Arsenic Trioxide-mediated Growth Inhibition in MC/CAR Myeloma Cells via Cell Cycle Arrest in Association with Induction of Cyclin-dependent Kinase Inhibitor, p21, and Apoptosis¹

Arsenic trioxide(As₂O₃) has recently been reported to induce complete remission in the patients with relapsed or refractory APL³without severe marrow suppression (1, 2). Although the mechanism of the antileukemic effect of As₂O₃ is not well understood, it is known that As₂O₃ is able to degrade a PML protein and a PML/RARα fusion protein in APL with a t(15;17)(3, 4, 5, 6). More recently, it has been shown that the antiproliferative effect of As₂O₃ is not limited to APL but can be observed in a variety of hematological malignancies without having the PML-RARα fusion protein (7, 8, 9, 10, 11, 12), suggesting that the antiproliferative effect of As₂O₃ might be independent on a PML or a PML-RARα fusion protein status. The accumulating evidences indicated that As₂O₃ could induce the apoptosis in leukemia and myeloma cells by modulating the apoptotic genes (2, 7, 9, 10, 11, 12, 13). As₂O₃-induced apoptosis was mediated via the down-regulation of Bcl-2 in APL cell line, NB4 cells(5), and molecular target of apoptosis was reported to be tubulin in myeloid leukemia cells (8).

The cell cycle in eukaryotes is regulated by CDKs. The cyclins, members of the cell cycle regulators, bind to and activate CDKs. Sequential formation, activation, and subsequent inactivation of cyclins and CDKs are critical for the control of cell cycle (14, 15, 16, 17). Recently, proteins of a new functional class that inhibit CDK activity,called CDKIs, have been identified. These CDKIs can play a key role in controlling the cell cycle progression by negatively regulating the CDK activities at an appropriate time in the cell cycle(18, 19, 20, 21). It has been indicated that lower doses of arsenical compound inhibit the proliferation of lymphoid malignant cells and NB4 cells through the cell cycle arrest in a G₁ phase (9) and a G₂-M phase (22), respectively. However, modulation of the cell cycle-regulatory proteins affected by As₂O₃ remains still open to question.

MM is a plasma cell neoplasm derived from clonal B cell lineage cells(23). Although many therapeutic advances such as combined chemotherapy and hematopoietic stem cell transplantation have been made to improve the survival rate of patients with MM, a higher proportion of patients cannot expect the long term remission due to drug-resistant disease, minimal residual disease, or infection. Therefore, a new potent therapeutic strategy is needed for the treatment of MM patients.

Recently, As₂O₃ was reported to inhibit the proliferation of human myeloma cells by induction of the apoptosis (7). However, little is known about the molecular mechanism of As₂O₃-induced apoptosis, as well as modulation of the cell cycle-regulatory proteins in MM. In the present study, we investigated cell cycle arrest and induction of the apoptosis better to understand the antiproliferative effect of As₂O₃ on various MM cell lines.

MM cell lines used in this study were MC/CAR, ARH-77, HS-SULTAN, IM-9,MOPC315, NCI-H929, RPMI8226, and U266, purchased from the American Type Culture Collection (Manassas, VA), and maintained according to the recommendation of the American Type Culture Collection. Cells were maintained at 37°C in an atmosphere of 5% CO₂in air.

As₂O₃ was purchased from Sigma Chemical Company (St. Louis, MO). As₂O₃ was dissolved in 1.65 m NaOH at 5 × 10⁻²m as a stock solution. The maximum concentration of NaOH in culture had no influence on cell growth of these cell lines. Human recombinant IL-6 was obtained from R&D Systems, Inc. And Z-VAD-FMK(caspase-3 inhibitor) obtained from Enzyme Systems Products (Livermore,CA) was dissolved in DMSO (Sigma).

In vitro growth inhibition effect of As₂O₃ on myeloma cells was determined by measuring MTT dye absorbance of living cells(24). Briefly, cells (2 × 10⁵ cells/well) were seeded in 96-well microtiter plates (Nunc, Roskilde, Denmark). After exposure to the drug for 72 h, 50 μl of MTT (Sigma) solution (2 mg/ml in PBS) were added to each well, and the plates were incubated for additional 4 h at 37°C. MTT solution in medium was aspirated off. To achieve solubilization of the formazan crystal formed in viable cells, 200 μl of DMSO were added to each well before absorbance at 570 nm was measured.

