CYT387, a potent IKBKE inhibitor, suppresses human glioblastoma progression by activating the Hippo pathway

The human glioblastoma cell lines U251 and LN229 came from the Institute of Biochemistry and Cell Biology (Shanghai, China). HEK293 cells were from the Institute of Biochemistry and Cell Biology (Shanghai, China). All cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA), and cultured at 37 °C in 5% CO₂. We established an IKBKE-shRNA lentiviral vector from GeneChem (Shanghai, China) with the sequence of 5ʹ-GCATCATCGAACGGCTAAATA-3ʹ. A GFP scrambled lentiviral vector with the sequence of 5ʹ-TTCTCCGAACGTGTCACGTTTC-3ʹ was used as the negative control. The shRNAs were transfected according to the manufacturer’s instructions. The Flag-IKBKE plasmid was purchased from Addgene (USA). The HA-TEAD2 and HA-YAP1 plasmids were from Hanbio Biotechnology (Shanghai, China). The IKBKE-overexpress lentivirus and its empty vector virus were from GeneChem (Shanghai, China). We selected Lipofectamine 3000 (thermo fisher scientific, USA) as transfection medium and the process of transfection was according to the manufacturer’s instructions. IKBKE rabbit mAb (No.2905,WB 1:1000; IP 1:100), c-myc rabbit mAb (No.13987,WB 1:1000), MMP9 rabbit mAb (No.13667,WB 1:1000), YAP1 mouse mAb (No.12395,WB 1:1000; IHC 1:400; IP 1:100), Bcl-2 rabbit mAb (No.2870,WB 1:1000), Phospho-YAP (Ser127) rabbit mAb (No.13008,WB 1:1000; IHC 1:2000), HA-Tag rabbit mAb (No.3724,WB 1:1000; IP 1:50), DYKDDDDK-Tag (Flag) rabbit mAb (No.14793,WB 1:1000; IP 1:50) and Axl rabbit mAb (No.8661,WB 1:1000;IHC 1:300) were purchased from Cell Signaling Technology (USA). IKBKE rabbit polyclonal antibody (ab7891,IHC 1:100), Cdk1 rabbit mAb (ab133327,WB 1:10000;IHC 1:300), Cdc25c rabbit mAb (ab32444,WB 1:2000), caspase-9 rabbit mAb (ab202068,WB 1:2000), Bax rabbit mAb (ab32503,WB 1:2000), CyclinB₁ rabbit mAb (ab32053,WB 1:5000;IHC 1:100), CyclinA₂ rabbit mAb (ab181591,WB 1:2000), c-myc rabbit mAb (ab32072,IHC 1:500), CyclinD₁ rabbit mAb (ab134175,WB:1:10000), TEAD2 rabbit polyclonal antibody (ab83670, WB 1:500) were purchased from Abcam (USA). TEAD2 rabbit polyclonal antibody (sc-67115, IP 1:50) was from Santa Cruz (USA). LATS2 rabbit polyclonal antibody (20276-1-AP, WB 1:500; IHC 1:50) were purchased from proteintech (USA). GAPDH mouse mAb (WB 1:2000) was from ZSGB-Bio (Beijing, China).

The cell total protein was extracted after treatment of CYT387, plasmids or IKBKE-shRNA using RIPA lysis buffer with protease and phosphatase inhibitor (MCE USA). The homogenates were clarified by centrifugation at 4 °C for 15 min at 12,000 rpm after cleavage by RIPA for 15 min, and the protein concentration was measured by BCA assay kit (Beyotime, Shanghai, China). At least 20 μg protein mixed with 4 × loading buffer was added into spacer gel and then separated by sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The protein bands were electrotransferred to PVDF membranes (Millipore, USA). Primary antibodies were incubated at 4 °C for overnight then HRP-conjugated secondary antibody (1:3000 dilution, ZSGB-Bio, Beijing, China) was used for 1 h at room temperature. The bands were detected by the G:BOX (Syngene Company, UK) using Chemiluminescent HRP Substrate (Millipore USA).

