Retinoic acid dramatically enhances the arsenic trioxide-induced cell ...

The Hematology Journal (2001) 2, 127 ± 135 ã 2001 The European Haematology Association All rights reserved 1466 ± 4680/01 $15.00 www.nature.com/thj

Retinoic acid dramatically enhances the arsenic trioxide-induced cell cycle arrest and apoptosis in retinoic acid receptor a-positive human T-cell lymphotropic virus type-I-transformed cells Nadine Darwiche*,1,2, Marwan El-Sabban3, Rana Bazzi2, Rihab Nasr2,4, Suha Al-Hashimi5, Olivier Hermine6, Hugues de TheÂ7 and Ali Bazarbachi*,2,4 1

Department of Biology, American University of Beirut, Beirut, Lebanon; 2Department of Biochemistry, American University of Beirut, Beirut, Lebanon; 3Department of Human Morphology, American University of Beirut, Beirut, Lebanon; 4Department of Internal Medicine, American University of Beirut, Beirut, Lebanon; 5Department of Pathology and Laboratory Medicine, American University of Beirut, Beirut, Lebanon; 6CNRS URA 1461 and Department of Hematology, Necker Hospital, Paris, France; 7UPR 9051 CNRS (Laboratoire associeÂ au comiteÂ de Paris de la ligue contre le cancer), Saint-Louis Hospital, Paris, France

Introduction: Adult T-cell leukemia/lymphoma, caused by the human T-cell lymphotropic virus type I, is an aggressive neoplasm of mature activated T cells that is generally resistant to conventional therapy. While arsenic trioxide (As) inhibits the growth and induces apoptosis in HTLV-I-infected T cells, synergistically, when combined with interferon-a, variable eects on growth with all trans retinoic acid treatment have been reported in ATL-derived cell lines and fresh ATL cells. In this study, we investigate the eects of ATRA alone or in combination with As in HTLV-I-transformed cells. Materials and methods: Four HTLV-I-transformed cell lines (HuT-102, MT2, C8166 and C91PL) were treated with dierent doses of ATRA alone or in combination with As for one to three days. Cell growth was assessed by cell count with 3H-thymidine incorporation. Cell cycle distribution was assessed by propidium iodine-labeled DNA content by ¯ow cytometry. Apoptosis was evaluated by Hoechst nuclear staining and annexin-V binding assays. Expression of retinoid receptors, the viral transactivator Tax, and the proteins bcl-2 and IkBa proteins, was analysed by Western blot. Results: Only C8166 cells were sensitive to the ATRA-induced growth inhibitory eect while HuT-102, MT2, and C91PL were resistant to ATRA treatment (up to 1075 M). The retinoid X receptor a and the retinoic acid receptor g (RARg) proteins were expressed in all four cell lines, while RARa protein was only detected in the HuT-102 and C8166 cells. The combination ATRA/As showed a highly synergistic eect on HuT-102 cells, and, to a lesser extent, on C8166 cells and resulted in a dramatic inhibition of cell proliferation and induction of massive apoptosis in HuT-102 cells, associated with caspase activation. While ATRA alone had no eect on Tax and IkB-a protein levels, ATRA increased the As-induced Tax degradation and the up-regulation of IkB-a protein. In contrast, the expression of bcl-2 protein was not signi®cantly aected by any of the treatments. Conclusion: Our data provide a rationale for combined ATRA and As-therapies in ATL patients refractory to conventional therapy and expressing RARa in their leukemic cells. The Hematology Journal (2001) 2, 127 ± 135 Keywords:

ATL; leukemia; retinoic acid; retinoic acid receptors; arsenic; apoptosis

Introduction Adult T-cell leukemia-lymphoma (ATL) is an aggressive malignancy of mature activated T cells associated *Correspondence: Nadine Darwiche, Biology Department, American University of Beirut, PO Box 11-0236, Beirut, Lebanon; Tel: +961 3 860548; Fax: +961 1 744461; E-mail: [email protected] or Ali Bazarbachi, Department of Internal Medicine, American University of Beirut, PO Box 113-6044, Beirut, Lebanon; Tel: +961 361 2434; Fax: +961 134 5325; E-mail: [email protected] Received 17 August 2000; accepted 17 December 2000

with human T-cell lymphotropic virus type I (HTLVI).1 HTLV-I infection appears to represent the ®rst event of a multi-step oncogenic process where the regulatory protein Tax plays a major role.2 ATL carries a poor prognosis because of its resistance to chemotherapy and a severe immunosuppression, predisposing patients to opportunistic infections. Important advances in the treatment of ATL have been reported with the combination of zidovudine and interferon-a (IFN), which improved the response rate

