Flavagline analog FL3 induces cell cycle arrest in urothelial carcinoma cell of the bladder by inhibiting the Akt/PHB interaction to activate the GADD45α pathway

Synthetic PHB ligands and FL3-conjugated compound 6 (an amino linker) were provided from The Laboratory for Therapeutic Innovation, University of Strasbourg (Strasbourg France). All of the drugs were stored at − 20 °C as solids, and were dissolved in dimethylsulfoxide (DMSO) before use.

UCB cell lines T24, BIU and 5637, and normal bladder uroepithelial cell line SV-HUC-1 were cultured at 37 °C with 5% CO₂ in RPMI 1640 media supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 μg/ml).

Cell viability was determined using the Cell Counting Kit-8 (CCK8) Kit (Dojindo, Kumamoto, Japan). After incubation with concentrations of FL3 and paclitaxel for indicated hours, absorbance was measured at 450 nm according to the manufacture’s introduction. Cell viability was expressed as the percentage of absorbance of cells treated with FL3 or paclitaxel compared with cells treated with DMSO.

Cells were plated as a density of 500 cells/well in 6-well plates and incubated overnight. Then different concentrations of FL3 were added to each well followed by incubation for 4–5 days. After a growth period, colonies were fixed in methanol for 30 min and stained with 0.1% crystal violet for 1 h.

Cells were harvested and lysed, followed by isolation of the supernatant and measurement of the protein concentration using the bicinchoninic method. Then the protein extracts were separated on 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels, and electrotransferred to a PVDF membrane at 250 mA for 2 h at room temperature. Then, the membrane was blocked in 5% bovine serum albumin, and incubated overnight at at 4 °C with primary antibodies (Akt, phospho-Akt Ser 473, prohibitin-1, phospho (Ser/Thr)-Akt substrate, and cyclin-dependent kinase antibodies were from Cell Signaling Technology, Danvers, MA, USA; REA antibody was from Millipore, Billerica, MA, USA; histone H3, Cox IV, GADD45α, and GAPDH antibodies were from Proteintech Group Inc., Rosemont, IL, USA). The membrane was washed and incubated with secondary antibody for 1 h at room temperature. The signal was measured using the ECL detection system (Tanon, Shanghai, China).

Cells (1 × 10⁸) were harvested after incubation with 0.5 μM FL3 or DMSO for 24 h. Then the cells were lysed with immunopreciptation (IP) lysate buffer (Beyotime, Shanghai, China) and the protein extracts were isolated. A volume of 20 μl Protein G-agarose beads/tube were washed three times with buffer, followed by incubation with anti-PHB antibody, anti-phospho-Akt antibody or IgG antibody at 4 °C with rotation for 3 h. Equal amounts of protein (21 μg) were incubated with beads overnight at 4 °C. After incubation, The beads were re-suspended in 1× LDS Sample Buffer and boiled for 5 min. Then the proteins were resolved on 10% SDS-PAGE gels for Western blot analysis.

FL3-Affigel beads were generated according to the protocol by Thuaud [16]. Briefly, a volume of 1 ml packed Affilgel 10 (Bio-Rad Laboratories, Hercules, CA, USA) was washed on a glass filter with acetonitrile (ACN) and added to a solution of FL3-conjugated compound 6 (3 mg) in ACN (0.8 ml) and triethylamine (20 μl). This suspension was gently rotated for 12 h at room temperature. Ethanolamine (50 μl) was added and after 4 h, the resin was thoroughly washed with ACN, ethanol and water. Coupled AffiGel-10 agarose beads were stored in PBS containing 0.02% NaN₃ at 4 °C until further use. PBS was used as the negative control (NC).

T24 cells (1 × 10⁸) cells were harvested to obtain total protein extracts. 30 μl FL3- or NC-Affigel beads/tube were incubated with total cell extracts at 4 °C with gentle rotation overnight. After incubation, the bound mixtures were re-suspended in 1× LDS buffer and subjected to SDS-PAGE for Western blot analysis with primary PHB antibody.

