Communication Vol. 267 No. 5 Issue of February 15 p 2856-2859 1992 THE JOURNAL OF BIOLOGICAL CHEMISTRY

0 1992 by The American Society fo; Biochemistry and Mbf&ar Biolo;, Inc. Printed in U.S.A.

The Role of Cytoplasmic Deoxycytidine Kinase in the Mitochondrial Effects of the Anti- human Immunodeficiency Virus Compound, 2’,3’- Dideoxycytidine”

(Received for publication, August 22,1991) Chin-Ho Chen and Yung-Chi ChengS From the Department of Pharmwobgy, Yak University School of Medicine, New Haven, Connecticut 06510

2’,3’-Dideoxycytidine (ddC) is a potent inhibitor of human immunodeficiency virus replication in vitro and shows beneficial effects in AIDS therapy. The compound inhibits mitochondrial DNA (mtDNA) syn- thesis at a clinically relevant concentration, which could be responsible for the side effects of ddC observed in the clinic. Thymidine (dThd), one of the substrates of mitochondrial deoxypyrimidine kinase (dPyd ki- nase), was not able to reverse the mitochondrial tox- icity of ddC in CEM cells. Furthermore, the cyto- plasmic deoxycytidine kinase (dCyd kinase)-deficient CEM cells were highly resistant to the mitochondrial toxicity of ddC. These data suggest a critical role for cytoplasmic dCyd kinase in the mitochondrial toxicity of ddC. The metabolites of ddC, but not ddC itself, were able to inhibit mtDNA synthesis in isolated mitochon- dria. The potency of the inhibitory effect was in the order of ddCTP > ddCDP > ddCMP > ddC. The lack of inhibition by ddC of mtDNA synthesis could be due to the inefficient ddC phosphorylation in mitochondria. Although the mitochondrial dPyd kinase was reported to phosphorylate ddC, the phosphorylation of ddC in isolated mitochondria was not detectable. The data suggest that ddC is phosphorylated to ddCTP in the cytoplasm and then transported into mitochondria to exert its inhibitory effect on mtDNA synthesis.

Human immunodeficiency virus (HIV)’ is believed to be the etiologic agent of acquired immunodeficiency syndrome (AIDS). Dideoxynucleoside analogs were shown to inhibit HIV replication (1). 2’,3’-Dideoxycytidine (ddC) is one of the most potent HIV inhibitors among the nucleoside analogs tested. This compound showed beneficial effects in AIDS patients during clinical trials (2). However, ddC treatment caused toxicity in vivo and in vitro. Patients who received long term ddC treatment developed a painful peripheral neu- ropathy (3). In cell culture, ddC was acutely toxic to CEM

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed. The abbreviations used are: HIV, human immunodeficiency virus;

ddC, 2’,3’-dideoxycytidine; dThd, thymidine; dCyd, deoxycytidine; ddCMP, 2’,3’-dideoxycytidine monophospate; ddCDP, 2’,3’-dideox- ycytidine diphosphate; ddCTP, 2’,3‘-dideoxycytidine triphosphate; dPyd, deoxypyrimidine; HPLC, high pressure liquid chromatography.

cells at high concentrations, and this may be due to the incorporation of ddC into nuclear DNA (4). In addition, the compound was shown to deplete mtDNA at very low concen- trations and induced delayed cytotoxicity in cell cultures (5). The depletion of mtDNA could be responsible for the delayed toxicity of ddC observed in clinic.

ddC exerts biological effects through its metabolite dideox- ycytidine triphosphate (ddCTP) (6). ddCTP is a very potent inhibitor of DNA polymerase y which is believed to be re- sponsible for mtDNA synthesis (7). Both cytoplasmic deoxy- cytidine (dCyd) kinase and mitochondrial deoxypyrimidine (dPyd) kinase can phosphorylate ddC to ddCMP (7), which is then further metabolized to ddCTP. In addition to phos- phorylating dCyd, mitochondrial dPyd kinase also catalyzes the formation of thymidine monophosphate (TMP) from thymidine (dThd) (8). It is known that dThd and dCyd are competitive substrates for the mitochondrial dPyd kinase and that the affinity of the enzyme for either dThd or dCyd is higher than for ddC (7). In contrast, the cytoplasmic dCyd kinase does not have the thymidine kinase activity. In view of the fact that ddC rather than ddCTP is likely to get into mitochondria, the mitochondrial dPyd kinase could be re- sponsible for the ddCMP formation in mitochondria. Thus, blocking the process that transforms ddC to ddCTP should protect the cell from the mitochondrial effect of ddC. If the mitochondrial kinases, rather than cytoplasmic kinases, are the major enzymes responsible for the mitochondrial ddCTP pool, dThd could have the potential to block the phosphoryl- ation of ddC in mitochondria and prevent the mitochondrial toxicity without compromising the anti-HIV activity of ddC. However, the results presented in this paper indicate that the cytoplasmic dCyd kinase is the major enzyme that phospho- rylates ddC. Thus, the phosphorylated ddC appears to be transported into mitochondria to exert its inhibitory effect on mtDNA synthesis. Consequently, dThd was not effective in antagonizing the ddC mitochondrial toxicity in CEM cells.

