EFFECT OF COMBINED ARGINASE AND NITRIC OXIDE DONOR TREATMENT ON NORMAL AND LEUKEMIC CELLS IN VITRO

Arginine deprivation has been recently suggested as a therapeutic approach against difficult to cure blood cancers. Herein, we investigated for the first time the combined effect of exogenous nitric oxide (NO) donor and recombinant human arginase (rhARG) as arginine-depleting agent on viability of several human leukemic cell lines and normal peripheral blood lymphocytes (PBL). We found that exogenous NO donor, sodium nitroprusside (SNP), at physiologically compatible dose did not counteract but augmented rhARG-mediated pro-apoptotic effect of arginine depletion in leukemic cells but not in resting lymphocytes. Thus, we hypothesize that NO deficiency resulting from arginine deprivation is not the primary cause of high leukemic cells sensitivity to the ac-tion of rhARG. The results of this study further support the notion that not arginine ca-tabolism but other cell response mechanisms must be involved in determining cell fate upon arginine restriction. SNP or alternative NO donors can be proposed as compo-nents of metabolic anti-leukemia therapy based on arginine deprivation. phytohaemagglutinin (PHA, “Sigma”, USA) and 20 U/ml recombinant inter-leukin 2 (rIL-2, “Millipore Corp.”, USA) in culture medium for 4 days. Resting PBL were cultured in the presence of 20 U/mL rIL-2 only. Resting and proliferatively active PBL were used in the experiments.


INTRODUCTION
Acute leukemia is a high-risk fast-proliferating cancer, which is associated with clonal proliferation and accumulation of abnormal cells derived from immature hematopoietic cells [16]. Current clinical strategy against childhood acute leukemia often employs combination of standard anticancer agents with asparaginase as a second line therapy [21]. This is because certain type of leukemia and lymphoma are auxotrophic for asparagine and thus susceptible to the treatment with recombinant asparaginase [23].
MTT assay. Concentration-and time-dependent SNP cytotoxicity was determined using the standard MTT assay. Cells were grown in 96-well plates with 0, 0.01, 0.1, 1 and 10 mmol/l of SNP in either complete medium (CM) or the CM with 2 U/ml rhARG. After different treatments, 20 µL of 5 mg/mL MTT solution ("Sigma", USA) was added to each well (0.1 mg/well) and incubated for 5 h. The supernatants were aspirated, the purple formazan crystals in each well were dissolved in 200 µL of dimethyl sulfoxide and optical density at 540 nm was read on a Microplate Reader ("Biotek", USA). The amount of SNP needed to kill 50 % of the cells in a culture was defined as the SNP inhibitory concentration (IC50).
Apoptosis detection. Cells were resuspended in Ringer solution with calcium ions to a final concentration of 1×10 6 /ml. Then 5 µl of cells suspension was mixed with 5 µL of Annexin V-FITC ("Sigma", USA, final concentration 3 µg/mL) followed by propidium iodide (final concentration 0.5 µg/mL) as a counter stain. The cells were then incubated at room temperature for 5-15 min in the dark. Finally, the cells were covered with a coverslip and examined by fluorescent microscopy (Carl Zeiss, Germany). The percentage of apoptotic cells was calculated using ImageJ software.
Statistical analysis. All experiments were repeated at least three times. Levels of significant differences between groups were determined by the Student's t-test. P values less then 0.05 were considered statistically significant.

