Structural design, biochemical properties, and evidence for improved therapeutic activity of 5-alkyl derivatives of 5-deazaaminopterin and 5-deazamethotrexate compared to methotrexate in murine tumor models.

Studies are described examining a new class of 4-aminofolate analogues modified by an N to C conversion and alkyl substitution at the N-5 position of aminopterin and methotrexate. All of these analogues were equivalent to aminopterin and methotrexate as inhibitors of tumor cell dihydrofolate reductase (Ki = 3.49-5.16 pM). N to C conversion at the N-5 position of aminopterin reduced its influx (inferred from the change in Ki) 3-fold, but the same modification increased influx of methotrexate 2-3-fold in Sarcoma 180 cells. Alkylation (methyl or ethyl) of this position on 5-deazaaminopterin increased influx 3-fold, while a similar alteration of 5-deazamethotrexate increased influx 4-5-fold. Influx of the methotrexate analogues was increased a total of 14-fold as a result of these modifications. Similar differences among these analogues were observed for inhibition of Sarcoma 180 cell growth in culture. Inhibitory potency was in the ascending order methotrexate less than 5-deazamethotrexate less than 5-deazaaminopterin less than aminopterin less than 5-alkyl (methyl or ethyl) analogues of 5-deazaaminopterin and 5-deazamethotrexate (the ethyl analogues were 2-fold more inhibitory than the methyl analogues). All of the analogues examined were equivalent in regard to efflux from Sarcoma 180 cells. Differences in transport alone did not account for all of the increased inhibitory potency (up to 33-fold) of the 5-alkyl-5-deaza analogues compared to the parent compounds. The extent of polyglutamylation of 5-deazaaminopterin and 5-deazamethotrexate and their 5-alkyl derivatives in Sarcoma 180 cells was substantially less compared to aminopterin and equivalent to methotrexate. Transport inward of 5-deazaaminopterin in isolated crypt cell epithelium from mouse small intestine was 2-fold lower than aminopterin (influx Km = 14.2 +/- 2 microM), while influx of 5-deazamethotrexate was 2-fold greater than methotrexate (influx Km = 98.6 +/- 23). However, transport inward of all of the 5-alkyl derivatives of these 5-deaza analogues was intermediate [influx Km = 44.4 +/- 11 (SEM) to 49.8 +/- 12 microM] between values for aminopterin and methotrexate. These differences accounted, to some extent, for the reduced toxicity of the 5-alkyl-5-deazaaminopterin analogues compared to aminopterin and the increased toxicity of 5-methyl-5-deazamethotrexate compared to methotrexate. All of the 5-alkyl derivatives of aminopterin and methotrexate were more active in vivo than methotrexate against four murine tumor models.(ABSTRACT TRUNCATED AT 400 WORDS)

Hydrolytic cleavage of methotrexate gamma-polyglutamates by folylpolyglutamyl hydrolase derived from various tumors and normal tissues of the mouse.

Folylpolyglutamate hydrolase (folyl hydrolase) activity derived from murine tumors and various normal tissues was measured by means of high performance liquid chromatography using methotrexate polyglutamates as substrate. Enzyme-mediated hydrolysis was considerably greater (10-20-fold) on a specific activity basis in extracts from all normal mouse tissues (kidney greater than bone marrow greater than small intestine approximately equal to liver) than from tumor cells (Sarcoma 180 greater than Ehrlich approximately equal to L1210 cells). Enzyme preparations from purified absorptive and crypt cell epithelium from mouse small intestine exhibited comparable levels of specific activity and were greater than that derived from the total organ. Folyl hydrolase from mouse kidney showed mixed endo- and exopeptidase activity while that derived from all other normal tissues and tumor cells was consistent with endopeptidase activity. Levels of cell-free folyl hydrolase activity derived from tumor cells varied substantially with the phase of growth in vivo. Also, levels were appreciably lower from the same cells grown in vitro. Hydrolysis by crude or partially purified enzyme preparations from mouse small intestine or tumor cells conformed to Michaelis-Menten kinetics (single saturable component). Rates of hydrolysis and Km values were proportional to gamma-glutamyl chain length in the case of L1210 cell-derived enzyme but not for enzyme derived from small intestine. Km values derived for 4-amino-10-methylpteroyldiglutamate were the same [Km = 80.4 +/- 9 (SE) microM] for small intestine and L1210 cells. However, with 4-amino-10-methylpteroyltetraglutamate Km values were 3-fold lower for tumor cell preparations and 8-fold lower for preparations derived from small intestine. Fourfold lower Km values for 4-amino-10-methylpteroyldiglutamate were obtained with enzyme derived from Sarcoma 180 cells as compared to the enzyme from L1210 or intestinal cells. Varying levels of folyl hydrolase activity for methotrexate polyglutamates in cell-free preparations from different tumor cells appeared to reflect differences in in situ hydrolytic activity shown for the same substrate when internalized. The relevance of these results to antifolate pharmacology and, specifically, to a role for polyglutamates of 4-aminofolate compounds in determining cytotoxicity and selective antitumor activity of these agents is discussed.

