69 



for it, has formed a ring with the purine. Naturally, for energy to 

 pass from the phosphate to the purine, the contact would have to 

 be an intimate one. H-bridges, possibly, formed between the O's 

 and N's might do, having been shown by Gergely and Evans that 

 H-bonding can establish relations in w^hich tt orbitals overlap. It is 

 thus possible that the ATP molecule, activated by myosin, connects 

 its two ends, thus opening the way for an energy' transmission 

 from one to the other. 



This, however, does not explain the role of the bivalent ions, 

 Ca and Mg, both of which can accelerate the ATP-ase activity of 

 myosin. The possible answer to this problem w^as given by a 

 chance observation. As discussed in Chapter 4, we may expect 

 energy transmitters to be fluorescent. Since there are violent shifts 

 in energ)' during contraction, the author expected to find in muscle 

 a fluorescent energy^ transmitter in high concentration, and pre- 

 pared alcoholic extracts of muscle, expecting them to show strong 

 fluorescence under the UV lamp. They showed none. However, if 

 a bivalent metal, as Mg, Ca, or Zn, was added (as chloride), an 

 intense blue fluorescence appeared. The fluorescent substance was 

 isolated, identified by McLaughlin, Schiffman, and the author and 

 found to be the metal complex of inosinediphosphate, IDF, the 

 substance produced from ATP by the loss of its terminal phos- 

 phate and its hydrolytic deamination. 



The probable structure of the metal complex is elucidated by 

 the close analogy of inosine and oxyquinoline (Fig. 17a and b). 

 The latter is known to form with Mg in alcoholic solution a very 

 stable, strongly fluorescent chelate ( Fig. 17b). Evidently, an 

 analogous chelate w^as formed by the inosine (Fig. 17a). 



Mg and Ca are known to form very stable, coordinative com- 

 plexes with polyphosphates, so the possibility is given that the 

 metals form with their four coordinative valencies a quadridentate 

 chelate, connecting the two ends of the ATP moleaile. This struc- 

 ture is illustrated by Fig. 18. The model in Fig. 16, shows that 

 there is just enough room left for an Mg between the two N's and 



