204 J. B. Neilands 



0-5 mg of iron/1. , it seems reasonable to conclude that it is an example of a 

 metabolic product which requires iron for its metabohsm and which, at the 

 same time, happens to be a very effective ferric ion complexing agent. It has 

 been repeatedly observed that itoic acid production cannot be obtained in 

 fermentation vessels constructed even of high quality stainless steel. Assuming 

 that itoic acid is made in undetectable amounts when there is a plentiful 

 supply of iron, its augmented production in iron deficiency would result in 

 the complexing and solubilization of iron from all available sources. The 

 soluble metal ion could then be transferred to, or into, the cell where it could 

 be released to the enzyme-forming systems by simple reduction. 



The requirement of many species of green plants for chelated iron has, of 

 course, been frequently encountered. In higher animals the stomach HCl 

 may, to some extent, replace the function of the complexing agents which 

 appear to be required by other forms of life. 



The demonstration that both ferrichrome and ferrichrome A contain 



bound "hydroxylamine" is of special interest in connection with the structure 



and function of these compounds. The presence of such an oxidized form of 



nitrogen proves the occurrence in the ferrichrome molecules of the following 



general structure: 



O— R" 



I 

 R'— N— R' 



The ferric chloride reaction and the ease of liberation of "free NHgOH" 

 suggest that R" = H and that either R' or R" is an acyl substituent. The 

 p^a of approximately 9 found in the metal-free compounds (Neilands, 1957) 

 also supports the conclusion that the NHgOH is present as a hydroxamic 

 acid. Acetylhydroxamic acid has been found in Tondopsis utilis by Virtanen 

 and Saris (1956). That the remaining substituent {R' or K") is a hydrogen 

 atom is rendered probable from the ease of formation of "free NH2OH" and 

 also from the fact that the ^Ka at pH 9 is spectrophotometrically operable 

 (Neilands, 1957). The spectral changes observed with iron-free ferrichromes 

 in dilute alkali, previously assigned to a keto-enol transformation, can in 

 all probability be accounted for by the hydroxamic-hydroximic transforma- 

 tion. Since there are three such residues per atom of iron and since there is 

 much evidence to indicate that ferrichrome is not made up of three separate 

 fragments, the characterization of the iron-binding centre of the molecule as 

 a polyhydroxamic acid seems proven beyond reasonable doubt. Experiments 

 are now in progress in which an attempt is being made to remove the hydroxyl- 

 amine under very gentle conditions so that the resultant carboxyl groups can 

 be esterified and subsequently split with excess hydroxylamine. The synthetic 

 ferrichrome obtained in this manner should resemble the natural product in 

 all respects. 

 The occurrence of bound hydroxylamine in natural products is quite 



