Arsenic induces apoptosis of multidrug-resistant human myeloid leukemia cells that express Bcr-Abl or overexpress MDR, MRP, Bcl-2, or Bcl-x_L

As₂O₃ and trichostatin A were purchased from Sigma (St. Louis, MO). As₂O₃ was dissolved in 1.65 mol/L sodium hydroxide (NaOH) to make a stock solution of 1 mmol/L, which was serially diluted in RPMI 1640. A monoclonal anti-Bcl-2 antibody was purchased from DAKO (Carpinteria, CA). Polyclonal anti-Bcl-x and anti-Bax antibodies and monoclonal anti-cyt c, anti-cIAP, and anti-Bcr-Abl antibodies were purchased from Pharmingen (San Diego, CA). Rabbit anti-DFF (DNA fragmentation factor),24 anti-Apaf-1,25 and anti-Bid antisera26 were kindly provided by Dr Xiaodong Wang (University of Texas Southwestern Medical Center, Dallas, TX).

Human leukemic cells HL-60/neo, HL-60/Bcl-2, HL-60/Bcl-x_L, HL-60/VCR, HL-60/AR, HL-60/Bcr-Abl, and the erythroid blast crisis CML K562 cells were cultured and passaged as previously described.15,18,27 28 

Logarithmically growing cells were exposed to various concentrations and exposure intervals (up to 7 days) of As₂O_3.After these treatments with As₂O_3, aliquots of cells were withdrawn and cell numbers were determined using a Coulter particle count and size analyzer (Coulter, Hialeah, FL). Suspension culture growth inhibition and the 50% inhibitory concentration values (IC₅₀) for As₂O₃ were determined as previously described.29 

The flow cytometric evaluation of the cell-cycle status and apoptosis was performed according to a modification of a previously described method.30 Briefly, untreated or As₂O₃-treated cells were centrifuged, washed in Hanks' balanced salt solution, and fixed in 70% ethanol. The tubes containing the cell pellets were stored at −20°C for at least 24 hours. After this, the cells were centrifuged at 800g for 15 minutes, and the supernatant was discarded to remove ethanol completely. The pellets were resuspended in 40 μL (for 2-3 × 10⁶ cells) of phosphate-citrate buffer at room temperature for 30 minutes. After this incubation, cells were washed with 4 to 5 mL phosphate-buffered saline (PBS) and stained with propidium iodide (PI) solution (20 μg/mL PI and 20 μg/mL RNAse A in PBS) for 30 minutes. The samples were read on a Coulter Elite flow cytometer using Elite software program 4.0 for 2-color detection. The percentage of cells in the apoptotic sub-G₁ and the G1-S phase and G2-M phases were calculated using Multicycle software (Phoenix Flow Systems, San Diego, CA).

Western analyses of Bcl-2, Bcl-x_L, Bax, Bid, Fas receptor (CD95), Fas ligand (Fas L), Bcr-Abl, DFF, cIAP, and β-actin were performed using specific antisera or monoclonal antibodies (see above), as described previously.15 18 Horizontal scanning densitometry was performed on Western blots by using acquisition into Adobe Photo Shop (Apple, Cupertino, CA) and analysis by the NIH Image Program (National Institutes of Health, Bethesda, MD). The expression of β-actin was used as a control.

Histones were acid-extracted from whole cells as described previously21 22; 20 μg-isolated histones were subjected to SDS-PAGE as above (15% gel). Ponceau stain (Sigma) visualization was used as a control for the amount of protein loading. Antibodies that specifically recognize the acetylated forms of histone H3 and H4 (Upstate Biotechnology, Lake Placid, NY) were used to detect hyperacetylated histones.

