Arsenic Trioxide Inhibits Growth of Human Multiple Myeloma Cells in the Bone Marrow Microenvironment ¹

MM³ remains an incurable plasma cell malignancy, despite the use of conventional or high-dose chemotherapy with stem cell support (1, 2). However, recent data suggest that drugs that target both the tumor cell and its microenvironment can overcome drug resistance. For example, thalidomide can achieve responses in 30% of patients whose disease is refractory to all other therapies (3) and appears to have multiple mechanisms of action: direct induction of G₁ growth arrest or apoptosis in MM cells (4); inhibition of secretion of IL-6, the major MM growth and survival factor, triggered by tumor cell binding to BMSCs (5); inhibition of VEGF-induced angiogenesis in the BM (6) as well as tumor cell proliferation and migration (7); and stimulation of natural killer cell anti-MM immunity (8). Proteasome inhibitors also directly induce apoptosis even of drug-resistant MM cells and act in the microenvironment to block MM cell binding to BMSCs, IL-6 secretion, and angiogenesis (9). These and other drugs offer great potential to improve outcome in patients with MM.

As₂O₃ is an old drug that achieves remarkable clinical responses in patients with APL (10), both by causing tumor cell differentiation at low concentrations (0.1–0.5 μm) and by inducing apoptosis at relatively high concentrations (0.5–2 μm; Ref. 11). Because the latter effect is not specific to APL (12, 13), preclinical studies and early clinical testing of As₂O₃ is ongoing in other malignancies, including myeloid leukemia (14), chronic B-cell leukemia (15), MM (16), and some solid tumors. In MM cells, recent evidence suggests that As₂O₃ induces apoptosis associated with caspase-3 cleavage (16, 17). Our prior studies have shown that γ-irradiation triggers MM cell apoptosis dependent on mitochondrial cytochrome c release and mediated via caspase-8, whereas Fas- and Dex-induced apoptosis occurs independent of cytochrome c release (18). Most recently, we showed that Dex-induced MM cell apoptosis is associated with second mitochondria-derived activator of caspases (Smac) release from mitochondria and complex formation of Smac with X-linked inhibitor of apoptosis protein, with resultant caspase-9 activation, downstream caspase-3 cleavage, and apoptosis (19). However, the pathway mediating As₂O₃-induced apoptosis in MM cells is not yet defined.

The BM microenvironment plays a crucial role in the pathogenesis of MM by influencing tumor growth, survival, and drug resistance (20, 21). Adhesion of MM cells to fibronectin confers protection from apoptosis (22), whereas binding of MM cells to BMSCs induces the production of cytokines including IL-6, VEGF, and TNF-α (5, 20, 23). IL-6 is the major known growth and survival factor for MM cells (24, 25), VEGF triggers growth and migration of MM cells (7) and stimulates BM angiogenesis (6), and TNF-α alters the adhesion molecule profile of MM cells and BMSCs to increase adhesion of MM cells to BMSCs and triggers IL-6 transcription and secretion in BMSCs (23). The proliferative effect of these cytokines in MM cells is mediated via Ras/Raf/mitogen-activated protein/ERK kinase/MAPK signaling (20, 25), whereas migration induced by VEGF is protein kinase C dependent (7). Importantly, adhesion of MM cells to BMSCs confers protection against drug-induced apoptosis (22), and IL-6 mediates drug resistance against Dex, but not irradiation (26, 27). We and others have begun to demonstrate the molecular mechanisms whereby both adherence to BMSCs and IL-6 protect against apoptosis (28). To date, however, whether As₂O₃ can overcome these MM survival advantages in the BM microenvironment is not known.

In the present study, we demonstrate that As₂O₃, like other novel biologically based agents, acts both directly on MM cells and in the BM microenvironment. As₂O₃ induces apoptosis even of drug-resistant MM cells, inhibits MM cell to BMSC binding, and abrogates IL-6 and VEGF production in the BM milieu. These studies provide the framework for clinical trials of As₂O₃ in MM.

As₂O₃ (5 mm in PBS) was provided by Cell Therapeutics Inc. (Seattle, WA) and stored at room temperature, and it was diluted in culture media (0.01–100 μm) just before use. Recombinant human IL-6 (Genetics Institute, Cambridge, MA) and TNF-α (R&D Systems, Minneapolis, MN) were reconstituted with sterile PBS containing 0.1% BSA and stored at −20°C.

