Preliminary experiment showed that DFMO (1–5 mM) inhibited Y79 cell growth in a dose-dependent manner (not shown). Therefore, we used 5 mM DFMO in the present study. The effects of 5 mM DFMO on Y79 cell growth were studied. After 72 h and 96 h, 5 mM DFMO inhibited Y79 cell growth 34.5 ± 3.97 and 53.5 ± 3.7%, respectively, compared to control without DFMO treatment (p < 0.001) (Fig. 1A). The growth suppression by 5 mM DFMO was completely restored by addition of 20 mM putrescine, indicating the specificity of the effect of DFMO on cell growth and the involvement of polyamines in Y79 cell growth. Viabilities of the cells treated with 5 mM DFMO were 96.6 ± 1.08 %, 96.67 ± 2.65 %, 95.88 ± 1.32 %, and 96.83 ± 1.58 % after 24 h, 48 h, 72 h, and 96 h respectively (not significant, compared to that observed at 0 h) (Fig. 1B). The results suggest that DFMO did not induce Y79 cell death. We next analyzed the effect of DFMO on ODC activities and polyamine levels in cells. DFMO (5 mM) treatment decreased ODC activities 96.7 ± 0.81 % and 99.6 ± 0.01 at 24 h and 48 h, respectively, compared to that observed at 0 h (p < 0.001) (Table 1). The levels of putrescine, spermidine, and spermine were significantly reduced at 24 h and 48 h (Table 2).

Cell cycle analysis was done using a flow cytometer. As shown in Fig. 2A and 2B, the percentages of G₁ phase cells were significantly increased at 48 h and 72 h. The percentages of G₂ phase cells were significantly reduced at 48 h and 72 h. These data suggest that DFMO treatment blocked cell cycle inside S phase in addition to G₁ arrest.

To study the mechanism of S and G₁ phase arrest, we studied the effect of DFMO on the expression of CDK inhibitors and pRb family proteins. The expression of p27 was markedly induced by DFMO treatment. The magnitude of induction at 24 h and 48 h was 235% and 268%, respectively, compared to that at 0 h (100%). Addition of putrescine rescued the effect of DFMO. In contrast, p16 and p21 were not significantly changed (Fig. 3A). We confirmed that Y79 cells did not express pRb using western blot analysis (Fig. 3B), and examined the expression of p107 and p130 proteins. Hypophosphorylated forms of p107 accumulated at 24 h and 48 h after DFMO treatment (Fig. 3B). On the contrary, the expression levels of each form of p130 protein did not change until 48 h after DFMO treatment (Fig. 3B).

Since p107 is known to be associated with E2F-4 2425, C-Myc 2627, and B-Myb 28, and suppress their transcriptional activities, we investigated whether DFMO induced binding of these proteins with p107. Whole cell extracts (100 μg protein) derived from cells treated with DFMO for 0 h, 24 h, and 48 h were immunoprecipitated with antibody against E2F-4 or C-Myc. Immunoprecipitated proteins were analyzed by western blotting using anti p107 antibody. The p107/E2F-4 complex was detected at 0 h, and DFMO treatment induced 1.6 ± 0.2- and 1.8 ± 0.2-fold increase in amount of the complex at 24 h and 48 h, respectively, compared to 0 h (p < 0.01) (Fig. 4A). Putrescine (20 μM) blocked the increase of the amount of p107/E2F-4 complex, indicating the specificity of the effect of DFMO on binding of E2F-4 with p107. Other E2F family proteins (E2F1, 2, 3, and 5) were detected in DFMO treated Y79 cells (Fig. 4B, lower panel), but the complex between p107 and these proteins were not detected (Fig. 4B, upper panel). Although C-Myc was detected in DFMO treated Y79 cells (Fig. 4C, right panel), the complex of p107 and c-Myc was not detected (Fig. 4C, left panel). Whole cell extracts (100 μg protein) derived from cells treated with DFMO for 0 h, 24 h, and 48 h were immunoprecipitated with antibody against p107. Immunoprecipitated proteins were analyzed by western blotting using anti B-Myb antibody and the complex of p107 and B-Myb was not detected (Fig. 4D, left panel). It has been reported that pRb and pRb family protein can directly regulate DNA replication by their association with MCM7/CDC47 29. Therefore, we analyzed the MCM7-immunoprecipitated whole cell extracts by western blotting using anti p107 antibody. The complex of p107 and MCM7 was detected, but it was not increased by DFMO treatment (Fig. 4E).

To clarify whether p107 is involved in repression of E2F by DFMO we performed gel shift assay using E2F consensus oligonucleotides. A fast migrating complex (Fig. 5A, b), and a complex migrating at slower rate (Fig. 5A, a) were apparent in DFMO treated Y79 cells. Complex b did not change significantly by DFMO treatment. The magnitudes of increase of complex a were 2.4 ± 0.3- and 2.4 ± 0.5-fold at 24 h and 48 h, respectively, compared to 0 h (p < 0.01) (Fig. 5A lanes 2 and 3). The binding activity in these complexes was due to E2F, as it was competed out by excess unlabeled wild type E2F consensus binding site oligonucleotides, but not by their mutants (Fig. 5A lanes 8 and 9). The addition of specific antibodies against E2F-4 and p107 to the reaction mixtures confirmed that the complex a contains E2F-4 and p107 (Fig. 5A lanes 6 and 7).

