Synthesis and Reactions of Water-Soluble Congeners of Hexamethylmelamine

Abstract

The pharmacological and chemical properties of hexamethylmelamine are reviewed and, in particular, attention is focussed upon its antitumour properties. These antitumour properties, although far from ideal, were attributed to the N-methyl groups present in the molecule and various workers were motivated to prepare analogues with enhanced water-solubility. The syntheses of some of these analogues are described together with the reduced antitumour properties observed for these compounds. Most of these analogues bore four or less N-methyl groups and it was proposed that six N-methyl groups were necessary for optimal activity. The quaternary salt, heptamethylmelamine chloride, was screened for antitumour activity which was found to be absent and the compound was found to be extremely toxic to mice. This toxicity was believed to be due to the presence of the trimethylammonium moiety which is present also in acetylcholine and may impart cholinergic properties to this quaternary salt of hexamethylmelamine. These observations founded the basis of this thesis and initially, attempts were made to prepare higher homologues of heptamethylmelamine chloride with larger groups on the ammonium moiety in order to suppress the possible cholinergic activity. The synthesis of these homologues proved to be impossible for all but the lowest homologues and this result coupled with the surprisingly higher toxicity of hexamethylethylmelamine chloride led to loss of interest in this series of analogues. The philosophy was extended to other congeners of hexamethylmelamine with a complement of six N-methyl groups but with enhanced water-solubility. The first successfully prepared analogue was 2,4,6-tris(dimethylamino)-1,3,5-oxadiazinium chloride which, although isoelectronic with hexamethylmelamine, was extremely water-soluble. The structure of this analogue was not immediately obvious but it was confirmed as 2,4,6-tris(dimethylamino)-1,3,5-oxadiazinium chloride by X-ray crystallography. Unfortunately, this modification sacrificed the antitumour properties and the compound was found to be extremely toxic. This oxadiazinium salt, however, is a pyrylium-type system and it was decided to compare and contrast its reactions with other pyrylium systems in their reactions with nucleophiles. It rapidly became apparent that oxadiazinium salt underwent the predicted reaction with the nucleophile but the products obtained were not those expected and several novel heterocyclic compounds were isolated. Reaction with ammonia gave a triazin-2-one, with primary amines an oxadiazine or one of two possible triazin-2-ones, with hydrazine to give a 1,2,4-triazole and with hydroxylamine to give a 1,2,4-oxadiazole. Reaction with phenylhydrazine gave a 1,2,4-triazole but it was not the expected phenyl homologue of the product obtained with hydrazine. The crystal structure of this product was determined because the product could not be characterised by any other means. Some of these products had modest water-solubility and were screened for antitumour properties but were devoid of favourable activity. Finally, various novel triazolotriazines with amino or dimethylamino substituents were successfully prepared but their antitumour properties have not as yet been evaluated.