In vivo inactivation by acivicin of carbamoyl-phosphate synthetase II in rat hepatoma

Revisiting the role of dihydroorotate dehydrogenase as a therapeutic target for cancer

Identified as a hallmark of cancer, metabolic reprogramming allows cancer cells to rapidly proliferate, resist chemotherapies, invade, metastasize, and survive a nutrient-deprived microenvironment. Rapidly growing cells depend on sufficient concentrations of nucleotides to sustain proliferation. One enzyme essential for the de novo biosynthesis of pyrimidine-based nucleotides is dihydroorotate dehydrogenase (DHODH), a known therapeutic target for multiple diseases. Brequinar, leflunomide, and teriflunomide, all of which are potent DHODH inhibitors, have been clinically evaluated but failed to receive FDA approval for the treatment of cancer. Inhibition of DHODH depletes intracellular pyrimidine nucleotide pools and results in cell cycle arrest in S-phase, sensitization to current chemotherapies, and differentiation in neural crest cells and acute myeloid leukemia (AML). Furthermore, DHODH is a synthetic lethal susceptibility in several oncogenic backgrounds. Therefore, DHODH-targeted therapy has potential value as part of a combination therapy for the treatment of cancer. In this review, we focus on the de novo pyrimidine biosynthesis pathway as a target for cancer therapy, and in particular, DHODH. In the first part, we provide a comprehensive overview of this pathway and its regulation in cancer. We further describe the relevance of DHODH as a target for cancer therapy using bioinformatic analyses. We then explore the preclinical and clinical results of pharmacological strategies to target the de novo pyrimidine biosynthesis pathway, with an emphasis on DHODH. Finally, we discuss potential strategies to harness DHODH as a target for the treatment of cancer.

Biochemical pharmacology and analysis of fluoropyrimidines alone and in combination with modulators

After more than three decades since their introduction, fluoropyrimidines, especially FUra, are still a mainstay in the treatment of various solid malignancies. The antitumor effects of fluoropyrimidines are dependent upon metabolic activation. FdUMP, FUTP and FdUTP were identified as the key cytotoxic metabolites that interfere with the proper function of thymidylate synthase and nucleic acids. The relevance of these metabolites is cell-type specific. Recently, fluorouridine diphospho sugars have been detected, but the precise function of this class of metabolites is currently unknown. In mammalian systems fluoropyrimidines and their natural counterparts share the same metabolic pathways since the substrate properties in enzyme-catalyzed reactions are frequently comparable. Ongoing studies indicate that the metabolism and action of fluoropyrimidines exhibit circadian rhythms, which appear to be due to variations in the activity of metabolizing enzymes. Essential for the expanding knowledge of the pathways and effects of fluoropyrimidines has been the constant improvement of analytical methods. These include ligand binding techniques, numerous dedicated HPLC systems and 19F-NMR. Because the overall response rates achieved with fluoropyrimidines are modest, strategies based on biochemical modulation have been devised to enhance their therapeutic index. Biochemical modulators include a wide range of various compounds with different modes of action. In recently completed clinical trials, combinations of FUra with leucovorin, a precursor for 5,10-methylene tetrahydrofolate, or with levamisole, an anthelminthic with immunomodulatory activity, appeared to be superior to FUra alone. At the preclinical level combinations of fluoropyrimidines with, e.g. interferons or L-histidinol were demonstrated to be interesting candidates for further testing. The future therapeutic utility of fluoropyrimidines will depend on both the improvement of combination regimens currently used in the treatment of cancer patients and the judicious clinical implementation of promising experimental modulation strategies. Moreover, novel fluoropyrimidines with superior pharmacological properties may become important as part of or instead of modulation approaches.