Arsenic; retinoic acid; and retinoic acid receptors in ATL N Darwiche et al

128

in patients and prolonged their survival.3 ± 5 Unfortunately, most of the patients relapsed resulting in a search for alternative or complementary therapies. Recently, promising in vitro results in HTLV-Itransformed cells were obtained with retinoids or arsenic compounds. Vitamin A and its derivative `retinoids' play an essential role in development, cell growth, and dierentiation.6,7 Retinoids have been widely used as chemopreventive and therapeutic agents because of their strong antiproliferative and dierentiative activities against certain types of cancers.8 In addition, retinoids can in¯uence the growth and dierentiation of various hematopoietic progenitor cells.9 Several lines of evidence suggest that ATRA and its receptors are involved in hematopoietic regulation, as judged by the RARa-mediated terminal dierentiation of HL-60 cells into granulocytes by ATRA.10,11 Two biologically active stereoisomers of retinoic acid have been identi®ed, ATRA and 9-cis RA (9C-RA), and shown to bind to speci®c nuclear receptors that are members of the steroid/thyroid hormone receptor superfamily.12,13 Two families of nuclear retinoic acid receptors (RAR) have been described, each comprising three receptor types (a, b, and g). The ®rst family of RAR binds ATRA and 9C-RA with high anity.14 Whereas, the second family, retinoid X receptors (RXR), only bind 9C-RA with high anity.15,16 Variable eects of ATRA have been reported in ATL-derived cells.17 Signi®cant growth inhibition by ATRA was observed in some HTLV-I-transformed cell lines but not in HTLV-I-negative lines. Moreover, ATRA slows down the growth of freshly isolated ATL cells but not that of normal peripheral blood mononuclear cells.18 In addition, ATRA, 9C-RA, and 13CRA have all been shown to induce cell growth inhibition, G0/G1 cell cycle arrest and apoptosis in some IL-2-dependent ATL cells.19 This eect is associated with increased p21 protein levels and retinoblastoma gene product hypophosphorylation.19 ATL patients were classi®ed into three groups: hypersensitive, sensitive, and resistant according to the dierence in response of their PBMC to ATRA treatment.18 All trans retinoic acid is currently used in combination with chemotherapy as a standard induction therapy in acute promyelocytic leukemia (APL).20 This combination induces a high response rate and cures the majority of APL patients. In vitro and in vivo results show that ATRA induces terminal dierentiation of APL blasts and APL-derived cell lines which is associated with degradation of the PML-RARa fusion oncoprotein.20 For the few APL patients who relapsed after ATRA/chemotherapy combination, a high response rate has recently been reported with the use of arsenic trioxide (As) which also degrades PML-RARa oncoprotein.21,22 Moreover, in a mouse model of APL, the combination of ATRA/As induces a long-term clinical and molecular remission that is con®rmed by a negative PCR for the APL-speci®c t(15;17) translocation.23 This suggests a strong synergy between ATRA The Hematology Journal

and As in APL blasts and the possibility of curing APL without chemotherapy. It has been shown that As inhibits proliferation and induces apoptosis in HTLV-Itransformed cell lines and in fresh ATL cells.24,25 However, patients with APL who are resistant to ATRA and conventional chemotherapy can still respond to As, and ATRA in combination with As is currently tested in these patients as a chemotherapy-free alternative. Therefore, we investigated the eects of ATRA/As combination on HTLV-I-transformed cells. In this report, we show a potentiating eect of ATRA on As inhibition of cell proliferation and induction of apoptosis in two RARa-expressing HTLV-I-transformed cells.

Materials and methods Cell lines and culture conditions HuT-102, MT2, C91PL and C8166 cell lines are HTLV-I-infected CD4+ T cells that constitutively express the interleukin-2 (IL-2) receptor alpha chain (CD25) and are independent of IL-2 for growth. The ®rst three produce HTLV-I proteins and virus while C8166 cells only produce the viral Tax protein. Cells were grown in RPMI-1640 culture medium supplemented with 10% heat inactivated fetal calf serum and antibiotics (GIBCO BRL, Gaithersburg, MD, USA). A seeding cell concentration of 16105 cells/ml was chosen for all experiments.

Reagents and drugs Arsenic trioxide (Sigma Chemical Co. St. Louis, MO, USA) was prepared as a stock solution (1074 M) in normal saline and stored at +48C. Working stocks were diluted in complete RPMI culture medium to a concentration of 1 mM that corresponds to the pharmacological level in vivo.21 All trans retinoic acid (Sigma) was prepared as a stock solution (361072 M) in dimethylsulfoxide (DMSO) and stored in amber tubes at 7808C. The concentration used in most experiments was 361076 M that is below the acceptable pharmacological level of 1075 M. The ®nal concentration of DMSO never exceeded 0.1% and this concentration showed no eect on the proliferation and CD25 expression in all tested cell lines. All experiments using ATRA were performed under yellow light (l4500 nm) to prevent photoisomerization of ATRA.

