Retinoblastoma gene‐independent G ₁ phase arrest by flavone, phosphatidylinositol 3‐kinase inhibitor, and histone deacetylase inhibitor

Abstract

In most human malignant tumors, retinoblastoma tumor‐suppressor gene (RB) product is inactivated by phosphorylation. Therefore, cancer preventive agents or molecular‐targeting agents can inhibit the tumor growth at G ₁ phase through RB reactivation. However, little is known about the effectiveness of RB reactivating agents against malignancies with mutated RB. We report here that chemopreventive agent flavone, phosphatidylinositol 3‐kinase (PI3K) inhibitor LY294002, and histone deacetylase (HDAC) inhibitor trichostatin A (TSA) also induce G ₁ phase arrest in malignant tumor cells with mutated RB. In human prostate cancer DU145 cells with mutated RB, flavone increased cyclin‐dependent kinase (CDK) inhibitors p21 and p27, and reduced cdk4 and cdk6, resulting in decrement of phosphorylated RB family proteins p130 and p107. LY294002 also dephosphorylated p107 and p130 proteins, whereas TSA dephosphorylated p130, but not p107. Furthermore, flavone induced G₁ phase arrest in both mouse embryo fibroblast (MEF) wild‐type and MEF RB ^−/− cells, but did not do so in RB, p107, and p130 triple‐knockout MEF cells. These results suggested that p130 and p107 contributed to G ₁ phase arrest by flavone in RB‐mutated cells. However, flavone induced tumor suppressor microRNA miR‐34a with reduction of E2F1 and E2F3, known to be downregulated by miR‐34a, raising the possibility that miR‐34a might partially contribute to G ₁ arrest by flavone. These results raise the possibility that RB reactivating chemopreventive agents or molecular targeting agents might also be effective against a variety of malignant tumor cells with mutant RB.

In most human malignant tumor cells, RB is known to be inactivated by mutations or phosphorylation. When RB is inactivated by phosphorylation, cancer preventive agents,1, 2, 3, 4 or most of the molecular targeting agents, such as MEK inhibitor, HDAC inhibitor, PI3K inhibitor, imatinib, and gefitinib,5, 6, 7, 8, 9 reactivate the RB function by dephosphorylation, resulting in G₁ phase arrest. However, when RB is inactivated by mutations in various malignant tumors, such as retinoblastoma, glioma, osteosarcoma, colon, breast, bladder, or prostate cancer cell lines,10, 11, 12, 13 it has not been reported whether or not these molecular‐targeting agents are also effective.

We previously found that flavone (2,3‐didehydroflavan‐4‐one) could induce the expression of CDK inhibitor p21 through its promoter and convert RB protein to the active form, resulting in G₁ phase arrest in human lung cancer A549 cells harboring wild‐type RB.1 We therefore examined whether flavone could cause G₁ arrest even in RB mutated cancer cells. We additionally examined whether other RB reactivating G₁ arrest inducers, PI3K inhibitor LY294002 and HDAC inhibitor TSA, could similarly cause G₁ arrest in RB mutated cancer cells.

In this report, we found that G₁ arresting agents such as flavone, PI3K inhibitor LY294002, and HDAC inhibitor TSA caused G₁ arrest in several RB mutated malignant tumor cells, with activation of RB family proteins, p107 and p130/RB2, and/or induction of a tumor‐suppressive microRNA (miRNA), miR‐34a.

Materials and Methods

Reagents

Flavone was purchased from Nacalai Tesque (Kyoto, Japan), TSA was purchased from Wako Pure Chemical Industries (Osaka, Japan), and LY294002 was purchased from Cell Signaling Technology (Beverly, MA, USA)

Cell culture

Human prostate cancer cell line DU145 was purchased from ATCC (Manassas, VA, USA). Human bladder cancer HTB9 cell line was a gift from Dr. R. Takahashi (Kyoto University, Kyoto, Japan).12 Human osteosarcoma Saos‐2 cell line was a gift from the Department of Orthopedic Surgery of Wakayama Medical College (Wakayama, Japan). The DU145 and Saos‐2 cells were maintained in DMEM supplemented with 10% FBS, 4 mM glutamine, 100 U/mL penicillin, and 100 μM streptomycin. HTB9 cells were maintained in RPMI‐1640 with 10% FBS, 2 mM glutamine, 100 U/mL penicillin, and 100 μM streptomycin. The MEF cells were maintained as previously described.14 All cells were incubated at 37°C in a humidified atmosphere containing 5% CO₂.