Approximately 1 × 10⁸ cells were harvested after incubation with DMSO (control) or FL3 0.5 μM for 24 h. Using the Nuclear and Cytoplasmic Protein Extraction Kit and Cell Mitochondria Isolation Kit (Beyotime), proteins of the nucleus, cytoplasm, and mitochondria were separately extracted according to the manufacturer’s introductions. Then the extracts were subjected for Western blot analysis.

The Lenti-Pac HIV Expression Packaging Kit (GeneCopoeia, Rockville, MD, USA) was used to construct stable low-expressing PHB T24 cell lines according to the manufacturer’s protocol. Stable PHB-silenced cells were selected with 0.1% puromycin (Invitrogen, Darmstadt, Germany).

The percentage of apoptotic cells was determined by using an Annexin V-PE apoptosis detection kit (KeyGen Biotech, Jiangsu, China) according to the manufacture’s introduction. In brief, a total of 10⁵ T24 and BIU cells were harvested after incubation with FL3 for 24 h, then stained these cells with Annexin V and PE subsequently. Finally, the cells were subjected to apoptosis analysis with the ACEA NovoExpress System (ACEA Biosciences Inc., CA, USA).

The Cell Cycle Assay Kit (Byeotime) was used to analyze the cell cycle distribution. According to the manufacture’s introduction, FL3-treated or control T24 cells were harvested and dealed with PI, RNase A, Triton X-100 subsequently, followed by cell cycle analysis with the MoFlo XDP Flow Cytometry System (Beckman Coulter, CA, USA).

Four-week-old Balb/c nude mice were purchased from Charles River Laboratories (Beijing, China). All of the animal experiments were approved by the Animal Ethics Committee of Sun Yat-sen University (Guangzhou, China). A total of 5 × 10⁵ cells were subcutaneously inoculated into the right flank of mice. When the tumor sizes reached to about 4–6 mm³, the mice were randomly divided into four groups with seven mice each, and intraperitoneally injected with FL3 (2 mg/kg, 5 mg/kg), palicetaxel (10 mg/kg, positive control), or DMSO (NC) every 2 days for a total of eight times. Body weights and tumor volume were measured and recorded at the time of administration. At the end of the study, the tumors and main organs including livers, kidneys, lungs, and hearts were removed for further immunohistochemistry experiments. The following formula was used to measure tumor volumes: tumor volume = 1/2 L × W², where L represents length and W denotes width.

The removed organs and tumors were fixed in formalin and then embedded in paraffin. Sections (4 μm thick) were cut and stained with hematoxylin and eosin (H & E). For further immunohistochemical analysis, sections were de-paraffnized in xylene, hydrated in graded alcohol, and blocked in 3% hydrogen peroxide to inhibit endogenous peroxidase activity. Antigen retrieval was completed by incubating the slides for 5 min in Ethylene Diamine Tetraacetic Acid (EDTA) buffer (pH 8.0). After incubation with 10% goat serum, the slides were incubated with anti-PHB antibody (1:400; Santa Cruz) overnight at 4 °C, followed by incubation with secondary goat anti-rabbit antibody at 37 °C for 30 min. Then, the slides were stained with DAB staining solution for less than 5 min, and re-stained with hematoxylin for 1 min followed by polarization for less than 10 s.

All statistical analyses were performed with IBM SPSS Statistics 19.0 (SPSS Inc., Chicago, IL, USA). All data both in vitro and in vivo are presented as mean ± S.D. and assessed by two-detailed Student’s t-test with three independent experiments. P values of < 0.05 was considered statistically significant.

To determine if Flavaglines had anti-tumor effects in UCB cells, we measured the cell viability of UCB T24 cells after treatment with various PHB ligands for 24 h. As shown in Fig. 1a, all of PHB ligands used decreased cell viability of T24 cells, in which FL3 exhibited the most potent effect to inhibit cell growth.

Next, we explored the inhibitory effect of FL3 on UCB cell proliferation in vitro. To this end, the cell viability assay was performed in three UCB cell lines after FL3 treatment. Paclitaxel, a well-known anticancer agent for some human cancers including UCB, was used as a positive control. As shown in Fig. 1b, FL3 dominantly inhibited the cell viability of three UCB cell lines in a time- and dose-dependent manner. Compared to paclitaxel, FL3 presented less cytotoxicity to normal bladder uroepithelial SV-HUC-1 cells (Fig. 1c). The colony formation assay showed that FL3 treatment significantly inhibited the colony formation ability of T24 and BIU cells compared to the control (Fig. 1d). Taken together, these results showed that FL3 suppressed UCB cell proliferation in vitro.