Proliferation assay Cell growth was assessed by the incorporation of 3Hthymidine (Amersham Pharmacia Biotech UK, Buckinghamshire, UK), cell count, and trypan blue dye exclusion protocols (hemocytometer). 3H-thymidine uptake was measured in cells grown in 96-well plates (Nunc, Naperville, IL, USA). The dierent test drugs were added at the indicated concentrations at the initiation of cultures. After 24, 48, or 72 h of


129

treatment, H-thymidine (1 mCi/well) was added for the last 4 h. Cells were harvested on ®lters using a PHD cell harvester (Cambridge Technology, Watertown, MA, USA) and dried overnight at 508C. Bio-HP liquid scintillation ¯uid (Fisher Scienti®c, Fair Lawn, NJ, USA) was then added to the ®lters and 3Hthymidine was measured using a Wallac 1409 liquid scintillation counter (Pharmacia-LKB, Uppsala, Sweden). DNA synthesis was quanti®ed by measuring cellular incorporation of 3H-thymidine and expressed as counts per minute. Incorporation of 3H-thymidine is shown as percentage of corresponding control. Results are representative of at least two independent experiments and performed in quadruplicate. Cell counts represent the mean of triplicate wells. 3

48C from treated and untreated cells. Protein concentrations were determined using DC protein assay (Bio-Rad Laboratories, Hercules, CA, USA). Total cellular proteins were boiled and loaded (30 mg/lane) onto a 10% SDS-polyacrylamide gel, subjected to electrophoresis, and transferred to nitrocellulose membranes (Bio-Rad). Another protein gel was run in parallel and stained with Coomassie blue to ensure equal loading. The blots were blocked for 1 h at room temperature in 5% skimmed milk in TBS (50 mM Tris-HCl and 150 mM NaCl). The membranes were then reacted for 2 h at 378C with the following antibodies: bcl-2 and IkB-a

Cell cycle analysis Cells were cultured in 6-well plates (Nunc) and were collected at 24 or 48 h post-treatment. Cells were then washed twice with cold phosphate-buered saline (PBS), ®xed in ice cold 100% ethanol, and stored for 24 h at 48C. Subsequently, they were rinsed with PBS, incubated for 1 h in Tris-HCl buer (pH 7.4) containing 50 units RNase A and then stained with propidium iodide (PI) (50 mg/ml) (Sigma, St. Louis, MO, USA). Cell cycle analysis was performed using a FACScan ¯ow cytometer (Becton-Dickinson, San Jose, CA, USA). Each sample was collected as 10 000 ungated events and the corresponding cell cycle distribution (including apoptosis) was determined using Mod®t software (Becton-Dickinson). Results are representative of three independent experiments.

Apoptosis studies Nuclear staining The nuclei of 48 h treated cells were labeled for 2 min at room temperature with Hoechst 33342 (Polysciences, Warrington, PA, USA). Cells were then observed under ¯uorescence microscopy using an ultraviolet ®lter. Selected ®elds are representative of stained nuclei. Annexin-V staining Cells were cultured in 6-well plates (Nunc) and collected at 48 h post-treatment. Apoptosis was assessed by Annexin-V staining. Cells were labeled with ¯uorescein isothiocyanate (FITC)-conjugated Annexin-V according to the manufacturer's recommendations (Boehringer, Mannheim, Germany) in the presence of 1 mM CaCl2. Propidium iodide (50 mg/ml) was added to counterstain the nuclei of cells. Cells were then observed under ¯uorescence microscopy and were further analysed by ¯ow cytometry (10 000 ungated cells). Apoptosis was estimated by the relative amount of FITC-positive-PI-negative cells. Necrotic or post-apoptotic cells are FITC-positive-PI-positive.

