Potent antitumour of the mTORC1/2 dual inhibitor AZD2014 in docetaxel‐sensitive and docetaxel‐resistant castration‐resistant prostate cancer cells

Abstract Recent studies indicate mammalian target of rapamycin (mTOR) may play an important role in PCa progression and drug resistance. Here, we investigated the effects of a novel mTORC1/C2 dual inhibitor, AZD2014, on naive and docetaxel (Doc)‐pre‐treated castration‐resistant PCa (CRPC) cells and explored its therapeutic potential in CRPCs. In the current study, AZD2014 has a greater inhibitory effect against 4EBP1 and AKT phosphorylation than rapamycin in CRPC cells and prevented the feedback activation of AKT signalling. Importantly, AZD2014 suppressed CRPC cell growth in vitro by suppressing proliferation, apoptosis, cell cycle arrest at G1 phase and autophagy to a greater extent than rapamycin. Moreover, AZD2014 was more efficacious than rapamycin in inhibiting migration, invasion and EMT progression in Doc‐sensitive and Doc‐resistant CRPC cells. Overall, AZD2014 showed significant antitumour effects. Thereby, the current study highlights a reliable theoretical basis for the clinical application of AZD2014 in both Doc‐sensitive and Doc‐resistant CRPCs.

be life-threatening. 4 Among the available treatments for CRPC, the efficacy of conventional chemotherapy remains limited. 5 Recently, cabazitaxel was designed to overcome Doc resistance in CRPC therapy, but cabazitaxel increased expression of the multidrug resistance 1 (MDR1) protein, induced anticancer drug efflux from the cell and caused a variety of serious side-effect. 6 Therefore, the need to find new methods to improve the efficacy of anticancer treatment in Doc-resistant patients is urgent.
The phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) pathway, one of the most commonly activated signalling pathways in many human cancers, is a key signalling pathway that controls cell growth, differentiation and metabolism. 7 The PI3K/Akt/mTOR pathway was found to be inappropriately activated in PCa tissues by immunohistochemistry and was more prevalent in metastatic sites, 8 as it was deregulated in 42% of localized disease cases and 100% of advanced-stage disease cases. 9 In addition to its role in regulating cell proliferation and invasion, the mTOR pathway is related to drug resistance. 10,11 Many investigations have shown that the PI3K/Akt/mTOR pathway is closely related to the progression of CRPC and the development of drug resistance. 12 Therefore, targeting mTOR signalling is considered a very promising treatment for CRPC. In addition, the PI3K/Akt/mTOR pathway is a critical target for many other cancer treatments 13,14 mTOR exists in complexes called mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2), which share subunits but have distinct cellular functions and localizations. The so-called rapalogs, such as everolimus, an mTORC1 inhibitor, have been approved to treat certain types of cancer, such as kidney cancer and pancreatic neuroendocrine tumours. 15,16 Everolimus combined with Doc was shown to control the progression of disease in CRPC patients. 17 However, androgens are vital factors that regulate tumour growth in CRPC patients and can activate the mTORC2 pathway, resulting in the activation of Akt and facilitating the survival of tumour cells. These results indicate a reciprocal feedback mechanism by which PI3K/Akt/mTOR signalling inhibits androgen receptor (AR) signalling, indicating a potential mechanism underlying the clinical inefficacy of mTOR inhibitors in CRPC. 18 Preclinical studies investigating specific inhibitors of PI3K or mTOR yielded promising results; however, the evidence from clinical trials remains less convincing.
There are some dual mTORC1/2 inhibitors, such as AZD8055, OSI-027, MTI-31 (LXI-15029), TAK-228 and AZD2014. Many studies have confirmed that AZD2014 can inhibit the growth of pancreatic neuroendocrine tumours resistant to everolimus and that AZD2014 is also a highly effective treatment for renal cell carcinoma and ER + breast cancer. [19][20][21] Therefore, it is reasonable to hypothesize that simultaneously inhibiting mTORC1 and mTORC2 would have a greater antitumour effect than rapalogs or even reverse Doc resistance in these kinds of PCa cell lines. In this study, we revealed the potential impact of AZD2014 in impeding both mTORC1 and mTORC2 signalling in CRPC cells. Analyses of AZD2014 in cultured human PCa cell lines revealed that AZD2014 has broad effects on cancer resistance across both Doc-sensitive and Doc-resistant CRPC cell lines. These effects include antiproliferative effects and the ability of AZD2014 to induce apoptosis; inhibit migration, invasion and EMT progression; and activate autophagy. In addition, parallel studies showed that AZD2014 is more efficient than rapamycin and the mechanism of action of AZD2014 as an antitumour agent. These promising results provide rationale for the clinical assessment of AZD2014 in PCa therapy.