EXPERIMENTAL PROCEDURES

Cell Culture-CEM cells, a lymphoblastoid cell line, were main- tained in RPMI 1640 medium supplemented with 5% fetal bovine serum. CEM/araC is a cytoplasmic dCyd kinase-deficient CEM line (a generous gift from Dr. A. Frieland, St. Jude Children’s Research Hospital). The cells were cultured under the same conditions as CEM cells.

Materials-ddC, ddCDP, and ddCTP were purchased from Phar- macia LKB Biotechnology Inc. ddCDP choline was a gift from Dr. D. Johns of the National Cancer Institute. ddCMP was synthesized by the procedure previously described (9). [3H]ddC (48 Ci/mmol), [3H]dCyd (18 Ci/mmol), and [3H]dATP was obtained from Moravek Biochemicals, Inc. Deoxynucleotides were from Sigma.

Determination of mtDNA Content-A procedure modified from Bresser et al. (10) was used to immobilize DNA from whole cells. Briefly, CEM cells (5 X lo4 for each sample) were collected and were freeze-thawed three times. The cell lysate was treated with RNase (10 pg/ml) a t 37 “C for 30 min, and then Proteinase K (100 pg/ml) was added to the cell lysate. The samples were incubated at 50 “C for 1 h, and then 0.8 volume of supersaturated NaI (2.5 g of NaI in 1 ml of hot water) was added. The cell lysate was heated at 90 “C for 5-10 min, and the DNA was immobilized on nitrocellulose paper by using a slot blot apparatus. The mtDNA on the nitrocellulose paper was detected with a mtDNA-specific probe described previously (5). The mtDNA probe spans from nucleotide 13,370 to 14,285 in the complete mitochondrial genome. The method used to estimate the mtDNA amount in total cellular DNA was previously described (5).

2856

Role of Cytoplasmic Deoxycytidine Kinase 2857

Assay of Delayed Cytotoxicity-CEM or CEM/araC cells were treated with different concentrations of ddC as indicated. The culture medium and ddC were refreshed at day 4. Cell number was estimated by a Coulter counter.

DNA Synthesis in Isolated Mitochondria-Mitochondria of CEM or CEM/araC cells were prepared by the “two-step” procedure de- scribed by Bogenhagen and Clayton (11). Incorporation of [3H]dATP into mtDNA was performed as described.jl2). The concentration of dTTP, dGTP, and dCTP used in this assay was 1 p~ and that of [3H]dATP was 0.3 pM (22 Ci/mmol). Mitochondria (0.1 mgof protein) in a total volume of 0.2 ml was used for each assay. The reaction was carried out a t 37 “C for 40 min and terminated by the addition of 1 N perchloric acid. The acid-insoluble pellet was washed three times with 1 N perchloric acid and resuspended in 20 pl of 10 mM Tris-HC1, pH 8.0,l mM EDTA, and 0.5% sarcosyl. The suspension was neutralized with 0.1 N KOH, which brought the suspension into a homogeneous solution. The radioactivity in each sample was measured by using a liquid scintillation counter.

Metabolism of Deoxycytidine or ddC in Isolated Mitochondria- Purified mitochondria were incubated with 10 pCi of [“C]dCyd or [3H]ddC at 37 “C for 30 min. The reaction condition used in this assay was reported previously by Mitra and Bernstein (13). The reaction mixture contained 0.1 p~ dCyd or ddC, 10 mM succinate, 2 mM ATP, 2 mM pyruvate, 1 mM malate, 2 mM nicotinic acid, 10 mM MgC12,2 mM KCl, 10 mM KHP04, pH 7.0, and 25 mM Tris-HCI, pH 8, in a total volume of 5 ml. Mitochondria (600 pg of protein) were added to start the reaction. After incubation, the mitochondria were centrifuged at 15,000 X g for 5 min. The mitochondria were washed three times with an ice-cold buffer containing 0.25 mM sucrose, 1 mM EDTA, and 1 p~ dCyd or ddC by microcentrifuge centrifugation. The acid-soluble materials from the mitochondria were extracted with 1 N perchloric acid and analyzed with anionic HPLC following the procedure previously described (5).

