Cephalosporin C (I), the amine (11), and derived cephalosporanic acids (111; R1 = H) are converted into new derivatives by replacing the acetoxygroup with various sulphur, nitrogen, and carbon nucleophiles.7-Aminocephalosporanic acid ( 11) is produced by the action of thiourea on 7-chloroacetamidocephalosporanic acid or spontaneously from 4'-chlorobutanamido-or 5'-chloropentanamidocephalosporanic acids as a result of intramolecular displacement.CONVERSION of the antibiotic cephalosporin C (I) into 7-aminocephalosporanic acid (11) and subsequent acylation to give a series of 7-acylaminocephalosporanic acids (111 ; R1 = H) have been de~cribed.l-~ We have converted such acids into their methyl esters (111; R1 = Me), lactones (IV), and I-oxides (VIII; Y = OAC).~ The stereochemistry of the oxides has not been studied. Hale, Newton, and Abraham described 5 the transformation of cephalosporin C (I) into the pyridinium betaine (VI; X = C,H5N+) and showed that a range of substituted pyridines formed similar derivatives (C, compounds) by replacing the acetoxy-group. The formation of a Bunte salt ( I X ; Y = S,O,Na R = [CH,],*CH(NH,)*CO,H), when cephalosporin C is treated with sodium thiosulphate was proposed 6 to account for the resulting increase in antibacterial activity. We have studied the substitution of the acetoxy-group in salts of 7-acylaminocephalosporanic acids (I11 ; R1 = H) by azide, NN-dimethyldithiocarbamate, thiobenzoate, and

The combined action of bases and acid anhydrides on 7-acylamidocephalosporanic acids, and the action of bases on their esters, set up equilibria in which the corresponding A2-isomers predominate. The isomerisation represents a prototropic shift probably favoured by the sulphur atom in the dihydrothiazine ring. The acetoxy-group in the A2-compounds can be replaced by nucleophiles.Treatment of methyl 3-acetoxymethyl-7~-phenylacetamidoceph-3-em-4-carboxylate 1 [-oxide with a base res u I ts i n d eca rboxy I a t i o n . Attempts to prepare 7 pp he n y I a ceta m i d o ce p h -2 -em -45car boxy I i c acid 1 toxi d e resulted in formation of its A3-isomer and some carbon dioxide.

By J. D. COCKER, T. G. HALSALL, and (in part) A. BOWERS. fi-Anisic acid, P-phenylpropionic acid, and two new diterpene acids have been isolated, in addition to straight-chain fatty acids and long-chain hydroxy-acids. One of the diterpenes, labdanolic acid, is a bicyclic hydroxy-acid, C2oH&3.The other is a bicyclic compound, C20H3203, which contains an @-unsaturated ketone grouping.The acid fraction from gum labdanum has been examined.

The structure of labdanolic acid, a constituent of gum labdanum, has IN Part I the isolation was described of labdanolic acid, a new bicyclic diterpene hydroxyacid, C,H3,03. It has now been shown to have structure (I).* Labdanohc acid was obtained by relatively mild hydrolysis of its methyl ester, showing that its carboxyl group was unhindered and not situated in the sterically hindered 4-position as in the majority of diterpene acids. The hydroxyl group, on the other hand, was unreactive, being acetylated only under vigorous conditions. On attempted mild oxidation of methyl labdanolate with chromic acid the ester was recovered unchanged. Reduction of the ester with lithium aluminium hydride gave the diol (111) which on mild acetylation afforded a monoacetate.Dehydration of methyl labdanolate with phosphoryl chloride in pyridine gave a homogeneous product (IV), the infrared spectrum of which indicated the presence of a vinylidene group (bands at 890 and 1645 cm.-l). This was confirmed by ozonolysis, formaldehyde and a keto-ester (V) being obtained. Reduction of the dehydration product, which was also formed when methyl labdanolate was acetylated under vigorous conditions, gave the alcohol (VI). This was characterised by conversion with osmium tetroxide in pyridine into a trio1 which is provisionally formulated as (VII), addition being assumed on the less hindered (a) side of the molecule (see below, however).The keto-ester (V) formed an oxime but not a 2 : 4-dinitrophenylhydrazone. These observations together with the carbonyl bands in the infrared spectrum at 1735 cm.-l (ester band) and 1712 cm.-l (aliphatic or 6-membered ring ketone band) suggested that the keto-ester contained a somewhat hindered keto-group in a six-membered ring. The formation of a hindered keto-group in a side chain appeared unlikely as it could only be accounted for by a branched side chain unlike any found in bicyclic diterpcnes. In turn, methyl labdanolate must have contained the grouping (VIII), with the hydroxyl group in the equatorial conformation to account for the formation of an exocyclic rather than an endocyclic double bond on dehydration.Dehydrogenation of the unsaturated alcohol (VI) with a platinum-charcoal catalyst gave a trimethylnaphthalene. It was identified as the 1 : 2 : 5-trimethyl isomer by comparison of its ultraviolet absorption spectrum with the spectra of the trimethylnaphthalenes described by Heilbronner, Frohlicher, and Plattner and by formation of the 1 : 3 : 5trinitrobenzene adduct. In addition 1 : 2 : 5 : 6-tetramethylnaphthalene was obtained and characterised as its 1 : 3 : 5-trinitrobenzene adduct.* The nomenclature used in this paper is based on the parent hydrocarbon (11) which is provisionally This It is to avoid confusion in been shown to be (I).The hydroxyl group was accordingly tertiary. called labdane and, on the advice of the Editor, is numbered on the principles used for steroids. hydrocarbon was previously called agathane and numbered differently. numbering that a new name is suggested.

No abstract