Pristimerin‐induced uveal melanoma cell death via inhibiting PI3K/Akt/FoxO3a signalling pathway

Abstract Uveal melanoma (UM) is a highly invasive intraocular malignancy with high mortality. Presently, there is no FDA‐approved standard for the treatment of metastatic UM. Pristimerin is a natural quinine methide triterpenoid compound with anti‐angiogenic, anti‐cancer and anti‐inflammatory activities. However, Pristimerin potential cytotoxic effect on UM was poorly investigated. In the present study, we found the migration and invasion of UM‐1 cells were inhibited by Pristimerin which also caused a rapid increase of ROS, decreased mitochondrial membrane potential, induced the accumulation of cells in G0/G1 phase, ending with apoptotic cell death. Pristimerin inhibited Akt and FoxO3a phosphorylation and induced nuclear accumulation of FoxO3a in UM‐1 cells, increased the expression of pro‐apoptotic proteins Bim、p27Kip1, cleaved caspase‐3, PARP and Bax, and decreased the expression of Cyclin D1 and Bcl‐2. LY294002 or Akt‐siRNA inhibited the PI3K/Akt/FoxO3a pathway and promoted the Pristimerin‐induced apoptosis, while Pristimerin effects were partially abolished in FoxO3a knockdown UM‐1 cell cultures. Taken together, present results showed that Pristimerin induced apoptotic cell death through inhibition of PI3K/Akt/FoxO3a pathway in UM‐1 cells. These findings indicate that Pristimerin may be considered as a potential chemotherapeutic agent for patients with UM.


| INTRODUC TI ON
Uveal melanoma (UM) is a common primary intraocular malignancy with an annual incidence of about 7 per million. 1 However, the mechanism underlying UM pathology is unclear. At present, there is no definite treatment for UM. Therefore, it is an important clinical need to study its underlying pathological mechanisms and formulate corresponding chemotherapeutic strategies.
Pristimerin ( Figure 1A) is a natural triterpenoid quinine compound, which is a traditional Chinese medicine isolated from the Celastraceae and Hippocrateacea plants. It has long been used as an anti-malarial, anti-inflammatory, anti-oxidant and insecticide. 2,3 Recent studies have shown that Pristimerin potently induced anti-proliferative and apoptosis activities in several human cancer cell lines, which originated from lung, breast, prostate, glioma, cervical, leukaemia and multiple myeloma tumours. 2,[4][5][6][7][8] Induction of apoptotic cell death by Pristimerin involved with different mechanisms, including caspase activation, proteasomes inhibition, mitochondrial dysfunction and different molecular mechanisms involved in the suppression of anti-apoptotic NF-κB, Akt and MAP kinases. [9][10][11] In addition, Pristimerin has been reported to activate the stress kinase, c-Jun N-terminal kinase(JNK) and the DNA damage sensor, poly (ADP-ribose) polymerase-1 (PARP-1) through the generation of reactive oxygen species (ROS). 12 Moreover, other studies indicated that Pristimerin inhibited cell cycle progression, tumour cell migration and angiogenesis. 5,[13][14][15] Unfortunately, the cytotoxic effects and the molecular mechanism by which Pristimerin affects UM-1 were poorly investigated and only one study reported that Pristimerin inhibited the malignant phenotypes of UM cells through inactivation of NF-κB pathway. 16 Here, we focus on the effect of Pristimerin on the PI3K/Akt signalling pathway in UM-1 cells.
The PI3K/Akt pathway is a highly conserved central regulator of growth, proliferation, motility and metabolism. The genes and proteins of this pathway have been extensively studied and found to be widely activated in human cancers. Inhibition of this pathway has been shown to cause better regression of human tumours and has been evaluated in preclinical research and clinical trials. 17,18 Moreover, the activation PI3K/Akt pathway represents one of the most frequent genetic alterations found in human disease. 19 PI3K/ Akt pathway dysregulation has also been associated with resistance to conventional chemotherapy. 20,21 The FoxO transcription factors are the direct downstream target of the PI3K/AKT pathway, comprising four highly related members-FoxO1, FoxO3a, FoxO4 and FoxO6. 22  assays; (D, E) UM-1 cells were exposed to various concentrations for 14 d, and clonogenic assay was employed to detect cell reproductive death. UM-1 cells were treated at indicated concentrations for 24 h, and then, the cells were stained with Hochest 33342 (F, G-apoptosis), FITC/PI (H, apoptosis), JC-1 (I, mitochondrial membrane potential) or DCFH-DA (J, K-ROS) followed by high-content screening or flow cytometry. The data were analysed by Flowjo 7.6. The results represent mean ± SD of three separate experiments ( * P < .05, ** P < .01 versus control group) proliferation, differentiation, DNA repair, defence against oxidative stress, apoptosis and autophagy. [23][24][25] They are also associated with multiple diseases, including cancer. 26,27 The activity of FoxO can be regulated post-translational by various signalling pathways in which the Akt serves as an important regulator. 28 Akt can directly phosphorylate FoxO transcription factors, promote their translocation into the cytoplasm from the nucleus and cause its functional inhibition. 29 In the absence of active Akt, FoxO proteins are localized in the nucleus, where they regulate the transcription of genes related to cell cycle arrest, reactive oxygen species (ROS) detoxification and regulate expression levels of apoptotic proteins such as Bim, p27 Kip1 , p21 and PUMA. [30][31][32] In the present study, we investigated Pristimerin-induced pro-apoptotic effects in the UM-1 cells cultures in relation to PI3K/ Akt/FOXO3a signalling pathway and found that Pristimerin induced apoptosis by inhibiting the PI3K/Akt/FoxO3a pathway. These results suggest that Pristimerin may be considered as a lead cytotoxic compound in the chemotherapy of uveal melanoma.