For As₂O₃-induced changes in mitochondrial membrane potential (ΔΨm) and ROS, 5 × 10⁵HL-60/neo, HL-60/Bcr-Abl, and K562 cells were incubated with 40 nmol/L 3,3 dihexyloxacarbocyanine iodide or 5 μmol/L dichlorodihydrofluorescein diacetate, respectively, and were analyzed by flow cytometry, as described previously.18,31 32 

HL-60/neo, HL-60/Bcr-Abl, and K562 cells were treated with various concentrations of As₂O₃ for up to 7 days. Cells were then washed with PBS and resuspended in 100 μL FACS wash buffer (PBS, 0.2% NaN₃, 0.1% bovine serum albumin, 2% human AB-positive serum, filtered by suction at 0.45 μm). After 10 μL PE antihuman CD11b, CD33, CD34, or HLA-DR antibody (Pharmingen, San Diego, CA) was added,33,34 the cells were incubated in the dark at 4°C for 30 minutes. Samples were then analyzed by flow cytometry. Alternatively, untreated or As₂O₃-treated cells were washed in PBS and intracellular hemoglobin levels were determined by a previously described method.35 

Untreated and As₂O₃-treated cells were harvested by centrifugation at 1000g for 10 minutes at 4°C. Cell pellets were washed once with ice-cold PBS and resuspended with 5 vol buffer (20 mmol/L HEPES-KOH, pH 7.5, 10 mmol/L KCl, 1.5 mmol/L MgCl₂, 1 mmol/L sodium EDTA, 1 mmol/L sodium EGTA, 1 mmol/L dithiothreitol, and 0.1 mmol/L phenylmethylsulfonyl fluoride), containing 250 mmol/L sucrose. Cells were homogenized with a 22-gauge needle, and the homogenates were centrifuged at 100,000g for 30 minutes at 4°C (S-100 fraction).18 Supernatants were collected, and the protein concentrations of S-100 were determined by the Bradford method (Bio-Rad, Hercules, CA). After that, 20 to 30 μg S-100 was used for Western blot analysis of cyt c, as described previously.15 18 

After treatment with or without As₂O₃, 50 × 10³ cells were washed with PBS (pH, 7.3) and resuspended in the same buffer. Cytospin preparations of the cell suspensions were fixed and stained with Wright stain. Cell morphology was determined by light microscopy. In all, 5 different fields were randomly selected for counting of at least 500 cells. The percentage of apoptotic cells was calculated for each experiment, as described previously.36 

After drug treatment, 5 × 105 to 1 × 10⁶ cells were washed in PBS and resuspended in 100 μL staining solution (containing annexin-V fluorescein and PI in a HEPES buffer, annexin-V-Fluos Staining Kit; Boehringer-Mannheim, Indianapolis, IN). After 15-minute incubation at room temperature, cells were analyzed by flow cytometry. Annexin-V binds to those cells that express phosphatidylserine on the outer layer of the cell membrane, and PI stains the cellular DNA of those that have a compromised cell membrane. This allows for the discrimination of live cells (unstained with either fluorochrome) from apoptotic cells (stained only with annexin-V) and necrotic cells (stained with both annexin-V and PI).37 

Significant differences between values obtained in a population of leukemic cells treated with different experimental conditions were determined by paired Student t test analyses. A 1-way analysis of variance was also applied to the results of the various treatment groups, and post hoc analysis was performed using the Bonferroni correction method.

Recent reports indicate a broad spectrum of antileukemic activity for As₂O_3.5,7 This prompted us to investigate its efficacy against a variety of human myeloid leukemia cell types that display a multidrug-resistant phenotype caused by diverse mechanisms.15,18 27 The growth inhibitory effects of 1, 2, and 10 μmol/L As₂O₃ were determined after 3-, 5-, and 7-day exposure intervals. Figure 1demonstrates that a dose-dependent growth inhibitory effect of As₂O₃ was evident against the control HL-60/neo, as well as against HL-60/AR, HL-60/Bcl-2, HL-60/Bcl-x_L, HL-60/Bcr-Abl, and K562 cells. After exposure to clinically achievable concentrations of As₂O₃(1 or 2 μmol/L) for 7 days, the relative degree of growth inhibition in the various cell lines was HL-60/neo > HL-60/VCR > HL-60/AR > HL-60/Bcr-Abl > K 562 > HL-60/Bcl-2 > HL-60/Bcl-x_L (Figure 1). In these cell lines, the IC₅₀ values for As₂O₃ were determined to be between 0.8 and 1.5 μmol/L. Figure2 shows the percentage of apoptotic cells observed in the various cell types after exposure to As₂O₃ (0.5 to 10 μmol/L) from 1 to 7 days. At the end of a 7-day exposure to 2 μmol/L As₂O_3, approximately 30% to 50% of cells showed either the morphologic features of apoptosis (data not shown) or the cell-surface phosphatidylserine expression detectable by annexin-V staining and flow cytometry (Table1). After exposure to 2 μmol/L As₂O₃ for 7 days, there was no significant difference in the apoptotic rate in HL-60/AR or HL-60/VCR versus HL-60/neo cells. However, a lower percentage of apoptotic cells was observed in identically treated HL-60/Bcl-2, HL-60/Bcl-x_L, HL-60/Bcr-Abl, and K562 cells.

