Our previous study showed that smenospongine (Figure 1) induced G1 arrest in K562 cells [12]. To investigate whether this compound has the same effect on other leukemia cells or not, cell cycle analysis for HL60 and U937 cells together with K562 cells was carried out. Various concentrations of smenospongine as DMSO solution were added to the randomly cultured K562, HL60, and U937 cells, respectively. After 24 h, the cells were collected for cell cycle analysis. As a result, smenospongine induced G1 phase arrest in K562 cells at 15 μM concentration (Figure 2A), which is consistent with our previous report [12]. G1 arrest was not found in HL60 and U937 cells, but sub-G1 accumulation, which indicates apoptosis, was observed. And the induction of apoptosis was shown to be dose-dependently (Figure 2B, 2C). In order to check whether this is attributed to the difference of the sensitivity for individual cells or not, we further analyzed with lower doses of smenospongine which did not induce apoptosis in both cells. However, no G1 accumulation was detected yet in both cells (data not shown). This suggests that smenospongine might affect different leukemia cells via the respective mechanism.

To demonstrate that the G1 phase arrest by smenospongine is attributed to p21 induction and inhibition of the phosphorylation of Rb, which located in the downstream of p21 and is known to play a key role in cell cycle progression from G1 to S phase, p21 expression and phosphorylation of Rb in K562 cells were examined. A 15 μM concentration of smenospongine was added to K562 cells. The cells treated for different length of time were collected, and the respective cell lysate was prepared for Western blotting analysis. The expression of p21 was detected after 12 h and increased in a time-dependent manner until 48 h (Figure 3). Accordingly, the underphosphorylated Rb protein (URb) was also detected after 24 h and highly increased until 48 h, indicating that the phosphorylation of Rb protein was inhibited by the smenospongine treatment for 24 to 48 h. This result suggests that the p21-Rb pathway might play an important role in the G1 phase arrest in K562 cells induced by smenospongine.

Since K562 cell is known as p53-negative cell [14], the p21 induction can be considered as p53-independent. Furthermore, the induction of p21 mRNA expression by smenospongine was also observed by RT-PCR analysis (data not shown). This suggested the possibility that the p21 expression was induced by trans-activating p21 promoter. In order to investigate whether smenospongine can trans-activate p21 promoter or not, we transfected K562 cells with pWWP, a human wild-type p21 promoter-luciferase fusion plasmid. Luciferase assay was carried out with the transfected cells after DMSO, smenospongine or trichostatin A (TSA) treatment. TSA was used as a positive control, since it was previously reported to activate p21 promoter [15,16]. As a result, the luciferase activity of the smenospongine-treated K562 cells was not significantly increased, while the TSA-treated cells showed about 200 fold enhancement over those DMSO-treated, indicating that smenospongine can not activate p21 promoter (Figure 4). Therefore, we postulate that smenospongine might act by stabilizing p21 mRNA to increase mRNA and protein expression and lead to G1 arrest in K562 cells via p21-Rb pathway. The mechanism of smenospongine to induce apoptosis in HL60 and U937 cells, including effect on p21 expression, is under investigation.

4–20% gel cassette was purchased from Daiichi Pure Chemicals Co., Ltd. DNA-Prep Reagents Kit was from Coulter Co. RPMI 1640 and DMEM Medium were from Nissui Pharmaceutical Co., Ltd. Polyvinylidene fluoride (PVDF) membrane was from Amersham Pharmacia Biotec. UK Ltd. Other reagents were from Sigma Co., Ltd. or Wako Pure Chemical Industries Co., Ltd unless stated specifically.

K562, HL60, and U937 cells, which were provided by RIKEN Cell Bank, were routinely maintained in the RPMI 1640 (for K562 and U937) or DMEM (for HL60) medium supplemented with 10% fetal bovine serum, kanamycin (100 μg/mL) and glutamine (0.44 mg/mL) at 37°C in a humidified atmosphere containing 5% CO₂.

Anti-Cip1/WAF-1/p21 antibody was purchased from Upstate Biotechnology Inc.; anti-human underphosphorylated Rb antibody was from BD Biosciences; anti-β-actin antibody was from Sigma Co., Ltd; anti-mouse horseradish peroxidase (HRP) conjugated antibody was from Nacalai Tesque Inc.

Smenospongine was isolated from the marine sponge Dactylospongia elegans (collected in Indonesia in 2001) as described previously [12]. Briefly, the dried marine sponge was cut into pieces and extracted with methanol. The resulting methanol extract was subjected to solvent partition to give hexane, 90% methanol, n-BuOH, and H₂O soluble portions. The 90% methanol soluble portion was separated by repeated SiO₂ column, reversed-phase column, and high-performance liquid chromatography (HPLC) to obtain smenospongine. The structure of smenospongine was identified by comparison of the mass and NMR data with those reported [17].

The cell suspension (2 × 10⁵ cells/2 ml/well) of K562, HL60 or U937 cells was placed in an 8-well plate and incubated for 24 h at 37°C under a 5% CO₂ atmosphere. The DMSO solution of the testing sample (10 μl) was added and further incubated for 24 h. The cells were harvested and collected, followed by staining with DNA-Prep Reagents Kit for 20 min. Then, the supernatant was removed by centrifugation, and the cell pellets were re-suspended in 500 μl of D-PBS (−). After being filtered with a 40–μm nylon mesh filter, the cell suspension was available for cell cycle analysis by flow cytometer (λex = 493 nm, λem = 630 nm). The analysis result was quantified by ModFit software (Verity Software, Topsham, ME).

The cell suspension (1 × 10⁶ cells/10 mL) of K562 cells was incubated in a flask together with 15 μM of smenospongine for the indicated length of time under a 5% CO₂ atmosphere at 37°C. The cells collected by centrifugation (1000 g for 3 min at 4°C) were washed with cold PBS and treated with lysis buffer (50 mM Tris-HCl, pH7.2; 1% NP-40; 0.25% sodium deoxycholate; 150 mM NaCl; 1 mM EDTA; 1 mM PMSF; 1% proteinase inhibitor cocktail) to furnish a cell lysate. Protein assay was carried out using a Bio-Rad protein assay kit. After boiling at 95°C for 5 min in the sample buffer (0.125 M Tris-HCl (pH 6.8), 10% 2-mercaptoethanol, 4% SDS, 10% sucrose, 5% bromophenol blue), the equal amounts of protein were subjected to SDS-Polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to PVDF membrane. The membrane was blocked with 5% milk TBS (Tween PBS) and exposed to anti-Cip1/WAF-1/p21, anti-human underphosphorylated Rb, or anti-β-actin antibody and then anti-mouse IgG HRP-conjugated antibody. The bound antibodies were visualized using an Enhanced Chemiluminescence (ECL) system.

K562 cells were transfected with a human wild-type p21 promoter-luciferase fusion plasmid, pWWP, by DEAE-Dextran (CellPhect Transfection Kit, Amersham Pharmacia Biotech). The cells were incubated at a density of 5 × 10⁴ cells/mL in a 12-well plate for 24 h. The 125 ng of the reporter plasmid DNA, pWWP, in DEAE-Dextran was added to perform transfection for 15 min. The transfected cells were further incubated for 24 h followed by DMSO, smenospongine or TSA treatment for another 24 h. Finally, the cells were collected for luciferase assay, which was carried out by use of the luciferase assay system (E1501, Promega) as described previously [18]. The luminescence was measured by using MICRO LUMAT Plus LB96V (BERTHOLD) and WING LOW software. The activation of p21 promoter was evaluated by the relative light intensity compared with that of the control (cells treated with DMSO only) [18].