Inhibition of glycinamide ribonucleotide formyltransferase and other folate enzymes by homofolate polyglutamates in human lymphoma and murine leukemia cell extracts.

In order to determine the biochemical basis for the cytotoxicity of homofolates, poly-gamma-glutamyl derivatives of homofolate (HPteGlu) and tetrahydrohomofolate (H4HPteGlu) were synthesized and tested as inhibitors of glycinamide ribonucleotide formyltransferase (GARFT), aminoimidazolecarboxamide ribonucleotide formyltransferase (AICARFT), thymidylate synthase, and serine hydroxymethyltransferase (SHMT) in extracts of Manca human lymphoma and L1210 murine leukemia cells. The most striking inhibitions are that of GARFT by (6R,S)-H4HPteGlu4-6 with IC50 values from 1.3 to 0.3 microM. Both diastereomers, (6R)-H4HPteGlu6 and (6S)-H4HPteGlu6, inhibit GARFT activity. In Manca cell extracts, the (6S) form is more potent than the (6R) form whereas in the murine system the reverse is true. The (6R,S)-H4HPteGlu polyglutamates are weak inhibitors of human AICARFT (IC50, 6-10 microM). Polyglutamates of HPteGlu, however, are more inhibitory to AICARFT, with HPteGlu4-6 having IC50 values close to 2 microM. Polyglutamates of HPteGlu and of H4HPteGlu are weaker inhibitors of thymidylate synthase (IC50, 8 microM for HPteGlu5-6 and greater than 20 microM for H4HPteGlu1-5). Polyglutamates of HPteGlu and of H4HPteGlu are poor inhibitors of SHMT (IC50, greater than 20 microM). Manca cell growth is inhibited 50% by HPteGlu and (6R,S)-5-methyl-H4HPteGlu at 6 and 8 microM, respectively. Both of these effects are reversed by 0.1 mM inosine. Trimetrexate at a subinhibitory concentration, 10 nM, antagonizes growth inhibition by HPteGlu, raising the IC50 from 6 to 64 microM, but enhances inhibition by (6R,S)-5-methyl-H4HPteGlu, lowering the IC50 from 8 to 5 microM. Our results support the view that homofolates become toxic after conversion to H4HPteGlu polyglutamates which block GARFT, a step in purine biosynthesis.

Similar specificity of membrane transport for folate analogues and their metabolites by murine and human tumor cells: a clinically directed laboratory study.

Mediated transport of folate compounds exhibited similar kinetic characteristics and structural specificity in a series of cultured murine and human tumor cells examined in a parallel fashion. In each case, influx was characterized by a single saturable component with an approach to steady-state conforming to a single exponential, while efflux was first order (poorly saturable). Both mediated fluxes exhibited high temperature dependence (Q10 27-37 degrees = 6 to 8). During competition studies with various analogues, it was found that positions 4, 5, 7, and 10 and the gamma-carboxyl position of the folate molecule were specified for influx in tumor cells from each species. Also, short-chain alkyl substitution at position 10 was specified in the case of N10, but not in the case of C10. None of the modifications at position 10 affected mediated efflux in either cell type. The linkage of additional glutamyl residues at the gamma-carboxyl-position resulted in reduced saturability (increased value for Ki) of influx in both murine and human tumor cells in a manner proportional to the number of glutamyl residues. Mediated influx in human ovarian carcinoma cells obtained from malignant effusions in several patients and in an established cell line derived from one of these patients showed similar kinetics for folate analogue transport and specificity for modification at position 10 of the 4-amino-folate molecule. Mediated entry of 10-deazaaminopterin and its 10-ethyl derivative compared to entry of methotrexate was 4- to 11-fold greater in murine tumor cells and 4- to 9-fold greater in human tumor cells in culture or when clinically derived. Mediated efflux was not specified for position 10 on the 4-amino folate structure in any tumor cell type. These findings appear to provide some basis for concluding that the results of studies of this type in model murine systems or with tumor cell lines established in culture have relevance to clinical cancer.

