- Email: [email protected]
Identification of a Novel Inhibitor (NSC 665564) of Dihydroorotate Dehydrogenase with a Potency Equivalent to Brequinar
JOBNAME: BBRC 223#3 PAGE: 1 SESS: 18 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
223, 654–659 (1996)
0950
Identification of a Novel Inhibitor (NSC 665564) of Dihydroorotate Dehydrogenase with a Potency Equivalent to Brequinar Emily S. Cleaveland,* Daniel W. Zaharevitz,† James A. Kelley,* Kenneth Paull,† David A. Cooney,* and Harry Ford, Jr.*,1 *Laboratory of Medicinal Chemistry, Building 37, Room 5C-24, Division of Basic Sciences, National Cancer Institute, National Institutes of Health, Bethesda Maryland, 20892; and †Information Technology Branch, Developmental Therapeutics Program, Division of Cancer Treatment, Diagnosis, and Centers, National Cancer Institute, National Institutes of Health, Rockville, Maryland, 20892 Received May 15, 1996 A novel inhibitor of dihydroorotate dehydrogenase (DHO-DH) has been discovered using data from the National Cancer Institute’s in vitro drug screen. Upon analysis of cytotoxicity results from the sixty tumor cell lines used in this screen, the COMPARE program predicted that NSC 665564 was likely to have the same mechanism of inhibition as brequinar, a known potent inhibitor of DHO-DH. We validated this prediction experimentally using MOLT-4 lymphoblast and found the IC50 of brequinar (0.5 mM) and NSC 665564 (0.3 mM) were comparable and that this induced cytotoxicity was reversed by either uridine or cytidine. The enzyme target of NSC 665564 was shown to be identical to that of brequinar when incubation with each drug followed by a 1 h pulse with [14C] sodium bicarbonate resulted in cellular accumulation of [14C]N-carbamyl-L-aspartic acid and [14C]L-dihydroorotic acid, with concurrent marked depletion of CTP and UTP. The Ki’s for NSC 665564 and brequinar were 0.14 and 0.24 mM, respectively, when partially purified MOLT-4 mitochondria (the site of DHO-DH) were used. These results show that mechanistic predictions obtained using correlations from the COMPARE algorithm are independent of structure since the structure of NSC 665564 is dissimilar to that of other established DHO-DH inhibitors. © 1996 Academic Press, Inc.
Inhibitors of dihydroorotate dehydrogenase (DHO-DH, EC 1.3.3.1) are finding wider application as chemotherapeutic agents. Initially explored as anticancer drugs (1,2), DHO-DH inhibitors are now being developed as antimalarials (3–5) and as immunosuppressant agents in transplantation medicine (6,7). These treatment modalities presumably act by inhibition of de novo pyrimidine biosynthesis at the step catalyzed by DHO-DH and prevent subsequent formation of uridine and cytidine nucleotides (8). Preferential cytotoxicity to rapidly dividing cells or parasitic organisms that use pyrimidine nucleotides produced predominantly by the de novo pathway is a hallmark to this type of mechanism. Among the most potent and extensively investigated of the DHO-DH inhibitors is brequinar sodium (Brequinar, Figure 1) which inhibits purified mouse leukemia L1210 mitochondrial DHO-DH with a Ki of z7 nM (9). In a previous publication we demonstrated the utility of the COMPARE program (which uses data from the NCI in vitro anticancer drug screening panel) to successfully predict and identify new DHO-DH inhibitors. These novel compounds were shown experimentally to antagonize pyrimidine biosynthesis at the level of DHO-DH and were, in simple models, structurally similar to previously identified DHO-DH inhibitors (10). In our current study we identify, by using correlations obtained solely from the COMPARE program, a novel DHO-DH inhibitor, NSC 665564 (Figure 1) whose potency is equivalent to brequinar’s but whose structure is dissimilar to any previously discovered DHO-DH inhibitor. Further utilization of this approach is certain to aid the discovery of new antipyrimidines. 1
Corresponding author. Fax: (301) 402-2275. Abbreviations: BNID, 1-(p-bromophenyl)-2-methyl-1H-naphth[2,3-d]imidazole-4,9-dione; DHO, L-dihydroorotic acid; DHO-DH, dihydroorotic acid dehydrogenase; NCI, National Cancer Institute. 654 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 223#3 PAGE: 2 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Structures of brequinar, 6-fluoro-2-(20-fluoro-1, 10-biphenyl-4-yl)-3-methyl-4-quinolinecarboxylic acid, sodium salt and NSC 665564, 2-acetyl-7-bromo-1-phenyl-1,2,3,4-tetrahydro-b-carboline.
