158 THE BIOCHEMISTRY OF B VITAMINS 



used. 127 It is now believed that the difference is due to the presence of 

 inactivated apocarboxylase and possibly other apoenzymes which com- 

 bine with some of the added coenzyme and render it unavailable to the 

 active apocarboxylase. 12<5 



That the pyrophosphoric acid group is at least partially responsible 

 for the association of the apoenzyme with the coenzyme is indicated by 

 the inhibition of carboxylase activity by other molecules containing 

 pyrophosphate groups — adenosine triphosphate 123 and thiazole pyro- 

 phosphate 12s Ca ++ ions interfere with the formation of the holoenzyme, 

 presumably by competing with the Mg ++ or Mn ++ ions. 129 Consequently, 

 it is desirable to use water free from calcium in the preparation of 

 apoenzymes. This interference may account for some of the discrepancies 

 found in the earlier studies on the recombination of the carboxylase 

 holoenzymes. 



Phosphorylated thiamine does not pass through cell membranes 

 easily. 130 This accounts for its relative inactivity as a thiamine source 

 for certain microorganisms, and explains why it is not as effective as 

 thiamine in stimulating decarboxylations by tissues from deficient ani- 

 mals. 131 The impermeability of cell membranes to the intact coenzyme 

 may account for the effects observed when thiamine and its pyrophos- 

 phate were tested for their relative activities in reversing the inhibition 

 of bacterial growth induced by pyrithiamine. The vitamin analogue more 

 effectively inhibits the coenzyme than it does the vitamin— a phenomenon 

 not ordinarily encountered. 132 To account for this, it was postulated that 

 thiamine is attached to the apoenzyme before it is phosphorylated, and 

 that the coenzyme so formed is more firmly bound than is preformed 

 coenzyme. An equally logical explanation is that the coenzyme added to 

 the medium must be hydrolyzed before absorption can take place, thus 

 necessitating subsequent resynthesis. The thiamine added in the free form 

 would be more rapidly absorbed than would the thiamine which had to 

 be first liberated from the coenzyme; hence the former would produce a 

 higher intracellular ratio of thiamine to pyrithiamine and would be the 

 more effective agent for reversing the inhibition. 



Reactions Catalyzed by the Thiamine Coenzyme. The enzymatic reac- 

 tions in which thiamine has been demonstrated to function in vivo are 

 limited to only two substrates, pyruvic acid and a-ketoglutaric acid, or 

 their degradation products. Some of these enzyme preparations have been 

 found to decarboxylate other a-keto acids — for example, a-ketobutyric 

 and a-ketovaleric acids. The latter compounds, though, have never been 

 shown to be a part of metabolic processes, and there seems little reason 

 to believe that these substances normally occur in vivo. Two other a-keto 

 acids which are important intermediates often formed during metabolism, 