Immunoblot analysis Protein lysates (0.25 M Tris-HCl (pH 7.4), 20% bmercaptoethanol, and 5% SDS) were prepared at

Figure 1 HuT-102, MT2 and C91PL cells were resistant to ATRA while C8166 are sensitive: (A) Eects of ATRA (361076 and 1075 M) treatment for 48 h on the 3H-thymidine incorporation of the dierent ATL cell lines HuT-102, MT2, C8166 and C91PL. Incorporation rates are expressed as percentage (mean+s.e.m.) of control, represent, and re¯ect both the number of cells and the individual DNA synthesis of three independent experiments. (B) Eects of ATRA on the growth of the HTLV-Ipositive cells. The dierent cell lines were treated with 361076 M ATRA for up to one week. Viable cell count, calculated from triplicate wells by trypan blue dye exclusion, is expressed as percentage (mean+s.d.) of controls and is representative of three independent experiments. The Hematology Journal


130

(Santa Cruz Biotechnology Inc. Santa Cruz, CA, USA), Tax (NIH `AIDS Research and Reagent Program'), and PARP (Boehringer Mannheim, Indianapolis, IN, USA). For Western blot analysis of the RARa, RARg, and RXRa (Santa Cruz Biotechnology Inc.), total cell lysates were extracted with RIPA buer (1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS in PBS) and loaded (30 mg/lane) onto a 10% SDSpolyacrylamide gel. Immunoblotting and blocking of membranes were performed as previously described.26 Binding of antibodies to speci®c proteins was visualized by enhanced chemiluminescence using the ECL system (Amersham, Buckinghamshire, UK) with horseradish peroxidase-conjugated secondary antibody (Bio-Rad) at a 1 : 3000 dilution. Subsequently, blots were stripped and reprobed with glyceraldehyde 3phosphate dehydrogenase (GAPDH) (Biogenesis, Stinford Fload, England) at a 1 : 2000 dilution.

Retinoids mediate their biological eects by activating two distinct types of nuclear receptors: RARs and RXRs. Screening for the retinoid receptors in ATL-cell lines showed that RXRa and RARg proteins were abundantly expressed in the four dierent cell lines while RARa was only detected in HuT-102 and C8166 cells (Figure 2).

All trans retinoic acid and arsenic trioxide synergize to inhibit the growth of RARa positive HTLV-I-transformed cells The eect of ATRA (361076 M), As (1 mM), and their combination treatments for 48 h was tested on the four

Results Most of the tested HTLV-I-transformed cells are resistant to pharmacological doses of all trans retinoic acid Cell growth was evaluated by 3H-thymidine uptake at 361076 and 1075 M ATRA (Figure 1A) and cell count at 361076 M ATRA (Figure 1B). Three out of four of the tested cell lines (HuT-102, MT2, and C91PL) were resistant to ATRA treatment at both concentrations. Moreover, HuT-102 cells consistently showed an increase in 3H-thymidine uptake up to 150%. Only C8166 cells were sensitive to ATRA treatment showing a 50 ± 60% decrease in 3H-thymidine uptake at 48 h for both concentrations (Figure 1A). Furthermore, the growth of the ATRA-treated C8166 cells was signi®cantly suppressed by the third day and was completely abolished by the ®fth day of ATRA treatment (Figure 1B).

Figure 2 HuT-102 and C8166 but not MT2 and C91PL cells express RARa proteins. Thirty mg of RIPA total lysates, of all tested four cell lines, were immunoblotted against RARa, RARg, and RXRa. Equal protein loading was veri®ed by stripping and reprobing for GAPDH. The Hematology Journal

Figure 3 Eects of ATRA, As, and ATRA/As on the growth of the dierent ATL cell lines HuT-102, MT2, C8166 and C91PL. ATRA and As were used for 48 h at concentrations of 361076 M and 1 mM, respectively. (A) 3H-thymidine incorporation rates are expressed as percentage of controls and re¯ects both the number of cells and the individual DNA synthesis. The mean (+s.e.m.) is calculated from three independent experiments. (B) the eect of ATRA, As, and ATRA/As on the growth of ATL cells as by trypan blue dye exclusion. The mean (+s.d.) is expressed as percentage of controls calculated from triplicate wells and is representative of three independent experiments.


131

HTLV-I-positive cell lines through incorporation of 3 H-thymidine (Figure 3A) and cell growth (Figure 3B). A variable sensitivity to As eect was noted for the dierent cell lines, with C91PL being the most sensitive and MT2 the most resistant. On the other hand, HuT102 and C8166 cells showed intermediate sensitivity to As. The ATRA/As combination was highly synergistic in HuT-102 and C8166 cells with almost a complete growth arrest by 48 h. MT2 cells were completely resistant to this combination while the growth of C91PL cells was completely inhibited by As alone was not further aected by the addition of ATRA. Similar

eects were observed at 72 h for the four dierent cell lines (data not shown). Interestingly, this observed synergy between ATRA and As treatment on HuT-102 cell proliferation was striking at 361076 M ATRA though it was highly signi®cant at 10- or 100-fold lower concentrations (Figure 4B). The synergy between ATRA and As in HuT-102 cells was not apparent at concentrations below 1 mM As (Figure 4A).

