372 4. ALLOXAN 



was found by Richardson and Cannan (1929) to be 0.048 at a pH of 1.08, 

 but unfortunately it was not possible to determine the value near neutra- 

 lity.* Electron-spin resonance studies show that free radicals are formed 

 during the reduction of alloxan by glutathione, and these free radicals 

 may play a role in the formation of alloxantin (Lagercrantz and Yhland, 

 1963). 



Stability of Alloxan 



Crystalline alloxan is not stable at room temperature in air, and de- 

 composes slowly to form alloxantin, oxalate, urea, and COg (and possibly 

 alloxanic acid) (Archibald, 1945). Several impurities may be present in 

 commercial samples of alloxan. Pure alloxan does not give a purple color 

 with Ba(0H)2 but several of the decomposition products do, so that this 

 may be used as a rough test of purity. In accurate work it is advisable to 

 prepare alloxan (Hartman and Sheppard, 1955), obtain a fresh prepara- 

 tion from commercial sources, or recrystallize it in the anhydrous form 

 from acetone or glacial acetic acid. It should be stored near 0° in the dark. 

 Since alloxan is reasonably stable below pH 3.5, acid solutions may be 

 made up and neutralized immediately before use, or in animal work injected 

 without neutralization. 



The instability of alloxan in neutral solutions has been long recognized, 

 and was noted by Dunn et al. (1943 b) in their initial studies on the dia- 

 betogenic action, neutralized solutions rapidly becoming inactive. Karrer 

 et al. (1945) found that a 7 mM alloxan solution kept at pH 7.5-8 and 37^ 

 contained no alloxan after 15 min, and Lazarow et al. (1948) stated the 

 half-life of alloxan in solution at pH 7.4 and 37^ to be about 1 min. In 

 solutions containing biological material, such as serum, blood, or tissue 

 extracts, alloxan is even more unstable due to other reactions not occur- 

 ring in simple solutions (Leech and Bailey, 1945) and disappearance of 

 alloxan may take place within a few seconds, even at a lower temperature 

 (Briickmann, 1946). The effects of blood components are well seen in the 

 results of Veksler (1956), but the rates of destruction are relatively slow 

 because of the low pH (Fig. 4-1). 



The instability is due to the conversion of alloxan to alloxanate by an 

 internal benzilic acid rearrangement (Schlieper, 1845). The rate of the con- 

 version was stated by Labes and Freisburger (1930) to be proportional to 

 the concentration of the alloxan anion, so that the rearrangement would 

 depend on enolization as in the scheme outlined by Klebanoff and Green- 

 baum (1954). However, Resnick and Wolff (1956) found that N, iV'-dimeth- 



* I regret being so perverse as to be critical of the customary formulations of these 

 compounds and their reactions, but after surveying the early literature on the struc- 

 tural characterizations of alloxan derivatives, I am not convinced that they are correct. 



