280 THE ANTIGENIC STRUCTURE OF BACTERIA 



chemical fractionation, namely that the treatment used may partly but fundamentally 

 alter the substance we are looking for. The pneumococcal polysaccharide as originally 

 isolated was not antigenic. By less chemically active fractionation, an acetylated form 

 of the substance is produced, which is immunologically far more significant (Avery and 

 Goebel 1933). 



Lancefield (1928) separated antigenic components from hsemolytic streptococci by 

 chemical methods. She found that the component that confers type -specificity is acid- 

 soluble and contains some 14 per cent, of protein nitrogen. In its purified state it is 

 hapten-like, in that it gives specific precipitation in the test-tube but fails to stimulate 

 antibody-production in vivo. In addition to this type-specific antigen there is a poly- 

 saccharide component that is shared by many types of hsemolytic streptococci, but dif- 

 ferentiates the species into large sub-groups (Lancefield 1933) ; and there is a nucleo- 

 protein antigen that is shared by a wide variety of streptococci, hsemolytic and non- 

 hsemolytic. Studies by Todd and Lancefield (1928) (see also Lancefield and Todd 1928) 

 have shown that the change from the virulent matt to the avirulent glossy form is associated 

 with the loss of the type-specific component, the polysaccharide component being retained. 



The capsulated bacUlus of Friedlander has given results entirely analogous to those 

 obtained with the pneumococcus (Heidelberger, Goebel and Avery 1925, Julianelle 1926). 

 The species may be divided into a number of sharply demarcated serological types. This 

 type-specificity is conferred by a polysaccharide hapten present in the capsule. The 

 body of the bacUlus contains a nucleo-protein antigen that is shared by all types. 



Polysaccharide haptens have also been isolated from the tubercle bacillus (Laidlaw 

 and Dudley 1925, Enders 1929), from Bact. lactis aerogenes (Tomcsik and Kurotchkin 

 1928), from organisms of the Salmonella group (Furth and Landsteiner 1928, 1929), from 

 cholera vibrios (Landsteiner and Levine 1927, Jermoljewa and Bujanowskaja 1930), from 

 Shiga's dysentery bacillus (Meyer 1930, Morgan 1931), from the anthrax bacillus (Tomcsik 

 1930, Schockaert 1928, Tomcsik and Szongott 1932) and from yeasts (Mueller and Tomcsik 

 1924, Stone and Garrod 1931, Duncan 1932a). 



Recently the isolation of antigens more nearly in the native form has been facilitated 

 by gentler and in some cases more specific methods of fractionation. Among these we 

 may mention trichloracetic acid (Boivin and Mcsrobeanu 1933) and diethylene glycol 

 (Morgan 1937) for the extraction of the main antigenic complexes from salmonella and 

 dysentery bacilli (see Chapters 29 and 30) ; Raistrick and Topley's (1934) method for the 

 same type of substance by tryptic digestion of acetone -extracted bacteria ; and Fuller's 

 (1938) formamide method of extracting group-specific components from streptococci. 

 Sonic vibrations apparently liberate antigens from bacteria in a more native form. Mudd 

 and his colleagues (1937) succeeded by this means in isolating from Str. pyogenes a labile 

 antigenic substance built up of smaller, serologically specific fractions extracted by Lance- 

 field. By similar methods, Sevag and his colleagues (1941 ) separated from the disintegrated 

 streptococcal cells large numbers of antigenic " macromolecular " particles containing 

 lipins, nucleic acid, protein, carbohydrate and a pigment. 



Julianelle and Wieghard (1934), and Thompson and Khorazo (1937), isolated carbo- 

 hydrate substances responsible for the type-specificity of Staph, aureus. By milder 

 procedures Verwey (1940) isolated an additional, type-specific protein antigen. According 

 to Menzel and Rake (1942) the type-specific substance of the Type II meningococcus 

 owes its serological activity to a carbohydrate-polypeptide complex differing from the 

 type-specific polysaccharide and type-specific proteins described for other species of cocci. 

 (For the characterization of protein antigens in the gonococcus, see Stokinger et at. 1944 ; 

 and of various antigenic fractions in C. diphtheria;, see Wong and T'ung 1939, 1940, Wong 

 1940, and Hoyle 1942.) 



The case of the anthrax bacUlus presents points of interest. The normal virulent form 

 of this bacillus is capsulated and gives a rough colony on agar ; the avirulent variant is non- 

 capsulated and gives a smooth colony on agar (Preisz 1904, Eisenberg 1912). This, then, is 

 one of the cases in which the normal relation between smoothness and virulence is reversed. 