Liposome-mediated delivery of pteridine antifolates to cells in vitro: potency of methotrexate, and its alpha and gamma substituents.

We have examined the growth-inhibitory potency of several pteridines encapsulated in negatively charged liposomes, including methotrexate, methotrexate-gamma-methylamide, methotrexate-gamma-dimethylamide, methotrexate-alpha-aspartate, and a lipophilic methotrexate-phosphatidylethanolamine conjugate. The potency of encapsulated methotrexate is greater than the potency of the free drug for CV1-P cells, but not for other cell lines. The potency of methotrexate-gamma-methylamide and methotrexate-gamma-dimethylamide is only minimally improved by encapsulation. The potency of methotrexate-alpha-aspartate is increased by encapsulation. In addition, the lipophilic methotrexate derivative has demonstrable potency when incorporated in liposomes. We have also examined the potency of several pteridines under conditions where the cells are exposed to the drug for periods shorter than the entire growth assay. Reduction of the exposure time decreases the potency of both encapsulated and free drugs. However, the difference in potency between the encapsulated and free drug is increased, because the potency of the encapsulated drug is affected less. Consequently, encapsulated methotrexate-gamma-aspartate is 300-fold more potent than free drug, if CV1-P cells are exposed to drug for 4 h. Moreover, encapsulated methotrexate is more potent than free methotrexate for growth inhibition of L929 fibroblasts, if the term of exposure is less than 8 h. Potency is least affected by reduction of exposure length for the lipophilic methotrexate derivative.

Cross-resistance studies of folylpolyglutamate synthetase-deficient, methotrexate-resistant CCRF-CEM human leukemia sublines.

CCRF-CEM human leukemia sublines resistant to short-term methotrexate (MTX) exposure as a result of decreased folylpolyglutamate synthetase (FPGS) activity were examined for their response to other cytotoxic agents. The R3/7 and R30dm sublines display 25 and 1%, respectively, of the FPGS activity of CCRF-CEM cells as measured with MTX in vitro. Response to agents in outgrowth experiments was examined under both continuous exposure (120 h, where MTX resistance is not observed) and short-term (6-14.5 h) exposure. During continuous exposure to various classes of agents, cross-resistance of R3/7 and R30dm that correlated with FPGS level was not observed, although some minor (< or = 3-fold) stochastic variations in sensitivity were noted. These agents included actinomycin D, Adriamycin, etoposide, vincristine, cisplatin, cytosine arabinoside, 5-fluorouracil, and some other antifolates. Cross-resistance during continuous exposure that did correlate with FPGS level was noted, however, to glutamate-containing thymidylate synthase inhibitors (including ICI D1694) and, to a minor extent, to 6-mercaptopurine and 5-fluorodeoxyuridine. Slight collateral sensitivity during continuous exposure that apparently correlated with FPGS level was noted to the lipid-soluble antifolate trimetrexate and to 5,8-dideazapteroyl-L-ornithine, an FPGS-specific inhibitor. In short-term exposures (where MTX resistance of the sublines is observed), the resistant sublines displayed sensitivity or cross-resistance to each agent that was qualitatively similar to that observed for the same agent in continuous exposure. Because of the requirement for reduced folates in the anti-DNA mechanism of action of fluoropyrimidines and the current clinical use of leucovorin (LV) to enhance their effects, the interaction of LV and fluoropyrimidines was examined. The results suggest that even highly FPGS-deficient cells are as sensitive to the effects of LV modulation as are wild-type cells even at fluoropyrimidine exposure times as short as 4 h.

Convenient synthesis of 10-deazaaminopterin via a pteridine ylide.

10-Deazaaminopterin, a potential antitumor agent now undergoing clinical trials, has been synthesized by a new approach involving the Wittig reaction. The ylide obtained by reaction of 6-(bromomethyl)-2,4-pteridinediamine with triphenylphosphine in Me2NAc, followed by treatment with NaOMe, underwent smooth reaction with diethyl N-(4-formylbenzoyl)-L-glutamate to give the vinyl precursor of the subject compound. Catalytic hydrogenation (Pt in glacial AcOH) of this product at ambient conditions led to uptake of 3 molar equiv of H2. Exposure to air during saponification of the ester groupings apparently gave the 7,8-dihydro compound according to UV spectral data, and further oxidation with H2O2 led to 10-deazaaminopterin.