Because PHB is the direct target of Flavaglines in Jurkat T cells [12], FL3-Affigel-10 beads were used to determine whether FL3 targets PHB protein in UCB cell lines using immunoprecipitation assay. As shown in Fig. 2a, PHB protein was pulled down by the FL3-Affigel-10 beads, but there was no PHB binding to the unconjugated-Affigel-beads, indicating that FL3 binds to PHB protein.

A recent study showed that FL3 treatment could influence the localization of PHB in cardiomyocytes by promoting PHB translocation from nucleus to mitochondria [13], however, it remains to be determined in UCB cells. We first measured the total PHB protein levels of FL3-treated T24 cells, and the result showed that the protein levels of PHB was slowly decreased by FL3 treatment (Fig. 2b). Although ubiquitously existed in multiple cellular compartments and each location stands for different functions, PHB mainly distributes in the mitochondria and participates in the stability of mitochondrial structure and functions [4, 7, 17‐19]. We detected the change of PHB protein levels in main cellular fractions after FL3 treatment, and the results revealed that mitochondrial PHB levels was dramatically decreased, whereas nuclear PHB expression and cytoplasmic PHB expression were not significantly changed (Fig. 2c), suggesting that FL3 dose not promote PHB translocation from mitochondria to nucleus and cytoplasm.

PHB is a substrate of Akt, it is activated and transferred to the mitochondria after phosphorylation by Akt at Thr-258 site [11, 20]. FL3 might impede PHB localization within the mitochondria by blocking the interaction of PHB with Akt, which subsequently inhibits the phosphorylation process of PHB. To test this, total proteins of FL3-treated T24 cells were subjected to immunoprecipitation assay with primary PHB or phospho-Akt (Ser 473) antibody followed by western blot analysis. When proteins were immunoprecipitated with PHB antibody, FL3 treatment resulted in significant decrease of pulled-down PHB, Akt and phospho-Akt (p-Akt) compared to the control (Fig. 2d). If the proteins were immunoprecipitated with p-Akt antibody, pulled-down Akt and phospho-Akt levels were not significantly different, whereas the pulled-down PHB was significantly decreased compared to control (Fig. 2e).

Due to the shortage of available antibody recognizing phosphorylated PHB protein, we indirectly detected phosphorylated levels of PHB with a phospho-(Ser/Thr) Akt substrate (PAS) antibody, which can preferentially recognize peptides and proteins containing phospho-Ser/Thr preceded by Lys/Arg at positions − 5 and − 3 in a manner largely independent of other surrounding amino acids. As shown in Fig. 2d, a protein between 25 and 35 kDa was pulled down by PAS, and its expression was clearly reduced by FL3 compared to the control. This result indirectly showed that FL3 treatment decreased PHB phosphorylated levels.

Taken together, these data showed that FL3 inhibited the interaction of Akt and PHB, which in turn decreased PHB phosphorylation and localization in the mitochondria.

A recent study has revealed the apoptotic activity of FL3 in certain cancer cells, such as HL60 and Hela cells, by triggering the translocation of Apoptosis Inducing Factor (AIF) and caspase-12 to the nucleus [16]. We also examined the apoptotic activity of FL3 in UCB T24 and BIU cells by flow cytometry assay. However, our results did not show a similar result that FL3 treatment did not significantly increase the apoptotic number of both T24 and BIU cells, suggesting that FL3 does not induce cell apoptosis in UCB cells (Additional file 1: Figure S1) .