Acivicin: a highly active potential chemotherapeutic agent against visceral leishmaniasis

Acivicin, a chlorinated amino acid antibiotic, is found to be remarkably effective in killing both the vector and the host form of the parasitic protozoa, Leishmania donovani, the causative agent for visceral leishmaniasis or Kala-azar. The ED50 (50 nM) for the pathogenic amastigote form in in vitro screening system is significantly lower than the reported values for other drugs under trial. The drug irreversibly inactivates both in vitro and in vivo carbamyl phosphate synthetase II, the first enzyme of the pyrimidine biosynthetic pathway. The irreversible inactivation of this sensitive target enzyme and lack of effective reversal by glutamine makes acivicin a preferred candidate for potential chemotherapy against increasing number of Kala-azar cases that are reported to be unresponsive to pentavalent antimonials.

Metabolism and action of amino acid analog anti-cancer agents

The preclinical pharmacology, antitumor activity and toxicity of seven of the more important amino acid analogs, with antineoplastic activity, is discussed in this review. Three of these compounds are antagonists of L-glutamine: acivicin, DON and azaserine; and two are analogs of L-aspartic acid: PALA and L-alanosine. All five of these antimetabolites interrupt cellular nucleotide synthesis and thereby halt the formation of DNA and/or RNA in the tumor cell. The remaining two compounds, buthionine sulfoximine and difluoromethylornithine, are inhibitors of glutathione and polyamine synthesis, respectively, with limited intrinsic antitumor activity; however, because of their powerful biochemical actions and their low systemic toxicities, they are being evaluated as chemotherapeutic adjuncts to or modulators of other more toxic antineoplastic agents.

Enzymic programs of rat bone marrow and the impact of acivicin and tiazofurin

The in vivo actions of two antimetabolites, acivicin (NSC-163501) and tiazofurin (NSC-286193), were examined on the enzymic programs of rat bone marrow. From the bone marrow of the femurs, 100,000 g supernatant fractions were prepared; enzymic activities were measured by isotopic assays, and cellularity was determined. In the normal bone marrow, the specific activities of pyrimidine de novo synthetic enzymes, CDP reductase, dTMP synthase, CTP synthase, carbamoyl-phosphate synthase II (synthase II), orotidine 5'-phosphate decarboxylase and aspartate carbamoyltransferase, were 1, 2.7, 5, 10, 63 and 601 nmol/hr/mg protein, respectively, whereas those of the salvage enzymes, deoxycytidine, thymidine, cytidine and uridine kinases were 3, 43, 149, and 367 nmol/hr/mg protein, respectively. In purine biosynthesis, the activities of the de novo synthetic enzymes, IMP dehydrogenase, formylglycinamidine ribonucleotide (FGAM) synthase, GMP synthase, amidophosphoribosyl-transferase (AT) and adenylosuccinate synthase were 16, 8, 107, 78 and 124 nmol/hr/mg protein, respectively, and those of the salvage enzymes, adenine, hypoxanthine and guanine phosphoribosyl-transferases, were 340, 407, and 1018 nmol/hr/mg protein, respectively. The sequence of events was elucidated after a single i.p. injection of acivicin (5 mg/kg) or tiazofurin (200 mg/kg). Within 2 hr after acivicin injection, CTP, GMP and FGAM synthases lost 85-90%, while AT and synthase II lost 50 and 80%, respectively, of their activities. The activities rose to near normal range by 72-96 hr. The bone marrow cellularity decreased, reaching a nadir at 24 and 48 hr, and returning to normal range by 72 and 92 hr; thymidine kinase activity followed a similar pattern. Tiazofurin injection depressed IMP dehydrogenase activity to 20% by 2 hr with a rebound to normal range by 48 and 72 hr. The cellularity decreased more slowly, reaching its lowest point at 24 hr and returning to normal range at 72 hr. For acivicin the marked depletion of the activities of the glutamine-utilizing enzymes and for tiazofurin that of IMP dehydrogenase might account, in part at least, for the bone marrow toxicity of these antimetabolites. Because of the presence in the bone marrow of high activities of purine and pyrimidine salvage enzymes, it should be possible to design methods utilizing nucleosides and nucleobases to protect the bone marrow from the action of antimetabolites.

