l.eukemla Research Vol 15. No 4. pp 20~-213. 1991 IH45-212e,.,'91 $3.00 + .00 Printed in Great Britain Pergamon Press pie

EFFECT OF T E T R A H Y D R O U R I D I N E AND D E O X Y T E T R A H Y D R O U R I D I N E ON THE INTERACTION

BETWEEN 2 ' -DEOXYCYTIDINE AND 1-fl-D- ARABINOFURANOSYLCYTOSINE IN HUMAN LEUKEMIA CELLS

STI-VEN GRANT,* KAPU, BHALLA and CARL MCCRADY

Division of Hematology/Oncology and Department of Pharmacology. Medical College of Virginia. Richmond. VA 23298. U.S.A. and Division of Hematology/Oncology, Medical University of South

Carolina, Charleston. SC. U.S.A.

(Received 13 March 19911. Revision accepted 31 August 1990)

Abstract--The interaction betwccn 2'-deoxycytidine (dCyd) and 1-fi-D-arabinofuranosylcytosine (ara- C), administered at pharmacologically achievable concentrations, was examined in four continuously cultured human leukemia cell lines, HL-60, KG-I. K-562, and CCRF-CEM. In three of the cell lines (HL-60, K-562, and CCRF-CEM), co-administration of 20 or 50 aM dCyd with 10 `aM ara-C reduced ara-CTP formation by at least 90% and incorporation of ara-C into DNA by at least 80%. In contrast, KG-I cells exhibited substantially smaller reductions in both ara-CTP formation and incorporation of ara-C into DNA under identical conditions. KG-1 cells were distinguished by the highest activity of the enzyme cytidine deaminase of the four lines assayed, and exhibited the smallest increments in the intracellular accumulation of both dCyd and dcoxycytidine triphosphate (dCTP) in response to exogenous dCyd. Co-administration of lmM tetrahydrouridine (THU) or 0.5mM deoxv- tetrahydrouridinc (dTHU) had little effect on the ability of dCyd to antagonize ara-C metabolism in HL-60, KG-I and K-562 cells. In contrast, these deaminase inhibitors substantially incrcased the intracellular accumulation of dCTP as well as the ability of dCyd to antagonize ara-CTP formation and incorporation of ara-C into DNA in KG-I cells. THU and dTtIU also permitted d('vd to antagonize ara-C growth inhibitory effects in KG-I cclls to thc extent observed in the other leukemic cell lines. These studies suggest that the intracellular deamination of exogenous deoxycytidine may influencc the degree to which this nucleoside antagonizes ara-C metabolism and toxicity in some leukemic cells. They also raise thc possibility that deaminase inhibitors may be employed to modulate. and perhaps to improve, the therapeutic selectivity of pharmacologically relevant concentrations of ara-C and dCyd in the treatment of acute leukemia in man.

Key word~': ara-C, deoxycytidine, tetrahydrouridine, deoxytetrahydrouridine, leukemia.

I N T R O D U C T I O N

1-/~-D-ARABINOFURANOSYLCYFOSINE (ara-C) is a deoxycytidine (dCyd) analog which is an effective antileukemic agent in man [1]. It is transported across cell membranes by a facilitated nucleoside diffusion mechanism [2]. and converted to its nucleotide

Abbreoiations: ara-C, 1-fi-D-arabinofuranosylcytosinc; ara-U, 1-fl-D-arabinofuranosyluracil; dCvd, 2'-deoxycy- tidine; dThd, thymidine; ara-CTP, 1-/%D-arabinofurano- sylcytosine triphosphate: dCTP, deoxycytidinc triphosphate; dUrd, deoxyuridine; dCMP, deoxycytidinc monophosphate; dCK, deoxycytidine kinase; THU, tetra- hydrouridine; dTHU. deoxytetrahydrouridinc; TCA. tri- chloroacetic acid.

* Supported by a Scholar award from the Leukemia Society of America. To whom correspondence should bc addressed.

2115

derivative, ara-CMP, by the pyrimidine salvage path- way enzyme, deoxycytidine kinasc (dCK) [3]. It is ultimately metabolized to its tr iphosphate form, ara- CTP, which inhibits DNA polymerase a [4] and is incorporated into elongating strands of DNA. Evi- dence currently suggests that ara-C DNA incor- poration, and resultant interference with chain initiation and elongation, as well as premature chain tcrmination, are primarily responsible for ara-C- mediated lethality in leukemic cells [5]. Opposing these anabolic processes is the degradative enzyme cytidine dcaminasc, which converts ara-C to an inac- tive derivative, 1-fl-D-arabinofuranosyluracil (ara-U) [6]. Cytidinc deaminase is found at high con- centrations in the plasma, as well as in several tissues including liver, gastrointestinal epithelium, and kid- ney [7]. It is also present in leukemic cells, although enzyme activity in this site has been reported to bc

206 S. GRANT et al.

dCyd

Ara-C

Cell Woll

Ii Nucleoside Tronsport

dUrd dUMP

dCK MPK DPK dCyd ~- dCMP m ~ dCTP -~

(~)dTTP { DANA Polymerose GdCTP ~ - ~ DNA

dCK MPK DPK ~-- ,&re- C Aro-CMP Aro-CTP.-) --~

I cyd ®drrP I dCMP ~dCTP I Deominase

T H e , eammose d T H ~ . ~

Aro- U Are- UMP

FIG. 1. A schematic diagram of the biochemical pathways of ara-C and dCyd metabolism. MPK, monophosphate kinase: DPK, diphosphate kinase. ((~) Positive allosteric

regulation; (G) negative allosteric regulation.

low [8]. In addition, ara-CMP may be deaminated to ara-UMP by the enzyme dCMP deaminase, although the extent to which this process occurs varies con- siderably among different cell types [9]. The relevant biochemical pathways involved in the intracellular metabolism of ara-C are shown in Fig. 1.

