The role of PKN1 in glioma pathogenesis and the antiglioma effect of raloxifene targeting PKN1

Abstract PKN1 (protein kinase N1), a serine/threonine protein kinase family member, is associated with various cancers. However, the role of PKN1 in gliomas has rarely been studied. We suggest that PKN1 expression in glioma specimens is considerably upregulated and positively correlates with the histopathological grading of gliomas. Knocking down PKN1 expression in glioblastoma (GBM) cells inhibits GBM cell proliferation, invasion and migration and promotes apoptosis. In addition, yes‐associated protein (YAP) expression, an essential effector of the Hippo pathway contributing to the oncogenic role of gliomagenesis, was also downregulated. In contrast, PKN1 upregulation enhances the malignant characteristics of GBM cells and simultaneously upregulates YAP expression. Therefore, PKN1 is a promising therapeutic target for gliomas. Raloxifene (Ralo), a commonly used selective oestrogen‐receptor modulator to treat osteoporosis in postmenopausal women, was predicted to target PKN1 according to the bioinformatics team from the School of Mathematics, Tianjin Nankai University. We showed that Ralo effectively targets PKN1, inhibits GBM cells proliferation and migration and sensitizes GBM cells to the major chemotherapeutic drug, Temozolomide. Ralo also reverses the effect of PKN1 on YAP activation. Thus, we confirm that PKN1 contributes to the pathogenesis of gliomas and may be a potential target for Ralo adjuvant glioma therapy.


| INTRODUC TI ON
Glioma is the most common primary brain tumour, especially its subtype, glioblastoma multiforme (GBM), a highly malignant tumour with a very poor prognosis frequently seen in adults. Glioma development is associated with many genetic and molecular aberrations and substantially changes major signalling pathways. The standard of care for glioma has improved greatly; however, the efficacy is still unsatisfactory, 1 and the median survival of GBM after diagnosis is only 12-15 months. 2,3 Therefore, the primary focus of glioma research is to elucidate glioma pathogenesis further and provide new clinical treatment approaches.
Protein kinase N1 (PKN1) belongs to the PKN family. PKN1 is a serine/threonine protein kinase with a catalytic domain homologous to protein kinase C (PKC) and has a regulatory region widely distributed in various tissues, including the brain. 4 PKN1 is implicated in the progression of prostate, bladder, liver and pancreatic cancer. [5][6][7][8] However, the expression and the role of PKN1 in glioma have not been elucidated.
The Hippo pathway is involved in tumorigenesis, and YAP (yesassociated protein) is an essential member of this pathway. YAP translocates and binds to the cell nucleus. Through transcriptional binding partners, YAP promotes glioma cell proliferation, migration and invasion. 9,10 However, the effect of PKN1 on YAP has not been reported.
Tamoxifen (Tamo), a first-generation selective oestrogen receptor modulator (SERM), is a chemotherapeutic drug to treat oestrogen receptor positive breast cancer. Tamo inhibits glioma cell proliferation, promotes apoptosis, 11,12 and increases phototherapy sensitivity in U251 and U87 cell lines. 13 Tamo induces apoptosis in Temozolomide (TMZ)-resistant glioma cell lines. 14 Although Tamo can improve the quality of life and performance status of patients, 15,16 its effects on time to progression and median survival time of patients with highgrade glioma are unsatisfactory. We used VIOD, 17 a bioinformatics service platform developed by the School of Mathematics, Tianjin Nankai University, to predict and analyse the potential drug targeting of PKN1 and found that Ralo, a second generation SERM, 18 might be directly used against PKN1. Ralo can prevent breast cancer and has fewer side effects than Tamo, 19,20 and reportedly inhibits prostate cancer, lung cancer, pituitary adenoma and acute lymphocytic leukaemia progression. [21][22][23][24] In addition, Ralo enhances the chemosensitivity of glioma cells to TMZ in vitro. 25 However, the therapeutic effect of Ralo on GBM is largely unknown. 26 In this study, the inhibitory effects of Ralo on GBM cells and its potential targeted protein, PKN1, were investigated, and the effect of Ralo sensitizing GBM cells to TMZ was further identified.

