THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Cytotoxic Mechanisms of Glutamine Antagonists in Mouse L 12 10 Leukemia*

Vol. 265, No. 19, Issue of July 5, pp. 11377-11381, 1990 Printed in U.S. A.

(Received for publication, January 29, 1990)

Stephen D. Lyons, Melissa E. Sant, and Richard I. ChristophersonS From the Department of Biochemistry, University of Sydney, Sydney, New South Wales 2006, Australia

The glutamine antagonists, acivicin (NSC 163501), azaserine (NSC 742), and 6-diazo-5-oxo-L-norleucine (DON) (NSC 7365), are potent inhibitors of many glu- tamine-dependent amidotransferases in vitro. Experi- ments performed with mouse L1210 leukemia growing in culture show that each antagonist has different sites of inhibition in nucleotide biosynthesis. Acivicin is a potent inhibitor of CTP and GMP synthetases and partially inhibits N-formylglycineamidine ribotide (FGAM) synthetase of purine biosynthesis. DON inhib- its FGAM synthetase, CTP synthetase, and glucosa- mine-6-phosphate isomerase. Azaserine inhibits FGAM synthetase and glucosamine-6-phosphate isomerase. Large accumulations of FGAR and its di- and triphos- phate derivatives were observed for all three antago- nists which could interfere with the biosynthesis of nucleic acids, providing another mechanism of cytotox- icity. Acivicin, azaserine, and DON are not potent in- hibitors of carbamyl phosphate synthetase II (gluta- mine-hydrolyzing) and amidophosphoribosyltransfer- ase in leukemia cells growing in culture although there are reports of such inhibitions in vitro. Blockade of de nouo purine biosynthesis by these three antagonists results in a “complementary stimulation” of de nouo pyrimidine biosynthesis.

L-Glutamine provides the amino group for biosynthesis of a variety of metabolites including the purine intermediates and derivatives, 5-phosphoribosylamine, N-formylglycineam- idine (FGAM),l GMP, NAD, and NADP, and the pyrimidine intermediates carbamyl phosphate and CTP, and glucosamine 6-phosphate (GlcN-6-P). The L-glutamine antagonists, aciv- icin, azaserine, and DON, have anti-cancer activity attributed

*This work was supported by Project Grant 880119 from the National Health and Medical Research Council, Grant A08415281 from the Australian Research Council, and Grants AI183 and AK123 from the Utah Foundation for HPLC detectors. The costs of publi- cation of this article were defraved in part bv the pavment of page _ _ __ - - charges. This article must therefore be hereby marked “advertise- ment” in accordance with 18 U.S.C. Section 1734 solelv to indicate this fact.

EXPERIMENTAL PROCEDURES AND RESULTS3

$ To whom correspondence should be addressed. ’ The abbreviations used are: FGAM. N-formvlplvcineamidine ri-

DISCUSSION

botide; P-Rib-PP, 5-phosphoribosyl ‘l-pyropgo&hate; lo-CHO- Metabolic crossover points have been used to define reac- H,folate. A”‘-formvl tetrahvdrofolate: FGAR. N-formvlglvcineamide tions of nucleotide metabolism in growing leukemia cells

” - ”

ribotide; AIR, 5-aminoimidazole ribotide; SAICAR, N-succino-5-ami- noimidazole-4-carboxamide ribotide; AICAR, 5-aminoimidazole-4- carboxamide ribotide; sAMP, adenylosuccinate; Ord, orotidine; Oro, orotate; OMP, orotidine 5’-monophosphate; amido PRTase, amido- phosphoribosvltransferase; CPSase, carbamvl phosphate svnthetase i1 (gcutaminehydrolyzing); DON, 6-diazo-5-dxolr,-norleuciie; FGAR- DP/TP, di- and triphosphate derivatives of FGAR: Henes. 4-(2- hydroxyethyl)-l-piperazineethanesulfonic acid; HPLC; high pies&e liquid chromatography; X,.,, wavelength of maximal ultraviolet ab- sorbance.