| Cell cultures and transfection
The human uveal melanoma cell line UM-1, the neuronal precursor cell line RGC-5 and the D407 cells were obtained from Shanghai Bioleaf Biotech Co., Ltd. UM-1 cell line clone was originally isolated from the uveal melanin tumour tissue. RGC-5 cell line that displays mouse retinal ganglion progenitor cell characteristics correspond to mouse photoreceptor cell line 661W, providing a valuable tool to study pathogenesis of retinal neurodegenerative diseases. The cell lines were maintained in DMEM supplemented with 10% foetal bovine serum (FBS). UM-1 cells were transfected with FoxO3a, FoxO3a siRNA or N1 (empty plasmid) plasmid using Lipofectamine 2000 according to the manufacturer's protocol. 25 At 36 hours after transfection, cells were treated as indicated.

| Cell viability assay
UM-1 cells were plated at a density of 1 × 10 4 per well in 96-well plates, and the cells were treated with various concentrations of Pristimerin for 24h. Then, cell viability was assessed using MTT and CCK-8 assays. All experiments were performed in 5 replicates and repeated for 3 times.

| Clonogenic assay
UM-1 cells were plated at a density of 200 cells/well in the 6-well plates and then treated with various concentrations of Pristimerin.
After 14 days, colonies were fixed with 4% paraformaldehyde, washed twice and stained with crystal violet (0.01% w/v). The colonies containing more than 50 cells were counted.

| Hoechst 33342 staining for apoptosis
After treatment with Pristimerin, UM-1 cells were fixed with 4% paraformaldehyde for 10 minutes. Then, the cells were washed twice with PBS and exposed to Hoechst 33342 (10 µg/mL) in PBS for10 minutes at room temperature. Chromatin staining pattern was analysed for individual cells by high content screening system (Array Scan VTI, Thermo Fisher Scientific, USA).

| Mitochondrial membrane potential assay
Mitochondrial membrane potential (∆ψ) was measured by the flow cytometer with JC-1dye. The cells were analysed by flow cytometry, and the data were evaluated using the Flowjo 7.6.1 software.
For this assay, UM-1 cells were treated with Pristimerin and then incubated with 50 µM DCFH-DA for 30 minutes at 37°C. Subsequently, the fluorescence was then detected using a high content screening system (ArrayScanVTI, Thermo Fisher Scientific) using an excitation wavelength of 480 nm. The experiments were performed at least three times.

| Cell cycle assay
We determined phases of cell cycle using flow cytometry. In brief, UM-1 cells were seeded on 6-well plates and treated with various concentrations of Pristimerin with/without LY294002 (30 µM) or Akt siRNA for 24 hours, and then, cells were collected after gentle trypsin treatment and fixed overnight in ice-cold 70% ethanol, after which cells were treated with RNase A and stained with propidium iodide (PI). Finally, the cell cycle distribution was analysed by a flow cytometer and the date were calculated using the Flowjo 7.6.1 software. All experiments were performed in triplicate.

| Real-time PCR
According to the manufacturer's instructions, total RNA was extracted with Trizol reagent (In Vitrogen). Reverse transcription was performed using Roche First Stand cDNA Synthesis Kit. PCRs were performed with the LightCycler ® 480 SYBR Green I using the following primers:

| Western blotting
Western blotting was performed according to protocols routinely used in our laboratory, as described by Zheng et al. 33 The protein bands were detected by ECL Blotting Detection Reagents. Blots were quantified using ImageJ analysis software. All experiments were performed in triplicate.