All trans retinoic acid and arsenic trioxide synergize to induce cell cycle arrest in HuT-102 cells Flow cytometry analysis of total DNA content on sensitive HuT-102 and resistant MT2 cells to the combined ATRA/As treatment (cells were treated with 361076 M ATRA and 1 mM As for 48 h) shows that both cell lines have the same baseline percentage of apoptotic cells, approximately 3% (Table 1). At 48 h of treatment, As alone caused an increase in the pre-G1 phase in HuT-102 cells (20%) but not in MT2 cells (4%). Retinoic acid treatment alone did not aect the cell cycle distribution of either line. However, combined ATRA/As treatment showed a clear synergy in the percentage of apoptotic HuT-102 cells (35%) while no eect was observed in MT2 cells (6%). Furthermore, the percentage of cycling cells (S+G2+M) was reduced by 54% in HuT-102 cells, with no detectable eect in MT2 cells.

Combined all trans retinoic acid and arsenic trioxide induces apoptosis in HuT-102 cells Histogram-related nuclear DNA content showed a signi®cant increase in the pre-G1 phase under ATRA/ As treatment (Table 1) suggesting that the eect of this combination is through induction of apoptosis. To further con®rm the role of apoptosis in the ATRA/As dramatic growth arrest in HuT-102 cells, annexin-V immunostaining was used as an early marker of apoptosis (Figure 5A) and Hoechst 33342 nuclear staining as an indicator of chromatin condensation (Figure 5B). All cells were treated with 361076 M ATRA and 1 mM As for 48 h. HuT-102 cells, treated with the ATRA/As combination, mostly displayed a pattern of condensed and fragmented nuclei typical of apoptotic cells (Figure 5B). These intensely ¯uorescent Table 1 Eects of RA, As and their combination on the cell cycle distribution of ATL cell lines HuT-102 and MT2

Figure 4 Eects of dierent concentrations of ATRA, As and ATRA/As on the 3H-thymidine incorporation at 48 h of HuT-102 cells. (A) The dierent tested concentrations of As were 0.2, 0.5 and 1 mM while ATRA was only used at 361076 M. The mean (+s.d.) is calculated from quadruplicate measurements and is representative of two independent experiments. (B) The dierent tested concentrations of ATRA ranged from 361078 M to 361076 M while As was only used at 1 mM. The mean (+s.e.m.) is expressed as percentage of controls and is calculated from three independent experiments.

Treatment 48 h Control As ATRA ATRA+As

Pre G1 (apoptosis) HuT-102 MT2 2.9 19.9 4.0 35.4

2.4 4.0 3.0 6.0

S+G2+M (cycling cells) HuT-102 MT2 45.4 32.0 37.6 24.7

46.2 40.6 47.6 41.4

Histogram-related nuclear DNA contents was measured by ¯ow cytometry. The pre-G1 percentages represent apoptotic cells. Cycling cells, the sum of (S+G2+M) phases, are a percentage of nonapoptotic cells The Hematology Journal


132

nuclei were also obvious under As treatment, although to a much lesser extent (data not shown). In addition, ATRA/As combination treatment resulted in less numerous cells and enhanced membrane blebbing of treated vs control cells (Figure 5A,B). The necrotic or post-apoptotic cells were indicated by the doublepositive FITC and PI staining, whereas the apoptotic cells were only FITC-positive. The apoptotic cells were numerous, and post-apoptotic cells less so, under combined ATRA/As treatment (Figure 5A), and to a lesser extent under As treatment alone (data not shown). The ATRA/As-induced apoptosis was associated with caspase activation, as shown by PARP cleavage (113 KD) into death-associated fragments (89 and 24 KD) (Figure 5C). Minor eects were observed with As, but not with ATRA alone.

Combined all trans retinoic acid and arsenic trioxide downregulates Tax and upregulates IkB-a To gain further insights into the possible mechanisms by which the combined ATRA/As treatment can cause cell cycle arrest and apoptosis in HuT-102 cells, we assessed the protein levels of the viral transactivator, Tax, of the NF-kB inhibitor, IkB-a, and of the antiapoptotic protein, bcl-2. Western blot analysis showed a sharp down-regulation of Tax protein and an up-regulation of IkB-a protein level by As, which was enhanced by the addition of ATRA (Figure 6). No signi®cant change in the protein level of bcl-2 was observed under the conditions investigated here (Figure 6).

