8-Amino-Adenosine Is a Potential Therapeutic Agent for Multiple Myeloma

Molecular Cancer Therapeutics 1411

8-Amino-adenosine is a potential therapeutic agent for multiple myeloma

Nancy L. Krett,1 Katharine M. Davies,1 membrane potential, an increase in caspase activity, Mary Ayres,2 Chunguang Ma,1 Chadi Nabhan,1 cleavage of caspase substrates, and an increase in cells 2 1 Varsha Gandhi, and Steven T. Rosen with a sub-G1 DNA content. Conclusion: Based on these results, we conclude that 8-NH -Ado may hold great 1 2 Division of Hematology Oncology, Department of Medicine, potential as a therapeutic agent for the treatment of MM. Robert H. Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University Medical School, Chicago, [Mol Cancer Ther 2004;3(11):1411–9] Illinois and 2Department of Experimental Therapeutics, University of Texas M.D. Anderson Cancer Center, Houston, Texas Introduction There are 14,600 cases of multiple myeloma (MM) Abstract diagnosed in 2003, with a median survival of only 4 to Multiple myeloma (MM) is a malignancy of clonal B-cells 5 years (1), which is only modestly improved with high- that accounts for 10% of all hematologic malignancies. dose chemotherapy and autologous stem cell transplanta- We have shown previously that a novel purine analogue, tion (2). The survival rate for MM has not changed 8-chloro-adenosine, has significant activity for MM in dramatically in the past 30 years. Therefore, new thera- preclinical studies. Objective: Using MM cell lines, we in- peutics need to be developed. vestigated the molecular mechanism of related congener Several nucleoside analogues are currently used for the treatment of acute and chronic hematologic malignancies of adenosine, 8-amino-adenosine (8-NH2-Ado). Methods: We employed biological and biochemical assays in MM that include tumors of myeloid and lymphoid lineage. Based on the extensive experience with hematologic cell lines to evaluate the clinical potential of 8-NH2-Ado. Results: In MM cell lines both sensitive and resistant to tumors, one would expect that nucleoside analogues that have been effective in leukemias and lymphomas should conventional chemotherapies, 8-NH2-Ado is cytotoxic, play a role in the treatment of MM (3, 4). The two with IC50 ranging from 300 nmol/L to 3 Mmol/L. A mouse leukemic cell line lacking adenosine kinase activity was nucleoside analogues that play a major role in the treatment of B-cell lymphocytic leukemia, hairy cell leukemia, and resistant to 8-NH2-Ado, indicating that phosphorylation indolent lymphomas are fludarabine and cladribine (5, 6). of 8-NH2-Ado to its triphosphate form is required for cytotoxicity. A 4-hour incubation of MM cells with Unfortunately, neither of these analogues has exhibited 10 Mmol/L analogue resulted in an accumulation of sufficient activity in vitro in myeloma cell lines or in clinical trials to justify continued clinical evaluation in this disease >7 mmol/L 8-NH2-ATP with a parallel decline in the (7). Results of phase II clinical trials in MM with endogenous ATP levels. Accumulation of 8-NH2-ATP was fludarabine (8, 9) and cladribine (3, 4) were completely dependent on both exogenous concentration of 8-NH2- negative. This may be due to poor phosphorylation of these Ado and incubation time. The accumulation of 8-NH2-ATP was accompanied by a decrease in both RNA and DNA agents by deoxycytidine kinase or the DNA-directed nature of these arabino-adenosine and deoxy-adenosine ana- synthesis. The mechanism of 8-NH2-Ado-mediated cytotoxicity was due to apoptosis as measured by an increase logues. However, promising preclinical studies with in Annexin V binding, a decrease in mitochondrial 8-chloro-adenosine (8-Cl-Ado) indicate that in contrast to the arabino-adenosine (fludarabine) and deoxy-adenosine (cladribine) analogues this adenosine congener may have efficacy in MM. This may be due to an efficient phosphorylation to the triphosphate by adenosine kinase, a decrease Received 9/15/03; revised 7/13/04; accepted 8/27/04. in endogenous ATP pools, and a specific inhibition of RNA Grant support: National Cancer Institute grant CA85915, Department of synthesis (10). These encouraging data provide an impetus Health and Human Services, and National Cancer Institute for Clinical Oncology Training Program 5T32 CA 79447 (C. Nabhan). to test related adenosine analogues. The costs of publication of this article were defrayed in part by the The 8-amino derivative of cyclic AMP has been shown payment of page charges. This article must therefore be hereby marked previously to be cytotoxic in a rat hepatoma cell line (11) advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. and breast cancer cell lines of human (12, 13) and mouse Note: C. Nabhan is currently at Oncology Specialists, S.C., Lutheran (14) origin. In these studies, the inclusion of a phosphodi- General Hospital Cancer Care Center, Park Ridge, IL 60056. esterase inhibitor to stabilize the cyclic AMP and to Requests for reprints: Nancy L. Krett, Division of Hematology Oncology, enhance its actions resulted in decreased cytotoxicity. Department of Medicine, Robert H. Lurie Comprehensive Cancer Center, Therefore, it was hypothesized that the cytotoxic com- Feinberg School of Medicine, Northwestern University Medical School, Chicago, IL 60611. Phone: 312-908-5919; Fax: 312-908-1372. pound was most likely a metabolite of 8-amino cyclic AMP E-mail: [email protected] [i.e., 8-amino-adenosine (8-NH2-Ado)]. Saito et al. (14) C Copyright 2004 American Association for Cancer Research. further investigated the actions of 8-NH2-Ado in an in vivo