to potent irreversible inhibition of amidotransferases involved in nucleotide biosynthesis (Livingston et al., 1970; Tso et al., 1980). Most of these amidotransferases are inhibited by all three glutamine antagonists in uitro. For example, Jayaram et al. (1975) found that of the seven amidotransferases in- volved in nucleotide metabolism, amidophosphoribosyltrans- ferase, GMP synthetase, carbamyl phosphate synthetase II (glutamine hydrolyzing) were not significantly affected by azaserine while FGAM synthetase, CTP synthetase, GlcN-6- P isomerase,2 and NAD synthetase were inhibited. DON inhibited all seven amidotransferases in uitro; acivicin inhib- ited six of the seven amidotransferases by more than 90% while amidophosphoribosyltransferase was inhibited by only 10%. Carbamyl phosphate synthetase II, CTP synthetase, amidophosphoribosyltransferase, FGAM synthetase, and GMP synthetase from rat hepatoma are potently inhibited by acivicin in uiuo when enzymic activities are subsequently assayed in cell extracts (Aoki et al., 1982; Weber et al., 1984; Elliott and Weber, 1985).

Inhibition of purified target enzymes in vitro and inactiva- tion of amidotransferases in uiuo with subsequent assay of enzymic activities in cell extracts determine all susceptible reactions but do not discriminate between possible mecha- nisms of cytotoxicity in uiuo. An antagonist may inhibit several amidotransferases in a pathway, but blockade of the most proximal reaction provides the primary cytotoxic effect; flux through the pathway to subsequent reactions is blocked and inhibition at distal amidotransferases would then be of minor significance. Data presented in this paper show that acivicin, DON, and azaserine each have unique modes of action in growing leukemia cells. Susceptible amidotransfer- ases for each antagonist in uiuo were identified as metabolic crossouerpoints, where cellular metabolites prior to the block- ade accumulate and subsequent metabolites deplete relative to those in untreated cells (Rolleston, 1972; Rognstad, 1979; Christopherson and Duggleby, 1983). The methodology de- scribed here can be generally applied to the determination of cytotoxic mechanisms of inhibitors of de nouo nucleotide biosynthesis in growing cells.

* GlcN-6-P isomerase catalyzes the conversion of Fru-6-P + GlcN- 6-P which is a precursor for the svntheses of UDP-GlcNAc and UDP- GalNAc: GlcG-6-P + GlcNAc-6-P ++ GlcNAc-1-P + UDP-GlcNAc ++ UDP-GalNAc.

3 Portions of this paper (including “Experimental Procedures,” “Results,” and Figs. 1 and 4) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

11377

by guest, on February 21, 2013

ww

w.jbc.org

Dow

nloaded from

Inhibition of Amidotransferases

25 ,#A ACIVICIN

NAD I

AMP

UMP j

AICAFI

IMP

UDphugars

NADP

CDP UDP

ADP

sAMP

SAICAR

FIG. 2. “Elution surfaces” showing effects of acivicin on leukemia cells. Two cultures (55 ml) were grown in the presence of [“Clformate (50 PM), and one was exposed to acivicin (25 pM) for 4 h. Metabolites from 50-ml samples were extracted and analyzed by