RESULTS AND DISCUSSION
Arginine deprivation enhances SNP cytotoxicity for leukemic cells. In our previous study we found that SNP, a NO donor, does not counteract apoptotic cell death and, in fact, increases cytotoxic effect of arginine deprivation for Jurkat leukemic cells in defined arginine-free medium [5]. We also demonstrated that rhARG treatment alone is relatively harmless for normal PBL [4]. Herein, we aimed to investigate how efficient can be combined rhARG and SNP treatment against leukemic cells of different origin. Three human leukemia cell lines were used as models for acute T-(Jurkat and CEM-T4) and B-cell (Namalva) leukemias. As control cells, we aplied primary isolated resting and PHA-activated human PBL from healthy donors. First, we examined cytotoxic and cytostatic effects of exogenous NO donor drug on normal and leukemic cells under arginine limitation. For this purpose, cells were treated with several increasing concentrations of SNP (0.01, 0.1, 1 and 10 mmol/l), either alone or in combination with purified rhARG in concentration of 2.0 U/mL. The SNP cytotoxic concentration (IC50) for leukemic and PBL cells upon different culture conditions were performed by the MTT assay. We showed that SNP cytotoxicity grew in a dose-and time-dependent manner and was considerably enhanced by arginine restriction ( Fig. 1). At 24, 48, and 72 h, the SNP IC50 in normal medium for leukemic cells were approximately similar in 2-5 mM range for Jurkat, CEM-T4, and Namalva cells ( Table 1). As shown in Fig. 1 and Table 1, in the medium with rhARG, SNP mediated cytotoxicity increased and IC50 at 72 h were significantly 3-7 times lower compared to IC50 values in complete medium (CM) and ranged between 0.4 and 1 mM for the three cell lines tested (p<0.05). Of note, Namalva cells appeared to be more resistant to SNP, which may be associated with their B-cell origin. However, why exogenous NO is more cytotoxic for non-proliferating arginine-deprived than for proliferating in the CM cancer cells remains to be elucidated. It is necessary to stress that similar effect was observed in our previous study on Jurkat cells incubated in defined arginine-free medium [5].
We next analyzed the SNP IC50 values for PBL in the same experimental conditions. Surprisingly, we observed that experimental normal cells were in general more sensitive to SNP as compared to malignant cells. Interestingly, SNP IC50 values for resting PBL did not differ drastically between control medium and medium with rhARG ( Fig. 1, Table). However, as also shown in Fig. 1, SNP-mediated cytotoxicity significantly differed (were lower) for the activated PBL in complete medium, as compared to resting cells (p<0.05). In contrast to leukemic cells, either proliferatively active or resting PBL were more sensitive to SNP treatment in complete medium.
Taken together, these data suggest, that nitric oxide donor is more toxic toward normal PBL, specially for proliferatively active PBL, but synergistically increases arginasemediated arginine deprivation cytotoxicity selectively for leukemic cells. Of note, SNP in very low concentrations (below 100 µM) that was proposed as therapeutic dose [5] was non-toxic for resting immune cells.
Low dose NO potentiates the arginine deprivation-mediated cytotoxicity against proliferatively active cells. NO-releasing compounds such as SNP are widely used to investigate the effects of NO on various physiological processes and molecular mechanisms of the cell [15]. In SNP, NO is coordinated as a nitrosyl group ligated to iron in a square bipyramidal complex, and is released spontaneously at physiological pH in one step reaction [13]. It was shown that NO formation in vivo is accompanied by cyanide (CN -) release which is rapidly metabolized by the liver to thiocyanate that is next released by the kidneys. However, despite its apparent toxicity, SNP is a popular drug, most effective in some difficult clinical circumstances [12].
To better clarify the role of NO and cyanide in SNP-mediated cytotoxic and cytostatic activity, we investigated the effects of potassium ferricyanide (K 3 [Fe(CN) 6 ]) and sodium thiosulphate (Na 2 S 2 O 3 ) as controls for the ferricyanide or cyanide moieties released by SNP [29]. As shown in Fig. 2, F vs A, 1 mM SNP caused significant growth   6 ] alone. Also, the cytotoxicity elicited by SNP was not abolished when cells were co-incubated with 1mM sodium thiosulphate -a substance known to bind CNmoiety (Fig. 2, F) [15]. According to this data, we suggest that only NO released by SNP is responsible for modulation of viability of experimental model cells used herein We further examined in more details the dynamics of viability of model cells under rhARG-mediated arginine starvation alone or in combination with low dose SNP in trypan blue dye exclusion assay. We have chosen one maximally tolerated dose of SNP (0.05 mM) that did not produce any apparent and unacceptable cytotoxicity in both arginine-containing and arginine-free media for normal resting PBL according to the obtained IC 50 values (Table). It is estimated by monitoring nitrite accumulation 0.05 mM SNP releases up to 20-30 µM of NO into the culture medium, the level compatible with physiologically observed in blood plasma (30-70 µM) [5,9].
