246 THE ANTIGEN-ANTIBODY REACTIONS 



The serum globulins as a whole behave in the ultracentrifuge as though they 

 were homogeneous, but they are separable, as we have already seen, into fractions 

 of differing solubilities. Antibodies are distributed variously among the fractions 

 (see, for example, Adair and Taylor 1936). The distribution depends on the 

 animal immunized, the nature of the antigen, and the mode of fractionation. 



It is still questionable whether the euglobuUns and pseudoglobulins are artefacts pro- 

 duced during fractionation of serum. Rabbit antisera to horse and human serum 

 proteins react to some extent separately with euglobuhns and pseudoglobulins, suggesting 

 that the two exist as such in the native serum (Harris and Eagle 1935). 



Marrack and Duff (1938) were unable to demonstrate an absolute serological distinction 

 between euglobulin and pseudoglobulin, which neither singly nor in mixture had the full 

 serological reactivity of native globulin. Treffers and Heidelberger (1941a) also found 

 only a partial serological distinction between the water-soluble and water-insoluble 

 globulins. 



The increase in serum globulin during immunization has been noted by several observers 

 (see Marrack 1938), though not all of it is attributable to the formation of antibodies 

 (Liu, Chow and Lee 1937, Boyd and Bernard 1937). In Boyd and Bernard's experiments 

 with a variety of antigens, the antibody never accounted for more than 35 per cent, of 

 the increase, though, as the authors point out, some of the serologically inert proteins 

 may have been antibodies of a reactivity too weak to be detected by their methods. 



Ando and his colleagues (Ando, Kee and Komiyama 1937, Ando, Kee and Manako 

 1937, Ando, Manako, Kee and Takeda 1937, Ando, Manako and Takeda 1938, Ando, 

 Takeda and Hamano 1938), by immunizing rabbits with antigen-antibody aggregates, 

 distinguished two types of globuhn, A and B in horse sera, A predominating in the water- 

 soluble, B in the water-insoluble fractions. Both normal sera and antisera contained 

 A and B globuUns. Immunization with diphtheria toxin, or Shiga's dysentery bacilli, 

 increased the A globulins ; B globulins were increased during immunization with pneumo- 

 cocci, Salm. typhi and other bacterial antigens. 



The globulin components revealed by Tiselius's electrophoretic apparatus undoubtedly 

 exist as such in native serum. Protein particles of similar molecular weight may prove 

 to be heterogeneous by electrophoresis, and Tiselius (19376) showed that antibody prepara- 

 tions of Heidelberger and Pedersen (1937), which resembled normal globuhns in the ultra- 

 centrifuge, differed strikingly from the normal in their rate of migration in an electrical 

 field. By comparing antibody to egg albumin and to pneumococci, before and after 

 absorption with the homologous antigens, Tiselius and Kabat (1939) demonstrated that, 

 in the rabbit and the monkey, antibody was electrophoreticaUy identical with the y com- 

 ponent of normal globulin ; horse antibody migrated as a new component, midway in 

 mobility between the ^ and the y components. In the hands of Moore, van der Scheer 

 and Wyckoff (1940) and van der Scheer, Lagsdin and WyckofF (1941) the high molecular 

 weight antibody in horse antipneumococcal sera migrated with the normal y globulin. 

 Van der Scheer, Wyckoff and Clark (1940, 19416) described in some antisera an increase 

 in y globuhn, in others a new component T, both associated with an increase in antibody. 

 The T component was midway in mobihty between the /5 and y components (see also Rothen 

 1941). It appears more readily in antitoxic, as distinct from antibacterial sera, though 

 not regularly so. Kekwick and Record (1941) found two antibody fractions in diphtheria 

 antitoxin from the horse, one associated with the /?, the other with the y component. 

 With optimal amounts of toxin, the ^ fraction flocculated slowly and formed a compound 

 with the connDosition (TAg),^. The y fraction flocculated 20 times as rapidly, and formed 

 compounds of the type (TA4),j. The y fraction, which is precipitated with the euglobulin 

 during salt fractionation, predominated in early bleedings from immunized horses ; in 

 the later bleedings, the ^ fraction constituted the main antitoxin content (Kekwick, 

 Knight, MacFarlane and Record 1941). 



We may note one more variety of antibody, recently described by Race (1944). The 



