One of the approaches (1) used for the synthesis of 3-[3-keto-7-(2-piperidyl)propyl]-4-quinazolone involved diketones of type I, but without the methoxyl

Early degradation experiments on the Hydrangea alkaloid (1) indicated that this molecule was a derivative of 4-quinazolone with a side chain on the 3-position containing one basic nitrogen and two oxygens. With this limited information on hand a number of 3-alkyl-4-quinazolones with functional groups in the side chain were synthesized in order to obtain (a) suitable models for degradation experiments (l), (b) compounds to be tested for antimalarial activity,' and (e) information on the synthesis and transformations of these functional groups.2There are a number of methods described in the literature for the synthesis of molecules of this type, three of which were employed. Clark and Wagner (7) ha,ve heated primary aromatic amines with isatoic anhydride (I) and ethyl orthoformate to form 3-aryl-4-quinazolones (11). This procedure has now been 0 0 0

In the previous communication of this series (1) 2-(5-carboxybutyl)thiophane-3,4-¿r

A synthesis of 3-[jS-keto-7-(2-piperidyl)propyl]-4-quinazolone (XI) was de-scribed in an earlier publication (1). This compound was the first of some sixty 3-alkyl-4-quinazolones (2) tested to show antimalarial activity, being about 1% as active as the Hydrangea alkaloid or about equal to quinine.1 Since this compound also duplicated all of the reactions of the natural alkaloid except those due to the hydroxyl function (3), a second type of synthesis was devised which would be applicable to the introduction of an hydroxyl group on certain positions of the piperidine moiety.The key reaction (4) was cyclization of a properly substituted 5-aminoketone to a tetrahydropyridine (X), followed by hydrogenation of the resultant double bond. The necessary diketone, IX, was obtained by an ester-ketone Claisen condensation. Attempted Claisen condensation between 3-acetonyl-4-quinazo-

One of the structures considered highly possible for the Hydrangea alkaloid was 3-[0-keto-7-(4-hydroxy-2-piperidyl) propyl]-4-quinazolone (XXVI) (1). At-tempts to synthesize this compound via the diketone approach were described in the preceding paper (2). The second type of approach used for these compounds involved 2-piperidineacetic acid as the key intermediate (3). This approach has been found feasible through 4-methoxy-2-piperidineacetic acid.Of the methods in the literature for preparation of 4-piperidols, the Dieckmann cyclization approach (4,5) appeared attractive. The requisite amino triester (IV) was obtained by the reductive amination of ethyl acetonedicarboxylate (I) with 0-alanine ester (II) in 40-45% yield or by addition of 0-alanine ester (II) to methyl glutaconate (III). Benzoylation to V followed by Dieckmann cyclization gave only 38-40% yields of the keto esters, VI and VII. Decarbalkoxylation was accompanied by debenzoylation under acid conditions (4), the benzoyl acid, VIII, being isolated in 8-12% yield after rebenzoylation. Hydrogenation of the carbonyl group at 120°in alkaline solution in the presence of a nickel catalyst gave a mixture of oily 4-piperidol isomers, IX. Acetylation and conversion to the acid chloride, X, allowed isolation of one of the isomers in crystalline form.Treatment of the crystalline isomer or the mixed isomers with diazomethane, then with hydrogen bromide gave a negligible yield of bromo ketone, XI. Sufficient material could not be obtained for studying this last reaction due to the consecutive low yields and the fact that larger scale preparation of IV from I gave greatly diminished yields. In some larger runs no triester, IV, could be isolated. The poor conversion of X to XI was subsequently found to be due to the benzoyl blocking group and probably could have been overcome by choice of a different N-blocking group. This point will be discussed later.The piperidine-2-acetic acids (XX) were then synthesized by a modification of the well known -aminoketone cyclization (6-8) followed by hydrogenation. The 5-carbobenzoxyaminovaleric acids (XII) (2) were converted to the ketomalonates, XIV, via the acid chlorides, XIII.The conversion of the ketomalonates (XIV) to the piperidine-2-acetic acids (XX) involved two variables, first, the new modification of the amino ketone ring closure by hydrogenolysis of the carbobenzoxy group, ring closure, and reduction to XVII and, secondly, the influence of the methoxy group (R = CH30) on this method. In order to eliminate the second variable and study only the reductive cyclization reaction, the series R = H was investigated since the intermediate XVIIa should be convertable to 1 -benzoyl-2-piperidineacetic acid (XXIIIa), a compound characterized in an earlier publication (3).