Regulation of apoptosis in human melanoma and neuroblastoma cells by statins, sodium arsenite and TRAIL: a role of combined treatment versus monotherapy
- First Online:
- Cite this article as:
- Ivanov, V.N. & Hei, T.K. Apoptosis (2011) 16: 1268. doi:10.1007/s10495-011-0649-2
Treatment of melanoma cells by sodium arsenite or statins (simvastatin and lovastatin) dramatically modified activities of the main cell signaling pathways resulting in the induction of heme oxygenase-1 (HO-1) and in a downregulation of cyclooxygenase-2 (COX-2) protein levels. Through heme degradation and the production of carbon monoxide and biliverdin, HO-1 plays a protective role in different scenario of oxidative stress followed by mitochondrial apoptosis. Both sodium arsenite and statins could be efficient inducers of apoptosis in some melanoma cell lines, but often exhibited only modest proapoptotic activity in others, due to numerous protective mechanisms. We demonstrated in the present study that treatment by sodium arsenite or statins with an additional inhibition of HO-1 expression (or activation) caused a substantial upregulation of apoptosis in melanoma cells. Sodium arsenite- or statin-induced apoptosis was independent of BRAF status (wild type versus V600E) in melanoma lines. Monotreatment required high doses of statins (20–40 μM) for effective induction of apoptosis. As an alternative approach, pretreatment of melanoma cells with statin at decreased doses (5–20 μM) dramatically enhanced TRAIL-induced apoptosis, due to suppression of the NF-κB and STAT3-transcriptional targets (including COX-2) and downregulation of cFLIP-L (a caspase-8 inhibitor) protein levels. Furthermore, combined treatment with sodium arsenite and TRAIL or simvastatin and TRAIL efficiently induced apoptotic commitment in human neuroblastoma cells. In summary, our findings on enhancing effects of combined treatment of cancer cells using statin and TRAIL provide the rationale for further preclinical evaluation.
Extracellular signal-regulated kinase
Fluorescence-activated cell sorter
Inhibitor of NF-κB
Inhibitor of nuclear factor kappa B kinase
Mitogen-activated protein kinase
Medium fluorescence intensity
Nuclear factor kappa B
Poly (ADP-ribose) polymerase-1
Prostaglandin-endoperoxide synthase 2
Reactive oxygen species
Tumor necrosis factor alpha
TNF-related apoptosis inducing ligand
Regulation of the optimal levels of reactive oxygen species (ROS) and suppression of oxidative stress in living cells requires precise function of numerous protective mechanisms, including induction of gene expression of several antioxidant enzymes, such as catalase, superoxide dismutase-2 (SOD-2), glutathione peroxidase, glutathione reductase and heme oxygenase-1 (HO-1). Inducible HO-1, as well as constitutive HO-2, catalyzes the first rate-limiting step of heme degradation producing carbon monoxide (CO) and biliverdin, which is further converted to bilirubin that possesses strong anti-oxidant activity [1, 2]. Furthermore, CO inhibits enzymatic activities of numerous hemoproteins resulting in downregulation of intracellular respiration and ROS production. The ability of HO-1 to catabolize free heme prevents induction of the mitochondrial apoptotic pathway that could operate as the final resolution to maintain homeostasis at the whole organism level via programmed death of particular cell populations [3–5]. Since inflammation is often linked with high levels of ROS production, the induction of HO-1 expression and enzymatic activity is also involved in the protective anti-inflammatory response [2, 6]. On the other hand, cancer development and progression (including melanoma) is linked with hypoxia, suppression of mitochondrial respiration, increased production of ROS, which are inducers of genomic instability, and establishing a general pro-inflammatory phenotype that is maintained in cancer cells through gene expression of the proinflammatory cytokines, such as IL6, IL1β and TNFα and the corresponding receptors . Paradoxically, suppression of the pro-inflammatory response of cancer cells could substantially block tumor development  and sensitize cancer cells to death receptor-mediated apoptosis .