Inactivation of Crithidia fasciculata carbamoyl phosphate synthetase II by the antitumor drug acivicin

In Crithidia fasciculata, carbamoyl phosphate synthetase II, which catalyses the first step of de novo pyrimidine biosynthesis, was separated from aspartate carbamoyltransferase by ammonium sulfate fractionation. The antitumor drug acivicin competitively inhibited the synthetase II activity with respect to L-glutamine, yielding an apparent Ki of 2 microM. In the absence of L-glutamine, acivicin resulted in a selective, time-dependent inactivation of L-glutamine-dependent activity of the enzyme, with an inactivation constant (Kinact) of 100 microM and a minimum inactivation half-time (T) of 0.2 min. L-Glutamine protected the enzyme from inactivation. These results are consistent with a postulate that acivicin is an active site-directed affinity analogue of L-glutamine, achieving irreversible inactivation. The inactivated enzyme retained ammonia-dependent activity. Acivicin stimulated the ammonia-dependent activity by increasing the Vmax value of the enzyme; apparent Km values for ammonia and MgATP were not affected. Differential action of acivicin on the Crithidia and mammalian synthetase II is discussed.

Kinetic properties of carbamoyl-phosphate synthetase II (glutamine-hydrolyzing) in the parasitic protozoan Crithidia fasciculata and separation of the enzyme from aspartate carbamoyltransferase

A high specific activity of carbamoyl-phosphate synthetase II (glutamine-hydrolyzing; EC 6.3.5.5) was demonstrated in extract of the cultured Crithidia fasciculata. The enzyme was separated from aspartate carbamoyltransferase by ammonium sulfate fractionation. Apparent Km for the synthetase for L-glutamine, NH4+, MgATP or bicarbonate was 0.27, 26, 1.7 or 1.7 mM at 2.0% dimethyl sulfoxide plus 0.3% glycerol. 8.6% dimethyl sulfoxide plus 1.4% glycerol decreased Km for L-glutamine to 0.10 mM, while Km for MgATP was unaffected. The higher solvent concentrations made Vmax markedly reduced, yielding the inhibition of the activity. These properties are unique to the Crithidia synthetase, compared with the mammalian enzyme.

Studies on turnover rates of rat gamma-glutamyltranspeptidase after chronic ethanol administration in vivo

Chronic ethanol administration to rats was shown to result in a significant increase of hepatic and serum GGT activities, contrasting to the decreased levels observed in pancreas, intestine, brain, and kidney by the new alcock regimen method. The kinetics of rat GGT synthesis and degradation in vivo among the different sources after chronic ethanol administration has been studied by use of acivicin, which irreversibly inactivates GGT. The comparison of kinetics of GGT return after acivicin injection showed that the kidney and serum GGT exhibits biphasic half-lives in contrast to liver, pancreatic, intestinal, and brain GGT half-lives in chronic ethanol-administered rats. The present studies on kinetics of GGT synthesis (Ks) and degradation (Kd) in vivo would seem to indicate the existence of three types of systems. That is, Ks rather than Kd may be preferential in liver and serum whereas Kd is apparently increased in kidney and intestine without noticeable change in Ks. The reverse phenomenon is also observed for pancreas and brain. These findings suggest that the contributions of alterations in the rates of GGT synthesis and degradation to changing levels of GGT have been evaluated as a mechanism for enzyme adaptation in animal tissues as a change from the control diet to the ethanol diet.

