| Summary: | <p>The introduction consists of a brief review of the subject with special reference to antimalarial agents and haematopoietic factors.</p> <p><strong>Part I. Some synthetic Biguanide derivatives</strong>.</p> <p>It has been suggested (Curd and Rose, <em>Nature</em>, 1946, <strong>158</strong>, 707) that the activity of paludrine as an antimalarial is due to the interference of a dimeric form with a porphyrin enzyme system essential for the growth of the plasmodia.</p> <p>Hawking (<em>Nature</em>, 1947, <strong>159</strong>, 409) also showed that paludrine was inactive as an antimalarial <em>in vitro</em>, but was converted into an unidentified biologically active material <em>in vivo</em>. Oxidation to a benziminazole is one possibility, and a series of 2-guanidinobenziminazoles, which are structurally very similar to paludrine, was synthesised for biological examination.</p> <p>The benziminazoles (1 to 15) were all made from the appropriate <em>o</em>-phenylenediamine dihydrochlorides and dicyandiamide, or <em>isopropyl</em>- or <em>n</em>-butyldicyandiamides in aqueous solution, and were usually purified through the <em>picrates</em> or copper derivatives. The hitherto undescribed alkylated dicyandiamides were prepared from sodium dicyanimide and the appropriate amine hydrochlorides in boiling <em>n</em>-butanol. <p>These 2-guanidinobenziminazoles, as their <em>hydrochlorides</em>, were tested against <em>F. gallinaceum</em> in chicks, or <em>P. relictum</em> in canaries, but only two, 7 and 14, had even slight activity. These results showed that 11 could not be Hawking's biologically active metabolite and do not support Curd and Rose's theory. The bimolecular form of 11 is no less like the porphyrin ring system than the postulated paludrine dimer.</p> <p>None of the 2-guanidinobenziminazoles showed large bacteriostatic action against <em>B. coli</em>, or <em>S. aureus</em> broth cultures.</p> <p>Attempts to synthesise the 2-guanidinobenziminazoles in the reverse direction failed. 2-Cyanamino-benziminazole was not formed from <em>o</em>-phenylenediamine and dicyanimide, and when prepared according to Pellizzari (<em>Gazzetta</em>, 1921, <strong>51</strong>, I, 140) could not be induced to combine with <em>iso</em>propylamine (experiments by P.C. Spensley).</p> <p>The product of the reaction of <em>o</em>-phenylenediamine with dicyanimide, 2;4-<em>diamino</em>-1:3:5-<em>triazabenzepine</em> (16), is a derivative of a new heterocyclic system. It was stable to dilute acid and alkali, and like its 1-<em>methyl derivative</em> (18) did not react with nitrous acid.</p> <p>A series of these benzepines was prepared from the appropriate diamines and dicyanimide, and one (20) is a possible paludrine metabolite.</p> <p>The compounds had no effect on the growth of <em>B. coli</em> or <em>S. aureus</em> broth cultures.</p> <p><strong>Part II. Alloxazines and <em>iso</em>alloxazines</strong>.</p> <p>This was a continuation of earlier work by the author and Prof. F.E. King (<em>J</em>., 1946, 681), on the preparation of alloxazine derivatives for biological testing.</p> <p>Piloty's synthesis of alloxazines (<em>Annalen</em>, 1904, <strong>333</strong>, 44) was extended to the <em>iso</em>alloxazine series, and violuric acid was found to react with 3-<em>dimethylamino</em>-β-<em>diethylaminoethylaniline</em> to give the corresponding <em>iso</em>alloxazine (21), isolated as the <em>violurate</em>. The constitution was proved by an independent synthesis from 2-amino-5-dimethylamino-β- diethylaminoethylaniline and alloxan, showing that the violuric acid condensation took place in the 4- position, rather than the 2- position of the benzene ring.</p> <p>The 3-dimethylamino-β-diethylaminoethylaniline was prepared from 3-<em>dimethylamino</em>-<em>p</em>-<em>toluenesulphonanilide</em> by alkylation and hydrolysis. The amine for the alloxan condensation was the reduction product of 2-<em>nitro</em>-5-<em>dimethylamino</em>-β-<em>diethylaminoethylaniline</em>, obtained from β-diethylaminoethylamine and 3:4-dinitrodimethylaniline.