170 ANTIBACTERIAL SUBSTANCES FOR TREATMENT OF INFECTIONS 



that the structure of the interfaces generally depends. The lipoid solvents (chloro- 

 form, toluene, etc.) fall into this category as do the phenols, cresols and soaps, 

 for, although these can act as protein coagulants, they are generally antiseptic 

 at higher dilutions than are necessary for coagulant action. Of greater interest 

 are the selective agents that are not classifiable in these general terms. 



Crystal violet, a triphenylmethane dye used for the surface treatment of wound 

 infections, is bacteriostatic in low concentrations — an effect, according to a number 

 of observers, which is due to the poising of the Eh of the bacterial environment at a 

 level too high for cell development (Dubos 1929, Ingraham 1933, Fildes 19406, 

 Hoffman and Kahn 1944 ; but see Steam 1930, for a contrary view). In a like man- 

 ner, quinones are antibacterial and redox-active. There is, however, little evidence, 

 either in synthetic quinones, or in natural quinones like citrinin and penicillic 

 acid (see below), that they owe their activity to interference with optimal Eh values 

 for bacterial growth, though Page and Kobinson (1943) found that the E'o values 

 of a series of quinones which were markedly active against Staph, axireus clustered 

 round a potential of about + 0-03 volt. 



Antibacterial activity in many compounds may be interpreted in terms of 

 their capacity to neutrahze acidic or basic groups in the bacterial cell forming 

 feebly ionized complexes (see Stearn and Stearn 1924). The anionic antiseptics, 

 like soaps and the " acid " dyes, combine with basic groups. 'The kationic anti- 

 septics, which include basic dyes Uke brilUant green or crystal violet, the amino- 

 acridine antiseptics, and the higher aliphatic amines hke cetyltrimethylammonium- 

 bromide combine with acidic groups, and since acidic groups preponderate in 

 most bacteria, particularly in Gram-positive organisms, it is to be expected that 

 the kationic antiseptics will be more active than anionic antiseptics, and that 

 Gram-positive organisms will be more readily killed than Gram-negative ; and 

 this is indeed the case. 



Another aspect of the difference in susceptibility of Gram-positive and Gram- 

 negative bacteria was pointed out by Miller and his colleagues (1942). Proceeding 

 from the fact that gramicidin, and the anionic detergents, acting only on Gram- 

 positive bacteria, were antagonized by certain phospholipins, they suggested that 

 Gram-negative bacteria owed their insusceptibility to the presence of such phospho- 

 Upins. In support of this hypothesis, they record experiments in which Bad. coli 

 was made susceptible to tyrothricin (a mixture of gramicidin and tyrocidin) by 

 the addition of protamine sulphate, a precipitant of phospholipins. Thus a slightly 

 inhibitory concentration of the protamine, together with a non-inhibitory concen- 

 tration of tyrothricin, completely inhibited the respiration of Bad. coli. 



In the mono-aminoacridine series of compounds, antibacterial activity increases 

 with basic strength up to a certain point, but compounds with a basic strength 

 greater than this are no stronger as antiseptics (Albert, Rubbo and Goldacre 1941). 

 It is probable that the compounds of extreme basicity are too active to form 

 non-ionized neutralization compounds in the bacteria. We have already noted 

 Albert and Goldacre's view (1942) that basicity and activity of the sulphonamides 

 are associated, and it is of interest that in both series of drugs those whose structure 

 permits resonance and induction of polarity tend to be the most active. 



In addition to basicity, surface activity may be a feature of a compound con- 

 tributing to the bactericidal power of a kationic antiseptic. Albert suggests 

 that the distinction between the acridines and the kationic detergent antiseptics, 

 the one group mainly bacteriostatic, the other bactericidal, may possibly lie in 



