HORSE-RADISH PEROXIDASE 425 



indicate that peroxidase contains carbohydrate. This carbohydrate 

 is probably present in the form of a uronic acid. If 18.4% is sub- 

 tracted from the molecular weight, a molecular weight of 35,900 is 

 obtained, which corresponds to about 288 moles of amino acids; a 

 figure of 287 is obtained by subtracting 5 humin nitrogen, 54 amide 

 nitrogen, 4 histidine nitrogen, 54 arginine nitrogen, and 12 lysine 

 nitrogen atoms from 415, the total number of nitrogen atoms. Only 

 14 nonamidized aminodicarboxylic acids + 2 hematin carboxylic 

 acids are present, as against 30 basic groups of arginine and lysine, 

 while the isoelectric point is 7.2. Therefore, acidic groups in the 

 carbohydrate fraction must neutralize some of the basic groups. On 

 transformation of peroxidase into paraperoxidase the greater part of 

 the carbohydrate is removed, with a resulting shift of the isoelectric 

 point to the alkaline side. At the same time an increase of nitrogen 

 content occurs, while the distribution of the nitrogen in the various 

 fractions remains unaltered. 



3.2.6. Linkage of Prosthetic Group and Protein. Theorell 

 {'£772) found that peroxidase can be split reversibly into protein and 

 hemin with acetone-hydrochloric acid at — 15° C. After neutrali- 

 zation, the protein recombines slowly with added protohematin, 

 giving active enzyme. Under certain conditions the activity can be 

 fully restored {2788). This was confirmed by G jessing and Sumner 

 {1008). They found that mesohematin and deuterohematin also 

 combine with the peroxidase protein to form active catalysts, while 

 porphyrin iron complexes which do not have the two propionic acid 

 side chains, such as rhodo- and pyrrohematin, combined, but yielded 

 inactive compounds; hematohematin did not combine. Their claims 

 that the compound with mesohematin was an even more active 

 peroxidase than that with protohematin, and that manganese por- 

 phyrins also gave an active peroxidase, have not been confirmed by 

 Theorell {2779). 



The problem of the mode of linkage of hematin to the protein has 

 been studied by Theorell {2776) and by Theorell and Paul {2789). 

 By differential titration of the enzyme and apoenzyme (hematin-free 

 protein), by study of the dissociation of fluoride peroxidase, and by 

 spectrophotometry, they arrive at the conclusion that the hematin 

 is bound to a carboxylic group of the protein, and that at the same 

 time one of the carboxylic acid groups of the hematin is bound to a 

 basic group of the protein of pK value 10.2. The latter is probably a 