Discussion

Figure 5 The combination ATRA/As induces apoptosis in HuT102 cells. ATRA, As and ATRA/As treatments were used for 48 h at concentrations of 361076 M and 1 mM, respectively. (A) FITC-conjugated Annexin-V membrane immunostaining was evident in HuT-102 cells treated with the ATRA/As combination (lower right) but not in the control cells (upper right). The PI nuclear counterstaining indicates the dead cells. The apoptotic cells are FITC+PI7 while the necrotic or post-apoptotic cells are FITC+PI+. The left panels indicate the corresponding light microscopy analysis. (B) Hoescht 33342 staining of control (upper right) and ATRA/As-treated HuT-102 cells (lower right). The latter display a pattern of condensed and fragmented nuclei characteristic of apoptosis (arrowheads) and were less numerous than control cells. The left panels indicate the corresponding light microscopy analysis. (C) Apoptosis triggered by ATRA/As treatment of HuT-102 cells may be caspase-dependent. ATRA (361076 M) and As (1 mM) and ATRA/As treatments were used The Hematology Journal

Adult T-cell leukemia remains of poor prognosis with a median survival of six months for patients with the acute form. Recent progress in its treatment has been achieved with the combination of AZT/IFN3 ± 5 and by the use of anti-CD25 monoclonal antibodies.27 In addition, the combination of As/IFN is highly synergistic in ATL cells, resulting in cell cycle arrest and induction of massive apoptosis.24 The eect of this combination is associated with Tax degradation and reversal of NF-kB activation.28 However, the search for new eective drugs in ATL is absolutely warranted to reduce or eliminate the emergence of resistant clones which is the principal cause of the poor outcome of ATL patients. Here, a potentiating eect between ATRA and As in the inhibition of cell proliferation, G0/G1 arrest and induction of apoptosis is reported. This synergistic eect was only seen in the RARa-positive HTLV-Ipositive cells; and HuT-102 cells and, to a lesser extent, the C8166 cells. This synergy was not observed in MT2

for 48 h. Total SDS protein lysates of HuT-102 cells were immunoblotted against PARP.


133

Figure 6 Eects of ATRA, As and ATRA/As on the expression of Tax, IkB-a and bcl-2 in HuT-102 cells treated for 48 h with ATRA and/or As at concentrations of 361076 M and 1 mM, respectively. Equal protein loading was assessed by hybridization with anti-GAPDH antibodies. For IkB-a expression, the upper band represents the 36 kD major IkB-a protein. The lower bands of 34 and 32 kD presumably represent N-terminal proteolytic forms of IkB-a29 up-regulated by ATRA/As treatment.

and C91PL cell lines lacking the RARa protein. The status of RAR and RXR protein expression in fresh ATL cells has yet to be determined and our results may provide a rationale for combined ATRA/As therapies in some refractory ATL patients, after con®rmation of RARa status. Furthermore, these combined treatments may lower the threshold of therapeutic agents needed during the treatment course. To our surprise even 10 ± 100-fold less ATRA concentrations, as low as 361078 M, were able to induce a potentiating synergy with As. On the other hand, a minimum threshold of 1 mM As was required to observe this synergy with ATRA. Retinoic acid alone or in combination with other drugs has been used in the treatment of several malignant disorders especially APL.20 In this acute leukemia, ATRA induces a high complete remission rate even in patients who are resistant to conventional chemotherapy. However, the use of ATRA alone is consistently associated with leukemia relapse due to the emergence of ATRA-resistant clones.20 Recently, the combination of ATRA/As showed a highly synergistic eect in vitro and in animal models of APL.23 This combination is currently being tested in clinical trials in relapsed or refractory APL patients. Con¯icting results are available on the eect of ATRA on ATL cells. Some authors suggest that HTLV-I-transformed cells are sensitive to ATRA which is associated with down-regulation of CD25,17 while others report less obvious or no eects.18,25 However, none of these studies test the RAR status