Hepatic carbamoyl phosphate metabolism. Role of cytosolic and mitochondrial carbamoyl phosphate in de novo pyrimidine synthesis

The interrelationship between the two carbamoyl phosphate pools in intact hepatocytes and intact liver was studied with respect to de novo pyrimidine synthesis by use of selective inhibitors of the mitochondrial and the cytosolic carbamoyl-phosphate synthetase. Inhibition of mitochondrial carbamoyl phosphate synthesis by 4-pentenoate was without effect on galactosamine-stimulated pyrimidine synthesis. Conditions favouring mitochondrial carbamoyl phosphate accumulation, like excess ammonium ions or L-norvaline, led to an increase in pyrimidine synthesis bypassing the feedback inhibition of cytosolic carbamoyl-phosphate synthetase by UTP. A stimulation of pyrimidine synthesis was not observed when the carbamoyl phosphate accumulation was due to aspartate deficiency in the presence of aminooxyacetate. The full response of pyrimidine synthesis to excess ammonium ions was restored, even in the presence of aminooxyacetate, when aspartate was substituted. This is explained by an inhibition of aspartate carbamoyltransferase flux [in view of the Km (aspartate = 0.7 mmol/l) of this enzyme] resulting from a 90% decrease in aspartate tissue levels. Acivicin, the inhibitor of cytosolic carbamoyl-phosphate synthetase, completely abolished the galactosamine-induced stimulation of pyrimidine synthesis, but was without effect on the stimulation of pyrimidine synthesis by ammonium ions and L-norvaline. It is concluded that experimental changes in mitochondrial carbamoyl phosphate content exert effects on de novo pyrimidine synthesis; however, it is considered unlikely that relevant amounts of mitochondrial carbamoyl phosphate are used for pyrimidine synthesis under physiological conditions. In addition the data point to a potential regulatory role of aspartate in hepatic pyrimidine synthesis.

In vivo inactivation of formylglycinamidine ribonucleotide synthetase in rat hepatoma

The antitumor drug acivicin, L-(alpha S,5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid, irreversibly inactivated in vivo formylglycinamidine ribonucleotide synthetase (FGAM synthetase, EC 6.3.5.3) in transplantable rat hepatoma 3924A while the activity in host liver remained unchanged. At acivicin doses of 1.0 and 5.0 mg/kg body weight, enzyme activity in the hepatoma decreased to 26 and 5%, respectively, after 2 hr. The activity of the in vivo inactivated hepatoma 3924A enzyme could not be restored by gel filtration or 40 hr of dialysis. In the absence of L-glutamine, acivicin in vitro inactivated both liver and hepatoma FGAM synthetase in a time-dependent fashion, with an inactivation constant Kinact = 66 microM and a minimum inactivation half-time T = 1.0 min. In the presence of L-glutamine, competitive inhibition was observed with a Ki = 5 microM. Protection against in vitro inactivation was observed in the presence of 1 mM L-glutamine, suggesting that L-glutamine concentrations are important in the selective toxicity of acivicin on hepatoma cells in vivo. Irreversible inhibition of FGAM synthetase by acivicin is consistent with the view that this antibiotic is an active site-directed affinity analog of L-glutamine and indicates that this enzyme is a sensitive target of acivicin action.

The glutamine analog acivicin as antipyrimidine. Studies on the interrelationship between pyrimidine and urea synthesis in liver

The inhibition of cytosolic carbamoyl-phosphate synthetase II by acivicin was used to study the role of the cytosolic carbamoyl phosphate pool as the exclusive substrate source for de novo pyrimidine synthesis in rat hepatocytes. De novo pyrimidine synthesis was stimulated: 1. by uridine triphosphate deficiency (incubation with D-galactosamine) leading to a stimulation of cytosolic carbamoyl phosphate synthesis, and 2. by accumulation and efflux of mitochondrial carbamoyl phosphate (incubation with ammonium ions and L-norvaline). The stimulated orotate formation from cytosolic carbamoyl phosphate in UTP depleted cells was completely blocked by acivicin. It was not influenced by an inhibition of mitochondrial carbamoyl phosphate synthesis mediated by 4-pentenoate, since mitochondrial carbamoyl phosphate did not participate in cytosolic pyrimidine synthesis even in the presence of ammonium ion concentrations maintaining physiological rates of urea synthesis. An excess of ammonium ions led to an artificial accumulation and efflux of mitochondrial carbamoyl phosphate, which could be avoided by 4-pentenoate. The non-regulated stimulation of pyrimidine synthesis from surplus mitochondrial carbamoyl phosphate was not inhibited by acivicin. Utilization of mitochondrial carbamoyl phosphate for de novo pyrimidine synthesis presumably does not occur under physiological conditions because mitochondrial CP efflux depends on the accumulation of this metabolite in the mitochondria under experimental or pathological circumstances. Acivicin inhibition of CPS II thus cannot be bypassed by mitochondrial CP. It is suitable as inhibitor of the physiological de novo pyrimidine synthesis.