A previous report18 from our laboratory demonstrates that the expression of Bcr-Abl in HL-60/Bcr-Abl (p185) and K562 (p210) produces resistance against antileukemic drug-induced mitochondrial ΔΨm and cyt c release from the mitochondria into cytosol, caspase activation, and apoptosis. In the current study, we compared the effect of As₂O₃ on the mitochondrial ΔΨm and the release of cyt c and on caspase activation in HL-60/neo versus HL-60/Bcr-Abl or K562 cells. Figure 3demonstrates that a 7-day exposure to 2 μmol/L As₂O₃ produced similar levels of accumulation of cyt c in the cytosol of HL-60/neo, HL-60/Bcr-Abl, and K562 cells. This was associated with the cleavage of p116 poly (ADP-ribose) polymerase (PARP) into its p85 and p31 fragments and with the degradation of p45 DNA fragmentation factor (DFF) into its p30 and p11 fragments (not shown). As has been reported,24 cleavage of PARP and DFF largely results from the generation of caspase-3 activity. DFF is known to be the inhibitory protein for the endonuclease (caspase-associated DNase), which produces the DNA fragmentation of apoptosis.38,39 Recently, a pathway has been elucidated by which the activity of the upstream caspases can cause the release of cyt c from the mitochondria to the cytosol, resulting in Apaf-1-mediated sequential cleavage of caspase-9 followed by caspase-3.40 Bid (p21), a BH3 domain containing proapoptotic members of the Bcl-2 family, was cleaved directly by caspase-8, and the C-terminal fragment (p14) acted on the mitochondria to trigger cyt c release.26 41 As shown in Figure 3, treatment with 1 or 2 μmol/L As₂O₃for 7 days produced p21 Bid cleavage into its p14 fragment in HL-60/neo and in HL-60/Bcr-Abl and K562 cells, though more Bid cleavage occurred in HL-60/neo cells treated with 1 μmol/L As₂O₃. These data indicated that treatment with As₂O₃ generated the activity of both the upstream (caspase-8) and the effector caspase-3.

Because the activated Bid acted on the mitochondria, we examined whether As₂O₃-induced Bid cleavage and the release of cyt c from mitochondria were associated with ΔΨm and the increase in ROS. Figure 4 demonstrates that the treatment with 1 or 2 μmol/L As₂O₃produced more ΔΨm and the generation of ROS in HL-60/neo versus HL-60/Bcl-Abl and K562 cells, though significant preapoptotic mitochondrial alterations were also produced in HL-60/Bcr-Abl and K562 cells after exposure to 2 μmol/L As₂O₃(Figure 4, panels A1 and B1 versus panels A2, A3, B2, and B3). Reduced mitochondrial effect of 1 μmol/L As₂O₃ in HL-60/Bcr-Abl and K562 cells was consistent with decreased Bid cleavage in these cells caused by this dose; however, after treatment with As₂O₃, the cytosolic accumulation of cyt c was similar in the 3 cell types (Figure 3). Table2 also demonstrates the cell-cycle effects of As₂O₃ (1 and 2 μmol/L for 7 days) determined by flow cytometry in the control HL-60/neo versus multidrug resistant HL-60/Bcr-Abl and K562 cells. As shown, treatment with 1 or 2 μmol/L As₂O₃ significantly increased the percentage of HL-60/Bcr-Abl and K562 cells accumulated in the G₂/M phase of the cell cycle. Although in HL-60/neo cells this was not obvious by flow cytometry (Table 2), mitotically arrested HL-60/neo cells were observed by Wright staining and light microscopy (data not shown). In comparison with HL-60/Bcr-Abl and K562, a higher percentage of HL-60/neo cells underwent apoptosis after treatment with As₂O₃. This could also partly explain the low percentage of HL-60/neo cells in the G₂/M phase observed after exposure to As₂O₃. Our data corroborate a recent report that demonstrates that As₂O₃-induced mitotic arrest in myeloid leukemia cells coincides with loss of viability.34 

