Autophagic degradation of CDK2 and CDK4 is responsible for NVP-BEZ235-induced G0/G1 cell cycle arrest in neuroblastoma

Cyclin-dependent kinases (CDKs) are regarded as promising targets for cancer therapy. CDK2 and CDK4 are the core molecules in the G0 and G1 phases of the cell cycle. Exploring the molecular mechanism of CDK2 and CDK4 protein degradation will provide clues and solutions to control the G0 and G1 phases in tumor cells. The expression levels of CDK2 and CDK4 were assessed by Western blotting and qRT-PCR. Cell viability, cell proliferation and growth were evaluated in CCK-8 and ow cytometric analysis. Protein interactions were analysed by immunoprecipitation and immunouorescence methods.


Background
The cell cycle process is orchestrated by the interplay of various cyclin/CDK combinations and other key cell cycle-related proteins at checkpoints to ensure unidirectional and irreversible cell division in eukaryotes [1][2][3]. As a critical regulator of cell proliferation, deregulation of the cell cycle universally leads to the development and progression of human cancers [4,5]. Hence, targeting cell cycle control, especially aiming at disrupting the related signaling pathways, may become a promising therapeutic approach.
Activation of the PI3K-Akt-mTOR pathway has recently attracted considerable attention due to its correlation with abnormalities of cell cycle regulation and tumor growth, and several targets of the PI3K-Akt-mTOR pathway are interrelated with cell cycle progression, including MDM2, GSK3β, p70S6K, and 4EBP1 [6,7]. These results suggest that inhibitors of this pathway may be a useful avenue for therapeutic intervention in cancer patients.
NVP-BEZ235 is a synthetic small molecule compound that potently and reversibly inhibits the catalytic activities of PI3K and mTOR by competing at their ATP-binding sites [8]. As a dual PI3K/mTOR inhibitor, NVP-BEZ235 shows the potential characteristics of overcoming resistance, which was observed in the negative feedback loop activation of Akt due to mTORC1 inhibitors, such as rapamycin [9], and activation of MAPK signaling with PI3K inhibitors, such as LY294002 [10]. NVP-BEZ235 is the most effective small molecule inhibitor of PI3K to date and has been a particularly attractive compound with the traits of a low dose requirement, good tolerance, a notable effect and greater selectivity compared with other dual inhibitors. Multiple studies have demonstrated the prominent inhibitory e cacy of NVP-BEZ235 on the proliferation in a wide variety of human malignant tumors in vitro and preclinical animal models and are currently in phase I/II clinical trials for advanced solid tumors [11,12]. It is generally believed that cell cycle arrest triggered by NVP-BEZ235 by antagonizing the PI3K/mTOR signaling pathway plays a pivotal role in its potent cytostatic effects. Previous observations have revealed that NVP-BEZ235 blocks the cell cycle at the G0/G1 phase, ultimately causing the interruption of cancer growth with the down regulation of cyclic D1 and pRb and the upregulation of p21 and p27 in multiple cancers, including neuroblastoma (NB) [13], osteosarcoma [14], B cell lymphoma [15] and nasopharyngeal carcinomas [16]. Additionally, it has been reported that NVP-BEZ235 can enhance the radiosensitivity of resistant prostate and glioblastoma cells through the attenuation of DNA double-strand break (DSB) repair and implementation of G2/M cell cycle arrest [17,18]. However, the precise mechanisms by which NVP-BEZ235 in uences cell cycle distribution via down regulation of cell cycle-related proteins remain largely unknown.
In the current research, NB models with excessive activation of the PI3K/Akt/mTOR pathway were used to evaluate the effect of NVP-BEZ235 on cell cycle arrest and investigate the concrete mechanisms by which NVP-BEZ235 induces the down regulation of cell cycle-related proteins. Our results indicated that NVP-BEZ235 facilitated autophagic degradation of CDK2 and CDK4, which can rapidly induce NB cell cycle arrest at G0/G1 state and proliferation inhibition.