in ATL-derived cells, or further correlate ATRA responsiveness to the availability of the receptor. In this report, neither of the two RARa-negative HTLV-I cell lines (MT2 and C91PL) were sensitive to ATRA alone, even at pharmacological doses. On the other hand, only one of the two RARa-positive lines (C8166) was sensitive to ATRA, while the other (HuT-102) was resistant. When the combination of ATRA/As was tested, a strong synergy was observed in RARapositive cells but not in the cells lacking this receptor. Our data suggest that the loss of RARa proteins may explain the lack of eect of the ATRA/As combination in some ATL cells. Preliminary results indicate that MT2 cells expressing RARa through retroviralmediated infection are more sensitive than their parental RARa-negative cells to ATRA/As combination treatment. However, the sensitization to ATRA/ As was less obvious in these RARa-infected cells than in HuT-102 and C8166 cells that naturally express RARa. In ATRA-sensitive ATL cells, treatment with ATRA/As causes G0/G1 arrest and induction of apoptosis. The mechanism of action of this combination remains unclear. We did not observe any signi®cant modi®cation of bcl-2 protein level but the induction of apoptosis by ATRA/As treatment was associated with PARP cleavage suggesting a classical caspase-dependent apoptotic pathway. Interestingly, the As/IFN treatment of ATL cells is caspaseindependent28 suggesting a dierent mechanism of action of the two tested As combination treatments. Moreover, our data con®rm a down-regulation of Tax protein and an up-regulation of IkB-a protein by As alone,28 a pattern which is enhanced by the addition of ATRA, similarly to the addition of IFN. Finally, the eects of ATRA/As combination on the retinoid receptor protein levels and its modulation of retinoid receptor-DNA-binding properties are currently under investigation. Our data suggest that screening for RARa status in fresh ATL cells should be conducted routinely and for those ATL patients who have sucient RARa expression in their leukemic cells, ATRA should be included among the drugs, such as arsenic trioxide, that are currently used. Acknowledgements The authors thank Dr Ghassan Dbaibo for his critical review of the manuscript. The work was done at the American University of Beirut Core Laboratory Facility. This work was supported by grants from the Lebanese National Council for Scienti®c Research, the American University of Beirut University Research Board, Medical Practice Plan, Diana Tamari Sabbagh and Terry Fox Cancer Research Funds and the Eli-Lilly International Foundation.

The Hematology Journal


134

References 1 Hinuma Y, Komoda H, Chosa T, Kondo T, Kohakura M, Takenada T, Kikuchi N, Ichimaru N, Yunoki K, Sato I, Matsuo R, Takiuchu Y, Uchino H, Hanaoka M. Antibodies to adult T-cell leukemia-virus-associated antigen (ATLA) in sera from patients with ATL and controls in Japan: A nation-wide sero-epidemiologic study. International Journal of Cancer 29: 631, 1982. 2 Franchini G. Molecular mechanisms of human T-cell leukemia/lymphotropic virus type I infection. Blood 86: 3619, 1995. 3 Hermine O, Bouscary D, Gessain A, Turlure P, LeBlond V, Franck N, Buzyn-Veil A, Rio B, Macintyre E, Dreyfus F, Bazarbachi A. Treatment of HTLV-I associated adult T-cell leukemia-lymphoma with a combination of zidovudine and alpha-interferon. New England Journal of Medicine 332: 1749, 1995. 4 Gill P, Harrington W, Kaplan M, Ribeiro R, Bennett J, Liebman H, Bernstein-Singer M, Espina B, Cabral L, Allen S, Kornblau S, Pike M, Levine A. Treatment of adult T-cell leukemia-lymphoma with a combination of interferon alpha and zidovudine. New England Journal of Medicine 332: 1744, 1995. 5 Bazarbachi A, Hermine O. Treatment by a combination of zidovudine and alpha interferon in naive and pretreated Adult T-cell leukemia/lymphoma patients. Journal of Acquired Immune De®ciency Syndrome and Human Retrovirology 13: s186, 1996. 6 De Luca LM. Retinoids and their receptors in differentiation, embryogenesis and neoplasia. Federation of the American Society of Experimental Biology Journal 5: 2924, 1991. 7 Pfahl M, Chytil F. Regulation of metabolism by retinoic acid and its nuclear receptors. Annual Reviews of Nutrition 16: 257, 1996. 8 Rosenthal D, Lancillotti F, Darwiche N, Sinha R, De Luca LM. Regulation of epithelial dierentiation by retinoids, p425. In: Blomho R (ed.) Vitamin A in Health and Disease. Marcel Decker, New York, 1994. 9 Zauli G, Visani G, Vitale M. All trans-retinoic acid shows multiple eects on the survival, proliferation and dierentiation of human fetal CD4+ haemopoietic progenitor cells. British Journal of Haematology 90: 274, 1995. 10 Breitman TR, Selonick SE, Collins SJ. Induction of dierentiation of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proceedings of the National Academy of Sciences USA 77: 2936, 1980. 11 Collins SJ, Robertson KA, Mueller L. Retinoic acidinduced dierentiation of HL60 myeloid leukemia is mediated directly through the retinoic acid receptor alpha. Molecular and Cellular Biology 10: 2154, 1990. 12 Chambon P. A decade of molecular biology of retinoic acid receptors. Federation of the American Society of Experimental Biology Journal 10: 940, 1996. 13 Mangelsdorf DJ, Evans RM. The RXR heterodimers and orphan receptors. Cell 83: 841, 1995. 14 Kastner P, Chambon P, Leid M. Role of nuclear retinoic acid receptors in the regulation of gene expression, p189 In: Blomho R (ed.) Vitamin A in Health and Disease. Marcel Decker, New York, 1994. 15 Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE, Kakizuka A, Evans RM. Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes and Development 6: 329, 1992. The Hematology Journal