As was expected, none of the cell lines incubated with rhARG up to 72 h exhibited significant growth (Fig. 2, A-C). As shown in Fig. 2 D, rhARG significantly inhibited proliferation of PHA-activated PBL but did not drastically decrease cell viability (p<0.05). Also, arginine restriction by rhARG had no effect on normal resting PBL viability   (Fig. 2, E). SNP at 0.05 mM concentration had no effect on leukemic cells growth (only slightly suppressed proliferation in complete medium; Fig. 2, A-C), but significantly suppressed PHA-stimulated proliferation of PBL (Fig. 2, D; p<0.05). In turn, as shown in Fig. 2, incubation of leukemic cells and PHA-activated PBL with 0.05 mM SNP in the presence of rhARG caused some decrease in cell viability. Jurkat T-lymphoblastic leukemia was more sensitive to this combinational treatment (Fig. 2, A). However, there was no statistically significant difference observed in the amount of viable cells in resting PBL (Fig. 2, E) under the combined treatment (p<0.05). Thus, low dose of SNP slightly, but evidently enhanced cytotoxicity of rhARG for all proliferatively active cells.
Normal resting PBL were highly resistant to the combined drugs action. Arginase and SNP treatment induces apoptosis selectively in leukemic cells. To better understand the mechanism of leukemic cell sensitivity to the combined rhARG and SNP treatment, apoptosis progression in malignant cells and PBL was investigated. During induction of apoptosis, as a target of active caspase-3, PARP to be cleaved into two fragments [14]. We therefore analyzed the appearance of cleaved form of PARP (poly(ADPribosyl)-polymerase) protein by western blotting. It was observed that rhARG treatment induced time-dependent apoptosis in three leukemic cell lines (Fig. 3, A). Importantly, supplementation with 0.05 mM SNP had no significant inhibitory effect on this process. This data was also corroborated by cells double staining with fluorochromes Annexin V-FITC (binds to phosphatidylserine on membranes of apoptotic cells) and propidium iodide (Fig. 3, B,C). It was shown that rhARG treatment induced morphological changes in leukemic cells such as cell shrinkage, nuclear condensation and phosphatidylserine externalization, which are typical hallmarks of apoptosis (Fig. 3, C). We observed a higher number of apoptotic cells upon 72 h-long rhARG treatment (Fig. 3, B) only for tested leukemic cell lines (30-46 %, p<0.05). Compared to rhARG treated values alone, the percentage of apoptotic cells considerably increased under both rhARG and SNP treatment (37-62 %). Importantly, there were no significant alterations in the number of apoptotic cells in normal resting PBL under all treatments: only spontaneous apoptosis was detected under both rhARG and Cav treatment (9-16 %; p<0.05). As shown in Fig. 3 B, C, minimal signs of apoptosis in PHA-activated PBL were observed after 72 h incubation of these cells in culture with rhARG alone (17 %). In agreement with cytotoxicity data above, the percentage of apoptotic cells significantly increased (34 %) under combinatory rhARG and SNP treatment in PHAactivated PBL. According to this data, we hypothesize that NO deficiency resulting from arginine deprivation is not the primary cause of high leukemic cells sensitivity to rhARG. In the opposite case, exogenous NO would rescue leukemic cells viability upon arginase treatment. This and other our studies on different cell models also support the notion that not arginine metabolism but other cell response signaling mechanisms must be potentially involved in determining cell fate upon arginine restriction [2,28].