Total RNA of glioblastoma cells (U87-MG and LN-229) after treatment with CYT387 in dose- and time-dependent manners was extracted by TRIzol reagent (Invitrogen, USA) following the manufacturer’s protocols and then reverse transcription was performed using GoScript™ Reverse Transcription System (Promega, USA). The quantitative real-time PCR was finished by GoTaq qPCR Master Mix (Promega, USA) according to the supplier’s instructions. The reaction conditions were as follows: 95 °C for 5 min and 40 cycles of 95 °C for 12 s and 60 °C for 40 s. The primers were synthesized by GENEWIZ (USA). The sequences of the primers were as follows: GAPDH: 5ʹ-GGAGCGAGATCCCTCCAAAAT-3ʹ (Forward primer) and 5ʹ-GGCTGTTGTCATACTTCTCATGG-3ʹ (Reverse primer); IKBKE: 5ʹ-GAGAAGTTCGTCTCGGTCTATGG-3ʹ (Forward primer) and 5ʹ-TGCATGGTACAAGGTCACTCC-3ʹ (Reverse primer); TEAD2: 5ʹ-GCCTCCGAGAGCTATATGATCG-3ʹ (Forward primer) and 5ʹ-TCACTCCGTAGAAGCCACCA-3ʹ (Reverse primer); YAP1: 5ʹ-TAGCCCTGCGTAGCCAGTTA-3ʹ (Forward primer) and 5ʹ-TCATGCTTAGTCCACTGTCTGT-3ʹ (Reverse primer). GAPDH was used as internal control.

U87-MG and LN-229 cells were seed in six-well plates (2 × 10³/well) divided into three groups as blank control, negative control (DMSO) and drug group (CYT387 with concentration of 6 μM). Growth medium was changed every 6 days. After 12 days, cells were fixed in 4% paraformaldehyde for 15 min and stained with crystal violet for 30 min. Colonies were scored after photographed.

For IC₅₀ measurement, we seeded U87-MG and LN-229 cells (5000/50 ul/well) into 96-well plates on the first day. After cells were adherent, we again added 50ul/well medium with different concentration of CYT387 to achieve the final drug concentration gradient of 0.5 μM, 1 μM, 2 μM, 4 μM, 8 μM, 16 μM, 32 μM and 64 μM. After treatment with CYT387 for 24, 48, and 72 h, 10 μl CCK8 reagent (dojindo, Japan) was mixed into each well and then 96-well plate was incubated for 2 h at 37˚C. The O.D. value was measured by Microplate reader at the wavelength of 450 nm.

For proliferative curve measurement, glioblastoma cells with normal medium, DMSO medium and 6 μM CYT387 medium were seeded (2000/100 µl/well) into 96-well plates. From first day to fifth day, 10 µl CCK-8 reagent (dojindo, Japan) was added into each well. The O.D. value was measured after incubated for 2 h at 37 °C in a 5% CO₂ atmosphere.

U87-MG and LN-229 cells were seeded in six-well plate and a straight wound was created with a sterile 100 µl pipette tip. Then we respectively added DMSO and CYT387 in DMSO group and drug group, making the drug concentration of 6 μM. After treatment for 24 h and 48 h, the wound healing area was detected by an inverted microscope.

Before experiment begun, matrigel with 3 times volume serum-free DMEM (total 80 µl/well) was coated on the upper surface of chamber. Then place it at 37 °C for 30 min, waiting for matrigel solidification. Then the cells (5 × 10⁴/well) were seeded into the transwell chambers with 200 μl serum-free DMEM while outer space was filled with 500 μl serum DMEM. After incubated at 37 °C in a 5% CO₂ atmosphere for 48 h, cells across the chamber membrane was fixed with 4% paraformaldehyde for 15 min, then stained with crystal violet for 5 min, counted and imaged under the microscope.

Before cell apoptosis assay, the cells were treated with CYT387 with concentration of 6 μM for 24 h. Cells were trypsinized without EDTA, washed with PBS twice and then stained using the Annexin V-FITC Apoptosis Detection kit from KeyGen Biotech (Nanjing, China) according to the manufacturer’s instructions. Flow cytometry analysis was finished by a FACS flow cytometer (Becton–Dickinson). Data were analyzed by CellQuest software. Before cell cycle analysis, the U87-MG and LN-229 cells was treated with CYT387 at the concentration of 6 μM for 72 h. Then the following protocols were performed as the previous article [9].

Co- immunoprecipitation (co-IP) was carried out as described previously [28].

Before animal studies, CYT387 purchased from selleck (USA) was dissolved in NMP (1-methyl-2-pyrrolidinone) to finally get the concentration of 120 mg/ml. Next, the CYT387 was diluted with 0.14 M Captisol (MCE, USA) to a concentration of 6 mg/ml. The tumor subcutaneous experiments method was carried out as described previously [9]. After subcutaneous tumour was shaped, the nude mice were fed with CYT387 (100 mg/kg/day).

