Cell proliferation and growth arrest could be regulated by the availability of nutritional factors. Amino acid degrading enzymes, e.g., asparaginase and methioninase, have been reported as potent in vivo anti-tumor agents, as well as being effective growth inhibitors of cancer cells in culture. Asparagininase continues as a therapy in T-cell acute lymphoblastic leukemia or lymphoma, although it produces serious side effects including anaphylactic shock and pancreatitis.1, 2 Methioninase significantly inhibits the growth of various cancer cells and shows anti-tumor activity in vivo since many tumor cells apparently have “an elevated growth requirement for methionine” compared to normal cells. This enzyme is undergoing clinical trials as an anti-tumor drug.3, 4 Arginine deiminase (ADI) catalyzes the hydrolysis of arginine into citrulline and ammonia, and is highly expressed in various microorganisms utilizing arginine as a major nonglycolytic energy source.5, 6 ADI derived from Mycoplasma induces apoptosis of human cancer cell lines7, 8, 9, 10 and inhibits angiogenesis.11, 12, 13 Its anti-tumor activity is most likely due to the depletion of arginine from the nutritional sources for cells because an exogenous supplement of arginine restores cell growth.7 In contrast, depletion of arginine had a cytoprotective effect in some cells. For example, arginase treatment suppressed neuronal cell death induced by multiple stimuli, such as glutathione depletion, staurosporine treatment and Sindbis virus infection.14 The cytoprotective mechanism of arginase in these apoptotic paradigms acts probably by inhibiting protein synthesis through arginine depletion.
Recently, we found that ADI purified from Mycoplasma arginini strongly inhibited proliferation of DU145 prostate cancer cells, and antagonized taxol-induced apoptotic cell death.15 In addition to the cytostatic effect that resulted from inhibition of protein synthesis, low levels of ADI arrested DU145 cells in G1. Cell arrest in the G1- and/or S0-phases by ADI was also reported to precede apoptosis of various cancer cells.9 DU145 cells arrested in G1 phase by ADI were more likely resistant to an anti-cancer agent such as taxol associated with G2/M phase arrest. This result was further supported by the fact that deferoxamine, which could arrest cells mainly at G1/S phases, decreased the number of cells arrested in G2/M phases and protected cells from taxol-induced apoptosis.16 Therefore, it was hypothesized that cancer cells arrested in G1 phase by ADI would be more sensitive to G1-specific anti-cancer drugs.
Glucocorticoids are well-characterized growth inhibitors of lymphoblastic leukemia CCRF-CEM cells that act through irreversible G1 arrest, leading often to cell death.17, 18 Furthermore, arginine is a more critical nutrient for the cultures of human T cells and T-lymphoblastoid cell lines than B and myeloid cell lines.7 These findings have led us to examine the synergistic anti-proliferative effects of ADI and dexamethasone (DEX) in vitro on CCRF-CEM cells, an acute human lymphoblastic leukemia cell line. In addition, we have explored the expression of cell cycle regulatory proteins, c-Myc and p27Kip1, which are heavily involved in G1 cell cycle arrest and cell death.
Human T-cell acute lymphoblastic leukemia CCRF-CEM (ATCC CCL119) cells were cultured in RPMI1640 (Gibco BRL, Rockville, MD) supplemented with 10% fetal bovine serum and 1% antibiotics-antimycotics in a humidified 5% CO2 incubator at 37°C. The DEX-resistant CEM (CEM-DR) cells were derived from the parental CCRF-CEM cells cultured for >10 passages in the presence of 1 μM DEX to select for cells resistant to this agent. The CEM and CEM-DR lines were routinely subcultured by 1:8 dilution of the cell suspension every 3–4 days to grow over the range 0.2–2 × 106 cells per ml. For viability assay, cell cultures were initiated at 2 × 105 /ml and treated with DEX (Sigma Chemical Co., St. Louis, MO) and/or ADI after 2 hr incubation. The number of viable cells was counted by the trypan blue dye exclusion assay with a hemocytometer. Cytotoxicity was determined using the lactate dehydrogenase (LDH) assay kit (Promega, Madison, WI). A portion of DEX and/or ADI-treated cells was centrifuged, and culture supernatant (50 μl) was incubated with an equal volume of LDH substrate solution in the dark for 30 min. The reaction was stopped with 50 μl of 1 M acetic acid, and the absorbance was determined at 492 nm.