In our previous investigations, we used sodium arsenite treatment (1–5 μM) for induction of apoptosis in human melanoma cells [10, 11]. Similar treatment was successfully used for therapy of acute promyelocytic leukemia (APL) and multiple myeloma (MM) [12, 13]. However, the mitochondrial apoptotic pathway was induced only at relatively low levels by clinically proved concentrations (2–5 μM) of sodium arsenite in most melanoma lines and required additional proapoptotic sensitization through specific suppression of the cell survival pathways, such as MEK-ERK, PI3K-AKT or IKK-NF-κB [10, 11].
In recent years, effects of statins (popular drugs that are widely used to down-regulate cholesterol production) on mitochondrial function, ROS production and HO-1 induction have been extensively investigated [14, 15]. In the present study, we further investigated a role of HO-1 suppression in the substantial upregulation of sodium arsenite- or statin-induced cell death in human melanoma cells. We found greater sensitization and killing of melanoma cells through combined treatment of cancer cells by the exogenous TRAIL and statins. Such dramatic sensitization of TRAIL-Receptor (TRAIL-R) mediated apoptosis by sodium arsenite  or by statins (the current study) offers a new modality for a possible therapy of melanoma, incident number of which progressively increased in USA and worldwide during the last 50 years .
In spite of the remarkable progress in the investigation of carcinogenesis and treatment of melanoma based on the application of specific inhibitors of permanently active mutated BRAF (V600E) [18–20], there is a critical necessity for alternative treatment of melanoma cells carrying wt BRAF (that represents 40–60% of melanoma cases)  and for overcoming resistance to specific inhibitors of BRAF (V600E) that could occur after several months of a successful treatment.
A role of HO-1 expression in protection of MEF against sodium arsenite-induced apoptosis
Upregulation of sodium arsenite-induced apoptosis in melanoma cells
As an alternative approach to increase apoptotic response, we used an inhibitor of enzymatic activity of HO-1, Zn(II) containing Protoporphyrin IX [PPIX(Zn)]. Sodium arsenite (5 μM), in combination with PPIX(Zn) at a dose of 20 μM, additionally increased (from 23 to 45%) apoptotic levels of WM793 cells 24 h after treatment (Fig. 3c). The combination of 5 μM sodium arsenite and 20 μM PPIX(Zn) was most efficient for apoptotic induction in several melanoma lines, including resistant LOX, while relatively mild effects were observed in normal fibroblasts and melanocytes (Fig. 3d, e). Hence, results of genetic and enzymatic inhibition of HO-1 demonstrated substantial upregulation of arsenite-induced apoptosis in human melanoma cells. There was a narrow window of concentrations of PPIX(Zn) (10–20 μM), when apoptotic response in melanoma cells induced by 5 μM sodium arsenite in the presence of PPIX(Zn) was substantially higher, than in normal cells. An additional increase in a concentration of PPIX(Zn) was quite toxic for normal cells and caused massive cell death by necrosis. Taken together, these results indicated that HO-1 activation was necessary, but not sufficient for protection of melanoma cells against apoptosis induced by sodium arsenite treatment. Suppression of HO-1, however, was linked with substantial upregulation of arsenic-induced apoptotic levels in human melanoma cell lines.
Sodium arsenite treatment further sensitizes melanoma cells to TRAIL-induced apoptosis: a probable role for COX-2 downregulation
Statins as inducers of HO-1 expression and cell death of cancer cells
To extend our observations regarding a regulatory role of HO-1 for apoptotic signaling, we used statins as inducers of HO-1 expression in melanoma cells. Statins are widely prescribed drugs for decreasing cholesterol levels, because they inhibit 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMG-CoA reductase), the rate-limiting enzyme, which controls production of mevalonate, an intermediate of the biosynthesis of cholesterol . Furthermore, mevalonate is a precursor of several lipid isoprenoid intermediates, such as geranylgeranyl pyrophosphate (GGPP) and farnesyl pyrophosphate (FPP), which are critically important for posttranslational modification and proper function of small GTP-binding proteins, including Rho, Rac and Ras that play pivotal role in cell signaling in normal and cancer cells . Furthermore, via control of cholesterol biosynthesis and cell signaling, statins demonstrated a strong anti-inflammatory potential that also included the induction of HO-1 expression in different cell systems [14, 29].