| Chemicals and antibodies
AZD2014, rapamycin and docetaxel were obtained from Selleck

| Cell lines and cell culture
The docetaxel-sensitive prostate cancer cell lines (C4-2, CWR22RV1) were purchased from the American type culture collection (ATCC),US. Drug gradient generated docetaxel-resistant prostate cancer cell lines (C4-2 DOCR and CWR22RV1 DOCR, which were provided by Chawnshang Chang from University of Rochester, US).
All the cell lines were grown in RPMI-1640 (Gibco) containing 10% foetal bovine serum and incubated in a humidified atmosphere of 5% CO 2 maintained at 37°C.

| Cell cycle
After 48 hours incubation with or without AZD2014 or rapamycin, cells were collected and fixed in 70% cold ethanol overnight at 4°C and then incubated with RNase A (Sigma) at 37°C for 30 min avoided from light and subsequently stained by propidium iodide (PI) (Nanjing KeyGen Biotech Co., Ltd.) staining solution. The samples were immediately were analysed on a FACSort flow cytometer (BD Biosciences). Data were processed by CELL quest software (BD Biosciences).

| Flow cytometry assay
After 48h incubation with or without AZD2014 or rapamycin, cells were harvested and resuspended in fixation fluid. 5 µl Annexin V-FIFC and 5µl propidium iodide were added to 500 µl cell suspension. Cell apoptosis was then determined by using FACSort flow cytometer (BD Bioscience). In the graphs, the quadrant, respectively, stands for dead cells, living cells, early apoptotic cells and late apoptotic cells.

| Wound-healing assay
Cells motility was determined by wound-healing assay. Cells were cultured in the presence or absence of AZD2014 or rapamycin for 48 hours and then spread within six well plate with 4 × 10 5 cells density per well and cultivated into 90% confluent cells. Then, a wound field was created by using a sterile 200 μl pipette tip, which resulted in a denuded area with a fixed width. Phosphate buffered saline (PBS) was used to wash cells off cell debris, and culture medium was added to the cell culture. The cells were incubated for 48 hours at 37°C with or without AZD2014 or rapamycin, and then, the migration of cells was monitored with a digital camera system at different timepoints (0, 24 or 48 hours).The cell migration distance (μm) was calculated by the software programme HMIAS-2000. Photographs were taken at 0 and 48 hours at the same position of the wound, and the distance between the edges was measured.

| Immunofluorescence
Cells were cultured in a 24-well plate and subjected to immunofluorescence analysis at 48 hours after incubation with or without Experiments were repeated at least three times.

| Acidic vesicular organelles staining
To determine the presence of autophagic vesicles in prostate cancer cells, C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were grown on cover slips inserted into a 6-well plate and incubation with DMSO, 3MA, AZD2014, rapamycin or both of two for 48h. The

| Western blot analysis
Cells were harvested at 48 hours following co-cultured with AZD2014 or rapamycin at a certain concentration. Proteins were separated by 10% SDS/PAGE and transferred onto PVDF membranes (Millipore). After blocking with non-fat dry milk for 1 hour at room temperature, the membranes were incubated overnight at 4°C with specific primary antibodies. After being washed three times in TBST, membranes were incubated with corresponding second antibody (1:5000) in blotting buffer for 60 min at room temperature and visualized by enhanced chemiluminescence (ECL) assay kit (Millipore). Western blot bands were exposed to X-ray films and were quantified by Quantity One software. The relative expression of protein was standardized by comparison with anti-GAPDH or anti-β-tubulin antibody. All antibodies used in this work were appropriately dilutions with 1:1000.

