390 Discussion 



There is, however, no particular reason why similar bonds could not occur in neutral 

 or acid cytochromes c, since all that is required is a concentration of hydrophobic 

 amino-acid side chains around the region where such an electrostatic bond may be 

 formed. The effect of such an 'umbrella' would be to change the dielectric constant at 

 the point where the electrostatic bond is to be formed, so as to render it more stable 

 than it would be in a normal water environment. Without the specific hydrophobic 

 microenvironment an electrostatic bond would be quite unstable even at neutral pH 

 in an aqueous medium. 



Structure and Redox Potentials of Cytochrome c 



The effect of denaturation on E^ values for cytochrome c 



Margoliash: To obtain what we have called the trichloroacetic acid modified or denatured 

 cytochrome c it is necessary to have the cytochrome c in solution during the denatura- 

 tion procedure, and not precipitated. This is done by carefully controlling the pH. 

 Under such conditions, or by using ethanol as a denaturing agent, one obtains a 

 product that can be separated into fractions of increasing degree of denaturation by 

 column chromatography on Amberlite IRC-50. These fractions have the full com- 

 plement of amino acids present in the original native protein. The more the degree of 

 denaturation, the lower the E^ value. The actual values obtained range from that of 

 native cytochrome c at -fO-255 V to not far from 0-0 V. 



The ligands to iron in cytochromes of the c-type 



Perrin: The difference of 0-54 V in Eq between cytochrome c and cytochrome c peptide 

 indicates that in the latter the ferric form is stabilized by a factor of 10^ relative to the 

 ferrous form, compared with cytochrome c. This represents a preferential stabiliza- 

 tion of the ferric cytochrome c peptide by a AG of about 13 kcal. This difference is 

 much greater than is found in the iron porphyrin haemochromes where the ligands 

 are neutral molecules and I should like to suggest that in the peptide the sixth link 

 to the metal is through a carboxyl group (e.g. of glutamic acid) or some other anion 

 because anions form much more stable complexes with ferric than with ferrous iron. 

 The very large difference between cytochromes c^ and C3 would then indicate that in 

 the former the sixth position ligand is neutral (e.g. amino-N or imidazole-N) while in 

 the latter it is carboxyl. Falk and I (p. 66) have indicated why we think that the 

 absorption spectra of such complexes are insensitive to the particular ligands other 

 than the porphyrin groups. 



Margoliash: I wish to reply to Perrin's suggestion that the lowering of the Eq value of 

 cytochrome c on denaturation may be due to an exchange of a — COO~ group for 

 one of the normal haemochrome-forming ligands in the native protein, during 

 denaturation as well as during proteolytic digestion to form the 'core'. 



I think that the spectrophotometric evidence rules out such a possibility. Whether 

 a ligand is acidic or basic, thespectrophotometrically-observed pkfor the dissociation 

 of a particular ligand would be at a pH below that of the normal pAT value for the 

 ligand in solution, since as pointed out previously by Perrin, one must take into con- 

 sideration the constants involved in the reaction of the ligand with the haem iron. In 

 the ferro- form of the cytochrome c 'core', the lowest apparent pA" value we have 

 observed spectrophotometrically, using the visible region of the spectrum, was at pH 

 values between 5 and 6. This clearly rules out the participation of a carboxyl as a 

 liaemochrome-forming group in the reduced form of the 'core'. 



With reduced denatured cytochrome c the results will naturally vary with the degree 

 of denaturation of the fraction studied, all the way from an apparent pA: at pA:2-3 

 for the native protein, to a behaviour essentially identical to the 'core' for the most 

 denatured material, again showing the lowest pA" at pH values of 5-6. Since as dis- 

 cussed previously, the ZTq' value will decrease with an increasing degree of denaturation 

 and one would expect that the introduction of a carboxyl group would result in a 

 decrease in Eq, Perrin's suggestion cannot be reconciled with the spectrophotometric 