Determination of NF-κB-dependent and GAS/STAT-dependent luciferase activity further demonstrated statin-dependent suppression of the general NF-κB transacting activity in both WM793 and WM9 cells. STAT-dependent luciferase activity was also suppressed in WM9 cells and was almost undetectable in the control and statin-treated WM793 cells (Fig. 5c, d). In summary, these results demonstrate a rapid sensitization of statin-treated cells to apoptosis through both the endogenous and exogenous pathways due to numerous changes in apoptosis-related proteins.
Induction of apoptosis by atorvastatin in melanoma cells was relatively low, compared to simvastatin (Fig. 6d). On the other hand, the HO-1 inhibitor, 20 μM PPIX(Zn), further upregulated levels of apoptosis induced by 20–40 μM atorvastatin and simvastatin in WM793 and WM9 melanoma cells (Fig. 6d). In contrast, zVAD-fmk (50 μM), a universal caspase inhibitor, substantially decreased levels of simvastatin-induced apoptosis in melanoma cells (Fig. 6d). Using control and HO1 deficient WM793 cells (see Fig. 2b) we also observed upregulation of apoptotic levels in HO1 deficient cells for a range of simvastatin concentrations (10–40 μM) (Fig. 6e). In summary, these results demonstrate pronounced statin-induced apoptosis (at high doses of statin) in WM793 and WM9 cells that could be further increased by suppression of HO-1 expression/activation. These data identified HO-1 expression, which operated in concert with other defensive mechanisms, as an important factor protecting against statin-induced apoptosis. Interestingly, simvastatin at relatively low concentration (5 μM) notably upregulated arsenite induced apoptosis, possibly via inhibition of ERK activation. U0126, a specific inhibitor of MEK-ERK pathway, demonstrated a similar effect (Supplementary Fig. 1).
We performed analysis of cell death induction by statins in normal human fibroblasts IMR-90, human melanocytes and several melanoma lines. Simvastatin at doses of 10–40 μM induced different levels of apoptosis in cancer cell lines, in the range between 10 and 70% 24 h after treatment: high levels of apoptosis were observed for WM793, FEMX and WM9 melanoma lines, while a slight response, which was similar to the response of normal cells, was detected for WM35 and HHMSX melanoma lines; a modest response was measured in WM164 and WM852 melanoma lines (Fig. 7d). While lovastatin produced a similar induction of apoptosis, apoptotic effect of atorvastatin (10–40 μM) was not well pronounced in human melanomas (Supplementary Fig. 2). We would like to highlight that WM793 and WM9 cells contain mutated BRAF (V600E), while FEMX and WM852 cells have normal BRAF . Our results indicated that melanoma cells with normal BRAF could also be treated and killed by simvastatin at relatively high concentrations.
Statin doses routinely used for decreasing cholesterol levels in human patients are 1–2 μM, while specific apoptotic effects for some melanoma cell lines were well pronounced only at doses 20–40 μM. Next we used statins at lower doses (5–20 μM) in a combination with TRAIL for an additional upregulation of TRAIL-induced apoptosis in cancer cells.
Statin treatment sensitizes melanoma cells to TRAIL-induced apoptosis
We further used a combination of TRAIL (50 ng/ml) and simvastatin (20 μM) for upregulation of apoptosis in additional melanoma lines (Fig. 8d). Normal fibroblasts and melanocytes, as well as WM164 and HHMSX melanoma cells, exhibited low levels of apoptosis after a combined treatment, due to low surface expression of TRAIL-R2/DR5. In contrast, WM793, LU1205, A375 and WM9 melanoma cells demonstrated enhanced apoptosis; medium apoptotic induction was observed in WM35, FEMX and WM852 cells (Fig. 8d). Additive effects of combined treatment with TRAIL and simvastatin on upregulation of apoptosis were likely dependent on several parameters such as the level of TRAIL-R surface expression, efficiency of the death-inducing signaling complex (DISC) formation , suppression of anti-apoptotic proteins, cFLIP-L and COX-2, and changes in NF-κB- and STAT3-dependent anti-apoptotic gene expression. Our results highlighted the usefulness of combined treatment for apoptotic induction in cell culture and could serve as a basis for subsequent in vivo experiments.