which become rate-limiting in the presence of a glutamine antagonist leading to cytotoxicity. These procedures enable determination in growing cells of an inhibited reaction which has substrates and products with significant ultraviolet ab- sorption (Figs. 1 and 2) and/or which incorporate 14C from a radiolabeled precursor (Figs. 3 and 4). At least two significant metabolic crossover points or sites of inhibition are induced by each glutamine antagonist investigated (Table II). Accu- mulation of FGAR and depletion of adenine and guanine nucleotides are consistent with inhibition of FGAM synthe- tase by acivicin, azaserine, and DON. Significant incorpora- tion of [Ylbicarbonate into 4-carboxy&aminoimidazole ri- botide and subsequently into ATP, after addition of acivicin (data not shown), indicates that this antagonist is not a potent inhibitor of FGAM synthetase in growing cells although El- liott and Weber (1985) demonstrated potent inhibition of FGAM synthetase by acivicin in vitro in assays of extracts from drug-treated rat hepatoma. The levels of FGAR deriva- tives accumulated may be a result of the relative proportions of FGAM synthetase and amidophosphoribosyltransferase in- activated by each antagonist. Accumulation of higher phos- phorylated derivatives of FGAR in response to azaserine treatment has been reported (Bennett and Smithers, 1964; Rosenbloom et al., 1968; Rowe et al., 1978); thin layer chro- matography was used for separating purine precursors, and the more highly charged “FGAR polyphosphates” remained unresolved at or near the origin of chromatograms after development with solvent. We have identified and quantified the di- and triphosphate derivatives of FGAR. Four hours after exposure to 25 pM azaserine, FGAR triphosphate reaches a cellular concentration of 9.5 mM (Fig. 4),3 which is nearly 3-fold higher than the normal concentration of ATP in mam- malian cells (-3.5 mM). The metabolic consequences of these higher phosphorylated forms of a natural purine intermediate are not known. However, it is known that azaserine induces morphological irregularities in the genetic material of malig- nant and normal cells with associated increases in cellular size and DNA content (Hansen and Vandevoorde, 1966; Liv- ingston et al., 1970). Incorporation of FGAR triphosphate into nucleic acids or direct inhibition of RNA or DNA polymerases by FGAR derivatives may be additional mechanisms of cyto- toxicity for all three glutamine antagonists. The greater ac- cumulation of FGAR derivatives induced by azaserine (Figs. 3 and 4) may lead to the greater host toxicity displayed by this antagonist (Catane et al., 1979; Earhart and Neil, 1985), possibly by interference of these metabolites with RNA syn- thesis in nondividing cells.

Accumulation of N-succino-5-aminoimidazole-4-carbox- amide ribotide, 5-aminoimidazole-4-carboxamide ribotide, IMP, and XMP and depletion of guanine nucleotides defines a metabolic crossover point at GMP synthetase in acivicin- treated leukemia cells (Figs. l-3). No [Ylbicarbonate is incorporated into GTP in the presence of acivicin indicating that inhibition of GMP synthetase is complete and the pri- mary site of inhibition in purine biosynthesis (data not shown). Because DON and azaserine are potent inhibitors of FGAM synthetase, purine intermediates distal to the blockade disappear and inhibition of GMP synthetase could not be observed. Purified GMP synthetase is totally inhibited by DON (Jayaram et al., 1975), but the site of inhibition most

HPLC. The representation of data as elution surfaces is described in the text. The full-scale deflection is 0.005 absorbance unit, and the colored contours have the following absorbance ranges: block, >O.OOO; cyan, >0.002; blue, >0.004; green, >0.006; yellow, >0.008; red, >O.OlO; magenta, >0.012.

by guest, on February 21, 2013

ww

w.jbc.org

Dow

nloaded from

Inhibition of Amidotransferases 11379

30 60 90 Time (min)

FIG. 3. Effects of glutamine antagonists on metabolites of leukemia cells which incorporate [Wlformate. Cultures (55 ml) were grown in the presence of [‘%]formate from 24 h after inoculation at a density of 4 X lo4 cells/ml. Cells were exposed to drug (25 pM) for 4 h, and metabolites from 50-ml samples were extracted and analyzed by HPLC. The eluate was monitored concurrently by ultra- violet absorbance and “C detectors; only 14C profiles are shown. A, B, and C are metabolites with ultraviolet absorbance which incorpo- rate [“Clformate and are unidentified. a, control; b, acivicin; c, DON; d, azaserine.

TABLE I

Effects of glutamine antagonists on nucleotide levels in leukemia cells Cultures (55 ml) were grown in the presence of [Wlformate (50

pM) and exposed to the indicated glutamine antagonist (25 pM) for 4 h. Cellular metabolites were extracted from a 50-ml culture sample and analyzed by HPLC as described under “Experimental Proce- dures.” Peaks representing nucleotides of interest in absorbance profiles, corresponding to the radioactivity profiles of Fig. 3, were integrated using Nelson software, and levels are expressed relative to untreated cells.

Nucleotide

ATP ADP GTP GDP UTP UDP CTP CDP NAD NADP UDP-Glc/Gal UDP-GlcNAc/GalNAc

Glutamine antagonist

Acivicin DON Azaserine

% of control/cell 86 39 29 78 48 46 52 35 21 58 61 49

180 290 67 140 190 66

3.9 4.5 50 67 14 80 94 97 68 95 99 66

160 350 100 170 50 34

proximal in the pathway induces the cellular deficiencies of nucleotides.