Human MM cell lines RPMI 8226 and U266 were obtained from the American Type Culture Collection (Manassas, VA). Dex-sensitive (MM.1S) and Dex-resistant (MM.1R) human MM cell lines were kindly provided by Dr. Steven Rosen (Northwestern University, Chicago, IL). Melphalan-resistant (LR5) and doxorubicin-resistant (Dox₄₀) RPMI 8226 cells were kind gifts from Dr. William Dalton (Moffitt Cancer Center, Tampa, FL). All cell lines were cultured in RPMI 1640 (Mediatech, Herndon, VA) containing 10% fetal bovine serum (Harlan, Indianapolis, IN), 2 mm l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin (Life Technologies, Inc., Grand Island, NY). Drug-resistant cell lines were cultured with Dex, melphalan, or doxorubicin to confirm their lack of drug sensitivity.

BM specimens were acquired from patients with MM after obtaining informed consent. Primary MM cells (96% CD38 positive and CD45RA negative) were purified from BM specimens, as described previously (29). MCs separated by Ficoll-Paque (Pharmacia, Piscataway, NJ) were used to establish long-term BMSC cultures, as in our prior studies (9). Briefly, BM MCs were cultured in Iscove’s modified Dulbecco’s medium (Sigma, St. Louis, MO) containing 20% fetal bovine serum, 8 mm l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin.

The inhibitory effect of As₂O₃ on growth of MM cell lines, MM patients’ cells, normal lymphocytes, and BMSCs from MM patients was assessed by MTT (Chemicon International, Temecula, CA) assay. In brief, 3 × 10⁴ cells were cultured in 96-well cell culture plates (Corning Inc., Corning, NY) for 48 h, and 10 μl of 5 mg/ml MTT reagent were then added to each well for 4 h. After 100 μl of isopropanol containing 0.04 n HCl was added, absorbance was measured at 570 nm using a spectrophotometer (Molecular Devices Corp., Sunnyvale CA).

Cell proliferation was measured by [³H]thymidine (NEN Life Science Products, Boston, MA) incorporation. MM cells (2 × 10⁴ cells/well) were incubated in 96-well culture plates at 37°C with or without As₂O₃, Dex, and IL-6. [³H]Thymidine (0.5 μCi) was then added to each well for the last 8 h of 48-h cultures. Cells were harvested onto glass filters with an automatic cell harvester (Cambridge Technology, Cambridge, MA) and counted using a MicroBeta Trilux counter (Wallac, Gaithersburg, MD).

Dual staining with annexin V-FITC and PI was used to detect apoptosis. After treatment with 2 μm As₂O₃ for 0, 6, 12, 24, and 36 h, 1 × 10⁶ MM cells were washed with PBS and resuspended in 100 μl of HEPES buffer containing annexin V-FITC and PI (annexin V-FLUOS staining kit; Roche Diagnostics, Mannheim, Germany). Cells were analyzed by Coulter Epics XL flow cytometry (Beckman Coulter, Fullerton, CA) for the annexin V-FITC-positive and PI-negative apoptotic cell population. Apoptosis was also confirmed by caspase cleavage, as described below.

MM cells were cultured with As₂O₃, harvested, washed, and lysed using lysis buffer containing 50 mm Tris-HCl (pH 7.4), 150 mm NaCl, 1% NP40, 2 mm Na₃VO₄, 1 mm NaF, and protease inhibitor mixture (Complete; Roche Diagnostics). Total cell lysates were subjected to SDS-PAGE and transferred to nitrocellulose membrane (Schleicher & Schuell, Keene, NH). The membranes were immunoblotted with antibodies against caspase-3 (PharMingen, San Diego, CA), caspase-8, cleaved caspase-9 (Asp315; Cell Signaling Technology, Beverly, MA), caspase-9, Bcl-2, Bcl-x_L, Mcl-1 (Santa Cruz Biotechnology, Santa Cruz, CA), and α-tubulin (Sigma). Anti-Fas antibody (CH-11; Medical and Biological Laboratories, Nagoya, Japan) was used as a positive control for induction of caspase-8-dependent apoptosis.

