ICI D1694, a Quinazoline Antifolate Thymidylate Synthase Inhibitor That Is a Potent Inhibitor of LI 210 Tumor Cell Growth in Vitro and in Vivo: A New Agent for Clinical Study1

AÃ--(5-|Ar-(3,4-dihydro-2-methyl-4-oxoquinazolin-6-ylmethyl)-/Vmethylamino]-2-thenoyl)-i.-glutamic acid (ICI D1694) is a water-soluble, folate-based thymidylate synthase (TS) inhibitor designed to be a less toxic and more potent analogue of the clinically tested /V'u-propargyl-5,8dideazafolic acid. Inhibition of isolated L1210 TS by ICI D1694 is mixed noncompetitive (although tending toward competitive), with a A', of 62 n\l (/T,,, = 960 n\l). The synthetic -y-polyglutamates are up to 2 orders of magnitude more potent as inhibitors of TS; e.g., the tetraglutamate (gin») has a A, of 1.0 IIM(A,,, = IS IIM).Although inhibitory activity of ICI D1694 toward rat liver dihydrofolatc reducÃ-asewas similar to that of TS (A, = 92 IIM:competitive inhibition) the polyglutamate derivatives did not show enhanced activity. ICI D1694 was also a very potent inhibitor of L1210 cell growth (50% inhibitory activity = 8 n>i). L1210 growth inhibition was not observed in the presence of thymidine, consistent with TS being the locus of action. Folinic acid antagonized LI210 growth inhibition in a competitive fashion such that the highest folinic acid concentration used (25 JIM) increased the 50% inhibitory activity 6000fold. \\ lien given as a 4-h delayed "rescue", folinic acid was much less effective in antagonizing growth inhibition. These observations are con sistent with folinic acid competing with ICI D1694 for uptake into the cell and/or intracellular polyglutamation. The LI 210:1565 cell line, which has greatly impaired reduced-folate/methotrexate transport and thus is resistant to methotrcxate, was significantly cross-resistant to ICI D1694 (121-fold), suggesting that ICI D1694 is dependent on this uptake mechanism for good cytotoxic potency in 1.121(1cells. 1.1210 cells that were incubated for 4 h with 0.1 MM'11-K I D1694 accumulated 1.5 n\} intracellular 'II. and the high performance liquid chromatography analy sis of the cell extracts demonstrated that 96% of the 'II was associated with the ICI D1694 polyglutamate fractions (principally glui). Upon resuspension in drug-free medium for 24 h, ~75% of the cellular 'H was retained, this being the higher polyglutamate pool (gluj 6). In mice, after a single bolus injection of 10 mg/kg of ICI D1694, TS was inhibited >80% for 24 h in ascitic L1210:NCI cells (as measured by the rate of 'II release from |5-3H|deoxyuridine). ICI D1694 cured the L1210:ICR ascitic tumor in mice at 0.4 mg/kg daily for 5 days (maximum tolerated dose, ~50 mg/kg). This antitumor activity was prevented by coadminisiral ion of thymidine, indicating that TS is the sole locus of action in rivo. Coadministration of folinic acid (20 mg/kg) with five daily injections of 10-mg/kg ICI D1694 also resulted in ablation of the antitumor activity. We conclude from the presented evidence that rapid cellular uptake of ICI D1694 and retention as more active polyglutamate metabolites are responsible for its excellent antitumor potency both in vitrn and in vivo. Received 4/25/91: accepted 8/8/91. 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. 1This work was supported by a grant to the Institute of Cancer Research: Royal Cancer Hospital from the Cancer Research Campaign and The Medical Research Council. 1To whom requests for reprints should be addressed, at Institute of Cancer Research. Drug Development Section, Block E. 15 Cotswold Road, Belmont. Sutton, Surrey SM2 5NG, United Kingdom. ' Present address: Cancer Research Unit. Medical School. University of New castle. Framlington Place, Newcastle-upon-Tyne. NE2 4HH. United Kingdom. INTRODUCTION Designing a drug that may be clinically useful based on inhibition of a target enzyme such as TS4 (EC 2.1.1.45) is complex. Potent inhibition of the isolated enzyme does not necessarily translate into potency in whole cell systems or in vivo. CB3717, the prototype folate-based TS inhibitor tested in humans, was a potent inhibitor of isolated TS (K¡ ~3 nM), with moderate potency against tumor cells in vitro or in vivo (1-10). Nevertheless, CB3717 had antitumor activity in humans, thus supporting the concept of TS being a useful target for antifolates (11-17). Unfortunately, CB3717 had a poor therapeutic index because of dose-limiting and life-threatening nephrotoxicity, a result of the poor aqueous solubility of CB3717 at urinary pH (18, 19). The non-dose-related and self-limiting liver toxicity may also have been related to this poor solubility. A number of very water-soluble analogues of CB3717 have been synthesized which are devoid of kidney toxicity in mice (10, 20-26). Gen erally, poor correlation has been found between their TS inhib itory potency and cytotoxic potency. It is clear that factors such as cell membrane penetration and substrate activity for FPGS must have an important role in determining cytotoxicity. For example, the 2-desamino-2-methyl analogue (ICI 198583) was 40-fold more cytotoxic than CB3717, which was related to its uptake into cells via the RFC (unlike CB3717). This, in turn, allowed polyglutamate formation to occur more readily (22). Among the most potent cytotoxic compounds synthesized was a series of 2-desamino-2-methyl-A"°-substituted-5,8-dideazafolate analogues in which the />-aminobenzoate was replaced with a thiophene or thiazole ring (27-29). These compounds also use the RFC in LI210 cells and are significantly better as substrates for isolated FPGS when compared with those pos sessing the benzene ring (30). Of these, the /V10-methyl thio phene compound (ICI D1694) (Fig. 1) has been identified as the most promising compound in vivo. In terms of potency, lack of nephroand hepatotoxicity, therapeutic index, and spectrum of activity against a range of tumor models including human tumor xenografts, ICI D1694 satisfies the in vivo selec tion criteria for a successor to CB3717 (31-33). This paper presents data summarizing the biochemical properties of ICI D1694 with regard to enzyme inhibition, locus of action in vitro and in vivo, the role of transport and polyglutamation in cyto4The abbreviations used are: TS. thymidylate synthase: ICI D1694, N-(S-[N(3,4-dihydro-2-methyl-4-oxoquinazolin-6-ylrnethyl)-A'-methylamino]2-lhenoyi)i.-glutamic acid: CB3717, A"°-propargyl-5,8-dideazafolic acid: gluw. glutamates where n = total number of glutamates; dito hcxaglutamate, ICI D1694 plus 15 extra glutamates: DHFR, dihydrofolate reducÃ-ase;IC50,50% inhibitory activity; MTX, methotrexate; HPLC, high performance liquid chromatography; MTD, maximum tolerated dose; FPGS, folylpolyglutamate synthetase; ICI 198583, 2desamino-2-methyl-,Vlo-propargyl-5.8-dideazafolic acid; RFC. reduced-folate/ methotrexate carrier; A'¡ap|J, apparent inhibition constant; 5,10-CH2-FH4, A"1'0methylenc tetrahydrofolic acid; dThd, thymidine; dUrd, 2'-deoxyuridine; Hx, hypoxanthine; DDATHF. 5.10-dideaza-5.6,7,8-tetrahydrofolic acid; TK, thymi dine kinase: L1210:ICR, LI 210 subline carried at the Institute of Cancer Research (previously called L1210:CBRI).