Compared with the apoptotic rate in Table 1, detected by annexin-V staining, Table 2 shows a lower rate of As₂O₃-induced apoptosis, represented by the percentage of sub-G₁ cells containing hypodiploid amounts of DNA. This is caused by the differences in the 2 methods used to detect apoptosis. Flow cytometry to detect apoptotic cells with hypodiploid DNA content may miss those apoptotic cells that undergo apoptosis in the G₂/M-arrested state and do not lose enough DNA from fragmentation to be detected as hypodiploid. Additional confirmation of the apoptosis data shown in Tables 1 and 2was obtained by performing TUNEL assays on untreated and As₂O₃-treated cells using the in situ cell death detection kit (Boehringer-Mannheim, Indianapolis, IN). TUNEL assay results (data not shown) were consistent with the flow cytometric findings of the percentage sub-G₁ cells containing hypodiploid amounts of DNA (Table 2). As₂O₃-induced mitochondrial ΔΨm, ROS, and apoptosis (as detected by annexin-V staining) were also determined after exposure to HL-60/neo, HL-60/Bcr-Abl, and K562 cells at time points earlier than 7 days (ie, 24 and 72 hours). As shown in Table3, (and compared with Tables 1 and 2), there was a time-dependent increase in the effects of As₂O₃ (1 or 2 μmol/L) on mitochondrial ΔΨm, ROS, and apoptosis of the 3 cell types. In general, the total loss of cell viability detected by PI staining occurred after the mitochondrial effects and annexin-V staining, as has been previously reported.30,35 36 

Previous reports18 indicate that ectopic expression of Bcr-Abl in HL-60 cells is associated with a marked down-regulation of Bcl-2 and an increased expression of Bcl-x_L in HL-60 cells. Treatment with 1 or 2 μmol/L As₂O₃ for 7 days did not cause any significant alterations in Bcl-2 in HL-60/neo or Bcl-x_L in HL-60/Bcr-Abl or K562 cells (Figure5A). Bax, Apaf-1, cIAP, Fas L, and Fas levels in the 3 cell types were also unaffected by treatment with As₂O₃ (Figure 5). However, it is important to note that a dose-dependent decline in p185 Bcr-Abl in HL-60/Bcr-Abl and p210 Bcr-Abl in K562 was clearly observed (Figure 5A). At the higher dose levels of As₂O₃, greater than or equal to 2 μmol/L, this was also observed after exposure to shorter intervals (48 hours) (data not shown). Because Bcr-Abl expression is known to exert resistance against apoptosis, the decline in Bcr-Abl levels may explain why As₂O₃ treatment induced apoptosis in HL-60/Bcr-Abl and K562 cells.

Recent reports42 indicate that hybrid polar compounds and sodium phenylbutyrate, which induce terminal differentiation or apoptosis of leukemic cells, concomitantly induce hyperacetylation of the histones. Based on this, we determined the effect of As₂O₃ on the acetylation status of histones H3 and H4 in HL-60/neo, HL-60/Bcr-Abl, and K562 cells. Immunoblot analysis in Figure 5B demonstrates that, after treatment with 1 or 2 μmol/L As₂O₃ for 7 days, approximately, a 3- to 4-fold increase in the amount of acetylated histones was observed in HL-60/neo, HL-60/Bcr-Abl, and K562 cells. This approximated the amount of histone hyperacetylation seen with treatment of these cells with trichostatin A, a known inhibitor of histone deacetylase (Figure 5B). Figure 5C demonstrates that the hyperacetylation of histones H3 and H4 was evident even after a shorter exposure to As₂O₃ (2.0 μmol/L for 24 hours).