Synthesis of potential inhibitors of hypoxanthine-guanine phosphoribosyltransferase for testing as antiprotozoal agents. 1. 7-Substituted 6-oxopurines.

Biological evidence indicates that the enzyme hypoxanthine-guanine phosphoribosyltransferase (EC 2.4.2.8) is vital for cell proliferation in malarial parasites but nonessential for mammalian cells. 7-Substituted guanines and hypoxanthines in which the 7 substituent bears functional or hydrophobic groups were prepared with the aim of finding a suitably constituted compound whose resemblance to the normal substrate allows it to compete for the reversible purine binding site of HGRPTase while allowing a substituent group of the inhibitor molecule to form a covalent bond or strong hydrophobic bond with appropriate sites on the enzyme. Multistep syntheses that began with hydroxyalkylations and alkylations of guanosine led to four key guanines substituted at the 7 position by the following chains: 2-aminoethyl, 3-amino-2-hydroxypropyl, 3-aminobenzyl, and 4-aminobenzyl. Similarly, 7-(4-aminobenzyl)hypoxanthine was prepared. Reactions at the side-chain amino groups with bromoacetic anhydride (or, alternatively, 4-nitrophenyl bromoacetate) and 3- and 4-(fluorosulfonyl)benzoyl chlorides afforded derivatives bearing functional groups capable of forming covalent bonds with enzymes through displacement reactions. 4-Chlorobenzyl derivatives were similarly prepared as potential inhibitors that might act through hydrophobic bonding. Three 7-substituted guanines whose side chains bear other functions (7-guanine-3-propranesulfonic acid, guanine-7-acetaldehyde, and the ethyl ester of 7-guanine-4-crotonic acid) were prepared as potential inhibitors and for possible use as intermediates. None of these compounds extended the life span of P. berghei infected mice or showed significant in vitro inhibition of HGPRTase from H.Ep.-2 cells.

A synthetic approach to poly(gamma-glutamyl) congugates of methotrexate.

Methotrexate poly(gamma-L-glutamate)s bearing two and three glutamate units above that present in methotrexate have been synthesized by extension of a previously described route used to synthesize the lower conjugate bearing one added glutamate unit. Key steps in the sequence are the peptide coupling of N-[4-[[(benzyloxy)carbonyl]-methylamino]benzoyl]-L-glutamic acid alpha-benzyl ester (5) with oligo(gamma-L-glutamate) benzyl esters, removal of blocking groups by catalytic hydrogenolysis, and introduction of the (2,4-diamino-6-pteridinyl)methyl grouping by alkylation with 6-(bromomethyl)-2,4-pteridinediamine hydrobromide. Elaboration of the required oligo(gamma-L-glutamate) chain was achieved one unit at a time, beginning with the coupling of L-glutamic acid dibenzyl ester with [(tert-butyloxy)carbonyl]-L-glutamic acid alpha-benzyl ester (7), followed by selective removal of the tert-butyloxycarbonyl grouping and another coupling step with 5 or 7 as required. Diphenylphosphoryl azide was used as the coupling reagent in each conversion producing a peptide linkage.

S-2,omega-Diaminoalkyl dihydrogen phosphorothioates as antiradiation agents.

To enable further structure-activity comparisons among radioprotective phosphorothioates, S-2,omega-diaminoalkyl dihydrogen phosphorothioates were synthesized from L-2,4-diaminobutyric acid, L-ornithine, L-lysine, and DL-2,7-diaminoheptanoic acid as homologues of S-2,3-diaminopropyl dihydrogen phosphorothioate (4) and as isomeric analogues of S-2-[(omega-aminoalkyl)amino]ethyl dihydrogen phosphorothioates (e.g., 1). The preferred route that evolved from exploratory trials retained optical activity and involved the reduction of methyl 2,omega-bis(benzoylamino)alkanoates with lithium borohydride, debenzoylation-bromodehydroxylation, and reaction of the resulting 1-(bromomethyl)-1,omega-alkanediamine dihydrobromides with trisodium phosphorothioate. The products of an alternative route that involved the reduction of phthaloylated intermediates with sodium borohydride were racemic. Exploratory conversions of N-(omega-alkenyl)phthalimides failed to provide suitable precursors of the target compounds. In terms of a protective index, these homologues were significantly more radioprotective than the parent phosphorothioate 4 when administered intraperitoneally to mice prior to whole-body gamma irradiation. The homologues derived from L-lysine also showed good peroral activity. No apparent difference was observed in the protection afforded by optically active homologues and the corresponding racemates.