Arginine deiminase (ADI)
The gene encoding ADI was cloned from Mycoplasma arginini and TGA tryptophan codon was changed to TGG by site-directed mutagenesis because TGA corresponds to the stop codon in E. coli, whereas TGG is a tryptophan codon. The modified ADI gene was subcloned in E. coli expression vector pET32a (Novagen, Madison, WI) and designated pET32a-ADI.19 ADI was purified to homogeneity on SDS-PAGE from inclusion bodies accumulated in the cytoplasm of E. coli NovaBlue(DE3) (Novagen, Madison, WI) transformed with pET32a-ADI and had a specificity of 20 units/mg protein. One unit of ADI is the amount of enzyme catalyzing 1 μmole of arginine to citrulline per min at 37°C under the assay conditions.15
Assay for DNA fragmentation
CEM or CEM-DR cells were seeded at 2 × 105/ml and treated with ADI in the presence or absence of DEX for 4 days. The cells were harvested and lysed with the genomic DNA lysis buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 25 mM EDTA and 0.5% SDS). The lysate was incubated with 20 μg/ml RNase A at 37°C for 1 hr, followed by 10 μg/ml proteinase K at 55°C for 16 hr. The DNA was extracted twice with phenol-chloroform and precipitated with an equal volume of isopropanol. The pellet was washed with 70% ethanol and air-dried. DNA (10 μg) was separated by 2% agarose gel electrophoresis with 1x TBE buffer.15
Total RNA isolation and Northern blot analysis
For c-myc mRNA expression, CEM cells (2 × 105/ml) were seeded and harvested at various time-points after treatment with ADI (100 ng/ml). Total RNA was isolated by the acid guanidium thiocyanate and phenol/chloroform extraction method of Chomczynski and Sacchi.20 Total RNA samples (10 μg) were size-fractionated in a 1.2% agarose gel containing 0.67 M formaldehyde and transferred to a Hybond N+ nylon membrane (Amersham, Arlington Heights, IL). The membrane was baked at 120°C for 30 min and hybridized with 32P-labeled human c-myc cDNA probe in hybridization buffer containing 6 × SSC, 1% SDS, 20 mM sodium phosphate, pH 7.0, 50% deionized-formamide and 0.5 × Denardt's solution. After 16 hr hybridization, the membrane was washed in 2 × SSC and 0.1% SDS at 50°C for 30 min, and twice in 0.2 × SSC containing 0.1% SDS at 55°C for 30 min. The membrane was exposed to X-ray film in an intensifying screen at −70°C for 2 days.
Western blot analysis
Protein samples were extracted from the CEM and CEM-DR cells treated with ADI, DEX and ADI plus DEX. Cells were washed with cold PBS and lysed in cell lysis-buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.02% sodium azide, 0.1% SDS, 100 μg/ml phenylmethylsulfonyl fluoride (PMSF), 1.0 μg/ml aprotinin, 1.0% Nonidet P-40 (NP-40) and 0.5% sodium deoxycholate. Protein samples (50 μg) were resolved by 10% SDS-polyacrylamide gel electrophoresis and transferred to a PVDF membrane (Bio-Rad, Hercules, CA). The membrane was blocked with 5% blocking solution for 1 hr and incubated with monoclonal anti-c-Myc antibody (Clone 9E10.3, 1:1000, NeoMarkers, Fremont, CA) or rabbit polyclonal anti-p27Kip1 antibody (C-19, 1:1,000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA). The immunoblot was developed with the Westzol-Blotting System (Intron, Seoul, Korea) using peroxidase-linked rabbit anti-mouse IgG and goat anti-rabbit IgG (Zymed, San Francisco, CA), respectively.