Methods
Reagent, cell lines and cell culture NVP-BEZ235 was purchased from Novartis (Selleck, S1009) and dissolved in DMSO to a stock concentration of 1 mmol/L. CQ (Sigma-Aldrich, C6628), and 3-MA (Sigma-Aldrich, M9281) were obtained and dissolved in sterile H 2 O to stock concentrations of 20 mmol/L and 100 mmol/L, respectively. CHX was purchased from Sigma (Sigma-Aldrich, R750107) and dissolved in ethanol at a stock concentration of 100 mg/ml. The stock solutions were wrapping in foil and maintained at -20°C.
SH-SY5Y and SK-N-MC cell lines were obtained from the Shanghai Institutes for Biological Sciences (Shanghai, China) and were cultured in RPMI 1640 medium containing 10% fetal bovine serum (FBS) (Gemini, 100-700) and 1% penicillin-streptomycin (Gibco, 15140122) under standard culture conditions (37°C and 5% CO 2 ). Cells in the logarithmic phase of growth were used in all experiments.
Cell Counting Kit-8 assay Cell proliferation was assessed using the Cell Counting Kit-8 reagent (Dojindo Molecular Technologies, CK04). SH-SY5Y and SK-N-MC cells were seeded in 96-well culture plates at 5×10 3 cells/well in triplicate, allowed to settle overnight, and treated with serial concentrations of NVP-BEZ235 (0, 100 and 200 nmol/l) in 200 μl of medium for 24 hours. Transfected cells were treated with 200 nmol/l NVP-BEZ235 or DMSO for 12 hours. The cells were then treated with CCK-8 solution and incubated for 2 hours according to the CCK-8 kit instructions. The absorbance was measured at 450 nm using a microplate reader (Bio-Rad, Hercules, CA, USA). Each experiment was performed at least 3 times independently.
Cell cycle by ow cytometric analysis NB cells in exponential growth were seeded in 6-well plates at a density of 1×10 6 per well for 24 hours attachment, after which they were treated with NVP-BEZ235 (0, 100 and 200 nM) for 24 hours, and the transfected cells were treated with 100 nM NVP-BEZ235 or DMSO for 12 hours. For cell cycle analysis, cells were digested with 0.25% trypsin, centrifuged at 2000 ×g for 5 minutes, washed three times with PBS and xed with ice-cold 75% (v/v) ethanol at -20°C for 2 hours. Then, after washing twice with PBS, the cells were resuspended in 50 μl of PI solution (PI 50 μg/ml and RNase A 100 μg/ml). After incubation for 15 minutes at room temperature in the dark, the cell cycle distribution was analyzed by a FACSCanto ow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).

Western blot analysis
Cultured cells or mouse tumor tissues were homogenized on ice in 1× RIPA lysis buffer containing phosphatase (Roche, 4906845001) and protease inhibitor cocktail (Roche, 4693116001), and the protein concentration was measured by a BCA Protein Assay Kit (Takara, T9300A). Equal amounts (50 μg) of isolated proteins were separated by SDS polyacrylamide gel electrophoresis and transferred to PVDF membranes. Membranes were blocked in 5% nonfat milk in Tris-buffered saline containing 0.1% Tween-20 (TBST, pH 7.4) for 1 h at room temperature and incubated overnight at 4°C at a dilution of 1:1000 with primary antibodies against Pi-RB, CDK2, CDK4, GAPDH, CTSB, CTSD, p62, LC3B and Beclin-1 (all from Cell Signaling Technology). Then, the membranes were probed for 2 hours at room temperature with speciesspeci c HRP-conjugated rabbit anti-mouse IgG (Cell Signaling Technology) at a dilution of 1:10000.

