RPMI-1640 medium, nitroblue tetrazolium (NBT), ATRA, tamoxifen and α-tocopherol were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). Dimethyl sulfoxide (DMSO), 1α,25-dihydroxy vitamin D₃ (VD3), and sodium butyrate were purchased from Wako Chemicals (Osaka, Japan). Fluorescein isothiocyanate (FITC)-labeled anti-CD11b antibody was obtained from Becton-Dickinson Immunocytometry Systems (San Jose, CA, USA). BMS195614 was purchased from Tocris Bioscience, R&D System Co. (Minneapolis, MN, USA). PD98059 was obtained from Calbiochem (La Jolla, CA, USA).

Leukemic bone marrow specimen was collected at diagnosis, after the patient gave his written informed consent for sample collection in accordance with institutional policy. A 45-year-old Japanese male was admitted to our hospital, presenting with purpura. Laboratory data on admission were WBC 8,720/μl, blasts 63.5%, and platelets 15,000/μl. Bone marrow aspiration revealed hyperplastic marrow cells (nucleated cell count 1,232,000/μl), 94.4% of which were blasts. APL was diagnosed on the basis of bone marrow morphology and the standard cytogenetic translocation, t(15;17), in 97% of cells by FISH analysis.

The HL-60 cell line, derived from an AML patient, and NB4 and HT93 promyelocytic leukemia cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum and 80 μg/ml gentamicin at 37°C in a humidified atmosphere of 5% CO₂ in air (9).

Heparinized bone marrow aspirations were diluted with RPMI-1640 medium supplemented with 10% fetal bovine serum, overlaid on 15 ml of Ficoll-Paque Plus (Amersham Biosciences, Uppsala, Sweden) and centrifuged at 500 × g for 30 min. Only specimens that contained ≥80% leukemia cells were studied. The mononuclear cells were washed twice and suspended in RPMI-1640 medium supplemented with 10% fetal bovine serum, plated in culture dishes at 5–12×10⁵ cells/ml, and incubated at 37°C in a humidified atmosphere of 5% CO₂ in air.

Suspensions of cells (5×10⁴ cells/ml) in 2 ml of culture medium were incubated with or without the test compounds in multidishes. Cell numbers were counted in a Model Z1 Coulter Counter (Beckman Coulter, Tokyo, Japan). Superoxide-generating oxidase was determined by the ability of the cells to reduce NBT upon exposure to 12-O-tetradecanoyl phorbol-13-acetate (9). Cells were incubated in 1 ml of RPMI-1640 medium containing NBT (1 mg/ml) and 12-O-tetradecanoyl phorbol-13-acetate (100 ng/ml) at 37°C for 50 min. The reaction was stopped by the addition of 5 M HCl (1 M, final concentration). The suspension was kept at room temperature for 20 min and then centrifuged. Formazan deposits were solubilized in dimethyl sulfoxide, and the absorption of the formazan solution at 560 nm/10⁷ cells was measured in a spectrophotometer. Morphological changes were examined in cell smears stained with May-Grünwald-Giemsa solution. Surface expression of CD11b was determined by monoclonal antibody labeling and flow cytometry using a FACScan (Becton-Dickinson, Mountain View, CA, USA), as described previously (10).

NB4 cells (3×10³ per dish) were plated into 3 ml of a semisolid medium with 0.8% methylcellulose and 20% fetal bovine serum in triplicate multiwell plates (6 wells, 9.6 cm² growth area/well) for 8 days. A solution of 0.1 ml of PBS containing various concentrations of drugs was added to the semisolid medium. Colonies were photographed under an inverted microscope. Colonies in enlarged photographs were measured and those >0.4 mm in diameter were counted.

Bone marrow plugs from a single femur of three Balb/c mice were pooled in RPMI-1640 medium. A single-cell suspension was prepared by vigorous pipetting. Nucleated cells (1,000 cells/ml/dish) were placed in 1 ml of a semi-solid medium containing haematopoietic growth factors (MethoCult^® GF M3434, Stem Cell Technology Inc., Vancouver, BC, Canada) in a 24-well plate and incubated for 8 days (11).

Cells were packed after being washed with cold PBS, and then lysed at 1.5×10⁷ cells/ml in sample buffer [63 mM Tris-HCl (pH 6.8), 15% glycerol, 2% sodium dodecyl sulfate (SDS), 5% 2-mercaptoethanol and 0.005% bromophenol blue]. The resultant lysates were resolved on 10% SDS-polyacrylamide gels. The proteins were transferred electrically from the gel to an Immobilon-P membrane (Millipore, Bedford, MA, USA) and immunoblotted with antibody. All antibodies were purchased from Cell Signaling Technology Japan (Tokyo). Horseradish peroxidase (HRP)-conjugated antibody (Cell Signaling Technology, CA, USA) was used as a secondary antibody (1:2,000 dilution). The bands were developed by treatment with the Immun-Star HRP Chemiluminescent kit (Bio-Rad Laboratories, Danvers, MA, USA) for 5 min at room temperature, and detected using a Fuji Lumino Image Analyzer LAS-4000 system (Fuji Film, Tokyo, Japan).

The results are expressed as means ± standard deviation (SD). Pairs of data were compared using Student's t-test. Significant differences were considered to exist for probabilities <5% (P<0.05).