Mol Cancer Ther 2004;3(11). November 2004

breast cancer mouse model and noted a significantly longer the MTS Cell Titer Aqueous assay (Promega, Madison, WI), survival time for the treated animals indicating efficacy. which measured the conversion of a tetrazolium compound Based on these observations with 8-amino cyclic AMP into formazan by a mitochondrial dehydrogenase enzyme and on our promising preclinical studies of 8-Cl-Ado in in live cells. The quantity of formazan product as measured MM, we investigated the cytotoxicity of 8-NH2-Ado in MM by the amount of 490 nm absorbance is directly propor- cell lines and further investigated the mechanism of action tional to the number of living cells in culture. The data were of 8-NH2-Ado in MM cell lines. expressed as the percentage of formazan produced by the cells treated with the control medium in the same assay. Measurement of Intracellular Nucleoside Triphos- Materials and Methods phates Cell Lines Nucleotides were extracted using perchloric acid and Experiments were conducted using MM cell lines that the extracts were neutralized with KOH as described pre- were developed previously by our group (15, 16). The viously (20) and stored at 20jC until analyzed. The original cell line (MM.1) was established from the pe- neutralized extracts were applied to an anion exchange ripheral blood of a MM patient that had been treated with Partisil-10 SAX column and eluted at a flow rate of 1.5 steroid-based therapy. By chronic exposure to glucocorti- mL/min with a 50-minute concave gradient (curve 7, Waters coid, a steroid-resistant variant was developed. MM.1S 600E System Controller, Waters Corp., Milford, MA) from cells are sensitive to the killing effects of glucocorticoids, 60% NH4H2PO4 (0.005 mol/L, pH 2.8) and 40% NH4H2PO4 whereas MM.1R cells are resistant to glucocorticoids. (0.75 mol/L, pH 3.6) to 100% NH4H2PO4 (0.75 mol/L, pH RPMI-8226 cells and the multidrug-resistant derivative 3.6). The column eluate was monitored by UV absorption at MDR10V MM cells were obtained from Dr. William S. 256 nm and the nucleoside triphosphate was quantitated Dalton (17). All cells were grown in RPMI 1640 (Invitrogen, by electronic integration with reference to external stand- Baltimore, MD) supplemented with 10% fetal bovine ards (21). The analogue triphosphate was identified by serum, glutamine (2 mmol/L), penicillin (100 units/mL), comparing the retention profile and absorption spectrum streptomycin (100 Ag/mL), and fungizone (2.5 Ag/mL). The with that of authentic standard. The intracellular concentra- L1210 and ED2 cell lines were obtained from Dr. Joseph tion of nucleotides contained in the extract was calculated G. Cory (18). The ED2 variant (lacking Ado kinase) of from a given number of cells of a determined mean volume. the mouse L1210 cell line was supplemented with 2,3- The cell number was determined using Coulter counter dihydro-1H-pyrazole[2,3a]imidazole (300 mol/L), Desferal (Coulter Electronics, Hialeah, FL). This equipment was (40 mol/L), erythro-9-(2-hydroxyl-3-nonyl)adenosine attached to a channelizer that was used to estimate the (5 mol/L), and deoxy-adenosine (400 mol/L). All cells mean volume of cells in a given cell population. This volume j were maintained in a 37 C incubator with 5% CO2. was used to quantitate the concentration of nucleotides. Drugs and Chemicals This calculation assumed that the nucleotides were uni- 8-NH2-Ado was purchased from RI Chemical, Inc. formly distributed in a total cell volume. The lower limit (Orange, CA). For high-performance liquid chromatogra- of sensitivity of this assay was 10 pmol in an extract of phy standards, triphosphates of these analogues were 5 106 cells corresponding to a cellular concentration of custom synthesized by Biolog (San Diego, CA). N-benzy- 1 Amol/L. loxycarbonly-Val-Ala-Asp-fluoromethyl ketone (ZVAD- Inhibition of Nucleic Acid Synthesis by Nucleoside fmk), Z-Leu-Glu(OMe)-His-Asp(OMe)-fluoromethyl ketone Analogues (LEHD-fmk), and Z-Ile-Glu(OMe)-Thr-Asp(OMe)-fluoro- Exponentially growing cells were incubated with the methyl ketone (IETD-fmk) were purchased from Enzyme indicated concentrations of nucleoside analogues for 4 Systems Products (Livermore, CA). Staurosporine was hours. After 3 hours, [3H]thymidine or [3H]uridine (2 ACi/ obtained from Alexis Chemical (San Diego, CA). CMXRos mL) was added to these cultures and incubation was (MitoTracker Red) was from Molecular Probes (Eugene, continued for an additional 1 hour in a multiscreen assay OR), whereas caspase-8 and caspase-9 fluorometric kits system (Millipore Corp., Bedford, MA). The cells were then were purchased from R&D Systems (Minneapolis, MN). 2,3- collected on multiscreen GV filters under vacuum and Dihydro-1H-pyrazole[2,3a]imidazole was obtained from washed four times each with ice-cold 8% trichloroacetic Dr. Robert J. Schultz (Chief, Drug Synthesis and Chemistry acid, water, and 100% ethanol. The radioactivity in the acid- Branch of the Developmental Therapeutics Program, insoluble material retained on the filters was measured by National Cancer Institute, NIH). All other chemicals were scintillation counting and expressed as the percentage of obtained from Sigma Chemical (St. Louis, MO). Annexin control (untreated) value of cells. V-FITC kit was obtained from Immunotech, a Beckman FlowCytometry Coulter company (Marseille, France). To determine the distribution of cells within the cell Cell Proliferation Assay cycle, aliquots of cells (1 106 each) were pelleted (500 g This assay was done as described previously (19). Briefly, for 5 minutes at 4jC) and washed twice in ice-cold PBS MM cells were cultured into 96-well dishes at a concentra- (8.1 g NaCl, 1.14 g Na2HPO4, 0.22 g KCl, 0.25 g/L KH2PO4), tion of 25,000 cells/well and incubated with the indicated fixed in ice-cold 70% ethanol, and stored at 4jC until drugs for 72 hours. Cell proliferation was determined using analyzed. Before analysis by flow cytometry, the fixed cells