Sodium arsenite or statin treatment sensitizes neuroblastoma cells to TRAIL-induced apoptosis
HTB-11 neuroblastoma cells were relatively resistant to statin treatment at a dose range of 1–20 μM. However, a combination of TRAIL (50 ng/ml) and simvastatin (20 μM) or TRAIL and atorvastatin (20 μM) demonstrated efficient for upregulation of apoptosis in HTB-11 neuroblastoma cells, compared to A375 melanoma cells (Fig. 9d). In contrast to A375 cells, HTB-11 cells produced and secreted high levels of the proinflammatory IL6. Simvastatin treatment did not decreased IL6 levels in the cell growth media, as we previously observed for WM793 cells (see Fig. 7c), indicating that anti-inflammatory effects of simvastatin were limited.
Hence, results obtained demonstrated a significant upregulation of TRAIL-mediated apoptosis by combined treatment with simvastatin not only in melanoma, but also in neuroblastoma cells.
Melanoma is the fifth most common cancer in men and the seventh most common cancer in women. It is estimated that 68,130 people in the United States were diagnosed in 2009, and 8,700 died because of it (Melanoma Cancer Overview by American Cancer Society, November 2010; www.cancer.org). In spite of substantial progress in understanding the molecular mechanisms of melanoma carcinogenesis and extensive development of new treatment modalities during the last 5 years, induction of apoptotic commitment in metastatic melanomas and overcoming resistance to new targeting treatment [for example a resistance to specific BRAF (V600E) inhibitors] remain significant problems [18, 37–39]. Our general expectation was that combined treatment based on activation of the TRAIL/TRAIL-R apoptotic pathway, which was quite specific for cancer cells, together with metabolic inhibitors could be potentially beneficial approaches for cancer therapy. In the present study, we have been focused first on suppression of inducible HO-1 activity that could result in further upregulation of ROS-induced apoptosis in human melanomas. Surprisingly, sodium arsenite and statins could affect many common targets, exhibiting anti-inflammatory effects through suppression of the IKK-NF-κB pathway , its critical transcriptional target COX-2  and a dramatic induction of HO-1 activity. Stress-induced HO-1 activation is often a result of mitochondrial targeting followed by release of heme-containing cytochrome c, which should be degraded: holoenzyme through the proteasome system  and heme by HO-1. Furthermore, HO-1 could be further involved in the maintenance of the anti-inflammatory response by suppression of COX-2 enzymatic activity and prostaglandin-E2 production. An alternative pathway that operates in parallel, a cytochrome c interaction with the apoptosome, results in the activation of the caspase-9-dependent apoptotic pathway that ultimately determines a life-or-death balance . Even though our results demonstrated enhanced upregulation of the mitochondrial apoptotic pathway by HO-1 suppression, a challenging problem that remains is the absence of suitable physiological inhibitors of HO-1, because Zn containing PPIX (widely used in cell culture experiments) is cytotoxic for in vivo treatment.
Our observations and published data also demonstrated that monotreatment with statin induced pronounced apoptotic death in some melanoma cell lines only at relatively high doses (5–20 μM) . Animal experiments with the B16F10 mouse melanoma model exhibited a suppression of tumor growth by treatment with 1.6 μM simvastatin . However, a regular statin treatment of patients for the control of cholesterol levels appears does not display a pronounced anti-melanoma protection [46, 47]. Furthermore, multiple side effects of statin treatment for normal cells linked with mitochondrial destruction, apoptosis and the subsequent myopathy were reported . Therefore, solid evidence does not exist that statin monotherapy would be efficient for in vivo treatment.