The naturally occurring nucleoside deoxycytidine serves as a substrate for each of these enzymes and antagonizes the metabolism of ara-C at several levels within cells [10] (Fig. 1). Antagonism may be direct, through competition at the level of transport, phos- phorylation by dCK or incorporation into DNA [11]. Alternatively, dCyd may indirectly interfere with ara-C phosphorylation as a consequence of feedback inhibition of dCK by deoxycytidine triphosphate (dCTP) [12]. Despite antagonism of ara-C metab- olism and cytotoxicity, there is evidence that co- administration of dCyd may improve the anti- leukemic selectivity of ara-C by preferentially pro- tecting normal hematopoietic cells from ara-C- mediated lethal effects. For example, Harris and Grahame-Smith reported that dCyd was inefficient in antagonizing ara-CTP formation in leukemic mye- loblasts displaying low levels of phosphorylating activity [13]. Buchman et al. demonstrated that co- administration of dCyd with otherwise lethal doses of ara-C in mice inoculated with the L1210 routine leukemia led to survival benefits superior to those obtained with the optimal dose of ara-C administered alone [14 I. More recently, Bhalla and co-workers have shown that dCyd preferentially antagonizes the metabolism and cytotoxicity of ara-C in normal vs leukemic myeloid progenitor cells, as determined by

soft-agar cloning assays [15]. Finally, preliminary results from a Phase I trial in humans suggest that dCyd administered at doses that yield plasma dCyd levels in excess of 20 ~tM substantially ameliorates the non-hematologic toxicity of continuously admin- istered high-dose ara-C without abrogating anti- leukemic activity [16].

A better understanding of the factors governing thc interaction bctwccn dCyd and ara-C in leukemic cells would be useful in attempting to optimize regi- mens employing these two agents. The present studies were undertaken in order to characterize the cellular pharmacology of ara-C and dCyd when these compounds arc administered at concentrations now known to be pharmacologically relevant [16]. In addition, the agcnts tetrahydrouridinc (THU) and deoxytetrahydrouridine (dTHU), inhibitors of cyti- dine deaminase [17] and dCMP deaminase [18] respectively, were employed to define the role of these enzymes in modulating this interaction. Our results suggest that the degree of deamination of dCyd by leukemic cells may have an important effect on its ability to antagonize ara-C metabolism and cytotoxicity.

MATERIALS AND METHODS

( 7~emicals

dCyd hydrochloride and ara-C hydrochloride were pur- chased from Sigma (St. Louis, MO). They were stored dessicated at 4°C and formulated in sterile medium prior to use. THU was provided by Nancita Lomax, Drug Synthesis and Chemistry Branch, Division of Cancer Treatment, National ('ancer Institute. dTItU was provided by Dr

Effect of deaminase inhibitors on ara-C and deoxycytidine metabolism 207

David Johns, Drug Synthesis and Chemistry Branch, National Cancer Institute. [3H]Ara-C (23 Ci/mmol) and [3H]dCyd (18 Ci/mmoi) were purchased from Amersham Radiochemicais (Arlington, ILl. Ara-CTP and dCTP were purchased from Sigma and stored dessicated at -70°C prior to formulation.

Cells HL-60 ceils were derived from the promyelocytic leu-

kemia cell line originally isolated by Collins et al. [19]. K- 562 cells were derived from the erythroleukemia cell line described by Lozzio and Lozzio [20]. Characterization of the myeloid leukemia cell line KG-1 and the lymphoblastic leukemia cell line CCRF-CEM have also been reported previously [21,22]. All cells were maintained in RPMI medium (GIBCO, Grand Island, NY) containing 1% sodium pyruvate, non-essential amino acids, and 15% heat inactivated bovine calf serum (Hyclone, Logen, UT) in a 37°C, 5% CO2, fully humidified incubator. Cells were passed twice weekly and evaluated monthly for myco- plasma contamination. Logarithmically growing cells (cell density <5 x 10 ~ cells/ml) were used in all expcrimcnts.

