RELATION OF STRUCTURE AND ANTIBACTERIAL ACTIVITY 167 



chloride. Here we have a similarity of structure between inhibitor and antagonizer, 

 but the antagonizer is not on the face of it likely to be an essential metabolite 

 for the bacteria concerned. It is, however, doubtful whether this mechanism 

 will serve as a model for all inhibitor-antagonist relationships, for Mcllwain (1944) 

 has recently shown, both with sulphonamide-jo-aminobenzoic acid and with pantoyl- 

 taurine-pantothenic acid systems, that there is no gross displacement of antagonizer 

 by inhibitor in the resting bacterium. Inhibition and antagonism take place 

 only in actively growing cells. 



There are at present insufficient data for judging the validity of these con- 

 flicting hypotheses, and fresh data may demand still different ones. It is clear 

 that the Woods-Fildes hypothesis, though it holds for certain classes of cell- 

 inhibitors, is not applicable to other classes. Further knowledge of the hypo- 

 thetical 7?-aminobenzoic acid metabolism, and its relation to other metabolic 

 functions inhibited by sulphonamides, may permit a modification of the hypothesis 

 to include discrepant facts. Meanwhile, its neatness and the brilliance of its 

 applications must not be allowed to justify the assumption of its universality. 



The Relation of Chemical Structure and Antibacterial Activity in the Sulphonamides. 



The activity of the sulphonamides depends on the integrity of the H2N<(' ^S 



radicle. Replacement of the ^ara-amino group produces inactive compounds, 

 except in the case of the ^ara-nitro (see King and Henschel 1941) and para- 

 hydroxylamine groups. 



The compound known as Marfanil (Fig. 27) is an exception to this generalization; 

 it is active, even though an ainino-methyl group is substituted for the amino group. Its 

 activity, however, is of a different kind from that of the sulphonamide series, since it 

 is not antagonized by ^'-aminobenzoic acid, or by substances present in blood and pus ; 

 nor is there any cross-resistance between Marfanil- and sulphonamide-resistant strains, 

 (see below) (Miller et al. 1940, Klarer 1941, Domagk 1943). 



The activity of hydroxylaminobenzene sulphonamide led Mayer (1937) to postu- 

 late that sulphanilamide acted by reason of its oxidation to this compound. Mellon 

 (1940) assumed that the hydroxylamino compound, by inhibiting catalase, produced 

 a lethal accumulation of HjOg in the bacteria. Burton and his colleagues (1940) 

 measured the in vitro activity of both the nitro- and hydroxylamino-forms, and 

 concluded that the " active " substance in sulphanilamide action, at least on 

 aerobic bacteria, was the hydroxylamino-form or an oxidation product inter- 

 mediate between it and the nitro-form. Green and Bielschowsky (1942) objected 

 to this interpretation on the grounds that the hydroxylamino-form was not antag- 

 onized by ^-aminobenzoic acid, an objection which is valid only if it is assumed 

 that the " active " substance must be formed prior to its absorption by the enzymes 

 for which j9-aminobenzoic acid competes. Nevertheless, competition between the 

 two is demonstrable (McLeod et al. 1944), and Mellon's interpretation of the signifi- 

 cance of the hydroxylamino group remains a possible one. 



The hypothesis of action by hydroxylamino compounds is in essence an extension 

 of a more general hypothesis of activity as a function of redox -potential developed by 

 the active substance. The latter hypothesis (see Dubos 1929, below) was elaborated 

 by von Jancso and von Jancso (1936) to cover chemotherapeutic interference in general, 

 and relates the activity to the power to poise the Eh of the cell and its environment at 

 a level unfavourable for cell growth or development. With sulphonamide, it is supposed 

 that the high lethal oxidizing intensity is achieved by the HjOj. 



