76 



own phosphate by means of the £* generated by the oxidoreduc- 

 tion in its alloxazine. To do this the alloxazine would have to have 

 three active groups in a vicinal position. It has three such groups, a 

 C=0 at position 2, an NH at 3, and another, C=0 at 4 (Fig. 

 20) and so the question is whether it is sterically possible to bring 

 the hypothetic triphosphate formed on the furanose into a position 

 in which one 0~ of each of the three phosphates touches upon one 

 of the three active groups of the alloxazine. 



The answer is given by Fig. 21, an alloxazine-triphosphate built 

 up of the Courtauld atomic model. The three atoms pointed at by 

 the single-headed arrows are 0"'s of the three phosphates while 

 the three double-headed arrows indicate the =0, NH, and =0 

 on the alloxazine. These are in close touch. If, in ATP, the sta- 

 tistical chances for the possibility of the formation of a complex 

 were remote, they are still more distant here and so we again can 

 say that the structure pleads for the assumption that the riboflavine 

 does not happen to be able to form such a complex but is made 

 that way. We can also add that any change on the riboflavine, as 

 the replacement of ribose witli ribofuranose or a different location 

 of the O's and NH on the isoalloxazine, would make the forma- 

 tion of such a complex impossible. So the structure of the ribo- 

 flavine molecule becomes accessible to a functional interpretation. 

 The CH3 groups at 6 and 7 may act as electron donors, resonating 

 with the CO's and NH. For this function we could expect them to 

 be located on the opposite end of the molecule, where they actually 

 are. 



Once the formation of a triphosphate is structurally possible, 

 the question arises whether this is possible energetically. The spec- 

 tral properties of riboflavine give us a lead. The riboflavine has 

 three absorption bands, one at 260, one at 375, and one at 445 

 vcifji. So if the molecule is excited by UV its electrons are raised to 

 one of the UV absorption bands and then drop in an "internal 

 conversion" to the lowest singlet level and emit from here the well 

 known greenish-yellow fluorescence with a maximum of its emis- 

 sion band at 540 m^M (Si -> G in Fig. 5). To excite electrons to 