| Glioma cell lines and cell culture
The human A172, U87, and LN229 GBM cell lines were purchased from Zhong Qiao Xin Zhou Biotechnology Co. Ltd (Shanghai, China).
The LN18, U118, U251, LN308 and SNB19 GBM cell lines were obtained from the Neuro-Oncology Laboratory, Tianjin Institute of Neurology. All cell lines were cultured in DMEM containing 10% FBS and in a 37°C 5% CO 2 incubator and subcultured every 2-3 days. The explant tissues were ground into a powder in liquid nitrogen, and 500 ul RIPA buffer was added. The tissues were then ultrasonically decomposed for 1 min, incubated on ice for 40 min, and centrifuged at 12000 rpm for 20 min. The supernatant was obtained, and appropriate 4× SDS-PAGE loading buffer was added, and boiled in a water bath for 5 min. The tissues were stored at −80°C for western blotting.

| IHC detection of tissue chip
The tissue chip was incubated overnight with the appropriate primary antibody (PKN1, 1:50 dilution) in a 4°C wet box. The chip was then incubated with a secondary antibody and stained using 3,3′-diaminobenzidine and haematoxylin. Dehydration and sealing of the chip followed, and the chip was scanned using a tissue chip scanner (Pannoramic MIDI, 3D HISTECH) and a semiquantitative analysis with histochemistry score (H-SCORE) was performed.

| Cell transfection and protein detection
The siR-PKN1 was transfected into A172 and U87 cells using an EntransterTM-R4000 when the cultured cell density reached 70-80%, and ADV-PKN1 was transfected into LN229 cells using ADV helper reagent according to the manufacturer's instructions. Proteins were extracted 48 h after transfection using RIPA lysis buffer (PMSF, 1:100), and cytoplasmic and nuclear proteins were extracted using Minute TM cytoplasmic and nuclear extraction kits. Western blotting detected PKN1, MMP2, Bcl2, PCNA, YAP nuclei and cytoplasm distribution.
Ralo was dissolved in DMSO to a concentration of 50 mM.
The 50% and 25% inhibition concentration (IC50 and IC25, respectively) values of Ralo in A172 and U87 cells were measured using the CCK-8 assay. A172 and U87 cells were then treated with corresponding concentrations of IC50 and IC25, and the proteins were extracted. PKN1, MMP2, Bcl2 and PCNA expression, and YAP expression in the cytoplasm and nuclei were analysed through western blotting.

| Glioma cell line proliferation analysis and colony-forming assay
Glioma cells were inoculated into 6-well plates, 500 cells/well, and cultured for 14 days. Cells were fixed with 4% paraformaldehyde for 10 min, then stained with 0.4% crystal violet for 10 min, and colony dots were photographed and counted.
For the CCK-8 assay, glioma cells were inoculated into 96-well plates and proliferation capacity was detected using the CCK-8 kit according to the manufacturer's instructions. Results are represented as the relative survival rate compared to the control.

| Real-time quantitative PCR to detect PKN1 and YAP expression
The total RNA was extracted from glioma cells using Trizol. RNA was reverse transcribed, and real-time quantitative PCR was conducted on a Bio-Rad Cycler system using the SYBER Premix DimerEraser.

| Flow cytometry to detect cell apoptosis
Cells of parental and positive groups in the logarithmic growth period were collected, centrifuged and resuspended in a binding buffer containing FITC/PI. Apoptosis was detected and analysed using FACSCalibur (Becton, Dickinson and Company) and FlowJo software (FlowJo LLC).

| Glioma cell invasion and migration
The bottom of the Transwell chambers was coated with diluted For the migration assay, 2 × 10 5 parental and positive cells were seeded on six-well plates and cultured. When the cell density reached 80-90%, "+" graph scratch wounds were made through the center of the plate using a 200 μL pipette tip, and cells were cultured.
At least three random visual fields were selected to be observed and photographed using an inverted light microscope at 0, 12 and 24 h.
The experiment was repeated thrice.