MATERIALS AND METHODS Chemicals. NSC 665564 (Interneuron Pharmaceuticals, Lexington, MA, N-0776), and brequinar sodium were obtained from the Developmental Therapeutics Program, NCI (Bethesda, MD). [14C]L-dihydroorotic acid, [14C]L-orotic acid, [14C]N-carbamyl-L-aspartic acid and [14C] sodium bicarbonate were purchased from Moravek Biochemicals, Brea, CA. All other reagents were of the highest grade available. Cell culture and growth inhibition studies (IC50). MOLT-4 lymphocytes (cat #1582-CRL, American Type Culture Collection, Rockville, MD) were cultured at 37°C in RPMI 1640 medium containing L-glutamine (Mediatech, Herndon, VA) and supplemented with 10% heat-inactivated (60 min, 56°C) fetal calf serum and 50 mg/ml gentamicin sulfate in 100% humidity and using 5% CO2 concentrations unless otherwise indicated in the text. Under all conditions used, MOLT-4 cells showed a doubling-time between 24 and 26 h. Growth inhibition studies were conducted by incubating these cells with various concentrations of NSC 665564 or brequinar for 18 or 24 h. Determination of ribonucleotide pools in MOLT-4 cells. Ribonucleotide pools were measured after an 18 hr incubation of MOLT-4 lymphoblast (5 × 105 cells/ ml × 50 ml) in the presence or absence of NSC 665564 (0.3 mM), or brequinar (1 mM). In parallel experiments cells were cultured in medium supplemented with 50 mM cytidine, uridine or saline to which was added NSC 665564 or brequinar after an interval of 1 h. Cells were collected by centrifugation, lysed, and extracted into ice-cold 60% methanol-water. After centrifugation at 12,000 × g for 6 min the nucleotide pools in the supernatant were measured by anion-exchange HPLC as previously described [10]. Effect of NSC 665564 and brequinar on [14C] bicarbonate flux into pyrimidine precursors and nucleotides. MOLT-4 cells, cultured for 18 h in, sodium bicarbonate free, RPMI 1640 medium, were incubated for 1 h in medium supplemented with [14C]sodium bicarbonate (4 mCi/ml). Cells were collected as described above, and the supernatant analyzed for nucleotide content and specific radioactivity using ion-exchange chromatography with on-line radiometric detection [10]. The concentration of L-dihydroorotate (DHO) and N-carbamyl-L-aspartic acid in cellular extracts was also determined using gradient elution ion-exchange chromatography as described previously [10]. Mitochondrial DHO-DH preparation and assay. A crude preparation containing intact mitochondria from MOLT-4 cells was prepared as previously described [10]. The apparent Ki determinations and the influence of brequinar and NSC 665564 on the enzymatic activity of bound mitochondrial DHO-DH were measured using DHO (50 mM) in the presence of the endogenous ubiquinone contained in the mitochondrial preparation; the rate of conversion of [14C]L-dihydroorotic acid to [14C]L-orotic acid was measured by ion-exchange chromatography as described above and taken as the index of enzyme activity. 655
JOBNAME: BBRC 223#3 PAGE: 3 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Drug screening and COMPARE program. Methodology concerning the NCI in vitro drug screen, data analysis and interpretation, and a description of the COMPARE program have been presented in detail elsewhere (11–13).