It has been shown that FL3 affects cell cycle distribution in Jurkat T cells [12], so we determined whether it has similar effects in UCB cells. As Fig. 3a shown, FL3 resulted in cell cycle arrest in the G2/M phase by increasing the percentage of cells in the G2 phase from 6.83% in control cells to 17.15% and 26.37% in FL3 -treated T24 cells, and the percentage of cells in S phase nonsignificantly ranged from 25.98% to 23.15% and 29.60%. In addition, protein expressions of cdc2 and cyclin B1, which are implicated with G2-M transition [21], were dramatically decreased (Fig. 3b). In contrast, protein expressions of G1-S transition related cyclins such as CDK2, CDK4, and cyclin D1 were unchanged. Also, several key regulators of cell cycle including p21, p27 and REA (repressor of estrogen receptor activity) [22‐24] were also analyzed, the results showed that their protein expressions were not significantly changed after FL3 treatment (Fig. 3b).

To clarify the action of FL3 on PHB signaling, we knocked down PHB expression in T24 cells to determine whether PHB knockdown would mimic the inhibitory effect of FL3 on cell cycle progression. Flow cytometry assay was performed and the results showed that PHB knockdown increased the percentage of cells in the G2 phase from 8.45% to 24.53%, accompanying with slight and insignificant increase of cells in the S phage from 21.55% to 29.87% (Fig. 3c). Meanwhile, the protein levels of cdc2 and cyclin B1 were decreased, whereas that of CDK2, CDK4, cyclin D1, p21, p27 and REA were not significantly changed (Fig. 3d). These data suggests that knockdown of PHB mimics the inhibitory effect of FL3 on cell cycle progression.

To identify the common pathways regulated by FL3 treatment or PHB knockdown, microarray analyses of mRNA expression were performed in FL3-treated or PHB knockdown T24 cells. Totally regulated genes were compared and presented as shown in Fig. 5e, a total of 181 genes were commonly regulated in both analyses. Since FL3 and PHB knockdown inhibits cell cycle progression in the G2/M phase, we tried to explore the underlying mechanism of this effect. A total of 84 cell cycle-related genes [25] were compared with each other in both analyses. Significantly regulated genes (more than 1.7-fold) were selected out in Additional file 2: Table S1, 28 genes were significantly regulated by FL3. Comparing the 28 genes in microarray analysis of PHB knockdown cells, 11 genes were commonly regulated by FL3 and PHB knockdown (Fig. 3f), among of which GADD45α (the growth arrest and DNA damage-inducible alpha), a checkpoint gene during the G2-M transition [26], was verified as the only significantly regulated gene (5.26 vs. 1.76-fold). Upregulation of GADD45α was further validated by western blotting and quantitative real time-PCR (qRT-PCR) (Fig. 3g-h). The results indicated that FL3-induced cell cycle arrest might be GADD45α-dependent.

To determine whether FL3-induced cell cycle arrest is GADD45α-dependent, a rescue experiment was performed by repressing GADD45α expression with siRNA in T24 cells. Flow cytometry assay demonstrated that FL3 treatment did not significantly change the percentage of cells in the G2 phase (from 7.70% to 11.36%) in GADD45α-silenced cells, whereas the percentage of cells in this phase increased from 10.58% to 24.42% in the control cells (Fig. 4a, b). Thus, repression of GADD45α rescued the inhibitory effect of FL3 on cell cycle progression. In addition, repression of GADD45α did not change the protein levels of PHB, cdc2 and cyclinB1 (Fig. 4c). Moreover, repression of GADD45α attenuated the inhibitory effect of FL3 on the colony formation ability of T24 cells (Fig. 4d), indicating that GADD45α is critical for UCB cell proliferation. Taken together, FL3-induced cell cycle inhibition is GADD45α-dependent.

To further verify the inhibitory role of FL3 on cell proliferation in vivo, xenograft tumor models were constructed (Fig. 5a). Both FL3 and paclitaxel treatments significantly inhibited the volumes and weights of tumors compared to the control (Fig. 5b, c). Furthermore, immunohistochemistry assay presented that FL3 treatment resulted in slightly decreased expression of PHB protein compared to the control (Fig. 5d). In contrast, the expression of GADD45α was slightly elevated. Meaningfully, there was no significant body weight loss in FL3-treated mice compared to the control (Fig. 5e). H&E staining of the livers, lungs, hearts, and kidneys showed no major organ-related toxicities from FL3 treatment compared to the control group (Fig. 5f).