Ara-CTP and dCTP pools Cells were centrifuged and the cell pellet resuspended

in fresh medium containing I(1% FCS at a cell density of 5 x 106cells/ml. Aliquots (5ml) of the cell suspension were placed in 15 ml centrifuge tubes to which were added 10 laM ara-C along with either 20 or 50 laM dCyd in the presence or absence of I mM THU or 0.5 mM dTIIU. The tubes were then placed in the 37°C, 5% C()~ incubator for 4 h after which the cell suspension was centrifuged at 4(X) x g for 7 rain at 4°C, and the cell pellet washed twice with ice-cold PBS. The pellct was then precipitated with 100 ~tl of 6 N TCA, allowed to stand on ice for 5 min, and the acid-soluble material transferred to pre-cooled glass centrifuge tubes. The T( 'A was then extracted with frcon- octylamine as previously described [23]. Thc neutralized cell extracts were subjected to high prcssure liquid chro- matographic analysis utilizing a minor modification [15] of the method of l.ilicmark and Plunkett [24]. Ribo- nucleotides were first eliminated by pcriodation as initially described by Garrett and Santi [25]. Separations were performed utilizing a Bio-Rad Model 700 I IPLC system employing a Milliporc 10.u Partisil SAX Radial-Pak car- tridge system and a Beckman Model 160 UV detector. The system was run isocraticallv for 22 min at a flow rate of 3 ml/ rain with 75~5 buffer A [(0.005 M (NH4)3PO4, pi i 2.8] and 25% buffer B 10.75M (NII~)3PO 4, pH3.7). The per- centage of buffer B was then increased linearly to 100% over the next 38 min. The retention times of dCTP and ara-CTP utilizing this system were 27.5 and 31.2rain respectively. Absorbance was monitored at 280 nM and dNTP levels quantitated by determining peak areas uti- lizing the Model 70(1 automatic integration program. Com- parisons were made with the peak areas of known standards (all Sigma), and dNTP values expressed as pmol/10 ~ cells.

Ara-C incorporation into DNA The incorporation of [3H]ara-(" into leukemic cell DNA

under various conditions was determined by a minor modi- fication of a previously described method [15]. A total of 5 x 106 cclls were suspended in 2 ml of medium containing 10% FCS along with 10 laM [31l]ara-C, 20 or 5(3 p.M dCyd, and THU or dTt lU as in the previous section. The cells

were then incubated for 4 h at 37°C in the 5% CO2 incu- bator, after which the suspension was centrifuged at 400 x g for 7 min at 4°C and the pellet washed twice with cold PBS to remove radiolabeled drug. DNA was then isolated and purified by pronase and RNAse digestion, phenol extraction, and ethanol precipitation as previously described [15]. After centrifugation at 9000 × g for 30 min at 4°C, the DNA was redissolved in Tris-HCl buffer, quantitated spectrophotometrically, and radioactivity determined utilizing a Beckman Model 7200 scintillation counter. [3H]Ara-C DNA incorporation for each condition was then expressed as pmol ara-C/I.tg DNA.

Intracellular dCyd concentrations (_?ells wcre incubated in plastic centrifuge tubes as

described in the previous section (20 × 106 cells/condition) along with 50 p.M dCyd and either 1 mM THU or 0.5 mM dTHU. After 4 h incubation at 37°C, the cell suspension was centrifuged at 40(1 x g for 7 min at 4°C and the pellet immediately placed on ice. The cells were then rapidly washed twice with ice-cold PBS and the cell pellet pre- cipitated with 0.1 ml of cold 6 N TCA. The TCA was then extracted with freon-octylamine as described above, and the neutralized extracts subjected to HPLC analysis uti- lizing a C-18 Bondapak (10 p. Radial compression column (Waters) and an isocratic 0.005 M (NH~)3PO~, pt t 3.2 buffer system with a flow rate of 0.5 ml/min. Using this system, dCyd exhibited a peak retention time of 11.5 rain and was clearly separable from other nucleosides. Absorbance was monitored at 280 nm and intracellular dCvd quantitated by comparing peak areas with those of known standards. Average cell volume was determined with a Model ZM Coulter Counter (t t ialeah. Fla) equipped with a Coulter Channelyzer, and intracellular dCyd con- centrations calculated by dividing the number of pmol dCwt recovered from 10 '~ cells bv the volume of the cells.

En=vme assavs

Assays for cytidinc dcaminase and dCK activity from cell extracts were performed utilizing minor modifications of the methods of Stueart and Burke [61 and Ives and Durham [12] rcspectively. A description of thcsc methods has been rcported in detail previously [26].

Growth inhibition studies

Logarithmically growing cells were suspended in RPMI medium containing 15% bovinc calf serum at a cell density of 105 cells/ml. Twenty milliliter aliquots of the cell sus- pension were placed in plastic T-flasks to which were added various concentrations of ara-C, dCyd, and either 1 mM THU or 0.5 mM dTHU. The tlasks were placed in the 37~(. ", 5% ( 'O , incubator for 72 h. At 24 h intervals, aliquots of the cell suspension were removed and cell densities assayed utilizing the Model ZM Coulter counter. A sim- ultaneous determination of cell viability was performed by assessing the percentage of cells excluding trypan blue. The total number of viable cells was then calculated for each experimental condition and expressed as a percentage relative to values obtained for untreated controls.

Statistical analysis

The significance of differences between experimental groups was determined utilizing the Student's t-test for paired observations.

208 S. GRANT et al.

KG-1 100 ] [] Control 100

. | II D0'"o .