| Statistical analyses
All experimental data from three independent experiments were analysed by Student's t test or chi-square test, and results were expressed as mean ± SD. In all cases, P-values <0.05 were considered to be statistically significant. All statistical tests were conducted by SPSS software (SPSS Standard version 19.0; SPSS Inc).

| AZD2014 restrains both mTORC1 and mTORC2 signalling in CRPC cells
AZD2014 is a dual inhibitor of mTORC1/2, and rapamycin is an mTORC1 inhibitor. We aimed to explore the biochemical activities of AZD2014 and rapamycin in controlling mTORC1 (p-4EBP1) and mTORC2 (AKT phosphorylated at Ser473) signalling in CRPC cells. were generated with a drug gradient) treated with AZD2014 and rapamycin at different concentrations. As shown by the results, in all the CRPC cell lines, AZD2014 effectively blocked the phosphorylation of 4EBP1 at Thr37/46 and that of AKT at Ser473 in a concentration-dependent manner, and this inhibitory effect was especially significant when the concentration was 100 nM or more. In contrast, rapamycin also had a certain weak inhibitory effect on 4EBP1 phosphorylation at Thr37/46 in a concentration-dependent manner but had little inhibitory effect on AKT phosphorylation at Ser473 even at a high concentration of 1000 nM ( Figure 1A-D). These results demonstrate that AZD2014 is a potent inhibitor of mTORC1 and mTORC2 in CRPC cells that blocks mTOR signalling more thoroughly than rapamycin. We performed a CCK8 assay to detect the

| AZD2014 inhibits proliferation and induces apoptosis in CRPC cells
To further explore the effects of AZD2014 and rapamycin on cell proliferation, we treated CRPC cell lines with AZD2014 at the ap- show certain specificity in CRPC and minimal side-effects in the normal prostate epithelium.

| AZD2014 induces cell cycle arrest at G0/G1 phase in CRPC cells
In the above experiments, we found that AZD2014 could signifi-

| AZD2014 activates autophagy in CRPC cells
Many studies have shown that the inhibition of mTOR signalling is associated with the induction of autophagy. 23 Therefore, we also evaluated whether AZD2014 affects this process. The CRPC cell lines were treated with AZD2014 and rapamycin, followed by monodansylcadaverine (MDC) staining to detect the formation of autophagic vacuoles. As shown in Figure 4A Figure 4B). In addition, AZD2014 treatment led to a more pronounced increase in the levels of LC3A/B-2 than rapamycin ( Figure 4B). These results suggest that the inhibition of mTORC1 by rapamycin had a less profound effect on the induction of autophagy than the blockade of mTORC1 and mTORC2 by AZD2014 in all four CRPC cell lines used here.  (Figure 5C-D). The effects of AZD2014 and rapamycin F I G U R E 2 AZD2014 inhibited proliferation and induces apoptosis in CPRC cells. C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were either treated with DMSO (CTR) or treated with the indicated concentrations of AZD2014 or rapamycin. The following experiments were performed after 48 hours. A-D, Cell growth curves of all prostate cancer cell lines were determined by CCK8 assay every 24 hours. E, G, I, K, Cell apoptosis was detected by Annexin V fluorescence-activated cell sorting (FACS), and the diagram shows representative apoptosis rates in prostate cancer cell lines treated with DMSO, AZD2014 or rapamycin. F,H,J,L, Statistical analysis of the apoptosis rates of prostate cancer cell lines. M, Western blot analysis of the relative expression of Bax and cleaved PARP. GAPDH served as a loading control. Data are presented as the mean ± SD from at least three independent experiments. (*P < .05, **P < .01) on CRPC cell invasion and migration were assessed in vitro, which showed that AZD2014 and rapamycin inhibited cell migration ( Figure 5E-F). In the process of EMT, the expression of intracellular epithelial markers (such as E-cadherin) is down-regulated, and some intracellular markers are up-regulated (such as N-cadherin and Vimentin) 28 We then determined whether AZD2014 and rapamycin could regulate EMT in CRPC cells ( Figure 6A). The expression of EMT markers was detected via immunofluorescence and Western blotting. The immunofluorescence results show that treatment of the CRPC cell lines with AZD2014 and rapamycin increased E-cadherin expression but decreased N-cadherin and vimentin expression.

