BERNARD L. STREHLER 235 



we and others have made on in vitro luminescence. Since it is pos- 

 sible to obtain luminescence with reduced flavins as substrate and 

 since the oxidation of this compound by molecular oxygen liberates 

 only 37 kcal (approx.) per mole of flavin oxidized (to water and 

 oxidized flavin), it seems unlikely that the oxidation of a single flavin 

 molecule would furnish enough energy to produce luminescence and 

 indeed this fact may be responsible for the fact that most flavin auto- 

 oxidase catalyzed reactions do not proceed with luminescence. 



Thus it seems highly probable that the energy released when more 

 than one flavin molecule is oxidized must be channeled into a single 

 flavin molecule (Strehler and Shoup, 1953; McElroy and Strehler, 

 1954 ) . How could this be accomplished? According to work of Drew, 

 the luminescent reaction involving peroxide and 3-aminophthalhydra- 

 zide is in fact a dismutation reaction between two peroxide molecules 

 with the fluorescent-chemiluminescent molecule acting only as an in- 

 termediary in the process (Drew, 1939). Similarly, riboflavin chemi- 

 luminesces in the presence of peroxide and a metallic activator. Inas- 

 much as the chemiluminescent emission of this reaction is quite similar 

 to the fluorescent emission of oxidized flavin, it seems rather un- 

 likely that the main chromophoric grouping of the flavin molecule is 

 destroyed during the reaction. Rather, flavin may be acting as a cata- 

 lyst for peroxide decomposition. 



In bacterial luminescence we would postulate that the reduced 

 flavin formed by reaction between DPNH2 and oxidized flavin is 

 oxidized with the intermediate formation of a compound equivalent 

 to but perhaps not identical with peroxide and that this peroxide 

 either reacts with another peroxide in intimate association with the 

 flavin molecule and constituent protein of the luciferase, or that the 

 peroxide analog thus formed oxidizes another reduced flavin molecule. 

 The dismutation of two peroxide molecules to form oxygen and water 

 hberates about 50 to 54 kcal, which is approximately the energy 

 required to excite a molecule to emit in the blue-green region of the 

 spectrumTThe scheme presented in Fig. 14 represents the most con- 

 sistent interpretation of the data we have been able to formulate. In 

 this scheme DPN is reduced by some metabolic intermediate such as 

 maHc acid or t>'pical Embden-Meyerhof components. Reduced DPN 

 thereupon reacts with oxidized FMN in the presence of a diaphorase 