0 ii ra-CTP dCTP arll-CTP dCTP arll- CTP dCTP

dCyd (gM) -I~ 0 20 50 deyd (~M) _~

HL-60 ] Control

[ ~ T H U

] dTHU

arm-CTP dCTP ara-CTP (:ICTP ar~.CTP dCTP

0 20 ,50

100

-i l°, dCyd (p.M) _~

K-562

ars-CTP dCTP

0

100

] Control

] THU

] dTHU

e~ 10

u

o

o

E I

CCRF-CEM

II r l -CTP dCTP

0

ara-CTP dCTP sr l i -c ' r P dCTP

20 50 dCyd (~M) _~.

FIG. 2. Cells were incubated for 4 h in the presence of the designated concentrations of ara-C and dCvd and either THU (1 mM) or dTHU (0.5 raM). Intracel(ular levels of ara-CTP and dCl-P were assayed by HPLC and expressed as pmol/' 10" cells. Values represent the means for duplicate determinations obtained in at least four separate experi-

ments ± 1 S.D.

] Co ntrol

~ THU

] dTHU

Jra-CTP dCTP i ra-CTP dCTP

20 50

RESULTS

The effects of exogenous dCyd and deaminase inhibitors on intracellular ara-CTP and dCTP pools in the four cell lines are displayed in the bar graphs in Fig. 2. The 4 h intracellular accumulation of ara- CTP ranged from 41 to 58 pmol/10 ~' cells. In three of the cell lines (HL-60, K562, and CCRF-CEM), co- administration of 20 or 50 gM dCyd reduced ara- CTP formation by 92-99.3%. In all cases, 50.aM dCyd produced greater reductions in ara-CTP for- mation than 20 gM dCyd. Increments in dCTP pools following exposure to 50.aM dCyd ranged from 84 to 120% in these cells. The most notable finding was that the response of KG-l cells to exogenous dCyd was distinctly different from that of the other cell lines in two major respects. First, co-administration of 20 or 50 gM dCyd alone resulted in only minor reductions in ara-CTP formation (10 and 38% respectively), unlike the significantly larger

reductions noted in the other cell lines. Second, KG- l cells were also unique in that they were the only cells that did not display a statistically significant increase in dCTP pools in response to 50 uM dCyd (p > 0.05). As a consequence, the reductions in ara- CTP/dCTP ratios in KG-I cells exposed to 211 and 50gM dCyd (27 and 55% respectively) were sub- stantially less than those observed in the other cell lines (e.g. ->98%" data not shown).

Addition of either THU or dTHU to cells exposed to 10 aM ara-C (but not dCyd) resulted in a parallel increase in both ara-CTP and dCTP pools in all cell lines, and little change in the ara-CTP/dCTP ratio (not shown). When dCyd was present, however, a major difference was observed between the response of KG-I cells and the other cell lines to these agents. While dCyd alone had a minor inhibitory effect on ara-CTP formation in KG-I cells, co-administration of THU or dTHU resulted in substantial reductions in ara-CTP accumulation (e.g. 94.7 and 98.1%

Effect of deaminase inhihitors on ara-C and deoxycytidine metabolism 209

100 ] KG-1 100 g [] c o . , t o , ._g - I

= 10 _c ~ 10

, .

'., =

1 < 1 0 2 0 5 0

dCyd (pM)

H L - 6 0 [] Control

0 2 0 5 0

dCyd (~M)

100

E Z~ ~ I0

(.~ ~

K - 5 6 2 100'

. co., ,o, . . o .°

~'J dTHU o O

-= 10"

< 1 0 2 0 5 0

dCyd (~M)

CCRF-CEM

~ Contro [ I THU [7} dTHU

0 2 0 5 0

dCyd (pM)

FIG. 3. Cells were incubated for 4 h in the presence of the designated concentrations of dCvd and 1(I [aM {~ll]ara-(" along with either 1 mM THU or 0.5 mM dTHU. Leukemic cell DNA was then isolated, and incorporation of ara-(" into DNA quantitated and expressed as pmol ara-C/ug DNA. Values represent the means for duplicate deter- minations obtained from at least four separate experi-

ments-+ 1 S.D.

respectively for 20 and 50~M dCyd). These reductions were similar to those observed in the other cell lines in the absence of THU and dTHU. The deaminase inhibitors also produced statistically significant increments in intracellular dCTP pools in KG-1 cells exposed to dCyd, whereas no increases were observed in the absence of these agents. For example, dCTP pools in KG-1 cells exposed to 50 ~tM dCyd increased from 4 . 3 - 0.8 to 8.4-+ 1.4pmol/ 106cells when 0.SmM dTHU was present (p < 0.05). These findings suggest that deamination of exogenous dCyd by KG-I cells limits both the expansion of dCTP pools and the resulting antag- onism of ara-C phosphorylation.