| AZD2014 inhibits cell migration, invasion and EMT in CRPC cells
Western blotting confirmed the increased expression of E-cadherin F I G U R E 3 AZD2014 induced cell cycle arrest at G0/G1 phase in CRPC cells. C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were treated with either DMSO (CTR) or AZD2014 or rapamycin at the indicated concentration and further cultured for several days. A-H, After 48 hours, flow cytometry was used to determine the proportion of all prostate cancer cells in distinct cell cycle phases. I, After 48 hours, Western blot analysis of the relative expression of CDK4 and cyclin D1 was carried out. β-Tubulin served as a loading control. The results are plotted as the mean ± SD of three independent experiments. (*P < .05,**P < .01; #P < .05, ##P < .01). and decreased expression of Vimentin in CRPC cell lines treated with AZD2014 and rapamycin compared with the CTR groups ( Figure 6B).
These results indicated that AZD2014 is more effective than rapamycin in inhibiting cell invasion, migration and EMT in CRPC cells. is affected by rapamycin treatment, while mTORC2 is thought to be rapamycin-resistant, although prolonged treatment with rapamycin inhibits its assembly in some cell types. 32 Some studies have indicated that mTORC2 is critical for the progression of PCa when PTEN is deleted in the prostate epithelium, while mTORC2 activity F I G U R E 4 AZD2014 activated autophagy in CRPC cells. C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were treated with either DMSO (CTR) or AZD2014 or rapamycin at the indicated concentration. The expression of related proteins was tested by Western blotting. A, Cells were treated with AZD2014 or rapamycin at concentrations of 200, 250, 400 and 600 nM for 48 hours in the presence or absence of 3-MA. The cells were then stained with MDC to detect the formation of autophagosomes and immediately observed with a confocal microscope. Bars, 50 μm. B, The expression of Beclin-1 and LC3A/B (LC3A/B-1 and LC3A/B-2) in prostate cancer cells treated as described above was assessed by immunoblot analysis is not essential in the normal prostate epithelium, 33 indicating that mTORC2-specific inhibitors may be promising therapeutic agents for PCa. Similar results were confirmed in the treatment of colon cancer through the targeted inhibition of mTORC. 34 Increasing preclinical evidence has shown the broad prospects of AZD2014 in antitumour therapy, and AZD2014 will soon enter clinical trials. 35 However, there are still many important issues, such as genetic determinants, dosage and the time of drug use, that will ultimately determine its clinical success. 36 Despite advances in the diagnosis and management of PCa, a large proportion of men progress to advanced or metastatic PCa, and morbidity from PCa remains high. 37 At present, although androgen deprivation therapy through either surgical castration or chemotherapy remains the first-line therapy for metastatic PCa, the effects of this treatment are temporary, and patients inevitably progress to CRPC. 38  preventing the feedback activation of Akt signalling. 45 Our research confirmed that the new mTORC dual inhibitor AZD2014 can solve the above problems. In the current study, we clarified that AZD2014, a novel small-molecular ATP-competitive tyrosine kinase inhibitor (TKI) blocked a wide range of mTOR complex functions in both Doc-resistant and Doc-sensitive CRPC cell lines.

