PEROXIDASE OF LEUCOCYTES 431 



the material which is richest in peroxidase. It is now evident that 

 the "indophenol oxidase" activity of leucocytes is not due to cyto- 

 chrome oxidase but to myeloperoxidase, hydrogen peroxide being 

 produced by autoxidation of the Nadi reagent. The literature is 

 fully reviewed in Agner's paper (26). 



Agner isolated the enzyme from the leucocytes of empyema and from 

 the blood of patients with myeloid leukemia. Because of its green color, he 

 first used the term "verdoperoxidase," and later the term "myeloperoxidase," 

 because of its origin. The preparation involves precipitation in a layer 

 between ether and ammonium sulfate solution, removal of impurities by 

 adsorption to barium sulfate, fractional alcohol and ammonium sulfate pre- 

 cipitations (the enzyme being precipitated at two-thirds saturation of the 

 latter), dialysis, and electrophoresis. At pVL 6.8 the enzyme, which has an 

 isoelectric point above pK 10, migrates to the cathode, while a red impurity 

 moves to the anode. If sufficiently pure, the enzyme is soluble in salt-free 

 solution. Agner calculates that the leucocytes contain 1 to 2% of the enzyme. 



Myeloperoxidase contains about 0.1% iron. It is resistant to alcohol- 

 formalin treatment. In the cells it is partly bound by acidic proteins. 



The absorption spectra of the enzyme and its compounds (cf. 

 Table VIII) show clearly that its prosthetic group is not a porphyrin 



TABLE VIII 

 Absorption Spectra of Myeloperoxidase and of Its Compounds" 



Absorption spectra 

 Compound Position of bands, van, and tmiyi (in parentheses) 



Fe3+ 690 (3.5), 625 (6.5), 570 (11.0), 495 (weak), 430 (70) 



Fe^^ 637 (17.3), 590 (11.0), 475 (65-80) 



Fe'+CN 634, 438 



Fe3+0H 628, 460 



Fe3+N, = Fe3+ 



Fe'+Nj (?) 615, 460, slow reaction 



Fe^+HjOj 625, unstable 



" According to Agner (£6) . 



hematin; the prosthetic group is firmly attached to the protein.* This, 

 together with the absorption spectra make it appear likely that this 

 peroxidase is related to choleglobin (c/. Chapter X). The visible 

 absorption spectra somewhat resemble those of choleglobin and also 

 those of biladienone hemochromes (c/. Chapter IV, Section 5.4.2.), 

 but the data are not sufficient to prove a close chemical relationship 

 to either. The band at 475 ran may be considered as a rather low 

 *Recent investigations {1699) have shown that, like choleglobin, verdoperoxidase 

 yields protohemochrome on heating in alkali in the presence of dithioriite. 



