Composition of Cytochrome c Oxidase 285 



be the prosthetic group of either cytochrome a or a^. In their last publication 

 Morrison, Connelly and Stotz (1958) do not record the relative amounts of 

 the two haemins, a^ and a^, which they separated. In the first experiments 

 where there was cross contamination the amounts were about equal (Morrison 

 and Stotz, 1955). 



The uncertainty regarding the number of haems a has been further compli- 

 cated by the discovery that porphyrin a may exist in two interconvertible 

 forms (Lemberg and Stewart, 1955). The porphyrin which is isolated and 

 purified is identified as «a. On treatment with aqueous hydrochloric acid 

 acf. is partly converted into porphyrin a(i which has the same absorption 

 spectrum, but which differs in its solvent distribution and chromatographic 

 behaviour. 



B. The Copper 



The view that copper might be a part of cytochrome c oxidase seems to be 

 contained in the accumulated studies of Warburg (1949) and his associates, 

 who concluded, however, that because of the hght-sensitivity of its carbon 

 monoxide compound, the enzyme could only contain iron. The earliest, as 

 well as the latest nutritional studies have supported an opposite view (Cohen 

 and Elvehjem, 1934; Gubler, Cartwright and Wintrobe, 1957). 



Keilin and Hartree (1938a, 1939) also briefly considered that cytochrome c 

 oxidase might be a copper enzyme. Hov/ever, when cytochrome a^, an 

 apparent haem-enzyme, satisfied all of the criteria of cytochrome oxidase, 

 they discarded the possibility. 



Our interest in the copper content of cytochrome c oxidase arose as a 

 consequence of the feeling that, since the response of the enzyme to carbon 

 monoxide was unusual, there must be an unusual component. We prepared 

 fractions from heart muscle mitochondrial fragments with deoxycholate and 

 found a good correlation between the copper content and the height of the 

 601 mil absorption of the reduced enzyme (Eichel, Wainio, Person and 

 Cooperstein, 1950). Even though in these experiments the correlation between 

 the copper content and the activity was not good (see also : Okunuki, Sekuzu, 

 Yonetani and Takemori, 1958), we concluded that the enzyme was a copper- 

 protein. An analysis of the metal content of the pyridine haemochrome of the 

 prosthetic group prompted us to return to the classical view (Person, Wainio 

 and Eichel, 1953). The view has persisted, however, that cytochrome c 

 oxidase must also be a copper-enzyme. 



Our position is supported by the recent results of Green, Basford and 

 Mackler (1956) and Mackler and Penn (1957), who find that only those mito- 

 chondrial fragments with a pronounced spectrum for cytochrome c oxidase 

 have considerable amounts of copper. It is also interesting to note that in 

 Mason's (1957) classification of four-electron transfer oxidases, of which 

 cytochrome c oxidase is one, the other three enzymes, laccase, the catechol 