Syntheses and evaluation as antifolates of MTX analogues derived from 2, omega-diaminoalkanoic acids.

Methotrexate (MTX) analogues 27a-c bearing 2, omega-diaminoalkanoic acids (ornithine and its two lower homologues) in place of glutamic acid were synthesized by routes proceeding through N2-[4-(methylamino)benzoyl]-N omega-[(1,1-dimethylethoxy)carbonyl]-2, omega-diaminoalkanoic acid ethyl esters and N2-[4-(methylamino)benzoyl]-N5-[(1,1-dimethylethoxy)carbonyl]-2, 5-diaminopentanoic acid followed by alkylation with 6-(bromomethyl)-2, 4-pteridinediamine hydrobromide. Reactions at the terminal amino group of 27-type analogues or of appropriate precursors led to other MTX derivatives whose side chains terminate in ureido, methylureido, N-methyl-N-nitrosoureido, N-(2-chloroethyl)-N-nitrosoureido, and 4-chlorobenzamido groups. Also prepared were unsymmetrically disubstituted ureido types resulting from addition of ethyl isocyanatoacetate and diethyl 2-isocyanatoglutarate to the ethyl esters of 27a,b. Of these ureido adducts (32a,b and 33a,b, respectively), only 33a was successfully hydrolyzed to the corresponding pure acid, in this instance the tricarboxylic acid 34, a pseudo-peptide analogue of the MTX metabolite MTX-gamma-Glu. Biological evaluations of the prepared compounds affirmed previous findings that the gamma-carboxyl is not required for tight binding to dihydrofolate reductase (DHFR) but is operative in the carrier-mediated transport of classical antifolates through cell membranes. High tolerance levels observed in studies against L1210 leukemia in mice suggest the reduced potency may be due not only to lower transport efficacy but also to loss of the function of intracellular gamma-polyglutamylation. The N-nitrosoureas 30 and 31 showed appreciable activity in vivo vs. L1210, but the activity did not appear to be due to antifolate action as evidenced by their poor inhibition of both L1210 DHFR and cell growth in vitro.

Synthesis and antitumor activity of 10-propargyl-10-deazaaminopterin.

Successive alkylation of dimethyl homoterephthalate with propargyl bromide and 2,4-diamino-6-(bromomethyl)pteridine followed by ester saponification at room temperature afforded 2,4-diamino-4-deoxy-10-carboxy-10-propargyl-10-deazapteroic acid. The 10-COOH was readily decarboxylated by heating in DMSO at a temperature of only 120 degrees C to yield the diamino-10-propargyl-10-deazapteroic acid intermediate. Coupling with diethyl L-glutamate and ester hydrolysis gave the title compound. The 10-propargyl analogue was about 5 times more potent than MTX as an inhibitor of growth in L1210 cells, but was only one-third as potent as an inhibitor of DHFR from L1210. The analogue was transported inward very effectively in L1210 cells showing a 10-fold advantage over MTX. At a dose of 36 mg/kg the 10-propargyl compound caused shrinkage of the E0771 solid murine mammary tumor to only 1% of untreated controls.

Synthesis and antifolate evaluation of 10-ethyl-5-methyl-5,10- dideazaaminopterin and an alternative synthesis of 10-ethyl-10- deazaaminopterin (edatrexate).

Previous findings suggesting that 5,10-dialkyl-substituted derivatives of 5,10-dideazaaminopterin warranted study as potential antifolates prompted synthesis of 10-ethyl-5-methyl-5,10- dideazaaminopterin (12a). The key step in the synthetic route to 12a was Wittig condensation of the tributylphosphorane derived from 6-(bromomethyl)-2,4-diamino-5-methylpyrido[2,3-d]pyrimidine (7a) with methyl 4-propionylbenzoate. Reaction conditions for the Wittig condensation were developed using the tributylphosphorane prepared from 6-(bromomethyl)-2,4-pteridinediamine (7b) as a model. Each of the respective Wittig products 8a and 8b was obtained in 75-80% yield. Hydrogenation of 8a and 8b at their 9,10-double bond afforded 4-amino-4-deoxy-10-ethyl-5-methyl-5,10-dideazapteroic acid methyl ester (9a) and 4-amino-4-deoxy-10-ethyl-10-deazapteroic acid methyl ester (9b). This route to 9b intersects reported synthetic approaches leading to 10-ethyl-10-deazaaminopterin (10-EDAM, edatrexate), an agent now in advanced clinical trials. Thus the Wittig approach affords an alternative synthetic route to 10-EDAM. Remaining steps were ester hydrolysis of 9a,b to give carboxylic acids 10a,b followed by standard peptide coupling with diethyl L-glutamate to produce diethyl esters 11a,b, which on hydrolysis gave 12a and 10-EDAM (12b), respectively. The relative influx of 12a was enhanced about 3.2-fold over MTX, but as an inhibitor of dihydrofolate reductase (DHFR) from L1210 cells and in the inhibition of L1210 cell growth in vitro, this compound was approximately 20-fold less effective than MTX (DHFR inhibition, Ki = 4.82 +/- 0.60 pM for MTX, 100 pM for 12a; cell growth, IC50 = 3.4 +/- 1.0 nM for MTX, 65 +/- 18 nM for 12a).

