MICROBIOLOGICAL PRODUCTION 



addition of sufficient oca'-dipyridyl, to inactivate all but a small 

 amount of iron, increased the yield of ribofiavine/o so that, for instance, 

 in the fermentation of corn-mash, a yield of 2 to 13 mg. of riboflavine 

 per litre was obtained in presence of 1-5 mg. of iron per litre. The 

 addition of sodium sulphite was claimed ^^ to result in consistently 

 high yields, although the maximum amount produced (2 mg. per g. 

 of dry solids) was not appreciably affected. The production of ribo- 

 flavine by fermentation of whey or skimmed milk with CI. aceto- 

 htityliciim was also protected by patents. ^2 Again, the concentration 

 of iron in the mash was found to be very critical, and it was stated 

 that, to obtain the best results, this should lie between i and 3 p.p.m. 

 Yields of 0-24 to 2-2 mg. per gram of dried material were claimed. 

 The sensitivity of the fermentation to iron was confirmed by A. 

 Leviton,^^ ^j^q found that less than 0-2 mg.-atom of ferrous (though 

 not of ferric) iron per litre prevented the formation of riboflavine and 

 destroyed any that was added to the fermentation liquor. Crystalline 

 catalase counteracted the effect of the iron, provided the concentration 

 of iron did not exceed 0-29 mg.-atom per litre, but in the presence of 

 0-33 mg.-atom per litre, catalase had little effect. Low concentrations 

 of sodium dithionite minimised the effect of iron, whilst potassium 

 iodide enhanced it. Riboflavine was destroyed by hydrogen peroxide 

 in the presence, though not in the absence, of iron, and the addition of 

 catalase or potassium iodide stabilised the riboflavine, although it 

 speeded up the destruction of the hydrogen peroxide. 



A. Leviton suggested that the synthesis and decomposition of ribo- 

 flavine during fermentation were intracellular processes involving the 

 formation of hydrogen peroxide. When iron was present, the hydrogen 

 peroxide destroyed the riboflavine inside the cells, so that catalase 

 had no effect. When less iron was present, hydrogen peroxide and 

 riboflavine accumulated in the cell and then diffused into the culture 

 fluid, where catalase destroyed the hydrogen peroxide and thus pro- 

 tected the riboflavine from destruction. 



The production of riboflavine by butyl alcohol-producing bacteria 

 has also been patented.^* 



Fermentation with Candida 



A large number of different strains of yeast also produced ribo- 

 flavine, although mostly in moderate amounts only.^^ Certain 

 strains, however, were found to produce 10 to 60 mg. per litre of 

 fermented liquor. ^^ In particular, most species and varieties of 

 Candida produced substantial amounts of riboflavine, though only 

 when glucose was used as the carbon source.^' The highest yields 

 were obtained from C. guillermondia and C. tropicalis var. Rhagii}^ 



149 