Analysis of in vitro cell growth

The DU145 cells were incubated with each agent at the concentrations indicated. The cells were treated with a ViaCount kit (Guava Technologies, Hayward, CA, USA), and viable cell numbers were measured with a Guava Easy‐Cyte plus flow cytometer (Guava Technologies) according to the manufacturer's instructions.

Cell cycle analysis

Cells were incubated with flavone at the concentrations indicated, and harvested. The cells were fixed in 0.1% TritonX‐100 containing RNase (Sigma, St Louis, MO, USA), and nuclei were stained with 100 μM of propidium iodide. Flow cytometry analysis was carried out with a FACSCalibur (Becton Dickinson, Franklin Lakes, NJ, USA). The data were analyzed with Modifit software (Becton Dickinson).

Western blot analysis

The DU145 cells were incubated with flavone at the concentrations indicated, and harvested. The cell lysate was prepared as described previously.15 Rabbit polyclonal anti‐p21, anti‐p27, anti‐E2F1, anti‐E2F3, anti‐p107, and anti‐CDK4 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and mouse monoclonal anti‐Rb2 (BD Biosciences, San Diego, CA, USA), anti‐CDK6 (Cell Signaling Technology), anti‐thymidylate synthase (Abcam, Cambridge, UK) and anti‐β‐actin (Sigma) antibodies were used as the primary antibodies. The blots were incubated with the appropriate HRP‐conjugated secondary antibody (GE Healthcare, Piscataway, NJ, USA), and signals were detected with Chemilumi‐One (Nacalai Tesque).

RNA analysis

Cells were incubated with flavone at the concentrations indicated, and harvested. Total RNA, containing miRNA, from the cells was extracted using a mirVana miRNA Isolation kit (Applied Biosystems, Foster, CA, USA), according to the manufacturer's instructions. For quantitative real‐time RT‐PCR, total RNA (2 μg) was reverse‐transcribed to cDNA in 20 μL reaction volume, using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems) according to the manufacturer's instructions. TaqMan probes for p21, p27, CDK4, CDK6, thymidylate synthase, and GAPDH were purchased from Applied Biosystems. The expression levels of mRNAs were quantified using Applied Biosystems 7300 Real‐Time PCR system according to the manufacturer's instructions. Total RNA (10 ng) was reverse‐transcribed to cDNA in 15 μL reaction volume using a TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems). The expression levels of miRNAs were quantified using the 7300 Real‐Time PCR system (Applied Biosystems) according to the manufacturer's instructions.

Statistical analysis

Data were expressed as the mean ± SD of three determinations. Statistical significance was assessed by Student's t‐test. A value of P < 0.05 was considered significant compared with untreated controls.

Results

Flavone causes G₁ phase arrest in human malignant tumor cells with mutated RB

Human prostate cancer DU145 cells have a mutation in exon 21 of the RB gene.11 The inhibitory effects of various concentrations of flavone on DU145 cells were detected by a Guava ViaCount system. Flavone inhibited the growth of DU145 cells in a dose‐ and time‐dependent manner 24–72 h after treatment (Fig. 1a). We next analyzed the effect of flavone on the cell cycle, and found that flavone markedly induced G₁ phase arrest in DU145 cells (Fig. 1b,c). We carried out further similar experiments using other RB mutated cell lines, HTB9 and Saos‐2, and found that flavone similarly induced G₁ phase arrest in these cells (Fig. 1d,e). However, flavone at 200 μM and/or 400 μM slightly increased the ratio of S and/or G₂/M phase (Fig. 1b,d,e).

Figure 1.

Flavone inhibits the growth of retinoblastoma gene (RB) mutated cells and causes G ₁ phase arrest. (a) Human prostate cancer DU145 cells were treated with the indicated concentrations of flavone for 24, 48, and 72 h. After incubation, viable cell numbers were determined using Guava EasyCyte plus. DU145 cells (b), human bladder HTB9 cells (d), and human osteosarcoma Saos‐2 cells (e) were treated with the indicated concentrations of flavone for 24 h. Cells were subjected to cell cycle analysis using FACSCalibur. Data represent the mean ± SD of three experiments. *P < 0.05, **P < 0.01 versus control. Representative histogram patterns from DU145 cells are shown in panel (c).