For intracranial orthotopic model, U87-MG transfected with luciferase-expressing lentivirus was injected intracranially into 6-week-old BALB/c-nu mice. After 7 days, mice started to be fed with CYT387 (100 mg/kg/day). The weight of mice was monitored every 2 days and the luminescence imaging of intracranial tumour was measured using an IVIS Lumina Imaging System (Xenogen) every 7 days.

After mice were sacrificed for subcutaneous and intracranial tumour, we made specimens embedded with paraffin. Then the paraffin-embedded tumours were sectioned and dewaxed. After antigen retrieval using 10 mmol/l citrate buffer, sections were incubated with 3% H₂O₂ and blocked with 5% BSA. Then the sections were added with primary antibodies at 4 °C overnight. After rewarming at room temperature the next day, the sections were incubated with secondary antibodies using two-step polymer HRP detection system (ZSGB-BIO, Beijing, China). The samples were colourated with DBA Kit (ZSGB-BIO, Beijing, China) and then counterstained with haematoxylin. After dehydration and sealing piece with neutral gum, the samples were detected and photographed by microscope (Olympus Japan).

All data were repeated at least three times. Quantitative data are shown as the mean ± standard deviation (SD). We used SPSS software (version 16.0) for the statistical analyses and P < 0.05 was considered statistically significant.

To evaluate the sensitivity of glioblastoma cells to CYT387 in vitro, two glioblastoma cell lines that highly express IKBKE, U87-MG and LN-229 [10], were treated with a concentration gradient of CYT387 (0.5 μM to 64 μM) to measure the half maximal inhibitory concentration (IC50) with a CCK-8 assay. A dose–response curve with data points for 24 h, 48 h, and 72 h of drug treatment is shown in Fig. 1a. The IC50 values of U87-MG cells were 6.395 ± 1.127 μM for 72 h, 17.68 ± 2.94 μM for 48 h, and 24.87 ± 3.63 μM for 24 h, and those of LN229 cells were 5.139 ± 0.501 μM for 72 h, 17.18 ± 2.61 μM for 48 h, and 25.46 ± 3.59 μM for 24 h. These data demonstrated that glioblastoma cells were sensitive to CYT387, especially after 72 h of treatment. Then, we carried out a CCK-8 assay to investigate the effect of CYT387 on glioblastoma cell proliferation using a drug concentration of 5 μM and incubating the cells for 5 days. The results (Fig. 1b) showed that the viability of U87-MG and LN229 cells treated with CYT387 was dramatically decreased compared with that of the cells in the blank control and DMSO (drug solvent) groups. Additionally, a colony formation assay was used to test whether CYT387 affects the ability of the two cell lines to form colonies over a period of 12 days. Figure 1c demonstrates that compared with those in the blank control and solvent groups, the cells treated with the drug at a concentration of 6 μM had a markedly decreased number of colonies, and the colonies in the groups treated with the drug were much smaller than those in the blank control and DMSO groups (P < 0.001). These data showed that CYT387 could significantly inhibit glioblastoma cell proliferation in vitro.

To explore whether CYT387 impacts tumor cell migration, a wound healing assay was adopted to assess the adherent tumor cell healing area at different times with drug treatment at a concentration of 5 μM. Compared with control or solvent treatment, CYT387 markedly inhibited the wound healing speed after only 24 h, suggesting a poor migratory ability in the CYT387 group (Fig. 2a). Furthermore, we investigated the influence of CYT387 on the glioblastoma cell invasive capacity using a Transwell assay. The experimental results (Fig. 2b) showed that the average number of invaded cells in the group treated with a drug concentration of 6 μM was significantly decreased compared to that in the blank control and solvent groups after 48 h, showing that CYT387 had an inhibitory effect on the invasive ability of U87-MG and LN229 cells. To further research the detailed mechanism, two important matrix metalloproteinases (MMPs), MMP2 and MMP9, were assessed by western blot analysis. As Fig. 2c shows, MMP2 and MMP9 expression levels were significantly reduced after tumor cells were treated with CYT387 for 48 h compared with control or DMSO treatment (Figure S2). These data fully demonstrated that CYT387 could effectively inhibit glioblastoma cell migration and invasion.