Cell cycle distribution was determined by staining DNA with PI (Sigma)as previously described (25). Briefly, 1 × 10⁶ cells were incubated with 10 μmbromodeoxyuridine (Sigma) at 37°C for 1 h. Cells then were washed in PBS and fixed in 70% ethanol. After incubation of cells with 1 ml of 2 n HCl containing Triton X-100 (Fisher Scientific,Fair Lawn, NJ) for 30 min at room temperature followed by two washes with PBS, cells were incubated with 20 μl of anti-bromodeoxyuridine for 30 min at room temperature. After two washes with PBS, cells were incubated with 1 μg of FITC-goat antimouse IgG (Caltag Laboratories,San Francisco, CA) for 30 min at room temperature. Cells were again washed with PBS and then incubated with 1 μg of PI. The percentage of cells in the different phases of the cell cycle was measured with FACStar flow cytometer (Becton Dickinson, San Jose, CA), analyzed by using Becton Dickinson software (Lysis II, Cellfit).

Total RNA was extracted by the TRI reagent (Molecular Research Center,Inc., Cincinnati, OH). RNA (15 μg/sample) was size fractionated through 1% agarose-formaldehyde gel and transferred to nylon membrane(Schleicher and Schuell, Dassel, Germany). The cDNA probes for p21,p27, p53, Bax, IL-6, and IL-6 receptor were labeled with[α-³²P]dCTP (ICN, Costa Mesa, CA) to high specific activity by random primer labeling. The filter was hybridized for 20–24 h at 42°C, washed, and exposed to Kodak XAR 5 film(Eastman Kodak, Rochester, NY). The filter was hybridized with theβ-actin cDNA to normalize for RNA loading.

Samples containing 30 μg of total protein were resolved by a 12%SDS-PAGE gel, transferred onto a nitrocellulose membrane (Bio-Rad,Hercules, CA) by electroblotting, and probed with anti-p21, anti-CDK2,anti-CDK4, anti-CDK6, anti-cyclin D1, anti-cyclin E, anti-cyclin A,anti-cyclin B1, anti-Rb, anti-cdc25B, anti-cdc25C, anti-Wee1,anti-PARP, anti-Bcl2, anti-Bax polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA), anti-p27 polyclonal antibody,anti-caspase 3 monoclonal antibody (Transduction Laboratories,Lexington, KY), anti-cdc2 polyclonal antibody (Oncogene Research Products, Cambridge, MA), and anti-cdc2 phosphate-specific monoclonal antibody (New England Biolabs, Inc., Beverly, MA). The blots were developed by using the ECL kit (Amersham, Arlington Heights, IL).

Samples of total protein (100 μg) were incubated with anti-CDK2,anti-CDK4, anti-CDK6, anti-cdc2, anti-cyclin A, anti-cyclin B1,anti-cyclin D1, and anti-cyclin E polyclonal antibodies for 2 h at 4°C, followed by incubation with protein A-agarose conjugates (Santa Cruz Biotechnology) for 1 h. The protein complexes were washed three times with immunoprecipitation buffer [50 mm Tris(pH 7.5), 0.5% NP40, 150 mm NaCl, 50 mm NaF,0.2 mm NaVO₄, 1 mm DTT,20 μg/ml aprotinin, 20 μg/ml leupeptin, 1 mmphenylmethylsulfonyl fluoride] and released from the agarose bead by boiling in 2× SDS sample buffer [125 mm Tris (pH 6.8),4% SDS, 10% β₂-mercaptoethanol, 2% glycerol,0.004% bromphenol blue] for 5 min, and the reaction mixture was resolved by a 12% SDS-PAGE gel, transferred onto a nitrocellulose membrane by electroblotting, and probed with anti-p21 and anti-p27 antibodies. The blot was developed by using the ECL kit.