Mol Cancer Ther 2004;3(11). November 2004

were pelleted, washed in PBS, and resuspended in ice-cold flow buffer (PBS containing 0.5% Tween 20, 15 g/mL propidium iodide, 5 g/mL DNase-free RNase). The stained cells were analyzed using an EPICS Profile II flow cytometer (Coulter Electronics). Immunoblotting Analysis MM1.S cells were cultured and grown at a concentration of 10 106 in tissue culture medium (30 mL) in 75 cm2 flasks. Cells were treated with 8-NH2-Ado (10 Amol/L) for the indicated times and harvested. Staurosporine (100 nmol/L) was used as a positive control for apoptosis. Cell pellets were homogenized in lysis buffer [10 mmol/L KPO4 (pH 7.0), 1 mmol/L EDTA, 5 mmol/L EGTA, 10 mmol/L Figure 1. Chemical structure of 8-NH2-Ado. h MgCl2,50mmol/L -glycerol phosphate, 1 mmol/L sodium orthovanadate, 2 mmol/L DTT, 1 mmol/L phenyl- methylsulfonyl fluoride, 0.5% NP40, 0.1% deoxycholate]. (10 Amol/L) or 8-Cl-Ado (10 Amol/L) for the indicated Homogenates were centrifuged at 4jC for 10 minutes at times. After collection, the cells were washed twice with 16,100 g and the supernatants were collected and stored cold PBS and resuspended with cell lysis buffer following at 20C until time of use. Protein concentration was the manufacturer’s protocol. Protein concentrations in all determined by Bio-Rad protein assay (Bio-Rad Laborato- cell lysates were determined using Bio-Rad protein assay ries, Hercules, CA). Protein at a concentration of 40 g was (Bio-Rad Laboratories) and total protein (100 Ag) was used mixed with sample buffer [125 mmol/L Tris (pH 6.8), 4% for each examined sample. Samples were added to the SDS, 20% glycerol, 100 mmol/L DTT, 0.05% bromophenol reaction buffer with the appropriate caspase fluorogenic blue] and fractionated on a precast 8% to 16% Tris-glycine substrate. Reactions were incubated for 2 hours at 37jC and gel (Invitrogen/Novex, Carlsbad, CA). Proteins were then fluorescence was determined using Packard FluoroCount transferred to a polyvinylidene difluoride membrane fluorescent microplate reader that allowed for light (Immobilon-P, Millipore). Following protein transfer, mem- excitation at 400 nm wavelength and collected emitted branes were blocked with 5% nonfat milk in PBS and light at a wavelength of 505 nm. subsequently incubated with the primary antibody at Mitochondrial Membrane Potential Assay 1:1,000 dilution overnight at 4C. To detect for poly(ADP- 2 6 ribose) polymerase (PARP), a mouse monoclonal antibody Cells were cultured in 25 cm flasks with 5 10 cells in A was used (PharMingen, San Diego, CA). The following day, 5 mL medium. A concentration of ZVAD-fmk (40 mol/L) blots were washed with PBS-T (PBS with 0.1% Tween 20) was added to the appropriate flasks 1 hour prior to drug A and incubated with horseradish peroxidase–linked sec- therapy. Cells were treated with 8-NH2-Ado (10 mol/L) A ondary antibody (Amersham, Arlington Heights, IL). After or 8-Cl-Ado (10 mol/L) for 24 hours. Cells were washed washing with PBS-T, the blots were developed using with cold PBS and samples were transferred into two tubes enhanced chemiluminescence (Amersham) and the signal with 1 mL PBS per tube. Samples were incubated with CMXRos fluorescent dye (10 nmol/L) for 15 minutes at was visualized with a X-ray film (Hyperfilm, Amersham). j Annexin V-FITC Assay 37 C and analyzed by flow cytometry using Coulter EPICS XL instrument with the data analyzed using the System II As cells undergo apoptosis, the integrity of the cell software package. A decrease in fluorescence indicates a membrane is disrupted and phosphatidylserine is exposed. decrease in the mitochondrial membrane potential (Dc ). Using a fluorescent conjugate of Annexin V, the presence m of phosphatidylserine is detected by Annexin V binding, which has a high selective affinity for phosphatidylserine. Results 2 MM.1S cells were grown in 25 cm flasks at a concentration Cytotoxicity of 8-NH2-Ado 6 of 2.5 10 cells in 5 mL medium. Cells were treated with The structure of 8-NH2-Ado is shown in Fig. 1. To 8-NH2-Ado (10 Amol/L), and in some conditions, the cells determine the growth inhibitory effect of 8-NH2-Ado, we received ZVAD-fmk (40 Amol/L) 1 hour prior to the examined MM cell lines that are sensitive (MM.1S) addition of drug. Cells were then harvested and washed and resistant (MM.1R and RPMI-8226; MDR10V) to twice with cold PBS and collected by centrifugation. traditional chemotherapies for this malignancy. The Samples were resuspended in diluted binding buffer MM.1R and RPMI-8226 cell lines are resistant to glucocorti- provided with the kit and cells were stained following coids, whereas the MDR10V cell line is a multidrug- manufacturer’s directions. Samples were analyzed by flow resistant MM cell line (17). Cells were incubated for cytometry using Coulter EPICS XL instrument and data 72 hours with the concentrations of 8-NH2-Ado indicated were analyzed using System II software package. in Fig. 2A. The results indicate that 8-NH2-Ado causes a Caspase-8 and Caspase-9 Fluorometric Assay decrease in cell number with an IC50 ranging from 300 Cells were cultured in 25 cm2 flasks with 5 106 cells nmol/L to 3 Amol/L. Normal human lymphocytes isolated in 5 mL medium. Cells were treated with 8-NH2-Ado from the peripheral blood mononuclear cell fraction had a

Mol Cancer Ther 2004;3(11). November 2004

an IC50 in the range of 1 to 3 Amol/L, whereas the ED2 cells are not affected even at 10 Amol/L analogue. These data indicate that adenosine kinase activity is required for 8-NH2-Ado cytotoxicity. Concentration-Dependent Metabolism of 8-NH2-Ado in MM Cells