Since apoptosis is critical to tumor regression, we performed Annexin V-FITC/PI staining using flow cytometry to test whether CYT387 can influence glioblastoma cell apoptosis. After cells were incubated with the drug for 24 h at a concentration of 5 μM, the tumor cell apoptosis rate was obviously increased compared with that observed in the blank control and DMSO groups (Fig. 3a, b). The apoptosis rates of U87-MG cells were 4.20 ± 0.127% in the blank group, 4.47 ± 0.287% in the DMSO group and 10.8 ± 1.20% in the drug group (P < 0.01), while the apoptosis rates of LN229 cells were 7.04 ± 0.176% in the blank group, 7.21 ± 0.138% in the DMSO group and 14.67 ± 0.960% in the drug group (P < 0.001). Moreover, the mechanism underlying the induction of apoptosis by CYT387 was assessed by western blot analysis. The protein expression of caspase-9 was decreased, but that of cleaved caspase-9 was increased in the groups treated with 5 μM drug for 24 h compared with the blank control and DMSO groups (Fig. 3c and Figure S2). Additionally, the antiapoptotic protein Bcl-2 and proapoptotic protein Bax were assessed. As shown in Fig. 3c, Bcl-2 expression was decreased, but Bax expression was increased after incubation with the drug (6 μM) for 24 h. All data indicated that CYT387 could obviously accelerate glioblastoma cell apoptosis.

We found that treatment with CYT387 (5 μM) for 3 days induced cell cycle arrest at the G2/M checkpoint. As shown in Fig. 4a, b, the proportions of cells at the G2/M checkpoint for U87-MG cells were 23.23 ± 0.54% in the drug group, 13.90 ± 0.50% in the blank control group and 14.47 ± 1.00% in the solvent group (P < 0.001), while those for LN229 cells were 37.03 ± 1.96% in the drug group, 15.11 ± 0.39% in the blank control group and 15.44 ± 0.79% in the solvent group (P < 0.001). In addition, the proportions of glioblastoma cells in the G1 and S phases were correspondingly decreased. In addition, we examined the expression levels of several cell cycle kinases including CyclinA2, CyclinD1, CyclinB1, Cdk1 and Cdc25c by western blot analysis. As shown in Fig. 4c, the expression of CyclinD1, an important regulator driving cell transition from the G1 phase into the S phase, exhibited a negligible change, while the levels of CyclinA2, CyclinB1, Cdk1 and Cdc25c, as several essential factors regulating the mitotic entry of cells from the G2/M checkpoint, were significantly decreased in the drug group (6 μM) compared to the blank control and DMSO groups (Figure S2). All this information suggested that CYT387 induced glioblastoma cell cycle arrest at the G2/M checkpoint to inhibit cell proliferation.

The Hippo pathway, which is composed of a cascade of phosphorylated kinases, can regulate cell proliferation, apoptosis and differentiation. The core Hippo pathway has been well established in mammals. MST1/2 directly phosphorylates LATS1/2 with the help of Sav1 and Mob1. Then, LATS1/2 directly interacts with and phosphorylates YAP1 at Ser127, resulting in YAP sequestration in the cytoplasm and degradation via ubiquitylation by 14-3-3 [24, 25]. When YAP1 is unphosphorylated, due to an ineffective Hippo pathway, it is transported into the nucleus to interact with TEAD1-4, creating transcription factors that induce the transcription of certain genes [29, 30], such as Axl [31], c-myc [32], Cyr61 [33, 34], and CTGF [34]. We showed that the levels of the core Hippo pathway effectors YAP1 and TEAD2 and their downstream factors including Axl or c-myc were decreased, while those of LATS2 and p-YAP1 (S127) were increased in the IKBKE-knockdown group compared with the blank control and scrambled groups by western blot analysis (Fig. 5a and Figure S3), indicating that IKBKE inhibition enhanced Hippo pathway activity by decreasing YAP1 and TEAD2 expression and increasing the expression of LATS2, which could accelerate the phosphorylation and degradation of YAP1. Additionally, overexpression of IKBKE increased YAP1 and TEAD2 expression, as well as the expression of the downstream factors Axl and c-myc, while reducing LATS2 and p-YAP1 (S127) expression (Fig. 5b and Figure S3). Furthermore, the effect of CYT387 on the Hippo pathway was confirmed to be mediated in a dose- and time-dependent manner by western blot analysis. As shown in Fig. 5c, with drug administration at a concentration of 6 µM for 0 h, 24 h, 48 h or 72 h, the expression of IKBKE, YAP1, TEAD2, Axl and c-myc gradually decreased over time; however, that of LATS2 and p-YAP1 (S127) progressively increased in U87-MG and LN229 cells, with an increase particularly noted at 72 h (Figure S3). Moreover, as shown in Fig. 5d, when increasing concentrations including 0 μM, 1 μM, 3 μM and 6 μM were administered for 72 h, the protein levels of IKBKE, YAP1, TEAD2, Axl and c-myc gradually decreased, while those of LATS2 and p-YAP1 (S127) increased, especially for the drug concentrations over 3 μM (Figure S3). However, it should be noted that the decrease in IKBKE expression resulting from CYT387 treatment was mainly due to posttranslational protein modification rather than effects at the transcriptional level, as IKBKE mRNA levels were determined to be negligibly changed by using real-time RT-PCR to evaluate dose- and time-dependent effects on U87-MG and LN229 cells (Additional file 1: Figure S1a, b). We next verified that CYT387 could reverse the inhibition of the Hippo pathway resulting from the overexpression of IKBKE. As shown in Fig. 5e, the expression levels of IKBKE, YAP1, TEAD2, Axl, and c-myc were increased with IKBKE overexpression and then decreased with 6 µM CYT387 treatment for 72 h. However, the expression trends for LATS-2 and p-YAP1 (S127) were contrary to those for YAP1 and TEAD2 (Figure S3). All of the above data demonstrated that CYT387 could enhance the activity of the Hippo pathway, which reduced YAP1, TEAD2 and downstream target protein expression, to inhibit glioblastoma progression by suppressing and inactivating IKBKE.

