Addition of methylmagnesium iodide/cuprous iodide or of lithium dimethylcuprate to 2-exo-and 2-endomethyl-1 -(phenylsulfonyl)bicyclo[ 1.1.0]butane and determination of the relative geometry of the methyls in the 2,3-dimethylcyclobutanes obtained allowed us to establish that addition of the methyl group occurred from the endo side of the bicyclic molecule. Similarly, lithium aluminum hydride reduction of 2-exo,3-and 2-endo,3dimethylbicyclobutane was also found to occur by endo-addition of the hydride ion at position 3 of the bicyclic system.

A series of 5,8-dideaza analogues of folic acid, isofolic acid, aminopterin, and isoaminopterin were evaluated for inhibition of thymidylate synthase, TS, from mouse L1210 leukemia cells with 10-propargyl-5,8-dideazafolic acid, CB3717, 4a, as the reference inhibitor. These compounds were also tested as inhibitors of human dihydrofolate reductase, DHFR, obtained from WIL2 cells. None of the analogues studied were as potent as 4a toward TS; however, 9-methyl-5,8-dideazaisoaminopterin, 6d, was only 2.5-fold less effective. Compound 4a was prepared by direct alkylation of the di-tert-butyl ester of 5,8-dideazafolic acid followed by hydrolysis of the resulting diethyl ester, which resulted from concomitant transesterification. It was found to be identical with a sample of 4a prepared by earlier methodology by using a variety of spectroscopic techniques. Its isomer, 9-propargyl-5,8-dideazaisofolic acid, 4b, which was synthesized by an analogous approach, was found to be dramatically less inhibitory toward TS than 4a. Each of the 2,4-diamino derivatives, including those possessing an allyl or propargyl group at N9, was an excellent inhibitor of DHFR, having a level of potency similar to that of methotrexate, MTX. However, many of these 5,8-dideazaaminopterin analogues were far more inhibitory toward TS than MTX.

A series of thirty eight 2,4‐diaminoquinazolines having diverse substitution patterns on the aromatic ring was evaluated for inhibitory activity against dihydrofolate reductase (DHFR) obtained from a human lymphoblast cell line. Many of these compounds were also evaluated as inhibitors of rat liver DHFR under the same experimental conditions. In most instances the results obtained with each enzyme were comparable indicating that the rodent enzyme is a suitable model for the human DHFR as far as the determination of I50 values is concerned. The results demonstrate that relatively simple 5‐substituted‐ or 5,6‐disubstituted‐2,4‐diaminoquinazolines can be potent DHFR inhibitors. The presence of a nonpolar substituent at position 7 or 8 was highly detrimental to inhibitory potency.

No abstract

Recently, we reported that appropriately substituted 2‐fluorobenzonitriles undergo cyclization with guanidine carbonate to afford 2,4‐diaminoquinazolines usually in good to excellent yield. This paper describes the preparation of a variety of new 2,4‐diaminoquinazolines substituted at positions five or seven. In addition, the reactions of selected 2‐fluorobenzonitriles with formamidine acetate or acetamidine acetate were examined. The results obtained demonstrate that the analogous 4‐amino‐ and 2‐methyl‐4‐aminoquinazolines can be prepared by this approach but that the yields are considerably lower than when guanidine carbonate is employed as the cyclization reagent.