To determine the metabolism of 8-NH2-Ado to its triphosphate, exponentially growing MM.1S cells were incubated with 8-NH2-Ado (1, 3, 10, or 30 Amol/L) for 4 hours. As shown in Fig. 3A, there was a concentration- dependent increase in 8-NH2-ATP with a maximal accumulation of 5,800 Amol/L. At 30 Amol/L exogenous drug concomitant with analogue triphosphate accumulation, there was a dose-dependent decrease in the endogenous level of ATP. At 4 hours with 8-NH2-Ado (30 Amol/L), the cellular concentration of ATP was decreased to 25% of the control value. Using a 10 Amol/L concentration of 8-NH2- Ado, the accumulation of 8-NH2-ATP was measured from 0 to 12 hours. As shown in Fig. 3B, an increase in 8-NH2-ATP concentration was noticed up to 6 hours where the level of triphosphate reached >8,000 Amol/L. After 6 hours, the f levels of 8-NH2-ATP decreased and returned to 2,500 Amol/L by 12 hours. The decrease is most likely attribut- able to cell death caused by 8-NH2-Ado. With the measured

Figure 2. Cytotoxicity of 8-NH2-Ado in human MM cell lines (A)or adenosine kinase – deficient or adenosine kinase – proficient cell lines (B). Cell lines were exposed to the indicated concentrations of 8-NH2-Ado for 72 hours and cell proliferation was determined by the MTS assay as described in Materials and Methods. Average of four independent determinations. Points, percentage of untreated control cells; bars, SD. The human myeloma cell lines tested are MM.1S (y), MM.1RL (n), RPMI- 8226 (4), and RPMI-MDR10V (o). The mouse leukemia cell lines were either adenosine kinase negative (ED2 cells; n) or adenosine kinase positive (L1210 cells; w ).

decrease in cell number, with an IC50 of 3 Amol/L in the 3- (4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2- (4-sulfophenyl)-2H-tetrazolium, inner salt, inner salt (MTS) assay (data not shown). This is a range of concentrations that has been achieved in plasma with another 8-substituted purine analogue (ref. 22; i.e., 8-Cl-Ado). Similar cytotoxicity has been measured in cells from epithelial cancers (prostate, breast, and cervix), indicating 8-NH2- Ado may also be beneficial in some solid tumors (data not shown). Cytotoxicity Requires Phosphorylation of 8-NH2-Ado To explore whether phosphorylation of 8-NH2-Ado is required for cytotoxicity in MM cell lines, we examined the effect of this compound in the ED2 mouse leukemia cell line that is deficient in adenosine kinase activity (18). The cytotoxicity of 8-NH2-Ado in ED2 cells was compared with the L1210 parental mouse leukemia cell line that retains Figure 3. Dose-dependent (A) and time course – dependent (B) accu- adenosine kinase activity and the results are shown in mulation of 8-NH2-ATP. Nucleotides were extracted using perchloric Fig. 2B. The ED2 (AK) and L1210 (AK+) cells were treated acid following either a 3-hour incubation with increasing concentrations with the indicated concentrations of 8-NH -Ado for 72 of 8-NH2-Ado (A) or increasing times of incubation with 8-NH2-Ado (10 2 Amol/L). High-performance liquid chromatography was used to quantify hours and the formazan absorbance was determined in the the triphosphates as described in Materials and Methods. The triphos- MTS assay. The L1210 cells are sensitive to 8-NH2-Ado with phates measured were 8-NH2-ATP (n) or ATP (E).

Mol Cancer Ther 2004;3(11). November 2004

ment results in a profound inhibition of both RNA and DNA synthesis (Fig. 4). This is in contrast to either deoxy- adenosine analogues currently in clinical use, which cause an inhibition of DNA synthesis, or the investigational adenosine analogue (8-Cl-Ado), which inhibits only RNA synthesis without an effect on DNA synthesis (10). Therefore, 8-NH2-Ado seems to be unique in its ability to profoundly inhibit both DNA and RNA synthesis. 8-NH2-Ado Induces Apoptosis in MM Cells The observed 8-NH2-Ado-induced cytotoxicity could be the result of apoptosis or decreased proliferation. Cells undergoing apoptosis have a reduced DNA content caused by cleavage and loss of small DNA fragments. Using cell cycle analysis, apoptotic cells are identified as those cells in the sub-G1 fraction of the cell cycle. As shown in Fig. 5, in a 24-hour treatment, 8-NH2-Ado (10 Amol/L) caused a significant decrease of cells in the G0/G1 and G2-M phases of the cell cycle and an increase to 66.1% in the number of cells in the sub-G1 fraction. 8-Cl-Ado at an equimolar concentration has a less marked effect, with only 19.4% of cells in sub-G1. Similar to the decreased cytotoxicity of 8-NH2-Ado in 8226-S myeloma cells, we measured a decreased percentage of these cells (18.1%) undergoing apoptosis (data not shown). Accumulation of the sub-G1 population of cells at the 10 Amol/L concentration of 8-NH2-Ado was blocked by the pan-specific caspase inhibitor ZVAD-fmk, indicating that caspase activity is required. These results indicate that 8-NH2-Ado stimulates a more robust response than 8-Cl-Ado causing cytolysis by Figure 4. Effect of 8-NH2-Ado on DNA and RNA synthesis in MM.1S (A) and MM.1R (B) cells (n = 3). Cells were incubated with 8-NH2-Ado the induction of apoptosis, an event that requires caspase (10 Amol/L) for the indicated times. One hour prior to harvesting the cells, activity. either [3H]thymidine or [3H]uridine was added to the medium as described in Materials and Methods. Incorporation of [3H] into either DNA or RNA. MM.1S cells were treated with 8-NH2-Ado for 24 hours Points, percentage of incorporation in untreated control cells; bars, SE. and Annexin V binding and PI staining were measured by Average of three determinations.

