274 E. BOERI, A. EHRENBERG, K. G. PAUL, H. THEORELL VOL. 12 (1953) 



the formation of four compounds, A, B, C, and D, in time sequence could be demon- 

 strated. 



As will be shown in this work, chloride ions form a compound with ferricytochrome 

 c in acid solution. Its properties have here been studied spectrophotometrically and 

 magnetometrically. A theory of the structure of the Cl-compound is advanced. It pre- 

 sents a sequence of dipoles resembling the arrangement in ionic crystals. 



METHODS 



Spectrophotometric measurements were made in a Beckman spectrophotometer model DU, in 

 most cases in i cm cells. In some preliminary experiments cells of varied depths were used in order to 

 check whether the absorption values at different hj'drogen ion and salt concentrations remained con- 

 stant at widely different concentrations of cytochrome c. It was found that a 300-fold variation of 

 this concentration did not influence the results. 



The pH-measurements were made with glass electrodes, calibrated against Pt-H.^ electrodes. 

 KCl-agar was used for connection in order to diminish the diffusion of chloride ions into the solution. 



All the experiments were carried out in rooms maintained at 2o°C. 



MATERIAL 



Four different specimens of cytochrome c were used. The first one, prepared according to 

 Theorell and Akeson*, contained 0.417% Fe, and the second and third ones, prepared according 

 to Keilin and Hartree^, 0.387 and 0.437%, respectively. Specimen 4, also prepared according to 

 Keilin and Hartree, gave only 0.317% Fe. During this preparation it was felt desirable to avoid 

 vacuum evaporation, hence all volumes were kept small to permit freeze-drying. This might explain 

 the incomplete removal of protein impurities. All preparations were examined for contaminating 

 hemoproteins by means of treatment with acid acetone in the cold. In preparation 3 about 5 % of 

 the total iron was found to be non-cytochrome heme-iron. The other three gave negative tests. 



The procedure of Tsou^'' was employed for the determination of the content of autoxidizable 

 cytochrome c. Preparation 3 contained 2 % of autoxidizable cytochrome; preparations 2 and 4 none. 

 Preparation i was not tested during the time it was used. Preparations i and 2 were used only for 

 spectrophotometric experiments, 3 and 4 for the magnetic. The content of auto.xidizable cytochrome 

 has been found to increase gradually when a preparation is stored at low temperature (0-4°). Specimen 

 I was found to contair. 50 % autoxidizable material five months later. It was therefore carefully 

 examined to determine whether this caused any changes in the light absorption in the Soret band 

 region at those acidities and chloride ion concentrations used in our experiments. Remarkably enough 

 it was found that the high auto.xidizability did not seem to have any influence upon the spectrophoto- 

 metric results. Moreover, it was noticed that exposure of an aliquot portion to low pH and high 

 chloride ion concentration for a few hours resulted in a reduction of the amount of autoxidizable 

 cytochrome c from 50 to 16%. This observation might be connected with the observation of P.\UL^ 

 that exposure of ferricytochrome c to chlorides at acid reaction {e.g. pH 1.63) caused a marked (30 %) 

 increase of its enzymic activity. For the removal of trichloroacetic acid from cytochrome c we generally 

 dialyze our preparations against dilute ammonia, while Keilin and Hartree prescribe dialysis 

 against 0.5 % KCl. Whereas we occasionally find a small percentage of autoxidizable cytochrome c in 

 our fresh preparations, Keilin and Hartree do not seem to find any: the difference may be attri- 

 buted to a protecting or restoring action of the chloride. 



RESULTS 



Experiments. It is seen from Fig. i that the ma.ximum of the Soret band, in agree- 

 ment with the earlier findings^ changes from 408 m/x in neutral solution ("type IIP") 

 to 395 m/A at pH 1.7 ("type 11"). If only these two forms were present we would e.xpect, 

 however, to see an isosbestic point in the region 400-405 mjn. This is not the case; the 

 variable intersection between the curves indicates that one or more other compounds 

 interfere with the results. 



It was found that the concentration of chloride ions had a decisive influence upon 

 the absorption spectrum. Fig. 2 illustrates how the addition of sodium chloride to cyto- 



References p. 282. 