The pattern of antagonism of ara-C incorporation into DNA by dCyd in the four cell lines is illustrated by the data presented in Fig. 3. In general, each dCyd concentration produced a greater relative decrease in ara-CTP levels than in ara-C DNA incorporation. For example, 50 .LtM dCyd reduced ara-CTP levels in HL-60 cells by 99.5% but DNA incorporation by only 93.2%. As in the previous example, KG-1 cells

exhibited a unique pattern of response both to dCyd and the deaminase inhibitors. For example, KG-I cells exposed to 20 and 50 [~M dCyd experienced only 52 and 60% reductions in ara-C DNA incorporation respectively. In contrast, reduction in values for the other cell lines ranged from 76 to 87% (2(1 .uM dCyd) and 87 to 93% (50 ~tM dCyd). These reductions were significantly larger than those observed in KG-1 cells (p < 0.05). As in the case of ara-CTP formation, co- administration of THU or dTHU permitted greater reductions in KG-1 ara-C incorporation into DNA at each dCvd concentration. For example, exposure of KG-I cells to 50 ~tM dCyd in conjunction with THU or dTHU resulted in 91 and 93% reductions in ara-C DNA incorporation respectively. In contrast, co-administration of THU and dTHU with either 20 or 5(1 p,M dCyd did not result in st~:tistically significant changes in incorporation of ara-C into DNA in any of the other cell lines.

Since dCyd antagonizes ara-C phosphorylation both directly (via competition at the level of dCK) and indirectly (through feedback inhibition of dCK

210 S. GRANI et al.

1

' 5 ; 2 ,

i

i i

• ' C3NT~OL

~ " U J i l C 5 M '

9,~, -- • d ' - d :bx lO 4M~

1"5 . i

f) " KG -I ~( 6 0 ~ - b 6 2 (:CR~ - C E M

FrO. 4. Cells were incubated for 4 h with 50 ~tM dCyd in the presence or absence of the designated concentrations of THU or dTHU. At the end of the incubation period, the cells were pelleted, a neutralized acid-soluble extract obtained, and intracellular dCyd concentrations deter- mined by HPLC as described in the tcxt. Values reprcsent the means for duplicate determinations performed on at

Icast three separatc occasions -+ I S.D.

by expanded dCTP pools), intracellular concen- trations of dCyd were determined in the four cell lines in the presence and absence of THU and dTHU (Fig. 4). A 4 h exposure to 50 gM dCyd produced intracellular dCyd concentrations of 40-50 [aM in HL- 60, K-562, and CCRF-CEM cells; these values did not differ significantly (p > 0.05). In contrast, the intracellular concentration of dCyd in KG-1 cells was only 20 pM, which was significantly less than values obtained for the other cell lines (p < 0.05). Co- administration of THU and dTHU increased dCyd concentrations in KG-1 cells to levels observed in the other lines. In contrast, THU and dTHU had negligible effects on dCyd concentrations in the other cells.

Activities of the enzymes cytidinc deaminase and deoxycytidine kinase were compared in the four cell lines and the results are shown in Table 1. Levels of cytidine deaminase activity were found to be con- siderably lower than values reported for more mature granulocytes [27], but similar to those prcviously obtained for a variety of continuously cultured human leukemic cell lines [28]. Cytidine deaminase activity was nearly three times greater in KG-I cells than in any of the other cell lines. In addition, the

ratio of dCK to cytidine deaminase kinase activity (K/D ratio) was only 10-30% of the values for the other cell types. This finding is consistent with the relative inability of KG-1 cells to expand intracellular dCTP pools in response to exogenous dCyd.

To determine whether the biochemical differences displayed by KG-1 cells might be reflected in their response to ara-C and dCyd, suspension culture growth studies were performed. A comparison between the response of KG-I and HL-60 cells to various agents is illustrated in Fig. 5. Both lines exhibitcd an approximately 90% reduction in the number of viable cells surviving at the end of the 72 h incubation period in thc presence of ara-C alone. In ttL-60 cclls, coadministration of 20[aM dCyd restored viability to nearly 50%, whilc addition of THU or dTHU did not appreciably increase survival further. In contrast, 21) [aM dCyd cxhibited a minimal capacity to reverse ara-C-mediated cytotoxicity in KG-I cells. However. co-administration of TIIU or dTHU led to substantial reversal of ara-C-mcdiated effects, resulting in a survival fraction similar to that observed in HE-60 cells. In separatc studies, K-562 and CCRF-CEM cells displayed survival patterns similar to thosc displayed by HL-60 cells (data not shown). These findings suggest that the biochcmical changes induced by THU and dTHU in KG-1 cclls (e.g. increased intraceilular dCyd and dCTP accumu- lation) result in an enhanced ability to antagonize ara-C inhibitory cffects.

DISCUSSION

The complexity of the interaction between dCyd and ara-C in leukemic cells makes it difficult to pre- dict a p r i o r i what effect a particular dCyd con- centration will have on ara-C metabolism. Factors such as competition at the level of transport, phos- phorylation, deamination, and incorporation into DNA all play potentially important roles in deter- mining the degree of antagonism. In addition, Plun- kett and co-workers have demonstrated that perturbations in intracellular dCTP pools may have a profound effect on ara-C phosphorylation as a consequence of feedback inhibition of the enzyme catalyzing the rate-limiting step in ara-C metabolism, dCK [24]. Previous studies designed to characterize the interaction of dCyd and ara-C in leukemic cells have generally employed low concentrations of dCyd (e.g. 1 [aM or less) [13] to reflect the low con- centration of this nucleoside present in the blood [29]. Consequently, little information exists con- cerning the factors influencing this interaction when dCyd is present at the high concentrations (e.g.