Inactivation by acivicin of rat brain CTP and GMP synthetases and depression of CTP and GTP concentrations

Evidence was provided that injection of acivicin (25 mg/kg, i.p.) into the rat inactivated brain CTP and GMP synthetases. Under the same circumstances, CTP and GTP concentrations in the rat brain decreased following the decline in the activities of CTP and GMP synthetases. The decrease in enzymic activities and nucleotide concentrations progressed with time. The decline in CTP and GMP synthetase activities and CTP and GTP concentrations caused by acivicin occurred more slowly and to a lesser extent than in liver and hepatoma 3924A. The delay in the expression of acivicin action in the rat brain was attributed to a possible slower entrance of acivicin and the lower concentration than might have been attained in the rat brain. These considerations are based on the rapid disappearance of acivicin from rat plasma noted earlier. The decline in CTP concentration in rat brain might interfere with neuronal function. The decline in GTP concentration might be expressed through the depletion of biopterins which are generated from GTP in the brain. The possible relevance to the biochemical basis of paranoid schizophrenia which occurs reversibly after high-dose acivicin or tiazofurin treatment was discussed.

Enzyme targets of antiglutamine agents in cancer chemotherapy

The modes of action of azaserine and acivicin were compared. The results were evaluated by assessing the impact of these drugs on primary targets, the activities of key enzymes, and on secondary and tertiary targets, the concentrations of pools of ribonucleotides and deoxyribonucleotides. It was observed that both drugs act as competitive inhibitors for glutamine-utilizing enzymes involved in the biosynthesis of purines and pyrimidines, but in addition acivicin exerts a direct inactivating effect (probably by alkylation) on the enzymes. The different tissues examined displayed varying sensitivity to the drugs which may be attributed in part at least to the tissue glutamine content. Acivicin markedly depleted the CTP pools, but ATP and UTP were unaffected. It also decreased the concentration of all 4 deoxynucleoside triphosphates. These biochemical targets serve as indicators of acivicin action in cancer cells and should also be helpful in the design of combination chemotherapy. On the basis of the biochemical action of acivicin, actinomycin and dipyridamole were selected for testing in combination chemotherapy. Both drugs acted synergistically with acivicin.

Inactivation by acivicin of carbamoyl-phosphate synthetase II of human colon carcinoma

The effect of the anti-tumor, anti-glutamine drug acivicin, L-(alpha S,5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid, was determined on the activity of the rate-limiting enzyme of de novo pyrimidine biosynthesis, carbamoyl-phosphate synthetase II (glutamine-hydrolyzing) (EC 6.3.5.5), in human colon carcinoma. The synthetase II activity in human colon carcinoma was elevated 2- to 3-fold over values of the normal colon mucosa, and the substrate kinetic constants were similar for the enzyme in normal and neoplastic colon. The Km for glutamine was 17 microM (colon carcinoma) and 23 microM (normal mucosa), whereas the Km for ATP was 2.1 and 1.7 mM in tumor and mucosa respectively. The synthetase II activity in colon carcinoma was inhibited to a similar extent by UMP, UDP and UTP (36-41%). The three uracil nucleotides were also equally effective in inhibiting the enzyme from normal mucosa (39-46%). Both enzymes were activated by PRPP (63 and 57%) in mucosa and carcinoma respectively. Acivicin in vitro selectively inactivated the glutamine-dependent synthetase II from human colon carcinoma, and it did not affect the ammonia-dependent activity. The acivicin inactivation constant (Kinact) was 100 microM, and the minimum inactivation half-time (T) was 0.7 min. Acivicin most likely exerts its effect against human colon synthetase II by acting as an active site directed affinity analogue of L-glutamine.