Total lysates were prepared and immunoprecipitated with anti-cdc2,anti-CDK2, anti-CDK4, and anti-CDK6 polyclonal antibodies as described above. The beads were washed three times in immunoprecipitation buffer and then three times in kinase buffer [10 mm Tris (pH 7.5), 2 mm MgCl₂]. The kinase reactions was carried out at 37°C for 30 min in 25 μl of kinase reaction buffer containing 2.5 mm EGTA, 0.1 mmNaVO₄, 1 mm NaF, 20 μl ATP, 5 μCi[γ-³²P]ATP, and 2 μg of histone H1 substrate. The reaction was stopped by adding 2× SDS sample buffer. After boiling for 5 min, the reaction products were electrophoretically separated on a 12% SDS-PAGE gel, and phosphorylated proteins were detected by autoradiography.

Apoptosis was determined by staining cells with annexin V-FITC and PI labeling, because annexin V can identify the externalization of phosphatidylserine during the apoptotic progression and therefore detect early apoptotic cells (26). To quantitate the apoptosis of cells, prepared cells were washed twice with cold PBS and then resuspended in binding buffer [10 mm HEPES/NaOH (pH 7.4), 140 mm NaCl, 2.5 mmCaCl₂] at a concentration of 1 × 10⁶ cells/ml. Then, 5 μl of annexin V-FITC(PharMingen, San Diego, CA) and 10 μl of 20 μg/ml PI (Sigma) were added to these cells, which were analyzed with FACStar flow cytometer(Becton Dickinson). Also, during the cell cycle analysis described above, cells were considered to be in apoptosis if they exhibited sub-G₁ DNA fluorescence and a forward angle light scatter the same as or slightly lower than that of cells in G₁. Mitochondrial transmembrane potential(Δψ_m) was determined by flow cytometry. Briefly, cells were washed twice with PBS and incubated with 0.1μg/ml Rhodamine 123 (Sigma) at 37°C for 30 min. Subsequently, PI (1μg/ml) was added, and Rhodamine 123 and PI staining intensity was determined by flow cytometry.

We examined the effect of As₂O₃ on the cell proliferation of the MM cell lines using MTT assay. Significant dose-dependent inhibition of cell growth was observed in all of eight cell lines used in this study following the treatment of As₂O₃ for 72 h (Fig. 1). As₂O₃ at 2μ m exhibited >50% inhibition of growth in five MM cell lines (MC/CAR, ARH-77, MOPC315, NCI-H929, and U266). Whereas RPMI8266 cells were the most resistant to As₂O₃ among the MM cell lines tested, NCI-H929 cells treated with As₂O₃ were inhibited more effectively (IC₅₀ 0.7 μm) in cell growth than any other MM cell lines tested in our study. With regard to MC/CAR cells, the concentration with a 50% growth inhibition to As₂O₃ was <2μ m. As₂O₃also inhibited the proliferation in primary MM cells in a dose-dependent manner (data not shown). These results showed that As₂O₃ was very potent in inhibiting the proliferation of MM cells in vitro.

The effect of As₂O₃ on the cell cycle was determined in MM cell lines by FACS analysis. As shown in Table 1, DNA flow cytometric analysis with three independent experiments indicated that As₂O₃induced a G₁ and/or a G₂-M phase arrest in these cell lines following 72 h of exposure. Whereas ARH-77, NCI-H929, RPMI8226, and U266 cells showed a significant increase in the number of cells in the G₁ phase of the cell cycle, HS-SULTAN and IM-9 cells significantly induced a G₂-M phase arrest of the cell cycle following the treatment of As₂O₃. MOPC315 and MC/CAR cells had a significant increase in cells of G₁ and G₂-M phases of the cell cycle. Compared with untreated control MC/CAR cells, the percentages of cells in G₁ phase and G₂-M phases were increased by 13.1 and 7.8%,respectively, accompanied with a concomitant reduction of cells at the S phase of cell cycle after treatment of MC/CAR cells with As₂O₃ for 72 h. Taken together, these results indicated that As₂O₃ inhibited the cellular proliferation of MM cells via a G₁and/or a G₂-M phase arrest of the cell cycle.

