166 ANTIBACTERIAL SUBSTANCES FOR TREATMENT OF INFECTIONS 



markedly, a fact difficult to reconcile with the hypothesis of direct competition 

 between the two (Hirsch 1944). (See p. 165, and Johnson et al. 1944, for com- 

 petition of relatively unrelated structures.) In the second place, sulphanilamide 

 may be antagonized, as we have seen, by substances that have no structural 

 relation to it. 



The hypothesis postulates that jo-aminobenzoic acid is an essential metabolite 

 in a wide variety of plants and animals (see Fildes 1940«, 1941). That it is a growth 

 factor for certain bacteria and animals is no more than suggestive ; at present 

 its wide distribution, and a similar distribution of enzyme systems that utilize 

 it, remain unproven. Two other of Henry's points, the limited value of the 

 evidence of (a) increased ^-aminobenzoic acid production by sulphonamide- 

 resistant strains, and (6) analogous systems of growth factors and inhibitors, we 

 have already dealt with. There remains the most cogent objection, that animal 

 and bacterial respiratory systems are inhibited by sulphanilamide. In two systems, 

 the sea urchin's egg (see Henry 1943) and perhaps Br. tularensis (Tamura 1944), 

 the inhibition is not reversed by j9-aminobenzoic acid. Sulphonamides appear to 

 have a direct action on the respiratory enzymes of bacteria, both aerobic and 

 anaerobic (Sevag and Shelburne 1942, Dorfman and Koser 1942, Berkman and 

 Koser 1943) ; on bacterial dehydrogenases (Macleod 1939, Fox 1942), and on 

 cocarboxylase (Sevag, Shelburne and Mudd 1942, Sevag et al. 1943). Sevag and 

 his colleagues conclude that sulphonamides inhibit oxidative enzymes (see also 

 Sevag and Green 1944) and, therefore, the growth of the bacteria. Henry groups 

 the sulphonamides with " indifferent " cell inhibitors like narcotics, inhibiting a 

 specific fraction of the total oxidative reactions of the cell upon which cell division 

 depends. Like the narcotics, the sulphonamides stimulate in low, and inhibit 

 in higher, concentrations (Finklestone-Sayliss et al. 1937, Green and Bielschowsky 

 1942, Lamanna and Shapiro 1943). Like narcotics, they act upon a wide variety 

 of tissues. For example, they inhibit the growth of tissue cultures of tomato 

 plants (Bonner 1942) and wheat and oat seedlings (Brian 1944, Jones 1944) ; and 

 they inhibit the reproductive division of flagellates (Lwoff et al. 1941). The 

 specificity of j9-aminobenzoic acid as an antagonizer does not imply that this sub- 

 stance, or any other antagonizer, necessarily acts by specific interference. The 

 antagonizer may act as a non-specific growth stimulant (see, for example, Rantz 

 and Kirby 1944fl), though it is pertinent to note that Lynch and Lockwood (1941) 

 distinguished clearly between the antagonistic action of peptone in a human 

 serum medium, which was due to growth stimulation, and that of ^-aminobenzoic 

 acid, which was not. Alternatively, antagonizers may combine directly with the 

 inhibitor, forming an inactive complex. This is unUkely to be the case with 

 ^-aminobenzoic acid and a sulphonamide, for they do not react in the absence 

 of bacteria. The antagonism of mercapto compounds to the disinfectant action 

 of HgClj (Fildes 19406) may be of this nature. Again, cationic antiseptics of the 

 long-chain fatty-acid type, which presumably act by disorganizing the hpoid 

 membrane of the bacterial cell, are antagonized by the addition of phosphoUpins 

 (Baker, Harrison and Miller 1941). Finally, antagonizers may shield the sus- 

 ceptible enzyme system from the inhibition, without blocking the enzyme action. 

 Here again cationic antiseptics provide a model for the hypothesis. (See also 

 Penicillic Acid, p. 178). Valko and DuBois (1944) reversed the antibacterial action 

 of a highly toxic cation, N-dodecyl dimethyl ammonium chloride, displacing it by 

 the addition of a relatively non-toxic cation like N-hexadecyl dimethyl ammonium 