To characterize inhibition of constitutive or IL-6-induced growth signaling by As₂O₃, immunoblotting was also done with anti-phosphorylated JAK1, anti-phosphorylated JAK2 (Biosource International, Camarillo, CA), anti-phosphorylated STAT3, anti-phosphorylated ERK (Santa Cruz), or anti-phosphorylated Akt (Cell Signaling Technology) antibodies, as well as antibodies against their nonphosphorylated counterparts. The immunoblots were visualized using the enhanced chemiluminescence detection system (ECL; Amersham Pharmacia Biotech, Piscataway, NJ). In our studies of signaling cascades, we used higher concentrations (25 μm) of As₂O₃ and IL-6 (10–25 ng/ml) than are clinically achievable, as well as short time (1 h) exposures, as in previous reports (30).

Cell adhesion assays were performed as described previously (9), with some modifications. Briefly, 5 × 10⁴ BMSCs were seeded in each well of 96-well plates and cultured for 24 h. After confirming the development of a confluent adherent monolayer of BMSCs, then As₂O₃ (0, 1, or 10 μm) was added to each well for 12 h, followed by TNF-α (5 ng/ml) for 2 h. After replacing culture media, 5 × 10⁴ MM cells labeled with Na₂CrO₄ (NEN Life Science Products) were added. After incubation for 1 h, each well was washed twice with media, and cells were lysed with 0.5% NP40. Radioactivity in supernatants was counted using a Wizard 3 gamma counter (Wallac).

MM cells (5 × 10⁴ cells/well) were cultured with MM patient-derived BMSC monolayers and As₂O₃ (0, 1, and 10 μm) in 96-well plates for 24 h. IL-6 and VEGF levels in culture supernatants were measured by ELISA (DuoSet; R&D Systems).

Flow cytometric analysis using the Coulter Epics XL (Beckman Coulter) was performed to determine the expression profile of adhesion molecules. BMSCs treated with As₂O₃ (10 μm for 12 h) and TNF-α (5 ng/ml for 4 h) were washed with PBS and incubated with FITC-conjugated anti-CD54 (Beckman Coulter) antibody for 20 min on ice. Cells were then washed, fixed with 1% paraformaldehyde, and evaluated by flow cytometry.

BMSCs were preincubated with As₂O₃ (25 μm) at 37°C for 1 h, and TNF-α (5 ng/ml) was then added. After incubation for 0, 10, or 20 min, BMSCs were harvested, and nuclear extracts were prepared by suspension in 1.0 ml of cold buffer A [10 mm HEPES (pH 7.9), 1.5 mm MgCl₂, 10 mm KCl, 0.1% NP40, 0.5 mm DTT, 1 mm phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, and 5 μg/ml aprotinin] on ice for 15 min. The lysate was vortexed for 10 s and centrifuged for 10 min at 850 × g. The sediment was resuspended in 25 μl of cold buffer B [20 mm HEPES (pH 7.9), 25% glycerol, 420 mm NaCl, 1.5 mm MgCl₂, 0.2 mm EDTA, 0.5 mm DTT, 0.5 mm phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, and 5 μg/ml aprotinin]. Samples were vigorously rocked for 30 min on ice, followed by centrifugation at 12,000 × g for 15 min at 4°C. Supernatants were stored at −70°C until use.

EMSA was carried out using nuclear extracts, as described previously (23). Double-stranded NF-κB consensus oligonucleotide probe (5′-GGGGACTTTCCC-3′; Santa Cruz Biotechnology) was end-labeled with [γ-³²P]ATP (50 μCi at 222 TBq/mm; NEN Life Science Products) and T4 polynucleotide kinase (New England Biolabs, Beverly, MA) in kinase buffer [70 mm Tris-HCl (pH 7.6), 10 mm MgCl₂, and 5 mm DTT] at 37°C for 1 h. Binding reactions containing 1 ng of oligonucleotide and 6 μg of nuclear protein were conducted at room temperature for 20 min in a total volume of 25 μl of binding buffer [10 mm Tris-HCl (pH 7.5), 50 mm NaCl, 1 mm MgCl₂, 0.5 mm EDTA, 0.5 mm DTT, 4% glycerol, and 0.5 μg of poly(dI-dC) (Pharmacia, Peapack, NJ)]. The samples were loaded on a 4% polyacrylamide gel, transferred to filter paper (Whatman International, Maidstone, United Kingdom), and visualized by autoradiography.