A general phenomenon has been observed for many inhib- itors of nucleotide biosynthesis; potent inhibition of either the de novo pyrimidine or purine pathway is accompanied by “complementary stimulation” of flux through the unaffected pathway (Lyons, 1989). Blockade of the de nouo pyrimidine pathway by dichloroallyl lawsone (NSC 126771; Kemp et aL., 1986) or pyrazofurin (Sant et al., 1989a) leads to complemen- tary stimulation of purine biosynthesis, particularly guanine nucleotides, which may be attributed to sparing of &phospho- ribosyl I-pyrophosphate from the orotate phosphoribosyl-

. . TABLE II

Inhibition of amidotransferases in growing leukemia cells by glutamine antagonists

Sites of inhibition were determined by comparison of metabolite levels prior to and 4 h after addition of drug to define a metabolic crossover point; - -, potent inhibition; -, partial inhibition; and o, no inhibition. AmidoPRTase, amidophosphoribosyltransferase; CPSase, carbamyl phosphate synthetase II (glutamine hydrolyzing).

Glutamine antagonist Amidotransferase

Acivicin DON Azaserine

AmidoPRTase 0 -0 0 FGAM synthetase -- GMP synthetase 0 0 NAD synthetase 0 0 0 CPSase 0 0 0 CTP synthetase -- 0 GlcN-6-P isomerase 0 -- -

u Partial inhibition of amidophosphoribosyltransferase by DON is consistent with the reduced accumulation of FGAR derivatives rela- tive to azaserine although inhibition of FGAM synthetase is complete for both antagonists.

transferase reaction and of glutamine from the CTP synthe- tase reaction. All three glutamine antagonists block the purine pathway (Table II) and induce accumulation of UTP relative to the level of ATP after drug treatment (Fig. 1, Table I). Acivicin and DON also inhibit CTP synthetase (Table II) which would enhance accumulation of UTP. Blockade of the purine pathway at FGAM synthetase or GMP synthetase would spare cellular glutamine from utilization in purine biosynthesis, and inhibition of amidophosphoribosyltransfer- ase would spare both glutamine and 5-phosphoribosyl l- pyrophosphate. Accumulation of these metabolites could stimulate pyrimidine biosynthesis, augmenting the accumu- lation of UTP induced by acivicin or DON and causing the relative increases in UTP and CTP induced by azaserine (Fig. Id), which does not inhibit CTP synthetase (Table II). While the physiological significance of complementary stimulation is yet to be determined, unbalanced nucleotide pools may lead to genetic miscoding. For example, methotrexate (NSC 740) induces incorporation of dUTP into DNA (Goulian et al., 1980). Increased levels of UTP accumulated in response to acivicin or DON treatment (Fig. 1) may be misincorporated into RNA.

Acivicin and DON block both de nouo purine and pyrimi- dine biosynthesis in growing cells, but different amidotrans- ferases are affected. Acivicin blocks progression of the cell cycle in G, or early S phase (Jayaram et al., 1975; Thornwaite and Allen, 1980), which suggests that RNA and/or DNA synthesis is affected. DON induces S phase arrest (Huber et al., 1987), consistent with inhibition of DNA synthesis. DON and acivicin both induce accumulation of UTP and depletion of ATP, GTP, and CTP (Fig. 1) although the depletion of ATP in acivicin-treated cells is minor compared with the corresponding depletion induced by DON (Table I). Acivicin has shown more promise than DON as an anti-cancer drug perhaps due to less accumulation of FGAR triphosphate, which may be generally toxic, and maintenance of ATP levels required for metabolic functions of normal resting (Go phase) cells.

Acivicin, DON, and azaserine exert their cytotoxic effects via inhibition of different groups of glutamine amidotransfer- ases (Table II); in contrast to results obtained in uitro (Jay- aram et al., 1975), amidophosphoribosyltransferase, carhamyl phosphate synthetase II, and NAD synthetase are not major sites of inhibition in growing cells. These antagonists are still being evaluated as potential anti-cancer agents long after their discovery. A detailed knowledge of their mechanisms of

by guest, on February 21, 2013

ww

w.jbc.org

Dow

nloaded from

11380 Inhibition of Amidotransferases

action against growing cancer cells should aid their use, per- haps in combination with glutaminase-asparaginase (Holcen- berg, 1979), as chemotherapeutic agents. Such knowledge may direct the syntheses of new glutamine antagonists which are enzyme-specific; a potent inhibitor of CTP synthetase has recently been synthesized which does not affect other ami- dotransferases (Rabinovitz and Fisher, 1988). Such specific inhibitors would enable precise control of nucleotide metab- olism in cancer cells giving the maximum chemotherapeutic effect.