Because As₂O₃ induced both G₁ phase and G₂-M phase arrests in MC/CAR cells, we wanted to determine the mRNA and protein levels of CDKI, p21, and p27, in MC/CAR cells exposed to 2μ m As₂O₃. The level of p21 protein increased with its mRNA in a time-dependent manner(Fig. 2). With regard to CDKI, p27, no detectable changes of this protein and mRNA were observed, although mild reduction was seen at 24 h after As₂O₃ treatment (Fig. 2). In particular, induction of p53 coincided with an increase in both protein and mRNA of p21 (Fig. 2,A). In vertebrate cells, the G₁ progression and the G₁-S transition are also regulated by d-type cyclins that bind to and activate CDK4, CDK6, and cyclin E and A which activate CDK2, respectively (14, 15, 16, 17). Treatment of MC/CAR cells with 2 μmAs₂O₃ resulted in a down-regulation of the protein level of CDK6 in a time-dependent manner, whereas the levels of CDK2 and CDK4 were not changed (Fig. 3,A). In addition, cyclin D1 and cyclin E were expressed constantly for 72 h without the reduction of their proteins (Fig. 3,A). In the case of cyclin A which is essential for CDK2 activation, its observed decrease was with time of treatment of As₂O₃ (Fig. 3 A).

Also in MC/CAR cells, we assessed the levels of G₂-M phase-related proteins to know how these proteins were regulated by As₂O₃. Treatment of As₂O₃ resulted in a down-regulation of the level of cdc2 protein (Fig. 3,B),which is an important role in the S phase and the G₂-M progression of the cell cycle (14, 15, 17). The expression of cyclin B1 which binds to and activates cdc2 was not changed at any times of incubation (Fig. 3,B). Furthermore, because the progression from the G₂phase into mitosis is negatively regulated by cdc2 phosphorylation on threonine 14 (T14) and tyrosine 15 (Y15) residues (15, 27, 28), we investigated the status of cdc2 phosphorylation on tyrosine 15 residues (cdc2-Y15) using a phosphorylation state-specific antibody capable of binding only to tyrosine 15-phosphorylated cdc-2 protein. In MC/CAR cells treated with 2μ mAs₂O₃ for 72 h, there was an increase of detectable cdc2-Y15 phosphorylation although the levels of cdc2-Y15 phosphorylation at the start time of As₂O₃ treatment were higher than expected as a basal level (Fig. 3,B). In addition, we examined the changes of Wee1 (protein kinase), cdc25B, and cdc25C(protein phosphatase), which regulate the cdc2-Y15 phosphorylation level. No detectable change in Wee1 protein content was observed in MC/CAR cells grown at 2 μmAs₂O₃ (Fig. 3,B). In contrast, cdc25B and cdc25C were substantially decreased in a time-dependent manner under the same condition (Fig. 3 B),suggesting that As₂O₃-induced up-regulation of cdc2-Y15 phosphorylation level may be mediated via down-regulation of cdc25B and cdc25C proteins.

Next, we questioned whether p21 induced by As₂O₃ (2 μmfor 72 h) could be detected in complexes with CDKs and cyclins in the cell cycle. As shown in Fig. 4,A, the complexes immunoprecipitated with anti-cdc2 and anti-CDK6 antibodies showed higher amounts of immunodetectable p21 protein from As₂O₃-treated cells than from control cells. However, the complexes immunoprecipitated with anti-CDK2 and anti-CDK4 antibodies showed no significant difference in the level of p21 protein between As₂O₃-treated and-untreated control cells. In addition, the amounts of cyclin A- and cyclin E-bound p21 forms were higher in As₂O₃-treated cells than in control cells (Fig. 4 B). In contrast to p21, the p27 protein which was not induced by As₂O₃ could not be detected in the complexes immunoprecipitated with all of the above cell cycle-related proteins in the presence or absence of As₂O₃ (data not shown). Collectively, these results suggested that p21 protein may play a key role in G₁ arrest and G₂-M phase arrests through its increased binding to cdc2, CDK6, cyclin A,and cyclin E proteins in As₂O₃-treated MC/CAR cells.

