230 THE ANTIGEN-ANTIBODY REACTIONS 



mid-piece unites directly with sensitized cells whereas end-piece unites with such 

 cells only in the presence of mid-piece. In various complement-fixation reactions, 

 mid-piece is mainly absorbed, the greater part of the end-piece being left in solution. 

 Both mid-piece and end-jjiece are inactivated by heat. 



It has long been known that a serum loses its complementary power after 

 absorption by yeast. Such a serum is rendered active again by adding serum 

 in which mid-piece and end-piece have been destroyed by heating to 56° C. (Coca 

 1914). This third heat-stable component is present in both mid-piece and end-piece, 

 as prepared by Liefmann's method of mid-piece precipitation by COg (Whitehead, 

 Gordon and Wormall 1925). 



Gordon, Whitehead and Wormall (1926a, b) demonstrated a fourth component 

 of complement which is heat- stable, but not absorbed by yeast. It is specifically 

 inactivated by treating the serum with ammonia. Though this component is 

 necessary for the hsemolytic action of complement, it is not necessary for the 

 complementary effect of normal serum in the opsonic reaction (see p. 236). The 

 existence of the fourth component was confirmed by Deissler (1932) and many 

 subsequent workers. 



The demonstration of components by inactivation with specific chemicals must be 

 viewed with caution, for activity is associated with proteins, and agents which denatm'e 

 the proteins will also destroy complement. The postulation of a benzene-inactivated 

 component by Tokano (1936), a dialysable component by Chow and Zia (1938) and of the 

 necessity for ionized calcium (Ottolenghi and Mori 1905) were probably based on observa- 

 tions of denaturation (see Jones and Ecker 1940, Pillemer and Ecker 1941a, Ecker and 

 PiUemer 1942). 



Complement is most active at a pH of about 7-3 in the presence of physiological saline. 

 Its activity is progressively reduced by increases and decreases of pH and salt concen- 

 tration. Its inactivation by various salts was studied by Gordon and Thompson (1933a, h), 

 who found that the activity of various ions fell into the general order of the Hofmeister 

 series (SCN' > I' > Br' = NO3' > S04"/2 > CI'). Some of the salts produced a reversible, 

 others an irreversible inactivation. It loses activity in a few days at 0° C. and in a few 

 minutes at 56° C. According to Wehmeyer (1941) the inactivation of complement at 

 any given temperature between 4° C. and 55° C. proceeds like one unimolecular reaction 

 between 0° C. and 37° C, and another between 50° G and 56° G. At the lower temperatures 

 only mid-piece is labile, at the higher, both mid-piece and end-piece are labile, end-piece 

 predominantly so ; in the 30°-50° C. range, the rate depends on the velocities of two 

 unimolecular reactions. 



Dry complement that has been kept for months is still active when reconstituted 

 (Craigie 1931). Filtration through a porcelain candle (Strong and Culbertson 1934) 

 inactivates complement mainly by retention of mid-piece and end-piece. 



Many attempts have been made to isolate the components of complement 

 (see Ecker and Pillemer 1942). The association of mid-piece activity and end- 

 piece activity with " globulin " and " albumin " fractions of serum is to a large 

 extent illusory, for, as Pillemer and Ecker (19416) have shown, the electrophoretic 

 patterns of both fractions show four distinct proteins, two of those in " globuUn " 

 fraction differing from any observed when normal guinea-pig serum is analysed 

 by this method. Inactivation of the third and fourth components has little effect 

 on these patterns These authors propose the symbols C'l, C'2, C'3 and C'4 for 

 the active principles in mid-piece, end-piece, third and fourth component respec- 

 tively. By carefully controlled methods, Pillemer, Ecker, Oncley and Cohn 

 (1941) have isolated three fractions with high C'l, C'2 and C'4 activity from 