| D ISCUSS I ON
Furthermore, the current study shows that AZD2014 had a more exhaustive inhibitory effect on mTORC1 than rapamycin and that the repression of mTORC2 prevented the activation of AKT signalling. by RNA interference technology reduced the ability of a variety of cancer cells to migrate and invade. 49,50 A recent study showed that the inhibition of mTORC1 or mTORC2 reduced cell mobility by preventing rearrangement of the actin cytoskeleton and caused the formation of plate-like pseudopods. 51 During cell EMT, TGF-β activates mTORC1 and mTORC2 signals through PI3K, resulting in increased protein synthesis, migration, invasion and epithelial-mesenchymal phenotypic transformation. 43 In this study, AZD2014, a dual inhibitor of mTORC1 and mTORC2, was more effective than rapamycin, which suppressed mTORC1 alone, in reversing the transformation of epithelial cells to mesenchymal cells. After CPRC cells were treated with AZD2014, the expression of E-cadherin was increased, while N-cadherin and vimentin expression was reduced. Thus, the concurrent inhibition of mTORC1 and mTORC2 may be a more effective strategy than the suppression of mTORC1 alone in PCa therapy and could especially be beneficial for patients who have become resistant to rapalogs.

F I G U R E 5
The effects of AZD2014 on migration and invasion in CRPC cells. C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were treated with either DMSO (CTR) or AZD2014 or rapamycin at the indicated concentration. A, C, The invasive abilities of prostate cancer cell lines were determined with Transwell assays (representative photographs are magnified 200×). B, D, Representative quantification of migrated and invaded cells. E, The migratory abilities of prostate cancer cells were determined with wound-healing assays (representative photographs are magnified 200×). F, Migrated cells were quantified. The inhibition of cell invasion by rapamycin and AZD2014 was observed in prostate cancer cells. The inhibition of cell migration by rapamycin or AZD2014 was observed in prostate cancer cells. Data are shown as the mean ± SD. (*P < .05, **P < .01) The effects of AZD2014 on the CRPC cell lines in this study would be more compelling if they were further demonstrated in Doc-sensitive and Doc-resistant models in vivo. Currently, the mechanism of AZD2014 remains unclear; thus, further studies will provide assurance. Future exploration of predictive biomarkers of PCa susceptible to mTOR inhibitors may be a promising direction.

| CON CLUS IONS
Our data demonstrated that AZD2014 was a highly potent inhibitor of both mTORC1 and mTORC2 and an effective antitumour agent for the treatment of CRPC in vitro, as treatment with AZD2014 led to the more profound inhibition of mTORC1 than rapamycin and F I G U R E 6 The effects of AZD2014 on EMT in CRPC cells. C4-2, CWR22RV1, C4-2 DOCR and CWR22RV1 DOCR cells were treated with either DMSO (CTR) or AZD2014 or rapamycin at the indicated concentration. The following experiments were performed after 48 hours. A, Immunofluorescence staining of prostate cancer cells treated as described above for E-cadherin and vimentin. Bars, 50 μm. B, Prostate cancer cells were treated as described above, lysed and immunoblotted. The expression of EMT markers was detected via Western blotting. Rapamycin and AZD2014 increased the expression of E-cadherin and decreased the expression of N-cadherin and vimentin in prostate cancer cells. Data are shown as the mean ± SD. (*P < .05, **P < .01) inhibited mTORC2 without activating AKT signalling. AZD2014 participated in the regulation of various biological processes in PCa cells in vitro and was more effective than rapamycin. The underlying mechanism of the antitumour activity of AZD2014 is associated with its complete inhibition of both mTORC1 and mTORC2 substrates, and the inhibition of mTORC2 impedes AKT signalling feedback inhibition. These lines of evidence provide a molecular basis for the clinical application of a dual mTORC1/2 inhibitor and indicate a useful anticancer strategy for treating patients with both CRPC and Doc-resistant CRPC.

ACK N OWLED G EM ENTS
We would like to express our gratitude to all those who financed the subject. This research was supported by grants from the National Natural Science Foundation of China (81870518 to Wei Yu; 81770755 to Jie Jin) and the Natural Science Foundation of Beijing Municipality (No. 7192208 to Jie Jin) and also supported by grants from AstraZeneca Company (No. C-2016-004 to Yi Song).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing financial interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.