THE MODIFICATION OF ANTIBODY BY PHYSICAL MEANS 247 



erythrocytes of certain human subjects contain an antigen D, which is one of the natural 

 Rh antigens (Chapter 49). An antibody to D occurs naturally in the sera of other human 

 subjects, and, hke other natural antibodies to the blood group antigens, it agglutinates 

 erythrocytes containing the corresponding antigen. In certain subjects, however, Race 

 found " incomplete " antibodies to D, which combined specifically with D-containing 

 erythrocytes, but failed to agglutinate them. The properties of this naturally occurring 

 "incomplete" antibody in many respects resemble the reacting but non-flocculating 

 albumin-antibody-globuUn complexes made artificially by Bawden and Kleczkowski 

 (1942) (see pp. 213, 244). 



The Modification of Antibody by Physical and Chemical Means. — Although 

 there is a general parallelism between gross alterations of the globulin nature of 

 antibodies, and their activity with respect to antigens, yet the protein can in 

 many cases be altered considerably without destroying its efficacy as an antibody. 



If, for instance, Type II pneumococcal antibody is " unfolded " in a mono-layer on 

 an air-water interface, it loses its capacity to combine with the specific carbohydrate of 

 the Type II pneumococcus (Danielli, Danielli and Marrack 1938). 



A similar unfolding of normal proteins appears to take place in urea solutions (see 

 Bernheim et al. 1942), possibly by the weakening of the hydrogen and other bonds respon- 

 sible for maintaining the association of adjacent polypeptide chains. In urea solutions at 

 pH 7-8 diphtheria antitoxin slowly loses its neutrahzing power for toxin, and more rapidly 

 at a higher or a lower pH (Wright 1944). These results suggest that the antibody activity 

 is closely associated with integrity of the general molecular structiu^e and that destruction 

 of the antibody configuration is associated with ionization of basic or acidic groups. 



Chemical modification by moderate treatment with formaldehyde, which reacts irre- 

 versibly with the protein at the surface of the molecule, has httle effect on antibody activity 

 (see Mudd and Joffe 1933, Ivanoff 1936, and Eagle 1938). As Eagle showed, stronger 

 treatment with formaldehyde, or with a reactive diazo compound, removed first the 

 precipitating and then the protective power of antipneumococcal sera and diphtheria 

 antitoxins, a fact we have already noted in discussion of the lattice hypothesis. Again, 

 the amino groups of an antibody globuUn molecule may be treated with ketene gas until 

 one-third of them are acetylated, without marked loss of the power to combine with 

 antigen (Goldie and Sandor 1937). The chemical groups on the surface of globuhn mole- 

 cules responsible for the antigenic power of the protein do not appear to form part of 

 the specific combining groups of the antibody. The destruction of activity of horse 

 antipneumococcal sera by photodynamic oxidation in the presence of methylene blue 

 only reduces the power of the proteins to precipitate with anti-horse serum (Ross 1938). 

 The independence in the antibody molecule of the antibody-combining group and the 

 groups that determine antigenic specificity has been demonstrated serologically by several 

 workers. Smith and Marrack (1930), and later Eagle (1936), demonstrated that diphtheria 

 antitoxin from the horse, which has combined with a rabbit anti-horse precipitin, would 

 react with toxin. Eagle in addition showed that antitoxin combined with toxin could 

 stiU react with the rabbit antibody. Treffers and Heidelberger (1941a, b) immunized 

 animals with specific precipitates formed by the combination of various antigens with 

 their homologous antibodies, both horse and rabbit. The reaction of the resulting anti- 

 antibodies was conditioned solely by the species of animal from which the antibody was 

 obtained ; any diflFerence between various antibodies that might be expected from the 

 fact that they each possessed different combining groups for their homologous antigens, 

 was not demonstrable by serological methods. 



The Enzymic Digestion of Antibodies. — Hydrolysis by enzymes ultimately destroys the 

 antibody activity of serum globulin (see Marrack 1938). Partial hydrolysis under con- 

 trolled conditions is selective in its action on the various serum proteins. For instance, 

 Pope (1939a, b), investigating the Parfentjev patent for the refinement of antitoxin, con- 



