ROliERT L. SEARLS AND I). R. SANADI 



59 



o 



CO 



< 



1 1 1 r 



400 



500 



m/i 



Fig 2. Effect of arsenite on the spectrum of dihydrothioctyl dehydrogenase re- 

 duced with a-ketoglutaric dehydrogenase. Curve A, 2.0 mg flavoprotein, 0.02 mg 

 CoA, 3 /xmoles cysteine, and 0.2 mg a-ketoglutaric dehydrogenase in 0.71 ml phos- 

 phate buffer, /;H 7.3. Curve B, same plus 2 ^umoles a-ketoglutarate. Curve C, re- 

 duced flavoprotein plus 0.4 /^imole arsenite. 



It is possible, by analogy with the recent work on other flavoproteins 

 (2, 3) , that the 535 ni/x peak associated with the reduced enzyme is 

 due to a semiquinoid free radical represented by Intermediate II or 

 something very closely related to it. Addition of arsenite or Cd++ 

 would eliminate Intermediate II by shifting the equilibrium to Inter- 

 mediate III, in which the flavin is in the oxidized state. When 

 DPXH is present in excess, the FAD would be reduced again (spec- 

 trum C, Fig. 1). When dihydrothioctate (or a-ketoglutarate) is the 

 redtictant, the FAD would remain oxidized (spectrum C, Fig. 2) since 

 arsenite blocks the disulfide group essential for this reaction. 



A flavin semiquinone stabilized by an — SH group has been pro- 

 posed as an intermediate based on inhibition studies with /?-chloro- 

 mercuribenzoate and Cu++ (15, 7) . However, these results do not 

 necessarily implicate a vicinal disulfhydryl group. 



A model reaction between a disulfhydryl compound and riboflavin- 

 5'-phosphate (FMN) leading to formation of a product with an ab- 

 sorption maximum at 535 m^, is shown in Fig. 3. The absorption 

 spectrum of a mixture of 1.8 X 10-^ M dihydrothioctate and 15 X 