Immunoprecipitation assays
After washing in cold PBS, cells or tissues were lysed in 500 μl of ice-cold RIPA lysis buffer (50 mM Tris-HCl pH 7.4, 1 mM EDTA, 0.5% sodium deoxycholate, 1% NP-40, 0.1% SDS, and 150 mM NaCl) containing protease and phosphatase inhibitors on ice for 30 minutes. Total cell lysates were clari ed by centrifugation at 12,000 ×g for 15 minutes at 4°C and were divided into two parts, with one (400 μl, 500 µg) used for the immunoprecipitation assays and the other (50 μl) used for total protein analysis. Lysate supernatants were incubated with 3 μg of anti-p62 antibody or rabbit IgG overnight at 4°C with rotation, followed by preclearing with a 50% protein A/G agarose slurry for 2 hours, and then the complexes were incubated with protein A/G agarose for 4 hours at 4°C. After washing 5 times with lysis buffer by centrifugation at 5,000 rpm for 5 minutes, samples were eluted in 2X SDS loading buffer by boiling for 8 min and subjected to immunoblot analysis with anti-CDK2, anti-CDK4 or anti-ubiquitin antibodies.
Immuno uorescence assays NB cells were cultivated in a 96-well plate exposed to various concentrations (0, 200 nM) of NVP-BEZ235 for 24 hours. After washing twice with PBS, the cells were xed with 4% paraformaldehyde/PBS for 15 min at room temperature. Cells were then washed three times with cold PBS, permeabilized with methyl alcohol for 10 min, and nally washed with 0.05% Tween-20/PBS for 5 min. The cells were blocked with 5% BSA/PBS (Sangon Biotech, A0332) for 2 hours at room temperature and then incubated with primary antibodies (1:100) overnight at 4°C. After washing three times with 0.05% Tween/PBS, the cells were incubated with a FITC-conjugated secondary antibody (ZF-0312, 1:100) for 2 hours at room temperature in a dark humid chamber. The cell nuclei were stained with DAPI (Sigma-Aldrich, D9542). The cells were visualized by a uorescence microscope (Nikon Eclipse Ti, Japan).

Results
NVP-BEZ235 could induce G0/G1 cell cycle arrest with the down regulation of the CDK 2 and CDK 4 proteins.
We analyzed the effect of NVP-BEZ235 on cell cycle progression in NB cells by ow cytometry using PI staining of the DNA content. After exposure to 100 nM and 200 nM NVP-BEZ235 for 24 hours, there was a signi cant increase in the percentage of the SH-SY5Y cells in the G0/G1 phase (82.28%±2.86% at 100 nM, 85.98%±1.93% at 200 nM) compared with the control group (66.82%±3.50%) (Fig. 1A). A signi cant increase in G0/G1 phase cells was also detected in SK-N-MC cells (Fig. 1B). To further investigate the correlation of NVP-BEZ235-induced G1 cell cycle arrest with some important G1-phase regulatory proteins in NB cells. The expression of endogenous CDKs was assessed when SH-SY5Y and SK-N-MC cells were treated with 200 nM NVP-BEZ235 for various times. CDK4 and CDK2 are signi cant G1 phase-associated regulatory proteins governing the restriction point of G1→S phase progression, and their inhibition or degradation lead to G1 phase arrest [20]. As indicated in Fig. 1C and D, there was a remarkable reduction or gradual decrease in CDK4 and CDK2 protein levels in both cell lines. Furthermore, phospho-Rb initially triggered by dependent kinases and then ampli ed by CDK2 complexes markedly decreased in a timedependent manner. This process is critical for the G1-to S-phase transition, as phospho-Rb results in the release of active E2F, which stimulates the transcription of the gene-mediated G1 phase. Based on the above results, we concluded that NVP-BEZ235 exerts notable G0/G1 cell cycle arrest in NB cells through the down regulation of several key G1-phase regulatory proteins.
NVP-BEZ235 attenuated the G0/G1 cell cycle-related protein levels by facilitating their protein degradation.
