RIBOFLAVINE 



R. Kuhn and F. Weygand ^^ condensed L-arabinamine and D- 

 xylamine with halogeno-o-nitrobenzene and halogeno-o-nitro xylene. 

 The compounds obtained on reducing the products with stannous 

 chloride in presence of alloxan had properties similar to those of 

 ribofiavine. In a later paper ^^ they described synthetic 6 : 7-dimethyl- 

 9-tetraacetyl-L-arabityl-isoalloxazine, m.p. 298° C, as resembling 

 the tetraacetyl derivative of riboflavine in absorption spectrum, colour 

 intensity, the effect of pH on the fluorescence and in optical rotation. 

 The activity of this substance as a substitute for Warburg's yellow 

 enzyme could not be tested, as it could not be adsorbed on the colloidal 

 carrier. 



Karrer et al}"^ obtained the same substance by a slightly different 

 route, in which arabinose was reductively condensed with N-mono- 

 acyl- or N-monocarbethoxy-o-phenylene diamine giving, after hydro- 

 lysis, N-((?-aminophenyl)-arabinamine ; other sugars gave analogous 

 compounds. These were all converted to the corresponding isoalloxa- 

 zines by condensation with alloxan. Karrer et al. also noted the close 

 resemblance of the arabityl compound to ribofiavine, but found slight 

 differences in m.p. and optical rotation. P. Karrer, K. Schopp and 

 F. Benz ^^ then synthesised the D-xylityl and D-ribityl compounds 

 and found that the latter was identical with ribofiavine. 



Ribofiavine therefore has the structure : 



H H H 



CH2 C C C— CH2OH 



I OH OH OH 



I II I I 



This paper is therefore the first to describe the synthesis of ribo- 

 fiavine, although a patent ^^ based on this method was ante-dated by 

 a patent covering the general reaction, filed earlier by the LG. (see 

 page 142). R. Kuhn ^o announced the successful synthesis five weeks 

 after Karrer and his colleagues. 



The identity of the synthetic and natural substances was con- 

 firmed by von Euler et al.,^^ who reported the results of growth tests 

 with synthetic ribofiavine and other homologues prepared from 

 different sugars. The D-ribityl compound was fully active, the 

 L-arabityl compound was slightly active, whilst the other derivatives 

 tested were inactive. These and other compounds subsequently 

 prepared are discussed in more detail on page 206. 



A better method of condensing alloxan with pentitylamino 



138 



