994) 249–257. G. Wolf, R. Elez, A. Doermer, U. Holtrich, H. Ackermann, H.J. Stutte, H.M. Altmannsberger, H. Rubsamen-Waigmann, K. Strebhardt, Prognostic significance of polo-like kinase (PLK) expression in nonsmall cell lung cancer, Oncogene 14 (1997) 543–549. W. Weichert, G. Kristiansen, K.J. Winzer, M. Schmidt, V. Gekeler, A. Noske, B.M. Muller, S. Niesporek, M. Dietel, C. Denkert, Polo-like kinase isoforms in breast cancer: expression patterns and prognostic implications, Virchows Arch. 446 (2005) 442–450.

229

[42] W. Weichert, G. Kristiansen, M. Schmidt, V. Gekeler, A. Noske, S. Niesporek, M. Dietel, C. Denkert, Polo-like kinase 1 expression is a prognostic factor in human colon cancer, World J. Gastroenterol. 11 (2005) 5644–5650. [43] K. Strebhardt, L. Kneisel, C. Linhart, A. Bernd, R. Kaufmann, Prognostic value of polo-like kinase expression in melanomas, JAMA 283 (2000) 479–480. [44] S. Yamada, M. Ohira, H. Horie, K. Ando, H. Takayasu, Y. Suzuki, S. Sugano, T. Hirata, T. Goto, T. Matsunaga, E. Hiyama, Y. Hayashi, H. Ando, S. Suita, M. Kaneko, F. Sasaki, K. Hashizume, N. Ohnuma, A. Nakagawara, Expression profiling and differential screening between hepatoblastomas and the corresponding normal livers: identification of high expression of the PLK1 oncogene as a poorprognostic indicator of hepatoblastomas, Oncogene 23 (2004) 5901– 5911. [45] L. Kneisel, K. Strebhardt, A. Bernd, M. Wolter, A. Binder, R. Kaufmann, Expression of polo-like kinase (PLK1) in thin melanomas: a novel marker of metastatic disease, J. Cutan. Pathol. 29 (2002) 354–358. [46] A. Ahr, T. Karn, C. Solbach, T. Seiter, K. Strebhardt, U. Holtrich, M. Kaufmann, Identification of high risk breast-cancer patients by gene expression profiling, Lancet 359 (2002) 131–132. [47] Y. Ito, Y. Nakamura, H. Yoshida, C. Tomoda, T. Uruno, Y. Takamura, A. Miya, K. Kobayashi, F. Matsuzuka, K. Kuma, K. Kakudo, A. Miyauchi, Polo-like kinase 1 expression in medullary carcinoma of the thyroid: its relationship with clinicopathological features, Pathobiology 72 (2005) 186–190. [48] U. Yamaguchi, K. Honda, R. Satow, E. Kobayashi, R. Nakayama, H. Ichikawa, A. Shoji, M. Shitashige, M. Masuda, A. Kawai, H. Chuman, Y. Iwamoto, S. Hirohashi, T. Yamada, Functional genome screen for therapeutic targets of osteosarcoma, Cancer Sci. (2009).