Effects of flavone on G1 phase regulatory proteins

We then investigated the mechanism of G₁ arrest by flavone, and found that flavone treatment for 24 h strongly induced the expressions of CDK inhibitors p21 and p27, and reduced those of thymidylate synthase, cdk4, and cdk6 in DU145 cells (Fig. 2a). This regulation was also observed at mRNA level except in p27 (Fig 2c). Interestingly, flavone converted RB family proteins, p107 and p130, to their hypophosphorylated active forms (Fig. 2b). We then carried out a time‐course study, and found that 100 μM flavone induced expression of p21 and p27 at 6 h, and reduced those of cdk4 and cdk6 at 12 h (Fig. 2d). Furthermore, flavone dephosphorylated p107 and p130 proteins at 6–12 h, and downregulated thymidylate synthase at 24 h (Fig. 2d,e).

Figure 2.

Effect of flavone on G ₁ cell cycle regulatory proteins in DU145 cells. (a,b) DU145 cells were treated with the indicated concentrations of flavone for 24 h, and G ₁ cell cycle regulatory proteins were analyzed by Western blotting. The blot of β‐actin is a loading control. (c) DU145 cells were treated with the indicated concentrations of flavone for 24 h, and real‐time RT‐PCR was carried out. The internal control was GAPDH. Data represent the mean ± SD of three experiments. *P < 0.05, **P < 0.01 versus control. (d,e) Six, 12, and 24 h after treatment with DMSO alone or 100 μM flavone, Western blot analysis was carried out. The blot of β‐actin is a loading control.

Molecular targeting agents induce cell cycle arrest at G₁ phase in RB mutated cancer cells

We also examined the effects of PI3K inhibitor LY294002 and HDAC inhibitor TSA on DU145 cells. As shown in Figure 3a, LY294002 and TSA induced G₁ phase arrest. Interestingly, LY294002 dephosphorylated p107 and p130 proteins, whereas TSA dephosphorylated p130, but not p107 (Fig. 3b).

Figure 3.

Molecular targeting agents cause cell cycle arrest at G ₁ phase. (a) DU145 cells were treated with various concentrations of LY294002 or trichostatin A (TSA) for 24 h. The cells were subjected to cell cycle analysis using FACSCalibur. Data represent the mean ± SD of three experiments. **P < 0.01 versus control. (b) Western blotting was carried out with lysate of cells treated for 24 h with 25 μM LY294002 or 50 nM TSA. The blot of β‐actin is a loading control.

RB family proteins, p107 and p130 are engaged in G₁ phase arrest by flavone

To determine whether dephosphorylation of p107 and p130 contributes to the G₁ phase arrest by flavone, we examined the effect of flavone on an RB, p107, and p130 triple‐knockout MEF cell line (MEF tko). As shown in Figure 4, flavone did not induce G₁ phase arrest in MEF tko cells but caused G₁ phase arrest in MEF wild‐type (wt) and MEF RB^−/− cells, whereas flavone at 100 and/or 200 μM slightly increased the ratio of S and/or G₂/M phase. The results suggest that the RB family proteins are responsible for G₁ arrest by flavone in MEF RB^−/− cells. Although MEFs are different from the three malignant tumor cell lines used in this study, these results with MEF cells support the possibility that p130 and/or p107 might be responsible for the flavone‐induced G₁ arrest in RB mutated malignant tumor cells.

Figure 4.

Flavone does not cause G ₁ arrest in retinoblastoma gene (RB), p107, and p130 triple‐knockout (tko) mouse embryo fibroblast (MEF) cells. The MEF wild‐type (wt) cells (a), MEF RB ^−/− cells (b), or MEF tko cells (c) were treated with the indicated concentrations of flavone for 24 h. Cells were then subjected to cell cycle analysis using FACSCalibur. Data represent the mean ± SD of three experiments. *P < 0.05, **P < 0.01 versus control.