RESULTS AND DISCUSSION The NCI in vitro anticancer drug discovery screen and its applications have recently been described in detail (13,14). Data obtained from screening new drug candidates against a panel of 60 different cultured human cancer cell lines generates descriptors of cellular toxicity from which a meangraph profile can be produced. When this is compared to the meangraph profile of similarly screened standard chemotherapeutic agents, for which the biochemical mechanism of action has been established, correlations can be made between meangraph profiles using the COMPARE program. The meangraphs for NSC 665564 and brequinar are shown in Figure 2. COMPARE correlations obtained for NSC 665564 and a number of pyrimidine biosynthesis inhibitors using brequinar as “seed compound” are listed in Table 1. As has been previously demonstrated, high COMPARE correlations (>0.5) reflect a similar meangraph profile and invoke a possible mechanistic similarity between the cytotoxicity of the drug candidate and the established chemotherapeutic agent used as the “seed compound”. MOLT-4 human lymphoblasts were used to test the premise that the cytotoxicity exhibited by NSC 665564 in the drug screen was due to antagonism at DHO-DH as suggested by the COMPARE evaluation. Parallel experiments were conducted using brequinar. In vitro, both NSC 665564 and brequinar inhibited the 24 hr logarithmic growth of MOLT-4 cells with IC50’s of 0.3 mM and 0.5 mM, respectively; performed uridine or cytidine reversed the antiproliferative effects of both drugs. NSC 665564 (1 mM) inhibited MOLT-4 cell growth by 97% during a 24 h incubation, yet, when simultaneously exposed to either uridine or cytidine (50 mM) a 35% and 12% inhibition of cellular growth ensued. Brequinar at 5 mM caused a 66% inhibition of cell growth; concurrent addition of either uridine or cytidine (50 mM) diminished the growth inhibition to 6%. Inhibition of DHO-DH, the fourth enzyme in the de novo pyrimidine biosynthetic pathway, precludes conversion of dihydroorotate (DHO) to orotic acid. Lacking this precursor the production of uridine-59-monophosphate and ultimately cytidine-59-triphosphate is reduced. When MOLT-4 cells were incubated for 18 h with NSC 665564, brequinar or saline and then pulsed for 1 hr with [14C] sodium bicarbonate, total unlabeled pyrimidine nucleotide triphosphate levels fell dramatically in the drug treated group. NSC 665564 (0.3 mM) decreases cellular pools of both CTP and UTP levels to 16% of control samples, while brequinar (1 mM) reduces CTP and UTP levels to 10% and 6%, respectively (Table 2). In this same study a decrease in the formation of the TABLE 1 COMPARE Correlations and Inhibition Constants for Some DHO-DH Inhibitors Compound
COMPARE correlationa
Apparent Ki (mM)b
1.00 0.84 0.75 0.72 0.67
0.24 0.53 NDd 0.33 0.14
Brequinar BNIDc Dichloroallyl lawsone Redoxale NSC 665564
a Correlations were obtained by a comparison of meangraph data from the NCI in vivo cancer screen using the computer algorithm COMPARE. Brequinar was used as the seed compound (12,13). b Except for brequinar and NSC 665564, the apparent Ki values were taken from reference 10. c 1-(p-bromophenyl)-2- methyl-1H-naphth[2,3-d]imidazole-4,9-dione (BNID) is structurally related to dichloroallyl lawsone, a naphthoquinone and well known competitive inhibitor of ubiquinone (16). d Not determined. A Ki of 0.027 mM for L1210 mitochondrial DHO-DH has been reported (16). e Redoxal (2,29-([3,39-dimethoxy[1,19-biphenyl]-4,49-diyl)diimino)]bis-benzoic acid) has structural features analogous to brequinar (10).
656
JOBNAME: BBRC 223#3 PAGE: 4 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Meangraphs for NSC 665564 and brequinar. The central vertical line represents the mean IC50 of all cell lines for each drug. Bars to the right indicate a more potent effect of the drug, bars to the left indicate a less potent effect than the average.
end-products of pyrimidine biosynthesis was accompanied by a large accumulation of [14C] Ncarbamyl-L-aspartic acid and dihydroorotic acid (DHO), the two antecedent intermediates in the de novo pathway. Incubation with NSC 665564 (0.3 mM) resulted in an 18-fold and 112-fold increase in labeled N-carbamyl-L-aspartic acid and DHO, respectively, while 1 mM brequinar showed an 11-fold and 64-fold increase in these two intermediates. After the 1-h pulse with the labeled bicarbonate, no [14C] incorporation was detected in the CTP and UTP pools for cells incubated with either NSC 665564 (0.3 mM) or brequinar (1 mM). The notable increases (121–156%, Table 2) in 657
JOBNAME: BBRC 223#3 PAGE: 5 SESS: 16 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 2 Cellular Triphosphate Levels in the Presence of Antipyrimidines Nucleotide concentration (pmol/106 cells)
Compound None Brequinar (1 mM) NSC 665564 (0.3 mM)
[CTP]
[UTP]
[ATP]
[GTP]
300 30 48
921 58 150
2393 2895 3650
272 420 424
MOLT-4 cells were incubated in the presence of brequinar and NSC 665564 for 18 hr. Cells were collected, extracted and analyzed by HPLC as described in Methods and Materials.