After exposure to 1 or 2 μmol/L As₂O₃, flow cytometric analyses of the cell-surface expression of CD11b, CD33, CD34, and HLA-DR in HL-60/neo, HL-60/Bcr-Abl, and K562 cells were performed. Figure 6 demonstrates that As₂O₃ markedly increased the percentage of cells expressing of the myeloid differentiation marker CD11b in all cell types. Taken together with the data in Table 3, these data demonstrate that there was a progressive increase in the percentage of cells expressing CD11b in all cell types after exposure to 2 μmol/L As₂O₃ from 24 hours to 7 days.These data do not exclude the possibility that thereis a CD11b-positive subgroup of leukemic cells that is relatively insensitive to As₂O₃ and is selectively expanded during treatment with As₂O_3. However, As₂O₃treatment did not affect the expression of CD33, CD34, and HLA-DR (data not shown). Because the differentiation of K562 cells was shown in a previous study35 to be associated with increased intracellular levels of hemoglobin, this was determined in the untreated and the As₂O₃-treated cells. Treatment with As₂O₃ did not induce hemoglobin production or morphologic differentiation in K562, HL-60/Bcr-Abl, or HL-60/neo cells (data not shown).

Data presented here clearly demonstrate that exposure to clinically achievable concentrations of As₂O₃ is able to induce growth inhibition and apoptosis of human myeloid leukemia cells resistant to multiple apoptotic stimuli. In these cells, apoptosis and multidrug resistance were shown to be secondary to diverse mechanisms, including the expression of Bcr-Abl or the overexpression of Bcl-2, Bcl-x_L, MDR, or MRP.15,18,27,29,34As₂O₃-induced apoptosis of HL-60/VCR or HL-60/AR cells was not significantly different from HL-60/neo cells (P > .05). This suggests that As₂O₃is not a substrate for the mdr-1 gene-encoded p-glycoprotein nor is it exported by MRP. Compared with HL-60/neo, however, As₂O₃-induced apoptosis was partially attenuated in HL-60/Bcl-2, HL-60/Bcl-x_L, HL-60/Bcr-Abl, and K562 cells. Consistent with recent reports, our findings demonstrate that As₂O₃ (2 μmol/L for 7 days) was able to induce mitochondrial ΔΨm and cytosolic accumulation of cyt c in HL-60/neo, HL-60/Bcr-Abl, and K562 cells.43,44As₂O₃ also generated PARP, DFF, and Bid cleavage activities of caspases in these cells.9,22,24 This suggests that As₂O₃ is able to induce the cleavage and activity of both the upstream (caspase-8) and the downstream executioner caspase (such as caspase-3) in HL-60/neo, HL-60/Bcr-Abl, and K562 cells.45 A recent report34 indicates that As₂O₃inhibits the binding of guanosine triphosphate to tubulin, its polymerization, and its microtubule formation, resulting in mitotic arrest of myeloid leukemia cells.34 This led to the apoptosis of leukemic cells. Our results confirmed that As₂O₃ treatment produces mitotic arrest (Table 2) and apoptosis (Table 1) of Bcr-Abl-positive and Bcr-Abl-negative myeloid leukemia cells. However, as for other antimicrotubule agents such as paclitaxel and vincristine, the precise mechanism by which the mitotic arrest induced by As₂O₃ is linked to preapoptotic mitochondrial events, cyt c release, and caspase activity remains to be elucidated.46 