To further clarify whether p107 is involved in transcriptional repression of E2F by DFMO we performed reporter assay using artificially synthesized four tandem repeats (E2WTx4) of the adenovirus E2 enhancer, which has two typical E2F-binding sites 30. The plasmids, pE2WTx4 and pRL-CMV were cotransfected into Y79 cells. After 24 h culture media were changed to RPMI 1640 media with or without 5 mM DFMO or 5 mM DFMO and 20 μM putrescine. Cell lysates were prepared at indicated times after the addition of DFMO, and examined for luciferase activity to monitor the promoter activity. Luciferase activity was reduced by DFMO treatment to 76.7 ± 3.3%, 66.7 ± 5.5%, 55.7 ± 4.1% at 24 h, 48 h, and 72 h, respectively, compared to control (p < 0.02) (Fig. 5B). The addition of 20 μM putrescine completely restored the luciferase activity to the level of control, indicating the specificity of DFMO effect on E2F promoter activity.

The human retinoblastoma cell line Y79 and the human T cell lymphoma cell line Jurkat were obtained from the RIKEN cell bank (Tsukuba, Japan). Cells were cultured in RPMI 1640 medium with 10% fetal bovine serum (FBS) (JRH Biosciences, Lenexa, KS) in humidified air with 5 % CO₂ at 37°C. DFMO was kindly provided by Dr. P. McCann, Merrell Dow Research Center. The antibodies against p16, p27, p107, p130, E2F-1, E2F-2, E2F-4, E2F-5, B-Myb, cytochrome c, and c-Myc were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The antibody against p21 was obtained from BD Transduction Laboratories (Franklin Lakes, NJ). The antibodies against E2F-3 and MCM7 were obtained from Neo Markers (Fremont, CA). Propidium iodide and protein A-sepharose CL4B are obtained from Sigma. The antibody against pRb was obtained from New England BioLabs (Beverly, MA). The wild and mutant E2F oligonucleotides were obtained from Santa Cruz Biotechnology. [γ-³²P]ATP and T4 polynucleotide kinase were purchased from Amersham Biosciences (Tokyo, Japan). The reporter plasmid pE2WTx4-Luc was kindly provided by Dr. Kiyoshi Ohtani 38. Dual luciferase reporter assay system and pRL-CMV vector were purchased from Promega (Madison, WI). FuGene™6 Transfection Reagent was purchased from Roche.

Cells (10⁵ cells/ml) were treated with or without 5 mM DFMO, or 5 mM DFMO and 20 μM putrescine for the indicated times. At 0 h, 24 h, 48 h, 72 h, and 96 h, the cells in each treatment group were collected. Viable cells were counted using trypan blue exclusion method.

ODC activity was determined using DL- [l-¹⁴C]-ornithine as a substrate, as described previously 39. Polyamine analysis was determined using high-performance liquid chromatography (HPLC) as described 40.

At 0 h, 24 h, 48 h, and 72 h, the cells were collected and washed twice with PBS and fixed with chilled 70 % ethanol on ice for 30 minutes, treated with RNase A (0.5 mg/ml) for 20 minutes at 37°C, and stained with 50 mg/ml propidium iodide for 30 minutes at 4°C. Stained cells were analyzed using EPICS XL flow cytometer (Beckman Coulter, Tokyo, Japan) and WinCycle software (Phoenix Flow Systems, Inc, San Diego, CA).

For immunoprecipitation, whole cell extracts (100 μg protein) were incubated at 4°C for 16 h with specific antibodies prebound to protein A-sepharose CL-4B prewashed in IP buffer (50 mM HEPES, pH 7.4, containing 150 mM NaCl, 10 mM EDTA, 100 mM NaF, and 2 mM sodium orthovanadate). For western blotting, extracted proteins (100 μg) were separated by SDS polyacrylamide gel electrophoresis, transferred to nitrocellulose membrane, probed with diluted antibody (1:1,000), and visualized by ECL (Amersham Biosciences). Quantification of the bands was performed using NIH image Version 1.62 (Wayne Rasband, National Institutes of Health, U.S.A). The protein concentration was measured using a Bradford assay kit (Bio-Rad, Hercules, CA).

At 0 h, 24 h, and 48 h, the cells were collected and washed twice with PBS. Nuclear extracts were prepared as described 41. The sequences of the E2F oligonucleotides used were 5'-ATTTAAGTTTCG

CG

AT

One μg of pE2WTx4-Luc and 10 ng of pRL-CMV was co-transfected into Y79 cells (5 × 10⁵) using FuGene™6 Transfection Reagent according to the manufacturer's instruction (Roche). After 24 h culture media were changed to RPMI 1640 media with or without 5 mM DFMO, or 5 mM DFMO and 20 μM putrescine. Promoter activities were analyzed by dual luciferase reporter assay system according to the manufacturer's protocol (Promega).

Data from two independent experiments performed in triplicate are presented as mean ± SD unless otherwise indicated. Statistical analysis was performed using analysis of variance, with comparison of different groups by Fisher's partial least-squares difference (PLSD) test (Statview 4; Abacus Concepts, Berkeley, CA). The P value less than 0.05 was considered to be significant.