net unlabeled adenosine-59-triphosphate and guanosine-59-triphosphate levels seen after incubation with either drug is similar to that observed by Lyons et al. (15) using mouse leukemia L1210 cells. These authors suggest that brequinar-induced inhibition promotes the build-up of components in the early stops of the pyrimidine de novo pathway (i.e. phosphoribosyl pyrophosphate, Lglutamine) and that these act to stimulate production of purine nucleotides. Direct inhibition of the DHO-DH by both NSC 665564 and brequinar was demonstrated using a crude preparation of MOLT-4 mitochondria. The apparent Ki values, determined from a Dixon plot, were 0.14 mM and 0.24 mM, respectively (Figure 3). Structurally, NSC 665564 is dissimilar to any previously reported DHO-DH inhibitor including brequinar and the lapachol family of compounds (i.e. dichloroallyl lawsone). Like brequinar, NSC
FIG. 3. Dixon plots for NSC 665564 (A) and brequinar (B). Ki values of NSC 665564 (0.14 mM, r2 0.988) and brequinar (0.24 mM, r2 0.987) were determined by a plot of 1/rate of formation of orotic acid at varying concentrations of each drug. 658
JOBNAME: BBRC 223#3 PAGE: 6 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
665564 does not structurally resemble either the substrate, product or associated DHO-DH cofactor, ubiquinone. Kinetic studies performed by other investigators suggest that brequinar exerts its inhibition by binding to a protein site distinctly different from the active or cofactor binding sites of DHO-DH. This alternate binding site is presumably specific to the mammalian enzyme in as much as DHO-DH from lower species are not inhibited by brequinar (9). Taken together, the evidence to date suggests that NSC 665564 is not binding to the primary catalytic site of the enzyme. Our discovery of this novel DHO-DH dehydrogenase inhibitor will allow us to construct computer models for the purpose of characterizing the structural requirements essential for biological inhibition of this enzyme. Like brequinar, an antitumor agent now being developed as an immunosuppressive agent, any new DHO-DH inhibitors may find potential application in cancer, transplantation medicine or possibly one or more parasitic diseases. This successful use of the COMPARE program to uncover the mechanism of action of a new, structurally unique and biologically active lead compound further validates the use of this analogous tool in drug discovery programs. REFERENCES 1. Kensler, T. W., and Cooney, D. A. (1989) in Design of Enzyme Inhibitors as Drugs, (Sandler, M., and Smith, H. J. Eds.), pp. 379–401. Oxford Univ. Press, Oxford, England. 2. Cody, R., Stewart, D., DeForni, M., Moore, M., Dallaire, B., Azarnia, N., and Gyves, J. (1993) Am. J. Clin. Oncol. 16, 526–528. 3. Ittrat, I., Asawamahasakda, W., and Meshnick, S. R. (1994) Exper. Parasitol. 79, 50–56. 4. Krungkrai, J., Krungkrai, S. R., and Phakanont, K. (1992) Biochem. Pharmacol. 43, 1295–1301. 5. Seymour, K. K., Lyons, S. D., Phillips, L., Rieckmann, K. H., and Christopherson, R. I. (1994) Biochemistry 33, 5268–5274. 6. Allison, A. C., and Eugui, E. M. (1993) Transplantation Proc. 25 (3) Suppl. 2, 8–18. 7. Makowka, L., Sher, L. S., and Cramer, D. V. (1993) Immunolog. Rev. 136, 51–70. 8. Chen, S. F., Ruben, R. L., and Dexter, D. L. (1986) Cancer Res. 46, 5014–5019. 9. Chen, S. F., Perrella, F. W., Behrens, D. L., and Papp, L. M. (1992) Cancer Res. 52, 3521–3527. 10. Cleaveland, E. S., Monks, A., Vaigro-Wolff, A., Zaharevitz, D. W., Paull, K., Ardalan, K., Cooney, D. A., and Ford, H. Jr. (1995) Biochem. Pharmacol. 49, 947–954. 11. Monks, A., Scudiero, D., Shoemaker, R., Paull, K., Vistica, D., Hose, C., Langley, J., Cronise, P., Vaigro-Wolff, A., Gray-Goodrich, M., Campbell, H., Mayo, J., and Boyd, M. (1991) J. Natl. Cancer Inst. 83, 757–5766. 12. Paull, K. D., Shoemaker, R. H., Hodes, L., Monks, A., Scudiero, D. A., Rubinstein, L., Plowman, J., and Boyd, M. R. (1989) J. Natl. Cancer Inst. 81, 1088–1092. 13. Paull, K. P., Hamel, E., and Malspeis, L. (1995) In Cancer Chemotherapeutic Agents (W. O. Faye, Ed.), Chpt. 2, pp. 1–40, American Chemical Society, Washington DC. 14. Boyd, M., and Paull, K. D. (1995) Drug Develop. Res. 34, 91–109. 15. Lyons, S. D., and Christopherson, R. L. (1990) Biochem. Int. 22, 939–949. 16. Bennett, L. L. Jr., Smithers, D., Rose, L. M., Adamson, J., and Thomas, H. J. (1979) Cancer Res. 39, 4868–4874.