Control of enzymic programs and nucleotide pattern in cancer cells by acivicin and tiazofurin

The mechanism of action of acivicin and tiazofurin was compared in hepatoma 3924A. The results were evaluated by assessing the impact of these drugs on primary targets, the activities of key enzymes, and on secondary and tertiary targets, the concentrations of pools of ribonucleotides and deoxyribonucleotides. The action of acivicin entails inhibition and inactivation of the key enzymes of glutamine utilization in the biosynthesis of purines and pyrimidines. As a result, the GTP and CTP pools were markedly depleted, whereas those of ATP and UTP were unaffected. Acivicin also markedly decreased the concentrations of all 4 deoxynucleoside triphosphates. The nucleotide pools returned to normal or near normal range within 2 to 3 days after a single acivicin injection. The pharmacologic targets of acivicin in anticancer chemotherapy include prominently the activities of glutamine-utilizing enzymes and the pools of GTP and CTP and all 4 dNTP's. These biochemical targets also serve as indicators of acivicin action in cancer cells. The action of tiazofurin in hepatoma cells entails the primary target, IMP dehydrogenase. The subsequent effects include marked enlargement of IMP and PRPP pools and depletion of the pools of GDP and GTP. The increased IMP concentration selectively inhibited the activities of hypoxanthine-guanine phosphoribosyltransferase, but did not affect that of adenine phosphoribosyltransferase. The markedly decreased GTP pool de-inhibited the activity of AMP deaminase which permitted the channeling of AMP to IMP. An important indicator of tiazofurin action is the prolonged depletion of dGTP pools and similar but less pronounced declines in the pools of dCTP and dATP. In contrast, dTTP pools were increased. The crucial biochemical targets and indicators of tiazofurin action in sensitive cancer cells include inhibition of IMP dehydrogenase, a decrease in the concentrations of GDP, GTP, dGTP, dCTP, dATP and marked rise in the pools of IMP, PRPP and dTTP. Measurements of the molecular targets and indicators of drug action should be helpful in identifying cancer cells and tissues sensitive or resistant to the action of acivicin or tiazofurin. Identification of the targets and indicators should also be helpful in the design of frequency of administration of the drugs in combatting animal and human neoplasia.

Combined action of acivicin and D-galactosamine on pyrimidine nucleotide metabolism in hepatoma cells

The glutamine antagonist acivicin, L-(alpha S, 5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid, strongly reduced CTP and GTP contents in AS-30D rat hepatoma cells in suspension. UTP only dropped to 63% of the respective control after 4 hr; however, by combining acivicin with the uridylate-trapping sugar analogue D-galactosamine, a synergistic decrease in UTP contents to 7% of control was induced. Incorporation of 14CO2 into purine and pyrimidine nucleotides followed by radio-high performance liquid chromatography showed marked inhibition of purine and pyrimidine biosynthesis de novo; the latter was reduced to 35% of control. The inhibitory potency of acivicin on glutamine-dependent carbamoyl-phosphate synthetase and consequently on de novo uracil nucleotide formation was also reflected by the complete suppression of the D-galactosamine-induced rise in total uridylate. Induction of UTP deficiency by interference with the first and rate-limiting step in pyrimidine biosynthesis de novo together with a trapping of uridylate by D-galactosamine may provide a promising approach to the chemotherapy of hepatocellular carcinoma.