increase in 8-NH2-ATP, there is a decrease in the endogenous ATP, ultimately approaching 0% of control by 12 hours. In the glucocorticoid-resistant MM.1R cells treated with 8-NH2-Ado, there is a comparable level of 8-NH2-ATP accumulation, indicating a similar mechanism of cytotoxicity (data not shown). In normal lymphocytes, which are less sensitive to the toxic effects of 8-NH2-Ado, there is a much lower accumulation of 8-NH2-ATP (13% of the accumulation in MM.1S cells) and a modest decrease in the endogenous ATP pools (data not shown). This indicates that sensitivity to the drug may be linked to the accumulation of analogue triphosphate. Effect of 8-NH2-ATP on Macromolecular Synthesis Because 8-NH2-ATP is an analogue of ATP, we investigated whether it has an effect on nucleic acid synthesis. Figure 5. 8-NH2-Ado induces apoptosis in MM.1S cells. Cells were Both MM.1S and MM.1R cells were incubated with 8-NH2- incubated with the indicated treatments for 24 hours. Cells were harvested Ado (10 Amol/L) for the indicated times; however, during and fixed overnight with ethanol and DNA content was stained with the last hour of incubation, the DNA and RNA precursors, propidium iodide and analyzed for cell cycle by flow cytometry. Control [3H]thymidine and [3H]uridine, respectively, were added (light gray columns); 10 Amol/L 8-NH2-Ado (black columns); 10 Amol/L 8-NH2-Ado + 40 Amol/L ZVAD-fmk (white columns); 10 Amol/L 8-Cl-Ado to the medium. In both glucocorticoid-sensitive and (hatched columns). Representative experiment of three individual experi- glucocorticoid-resistant MM cell lines, 8-NH2-Ado treatments. Bars, SE.

Mol Cancer Ther 2004;3(11). November 2004

Another hallmark of apoptosis is the decrease in Dcm (23) as measured by decreased binding of the mitochondrial cationic fluorescent dye CMXRos. Using this approach, we have examined the Dcm in MM.1S cells following 24 hours of treatment with 8-NH2-Ado (10 Amol/L; Fig. 6B). In staurosporine-treated cells, a positive control for apoptosis, 95.4% of the cells are in the low fluorescence population, indicating a loss of Dcm. Similarly, the 8-NH2-Ado treatment causes a loss of Dcm with 95.4% of the cells expressing low fluorescence, whereas 8-Cl-Ado treatment is less effective with 37.8% of the cells expressing low fluorescence. The loss of Dcm is unchanged by the addition of ZVAD-fmk, indicating that caspase activity is not required for the loss of Dcm. Taken together, the data in Fig. 6 indicate that 8-NH2-Ado is causing apoptosis in MM.1S cells and these early apoptotic events do not require the activation of caspases. 8-NH2-Ado Promotes Activation of Caspases We have shown that 8-NH2-Ado-induced cell death is associated with caspase activation (Fig. 5). Here, we show cleavage of a caspase substrate and further characterize which initiator caspases are activated. PARP is a substrate of caspase-3 that is cleaved during apoptosis. MM.1S cells were treated with 8-NH2-Ado (10 Amol/L) for times indicated in Fig. 7A and whole cell lysates were examined by immunoblotting with an antibody against PARP. There is significant cleavage of PARP by 6 hours of drug incubation, which is complete by 12 hours. The addition Figure 6. 8-NH2-Ado induces markers of apoptosis in MM.1S cells. of ZVAD-fmk (40 Amol/L) blocks PARP cleavage, com- MM.1S cells were treated with 8-NH2-Ado (10 Amol/L) or 8-Cl-Ado (10 Amol/L) for 24 hours and stained with Annexin V to evaluate phosphati- pletely indicating that caspase activity is required for this dylserine location (A) or stained with CMXRos to evaluate Dcm (B). Stained action. These results indicate that caspases are rapidly cells were analyzed by flow cytometry as described in Materials and Methods. Representative of three replicates. Bars, SE.

flow cytometry with the results shown in Fig. 6A. Increased Annexin V binding correlates with the movement of phosphatidylserine to the outer leaflet of the plasma membrane during early events in apoptosis. As a positive control for Annexin V binding, MM.1S cells were treated with staurosporine (100 nmol/L) for 24 hours with 50.9% of the cells binding Annexin V. 8-NH2-Ado treatment results in an increase in Annexin V binding to 57.6% in contrast to the untreated control cells with 9.5% of cells staining positive for Annexin V. In comparison, only 18.4% of 8-Cl- Ado-treated cells stained positively for Annexin V, indicating a decreased responsiveness. The Annexin V binding in the 8-NH2-Ado treatment group is further partitioned into 46.4% positive for Annexin V binding alone represent- ing the percentage of cells in early apoptosis and 11.0% of the cells both bind Annexin V and stain with PI represent- ing cells in late apoptosis or necrosis. The addition of Figure 7. 8-NH -Ado activates caspase activity in MM.1S cells. MM.1S ZVAD-fmk 1 hour prior to the purine analogue does not 2 cells were treated with 8-NH2-Ado or the indicated treatment for the decrease the Annexin V binding, indicating that caspase indicated times. A, whole cell extracts were prepared and immunoblotted activity is not required for the movement of phosphatidyl- for PARP as described in Materials and Methods. B, caspase-8 (gray serine in the plasma membrane. These results emphasize columns) or caspase-9 (black columns) were measured by fluorometric assay as described in Materials and Methods. Data are normalized to that the majority of cell death by 8-NH2-Ado is due to expression in untreated (Control) cells and are reported as fold increase apoptosis. over control. Representative of three replicate experiments. Bars, SE.