659
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
223, 654–659 (1996)
0950
Identification of a Novel Inhibitor (NSC 665564) of Dihydroorotate Dehydrogenase with a Potency Equivalent to Brequinar Emily S. Cleaveland,* Daniel W. Zaharevitz,† James A. Kelley,* Kenneth Paull,† David A. Cooney,* and Harry Ford, Jr.*,1 *Laboratory of Medicinal Chemistry, Building 37, Room 5C-24, Division of Basic Sciences, National Cancer Institute, National Institutes of Health, Bethesda Maryland, 20892; and †Information Technology Branch, Developmental Therapeutics Program, Division of Cancer Treatment, Diagnosis, and Centers, National Cancer Institute, National Institutes of Health, Rockville, Maryland, 20892 Received May 15, 1996 A novel inhibitor of dihydroorotate dehydrogenase (DHO-DH) has been discovered using data from the National Cancer Institute’s in vitro drug screen. Upon analysis of cytotoxicity results from the sixty tumor cell lines used in this screen, the COMPARE program predicted that NSC 665564 was likely to have the same mechanism of inhibition as brequinar, a known potent inhibitor of DHO-DH. We validated this prediction experimentally using MOLT-4 lymphoblast and found the IC50 of brequinar (0.5 mM) and NSC 665564 (0.3 mM) were comparable and that this induced cytotoxicity was reversed by either uridine or cytidine. The enzyme target of NSC 665564 was shown to be identical to that of brequinar when incubation with each drug followed by a 1 h pulse with [14C] sodium bicarbonate resulted in cellular accumulation of [14C]N-carbamyl-L-aspartic acid and [14C]L-dihydroorotic acid, with concurrent marked depletion of CTP and UTP. The Ki’s for NSC 665564 and brequinar were 0.14 and 0.24 mM, respectively, when partially purified MOLT-4 mitochondria (the site of DHO-DH) were used. These results show that mechanistic predictions obtained using correlations from the COMPARE algorithm are independent of structure since the structure of NSC 665564 is dissimilar to that of other established DHO-DH inhibitors. © 1996 Academic Press, Inc.
Inhibitors of dihydroorotate dehydrogenase (DHO-DH, EC 1.3.3.1) are finding wider application as chemotherapeutic agents. Initially explored as anticancer drugs (1,2), DHO-DH inhibitors are now being developed as antimalarials (3–5) and as immunosuppressant agents in transplantation medicine (6,7). These treatment modalities presumably act by inhibition of de novo pyrimidine biosynthesis at the step catalyzed by DHO-DH and prevent subsequent formation of uridine and cytidine nucleotides (8). Preferential cytotoxicity to rapidly dividing cells or parasitic organisms that use pyrimidine nucleotides produced predominantly by the de novo pathway is a hallmark to this type of mechanism. Among the most potent and extensively investigated of the DHO-DH inhibitors is brequinar sodium (Brequinar, Figure 1) which inhibits purified mouse leukemia L1210 mitochondrial DHO-DH with a Ki of z7 nM (9). In a previous publication we demonstrated the utility of the COMPARE program (which uses data from the NCI in vitro anticancer drug screening panel) to successfully predict and identify new DHO-DH inhibitors. These novel compounds were shown experimentally to antagonize pyrimidine biosynthesis at the level of DHO-DH and were, in simple models, structurally similar to previously identified DHO-DH inhibitors (10). In our current study we identify, by using correlations obtained solely from the COMPARE program, a novel DHO-DH inhibitor, NSC 665564 (Figure 1) whose potency is equivalent to brequinar’s but whose structure is dissimilar to any previously discovered DHO-DH inhibitor. Further utilization of this approach is certain to aid the discovery of new antipyrimidines. 1
Corresponding author. Fax: (301) 402-2275. Abbreviations: BNID, 1-(p-bromophenyl)-2-methyl-1H-naphth[2,3-d]imidazole-4,9-dione; DHO, L-dihydroorotic acid; DHO-DH, dihydroorotic acid dehydrogenase; NCI, National Cancer Institute. 654 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 223#3 PAGE: 2 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Structures of brequinar, 6-fluoro-2-(20-fluoro-1, 10-biphenyl-4-yl)-3-methyl-4-quinolinecarboxylic acid, sodium salt and NSC 665564, 2-acetyl-7-bromo-1-phenyl-1,2,3,4-tetrahydro-b-carboline.