To determine whether the increased p21 and the decreased cdc2, CDK6,and cyclin A proteins result in the inhibition of CDK activity in As₂O₃-treated MC/CAR cells,we performed in vitro CDK activity assay on histone H1 substrate in immunoprecipitates with anti-cdc2, -CDK2, -CDK4, and -CDK6 antibodies. Treatment with As₂O₃ (2μ m for 72 h) demonstrated a dramatic decrease of cdc2- and CDK6-associated kinase activity on histone H1 substrate compared with As₂O₃-untreated control cells (Fig. 5,A). However, there is no significant difference of CDK2- and CDK4-associated histone H1 kinase activity between As₂O₃-treated cells and-untreated cells. In addition, the decrease of CDK6-associated kinase activity was associated with the underphosphorylation of Rb protein (Fig. 5 B). Taken together, these results suggested that As₂O₃ induced a G₁ and a G₂-M phase arrest via the reducing activity of CDK6-associated kinase and cdc2-associated kinase, respectively, in association with increased p21 protein.

We performed in vitro apoptosis detection assay to know whether As₂O₃ treatment could induce the apoptosis in MC/CAR cells by FACS analysis. Apoptosis was induced in a dose-dependent manner except at a dose of 1μ mAs₂O₃ for 72 h (Fig. 6,A). To confirm and evaluate the induction of apoptosis, we performed the staining of cells with annexin V and PI (Fig. 6 B). As with the percentages of sub-G₁group by FACS analysis, the proportion of apoptotic cells in the lower right quadrant was increased in a dose-dependent manner. Especially, at 2 μmAs₂O₃, the percentages of apoptotic and dead cells were ∼72.9%. Taken together, these results indicate that induction of apoptosis can be another mechanism of the antiproliferative effect of As₂O₃ besides G₁ and G₂-M phase arrests of the cell cycle in MC/CAR myeloma cells.

Concerning the relationship between Bcl-2 and Bax regulation during apoptosis, the Bcl-2 protein level was down-regulated in a dose-dependent manner (Fig. 6,C) and in a time-dependent manner (Fig. 7,A). However, Bax protein was constitutively expressed in the absence of change of its protein level (Figs. 6,C and 7,A), suggesting that apoptosis by As₂O₃ may be mediated through the down-regulation of Bcl-2 apoptotic protein in MC/CAR cells. As shown in Fig. 7,B,As₂O₃ also induced the subsequent loss of mitochondrial transmembrane potential(Δψ_m). Next, we wanted to investigate whether caspase-3 might be activated during the induction of apoptosis by As₂O₃, because the ICE/caspase family plays an essential role in apoptosis(29, 30, 31, 32). The 32-kDa precursor (procaspase-3) was degraded in a time-dependent manner (Fig. 7 A), implying that induction of apoptosis by As₂O₃ was related to activation of caspase-3 enzymatic protein. Concerning poly(ADP-ribose)polymerase (PARP) protein, which is a major substrate for active caspase-3 and a hallmark of apoptosis, Western blotting showed that the intact 113-kDa moiety of PARP was degraded, as evidenced by increased 89-kDa cleavage products (Fig. 7A). Conclusively, these results indicated that As₂O₃-induced apoptosis of MC/CAR cells was accompanied by the changes of Bcl-2 protein levels,Δψ _m and caspase-3 activity.

In the present study, we have shown that As₂O₃ inhibited the cell proliferation of MM cells by inducing the cell cycle arrest as well as triggering the apoptosis. As₂O₃ exhibited a dose-dependent inhibition of cellular proliferation in primary MM cells as well as MM cell lines, although there is a little different susceptibility to As₂O₃among the MM cells. We found that IC₅₀ of As₂O₃ was <2μ m in five cell lines among eight MM cells, which was somewhat similar to that of As₂O₃ in growth inhibition of APL cells (3, 4, 5, 6). Very recently, Rousselot et al. (7) reported that pharmacological concentration of As₂O₃ caused the inhibition of growth in MM cell lines as well as primary MM cells isolated from patients. They also showed almost same IC₅₀ of As₂O₃ as our results in NCI-H929 and U266 cells. These results suggest that As₂O₃ may be clinically useful in patients with MM at the therapeutic and pharmacological concentration, as demonstrated by studies on the treatment of APL patients (1, 2).