16 Rowe A. Retinoid X receptors. International Journal of Biochemistry and Cellular Biology 29: 275: 1997. 17 Miyatake JI, Maeda Y. Inhibition of proliferation and CD25 down-regulation by retinoic acid in human adult T cell leukemia. Leukemia 11: 401, 1997. 18 Maeda Y, Miyatake J, Sono H, Matsuda M, Tatsumi Y, Horiuchi F, Irimajiri K, Horiuchi A. 13-cis retinoic acid inhibits growth of adult T-cell leukemia cells and causes apoptosis. Possible new indication for retinoid therapy. Internal Medicine 35: 180, 1996. 19 Fujimara S, Suzumiya J, Anzai K, Ohkubo K, Hata T, Yamada Y, Kamihira S, Kihuchi M, Ono J. Retinoic acids induce growth inhibition and apoptosis in adult Tcell leukemia (ATL) cell lines. Leukemia Research 22: 611, 1998. 20 Warrell R, de TheÂ H, Wang Z, Degos L. Acute promyelocytic leukemia: recent advances in biology and treatment. New England Journal of Medicine 329: 177, 1993. 21 Chen CQ, Zhu J, Shi XG, Ni JH, Zhong HJ, Si GY, Jin XL, Tang W, Li XS, Xong SM, Shen ZX, Sun GL, Ma J, Zhang P, Zhang TD, Gaxin C, Naoe T, Chen SJ, Wang ZY, Chen Z. In vitro studies on cellular and molecular mechanisms of arsenic trioxide (As2O3) in the treatment of acute promyelocytic leukemia: As 2O3 induces NB4 cell apoptosis with downregulation of bcl-2 expression and modulation of PML-RAR alpha/PML proteins. Blood 88: 1052, 1996. 22 Zhu J, Koken M, Guignon F, Chelbi-Alix MK, Degos L, Yi Wang Z, Chen Z, de TheÂ H. Arsenic-induced PML targeting onto nuclear bodies implications for the treatment of acute promyelocytic leukemia. Proceedings of the National Academy of Sciences of the USA 94: 3978, 1997. 23 Lallemand-Breitenbach V, Guillemin MC, Janin A, Daniel MT, Degos L, Kogan SC, Bishop JM, de TheÂ H. Retinoic acid and arsenic synergize to eradicate leukemic cells in a mouse model of acute promyelocytic leukemia. Journal of Experimental Medicine 189: 1043, 1999. 24 Bazarbachi A, El-Sabban ME, Nasr R, Quignon F, Awaragi C, Kersual J, Dianoux L, Zermati Y, Haidar J, Hermine O, De TheÂ H. Arsenic trioxide and interferon-a synergize to induce cell cycle arrest and apoptosis in human T-cell lymphotropic virus type I-transformed cell. Blood 93: 278, 1999. 25 Ishitsuka K, Hanada S, Suzuki S, Utsunomiya A, Chyuman Y, Takeuchi S, Takeshita T, Shimotakahara S, Uozumi K, Makino T, Arima T. Arsenic trioxide inhibits growth of human T-cell leukaemia virus type I infected T-cell lines more eectively than retinoic acids. British Journal of Haematology 103: 721, 1998. 26 Darwiche N, Scita G, Jones C, Rutberg S, Greenwald E, Tennenbaum T, Collins SJ, De Luca LM, Yuspa SH. Loss of retinoic acid receptors in mouse skin and skin tumors is associated with activation of the ras(Ha) oncogene and high risk for premalignant progression. Cancer Research 56: 4942, 1996. 27 Waldmann TA, White JD, Goldman CK, Top L, Grant A, Bamford R, Roessler E, Horak I, Zaknoen S, Kasten C, England R, Horak E, Mishra B, Dipre M, Hale P, Fleisher T, Junghans R, Jae E, Nelson D. The interleukin-2 receptor: a target for monoclonal antibody treatment of human T-cell lymphotrophic virus I-induced adult T-cell leukemia. Blood 82: 1701, 1993.


135

28 El-Sabban ME, Nasr R, Dbaibo G, Hermine O, Abboushi N, Quignon F, Ameisen JC, Bex F, de TheÂ H, Bazarbachi A. Arsenic/Interferon-a triggered apoptosis in HTLV-I transformed cells is associated with Tax downregulation and reversal of NF-kB activation. Blood 96: 2849, 2000.

29 Chou-Chi H, Ren-Ming D, Longo DL. Inactivation of NF-kB inhibitor IkB-a: ubiquitin-dependent proteolysis and its degradation product. Biochemical and Biophysical Research Communications 215: 292, 1995.

The Hematology Journal