MATERIALS AND METHODS Chemicals. NSC 665564 (Interneuron Pharmaceuticals, Lexington, MA, N-0776), and brequinar sodium were obtained from the Developmental Therapeutics Program, NCI (Bethesda, MD). [14C]L-dihydroorotic acid, [14C]L-orotic acid, [14C]N-carbamyl-L-aspartic acid and [14C] sodium bicarbonate were purchased from Moravek Biochemicals, Brea, CA. All other reagents were of the highest grade available. Cell culture and growth inhibition studies (IC50). MOLT-4 lymphocytes (cat #1582-CRL, American Type Culture Collection, Rockville, MD) were cultured at 37°C in RPMI 1640 medium containing L-glutamine (Mediatech, Herndon, VA) and supplemented with 10% heat-inactivated (60 min, 56°C) fetal calf serum and 50 mg/ml gentamicin sulfate in 100% humidity and using 5% CO2 concentrations unless otherwise indicated in the text. Under all conditions used, MOLT-4 cells showed a doubling-time between 24 and 26 h. Growth inhibition studies were conducted by incubating these cells with various concentrations of NSC 665564 or brequinar for 18 or 24 h. Determination of ribonucleotide pools in MOLT-4 cells. Ribonucleotide pools were measured after an 18 hr incubation of MOLT-4 lymphoblast (5 × 105 cells/ ml × 50 ml) in the presence or absence of NSC 665564 (0.3 mM), or brequinar (1 mM). In parallel experiments cells were cultured in medium supplemented with 50 mM cytidine, uridine or saline to which was added NSC 665564 or brequinar after an interval of 1 h. Cells were collected by centrifugation, lysed, and extracted into ice-cold 60% methanol-water. After centrifugation at 12,000 × g for 6 min the nucleotide pools in the supernatant were measured by anion-exchange HPLC as previously described [10]. Effect of NSC 665564 and brequinar on [14C] bicarbonate flux into pyrimidine precursors and nucleotides. MOLT-4 cells, cultured for 18 h in, sodium bicarbonate free, RPMI 1640 medium, were incubated for 1 h in medium supplemented with [14C]sodium bicarbonate (4 mCi/ml). Cells were collected as described above, and the supernatant analyzed for nucleotide content and specific radioactivity using ion-exchange chromatography with on-line radiometric detection [10]. The concentration of L-dihydroorotate (DHO) and N-carbamyl-L-aspartic acid in cellular extracts was also determined using gradient elution ion-exchange chromatography as described previously [10]. Mitochondrial DHO-DH preparation and assay. A crude preparation containing intact mitochondria from MOLT-4 cells was prepared as previously described [10]. The apparent Ki determinations and the influence of brequinar and NSC 665564 on the enzymatic activity of bound mitochondrial DHO-DH were measured using DHO (50 mM) in the presence of the endogenous ubiquinone contained in the mitochondrial preparation; the rate of conversion of [14C]L-dihydroorotic acid to [14C]L-orotic acid was measured by ion-exchange chromatography as described above and taken as the index of enzyme activity. 655
JOBNAME: BBRC 223#3 PAGE: 3 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Drug screening and COMPARE program. Methodology concerning the NCI in vitro drug screen, data analysis and interpretation, and a description of the COMPARE program have been presented in detail elsewhere (11–13).
RESULTS AND DISCUSSION The NCI in vitro anticancer drug discovery screen and its applications have recently been described in detail (13,14). Data obtained from screening new drug candidates against a panel of 60 different cultured human cancer cell lines generates descriptors of cellular toxicity from which a meangraph profile can be produced. When this is compared to the meangraph profile of similarly screened standard chemotherapeutic agents, for which the biochemical mechanism of action has been established, correlations can be made between meangraph profiles using the COMPARE program. The meangraphs for NSC 665564 and brequinar are shown in Figure 2. COMPARE correlations obtained for NSC 665564 and a number of pyrimidine biosynthesis inhibitors using brequinar as “seed compound” are listed in Table 1. As has been previously demonstrated, high COMPARE correlations (>0.5) reflect a similar meangraph