| Statistical analysis
Statistical analysis was performed using SPSS Statistics 17.0 (SPSS, Chicago, IL, USA) of anova analysis or Student's test. *p < 0.05 was considered statistically significant. The data are presented as the means±standard deviation, and the graphs were drawn using GraphPad Prism 6.0 (GraphPad Software).
PKN1 expression in GBM cell lines was higher than that in NB tissues. PKN1 expression in A172 cells was the highest, LN229 cells were the lowest and U87 cells were medium ( Figure 1A). was presented in the cytoplasm and nucleus with increasing tumour grade. The semi-quantitative scores of low-and high-grade glioma tissues were higher than in the control group, and the higher the glioma grade, the higher the H-SCORE ( Figure 1D,E). These findings indicated that the PKN1 expression level was positively correlated with the tumour grade.

| PKN1 plays a vital role in glioma pathogenesis
When siR-PKN1 was transfected into A172 and U87 cells with higher PKN1 expression, PKN1 expression was inhibited. In contrast, in LN229 cells, which had low PKN1 expression, when transfected with ADV-PKN1, PKN1 expression was upregulated. The PKN1 expression level in GBM cell lines was detected using western blotting and real-time PCR (Figures 2A and 3A). Four different ADV-PKN1 concentrations (T1-T4) were tested; T4 appeared to be the most effective and was identified as a working concentration in the followup studies ( Figure 3A).
The proliferation viability of the siR-PKN1 groups significantly decreased compared to the siR-NC groups ( Figure 2B,D), whereas the GBM cells proliferation in the ADV-PKN1 group was enhanced ( Figure 3B,E).
In the wound healing assay, the migration ability of PKN1 knockdown cells was significantly reduced, and PKN1 overexpression drastically increased the mobility of GBM cells (Figures 2E and 3D) compared with that in the control group. The number of invading cells in A172 and U87 cells transfected with siR-PKN1 was significantly decreased, whereas the invasion of LN229 cells transfected with ADV-PKN1 was increased compared to the control group ( Figures 2C and 3C).
Moreover, the apoptosis of A172 and U87 cells in the siR-PKN1 group was increased compared with the control cell group, indicating that PKN1 knockdown induced apoptosis, whereas PKN1 overexpression suppressed LN229 cell apoptosis (Figures 2F and 3F).
MMP2, PCNA and Bcl2 expression was significantly decreased in A172 and U87 cells transfected with siR-PKN1, but their expression was increased in LN229 cells transfected with ADV-PKN1 ( Figure 2G and 3G). These results coincided with the effect of PKN1 on cell proliferation, invasion and apoptosis, as described above.

| The effects of Ralo targeting PKN1 on the biological behaviour of GBM cells
The IC50 value, the drug concentration required for 50% cell growth inhibition, of Ralo for A172 and U87 cells were 34 μM and 32 μM, respectively (Figure 4-1A). The IC25 value, the drug concentration required for 25% cell growth inhibition, of Ralo for A172 and U87 cells were 25 μM and 15 μM, respectively (Figure 4-1A). Therefore, F I G U R E 1 PKN1 expression in GBM cell lines and glioma tissues (*p < 0.05, **p < 0.01; compared with NB tissues) (A) The protein expression level of PKN1 in GBM cell lines (A172, U87, LN229, LN18, U118, U251, LN308, and SNB19) was detected using western blotting, and the protein rations for all lanes were included in western blots; the NB tissues served as a negative control. (B) The protein expression level of PKN1 in 37 glioma specimens and 10 NB tissues was detected using western blotting, and the protein rations for all lanes were included in western blots.