Statistical significance of differences observed in drug-treated versus control cultures was determined using the Mann-Whitney U test. The minimal level of significance was P < 0.05.

We first investigated the effects of As₂O₃ at various concentrations on proliferation of MM cell lines. As₂O₃ inhibited the growth of MM cells in a dose-dependent (Fig. 1, A–C) as well as time-dependent (data not shown) fashion. Fifty percent growth inhibition (IC₅₀) in U266 cells and RPMI 8226 cells at 48 h was observed at 3.2 and 2.4 μm As₂O₃, respectively (Fig. 1A). To examine whether there was cross-resistance between As₂O₃ and conventional therapies, we similarly studied drug-resistant MM cell lines. Proliferation of RPMI 8226/DOX₄₀ was not inhibited by doxorubicin (400 nm), whereas RPMI 8226/LR5 cells could be maintained in the presence of melphalan (5 μm; data not shown). Importantly, these cell lines retained sensitivity to As₂O₃, with an IC₅₀ of 3.1 μm for RPMI 8226/Dox₄₀ and 5.8 μm for RPMI 8226/LR5 (Fig. 1A), within the range of clinically achievable concentrations (5–7 μm) after i.v. administration (10). As₂O₃ also suppressed the proliferation of both Dex-sensitive (MM.1S) and Dex-resistant (MM.1R) MM cell lines with IC₅₀ values of 1.8 and 1.1 μm, respectively (Fig. 1B). These data demonstrate that As₂O₃ effectively inhibits the growth of drug-resistant MM cells at clinically achievable levels and can overcome resistance to doxorubicin, melphalan, and Dex.

We next investigated the effects of As₂O₃ on patients’ MM cells. As₂O₃ inhibited the proliferation of MM cells from three patients in a dose-dependent manner, with an IC₅₀ < 5 μm (Fig. 1C). In contrast, peripheral blood MCs from two normal donors were less sensitive (IC₅₀ > 30 μm) than MM cell lines and tumor cells from MM patients (Fig. 1D). These results demonstrate enhanced sensitivity of MM cells versus normal cells to As₂O₃.

To investigate the mechanism of cytotoxic effects caused by As₂O₃, we next performed dual staining with annexin V-FITC and PI. MM cells were incubated with As₂O₃ (2 μm), stained with annexin V-FITC and PI, and examined using flow cytometry. As shown in Fig. 2A, the annexin V-FITC-positive and PI-negative cell fraction gradually increased, with transition to annexin V/PI double-positive component. All MM cells, including drug-resistant cell lines, showed similar profiles (data not shown). Because this result indicated that As₂O₃ induced apoptosis in MM cells, we next investigated the activation of caspases mediating apoptosis (31, 32). Moreover, in both MM cell lines and MM patients’ cells, we observed the cleavage of procaspase-9, but not procaspase-8, followed by cleavage of procaspase-3 (Fig. 2B), suggesting that As₂O₃ caused activation of caspase-3 mainly via caspase-9.

We next investigated the effects of As₂O₃ on constitutive activation of STAT3, which mediates MM cell survival (33). As shown in Fig. 3A, both STAT3 and ERK are constitutively activated in U266 cells, and As₂O₃ inhibited the activation of STAT3 in a dose-dependent manner (Fig. 3A). This effect was also dependent on the duration of As₂O₃ treatment, with almost complete abrogation of constitutive STAT3 activation at 12 h (Fig. 3B). In contrast, activation of ERK was not affected by As₂O₃. This selective inhibition of STAT3 persists for at least 48 h (data not shown).

Because activation of STAT3 mediates up-regulation of Bcl-x_L and Mcl-1 that confer survival in MM cells (33, 34), we next examined the effects of As₂O₃ on members of Bcl-2 family of proteins. As shown in Fig. 3C, As₂O₃ decreased the expression level of Mcl-1, with little, if any, effect on Bcl-x_L or Bcl-2.