To investigate the detailed mechanism underlying the impact of IKBKE on the Hippo pathway, we first confirmed that IKBKE regulates YAP1 and TEAD2 at the posttranslational level. As shown in Fig. 6a, the mRNA expression of TEAD2 and YAP1 was hardly changed in the IKBKE-knockdown group compared to the blank control and scrambled groups, showing that inhibition of IKBKE decreased YAP1 and TEAD2 levels via posttranslational modification rather than changes in the mRNA levels. Next, we verified that IKBKE can directly interact with TEAD2 and YAP1 using endogenous co-IP (Fig. 6b and Figure S4). Then, we used Flag-IKBKE and HA-TEAD2 plasmids to perform exogenous co-IP. As shown in Fig. 6c, Flag-IKBKE could directly interact with HA-TEAD2 (Figure S4). We also performed exogenous co-IP using Flag-IKBKE and HA-YAP1 plasmids, showing that Flag-IKBKE interacted with HA-YAP1 (Fig. 6d and Figure S4). Additionally, we demonstrated that inhibition of IKBKE suppressed TEAD2 and YAP1 translocation into the nucleus using western blot analysis (Fig. 6e and Figure S4). All these data showed that IKBKE may directly interact with YAP1 and TEAD2 to promote YAP1 and TEAD2 transport into the nucleus. The detailed pathway diagram is shown in Fig. 6f.

To confirm whether CYT387 can inhibit tumor growth in vivo, we first established a subcutaneous nude mouse model via U87-MG cell inoculation. We used 6 mice fed the same concentration of solvent as the negative control (NC) group and 6 mice fed CYT387 at a dose of 100 mg/kg/day as the experimental group. Tumor volume was monitored every two days, and all mice were euthanized to measure implanted tumor weight on the 30th day. As time passed, the size of the tumors in the mice treated with CYT387 was obviously decreased (Fig. 7a, b) (P < 0.05), and the weight of these tumors was lower than that of the tumors in the NC group (Fig. 7c) (P < 0.05). Then, the expression of IKBKE, YAP1, Axl and c-myc was found to be downregulated, while that of LATS2 and p-YAP1 (S127) was shown to be increased in the tumors in the CYT387 (100 mg/kg/day) group compared with those in the NC group by immunohistochemical staining (Fig. 7d). Additionally, the levels of cell cycle markers such as Cdk1, CyclinA2 and CyclinB1 were reduced after mice were fed CYT387 compared with NC treatment (Fig. 7d). To explore whether CYT387 has an inhibitory effect on intracranial glioblastoma-like subcutaneously implanted tumors, we next established an orthotopic xenograft model with nude mice divided into two groups: an NC group fed the same concentration of solvent and a drug group fed CYT387 (100 mg/kg/day); U87-MG cells infected with a luciferase-expressing lentivirus were used. Imaging of intracranial tumor size was performed every 7 days after orthotopic xenotransplantation, and mouse weight was measured every 2 days. As shown in Fig. 7e, luminescence imaging showed no obvious differences between the NC group and the drug group at 7, 14, and 21 days. The weights and survival rates of the two groups of mice also showed few significant differences (Fig. 7f, g). Furthermore, we assessed IKBKE, LATS2, YAP1, and p-YAP1 (S127) expression levels by immunohistochemical staining, showing that there were few significant differences in these targets between the drug group and the NC group (Fig. 7h). These data suggested that CYT387 could inhibit glioblastoma progression in subcutaneous tumors but had little effect on intracranial orthotopic xenografts.