| Ralo enhances TMZ sensitivity by targeting PKN1
The IC50 values of TMZ in the NC, siR-PKN1, and ADV-PKN1 groups of A172 cells were 1306.4, 862.3, and 1409.8 μM, respectively, whereas, in U87 cells, the IC50 values were 1877.9, 1226.9 and 2489.9 μM, respectively ( Figure 6A). The IC50 value of TMZ in the PKN1 knockdown group was significantly lower than in the control group. The IC50 value in the PKN1 overexpression group was higher than in the control group. Therefore, PKN1 could enhance TMZ resistance in GBM cells.
We administered Ralo combined with TMZ to GBM cells and  In the present study, PKN1 expression was upregulated in glioma cell lines and specimens. Furthermore, a positive correlation was observed between the PKN1 expression level and histopathological tumour grades. In addition, IHC staining of glioma tissue microarray showed that PKN1 was detected in the cytoplasm of low-grade specimens 8 ; however, PKN1 was detected in the cytoplasm and nucleus in high-grade gliomas. This finding indicates that PKN1 translocates from the cytoplasm to nuclei in the malignant glioma cell transformation.
To further elucidate the role of PKN1 on glioma pathogenesis, we knocked down PKN1 in GBM cell lines and found that proliferation, migration and colony formation of GBM cells were suppressed.
In contrast, apoptosis was induced when PKN1 was knocked down. reportedly achieve additional benefits when combined with TMZ in vitro. 25 We identified that Ralo suppressed proliferation, invasion, mi-

F I G U R E 5
The effect of PKN1 and Ralo treatment on YAP expression in GBM cell lines (*p < 0.05, **p < 0.01; compared with siR-NC group and DMSO group). (A) After siR-PKN1 transfection in A172 and U87 cells and ADV-PKN1 transfection in LN229 cells, the total YAP expression and its distribution in the cytoplasm and nucleus were examined through western blotting. (B) RT-PCR detected the YAP expression in GBM cells transfected with siR-PKN1 or ADV-PKN1. GAPDH was the loading control. (C) RT-PCR detected the YAP expression in A172 and U87 cells treated with Ralo; GAPDH was the loading control. (D) YAP expression and its distribution in cytoplasm and nucleus after Ralo treatment. (E) ADV-PKN1 was transfected into Ralo-treated GBM cells to detect the total YAP expression and its distribution in the cytoplasm and nucleus.
As a tumour with a poor prognosis, the median survival of GBM is less than 1 year. TMZ is the first-line chemotherapeutic drug in the standard of care for GBM to improve patient survival. 31 TMZ resistance weakens the TMZ response in GBM 32 ; the anti-tumour effect of TMZ with a sensitizer will be more effective. Searching for drugs that enhance the sensitivity of GBM to chemotherapy is the focus of treatment. In this study, the CI value 33 of Ralo combined with TMZ was less than 1, indicating that the combination of Ralo and TMZ suppressed GBM cell proliferation more significantly than TMZ alone, and Ralo enhanced the sensitivity of GBM cells to TMZ.
The O-6-methylguanine-DNA methyltransferase (MGMT) is responsible for the repair of the DNA damage induced by TMZ. MGMT repairs O6-meG lesions by transferring the alkyl group from guanine to a cysteine residue. 34 MGMT overexpression contributes to TMZ resistance, whereas MGMT downregulation in glioma cells increases the tumour sensitivity to the cytotoxic effects of TMZ. We also identified that Ralo repressed MGMT and increased tumour cell sensitivity to TMZ by inhibiting PKN1.

| CON CLUS IONS
We identified that PKN1 is upregualted in glioma, and the promoting role of PKN1/YAP in glioma pathogenesis. Rlao has been proved to inhibit glioma progression by targeting PKN1, and its synergistic

ACK N O WLE D G E M ENTS
We thank Prof Ruan Jishou, from School of Mathematics, Tianjin Nankai University, for bioinformatic analysis assistance.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are available from the corresponding author upon reasonable request.

E TH I C S S TATEM ENT
The Tianjin Medical University General Hospital Ethical Committee