Synthesis and antifolate evaluation of the 10-propargyl derivatives of 5-deazafolic acid, 5-deazaaminopterin, and 5-methyl-5-deazaaminopterin.

5-Deaza-10-propargylfolic acid (4), an analogue of the thymidylate synthase (TS) inhibitor 10-propargyl-5,8-dideazafolic acid (PDDF, 1), was prepared via alkylation of diethyl N-[4-(propargylamino)benzoyl]-L-glutamate (7) by 2-amino-6-(bromomethyl)-4(3H)-pyrido[2,3-d]pyrimidinone (15). Bromomethyl intermediate 15 was prepared from the corresponding hydroxymethyl precursor 14 by treatment with 48% HBr. Hydroxymethyl compound 14 was obtained by deamination of reported 2,4-diaminopyrido[2,3-d]pyrimidine-6-methanol (12a) in refluxing 1 N NaOH. Both 12a and its 5-methyl-substituted analogue 12b were converted to versatile 6-bromomethyl intermediates 13a and 13b from which important antifolates may be readily derived. Alkylation of 7 by 13a,b led to 10-propargyl-5-deazaaminopterin (5) and 5-methyl-10-propargyl-5-deazaaminopterin (6). As an inhibitor of TS from H35F/F cells, 4 gave an IC50 value showing it to be approximately 6-fold less inhibitory than PDDF (90 nM for 4 vs 14 nM for PDDF). In in vitro studies, IC50 (microM) values obtained for 4 vs L1210 and S180 of 1.50 and 2.35, respectively, were similar to those obtained for PDDF (2.61 and 1.97). Against HL60 cells, 4 was about 7-fold more cytotoxic than PDDF (IC50 values 0.72 and 5.29 microM). Inclusion of thymidine did not establish TS as the site of cytotoxic action for either 4 or PDDF in the cell lines used. In in vivo tests against L1210 in mice, 4 failed to show therapeutic effect. The 2,4-diamino compounds 5 and 6 were as potent inhibitors of DHFR from L1210 cells as MTX and 7- and 35-fold, respectively, more inhibitory than MTX toward L1210 cell growth. In mediated influx into L1210 cells, 5 and 6 were transported 2.7- and 8.5-fold, respectively, more readily than MTX. Against the EO771 mammary adenocarcinoma in mice, 6 produced greater antitumor effect than MTX. A dose of 36 mg/kg per day for 5 days caused no toxic deaths while the average tumor volume among 10 mice was reduced to 8-9% of that of the control, and 20% of the test animals were rendered tumor free.

Analogues of 10-deazaaminopterin and 5-alkyl-5,10-dideazaaminopterin with the 4-substituted 1-naphthoyl group in the place of 4-substituted benzoyl.