Salvage capacity of hepatoma 3924A and action of dipyridamole

The role and behavior of the salvage enzymes in the biosynthesis of purines (adenine and hypoxanthine-guanine phosphoribosyltransferases) and pyrimidines (uridine-cytidine, deoxycytidine and thymidine kinases) were elucidated. In liver purine metabolism the transferase activities were orders of magnitude higher than the activities of the enzymes of de novo biosynthesis. In both purine and pyrimidine biosynthesis the activities of the enzymes of the de novo pathways were low (23 pmol to 70 nmol/hr/mg protein), whereas those of salvage synthetic pathways ranged from 0.8 to 1,470 nmol/hr/mg protein. In purine metabolism the salvage enzymes had markedly higher affinity to the shared substrate PRPP (4 to 40 microM) than the rate-limiting enzyme of de novo synthesis, amidophosphoribosyltransferase (900 microM). In rapidly growing hepatoma 3924A the activities of the enzymes of de novo purine biosynthesis increased, whereas those of the salvage pathway changed little. However, the activities of the enzymes of the salvage pathways remained much higher than those of the enzymes of de novo purine production. In pyrimidine production in the hepatomas the activities of both de novo and salvage enzymes markedly increased. However, the activities of the salvage enzymes far outstripped those of the enzymes of the de novo pathways. To inhibit the operation of the salvage pathways, the action of the transport inhibitor, dipyridamole, was examined. In tissue culture, dipyridamole inhibited the transport of purine and pyrimidine nucleosides with an IC50 of 10(-6) or 10(-7) M. As measured by colony-forming assay, dipyridamole killed hepatoma cells with an IC50 of 20 microM. Dipyridamole markedly depressed the pools of ATP, GTP, CTP and UTP; in combination chemotherapy with acivicin, an anti-glutamine agent, synergistic action was observed on the pools of nucleotides in hepatoma 3924A in vivo. These investigations emphasize the importance of the capacity to utilize precursors by the salvage enzymes and may explain, in part at least, the failure of inhibitors of the de novo pathways to yield lasting chemotherapeutic results. Combination chemotherapy of inhibitors of the de novo pathways with an inhibitor of the salvage pathways (dipyridamole) should impact on our understanding of the contribution of salvage pathways and provide a rational basis for successful combination chemotherapy of neoplastic diseases.

Biochemical pharmacology of acivicin in rat hepatoma cells

The antiglutamine agent acivicin, L-(alpha S,5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid, inhibited the growth of hepatoma 3924A cells in culture. After 7 days of incubation with the drug, an LC50 of 1.4 microM was observed by determination of colony forming ability. A combination of cytidine (1 mM), deoxycytidine (10 microM) and guanosine (10 microM) completely protected the hepatoma cells against the cytotoxic action of acivicin, but each nucleoside by itself had no effect. Acivicin (0.1 mM) inhibited the incorporation of uridine and thymidine into macromolecules, but not that of leucine. Acivicin depressed the pools of CTP, GTP, dCTP, dGTP and dTTP to 46, 62, 40, 64 and 53%, respectively, but it increased UTP level to 152% of the values of untreated cancer cells. The activity of a highly purified CTP synthetase (EC 6.3.4.2) from rat liver and hepatoma 3924A was inhibited by acivicin. The inhibition was competitive with respect to L-glutamine, and the Ki values with liver and hepatoma enzymes, determined by Dixon and reciprocal plots, were 1.1 and 3.6 microM respectively. The hydroxy analog of acivicin was also a competitive inhibitor, but it was less effective than acivicin, with a Ki value of 1.8 mM for the hepatoma enzyme. Our observations on the impact of acivicin on the behavior of pools of ribonucleotides and deoxyribonucleotides and the competitive inhibition of purified CTP synthetase from hepatoma cells suggest that a major mechanism of action for this drug is the inhibition of CTP synthetase and GMP synthetase (EC 6.3.5.2).