Flow cytometry analysis
For cell cycle analysis, 2 × 106 cells were centrifuged at 1,000 rpm for 10 min at 4°C. Cell pellets were suspended in 10 ml sample buffer (1 g/L glucose in PBS, filtered through a 0.22 μm filter) and washed twice, suspended in 70 % ethanol and fixed overnight at 4°C. Fixed cells were briefly vortexed and centrifuged at 3,000 rpm for 5 min. The ethanol was discarded, and pellets were stained with 1 ml of propidium iodide (PI) solution (50 μg/ml PI in sample buffer containing 100 U/ml of RNase A). Before analysis by flow cytometry, each sample was incubated for 30 min at room temperature. PI-DNA complex in each cell nucleus was measured with FACS (Beckton & Dickinson Co., CA). The percentage of cells in sub-G1, G1, S and G2/M phases was analyzed with ModiFIT software.
Results were expressed as mean ± S.D, for at least 3 treatments and n = 3 or 4 per treatment. Statistical significances were evaluated by Student's t-test; p-values less than 0.05 were considered to be significant.
Growth inhibition of CEM cells by ADI
Since ADI was previously shown to inhibit growth of various human tumor cells and vascular endothelial cells,9, 21 we examined the growth-inhibitory effect of ADI on acute lymphoblastic leukemia CEM cells by measuring LDH release and cell number. CEM control cells grew exponentially over 4 days. ADI treatment significantly inhibited the rate of cell growth in a dose dependent manner from 4 to 100 ng/ml (Fig. 1a). The growth inhibitory effect of ADI was seen as cycle arrest and cell death. Although the value of LDH released in culture media was not altered at day 4 after treatment with ADI, LDH release/cell number dramatically increased in a dose-related fashion (Fig. 1b). Since the number of cells was considerably different by day 3 after ADI treatment, LDH release/cell number on day 4 would be more informative than LDH release by itself. ADI-treated cells appeared smaller, exhibiting cell shrinkage, nuclear and chromatin condensation, compared to untreated control cells (data not shown). In addition, ADI treatment induced DNA fragmentation, indicating apoptotic features of cell death (Fig. 1c). As changes in cell cycle and progression towards apoptosis are associated with the expression of c-myc and p27Kip1, we examined the products of these 2 genes. The level of c-myc mRNA was downregulated as soon as 2 hr after ADI treatment (100 ng/ml), while c-Myc protein was suppressed after 12 hr (Fig. 2a). The expression of p27Kip1, one of target molecules of c-Myc, was induced at 3 days after ADI treatment, also in a dose-dependent manner (Fig. 2b). The data confirm that ADI arrests growth and subsequently induces cell death in CEM cells, and these effects appear to be directly related to c-myc downregulation and p27Kip1 upregulation.
ADI enhanced DEX-induced cytotoxicity of CEM cells
ADI arrests cancer cells at in the G1- and/or S0-phases prior to apoptotic cell death.9 Cells arrested in G1 phase are more likely to be resistant to an anti-cancer agent, such as taxol, associated with G2/M phase arrest.15 We surmised that cancer cells arrested in G1 phase by ADI should be more sensitive to G1 cycle-specific anti-cancer drugs. DEX inhibits proliferation of lymphoblastic leukemia CEM cells through G1 arrest and apoptosis.17, 18 It is plausible that ADI and DEX act synergistically in causing cell death. To test this possibility, cell viability, DNA fragmentation and apoptotic cell debris, as well as expression of c-myc and p27Kip1, were analyzed in CEM cells after cotreatment with 100 ng/ml ADI and 1 μM DEX. This combined treatment caused 1) a decrease of cell numbers over 4 days and 2) an increase of LDH release on day 4, significantly greater than was found in controls or in cells treated with either ADI or DEX alone (Fig. 3a,b). When CEM cells are individually treated with 100 ng/ml ADI or μM DEX for 2 days, cell debris in the sub-G1 fraction was mildly increased compared to untreated control cells. On the other hand, cotreatment synergistically increased the number of cell debris compared to ADI or DEX treatments (Fig. 4), in parallel with DNA fragmentation (data not shown). In addition, cotreatment with ADI and DEX suppressed c-Myc protein by day 1 and increased p27Kip1 expression by day 3 (Fig. 5).