Because Dex is one of the most common therapeutic agents for MM, we next determined whether Dex enhanced the inhibitory effect of As₂O₃ on proliferation of MM cells. Dex-sensitive MM.1S cells were incubated with various concentrations of As₂O₃ and Dex for 48 h. The addition of Dex (0.01–0.1 μm) to As₂O₃ (0.01–1 μm) significantly enhanced its inhibitory effect on the growth of MM cells in a dose-dependent fashion (Fig. 4), P = 0.007 at 0.01 m Dex. Although As₂O₃ inhibited proliferation of Dex-resistant MM.1R cells, it did not reverse resistance of MM.1R cells to Dex (data not shown).

IL-6 is a major growth factor for MM cells (24) and protects against Dex-induced apoptosis (28, 35); therefore, we next examined the effects of As₂O₃ in the presence of IL-6. Although IL-6 (50 ng/ml) induced proliferation (stimulation index 1.8) in MM.1S cells, it did not rescue MM cells from apoptosis triggered by As₂O₃. Specifically, As₂O₃ overcomes the protective effects of IL-6 in a dose-dependent manner (Fig. 5A).

Having shown that As₂O₃ can overcome the effects of IL-6, we next investigated the effect of As₂O₃ on IL-6-induced signaling cascades in MM.1S cells. IL-6 (25 ng/ml) was added to MM.1S cells pretreated with As₂O₃ (25 μm) for 1 h, and total cell lysates were prepared at 0, 10, 30, and 60 min. To define the effects on signaling pathways in viable cells, we used higher concentrations of IL-6 and As₂O₃ and shorter time exposures. Trypan blue dye exclusion confirmed >90% viability of treated cells (data not shown). As shown in Fig. 5B, IL-6 triggered phosphorylation of STAT3 in MM.1S cells, and pretreatment of MM.1S cells with As₂O₃ inhibited this effect of IL-6. Because IL-6 activates STAT3 via the gp130 receptor complex and protein tyrosine kinases (25, 36), we next examined whether As₂O₃ pretreatment also inhibited IL-6-induced activation of JAKs. As₂O₃ inhibited IL-6-induced phosphorylation of JAK1 and JAK2 at early time points (1 and 5 min; data not shown), but this inhibitory effect did not persist beyond 10 min (Fig. 5B). Although IL-6 also stimulated MAPK and AKT signaling, As₂O₃ did not inhibit the activation of these pathways (Fig. 5B). These studies suggest selective effects of As₂O₃ on IL-6-triggered signaling cascades.

We next examined the effects of As₂O₃ on the BM microenvironment, where MM cells bind to extracellular matrix proteins and BMSCs, thereby enhancing their growth and survival due to both adhesion and resultant cytokine secretion (20, 22, 37, 38). We first examined the cytotoxic effects of As₂O₃ on BMSCs. As shown in Fig. 6A, BMSCs from two MM patients were resistant to As₂O₃, assayed by MTT at 48 and 72 h in the presence of high concentrations (100 μm) of As₂O₃. To assess whether As₂O₃ inhibits the growth of MM cells adherent to BMSCs, we measured [³H]thymidine uptake of MM cells adherent to BMSCs in the presence of increasing concentrations (0–2 μm) of As₂O₃. As can be seen in Fig. 6B, adherence of MM cells to BMSCs triggered a 1.8-fold increase in MM cell proliferation relative to MM cells alone; importantly, however, this increased proliferation of adherent MM cells was completely inhibited by As₂O₃.

We have reported previously that adhesion of MM cells to BMSCs increases the secretion of IL-6 (38) and VEGF (5, 7) in BMSCs, mediating proliferation and survival of MM cells, as well as angiogenesis in the BM milieu. Binding of MM.1S cells to BMSCs induced significant increases in IL-6 (Fig. 7A) and VEGF (Fig. 7B) secretion compared with BMSCs alone. The cytokine secretion by MM cells alone was negligible. As₂O₃ (1–10 μm) reduced the secretion of both cytokines in cultures of MM cells adherent to BMSCs.

We have reported previously that TNF-α increases the expression of adhesion molecules on both MM cells and BMSCs via activation of NF-κB, resulting in increased MM cell to BMSC binding (23). As shown in Fig. 8A, As₂O₃ did not alter constitutive binding of MM cells to BMSCs. However, As₂O₃ blocked the increased binding of MM.1S cells to BMSCs triggered by TNF-α. To delineate the mechanism of this effect of As₂O₃, we first demonstrated that As₂O₃ blocked TNF-α-induced up-regulation of ICAM-1 on BMSCs (Fig. 8B). Given that TNF-α-induced ICAM-1 expression is conferred via NF-κB (23), we next examined the effect of As₂O₃ on TNF-α-induced NF-κB nuclear translocation in BMSCs. Specifically BMSCs, with or without As₂O₃ pretreatment, were stimulated with TNF-α (5 ng/ml) for 0, 10, and 20 min, and EMSA was performed with nuclear extracts. As can be seen in Fig. 8C, TNF-α-induced nuclear translocation of NF-κB in BMSCs was completely blocked by As₂O₃.