Flavone induces expression of miR‐34a

Flavone reduced the expression of transcription factors E2F1 and E2F3 (Fig. 5a). It is known that miR‐34a reduces E2F1 and E2F3 proteins.16 Therefore, we examined the effect of flavone on miR‐34a expression, and found that flavone increased miR‐34a (Fig. 5b). To evaluate the contribution of miR‐34a induction on flavone‐induced G₁ phase arrest, we examined the effect of miR‐34a knockdown using miR‐34a inhibitor. As shown in Figure S1(a), miR‐34a knockdown could slightly but significantly reduce the flavone‐induced G₁ phase arrest. We next examined the effect of miR‐34a knockdown using miR‐34a inhibitor on the reduction of E2F1 and E2F3 by flavone in DU145 cells, and found that the reduction of E2F1 and E2F3 by flavone was slightly suppressed by the miR‐34a inhibitor (Fig. S1b). We therefore consider that RB‐independent G₁ arrest by flavone is mainly due to activation of RB family proteins, and the contribution of upregulation of miR‐34a is rather weak.

Figure 5.

Effect of flavone on miR‐34a expression in DU145 cells. (a) DU145 cells were treated with the indicated concentrations of flavone for 24 h, and E2F1 and E2F3 proteins were analyzed by Western blotting. The blot of β‐actin is a loading control. (b) DU145 cells were treated with the indicated concentrations of flavone for 24 h, and real‐time RT‐PCR was carried out. The internal control was RNU48. Data represent the mean ± SD of three experiments. *P < 0.05 versus control.

Discussion

It is well known that cancer preventive agents or most clinically used molecular targeting agents ultimately reactivate the RB function, suppressing tumor growth.1, 2, 3, 4, 5, 6, 7, 8, 9 However, no‐one has clarified whether or not G₁ arresting agents are effective against malignant tumors with mutations of RB. We therefore examined whether flavone, PI3K inhibitor LY294002, and HDAC inhibitor TSA, known to reactivate RB function with the G₁ phase arrest in RB wild‐type malignant tumor cells, induced G₁ phase arrest in three RB mutated malignant tumor cells. We found that all of the agents clearly caused G₁ phase arrest.

As described, we previously reported that flavone could cause G₁ phase arrest by inducing the expression of p21 in human lung cancer A549 cells with wild‐type RB.1 In the present study, we additionally found that flavone upregulated the expression of p27 as well as p21 with downregulation of cdk4 and cdk6, resulting in G₁ phase arrest in DU145 cells with mutated RB. We then examined the status of RB family proteins, and found that flavone converted p130 and p107 to the active forms, consistent with reports that other food factors, such as silibinin and inositol hexaphosphate, similarly induced G₁ phase arrest.17, 18 We additionally studied the significance of the activation of RB family proteins, and clarified that flavone could not cause G₁ phase arrest in MEFs without RB, p130, and p107. Taken together, we conclude that flavone, PI3K inhibitor LY294002, and HDAC inhibitor TSA can cause G₁ phase arrest even in RB mutant malignant tumor cells, possibly through activation of RB family proteins.

In the present study, flavone reduced the levels of cdk4, cdk6, E2F1, and E2F3 proteins, and induced expression of tumor suppressive miRNA miR‐34a. The latter is known to induce G₁ phase arrest by downregulating cdk4, cdk6, E2F1, and E2F3 protein levels.16, 19, 20, 21, 22 Therefore, we postulate that induction of miR‐34a might also partially contribute to G₁ phase arrest by flavone in addition to activation of RB family proteins.

In conclusion, we report here that flavone, PI3K inhibitor LY294002, and HDAC inhibitor TSA induce G₁ phase arrest in RB mutant malignant tumor cells. Although mutations of the RB gene are common events in human cancer, mutations of other members of the RB family, p107 and p130, rarely occur,23, 24 and malignant tumor cells with mutations of all three RB family genes have not been described. This raises the possibility that all RB reactivating chemopreventive agents or molecular targeting agents could be used against a variety of malignancies with mutated RB.

Disclosure Statement

The authors have no conflict of interest.

Abbreviations

CDK

Cyclin‐dependent kinase

HDAC

Histone deacetylase

MEF

Mouse embryo fibroblast

miRNA

microRNA

PI3K

Phosphatidylinositol 3‐kinase

RB

Retinoblastoma gene

TSA

Trichostatin A

Supporting information

Fig. S1. Effect of miR‐34a inhibitor on G ₁ phase arrest and the reduction of E2F1 and E2F3n induced by flavone in DU145 cells.

(Cancer Sci 2012; 103: 2139–2143)