To elucidate whether NVP-BEZ235-mediated down regulation of CDK4 and CDK2 proteins results from transcriptional or posttranscriptional regulation, their mRNA levels were evaluated by qRT-PCR. As shown in Fig. 2A and B, the mRNA levels of CDK4 and CDK2 were not drastically altered in the presence of 200 nM NVP-BEZ235. This indicates that NVP-BEZ235 may affect the protein stability of CDK4 and CDK2 that were detected via CHX-chase analysis. As showed in Fig. 2C-H, NB cells were cotreated with 10 µg/ml CHX for different time followed by pretreatment with 200 nM NVP-BEZ235 for 1 hour. NVP-BEZ235 obviously attenuated the half-life of these proteins in both cell lines. These data suggested that NVP-BEZ235 reduces CDK4 and CDK2 protein stability and promotes their degradation.
Autophagy-lysosome inhibitors restored the degradation of CDK2 and CDK4 proteins in NVP-BEZ235treated NB cells.
To clarify whether the ubiquitin-proteasome pathway is involved in the degradation of CDK2 and CDK4 proteins, we pretreated SH-SY5Y and SK-N-MC cells with the 26S proteasome inhibitor MG132 (10 µM) for 1 hour, followed by treatment with 200 nM NVP-BEZ235 for 12 hours. Compared with the NVP-BEZ235 single factor treatment group, the CDK2 and CDK4 proteins in the MG132 and NVP-BEZ235 dual factor treatment groups did not change signi cantly (Fig. 3A and B). In addition to the effect on cell proliferation, the PI3K/Akt/mTOR pathway is also involved in the autophagy process. We further assessed autophagy-related hallmark proteins in both cell lines treated with NVP-BEZ235. As showed in Fig. 3C and D, NVP-BEZ235 enhanced the conversion of LC3-II in both tested cell lines at an early stage. Given that the activation of autophagy almost exactly synchronized with the down regulation of CDK4 and CDK2 proteins, we speculated that autophagy induced by NVP-BEZ235 might play a role in cell cycle arrest. To prove our hypothesis, the effects of the autophagy inhibitors CQ and 3-MA were analyzed. Intriguingly, cotreatment with CQ or 3-MA for 12 hours rescued the expression of CDK2 and CDK4 (Fig. 3E-H).
NVP-BEZ235-induced G0/G1 cell cycle arrest and proliferation inhibition by autophagy activation in NB cells.
Autophagy is one of the two major key intracellular protein degradation routes within eukaryotic cells [21]. To further verify whether NVP-BEZ235-induced autophagy in uences NB cells, we tested the effect of inhibition of the autophagic process on CDK4 and CDK2 degradation, cell viability and cell cycle distribution via knockdown of autophagy-related genes. Accordingly, SK-N-MC cells that were transfected with shRNA for the knockdown of Beclin-1 were subsequently treated for 12 hours with vehicle or 200 nM NVP-BEZ235. CDK4 and CDK2 were analyzed by immunoblotting, cell viability was assessed using a CCK-8 assay, and cell cycle distribution was demonstrated with ow cytometry analysis. We found that blockade of autophagy (Fig. 4A) markedly reversed the CDK2 and CDK4 protein levels (Fig. 4B). Simultaneously, the viability of NVP-BEZ235-treated NB cells was obviously rescued (Fig. 4C), and the percentage of NVP-BEZ235-treated cells in the G0/G1 phase was strikingly increased (Fig. 4D) due to autophagy inhibition. These results demonstrated that the autophagy-mediated degradation pathway plays roles in NVP-BEZ235-induced G0/G1 cell cycle arrest and proliferation inhibition in NB cells.
NVP-BEZ235-induced CDK2 and CDK4 binding to the autophagy adaptor protein p62 in NB cells.
P62 (known as SQSTM1) is implicated in autophagic cargo recognition and is degraded in the nal stages of autophagy during autolysosome degradation [22]. A time-dependent reduction in p62 expression was corroborated in both NB cell lines treated with NVP-BEZ235 ( Fig. 5A and B), consistent with the upregulation of LC3 II levels ( Fig. 3C and D). To further validate the link between p62 and the CDK2 and CDK4 proteins, the levels of the p62 protein were measured in ATG7-deleted SH-SY5Y cells and Beclin-1-deleted SK-N-MC cells after treatment with 200 nM NVP-BEZ235 for 12 hours. As showed in Fig. 5C and D, there was concomitant restoration of p62 in NVP-BEZ235-treated autophagy-defective cells compared with the negative control. Moreover, immunoprecipitation assays indicated that the interactions between p62 and the CDK2 and CDK4 proteins were enhanced after exposure to NVP-BEZ235 in NB cells ( Fig. 5E and F). On this basis, the degradation of the CDK2 and CDK4 proteins induced by NVP-BEZ235 was restored when p62 was silenced in NB cells ( Fig. 5G and H). These ndings indicated that p62 is involved in autophagic degradation of the CDK2 and CDK4 proteins in NB cells.