Rapid in vivo inactivation by acivicin of CTP synthetase, carbamoyl-phosphate synthetase II, and amidophosphoribosyltransferase in hepatoma

A single injection of the anti-glutamine drug, acivicin (NSC 163501), in tumor-bearing rats in 30 min decreased the activities of amidophosphoribosyltransferase, carbamoyl-phosphate synthetase II and CTP synthetase to 56, 50, and 7% of those of the controls. By 1 hr the activities were down to 32, 13 and 3% and they remained low for 12 hr, after which they slowly returned towards normal range in 72 hr. The decline of the activity of CTP synthetase (a loss of 80% in 10 min) was the most rapid, and the activity only returned to 60% of the controls by 3 days after the acivicin injection. In the hepatoma the concentrations of ATP and UTP changed little, but those of GTP and CTP rapidly decreased, reaching at the lowest point 32 and 2%, respectively, of control values 2 hr after acivicin; concentrations started to rise at 12 hr, reaching normal levels by 48 hr. The drop in enzyme activities preceded the decline in the pools of GTP and CTP. The behavior of enzyme activities and nucleotide concentrations in the host liver had a pattern similar to that in the hepatoma; however, the changes were less extensive than those in the tumor. The differential response between tumor and liver is attributed, in part at least, to the tissue L-glutamine concentration which in the hepatoma (0.5 mM) was 9 times lower than in the liver (4.5mM). The selectivity of acivicin action in inhibiting glutamine-utilizing enzymes is also demonstrated by the lack of effect on aspartate carbamoyltransferase, an enzymic activity which resides in the same complex as that of carbamoyl-phosphate synthetase II. The rapid decline in the activities of glutamine-utilizing enzymes is attributed to an inactivation of the enzymes by acivicin which functions as an active sitedirected affinity analog of L-glutamine. The rapid modulation of the enzymic phenotype and ribonucleotide concentrations by acivicin provides a useful tool for elucidating the role of enzymic and nucleotide imbalance in the commitment of cancer cells to replication and in the targeting of anticancer chemotherapy.

Multi-enzyme-targeted chemotherapy by acivicin and actinomycin

On the basis of our observation of the increased specific activities of glutamine-utilizing enzymes in purine and pyrimidine metabolism in hepatoma 3924A, and because the concentration of glutamine is ten times lower in the hepatomas than in the liver, the biochemical pharmacology of the anti-glutamine agent, acivicin, was examined. (1) Acivicin competitively inhibited the activities of amidophosphoribosyl-transferase, CTP synthetase and carbamoyl-phosphate synthetase II from extracts of liver and hepatoma 3924A. (2) In addition to the competitive inhibition exerted by acivicin, evidence was obtained that this drug also irreversibly inactivated in vitro the glutamine-utilizing enzymes. It is particularly relevant for the selectivity of acivicin that the activity of aspartate carbamoyltransferase, an enzyme present in the same complex as carbamoyl-phosphate synthetase II, was not affected by the anti-glutamine agent. (3) Acivicin in vivo brought down the activities of glutamine-utilizing enzymes in a period of 10 min to 1 hr after injection. CTP synthetase activity declined to less than 10% of that observed in the uninjected rats. The decreases were not reversible by various in vitro methods, but in vivo the activities returned to normal range in 72 hr. (4) The activity of aspartate carbamoyltransferase, which exists as a multi-enzyme complex with synthetase II, was not altered by acivicin injection. Similar results were observed in transplantable sarcoma in the rat. (5) The acivicin-induced decrease in enzymic activities could not be restored by purification of the enzymes. (6) In vitro studies indicated that addition of acivicin to liver or hepatoma extracts or purified enzymes rapidly decreased enzymic activities; the activities could not be restored. These results are consistent with an interpretation that acivicin acts either as a tight-binding inhibitor or as an inactivator through alkylation of the enzymes of glutamine utilization. (7) Acivicin in combination with actinomycin provided a synergistic kill of hepatoma cells in tissue culture and also inhibited the growth of transplantable solid hepatoma 3924A in the rat. (8) The synergistic biological results of combination chemotherapy with acivicin and actinomycin can be accounted for by the action of acivicin in inhibiting GMP and CTP synthetases, resulting in a decrease in GTP and CTP content, and by the actinomycin-caused inhibition of RNA polymerase in selectively blocking the utilization of GTP and CTP.