Various compounds induce human myeloid leukemia HL-60 cells to undergo differentiation into granulocytes and monocytes (12). We examined the effects of tamoxifen on the differentiation and growth of HL-60 cells. Tamoxifen concentration-dependently inhibited the proliferation of HL-60 cells and induced the expression of CD11b, a surface marker of myelomonocytic differentiation (Fig. 1A). Since our previous report indicated that α-tocopherol prevents the growth-inhibitory effect of tamoxifen in pancreatic cancer cells (11), we examined the effect of α-tocopherol on CD11b expression in HL-60 cells (Fig. 1A). α-tocopherol significantly suppressed tamoxifen-induced CD11b expression, but had little effect on ATRA-induced CD11b expression. Tamoxifen also induced the reduction of NBT, a functional marker of differentiation (Fig. 1B). However, tamoxifen scarcely affected morphologic changes (Fig. 1C). These results indicate that tamoxifen is not a potent inducer of differentiation. Tamoxifen effectively enhanced the growth-inhibitory and NBT-reducing effects of VD3 in HL-60 cells (Fig. 1B). Tamoxifen also enhanced the morphologic differentiation of HL-60 cells induced by VD3 (Fig. 1C). Next, we examined the effects of tamoxifen on the growth-inhibitory and NBT-reducing effects of ATRA, butyrate, and DMSO (Fig. 2). Tamoxifen enhanced ATRA-induced NBT reduction in the APL cell line NB4 (Fig. 3A). Although ATRA significantly inhibited the proliferation of HT93 cells, viable cells were still observed among HT93 cells treated with 900 nM ATRA, as judged by the MTT assay (Fig. 3B). Tamoxifen and ATRA co-operatively inhibited the viability of HT93 cells, and no viable cells were observed in a culture with a combination of 6 μM tamoxifen and 900 nM ATRA for 6 days.

The growth-inhibitory effect of tamoxifen on NB4 cells was prevented by treatment with α-tocopherol, even in the presence of ATRA (Fig. 3C). ATRA-induced inhibition of the growth of NB4 cells was concentration-dependently counteracted by a selective antagonist of retinoic acid receptor-α, BMS195614, while this antagonist hardly affected the growth-inhibitory effect of tamoxifen (Fig. 3D).

Several inducers including ATRA activate ERK in leukemia cells before inducing myelomonocytic differentiation, and this activation is needed to elicit differentiation and growth arrest (13,14). Treatment with PD98059, a specific inhibitor of MAPK kinase (MEK) phosphorylation and activation (15), resulted in a loss of ERK activity from leukemia cells and blockade of the differentiation of leukemia cells (16). ATRA significantly increased ERK activity and this increased ERK activity was reduced by pretreatment with PD98059 (13). This inhibitor also suppressed the NBT reduction induced by ATRA, suggesting that ERK activation is required for ATRA-induced differentiation (13). PD98059 also inhibited tamoxifen-induced NBT reduction in HL-60 cells (Fig. 3E), and kinase activity was enhanced in tamoxifen-treated cells (Fig. 3F). Fig. 3E shows that treatment with PD98059 reduced the NBT reduction induced by ATRA plus tamoxifen. These results suggest that synergism between ATRA and tamoxifen in the inhibition of proliferation and induction of differentiation of myeloid leukemia cells is, at least in part, due to ERK activation by tamoxifen.

A colony-forming assay indicated that, while ATRA effectively inhibited the formation of large colonies, 30 nM ATRA could not completely eliminate small colonies (Fig. 4). The mean colony numbers in untreated and ATRA-, tamoxifen-, and ATRA plus tamoxifen-treated cultures were 476.5±51.6, 218.1±23.2, 132.3±22.6, and 0.6±0.1 (± SD), respectively. Combined treatment with ATRA and tamoxifen completely suppressed colony formation by NB4 cells, although several clusters (<0.1 mm in diameter) were observed in cultures treated with this combination (Fig. 4D), suggesting that the combined treatment was not cytotoxic. Next, the combined effects of ATRA and tamoxifen on colony formation in mouse bone marrow cells were compared to those in NB4 cells. The combination of 30 nM ATRA plus 6 μM tamoxifen significantly suppressed colony formation by NB4 cells but hardly affected colony formation by normal mouse bone marrow cells. Both the number of colonies and the colony size in the treated cultures were similar to those in untreated cultures (Fig. 4E and F). These results suggest that APL cells were more sensitive to combined treatment with ATRA and tamoxifen than normal hematopoietic cells.

APL cells were freshly isolated and incubated with various concentrations of ATRA and tamoxifen for 3 days. We analyzed morphological changes in APL cells in primary culture. Most untreated APL cells were still blastic after 3 days of culture (Fig. 5A). Tamoxifen did not affect the morphological changes in APL cells (Fig. 5B). Although ATRA at 300 nM induced APL cells to differentiate into myelocytes and metamyelocytes, 9 nM ATRA did not significantly affect the morphology of APL cells (Fig. 5C). Combined treatment with 9 nM ATRA and 2 μM tamoxifen significantly induced the granulocytic differentiation of APL cells (Fig. 5D), suggesting that tamoxifen enhances the granulocytic differentiation of APL cells induced by ATRA.

Under treatment with ATRA for 3 days, although malignant cells did not grow, most of the cells were still viable even at a high concentration (300 nM). In the presence of 6 μM tamoxifen, ATRA concentration-dependently reduced the numbers of viable cells (Fig. 5E).