Lysosomal cathepsin B was indispensable for autophagic degradation of the CDK2 and CDK4 proteins induced by NVP-BEZ235.
To further substantiate CDK2 and CDK4 protein degradation due to autophagic lysosomes, we performed immuno uorescence staining for CDK2 and CDK4 and the lysosomal membrane marker LAMP2A in SH-SY5Y cells and observed a strong colocalization between them after NVP-BEZ235 treatment (Fig. 6A and  B). Lysosomal cathepsins have been implicated in autophagic ux facilitation and in mediating substrate protein degradation [23] in the nal stage of autophagy. To validate the functional role of cathepsins in the degradation of CDK2 and CDK4 induced by NVP-BEZ235, lysosomal-speci c protease inhibitors E64d and pepstatin A were administered to SH-SY5Y cells. Importantly, suppression of cathepsins stabilized the CDK2 and CDK4 protein levels and reversed NVP-BEZ235-mediated degradation of CDK2 and CDK4 proteins (Fig. 6C). Herein, we mainly focused on the regulatory effects of cathepsin B (CTSB) and cathepsin D (CTSD) on CDK2 and CDK4 degradation. We observed that silencing CTSD expression by utilizing a siRNA approach did not rescue CDK2 and CDK4 degradation (Fig. 6D), whereas attenuation of CTSB by siRNA targeting resulted in a partial rescue of the NVP-BEZ235-induced CDK2 and CDK4 down regulation in SH-SY5Y cells (Fig. 6E). Altogether, these results indicated that CDK2 and CDK4 enter lysosomes and undergo degradation in a CTSB-dependent manner. Autophagy-lysosome system mediates NVP-BEZ235-triggered CDK2 and CDK4 degradation in NB xenograft tumors.
To determine the association of its autophagy with CDK2 and CDK4 protein degradation, an NB xenograft mouse model was generated. In agreement with the blockade of the G1 cell cycle in vitro, NVP-BEZ235 also reduced the expression of CDK2 and CDK4 in xenograft tumors, accompanied by a decrease in the autophagic protein marker p62 (Fig. 7A). Furthermore, the enhanced interactions between p62 and the CDK2 and CDK4 proteins were similarly assessed by immunoprecipitation in the NVP-BEZ235-treated group (Fig. 7B), which demonstrated that the autophagy-lysosome system mediates NVP-BEZ235triggered CDK2 and CDK4 degradation. Together, these results revealed that NVP-BEZ235-induced autophagy plays a critical role in CDK2 and CDK4 degradation and tumor proliferation inhibition in NB xenografts.

Discussion
The PI3K/Akt/mTOR signaling pathway is vital in maintaining cell growth, proliferation, and survival [24]. NVP-BEZ235, as a potent novel dual PI3K and mTOR kinase inhibitor, exerts strong anticancer effects via multiple mechanisms, including disturbing the cell cycle and differentiation and promoting apoptosis, autophagy, angiogenesis and EMT [17,25,26]. In the present study, we demonstrated that NVP-BEZ235 inhibited NB cell proliferation by inducing G0/G1 cell cycle arrest and that inducing G0/G1 cell cycle arrest was correlated with autophagic degradation of the CDK2 and CDK4 proteins.