10-Deaza modifications of classical antifolate analogues bearing the 1,4-disubstituted naphthalene ring in place of the 1,4-disubstituted benzene ring were prepared and tested for antitumor activity. Naphthalene analogues (9a-c, respectively) of 10-deazaaminopterin, 5-methyl-5, 10-dideazaaminopterin, and 5-ethyl-5,10-dideazaaminopterin were prepared by a route consisting of C-alkylations of the anion derived from 4-carboxyl-1-naphthaleneacetic acid dimethyl ester (2) by 6-(bromomethyl)-2,4-diaminopteridine (1a) and 6-(bromomethyl)-2,4-diamino-5-methyl- and -5-deazapteridines (1b and 1c, respectively) followed by ester hydrolysis and subsequent decarboxylation to give naphthalene analogues (7a-c, respectively) of 4-amino-4-deoxy-10-deazapteroic acid and 4-amino-4-deoxy-5- methyl- and -5-ethyl-5,10-dideazapteroic acids. Peptide coupling of 7a-c with L-glutamic acid dialkyl ester followed by mild ester hydrolysis gave target compounds 9a-c. The key advantage of this route is circumvention of a hydrogenation step requiring selectivity as in earlier approaches involving 9,10-olefinic precursors. Steric limitations thwarted plans to prepare the naphthalene analogue of 10-ethyl-10-deazaaminopterin; attempted alkylations of 2-(4-carboxy-1-naphthyl)butyric acid dimethyl ester with 1a failed as did attempted further alkylation (by EtBr) of the product derived from 1a and 2. Growth inhibition tests against three tumor cell lines (L1210, S180, and HL60) showed 9a to be 4-6-fold more inhibitory than methotrexate but not as inhibitory as 10-ethyl-10-deazaaminopterin; 9b and 9c were no more inhibitory than MTX. In tests against the EO771 mammary adenocarcinoma in mice, 9a was less active than MTX.

Hydroxy derivates of S-2-(3-aminopropylamino)ethyl dihydrogen phosphorothioate and related compounds as antiradiation agents.

The high antiradiation activity and low toxicity of sodium 3-amino-2-hydroxypropyl hydrogen phosphorothioate (1) suggested the introduction of hydroxyl groups into other types of radioprotective phosphorothioates. A number of such compounds were synthesized, including S-3-(3-aminopropylamino)-2-hydroxypropyl dihydrogen phosphorothioate (11, n equals 3), S-2-(3-amino-2-hydroxypropylamino)ethyl dihydrogen phosphorothioate (20) and its propyl homolog 26, N,N'-(2-hydroxytrimethylene)bis(S-2-aminoethyl dihydrogen phosphorothioate) (40), S-2-[3-(2-hydroxyethylamino)propylaminoi1ethyl dihydrogen phosphorothioate (44), and sodium S-2-amino-2-(hydroxymethyl)-3-hydroxypropyl hydrogen phosphorothioate (49). Compounds 11 (n equals 3), 20, 26, and 49 were highly protective when administered intraperitoneally but were generally ineffective when given perorally, as were the other hydroxylated phosphorothioates prepared. The introduction of hydroxyl groups significantly enhanced the radioprotective properties of nonhydroxylated parent compounds, however, only in the case of intraperitoneally administered.

Syntheses of alpha- and gamma-substituted amides, peptides, and esters of methotrexate and their evaluation as inhibitors of folate metabolism.

N-[4-[[(Benzyloxy)carbonyl]methylamino]benzoyl]-L-glutamic acid alpha-benzyl ester (2) and gamma-benzyl ester (6) served as key intermediates in syntheses of precursors to amides and peptides of methotrexate (MTX) involving both the alpha- and gamma-carboxyl groupings of the glutamate moiety. Coupling of 2 and 6 at the open carboxyl grouping with amino compounds was affected by the mixed anhydride method (using isobutyl chloroformate); carboxyl groupings of amino acids coupled with 2 and 6 were protected as benzyl esters. N-[4-[[(Benzyloxy)carbonyl]methylamino]benzoyl]-L-glutamic acid gamma-methyl ester (5), a precursor to MTX gamma-methyl ester, was prepared from L-glutamic acid gamma-methyl ester and 4-[[(benzyloxy)carbonyl]methylamino]benzoyl chloride (1) in a manner similar to that used to prepare 2 and 6. The precursor to MTX alpha-methyl ester was prepared from gamma-benzyl ester 6 by treatment with MeI in DMF containing (i-Pr)2NEt. Benzyl and (benzyloxy)carbonyl protective groupings were removed by hydrogenolysis, and the deprotected side-chain precursors were converted to alpha- and gamma-substituted amides, peptides, and esters of MTX by alkylation with 6-(bromomethyl)-2,4-pteridinediamine hydrobromide (12). Biochemical-pharmacological studies on the prepared compounds aided in establishing that the alpha-carboxyl grouping of the glutamate moiety contributes to the binding of MTX to dihydrofolate reductase while the gamma-carboxyl does not. Other studies on the peptide MTX-gamma-Glu (13h) are concerned with the contribution toward antifolate activity of this metabolite of MTX. The compounds prepared were also evaluated and compared with MTX with respect to cytotoxicity toward H.Ep.-2 cells and effect on L1210 murine leukemia.

Synthesis of potential inhibitors of hypoxanthine-guaine phosphoribosyltransferase for testing as antiprotozoal agents. 2. 1-Substituted hypoxanthines.