Mol Cancer Ther 2004;3(11). November 2004

activated by 8-NH2-Ado. Caspase-3 is an effector caspase response than 8-Cl-Ado. The mechanism of cytotoxicity and can be activated by the initiator caspase-8 or caspase-9. requires phosphorylation in the cells by adenosine kinase, We examined which initiator caspases are activated in ultimately resulting in accumulation of 8-NH2-ATP to response to 8-NH2-Ado to gain insights into the apoptotic levels of 7 mmol/L, whereas endogenous ATP levels are pathways used by 8-NH2-Ado. dramatically decreased. The cellular consequences of the In apoptosis, there are two classic pathways that lead to accumulation of amino-modified ATP include the inhibi- activation of initiator caspases. Activation of the extrinsic tion of both RNA and DNA synthesis. Cell death occurs by death receptor pathway leads to caspase-8 activation and apoptosis as measured by increased Annexin V binding, the intrinsic mitochondrial injury pathway leads to decreased Dcm, increased activity of both caspase-8 and activation of caspase-9; however, there may be cross-talk caspase-9, and cleavage of cellular substrates including between these two pathways (24). MM.1S cells were treated PARP and DNA. with 8-NH2-Ado (10 Amol/L) for the times indicated in We have found that 8-NH2-Ado has characteristics that Fig. 7B and caspase-8 and caspase-9 activity was measured are similar to 8-Cl-Ado; however, the amino compound is in an enzymatic assay. Both caspase-8 and caspase-9 are more effective in causing cell death and also shows unique activated by as early as 6 hours and activation increases up qualities. The range of concentrations required for cytolysis to 24 hours. The activation of both caspase-8 and caspase-9 is 10-fold lower for 8-NH2-Ado in comparison with that is completely blocked by the addition of ZVAD-fmk achieved with 8-Cl-Ado (10), indicating that 8-NH2-Ado (40 Amol/L), the pan-specific caspase inhibitor. 8-Cl-Ado may be more effective in causing cell death. The glucocor- also increases activation of both caspase-8 and caspase-9 ticoid-resistant MM.1R cells require a slightly higher albeit to a lesser extent. concentration of 8-NH2-Ado to achieve cytotoxicity than We further investigated the possible mechanism for the glucocorticoid-sensitive MM.1S cells, perhaps indicat- 8-NH2-Ado activation of caspase-8 by examining whether ing some amount of cross-resistance to 8-NH2-Ado. the Fas pathway is used. Using ZB4, an antibody that Although the primary mechanism of cell death for both blocks Fas, we examined whether the ability of 8-NH2-Ado compounds is via apoptosis, the apoptosis achieved with to induce apoptosis in MM.1S cells was impaired. ZB4 the amino compound is more rapid with significant PARP treatment did not block the ability of 8-NH2-Ado to induce cleavage by 6 hours of incubation as opposed to 24 hours apoptosis in MM.1S cells (data not shown), indicating that for 8-Cl-Ado (data not shown). In addition, the amplitude the 8-NH2-Ado-induced activation of caspase-8 does not of the apoptotic response is more robust for the amino require Fas activity but rather occurs by another mecha- compound. Following 24 hours of incubation with a nism. Caspase-8 may also be activated secondary to stim- maximal concentration of the analogues, for the amino ulation of the intrinsic pathway. In this scenario, activation compound, 66.1% of the cells are in the sub-G1 fraction of of caspase-9 would ultimately lead to activation of caspase-8. the cell cycle and 95% of the cells exhibit a decreased Dcm To examine this possibility, we used specific inhibitors for as compared with 8-Cl-Ado where 19.4% of the cells are in both caspase-9 (LEHD-fmk) and caspase-8 (IETD-fmk) to sub-G1 and 37% of the cells have a decreased Dcm. determine whether subsequent activation of caspase 8-NH2-Ado requires phosphorylation of the compound activity is impaired. When caspase-9 activation is blocked to its triphosphate form for the toxicity observed. The initial by LEHD-fmk, there was a concomitant decrease in the reaction is the conversion to the monophosphate form by activation of caspase-8. This result would indicate that adenosine kinase and ultimately to the triphosphate. Those perhaps caspase-8 activation is subsequent to caspase-9 cells with higher adenosine kinase activity will most likely activation. However, when caspase-8 activation was be more susceptible to the cytotoxic actions of 8-NH2-Ado. blocked with IETD-fmk, there was no measurable activa- In addition, the high specific activity of adenosine kinase in tion of caspase-9, indicating that the activation of caspase-9 lymphocytes may limit toxicity in other tissues (26). We and caspase-8 by 8-NH2-Ado is complex and that there is plan to examine the effects of 8-NH2-Ado in normal bone likely a cross-talk between the activation pathways. marrow progenitor cells to assess potential toxicity issues. A comparison of the reported literature for triphosphate accumulation with other adenosine-based nucleoside ana- Discussion logues is shown in Table 1. There is a 40- to 400-fold higher We have reported previously promising preclinical studies accumulation of 8-NH2-ATP when compared with the in MM for the RNA-directed nucleoside analogue 8-Cl-Ado triphosphate accumulation for 8-Cl-Ado, cladribine, fludar- (10) and this drug is progressing to phase I clinical trials. abine, and clofarabine. The formation of 8-NH2-ATP seems Based on our earlier observations on the activity of 8-Cl- to be the favored reaction over the production of ATP Ado (10, 25) in MM cell lines and on the literature because the concentration of endogenous ATP is signifi- hypothesizing the activity of 8-NH2-Ado activity in cell cantly reduced in the face of elevated 8-NH2-ATP forma- lines derived from solid tumors (11–14), we have tested the tion. These changes in cellular ATP levels not only activity of 8-NH2-Ado in MM cell lines and determined the contribute to apoptosis by affecting nucleic acid synthesis mechanism of action of this compound. We report here that but also may affect the bioenergetics of the cell, as the 8-NH2-Ado causes cell death at a lower concentration and endogenous ATP levels are depleted by 12 hours of at a faster time course and induces a more robust apoptotic incubation with 8-NH2-Ado.