We have shown that there was a remarkable reduction or gradual decrease in cyclin D1 and E1 protein levels in NVP-BEZ235-treated NB cells [19]. Previous evidence that the overexpression of cyclin D1, cyclin E1, CDK2 and CDK4 has been associated with poor prognosis and decreased survival in patients, and deregulated degradation appears to be responsible for their increased levels in various cancers [27][28][29]. Therefore, our results are signi cant and suggest that NVP-BEZ235 triggered a decrease in these proteins. Given the signi cance of these G1 phase-related proteins, we then probed the mechanism of NVP-BEZ235-mediated down regulation of CDK2 and CDK4 in NB. The results indicated that NVP-BEZ235 did not obviously change the mRNA levels but obviously attenuated the half-life of these proteins and promoted their degradation, as con rmed by the CHX-chase analysis.
In our work, we identi ed for the rst time that the autophagic process induced by NVP-BEZ235 is involved in the degradation of CDK2 and CDK4 in NB cells, as there is evidence that ablating autophagyrelated genes and autophagy-lysosomal inhibitors could obviously reverse CDK2 and CDK4 in NVP-BEZ235-incubated NB cells. With regard to CDK2 and CDK4, there are several potential molecular weight compounds that could in uence their regulation, mainly through elimination of CDK catalytic activity and induction of their degradation [30][31][32][33]. Disappointing results were observed in CDK inhibitors in terms of low speci city, serious side effect pro les and a lack of robust criteria for selecting patients [34]. Thus, targeting compounds to induce CDK-sustained degradation is inspiring and imperative. Our studies provide evidence for a novel mechanism involving autophagic degradation of CDK2 and CDK4, which is distinct from previous reports on the degradation of CDK2 and CDK4 [32,33,35,36].
Our subsequent experiments elaborate the particular modulatory mechanism of autophagy in NVP-BEZ235-induced CDK2 and CDK4 degradation. Autophagy is an evolutionarily conserved pathway that delivers substrates to lysosomes for degradation. P62/SQSTM1, as a cargo receptor protein, can bind directly to LC3 to promote the degradation of polyubiquitinated protein targets via macro-autophagy [37]. Obviously, p62 is required for constitutive autophagic degradation of CDK2 and CDK4, as illustrated in our results in which an enhanced interaction between p62 and the CDK2 and CDK4 proteins was observed and the degradation of CDK2 and CDK4 induced by NVP-BEZ235 was restored when p62 was silenced in NB cells. Analogous results regarding the correlation of p62 and the CDK2 and CDK4 proteins were observed in an NB xenograft mouse model in vivo. Moreover, intensive studies have shown that CTSB, as a notable lysosomal protease, is broadly up regulated in cancer and is essential for autophagic ux [38,39]. Our data also implicated CDK2 and CDK4 as downstream substrates of CTSB in the autophagy process, as supported by our experiments in which selective knockdown of CTSB resulted in the partial reversal of CDK2 and CDK4 degradation. Based on these ndings, we demonstrate that NVP-BEZ235 induces CDK2 and CDK4 binding to the autophagy adaptor protein p62 at the autophagosome site to enter the autolysosome, which are then degraded in a CTSB-dependent manner.

Conclusion
Taken together, these results clearly indicated that NVP-BEZ235 can mediate the autophagic degradation of CDK2 and CDK4 in a CTSB/p62-dependent manner. These events preserve the cells in G0/G1 status and alleviate growth. Therefore, NVP-BEZ235 can exhibit a strong anti-proliferation effect in NB cells (Fig. 7C). This nding provides the experimental basis that NVP-BEZ235 intervenes in the regulation of CDK2 and CDK4 at the posttranscriptional level. At the same time, these ndings also provide a new clue for the rational utilization of NVP-BEZ235 in neuroblastoma.

Consent to publish
Not applicable.

Availability of data and materials
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.     Figure  4B). (C) Cell viability was assessed using a CCK-8 assay. (D) Cell cycle distribution is shown by a ow cytometric analysis.   NVP-BEZ235 induces the interaction of p62/SQSMT1 with CDK2 or CDK4 and the subsequent autophagic degradation of CDK2 and CDK4 in a CTSB-dependent manner. In addition, CDK2 and CDK4