profile and invoke a possible mechanistic similarity between the cytotoxicity of the drug candidate and the established chemotherapeutic agent used as the “seed compound”. MOLT-4 human lymphoblasts were used to test the premise that the cytotoxicity exhibited by NSC 665564 in the drug screen was due to antagonism at DHO-DH as suggested by the COMPARE evaluation. Parallel experiments were conducted using brequinar. In vitro, both NSC 665564 and brequinar inhibited the 24 hr logarithmic growth of MOLT-4 cells with IC50’s of 0.3 mM and 0.5 mM, respectively; performed uridine or cytidine reversed the antiproliferative effects of both drugs. NSC 665564 (1 mM) inhibited MOLT-4 cell growth by 97% during a 24 h incubation, yet, when simultaneously exposed to either uridine or cytidine (50 mM) a 35% and 12% inhibition of cellular growth ensued. Brequinar at 5 mM caused a 66% inhibition of cell growth; concurrent addition of either uridine or cytidine (50 mM) diminished the growth inhibition to 6%. Inhibition of DHO-DH, the fourth enzyme in the de novo pyrimidine biosynthetic pathway, precludes conversion of dihydroorotate (DHO) to orotic acid. Lacking this precursor the production of uridine-59-monophosphate and ultimately cytidine-59-triphosphate is reduced. When MOLT-4 cells were incubated for 18 h with NSC 665564, brequinar or saline and then pulsed for 1 hr with [14C] sodium bicarbonate, total unlabeled pyrimidine nucleotide triphosphate levels fell dramatically in the drug treated group. NSC 665564 (0.3 mM) decreases cellular pools of both CTP and UTP levels to 16% of control samples, while brequinar (1 mM) reduces CTP and UTP levels to 10% and 6%, respectively (Table 2). In this same study a decrease in the formation of the TABLE 1 COMPARE Correlations and Inhibition Constants for Some DHO-DH Inhibitors Compound
COMPARE correlationa
Apparent Ki (mM)b
1.00 0.84 0.75 0.72 0.67
0.24 0.53 NDd 0.33 0.14
Brequinar BNIDc Dichloroallyl lawsone Redoxale NSC 665564
a Correlations were obtained by a comparison of meangraph data from the NCI in vivo cancer screen using the computer algorithm COMPARE. Brequinar was used as the seed compound (12,13). b Except for brequinar and NSC 665564, the apparent Ki values were taken from reference 10. c 1-(p-bromophenyl)-2- methyl-1H-naphth[2,3-d]imidazole-4,9-dione (BNID) is structurally related to dichloroallyl lawsone, a naphthoquinone and well known competitive inhibitor of ubiquinone (16). d Not determined. A Ki of 0.027 mM for L1210 mitochondrial DHO-DH has been reported (16). e Redoxal (2,29-([3,39-dimethoxy[1,19-biphenyl]-4,49-diyl)diimino)]bis-benzoic acid) has structural features analogous to brequinar (10).
656
JOBNAME: BBRC 223#3 PAGE: 4 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Meangraphs for NSC 665564 and brequinar. The central vertical line represents the mean IC50 of all cell lines for each drug. Bars to the right indicate a more potent effect of the drug, bars to the left indicate a less potent effect than the average.
end-products of pyrimidine biosynthesis was accompanied by a large accumulation of [14C] Ncarbamyl-L-aspartic acid and dihydroorotic acid (DHO), the two antecedent intermediates in the de novo pathway. Incubation with NSC 665564 (0.3 mM) resulted in an 18-fold and 112-fold increase in labeled N-carbamyl-L-aspartic acid and DHO, respectively, while 1 mM brequinar showed an 11-fold and 64-fold increase in these two intermediates. After the 1-h pulse with the labeled bicarbonate, no [14C] incorporation was detected in the CTP and UTP pools for cells incubated with either NSC 665564 (0.3 mM) or brequinar (1 mM). The notable increases (121–156%, Table 2) in 657
JOBNAME: BBRC 223#3 PAGE: 5 SESS: 16 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 2 Cellular Triphosphate Levels in the Presence of Antipyrimidines Nucleotide concentration (pmol/106 cells)
Compound None Brequinar (1 mM) NSC 665564 (0.3 mM)
[CTP]
[UTP]
[ATP]
[GTP]
300 30 48
921 58 150
2393 2895 3650
272 420 424
MOLT-4 cells were incubated in the presence of brequinar and NSC 665564 for 18 hr. Cells were collected, extracted and analyzed by HPLC as described in Methods and Materials.