Effect of deaminase inhibitors on a r a - C a n d d e o x y c y t i d i n e m e t a b o l i s m

T A B L E 1. C Y T I D I N E DEAMINASE AND d C K A(SrlVITY IN HUMAN LEUKEMIA ( 'ELLS

211

Cytidine deaminase dCK Cell (nmol dUrd/h/mg protein) (nmol dCMP/h/mg protein) K/D (× 10 ~)

KG-1 7.3 -+ 1.2 0.22 -+ 0.02 3(1 HL-60 2.5 ± 0.7 1.14 '- 0.9 456 K-562 1.4 ± 0.4 0.77 ± 0.6 55(1 CCRF-CEM 2.8 ± 0.6 0.6o _+_ 0.5 214

Cell extracts were obtained as described in the text and dCK assayed. Values represent the means for triplicate determinations in experiments performed on at least three separate occasions ± 1 S.D.

I0 9 8

N 7 -3 6

" ' 5

_3 0 4

0 3 -

m 2 _<

J

uJ 8 ~ 7

7- 6 -

~ 5 - Z _o 4 -

b c~ 3 - u_

~ 2 r r

(/3

I - - T i i

24 48 72

HOURS

FIG. 5. Logarithmically growing cells were seeded at 10~cells/ml in thc presence of 10 ~.M ara-C and 2(1 aM dCyd ± 10 ) M TItU or 5 x 10 4 M dTHU. At 24 h inter- vals, cell viability and cell density determinations were made. Values for each condition are expressed as a survival fraction (relative to untreated controls). Values represent the means for at least three separate experiments per- formed in duplicate +- I S.D. (C)~D) ara-C; (B---II) ara- C + dCyd: (A---A) ara-C + dCyd + THU: (V- -V) ara-

(" + dCyd + dTHU.

->20 ~tM) that arc now achievable in the plasma of humans [16].

The current studies suggest that the intraceilular deamination of dCyd may play an important role in

modulating the ability of this nucleoside to anta- gonize ara-C metabolism in leukemic cells. This hypothesis is supported by the finding that KG-1 cells, characterized by higher levels of cytidine deaminase and a lower K/D ratio than the other cell lines, exhibited the smallest increments in intra- cellular dCyd and dCTP levels in response to exogen- ous dCyd. The relatively small increases in these intraccllular metabolites permitted extensive phos- phorylation of ara-C as well as incorporation of ara- C into DNA, despite high extracellular concen- trations of dCyd. In contrast, increments in both intracellular dCyd and dCTP levels were con- siderably larger in the other cell lines, and the reduction in ara-C nucleotidc formation and DNA incorporation correspondingly greater. Since dCyd exhibits greater affinity for dCK than ara-C [3(I], and dCK is strongly inhibited by elevations in dCTP levels [12], both factors are likely to contribute to the substantial antagonism of ara-C metabolism in these cells. Although we cannot ascertain from these studies whether the direct effect (i.e. competition at the level of dCK) or the indirect effect (feedback inhibition by dCTP) predominates, it is noteworthy that Lilliemark has reported a Ki for dCTP with respect to dCK of 5.9 btM [24]. The observation that 50.uM dCvd generally led to greater inhibition of ara-C metabolism than 20 ~tM dCyd (despite the lack of further increases in dCTP pools) suggests that above certain dCvd concentrations, direct com- petition prevails.

Since THU and dTHU were equally effective in enhancing the ability of dCyd to antagonize ara-C metabolism in KG-1 cells, a single mechanism might be sufficient to account for their effects in our system. While THU selectively inhibits the enzyme cvtidinc dcaminasc [17], dTHU is also a potent inhibitor of dCMP dcaminasc [18]. Mancini & Cheng have reported that ara-CMP is a poor substratc for enzyme isolated from human leukemic myeloblasts and sug- gest that this enzyme plays a minor role in ara-C metabolism in such cells [31]. In contrast, Ellims et al. 132] and Fridland and Verhocf [9] have shown

212 S. GRANT et al.

that dTHU potentiates ara-C metabolism in T-lym- phoblastoid cell lines such as CCRF-CEM. The pres- ence of high concentrations of dCyd makes analysis of these interactions considerably more complex, since the deaminase inhibitors could exert differential effects on dCyd metabolism. However, in our system T t tU and dTHU produced similar alterations in dCyd (and ara-C) metabolism in KG-I cells, and it is therefore possible that inhibition of cytidine deaminase might account for the observed changes. Whether T H U and dTHU also exert direct effects on the intracellular metabolism of ara-C in the pres- ence of high concentrations of dCyd, and whether inhibition of dCMP deaminase is of independent significance, cannot presently be determined.