REFERENCES

Aoki, T., Sebolt, J., and Weber, G. (1982) Biochem. Pharmacol. 31, 927-932

Bennett, L. L., Jr., and Smithers, D. (1964) Biochem. Pharmacol. 13, 1331-1339

Catane, R., Von Hoff, D. D., Glaubiger, D. L., and Muggia, F. M. (1979) Cancer Treat. Rep. 63, 1033-1038

Christopherson, R. I., and Duggleby, R. G. (1983) Eur. J. Biochem. 134,331-335

Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M. (1986) Data for Biochemical Research, 3rd Ed., p. 426, Clarendon Press, Oxford

Earhart, R. H., and Neil, G. L. (1985) Adu. Enzyme Regul. 24, 179- 205

Elliott, W. L., and Weber, G. (1985) Biochem. Pharmacol. 34, 243- 248

Goulian, M., Bleile, B. M., and Tseng, B. Y. (1980) Proc. Nutl. Acud. Sci. U. S. A. 77.1956-1960

Hansen, H. J., and Vandevoorde, J. P. (1966) Proc. Sot. Exp. Biol. Med. 123,209-212

Holcenberg, J. S. (1979) Cancer Treat. Rep. 63, 1109-1114 Huber, K. R., Mayer, E. P., Mitchell, D. F., and Roberts, J. (1987)

Br. J. Cancer 65,653-656 Jayaram, H. N., Cooney, D. A., Ryan, J. A., Neil, G., Dion, R. L., and

Bono, V. H. (1975) Cancer Chemother. Rep. 69,481-491 Kemp, A. J., Lyons, S. D., and Christopherson, R. I. (1986) J. Biol.

Chem. 261,14891-14895 Livingston, R. B., Venditti, J. M., Cooney, D. A., and Carter, S. K.

(1970) Adu. Phurmucol. Chemother. 8, 57-120 Lyons, S. D. (1989) Inhibition of de Novo Nucleotide Biosynthesis in

Mouse LI210 Leukemia. Ph.D. thesis, University of Sydney Rabinovitz, M., and Fisher, J. M. (1988) Mol. Pharmucol. 34, 401-

406 Rognstad, R. (1979) J. Biol. Chem. 254, 1875-1878 Rolleston, F. S. (1972) Curr. Top. Cell. Regul. 6,47-75 Rosenbloom, F. M., Henderson, J. F., Caldwell, I. C., Kelley, W. N.,

and Seegmiller, J. E. (1968) J. Biol. Chem. 243, 1166-1173 Rowe, P. B., McCairns, E., Madsen, G., Sauer, D., and Elliott, H.

(1978) J. Biol. Chem. 253,7711-7721 Sant, M. E., Lyons, S. D., McClure, L. K., Szabados, E., and Chris-

topherson, R. I. (1989a) Cancer Res. 49, 2645-2650 Sant, M. E., Poiner, A., Harsanyi, M. C., Lyons, S. D., and Christo-

pherson, R. I. (1989b) Anal. Biochem. 182, 121-128 Thornwaite, J. T., and Allen, L. M. (1980) Res. Commun. Chem.

Puthol. Phurmucol. 293, 393-396 Tso, J. Y., Bower, S. G., and Zalkin, H. (1980) J. Biol. Chem. 255,

6734-6738 Weber, G., Natsumeda, Y., Lui, M. S., Faderan, M. A., Liepnieks, J.

J., and Elliott, W. L. (1984) Adu. Enzyme Regul. 22, 69-93

by guest, on February 21, 2013

ww

w.jbc.org

Dow

nloaded from

Inhibition of Amidotransferases 11381

by guest, on February 21, 2013

ww

w.jbc.org

Dow

nloaded from