In the present study, we show that As₂O₃ has direct effects not only on MM cells, including drug-resistant MM cells, but also on their growth in the BM microenvironment. As₂O₃ induces apoptosis of MM cells via caspase-9 and overcomes the protective effect of IL-6 in the BM milieu by inhibiting JAK-STAT survival signaling in tumor cells. Furthermore, As₂O₃ reduces TNF-α-induced adhesion to BMSCs and the resultant induced secretion of cytokines (IL-6 and VEGF) that promote MM cell growth, survival, and migration. Importantly, As₂O₃ inhibits the growth even of MM cells adherent to BMSCs at clinically achievable concentrations. These studies provide the rationale for clinical trials of As₂O₃ in MM.

We first demonstrate that As₂O₃ acts directly on MM cells, inhibiting the proliferation of both MM cell lines and patients’ MM cells. As₂O₃ induced apoptosis in MM cells, as evidenced by an increase in the annexin V-positive and PI-negative apoptotic cell population. These results confirm and extend previous investigations (16, 17). Furthermore, our study shows that As₂O₃ activates caspase-9, but not caspase-8, followed by cleavage of procaspase-3, in both MM cell lines and MM patients’ cells. Two major pathways of activation of caspase-3 are reported, i.e., via caspase-8 and/or via caspase-9 (31, 32). Both have been implicated in As₂O₃-triggered apoptosis: caspase-9 in head and neck cancer (39) and caspase-8 in APL (40). Other reports suggested that As₂O₃-induced apoptosis is associated with accumulation of hydrogen peroxide (41), which is enhanced by ascorbic acid (42) or collapse of mitochondrial transmembrane potential (43), followed by release of cytochrome c and consequent activation of caspase-9. Our recent demonstration that Dex induces caspase-9 activation in MM cells (19), coupled with this report implicating caspase-9 in As₂O₃-associated apoptosis, provides further support for the additive anti-MM activities in vitro of Dex and As₂O₃ and the potential clinical utility of combination therapy.

As₂O₃ inhibited the proliferation even of doxorubicin, melphalan, or Dex-resistant MM cell lines. These data suggest that As₂O₃ can overcome conventional drug resistance. As₂O₃ has previously been reported to overcome drug resistance by inducing hyperacetylation of histones H3 and H4 in APL (14) or by inhibiting activation of NF-κB when used with IFN-α in human T-cell lymphotropic virus type I-associated adult T-cell leukemia/lymphoma (44). In MM cells, constitutive STAT3 activation and up-regulation of Bcl-x_L confer resistance to apoptosis and a multidrug-resistant phenotype (33, 45). Moreover, IL-6-induced up-regulation of Mcl-1 is mediated via JAK-STAT signaling, thereby promoting MM cell survival (34, 46). In our study, As₂O₃ reduced IL-6-induced phosphorylation of STAT3 and expression of Mcl-1, but not Bcl-x_L. Mcl-1 is a prosurvival member of the Bcl-2 family (47) that confers resistance to apoptosis induced by several cytotoxic agents in vitro (48), and it is elevated at the time of recurrence in acute leukemia (49). In this context, the suppression of Mcl-1 expression in MM cells may contribute, at least in part, to overcoming conventional drug resistance.