RESULTS

Protective Effect of Normal Nucleosides on the Delayed Cytotoxicity of ddC-Exposure of CEM cells to 0.2 PM ddC resulted in a typical delayed cytotoxicity (Fig. la). The drug did not affect cell growth within the first 3 days. However, the cell growth was retarded between day 3 and day 6 after ddC treatment. Several nucleosides were tested for their abil- ity to antagonize the ddC mitochondrial toxicity. The mito- chondrial toxicity refers to the delayed cytotoxicity and the loss of mtDNA caused by ddC (5). Some of the nucleosides tested in this study, dCyd, dThd, dUrd, and dGuo, were toxic to CEM cells at a concentration of 20 PM but not at 2 PM. Thus, the nucleosides were examined for their potential to protect the CEM cells from ddC mitochondrial toxicity at 2 PM. Only dCyd was able to reverse the delayed cell growth inhibition caused by ddC. Other nucleosides did not protect CEM cells from the delayed growth inhibition of ddC (Table I). The mtDNA of CEM cells was severely depleted after 6 days of ddC treatment (Fig. 2). Among the tested nucleosides, only dCyd was able to partially protect the mtDNA depletion caused by ddC treatment. dCyd at 2 PM could protect 30% of mtDNA from depletion by ddC at day 6. These data indicate that the cells do not require 100% of mtDNA to have a normal growth rate, and this is consistent with our previous obser- vations (5).

Resistance of CEMlaraC Cells to the Delayed Cytotoxicity of ddC-To examine the role of cytoplasmic dCyd kinase in the mitochondrial toxicity of ddC, a cytoplasmic dCyd kinase- deficient CEM cell line, CEM/araC, was exposed to ddC. The mtDNA of CEM cells was severely depleted after exposure to 0.05 PM ddC for 4 days (5). Under the same conditions, the mtDNA content of CEM/araC cells was not altered until the concentration of ddC reached 10 PM (Fig. lb). The total cellular DNA content was not significantly affected by the ddC treatment, which was shown by ethidium bromide stain- ing of the total cellular DNA in each lane of Fig. 1, b and c (data not shown). In addition, CEM/araC cells were approx-

a.

A - a W

i - Y

3 S ..

b.

C.

10 0 2 4 6 8 10

ddc(pM) 0 1 10 loo

1 2 3 4

“mtDNA

FIG. 1. Resistance of the CEM/araC cells to ddC. a, CEM cells or CEM/araC (dck) cells (2 X lo4 cells/ml) were treated with different concentrations of ddC as indicated for 8 days. The doubling time of the cells was estimated every 2 days. The medium and ddC were changed at day 4, and cells were resuspended in a density of 2 X lo4 cells/ml. b, the mtDNA level of CEM/araC cells in the presence of different concentrations of ddC. Total cellular DNA was extracted from CEM/araC cells after 4 days of ddC treatment. The mtDNA contents were analyzed by Southern blot hybridization using a mtDNA fragment as probe (see “Experimental Procedures”). c, CEM or CEM/araC cells were treated with ddC for 6 days. The mtDNA content of CEM or araC cells was determined as described in b. Lane 1, CEM cells without ddC treatment; lane 2, CEM cells treated with 0.2 p~ ddC; lane 3, CEM/araC cells without ddC treatment; lane 4, CEM/araC cells with 0.2 p~ ddC treatment.

imately 1000-fold more resistant to the delayed cytotoxicity of ddC than CEM cells (Fig. la).

Effect of ddC and Its Metabolites on mtDNA Synthesis-To determine whether exogenous ddCTP could exert an inhibi- tory effect on mtDNA synthesis, ddCTP was tested for its effect on isolated mitochondria. ddCTP inhibited mtDNA synthesis in a dose-dependent manner. In contrast, ddC did not affect mtDNA synthesis under the same experimental conditions (Fig. 3a). Other ddC metabolites were also tested for their effect on mtDNA synthesis in isolated mitochondria. The potency of the ddC metabolites on mtDNA synthesis is in the order of ddCTP > ddCDP > ddCMP > ddC = ddCDP choline (Fig. 3b). Aphidicholine (25 pg/ml), a specific inhibitor of DNA polymerase a, was used as a control to demonstrate the lack of nuclear DNA synthesis under the assay conditions

The Metabolism of ddC in Isolated Mitochondria-Although mitochondrial dPyd kinase has been shown to catalyze the formation of ddCMP from ddC (7), whether ddC is taken up and phosphorylated by isolated mitochondria is unknown. HPLC analysis of ddC metabolites formed in isolated mito- chondria of CEM cells revealed no detectable ddCMP,

(14).

