MECHANISM OF THE DIABETOGENIC ACTION 413 



it is hard to account for the selective action. There is today very little 

 evidence in favor of this theory. 



(B) Inhibition of enzymes other than hexohinase. Very few enzymes mark- 

 edly susceptible to alloxan have been found, as discussed earlier (page 

 400). Some of the cycle enzymes are certainly inhibited at low concentra- 

 tions (Younathan, 1962) but if this is the site of action in the /5-cells, some 

 further assumptions must be made to account for the selective damage 

 to these cells. Since the mitochondria used in this work were from kidney, 

 one might expect on this basis a rather marked renal toxicity, whereas 

 only doses higher than the diabetogenic seem to affect the kidney. Bur- 

 gen and Lorch (1947) attempted to correlate diabetogenic activity with 

 the inhibition of alkaline phosphatase, but the fact remains that no in- 

 hibition of this enzyme could be demonstrated in vivo. The inhibition of 

 UMP kinase demonstrated by Harris (1963) could hardly account for the 

 initial /?-cell damage, nor could inhibition of acetylating systems as re- 

 ported by Cooperstein and Lazarow (1958). The malate dehydrogenase 

 of islet tissue is unaffected by diabetogenic doses of alloxan up to 12 hr, 

 but a steady loss of activity is then observed, probably due to secondary 

 changes resulting from necrosis (Dixit et al., 1963). Although malate dehy- 

 drogenase is about 3 times more active in the islets than in acinar tissue, 

 it is certainly not involved in the alloxan action. The most interesting 

 approach of this type has been by Lazarus et al. (1962). They studied his- 

 tochemically the effects of diabetogenic doses of alloxan on several enzymes 

 in the /5-cells. No early changes were noted for succinate dehydrogenase, 

 cytochrome oxidase, glucose-6-P dehydrogenase, NADP diaphorase, glu- 

 cose-6-phosphatase, or acid phosphatase; during the later necrotic stages, 

 of course, the enzyme activities fall markedly. The one enzyme tested 

 showing a rapid inactivation is the extramitochondrial ATPase. It was 

 stated that cortisone administration causes this enzyme to disappear from 

 the /5-cells and simultaneously there is an increase in the resistance to al- 

 loxan, which is strange since, if the ATPase is of such importance to the 

 cells, one would expect its disappearance to be detrimental. Possibly cor- 

 tisone in some manner alters the /?-cell metabolism so that this enzyme 

 is not necessary. If this ATPase were located in the /5-cell membrane, its 

 inhibition might have disastrous effects on the membrane structure. There 

 is need for a thorough investigation of various ATPases in vitro and pos- 

 sibly this could be done using toadfish or sculpin islets. 



(C) Inactivation of glutathione. For some years Lazarow (1954 a) empha- 

 sized the imjiortance of glutathione in the diabetogenic action of alloxan, 

 inasmuch as the reaction between them occurs so readily, and it was quite 

 easy to imagine how glutathione - might be of importance in the y5-cells, 

 particularly relative to certain enzymes such as 3-phosphoglyceraldehyde 