Mol Cancer Ther 2004;3(11). November 2004

Table 1. Comparison of analogue triphosphate in hematologic cell lines

Commercial name Chemical name Cell line Triphosphate accumulation (Amol/L) Reference

8-Amino-adenosine 8-NH2-Ado MM.1S 5,800 Fig. 3A 8-Chloro-adenosine 8-Cl-Ado MM.1S 150 (10) Cladribine 2CdA Raji 15 (33) Fludarabine F-ara-9-h-D-arabinofuranosyladenine CCRF-CEM 30 (33) Clofarabine Cl-F-ara-9-h-D-arabinofuranosyladenine K562 52 (34)

The extensive cellular accumulation of 8-NH2-ATP has therapy has been effective in several lymphoid malignan- an inhibitory effect on both RNA and DNA synthesis in cies, the mechanism of action of which occurs through both glucocorticoid-sensitive and glucocorticoid-resistant metabolic conversion to the triphosphate form (nucleotide) myeloma cell lines. This is in contrast to the deoxy- causing an inhibition of DNA synthesis by direct incorpo- adenosine analogues (cladribine and pentostatin), which ration and chain termination or by interference with DNA inhibit DNA synthesis without effect on RNA synthesis. polymerase and nucleotide reductase. Nucleoside ana- Fludarabine is an arabino-adenosine analogue that inhibits logues with closely related structures and that share DNA synthesis in proliferating cells; however, in indolent metabolic pathways and inhibit similar target enzymes cells such as chronic lymphocytic leukemia cells, fludar- can still exhibit a diverse spectrum of anticancer activity in abine has an inhibitory effect on RNA transcription (27). human tumor types in the clinic. Although the current The experimental adenosine analogue 8-Cl-Ado inhibits DNA-directed purine nucleoside analogues in clinical use RNA synthesis without effect on DNA synthesis in MM cell (cladribine, pentostatin, and fludarabine) have not been lines. Because MM cell proliferate very slowly, it is effective in MM patients, novel analogues may still be plausible that inhibition of RNA synthesis may be an effective in MM and should be tested for clinical efficacy. important growth inhibitory effect in indolent cells, thus Based on our favorable data presented here, 8-NH2-Ado contributing to the 8-NH2-Ado-induced cytotoxicity in MM seems to have a unique mechanism of action and holds cell lines. We hypothesize that the 8-NH2-Ado-induced great promise as a potential therapeutic agent for the decrease in DNA synthesis is secondary to the inhibition of treatment of MM. RNA synthesis and the ensuing cell death. Future studies are planned to examine this issue in detail. Acknowledgments Based on the dramatic reduction in the Dcm and depleted We thank Dr. Joseph G. Cory (East Carolina University, Greenville, NC) for endogenous ATP levels, we speculate that 8-NH2-Ado may the L1210 and ED2 cell lines and Dr. William S. Dalton (H. Lee Moffitt have a direct effect on the mitochondria to promote Cancer Center, Tampa, FL) for the 8226 and MDR10V cell lines. apoptosis as has been reported for other chemotherapeutics (28). To examine this hypothesis, we investigated the References activation of initiator caspases. As anticipated, we mea- 1. Jemal A, Murray T, Samuels A, et al. Cancer statistics, 2003. CA Cancer J Clin 2003;53:5–26. sured 8-NH2-Ado-induced activation of caspase-9 indicating the involvement of mitochondrial damage. However, 2. Attal M, Haroussequ JL, Stoppa AM, et al. A prospective, randomize trial of autologous bone marrow transplantation and chemotherapy in we did not anticipate the activation of caspase-8. To multiple myeloma. N Engl J Med 1996;335:91–7. understand more about the mechanism of caspase-8 activa- 3. Hjertner O, Borset M, Waage A. Comparison of the effects of tion, we investigated whether Fas activation was involved 2-chlorodeoxyadenosine and melphalan in myeloma cell lines. Leuk Res and determined that Fas activation was not required for 1996;20:155–60. caspase-8 activation. Fas-independent activation of cas- 4. Oken M. New agents for the treatment of multiple myeloma and non- Hodgkin lymphoma. Cancer 1992;70:946–8. pase-8 by chemotherapeutic compounds has been reported 5. Gandhi V, Plunkett W. Combination strategies for purine nucleoside in cell lines from breast cancer (29), hepatomas (30), and analogs. New York: Marcel Dekker, Inc.; 2001. p. 195–208. Jurkat cells (31). In addition, activation of caspase-8 inde- 6. Plunkett W, Gandhi V. Purine and pyrimidine nucleoside analogs. pendent of Fas has been reported in B-lymphoid cells and Cancer Chemother Biol Response Modif 2001;19:21–45. found to occur downstream of caspase-3 activation (32). 7. Nagourney RA, Evans SS, Messenger JC, Su YZ, Weisenthal LM. 2 Chlorodeoxyadenosine activity and cross resistance patterns in primary Studies using caspase-8 and caspase-9 specific inhibitors cultures of human hematologic neoplasms. Br J Cancer 1993;67:10–4. were inconclusive; therefore, further studies are needed 8. Lichman SM, Mittelman A, Budman DR, et al. Phase II trials of to determine the mechanism of caspase-8 activation by fludarabine phosphate in multiple myeloma using a loading dose and continuous infusion schedule. Leuk Lymphoma 1991;6:61–3. 8-NH2-Ado in myeloma cells. MM is an incurable hematologic malignancy, which 9. Kraut EH, Crowley JJ, Grever MR. Phase II study of fludarabine phosphate in multiple myeloma. Invest New Drugs 1990;8:199–200. would benefit from the development of new therapeutic 10. Gandhi V, Ayres M, Halgren RG, et al. 8-chloro-cAMP and 8-chloro- approaches. Myeloma cells are slowly replicating, making adenosine act by the same mechanism in multiple myeloma cells. Cancer them candidates for agents that affect transcription and Res 2001;61:5474–9. cellular bioenergetics. DNA-directed purine nucleoside 11. Koontz JW, Wicks WD. Cytotoxic effects of two novel 8-substituted