An alternative approach was an upregulation of TRAIL-induced apoptosis in melanoma cells with additional sensitization of cancer cells by sodium arsenite  or statins (in the present study). Specific inhibitors of IKK-NF-κB  or JAK2-STAT3 , which exhibited a strong antiinflammatory response, specific inhibitors of pro-inflammatory regulators, such as COX-2 inhibitors, and more general inhibitors of the inflammatory response, like resveratrol [33, 50], curcumin  and several others, could be efficient sensitizers of cancer cells to TRAIL-induced apoptosis. Statins with certain antiinflammatory activities [29, 41] might play a similar role in upregulation of death-receptor-mediated apoptosis in melanoma and neuroblastoma cells. Costimulatory effects of statins on TRAIL-induced apoptosis in glioblastomas  and on FasL-mediated apoptosis of smooth muscle cells were previously observed . Interestingly, that through inhibition of COX-2 activity, HO-1 alone had a proapoptotic role in TRAIL-mediated apoptosis in melanoma cells. That was in contrast to the antiapoptotic role of HO-1 in the mitochondria-dependent apoptosis induced by sodium arsenite or statins at higher doses. The proapoptotic effect of HO-1 was very similar with NS398-mediated suppression of COX-2 activity, which also sensitized melanoma cells to TRAIL-mediated apoptosis (see Fig. 4e). The analogous effects of COX-2 inhibitors on TRAIL-induced apoptosis in different cancer models were previously described .
Even though proapoptotic effects of statins for melanoma cells have been previously reported in several publications [44, 55, 56], sensitization of melanoma cells to TRAIL-induced apoptosis by statins, which was investigated in our study, potentially could be a new modality for melanoma therapy. It was especially important, that combined treatment of statin and TRAIL was also quite effective for killing BRAF wt melanoma cell lines, such as WM852 and FEMX (see Fig. 8), where mutation specific (BRAF V600E) inhibitors did not work.
Furthermore, numerous data are available about proapoptotic effects of arsenic , TRAIL alone and TRAIL in different combinations for induction of apoptosis in human neuroblastoma cells . Combined treatment of neuroblastoma cells with TRAIL and sodium arsenite or TRAIL and simvastatin that was very efficient in our in vitro study (see Fig. 9) might also serve as a potential treatment modality for induction of high levels of apoptosis in this type of tumor. In summary, our results further highlighted significance of combined targeting of cell signaling pathways for efficient and specific elimination of cancer cells.
Materials and methods
Sodium arsenite, simvastatin, atorvastatin, mevalonic acid (lithium salt) and GGPP were obtained from Sigma-Aldrich (St. Louis, MO, USA). Human Killer-TRAIL was purchased from Axxora (San Diego, CA, USA). Lovastatin was purchased from Calbiochem/EMD Chemicals (San Diego, CA, USA). COX-2 inhibitor NS398 was obtained from Cayman Chemical (Ann Arbor, MI, USA). Human recombinant IL6 was obtained from R&D Systems (Minneapolis, MN, USA). Zn(II) containing Protoporphyrin IX, an inhibitor of HO-1, was purchased from Frontier Scientific (Logan, UT, USA).
The primary human embryonic lung fibroblasts IMR-90 (Coriell Cell Repository, Camden, NY, USA) were maintained in medium supplemented with 15% fetal bovine serum, vitamins, non-essential amino acids and antibiotics. Normal human melanocytes were obtained from the Department of Dermatology, Yale University (New Haven, CT, USA) and maintained in TICVA medium. Human melanoma cell lines LU1205 (also known as 1205lu), WM9, WM35, WM164, WM793, WM852 , FEMX, LOX, HHMSX  and A375 and human neuroblastoma cell line HTB-11 (SK-N-SH) were maintained in DMEM medium supplemented with 10% fetal bovine serum, l-glutamine and antibiotics. Cells were grown at 37°C with 5% CO2. A375 and HTB-11 cell lines were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA).
Transfection and luciferase assay
The NF-κB luciferase reporter containing two κB binding sites, and STAT-Luc reporter containing three repeats of GAS sites from the Ly6A/E promoter were used to determine NF-κB and STAT transactivation, respectively. Transient transfection of different reporter constructs (1 μg) together with pCMV-β-galactosidase (0.25 μg) into 5 × 105 cells was performed using Lipofectamine (Life Technologies-Invitrogen, Carlsbad, CA, USA). Proteins were prepared for β-Gal and luciferase analysis 16 h after transfection. Luciferase activity was determined using the Luciferase assay system (Promega, Madison, WI, USA) and was normalized based on β-galactosidase levels.