The possibility that dCyd may antagonize ara-C metabolism by interfering with ara-C transport was not directly addressed in our studies, but is unlikely to have played a major role. White et al. have previously demonstrated that while transport may be rate lim- iting for ara-C phosphorylation at low concentrations (e.g. <1 uM), its importance declines dramatically at ara-C concentration of I0 uM or greater [33]. This group has reported that dCyd concentrations as high as 50 uM had a negligible effect on the transport of higher concentrations of ara-(" in Ehrlich cells [34]. In the present studies, high concentrations of THU and dTHU, which share thc facilitated nucleoside diffusion carrier, did not inhibit ara-C metabolism. It is therefore likely that the inhibitory effects of dCyd on ara-C metabolism result from interference with ara-C phosphorylation by dCK, and do not involve transport related factors.

The finding that the deaminase inhibitors THU and dTHU preferentially enhance dCyd protective effects in cells displaying higher levels of cytidine deaminase activity has implications for attempts to improve the therapeutic index of the ara-C/dCyd regimen. Most patient-derived leukemic myelo- blasts, like the majority of myeloid leukemic cell lines [28], exhibit low levels of cvtidine deaminase activity [8]. In contrast, several target host tissues, such as gastrointestinal epithelium, contain high levels of this enzyme [7]. THU or dTI-|U might therefore selectively enhance the protective effects of dCyd in these normal tissues, without producing a parallel reduction in ara-C antileukemic efficacy. Recent clinical studies have now established that plasma dCyd levels in excess of 20uM can be achieved in leukemic patients, and that such levels substantially ameliorate the non-hematologic toxicity of continuously administered high-dose ara-C with- out eliminating antileukemic effects 116]. In view of the possibility that THU or dTHU might permit even higher ara-C doses to be safely administered under

these conditions, without compromising antileu- kemic efficacy, animal studies designed to test this hypothesis appear warranted.

Acknowledgements--We gratefully acknowledge the expert technical assistance of Amy Turner in the per- formance of these studies. We appreciate the excellent secretarial assistance of Mrs Fran Hamilton. This work was supported by Award 2-RO1-CA-35601 from the NIH. Portions of this work have been presented in preliminary form at the American Association of Cancer Research, San Francisco, ('A. 1989.

REF EREN CES

I. Ellison R. R., I[olland J. 1-., Weill M., Jacquillat C., Boiron M., Bernard J., Savitsky A., Rosner E., Sussoff B., Sib,'crs R. T.. Karanas A.. Cuttner J., Spurr C., flares D. M., Biota J., Leone L. A., Ilaurani F., Kylc R., llutchison J.. Forcier R. J. & Moon J. II. (1968) Arabinofuranosylcytosinc: a useful agent in the treat- ment of acute leukemia in adults. Blood 32, 507.

2. Plagemann G. G. W., Marz R. & Wohlhueter R. M. (1978) Transport and metabolism of deoxycytidine and l-fi-l) arabinofuranosylcytosine into cultured Norvikoff rat hepatoma cells, relationship to phosphorylation, and regulation of triphosphate synthesis. ("ancer Res. 38. 978.

3. Momparler R. L., Brent T. I'., Labitan A. & Krygier V. (1971) Studies on the phosphorylation of cytosine arabinoside in mammalian cells. Mol. Pharmac. 8, 362.

4. Furth J. & Cohen S. (1968) Inhibition of mammalian DNA polymerase by the 5'-triphosphate of 9-fl-D ara- binofuranosylcytosine and the 5'-triphosphate of 9-fl- arabinofuranosyladenine. Cancer Res. 28, 2(161.

5. Kufe D. W., Major P. P., Egan E. M. & Beardsley G. P. (1980) Correlation of cytotoxicity with incorporation of ara-C into DNA. J. biol. Chem. 225, 8997.

t~. Stcuart ('. D. & Burke P. J. (1971) ('vtidine deaminase and the development of resistance to arabinosyl cyto- sine. Nature New Biol. 233, 102.

7. Camiener G. W. & Smith ('. G. (1965) Studies of the enzymatic deamination of cytosine arabinoside. I. Enzyme distribution and species specificity. Biochem. Pharmac. 14, 1405.

8. Meier J. & Nygaard P. (1978) Cytosine arabinoside phosphory[ation and dcamination in acute myeloblastic leukemia, l,eukemia Re.s. 2, 127.

9. Fridland A. & Verhoef V. (1987) Mechanism tk)r ara- CI'P catabolism in human leukemic cells and effect of deaminase inhibitors on this process. Sere. Oncol. 14 (Suppl. 1), 262.

10. Biro U. & Goldenberg D. M. (1971) Hematologic toxicity of I-/~-D arabinofuranosylcytosine, and its pre- vention by deoxycytidine in the mouse. Chemotherapy 16. 29.

11. Chu M. Y. & Fischer G. A. (1968) A proposed mech- anism of action of 1-/3-o arabinofuranosylcytosine as an inhibitor of the growth of leukemic cells. Biochem. Pharmac. 11,423.

12. Ives D. H. & Durham J. P. (1970) Deoxycytidine

Effect of deaminase inhibitors on ara-C and dcoxycytidine metabolism 213

kinase. IIl: Kinetic and allosteric regulation of the calf thymus enzyme. J. biol. Chem. 245, 2285.