Mol Cancer Ther 2004;3(11). November 2004

cyclic nucleotide derivatives in cultured rat hepatoma cells. Mol Pharmacol 24. Li H, Zhu H, Xu C-J, Yuan J. Cleavage of BID by caspase 8 mediates 1980;18:65–71. the mitochondrial damage in the Fas pathway of apoptosis. Cell 12. Boe R, Gjertsen BT, Doskeland SO, Vintermyr OK. 8-Chloro-cAMP 1998;94:491–501. induces apoptotic cell death in a human mammary carcinoma cell (MCF-7) 25. Halgren RG, Traynor AE, Pillay S, et al. 8Cl-cAMP cytotoxicity in both line. Br J Cancer 1995;72:1151–9. steroid sensitive and insensitive multiple myeloma cell lines is mediated 13. Vintermyr OK, Boe R, Brustugun OT, et al. Cyclic adenosine by 8Cl-adenosine. Blood 1998;92:2893–8. monophosphate (cAMP) analogs 8-Cl- and 8-NH2-cAMP induce cell death 26. Pawelczyk T, Sakowicz M, Podgorska M, Szczepanska-Konkel M. independently of cAMP kinase-mediated inhibition of the G1-S transition in Insulin induces expression of adenosine kinase gene in rat lymphocytes by mammary carcinoma cells (MCF-7). Endocrinology 1995;136:2513–20. signaling through the mitogen-activated protein kinase pathway. Exp Cell 14. Saito M, Nasu A, Kataoka S, Ohshita K, Yamaji N. Cytotoxic effect of Res 2003;286:152–63. 8-amino adenosine 3V,5V-cyclic monophosphate on FM3A mouse mam- 27. Huang P, Plunkett W. Action of 9-h-D arabinofuranosyl-2-flurorade- mary tumor cells. J Pharmacobiodyn 1989;12:357–62. nine on RNA metabolism. Mol Pharmacol 1991;39:449–55. 15. Goldman-Leikin RE, Salwen HR, Herst CV, et al. Characterization of 28. Costantini P, Jacotot E, Decaudin D, Kroemer G. Mitochondrion as a novel myeloma cell line. J Lab Clin Invest 1989;113:335–45. a novel target of anticancer chemotherapy. J Natl Cancer Inst 2000; 16. Moalli PA, Pillay S, Weiner D, Leikin R, Rosen ST. A mechanism of 92:1042–53. resistance to glucocorticoids in multiple myeloma: transient expression of a truncated glucocorticoid receptor mRNA. Blood 1992;79:213–22. 29. Pirnia F, Schneider E, Betticher DC, Borner MM. Mitomycin C induces apoptosis and caspase-8 and -9 processing through a caspase-3 and Fas- 17. Bellamy WT, Dalton WS, Gleason MC, Grogan TM, Trent JM. independent pathway. Cell Death Differ 2002;9:905–14. Development and characterization of a melphalan-resistant human multiple myeloma cell line. Cancer Res 1991;51:995–1002. 30. You KR, Shin MN, Park RK, Lee SO, Kim DG. Activation of caspase- 8 during N-(4-hydroxyphenyl)retinamide-induced apoptosis in Fas-defec- 18. Cory JG, Carter GL. Leukemia L1210 cell lines resistant to tive hepatoma cells. Hepatology 2001;34:1119–27. ribonucleotide reductase inhibitors. Cancer Res 1988;48:839–43. 19. Krett NL, Zell JL, Halgren RG, et al. Cyclic adenosine-3V,5V-mono- 31. Ferrari D, Stepczynska A, Los M, Wesselborg S, Schulze-Osthoff K. phosphate-mediated cytotoxicity in steroid sensitive and resistant myelo- Differential regulation and ATP requirement for caspase-8 and caspase-3 ma. Clin Cancer Res 1997;3:1781–7. activation during CD95- and anticancer drug-induced apoptosis. J Exp Med 1998;188:979–84. 20. Xie C, Plunkett W. Metabolism and actions of 2-chloro-9-(2-deoxy-2- fluoro-h-D-arabinofuranosyl)-adenine in human lymphoblastoid cells. Can- 32. Wieder T, Essmann F, Prokop A, et al. Activation of caspase-8 in cer Res 1995;55:2847– 52. drug-induced apoptosis of B-lymphoid cells is independent of CD95/Fas 21. Gandhi V, Danhauser L, Plunkett W. Separation of 1-h-D-arabinofur- receptor-ligand interaction and occurs downstream of caspase-3. Blood anosylcytosine 5V-triphosphate and 9-h-D-arabinofuranosyl-2-fluoroade- 2001;97:1378–87. nine 5V-triphosphate in human leukemia cells by high-performance liquid 33. Wilson PK, Szabados E, Mulligan SP, Christopherson RI. Comparative chromatography. J Chromotogr 1987;413:293–9. effects of cladribine, fludarabine and pentostatin on nucleotide metabo- 22. Gandhi V, Chen W, Ayres M, et al. Plasma and cellular pharmacology lism in T- and B-cell lines. Int J Biochem Cell Biol 1998;30:833–42. of 8-chloro-adenosine in mice and rats. Cancer Chemother Pharmacol 34. Parker WB, Shaddix SC, Chang CH, et al. Effects of 2-chloro-9-(2- 2002;50:85–94. deoxy-2-fluoro-h-D-arabinofuranosyl)adenine on K562 cellular metabolism 23. Ferri KF, Kroemer G. Mitochondria—the suicide organelles. Bioessays and the inhibition of human ribonucleotide reductase and DNA poly- 2001;23:111–5. merases by its 5V-triphosphate. Cancer Res 1991;51:2386–94.

Mol Cancer Ther 2004;3(11). November 2004

Nancy L. Krett, Katharine M. Davies, Mary Ayres, et al.

Mol Cancer Ther 2004;3:1411-1420.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/3/11/1411

Cited articles This article cites 32 articles, 12 of which you can access for free at: http://mct.aacrjournals.org/content/3/11/1411.full#ref-list-1

Citing articles This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/3/11/1411.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/3/11/1411. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.