Our cell cycle analysis has revealed that As₂O₃ was able to prominently induce a G₁ phase and/or a G₂-M phase arrest of MM cells after their exposure to As₂O₃. These results were consistent with those of other investigators who showed that antiproliferative action of arsenical compound was linked to a G₁ phase arrest in lymphoid neoplasms(9) and a G₂-M phase arrest in NB4 cells (22) at lower doses. Therefore, it is likely that As₂O₃ may induce the cell cycle arrest of a G₁ phase and/or a G₂-M phase depending on the cell type, suggesting that the molecular mechanisms of cell cycle arrest by As₂O₃ have great variety. In this study, the G₁ and G₂-M phase arrests in MC/CAR cells were associated with a marked up-regulation of p21 protein and mRNA,suggesting transcriptional and translational regulation of p21 gene by As₂O₃. The p21 can be up-regulated by both p53-dependent and p53-independent pathways(33, 34). The expression of high level of the p53 can give rise to a G₁ arrest alone, a G₂-M arrest, or both G₁ and G₂-M arrests depending on the cell type(35, 36, 37). The expression of p53 mRNA was elevated in association with up-regulation of p21 mRNA. Therefore, it is probable that p21 induction and the cell cycle arrest of both G₁ and G₂-M phases in As₂O₃-treated MC/CAR cells may be mediated by p53. Among CDKs that regulate the cell cycle, CDK2,CDK4, and CDK6 are activated in association with d-type cyclins or cyclin E during the G₁ progression and the G₁-S transition. We found that the expression of CDK6 protein was decreased in a time-dependent manner in As₂O₃-treated MC/CAR cells,but those of CDK2, CDK4, cyclin D1, and cyclin E were not. In addition,the accumulated p21 protein in association with G₁ arrest was detected largely in complexes with CDK6 and cyclin E without its increased binding to complexes with CDK2,CDK4, and cyclin D1, supporting the idea that only CDK6-associated kinase activity was markedly decreased in our kinase assay. Furthermore, reduced kinase activity of CDK6 was accompanied with the underphosphorylation of Rb protein, which is known to sequester the transcription factor, E2F, thereby preventing cells from further entering the cell cycle progression. These results suggest that As₂O₃-induced p21 binds specifically to CDK6 and cyclin E proteins and inhibits the kinase activity of CDK6, ultimately resulting in hypophosphorylation of Rb protein. Taken together, G₁ blocking MC/CAR cells from entry into S phase is mediated by down-regulation of CDK6-associated kinase activity in association with induction of CDKI,principally p21.

Cdc2 kinase is activated primarily in association with cyclin A and B in G₂-M phase progressions. In this study, we have demonstrated that cdc2 and cyclin A proteins were decreased following treatment with As₂O₃. Rb-bound E2F suppresses a number of key genes needed for S phase progression including cyclin A (38, 39, 40) which is required in both S phase progression as well as the G₂-M transition(14, 15, 16, 17). Therefore, the decrease of cyclin A by As₂O₃ might be mediated via E2F sequestered by hypophosphorylation of Rb. The increased p21 was seen much higher in complexes with cdc2 and cyclin A in As₂O₃-treated MC/CAR cells. These effects could account for the decreased activity of cdc2-associated kinase in our kinase assay. Alternatively, cdc2 activity can be negatively regulated by cdc2 phosphorylation on threonine 14 and tyrosine 15 (15, 27, 28). These phosphorylations are enforced by protein kinases including Wee1(41) and are also retarded by cdc25 phosphatases,especially cdc25C (42, 43, 44). Therefore, tyrosine phosphorylation of cdc2 at tyrosine 15 (cdc2-Y15 phosphorylation)inhibits kinase activity and is one of the mechanisms by which human cells inhibit mitosis after exposure to DNA-damaging agent(45). In MC/CAR cells treated with As₂O₃, there was an increase of detectable cdc2-Y15 phosphorylation. This result might be due to the down-regulation of cdc25B and cdc25C phosphatases, because no detectable change in Wee1 protein was observed. Thus, it is likely that the increased phosphorylation of cdc2 via down-regulation of cdc25B and cdc25C enhances the decreased activity of cdc2 kinase by p21 protein during a G₂-M arrest. Conclusively,As₂O₃-induced cell cycle arrest in MC/CAR cells resulted from the inactivation of CDK6 in a G₁ phase and cdc2 in a G₂-M phase through the induction of p21.