13. Harris A. L. & Grahame-Smith D. G. (1981) Cytosine arabinoside triphosphate production in human leu- kemic myeloblasts: interactions with deoxycytidine. Cancer Chemother. Pharmac. 5, 185.

14. Buchman V. M., Belyanchikova N. I., Mkheidze D. M., Litovchenko T. A., Lichinitser M. R., Barkotkina M. F. & Svet-Moldavsky G. J. (1979) 2'-Deoxycytidine hydrochloride protection of mice against the lethal toxicity of cytosine arabinoside. Cancer Chemother. Pharmac. 3, 229.

15. Bhalla K., MacLaughlin W., Cole J., Arlin Z., Baker M.. Graham G. & Grant S. (19871 Deoxycytidine preferentially protects normal versus leukcmic myeloid progenitor cells from cytosine arabinoside mediated cytotoxicity. Blood 70, 568.

16. Bhalla K., Birkhofer M., Baker M., Cole J., Lutzyky J. & Grant S. ( 19891 Phase I and pharmacologic study of continuous infusion high-dose ara-C and deoxv- cytidine in refractory leukemia. Blood (Suppl.) 74,579.

17. Stoller R. G., Myers C. F. & Chabner B. A. (1978) Analysis of cytidine deaminase and tetrahydrouridinc interaction by the use of ligand techniques. Biochem. Pharmac. 27, 53.

18. Maley F. & Maley (;. F. (1971) Tetrahvdro- deoxyuridylate: a potent inhibitor of deoxycyticlylate deaminase. Archs. Biochem. Biophys. 144, 723.

19. Collins S. J., Gallo R. C. & Gallegher R. E. (1977) Continuous growth and differentiation of human myeloid leukemic cells in suspension culture. Nature. Lond. 270, 347.

20. Lozzio C. B. & Lozzio B. B. (1975) tluman chronic myelogenous leukemia cell line with positive Phi- ladelphia chromosome. Blood 45, 321.

21. Koeffler H. P. & Golde D. W. (19781 Acute mve- logenous leukemia: a human cell line responsive to colony stimulating activity. Science 200, 1153.

22. Foley G. E., Lazarus H., Farber S.. Uzman B. G., Boone B. A. & McCarthy R. E. (1965) Continuous culture of human lymphoblasts from peripheral blood from a child with leukemia. Cancer. Philadelphia 18, 552.

23. Khym J. X. (1975) An analytic system for the rapid separation of tissue nucleotidcs at low pressure on conventional ion exchangers. Clin. Chem. 21, 1245.

24. Liliemark J. & Plunkett W. (1986) Regulation of l-fl- D arabinofuranosylcytosine-5'-triphosphate accumu- lation in human leukemia cells by deoxycytidine 5'- triphosphate. Cancer Res. 46, 1079.

25. Garrett C. & Santi D. (1979) A rapid and sensitive high pressure liquid chromatographic for deoxy- ribonucleoside triphosphates in cell extracts. Analvt. Biochem. 99, 268.

26. Bhalla K.. Nayak R. & Grant S. (19841 Isolation and characterization of deoxycytidine kinase deficient, human promyelocytic leukemic cell line highly resistant to 1-fl-t) arabinofuranocylcytosine. Cancer Res. 44, 5505.

27. Chabner B., Johns D., Coleman C.. Drake J. C. & F, vans W. tt. (1974) Purilication and properties of cytidine deaminase from normal and leukemic gra- nulocytes. J. olin. lnoest. 53, 922.

28. Nygard P. & l,undstrom C. (1987) Low cytidine deami- nase levels in human hematopoietic cell lines. Leu- kemia Res. !1,681.

29. Danhauser k. k. & Rustum Y. M. (198(I) Effect of thymidinc on the toxicity, antitumor activity and metabolism of I-/3-D arabinofuranosylcytosine in rats bearing a chemically induced colon carcinoma. Cancer Res. 40, 1274.

30. Coleman C. N., Stoller R. G., Drake J. C. & Chabner B. (19751 Deoxycytidine kinase: properties of the enzyme from human leukemic granulocytes. Blood 46, 791.

31. Mancini W. R. & Cheng Y.-C. (1982) tluman deoxy- cvtidvlate deaminase: substrate and regulator speci- ficities and their chemotherapeutic implications. Mol. Pharmac. 23, 159.

32. FAlims P. 1t., Kao A. Y. & ( 'habner B. A. (1983) Kinetic behavior and allosteric regulation of human deoxvcvtidvlatc deaminase derived from leukemic cells." ,(4ole'c. (?ell Biochem. 57, 185.

33. Whitc J. C., Rathmell J. P. & Capizzi R. L. (1987) Membrane transport influences the rate of accumu- lation of cytosine arabinoside in human leukemia cells. J. olin. lmwst. 79, 38(I.

34. White J. C., Ilines L. H. & Rathmell J. P. (1985) Inhibition of 1-/:I-D arabinofuranosylcytosine transport and net accumulation by teniposide and ctoposide in Ehrlich aseitcs cells and human leukemia blasts. Cancer Res. 45, 3070.