Bcr-Abl expression mediates resistance to apoptosis.47 we have reported that HL-60/Bcr-Abl and K562 cells are highly resistant to high-dose ara-C and etoposide-induced mitochondrial ΔΨm, cytosolic accumulation of cyt c caspase activation, and apoptosis.18In addition, ara-C and etoposide fail to alter p210 or p185 Bcr-Abl levels in these cells.18 In contrast, treatment with As₂O₃ significantly down-regulates Bcr-Abl levels in both cell types, which may explain why As₂O₃ causes mitochondrial ΔΨm, accumulation of cyt c in the cytosol, and apoptosis of HL-60/Bcr-Abl and K562 cells. These findings are consistent with a previous report48 demonstrating that the abrogation of Bcr-Abl expression by antisense oligonucleotides selectively eliminates CML blast cells. In addition, the abrogation of Bcr-Abl activity by a relatively specific tyrosine kinase inhibitor CGP57 148B recently has been shown to cause in vitro and in vivo eradication of human Bcr-Abl positive leukemia cells.49 Although ectopic or endogenous Bcr-Abl expression is associated with the up-regulation of Bcl-x_L,18,35,47 our current data show that As₂O₃-mediated declines in Bcr-Abl levels do not down-regulate Bcl-x_L levels. Recently, it has been shown that Bcl-x_L is a caspase-3 substrate and that the cleavage of Bcl-x_L in the loop region releases a c-terminal product that lacks the BH4 homology domain and induces cell death.50,51 However, As₂O₃-induced apoptosis of HL-60/Bcr-Abl and K562 cells was not seen to be associated with Bcl-x_L cleavage. Furthermore, As₂O₃also did not alter Bax, Apaf-1, Fas L, Fas R, and cIAP levels. Collectively, these findings point to As₂O₃-mediated down-regulation of Bcr-Abl as the key perturbation responsible for facilitating the apoptosis of HL-60/Bcr-Abl and K562 cells. Bcl-2 or Bcl-x_Loverexpression (as in HL-60/Bcl-2 or HL-60/Bcl-x_L cells) also inhibits preapoptotic mitochondrial events.9 However, the relative potency of Bcl-2 or Bcl-x_L versus Bcr-Abl for exerting this effect is unknown. As shown in the current study, though As₂O₃ does not lower Bcl-2 and Bcl-x_L levels in HL-60/neo, HL-60/Bcl-2, and HL-60/Bcl-x_L, the mitochondrial toxic effects of As₂O₃ may be potent enough to overcome the inhibitory effects of Bcl-2 and Bcl-x_L and may induce apoptosis of HL-60/neo, HL-60/Bcl-2, and HL-60/Bcl-x_Lcells. In contrast, As₂O₃-induced down-regulation of Bcr-Abl may be necessary for facilitating apoptosis of HL-60/Bcr-Abl and K562 cells.

Recent studies have firmly established that targeted acetylation of the internal lysine residues in the amino-terminal tails of histones relieves nucleosomal repression, which limits the access of the transcriptional machinery to the DNA template and facilitates transcriptional activation.23,50 Histone acetylation is a reversible process.23,52 Histone acetyltransferases transfer the acetyl moiety from acetyl coenzyme A to the lysine residues neutralizing the positive charge and increasing hydrophobicity, whereas histone deacetylases remove the acetyl groups and reestablish the positive charge in the histones.23,52Recently, a class of hybrid polar compounds and sodium phenylbutyrate, which induce terminal differentiation or apoptosis, were shown to induce concomitantly the hyperacetylation of histones by inhibiting histone deacetylase.21,22 Our data demonstrate, for the first time, that clinically relevant concentrations of As₂O₃ induce the hyperacetylation of histones H3 and H4. It is unclear whether this is a direct or an indirect effect mediated through the modulation of transcriptional coactivators or corepressors that may have histone acetyltransferase or deacetylase activity, respectively.42,53 The effect of these corepressors and coactivators is selective. Some promoters and transcription factors are blocked, whereas others are not, by this recruitment of histone deacetylase or histone acetyltransferase.42,53 54As₂O₃-induced histone hyperacetylation may be responsible for altering the transcription of a number of genes, which may collectively mediate As₂O₃-induced growth inhibition and apoptosis. Our studies do not establish whether the As₂O₃-induced down-regulation of Bcr-Abl is transcriptionally or posttranscriptionally regulated. If As₂O₃ affects the transcription of the bcr-abl fusion gene, this may also be mediated directly or indirectly by altered gene-transcription and expression brought about by As₂O₃-induced histone hyperacetylation. These mechanistic issues would have to be resolved by future studies.

In summary, this article highlights the activity of As₂O₃ against leukemic cells that are resistant to apoptotic stimuli either because of the expression of Bcr-Abl or the overexpression of Bcl-2, Bcl-x_L, MDR, or MRP proteins. These findings, as well as As₂O₃-induced declines in Bcr-Abl, suggest that As₂O₃ should be investigated for potential in vivo activity against refractory acute myelocytic leukemia and CML.