Surface levels of TRAIL-R1/DR4 and TRAIL-R2/DR5 were determined by staining with the PE-labeled mAbs from eBioscience (San Diego, CA, USA). A FACS Calibur flow cytometer (Becton Dickinson, Mountain View, CA, USA) combined with the CellQuest program was used to perform flow cytometric analysis.
For induction of apoptosis, cells were exposed to sodium arsenite (1–20 μM) or statin (1–80 μM) alone or in the presence of protoporphyrin IX (ZnII) at a dose 10–40 μM. Cells were also exposed to soluble TRAIL (50 ng/ml) alone or in a combination with sodium arsenite (5 μM) or simvastatin (5–20 μM). Apoptosis was then assessed by PI staining and quantifying the percentage of hypodiploid nuclei (pre-G1) using FACS analysis.
Western blot analysis
Total cell lysates (50 μg protein) were resolved on SDS-PAGE, and processed according to standard protocols. The monoclonal antibodies used for Western blotting included: anti-β-Actin (Sigma, St. Louis, MO, USA); anti-FLIP (NF6) (Axxora, San Diego, CA, USA); anti-caspase-8, anti-caspase-9, anti-caspase-3 (Cell Signaling, Danvers, MA, USA); anti-COX-2 (Cayman Chemical, Ann Arbor, MI, USA). The polyclonal antibodies used included anti-phospho-p44/p42 MAP kinase (Thr202/Tyr204) and anti-p44/p42 MAP kinase; anti-phospho-AKT (Ser473) and anti-AKT; anti-p65 NF-κB, anti-STAT3, anti-phospho-STAT3 (Tyr705); anti-PARP-1; anti-Bid (Cell Signaling, Danvers, MA, USA); anti-HO-1 (Enzo Life Sciences, Plymouth Meeting, PA, USA). The secondary antibodies were conjugated to horseradish peroxidase; signals were detected using the ECL system (Thermo Scientific, Rockford, IL, USA).
Antibody pairs used in sandwich ELISA for this study were all commercially available. Kits to detect IL6, TNFα and IL1β were from Life Technologies/Invitrogen (Carlsbad, CA, USA).
RNAi targeting of HO-1 mRNA
The empty vector pSR-GFP/Neo was obtained from Oligoengine (Seattle, WA, USA). RNAi of 19 nucleotides (AGATTGCCCAGAAAGCCCT), designed to target human HO-1 mRNA within nucleotides 526–544 were expressed using pSR-GFP/Neo (HO-1-RNAi) plasmid construct, which also produced a marker GFP protein. Human melanoma cells WM793 have been used for HO-1 targeting. Melanoma cells were transfected with indicated expression vectors using Lipofectamine (Life Technologies/Invitrogen, Carlsbad, CA).
For detection of apoptotic levels after treatment with sodium arsenite or statin, GFP-positive stably transfected cells were stained by PE-labeled Annexin-V (red) (BD-Pharmingen, San Diego, CA, USA). A FACS Calibur flow cytometer (Becton Dickinson, Mountain View, CA, USA) combined with the CellQuest program was used to perform flow cytometric analysis of double-stained cells.
Data from 3 to 4 independent experiments were calculated as means and standard deviations. Comparisons of results between treated and control groups were made by the Students’ t tests. A P value of 0.05 or less between groups was considered significant.
We would like to thank Dr. S. Snyder for HO-1 Null fibroblasts, Drs. M. Herlyn and Z. Ronai for melanoma cell lines, Dr. Y. Chai for ELISA detections of cytokines, Drs. H. B. Lieberman and J. A. Meador for a critical reading of the manuscript and discussion. This research was supported by Superfund Grant ES 10349.
Conflicts of interest
No potential conflicts of interest were disclosed.