</p> <p>The reaction of violuric acid with 3-amino-N-methylaniline was investigated as this could give an alloxazine and/or an isoalloxazine, assuming that this condensation also takes place in a <em>r</em>- position of the benzene ring. A comparison of the ultra-violet absorption spectrum of the product with those of 7-<em>dimethylaminoalloxazine</em>, and 7-<em>dimethylamino</em>-9-<em>methylisoalloxazine</em>, prepared by Piloty's method, showed that it was inhomogeneous. Its sulphuric acid solution was therefore treated with sodium nitrite. The precipitated 7-(N-<em>nitroso</em>-<em>[illegible]</em>-<em>methylamino</em>)-<em>alloxazine</em> was collected and the filtrate gave an olive green <em>azo</em>-<em>dye</em> with β-naphthol in dilute alkali. From the respective yields of these products the ratio of alloxazine to <em>iso</em>alloxazine in the mixture was approximately 10:60. The first stage of the Piloty reaction is probably the attack of the 4-carbonyl of the violuric acid on an amino group of the diamine. If the first stage were the attack of the oximino group in violuric acid on the benzene nucleus the initial product would be an anil of a type that does not cyclise to an alloxazine derivative (Tischler, Wellman and Ladenburg, <em>J. Amer. Chem. Soc.</em>, 1945, <strong>67</strong>, 2165).</p> <p><strong>Part III. The synthesis of some Benziminazoles</strong>.</p> <p>The reaction of phenylacetiminomethylether with <em>o</em>-phenylenediamine N to give 2-benzylbenziminazole was investigated. The free iminoether reacted with the diamine only at elevated temperatures to give poor yields of the benziminazole, but the reaction was acid catalysed. In the presence of one, or two equivalents of acid good yields of the benziminazole were obtained, but if three equivalents were used the yield dropped. This was also found with phenylacetiminothiobenzylether, and the reaction mechanism has been interpreted as an attack of the iminoether cation on an uncharged aromatic amino group.</p> <p>The reaction was extended to various mono-N-alkylated <em>o</em>-phenylenediamines with good results.</p> <p>Three benziminazoles (22–24) were prepared, for antimalarial testing, from the appropriate diamines by the Phillips method; 22 and 23 were inactive against <em>P. gallinaceum</em> in chicks, and the results for 24 are awaited.</p> <p>Acetiminothiobenzylether hydrochloride with 2-amino-5-chloro-γ-diethylaminopropylaniline also gave 23, but only small quantities of 22 were isolated from the reaction of acetiminomethylether hydrochloride with 2-amino-4-chloro-γ-diethylaminopropylaniline under various conditions. The main product of the reaction was provisionally regarded as 2-(N-<em>acetamidino</em>)-4-<em>chloro</em>-γ-<em>diethylaminopropylaniline</em>. In support of the idea that the electrophilic chlorine atom <em>para</em> to the secondary amino group in the amine was reducing benziminazole formation the corresponding 2-amino-4-methoxy-γ-diethylaminopropylaniline gave a much larger yield of 1-(γ-diethylaminopropyl)-2-methyl-4-methoxy-benziminazole with acetiminomethylether hydrochloride than the chloro derivative.</p> <p><strong>Part IV. The synthesis of Pteroylglutamic acid and Related Compounds</strong>.</p> <p>The benziminazole analogue (25) of ethyl pteroate was prepared from chloroacet-2-nitroanilide, which reacted with ethyl <em>p</em>-aminobenzoate to give 4-<em>carbethoxyphenylglycine</em>-2-<em>nitroanilide</em>. On hydrogenation this gave the corresponding amine which cyclised to ethyl N-(2-benziminazole)-4-aminobenzoate (25) under acid conditions. Others completed work in this series, and details are in press (King, Spensley, and Nimmo-Smith, <em>Nature</em>, 1948).</p> <p>N-(2-<em>quinoxaline</em>)-4-<em>aminobenzoic acid</em> (26) was prepared from 2-chloroquinoxaline and <em>p</em>-aminobenzoic acid. It had little effect on the growth of <em>S. lactis R</em>.</p> <p>The pteroylglutamic <em>analogue</em> (28) was prepared by the hydrolysis of <em>ethyl</em> N<sup>4</sup>-(2-<em>quinoxaline</em>)-4-<em>aminobenzoyl</em>-<em>l</em>-<em>glutamate</em> (27), the condensate of 2-chloroquinoxaline and ethyl <em>p</em>-aminobenzoyl-<em>l</em>-glutamate. Its inhibitory action on the growth of <em>S. lactis R</em> was prevented by pteroylglutamic acid, and quantitative biological results are awaited.</p> <p>Many attempts to synthesise N-(quinoxaline-2-methyl)-4-aminobenzoic acid (29) and some of its derivatives were made. It was not found possible to prepare 2-bromomethyl-quinoxaline by any route examined, but quinoxaline-2-aldehyde reacted with <em>p</em>-aminobenzoic acid, and its ethyl ester, to give the corresponding <em>anils</em>. Nothing homogeneous could be isolated from the reduction of N-(quinoxaline-2-methylene)-4-aminobenzoic acid, but a poor yield of <em>ethyl</em> N-(<em>quinoxaline</em>-2-<em>methyl</em>)-4-<em>aminobenzoate</em> (30) was obtained from the reduction of the corresponding ester. One attempt at hydrolysis failed.</p> <p>Methylglyoxal on bromination gave <em>[illegible]</em>, which reacted with 2:4:5-triamino-6-hydroxypyrimidine and <em>p</em>-aminobenzoyl-<em>l</em>-glutamic acid to give a product containing about 5% of pteroylglutamic acid.</p> <p><em>[For the figures to accompany this abstract, please consult the PDF]</em></p></p>
|
|---|