2858 Role of Cytoplasmic Deoxycytidine Kinase TABLE I

Effect of nucleosides on ddC (0.2 pM)-treated CEM cells CEM cells (2 X 10‘ cells/ml) were treated with 0.2 p~ ddC and 2 p~ nucleoside as indicated for 8 days.

Nucleosides, ddC, and medium were changed at day 4. The doubling of the cells was estimated every 2 days. Nucleosides (2 PM)

dC T U dU C dG Cell growth +O - b - - mtDNA

- - c - - - - -

a Cell growth was not significantly different from that of control cells, i.e. 100% protection. Cell growth or mtDNA content was not significantly different from that of ddC-treated cells. Partial protection (30%).

A B

ddC - i-

FIG. 2. Effect of nucleosides on the mitochondrial toxicity of ddC. The slot blot procedure was described under “Experimental Procedures.” A , an mtDNA-specific probe was used to detect the mtDNA on the slot blot. B, the mtDNA probe on the blot paper of panel A was removed and reprobed with a 32P-labeled topoisomerase I cDNA fragment (5). A densitometer (Visage 2000) was used to determine the relative intensity of the DNA bands.

0.1 1 10

Concentration (1“)

1

FIG. 3. mtDNA synthesis in isolated mitochondria. The in- corporation of [3H]dATP into mtDNA in each case is expressed as a percent of that without the treatment of ddC metabolites. a, dose- dependent effect of ddCTP on mtDNA synthesis. Different concen- trations of ddC or ddCTP were tested for their effect on mtDNA synthesis. The mtDNA synthesis was determined under the condi- tions described under “Experimental Procedures.” b, the effect of ddC metabolites on mtDNA synthesis. The mtDNA synthesis was assayed in the presence of 10 p~ ddC, ddCMP, ddCDP, or ddCTP. The conditions used for the assay were the same as those in a.

.E 3 0 I I i dCydddC ””_ dCyd metabolites 1

6 1 s F s 2 0

- ddC metabolites I I: dCMP

0 10 2 0 3 0 4 0

Minutes

FIG. 4. Anionic HPLC profiles of dCyd or ddC metabolites in isolated mitochondria. A Partisil 10 SAX column (Whatman) was used to analyze the ddC metabolites. The column was eluted with 0.15 M potassium phosphate buffer, pH 6.6. The fractions were collected at 1 ml/min.

ddCDP, or ddCTP (Fig. 4). In contrast, the isolated mito- chondria were able to metabolize dCyd to dCMP, dCDP, and dCTP.

DISCUSSION

The limiting toxicity of ddC in AIDS therapy is delayed adverse effects such as peripheral neuropathy (3). The delayed toxicity was suggested to be due to the potent effect of ddC on mtDNA synthesis in affected organs (5). The delayed onset of the mitochondrial toxicity might be due to the presence of multiple copies of mtDNA in a cell (15). Avoiding the mito- chondrial toxicity without jeopardizing the anti-HIV effects of ddC could optimize the clinical usefulness of ddC. Under- standing the metabolism of ddC and nucleosides in mitochon- dria is an important step in preventing the mitochondrial toxicity. It is thought that the mitochondrial deoxynucleotide pools are compartmentalized and that there are unique nucle- oside kinases which are capable of providing the nucleotide pools in mitochondria (16). The nucleotides in the mitochon- dria are likely to be the immediate precursors for mtDNA synthesis. This notion would suggest that the cytoplasmic dCyd kinase would not be important in the action of ddC against mitochondria. However, the results of this study are not consistent with this idea. It appears that ddC was phos- phorylated in the cytoplasm and then transported into the mitochondria to exert its inhibitory effect on mtDNA synthe- sis. I t is not clear how ddCTP enters the mitochondria. ddCTP could enter the mitochondria through less hydrophilic metab- olites such as ddCDP choline. However, ddCDP choline did not affect mtDNA synthesis in isolated mitochondria, which suggests that it is unlikely to be the intermediate for the ddCTP transport into mitochondria. Both carrier-mediated transport and passive diffusion could be responsible for the