net unlabeled adenosine-59-triphosphate and guanosine-59-triphosphate levels seen after incubation with either drug is similar to that observed by Lyons et al. (15) using mouse leukemia L1210 cells. These authors suggest that brequinar-induced inhibition promotes the build-up of components in the early stops of the pyrimidine de novo pathway (i.e. phosphoribosyl pyrophosphate, Lglutamine) and that these act to stimulate production of purine nucleotides. Direct inhibition of the DHO-DH by both NSC 665564 and brequinar was demonstrated using a crude preparation of MOLT-4 mitochondria. The apparent Ki values, determined from a Dixon plot, were 0.14 mM and 0.24 mM, respectively (Figure 3). Structurally, NSC 665564 is dissimilar to any previously reported DHO-DH inhibitor including brequinar and the lapachol family of compounds (i.e. dichloroallyl lawsone). Like brequinar, NSC
FIG. 3. Dixon plots for NSC 665564 (A) and brequinar (B). Ki values of NSC 665564 (0.14 mM, r2 0.988) and brequinar (0.24 mM, r2 0.987) were determined by a plot of 1/rate of formation of orotic acid at varying concentrations of each drug. 658
JOBNAME: BBRC 223#3 PAGE: 6 SESS: 15 OUTPUT: Sat Jun 29 11:10:50 1996 /xypage/worksmart/tsp000/72467k/33
Vol. 223, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
665564 does not structurally resemble either the substrate, product or associated DHO-DH cofactor, ubiquinone. Kinetic studies performed by other investigators suggest that brequinar exerts its inhibition by binding to a protein site distinctly different from the active or cofactor binding sites of DHO-DH. This alternate binding site is presumably specific to the mammalian enzyme in as much as DHO-DH from lower species are not inhibited by brequinar (9). Taken together, the evidence to date suggests that NSC 665564 is not binding to the primary catalytic site of the enzyme. Our discovery of this novel DHO-DH dehydrogenase inhibitor will allow us to construct computer models for the purpose of characterizing the structural requirements essential for biological inhibition of this enzyme. Like brequinar, an antitumor agent now being developed as an immunosuppressive agent, any new DHO-DH inhibitors may find potential application in cancer, transplantation medicine or possibly one or more parasitic diseases. This successful use of the COMPARE program to uncover the mechanism of action of a new, structurally unique and biologically active lead compound further validates the use of this analogous tool in drug discovery programs. REFERENCES 1. Kensler, T. W., and Cooney, D. A. (1989) in Design of Enzyme Inhibitors as Drugs, (Sandler, M., and Smith, H. J. Eds.), pp. 379–401. Oxford Univ. Press, Oxford, England. 2. Cody, R., Stewart, D., DeForni, M., Moore, M., Dallaire, B., Azarnia, N., and Gyves, J. (1993) Am. J. Clin. Oncol. 16, 526–528. 3. Ittrat, I., Asawamahasakda, W., and Meshnick, S. R. (1994) Exper. Parasitol. 79, 50–56. 4. Krungkrai, J., Krungkrai, S. R., and Phakanont, K. (1992) Biochem. Pharmacol. 43, 1295–1301. 5. Seymour, K. K., Lyons, S. D., Phillips, L., Rieckmann, K. H., and Christopherson, R. I. (1994) Biochemistry 33, 5268–5274. 6. Allison, A. C., and Eugui, E. M. (1993) Transplantation Proc. 25 (3) Suppl. 2, 8–18. 7. Makowka, L., Sher, L. S., and Cramer, D. V. (1993) Immunolog. Rev. 136, 51–70. 8. Chen, S. F., Ruben, R. L., and Dexter, D. L. (1986) Cancer Res. 46, 5014–5019. 9. Chen, S. F., Perrella, F. W., Behrens, D. L., and Papp, L. M. (1992) Cancer Res. 52, 3521–3527. 10. Cleaveland, E. S., Monks, A., Vaigro-Wolff, A., Zaharevitz, D. W., Paull, K., Ardalan, K., Cooney, D. A., and Ford, H. Jr. (1995) Biochem. Pharmacol. 49, 947–954. 11. Monks, A., Scudiero, D., Shoemaker, R., Paull, K., Vistica, D., Hose, C., Langley, J., Cronise, P., Vaigro-Wolff, A., Gray-Goodrich, M., Campbell, H., Mayo, J., and Boyd, M. (1991) J. Natl. Cancer Inst. 83, 757–5766. 12. Paull, K. D., Shoemaker, R. H., Hodes, L., Monks, A., Scudiero, D. A., Rubinstein, L., Plowman, J., and Boyd, M. R. (1989) J. Natl. Cancer Inst. 81, 1088–1092. 13. Paull, K. P., Hamel, E., and Malspeis, L. (1995) In Cancer Chemotherapeutic Agents (W. O. Faye, Ed.), Chpt. 2, pp. 1–40, American Chemical Society, Washington DC. 14. Boyd, M., and Paull, K. D. (1995) Drug Develop. Res. 34, 91–109. 15. Lyons, S. D., and Christopherson, R. L. (1990) Biochem. Int. 22, 939–949. 16. Bennett, L. L. Jr., Smithers, D., Rose, L. M., Adamson, J., and Thomas, H. J. (1979) Cancer Res. 39, 4868–4874.
659