VII. ESTIMATION 4G7 



liter of .sample extract, and treatment of the inoculum must 1)0 carefully 

 controlled. A detailed i)iocedure for the cup-plate assay with L. laclis 

 (8000) has been recently published by Cuthbertson, Pegler, and Jjloyd.^' 

 The basal medium is not chemically defined in that it contains tomato 

 juice. Ascorbic acid is added as a reducing agent. Only a limited series of 

 assays have been reported with this method. It does not appear to be 

 applical)le to Sfreptomyccs fermentation lic]uors which contain vitamin 

 Bi2c-^^ Ii^ the plate test vitamin Bi2c was three to four times as active as 

 vitamin Bia, although clinically its activity is approximately the same.^^ 

 The.se authors did not attempt to eliminate interference due to dcsoxy- 

 ribosides, since such interference can be readily detected visually. Desoxy- 

 ribosides gave faint diffuse zones, and solutions containing both vitamin 

 Bi2 and desoxyribosides gave a dense zone of growth in the center sur- 

 rounded by a diffuse area due to desoxyribosides. At very high concentra- 

 tions of desoxyribosides the center vitamin B12 growth zone is obscured. 



Bessell et al.^^ have briefly described the application of the Davis E. coli 

 mutant to the assay of vitamin B12, using the cup-plate technique. These 

 authors report that the method is relatively free of the usual assay troubles 

 encountered with lactobacilli, and the reproducibility is comparable. The 

 simple medium, inorganic salts and glucose, was a distinct advantage. 

 Methionine at a concentration of 10 to 100 mg. per milliliter gave a growth 

 zone. More recently Harrison et al.^'^ reported in detail on the E. coli cup- 

 plate method, pointing out the superiority of the method for determination 

 of the naturally occurring vitamin Bi2c (nitrosocobalamin). Vitamin Bi2c 

 which occurs in Strepiomyces lic^uors and liver extracts gave assay values 

 four to five times as high by the L. lactis Dorner plate method as those 

 obtained by tube methods. With a L. leichmannii 313 plate method vitamin 

 Bi2c gave fuzzy, poorly defined zones. With the E. coli cup-plate assay, 

 however, vitamin B]2c is approximately equivalent on a weight basis to 

 vitamin B12. The simple salts-glucose medium used earlier for the E. coli 

 cup-plate assay^^ was fortified with a variety of trace minerals and with 

 asparagine which increased the density of the growth zones. The method 

 appears to be remarkably free of effects of non-specific interfering sub- 

 stances and inexplicable variations in growth at various levels of the same 

 sample commonly observed in lactobacilli methods. The method is some- 

 what less sensitive than the E. coli pad-plate method (Table XII) requiring 

 10 m7 per milliliter of sample extract. Methionine at a concentration of 

 1 mg. per milliliter of extract interferes with the assay. 



" W. F. J. Cuthbertson, H. F. Pegler, and J. T. Lloyd, Analyst 76, 133 (1951). 



" E. L. Smith, Proc. Roy. Soc. Med. 43, 535 (1950). 



" C. C. Ungley, D. L. Mollin, and J. V. Dacie, Lancet I, 353 (1950). 



*8 C. J. Bessell, E. Harrison, and K. A. Lees, Chemistry & Industry 1950, 561. 



" E. Harrison, K. A. Lees, and F. Wood, Analyst 76, 696 (1951). 