Evidence incicating that effective in vivo inhibition of hypoxanthine-guanine phosphoribosyltransferase (HGPRT, EC 2.4.2.8) should produce antiprotozoal activity without significant toxic effects on mammalian hosts prompted syntheses of 1-substituted hypoxanthines bearing functionalized side chains whose groupings might interact with appropriate groupings of HGPRT to form covalent bonds or strong hydrophobic bonds. 3-(Fluorosulfonyl)benzoyl, 4-(fluorosulfonyl)benzoyl, 4-chlorobenzoyl, and bromacetyl derivatives of two parent amines, 1-(2-aminoethyl)-hypoxanthine and 1-(4-aminobenzyl)hypoxanthine, were synthesized for evaluation in this connection. None of these compounds extended the life span of Plasmodium berghei infected mice or showed significant in vitro inhibition of HGPRT from H.Ep.-2 cells, but 1-[2-(bromoacetamido)ethyl]hypoxanthine displayed in vivo activity against Trypanosoma rhodesiense.

Analogues of methotrexate.

Analogues of methotrexate (MTX) were prepared by alkylation of side-chain precursors with 6-(bromomethyl)-2,4-pteridinediamine followed, where necessary, by saponification of the intermediate esters and, in two cases, by electrophilic substitution reactions in the pyridine ring portion of 3-deazamethotrexate. Effects of the various modifications on their ability to inhibit dihydrofolate reductase, cytotoxicity, and activity against L1210 leukemia in mice were examined in light of recent findings concerning active transport of MTX and related compounds and the binding features of the MTX-dihydrofolate reductase complex.

Synthesis and antifolate activity of 5-methyl-5,10-dideaza analogues of aminopterin and folic acid and an alternative synthesis of 5,10-dideazatetrahydrofolic acid, a potent inhibitor of glycinamide ribonucleotide formyltransferase.

The title compounds were prepared in extensions of a general synthetic approach used earlier to prepare 5-alkyl-5-deaza analogues of classical antifolates. Wittig condensation of 2,4-diaminopyrido[2,3-d]pyrimidine-6-carboxaldehyde (2a) and its 5-methyl analogue 2b with [4-(methoxycarbonyl)benzylidene] triphenylphosphorane gave 9,10-ethenyl precursors 3a and 3b. Hydrogenation (DMF, ambient, 5% Pd/C) of the 9,10-ethenyl group of 3b followed by ester hydrolysis led to 4-[2-(2,4-diamino-5-methylpyrido[2,3-d]pyrimidin-6-yl)ethyl]ben zoi c acid (5), which was converted to 5-methyl-5,10-dideazaaminopterin (6) via coupling with dimethyl L-glutamate (mixed-anhydride method using i-BuOCOCl) followed by ester hydrolysis. Standard hydrolytic deamination of 6 gave 5-methyl-5,10-dideazafolic acid (7). Intermediates 3a and 3b were converted through concomitant deamination and ester hydrolysis to 8a and 8b. Peptide coupling of 8a,b (using (EtO)2POCN) with diesters of L-glutamic acid gave intermediate esters 9a and 9b. Hydrogenation of both the 9,10 double bond and the pyrido ring of 9a and 9b (MeOH-0.1 N HCl, 3.5 atm, Pt) was followed by ester hydrolysis to give 5,10-dideaza-5,6,7,8-tetrahydrofolic acid (11a) and the 5-methyl analogue 11b. Biological evaluation of 6, 7, 11a, and 11b for inhibition of dihydrofolate reductase (DHFR) isolated from L1210 cells and for growth inhibition and transport characteristics toward L1210 cells revealed 6 to be less potent than methotrexate in the inhibition of DHFR and cell growth. Compounds 6, 11a, and 11b were transported into cells more efficiently than methotrexate. Growth inhibition IC50 values for 11a and 11b were 57 and 490 nM, respectively; the value for 11a is in good agreement with that previously reported (20-50 nM). In tests against other folate-utilizing enzymes, 11a and 11b were found to be inhibitors of glycinamide ribonucleotide formyltransferase (GAR formyltransferase) from one bacterial (Lactobacillus casei) and two mammalian (Manca and L1210) sources with 11a being decidedly more inhibitory than 11b. Neither 11a nor 11b inhibited aminoimidazolecarboxamide ribonucleotide formyltransferase. These results support reported evidence that 11a owes its observed antitumor activity to interference with the purine de novo pathway with the site of action being GAR formyltransferase.