| Reporter gene assay
UM-1 cells were co-transfected with FoxO3a-pGL3 plasmid for firefly luciferase or the pGL2-basic empty vector and pRL-TK-luc plasmid encoding for Renilla luciferase. Cells were incubated for 36h and treated with Pristimerin for additional 6, 12 or 24 hours, after which the activity of firefly, and Renilla luciferase was measured using the Dual Luciferase reporter assay system (Promega). Values for firefly luciferase activity were normalized to Renilla luciferase activity in the corresponding well. The experiments were repeated three times.

| Immunofluorescence
After various treatments, UM-1 cells were washed three times with PBS, fixed with 4% paraformaldehyde at room temperature for

| Statistical analysis
The data are presented as the mean ± SEM of 3-5 cultures, and experiments were performed at least three times. Statistical comparison between groups was performed by ANOVA, followed two-sided t-test using SPSS 13.0 program. Differences were considered significant when P < .05.

| Pristimerin induced UM-1 cells apoptotic death
To The decline of mitochondrial transmembrane potential is another important indicator in early apoptosis. To assess a potential effect of Pristimerin on mitochondrial transmembrane potential, UM-1 cells were incubated with JC-1 and the change of fluorescence from red to green, reflecting the decline of the membrane potential, was measured.
As shown in Figure 1I, Pristimerin induced a decline of mitochondrial membrane potential in a dose-dependent fashion. Moreover, to assess whether the mitochondrial dysfunction was correlated to increased ROS level, the intracellular ROS level was evaluated using high content screening system. UM-1 cells were treated with different concentrations of Pristimerin, stained with DCFH-DA and the percentage of stained cells was measured compared with control. The results indicated that intracellular ROS level was increased in a concentration-dependent manner in UM-1 cells ( Figure 1J,K), providing an explanation to the findings of Pristimerin-induced mitochondrial dysfunction. Cumulatively, the findings described in Figure 1 indicated that the UM-1 cells are more sensitive than neuronal and epithelial cells towards the cytotoxic-apoptotic cell death effect of Pristimerin.  with the different Pristimerin concentrations ( Figure 4A,B,E,F).

| Pristimerin inhibited migration and invasion of UM-1 cells
Consequently, the phosphorylation level of transcriptional factor FoxO3a at Ser253 was also found to be significantly inhibited at 3-36 h from the start of treatment with different Pristimerin concentrations ( Figure 4A,C,E,G). Interestingly, the level of total FoxO3a significantly increased from 12-36 h upon treatment with Pristimerin concentrations from 1 to 30 µM ( Figure 4A,E,D,H). Meanwhile, we also found that Pristimerin in a dose-dependent fashion increased cleaved caspase-3, PARP and Bax levels of expression, but reduced the level of expression of Bcl-2 and as a consequence the ratio of Bcl-2/Bax was also decreased ( Figure 5). These results indicate an apparent direct correlation among Pristimerin-induced cytotoxic effects and inhibition of Akt/FoxO3a phosphorylation activities and cleaved caspase-3, PARP, Bax and Bcl-2 expression.

| PI3K/Akt/FoxO3a inhibitors differentially modulated protein expression of cell cycle and apoptosis-associated proteins in UM-1 cells induced by Pristimerin
In the next step, we investigated the effects of PI3K/Akt and FoxO3a inhibitors on Pristimerin effects on phosphorylation and expression protein level of FoxO3a, a transcriptional factor that regulates gene expression, essential for cell cycle arrest and apoptosis in its dephosphorylated form in a nuclear location. 35,36 Figure