Effect of ADI on DEX-resistant CEM cells
Glucocorticoids are currently employed in anti-leukemic therapy due to their ability to block proliferation and induce lysis of susceptible lympholasts. However, the occurrence of primary or secondary glucocorticoid-resistance limits their clinical usefulness. To investigate whether ADI also inhibits proliferation of glucocorticoid-resistant cells, we selected DEX-resistant CEM (CEM-DR) cells by culturing the parental CCRF-CEM cells in the presence of 1 μM DEX. The CEM-DR cells were routinely subcultured in suspension in RPMI1640 media containing 10% FBS and 1 μM DEX, and maintained by 1:8 dilution every 3–4 days to keep cell densities of 0.2–2 × 106 cells per ml. Although DEX treatment inhibited proliferation of parental CEM cells, growth of CEM-DR cells and LDH release/cell number were not influenced by 1 μM DEX treatment (Fig 1a, 6a,b). In addition, DEX did not suppress c-Myc expression in CEM-DR cells, whereas its downregulation was seen in parental CEM cells by DEX treatment (Figs. 5 and 6c). However, ADI significantly inhibited growth of CEM-DR cells in a dose-dependent manner, as in parental CEM cells (Fig 6a). LDH release/cell number and apoptotic bodies in sub-G1 fraction were increased in CEM-DR cells by treatment with 100 ng/ml ADI alone, although they were not synergistically enhanced by cotreatment with DEX in contrast to the synergism seen in parental CEM cells (Figs. 1b, 4, 6b,d). In addition, DEX produced neither suppression of c-Myc protein nor increase of p27Kip1 expression in CEM-DR cells. However, ADI inhibited c-Myc and upregulated p27Kip1 expression in CEM-DR cells in much the same way as in parental CEM cells (Fig. 5 and 6c).
Synergism of ADI and DEX on anti-leukemic activity
ADI induces apoptotic cell death of a variety of human cancer cells in vitro as well as have a potent anti-tumor activity in vivo, indicating its pro-cytotoxic activity.7, 9, 10 However, depletion of arginine with ADI arrested cells in G1 phase, and prevented taxol-induced cytotoxicity of DU145 prostate cancer cells, suggesting a cytoprotective activity.15 This discrepancy suggests that it should not be used for the combination therapy with an anti-cancer agent (e.g., taxol) associated with G2/M phase arrest, although ADI itself induces apoptosis and does exert an anti-tumor activity in vivo. Conversely, we might assume that cancer cells arrested in G1 phase with ADI are more sensitive to glucocorticoid in anti-lymphoblastic leukemia therapy, in which it also induces G1 arrest and apoptotic cell death. Our results show that recombinant ADI purified from E. coli inhibited proliferation and resulted in apoptotic cell death through cycle arrest in acute human lymphoblastic leukemic CCRF-CEM cells. Consequently, synergism of the anti-leukemic effect of ADI in conjunction with DEX was worth investigating. Indeed, ADI treatment proved an enhancer of DEX-mediated cytotoxicity in CEM cells, as shown by viable cell counting, DNA fragmentation, and cell-cycle analysis.