In addition, our data showed that As₂O₃ markedly induced the apoptosis in a dose-dependent manner in all of the cell lines [ARH-77,HS-SULTAN, NCI-H929, and U266 (data not shown) including MC/CAR]tested. These results strongly support the notion that apoptosis induced by As₂O₃ can occur independently of PML-RARα fusion protein. To gain insight into understanding the molecular mechanism involved in apoptosis by As₂O₃, expression of the antiapoptotic protein, Bcl-2, was assessed in MC/CAR cells. It has been reported that nonorganic(As₂O₃) or organic(melarsoprol) arsenical compounds efficiently induced the apoptosis in myeloid and lymphoid cell lines through the down-regulation of Bcl-2 gene expression (9, 10, 11, 12). Similarly, we showed that the induction of apoptosis was accompanied with the down-regulation of Bcl-2 protein, supporting the idea that alteration of Bcl-2 is directly or indirectly involved in the apoptotic effect of As₂O₃ in MC/CAR cells. By contrast, no detectable changes in Bcl-2 were reported during apoptosis in the T cell line and MM cell lines (NCI-H929 and U266)(7, 46). In addition to MC/CAR cells, we examined the modulation of Bcl-2 in ARH-77, NCI-H929, and U266 cells,resulting in the absence of any modification in Bcl-2 protein (data not shown). This discrepancy may be due to the existence of more than one distal pathway of apoptosis or different cell lineage specificity. In regard to regulation of the proapoptotic Bax gene, any induction of Bax protein was not observed during treatment with As₂O₃. However, it should be emphasized that the ratio of Bcl-2 to Bax determines the amount of Bcl-2/Bax heterodimers versus Bax/Bax homodimers and is important in determining susceptibility to apoptosis (47).

The ICE/caspase family plays a crucial role in apoptosis(29, 30, 31, 32). In particular, caspase-3 (CPP32/Yama/apopain)has been shown to be a key component of the apoptotic machinery. Recently, it was reported that As₂O₃ appeared to induce apoptosis, coincident with conversion from inactive precursors to activated enzymes, especially caspase-3, in NB4 cells (9),primary APL cells (2), and mouse B cell leukemia cells(11). Similarly, our data demonstrated that caspase-3 was activated and PARP protein was degraded by As₂O₃. Although the collapse of mitochondrial transmembrane potential(Δψ_m) resulting from the low ratio of Bcl-2 to Bax was believed to cause the activation of caspase-3 in our study,the mechanism by which As₂O₃ activated caspase-3 protease remains to be elucidated. Furthermore, MC/CAR cells were protected from the apoptotic effect of As₂O₃ when treated with Z-VAD-FMK (50 μm) of the caspase-3 inhibitor(data not shown). Therefore, these results strongly provide the evidence that caspase-3 might be one of the critical steps in As₂O₃-induced apoptosis.

MM cells are tightly related to various cytokines including IL-6,recognized as a myeloma growth factor (48, 49, 50). It has been shown that IL-6 protects the myeloma cells against dexamethasone-and Fas-induced apoptosis (49, 50). However, mild cellular proliferation was noted in HS-SULTAN, NCI-H929, and U266 cells treated with IL-6 (10 ng/ml), and IL-6 could not inhibit As₂O₃-induced apoptosis of these MM cells (data not shown), which was similar to the result of Rousselot et al. (7). Thus, the mechanism by which As₂O₃induces apoptosis in MM cells may be different from that of dexamethasone- or Fas-induced apoptosis. Furthermore, we could not identify the detectable mRNA changes of IL-6 and IL-6 receptor genes in all of the MM cell lines during the As₂O₃-induced cell cycle arrest and apoptosis (data not shown), suggesting that the antiproliferative effect of As₂O₃ in MM cells may not be mediated via the IL-6 signaling pathway. However, whether As₂O₃ disturbs the downstream signal transduction of IL-6/IL-6 receptor complex remains to be elucidated.

In summary, As₂O₃ inhibits the cell proliferation of MM cell lines by not only inducing the cell cycle arrest through p21 but also triggering the apoptosis through caspase-3, especially in MC/CAR cells. Finally, these results suggest that As₂O₃ may be useful as one of the investigational drugs in the treatment of MM patients. We are currently investigating the in vivo effect of As₂O₃ on MM using SCID mice.