| Pristimerin inhibited cytoplasmic-nuclear shuttling and phosphorylation activity of FoxO3a
We used immunofluorescence microscopy to observe the shuttling To further corroborate these findings, we extracted cytoplasmic and nuclear proteins and the level of non-phosphorylated and phosphorylated FoxO3a was measured in both control and Pristimerintreated group by Western blotting using respective antibodies and considering Lamin B1 and β-actin as markers of nucleus and cytoplasm, respectively. We found that Pristimerin increased the nuclear levels of FoxO3a protein but decreased its levels of phosphorylation in a time-dependent manner ( Figure 8B,D), an effect potentiated by LY294002 ( Figure 8H-K). Furthermore, the reporter gene assay indicated that Pristimerin promoted FoxO3a transactivation activity ( Figure 8F). Taken together, these results indicate an apparent correlation between Pristimerin-induced apoptotic cell death and F I G U R E 6 Effects of PI3K/Akt inhibition on Pristimerin-induced cell cycle distribution, cells apoptosis and mitochondrial membrane potential in UM-1 cell. (A-D) apoptosis; (E, F) mitochondrial membrane potential; (G, H) cell cycle distribution. UM-1 cells were first pretreated with 30 µM LY294002 for 1 h or 10 µM Akt siRNA for 36 h and thereafter exposed for 24 h to 3 µM Pristimerin. Thereafter, the cultures UM-1 cells were analysed by flow cytometry upon staining with FITC/PI (A, C) or JC-1 (E). The results represent mean ± SD of three independent experiments ( * P < .05 versus control group; # P < .05 versus Pristimerin group) inhibition of PI3K/Akt phosphorylation and FoxO3a nuclear-cytoplasm shuttling inhibition and transactivation.

| D ISCUSS I ON
Uveal melanoma (UM) is considered a rare disease, but in fact, it is the most common primary intraocular malignant tumour in adults. 38 Paradoxically, even combined with surgery, chemotherapy and radiotherapy, the median survival time of patients is did not improve significantly. 39 Natural products derived from medicinal plants have been used since ancient times for the treatment of many diseases and have an important contribution to the discovery and development of new drugs with therapeutic potential against tumours. 40,41 Pristimerin, a triterpenoid quinone methide molecule, is characterized by beneficial pharmacological properties such as anti-inflammatory, anti-oxidant, anti-tumour, anti-malaria and anti-microbial activities. However, Pristimerin-induced cell death in UM-1 cells was poorly investigated. In the present study, we found that Pristimerin induced a pro-apoptotic effect in the UM-1 cells through modulation of the PI3K/Akt/FoxO3a signalling pathway. We found that Pristimerin increased ROS, decreased the mitochondrial membrane potential, promoted accumulation of cells in G0/G1 phase of the cell cycle and induced apoptotic cell death.
In recent years, it has reported that Pristimerin could affect many tumour-related processes, such as autophagy, apoptosis, vasculogenesis, migration and invasion, and drug resistance. 42 In human breast cancer cells, Pristimerin-triggered apoptosis through caspase activation, which could be completely prevented by benzyloxycarbonyl Val-Ala-Asp-fluoromethyl ketone, a pan-caspase inhibitor. 10  Mitochondria have been shown to play a key role in the apoptotic process due to the release of the pro-apoptotic proteins. 45 Pristimerin was found to induce mitochondria depolarization 46 and in breast tumour cells caused a rapid release of cytochrome c, triggering caspase activation and decrease of the mitochondrial membrane potential. 10 Here, we confirmed these findings since treatment of UM-1 cells with Pristimerin indicated decrease of mitochondrial membrane potential in correlation to the cytotoxic effect. Mitochondrial reactive oxygen species (ROS) production play a crucial role in induction of both intrinsic and extrinsic apoptotic cell death. 47 Many studies have showed that some anticancer agents, such as Epirubicin and Daunomycin, induced apoptosis in part with the generation of ROS and the disruption of redox homeostasis. 48 The increase production of ROS may lead directly to the mitochondrial permeability transition activation and induce the loss of mitochondrial membrane potential. 49 In line with this concept, Pristimerin induced mitochondrial cell death through ROS generation in cervical tumour cells. 7,12 However, the generation of ROS in breast cancer cells was not affected by Pristimerin, 10,12 indicating differences between tumour types. Here, we found that Pristimerin stimulated production of ROS in UM-1 cells, in correlation with the decrease in the mitochondrial membrane potential, emphasizing the role of ROS-mediated mitochondrial dysfunction in its cytotoxic effect. These data are reminiscent of previous reports indicating that Pristimerin anticancer effects involve ROS-mediated mitochondrial dysfunction. 8,12,50 The PI3K/Akt signalling pathway is necessary for cell growth and survival in many human cancers and is also activated in uveal (e) Pristimerin-induced FoxO3a translocation into the nucleus by inhibition of its cytoplasmic-nuclear shuttling and phosphorylation activity.
In conclusion, we propose that Pristimerin induced apoptosis by inhibiting the PI3K/Akt/FoxO3a signalling pathway in UM-1 cells.
These in vitro findings propose considerations of Pristimerin as a lead cytotoxic compound in the chemotherapy of uveal melanoma.
Further studies in animal models as well as clinical studies are required to characterize the therapeutic potential of Pristimerin towards UM tumours.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data are available from the corresponding author on reasonable request.