ADI specifically hydrolyzes arginine to citrulline, leading to depletion of an essential amino acid. Arginine is known to be the first amino acid depleted by normal cell metabolism and is less efficiently recycled from metabolized proteins than other amino acids.22 Deprivation of arginine by ADI therefore disrupts many biochemical pathways, leading more quickly to cell death than depletion of other single essential amino acids.23, 24 However, there is a problem regarding reutilization of citrulline in ADI treatment, since generation of equimolar levels of arginine from it should only partly restrict growth in cells that have adequate arginosuccinate synthetase levels, and clearly availability remains restricted.24, 25 The anti-leukemic action of the enzyme is however principally due to a drastic limitation of free arginine. As a consequence, it reduces polyamine, proline and glutamate production because arginine is also a substrate for their syntheses in mammals.26 Because polyamines are essential for proliferation of cancer cells,27 their reduced syntheses simply adds further weight to the anti-proliferative effect. Our recent results show that depletion of polyamines alone by α-difluoromethylornithine (DFMO) induces G1 arrest and increases DEX-mediated G1 arrest in CEM cells.28 Pretreatment with sublethal concentrations of sequential polyamine pathway inhibitors, DFMO and methylglyoxal-bis-guanylhydrazone (MGBG), an inhibitor of S-adenosylmethionine decarboxylase, synergistically enhanced the extent and onset of apoptotic cell death induced by DEX,29 although MGBG did not act synergistically with another enzyme (arginase) that inhibited tumor cell growth in vitro (Wheatley and Lamb, unpublished data). This simply illustrates that inhibition of protein synthesis has a much more devastating effect than polyamine synthesis in bringing about not just cell arrest here but causing cell death, which is far more important if the treatment is to kill tumor cells rather than merely arrest them. Reduced polyamine synthesis due to ADI treatment will contribute to growth arrest, but a synergistic anti-leukemic effect in conjunction with DEX is probably not going to make any significant difference in terms of cell kill. Recently, Miller et al.,29 reported that DEX reduced the level of putrescine by ODC inhibition, but the downstream polyamines, spermidine and spermine increased during the time-frame in which DEX proceeded to initiate apoptosis. Thus, it might be worth determining the intracellular levels of polyamines after depletion of arginine by ADI treatment to see if it does seriously assist the synergistic response.
Growth inhibitory effect of ADI on DEX-resistant CEM-DR cells
Although glucocorticoids are currently used in anti-leukemic treatment regimes, glucocorticoids-resistance can limit their clinical usefulness. DEX alone did not inhibit proliferation of CEM-DR cells, although the expression of functional glucocorticoid receptors was not determined. However, the growth arrest and/or apoptotic cell death of CEM-DR cells was clearly demonstrated by ADI, alone and in conjunction with DEX. Suppression of the proto-oncogene c-myc is one of early responses in association with growth arrest and/or apoptosis of CEM leukemic cells induced by DEX treatment.30, 31, 32 In fact, c-Myc expression was suppressed in CEM cells but not in CEM-DR cells by 1 μM DEX. However, dose-dependent reduction in c-Myc expression was detected not only in CEM cells but also in CEM-DR cells after ADI treatment.
We also examined the expression of p27Kip1 since Myc suppressed expression of cell cycle/growth arrest gene such as p27Kip1,33 and induction of p27Kip1 is associated with the G1 arrest induced by DFMO treatment in CEM cells.28 As expected, ADI treatment induced p27Kip1 in both CEM and CEM-DR cells and this correlated with suppression of c-Myc, but 1 μM DEX had no significant effect on expression of c-Myc and p27Kip1, cell proliferation, LDH release and DNA fragmentation in CEM-DR cells. Vulnerability of cancer cells to essential amino acid deprivation is known to be associated with low stringency of the cell cycle checkpoints. After arginine deprivation, the majority of normal diploid fibroblasts can survive for several weeks in a quiescent state (Go) after exiting from the cell cycle, whereas tumor cells lacking stringent regulation of G1 checkpoint fail to enter quiescence and keep on dividing up to apoptotic cell death. This clearly indicates a selective potential advantage in cancer therapy.22, 24 In conclusion, the synergistic effect by cotreatment of ADI plus DEX may result from G1 cell cycle arrest leading to cell death in both CEM and CEM-DR cells. The cooperation between suppression of c-Myc and upregulation of p27Kip1 leads to early apoptotic events in CEM cells. These results suggest that ADI, which is known to exert anti-tumor activity and has few side effects, may be developed to improve the therapy of leukemia with lower overall toxicity in combination with DEX, which is widely used in the therapy of various lymphoid malignancies. ADI might even be effective in DEX-resistant cells. The importance of these in vitro findings encourages us to turn now to an appropriate in vivo model and establish whether it can be as effective a treatment as in vitro.
We thank Miss Jane Sarginson and Kyoung-Yeon Chae for assistance with cell culturing.