IL-6 can completely rescue MM cells from Dex-induced apoptosis (28, 35), but the present study showed no similar protective effect of IL-6 against As₂O₃, as has recently been reported (16). Our previous studies show that IL-6 protects against Dex-induced apoptosis by activation of the SH2 domain-containing protein tyrosine phosphatase (28) and phosphatidylinositol 3-kinase/AKT signaling (50). To delineate the mechanisms whereby As₂O₃ blocks protection conferred by IL-6, we investigated the effect of As₂O₃ on IL-6 signaling pathways in MM cells. Specifically, IL-6 activates Ras/Raf/mitogen-activated protein/ERK kinase/MAPK (24), JAK/STAT (20, 25), and phosphatidylinositol 3-kinase/AKT (50, 51) signaling pathways in MM cells. Our results show that As₂O₃ inhibits the IL-6-induced phosphorylation of STAT3, thereby inhibiting its DNA binding (52, 53), and also blocks early activation of JAK1 and JAK2 in the JAK/STAT pathway, known to mediate cell survival (20, 33). In contrast, our data show little, if any, suppressive effect of As₂O₃ on MAPK and AKT signaling in MM cells induced by IL-6. Our recent studies indicating that IL-6 confers protection against Dex-induced apoptosis via AKT signaling (50), coupled with the present study showing selective inhibition of As₂O₃ on IL-6-induced STAT3 (but not MAPK or AKT) pathways, suggest distinct mechanisms of As₂O₃ versus Dex killing in MM cells. These preclinical studies provide the rationale for protocols using combinations of As₂O₃ and Dex to enhance clinical anti-MM activity.

Many novel biologically based drugs, including thalidomide and its immunomodulatory derivatives as well as proteasome inhibitors, act on both MM cells and the BM microenvironment to alter growth and survival signals (4, 9), and our study suggests that As₂O₃ has similar effects. First, adhesion of MM cells to fibronectin and BMSCs both localizes MM cells in BM and promotes tumor growth, survival, and drug resistance (20, 21). Although As₂O₃ does not inhibit the constitutive binding of MM cells to BMSCs, we show that As₂O₃ blocks the increase in MM to BMSC adhesion induced by TNF-α, as we have previously shown using proteasome inhibitor PS-341 (9). This is related, at least in part, to blockade of NF-κB-dependent up-regulation of ICAM-1 expression on BMSCs (23). Because adhesion confers resistance to drug-induced apoptosis (22), this inhibition of tumor cell binding may restore drug sensitivity. Second, our previous studies have shown that adhesion of MM cells to BMSCs triggers NF-κB-dependent up-regulation of IL-6 and VEGF secretion, with related MM cell growth and migration (7, 37, 38). In this study, we showed that As₂O₃, like thalidomide (5) and proteasome inhibitor PS-341 (9), abrogates the production of both cytokines triggered by MM cell to BMSC binding. Third, we show here that proliferation even of adherent MM cells is inhibited by As₂O₃. Fourth, we have recently shown that VEGF mediates migration of MM cells and BM neovascularization (6, 7), suggesting that As₂O₃ may also block these processes. A prior report (54) has shown that As₂O₃ also decreases VEGF production by the leukemic cell line HEL, with related inhibition of capillary tube formation. Fifth, both IL-6 (55) and VEGF (56) inhibit the antigen-presenting function of DCs by blocking the differentiation of monocytes to DCs and by inhibiting DC maturation, respectively. Therefore, As₂O₃ may improve the immune-compromised status in MM by overcoming these effects on DCs. Finally, it has also been reported that As₂O₃, as we have observed with thalidomide (8), augments lymphokine activated killer-mediated cytotoxicity (57), suggesting that As₂O₃ also has the potential to enhance both specific and nonspecific immunological reactivity against MM cells.

Recently, preliminary results of Phase I/II clinical trials of As₂O₃ alone or coupled with ascorbic acid in patients with refractory or recurrent MM have been reported (58, 59). In a study with As₂O₃ alone, 7 of 10 evaluable patients had minor decreases in M-protein or stable disease; grade 3 leukopenia, anemia, abdominal pain and diarrhea, fever, and fatigue were observed (58). Another study coupled As₂O₃ with ascorbic acid and reported that six of six patients achieved minor responses or stabilization of disease, with similar toxicities (59). As shown in our results (Fig. 4), combination with Dex enhances the effects of As₂O₃ in vitro; higher response rates and less toxicity are expected using this combination, and a clinical trial is currently on going. Additional large-scale clinical studies are needed to evaluate the clinical efficacy of As₂O₃.

In summary, our data demonstrate multiple direct and indirect effects of As₂O₃ on MM cells in vitro and provide the framework for use of As₂O₃, either alone or coupled with Dex, to overcome drug resistance and improve outcome for patients with MM.