Role of Cytoplasmic Deoxycytidine Kinase 2859

mitochondrial entry of ddCTP. Inhibition of the ddCTP transport in mitochondria might prevent the mitochondrial toxicity of ddC. As a matter of fact, any mechanism that decreases the mitochondrial ddCTP level could alleviate the ddC toxicity. In the case of CEM/araC cells, the lower levels of ddCTP formed in cytoplasm could be responsible for their resistance to the ddC mitochondrial toxicity. A defect in the transport of ddC in the CEM/araC cells is an alternative explanation for the resistance. However, the ddC uptake in the CEM/araC cells is comparable with that of CEM cells. Moreover, ddC uptake was similar in the isolated mitochon- dria derived from each cell line.' Despite the fact that ddC can enter the mitochondria, the phosphorylation of ddC was not detected. The limited ddC phosphorylation in mitochon- dria may be due to the presence of dThd and dCyd in the mitochondria. The dPyd kinase has higher affinity for dThd and dCyd compared with ddC. Other mechanisms, such as the feedback regulation of the mitochondrial kinases by deoxyn- ucleotides, could affect the metabolism of ddC. Although the inefficiency of ddC metabolism in mitochondria is clearly demonstrated, it should be noted that CEM is a proliferating cell line. The metabolism of ddC in quiescent cells could be different from that of CEM cells. The lower activity of nucle- oside kinases in nonproliferating cells may change the role of the mitochondrial enzymes in the metabolism of ddC.

In summary, the cytoplasmic dCyd kinase is responsible for the mitochondrial ddCTP pool which inhibits the mtDNA synthesis in CEM cells. Thus the activity of nucleoside analog triphosphates against DNA polymerase y might not be the only factor that determines the mitochondrial effect of nucle- oside analogs. Mitochondrial nucleoside and nucleotide trans-

' C.-H. Chen and Y.-C. Cheng, unpublished data.

port, and cytoplasmic and mitochondrial metabolism play important roles in the mitochondrial effects of the nucleoside analogs.

1.

2.

3.

4.

5.

6.

7.

8. 9. 10. 11.

12.

13.

14.

15. 16.

REFERENCES Mitsuya, H. & Broder, S. (1986) Proc. Natl. Acad. Sci. U. S. A.

Yarchoan, R., Thomas, R. V., Allain, J-P., McAtee, N., Dubinsky, R., Mitsuya, H., Lawley, T. J., Safai, B., Myers, C. E., Perno, C. F., Klecker, R. W., Wills, R. J., Fischl, M. A., McNeely, M. C., Pluda, J. M., Leuther, M., Collins, J. M. & Broder, S. (1988) Lancet 1, 76-81

Dubinsky, R. M., Yarchoan, R., Dalakas, M. & Broder, S. (1989) Muscle & Nerve 12, 856-860

Vazquez-Padua, M. A., Starnes, M. C. & Cheng, Y. C. (1990) Cancer Commun. 2.55-61

Chen, C. H. & Cheng, Y. C. (1989) J. Biol. Chem. 264, 11934- 11937

Cooney, D. A., Dalal, M., Mitsuya, H., McMahon, J. B., Nadkarni, M., Balzarini, J., Broder, S. & Johns, D. G. (1986) Biochern. Pharmacol. 35, 2065-2068

Starnes, M. C. & Cheng, Y. C. (1987) J. Biol. Chem. 262, 988- 991

Lee, L. S. & Cheng, Y. C. (1976) Biochemistry 15,3686-3690 Ruth, J. L. & Cheng, Y. C. (1981) Mol. Pharrnacol. 20, 415-422 Bresser, J., Doering, J. & Gillespie, D. (1983) DNA 2, 243-253 Bogenhagen, D. & Clayton, D. A. (1974) J. Biol. Chem. 249,

Simpson, M. V., Chin, C. D., Keilbaugh, S. A., Lin, T. S. & Prusoff, W. H. (1989) Biochem. Pharmacol. 38, 1033-1036

Mitra, R. S. & Bernstein, I. A. (1970) J. Biol. Chem. 245, 1255- 1260

Zimmermann, W., Chen, S. M., Bolden, A. & Weissbach, A. (1980) J. Biol. Chem. 255,11847-11852

Robin, E. D. & Wong, R. (1988) J. Cell. Physiol. 136,507-513 Bestwick, R. K., Moffett, G. L. & Mathews, C. K. (1982) J. Biol.

83,1911-1915

7991-7995

Chern. 257,9300-9304