THE ANTIGENIC STRUCTURE OF THE HEMOLYTIC STREPTOCOCCI 677 



made it possible to differentiate and identify many different antigenic types among 

 hsemolytic streptococci isolated from human infections ; and the more detailed 

 and extensive studies of Griffith (1926, 1927, 1928, 1934, 1935), Smith (1926, 1927), 

 James (1926), McLachlan and Mackie (1928), Gunn and Griffith (1928), and Dora 

 Colebrook (1935) have served to establish many of these types on a satisfactory 

 serological basis. Griffith, who has been the main contributor to this particular 

 problem, has differentiated by means of the slide agglutination reaction 27 types 

 of pathogenic hsemolytic streptococci isolated from various lesions in man. Of 

 these, 23 were later found to be members of Group A ; Types 7, 20 and 21 fell 

 into Group C, and Type 16 into Group G. Before he died, Griffith added three 

 more types to Group A, bringing the total numbered types up to thirty. Further 

 types have since been added. 



Lancefield (1928, 1933) attacked the problem from another angle. She extracted 

 type-specific antigens from the cocci by hot acid, and identified them by the pre- 

 cipitin test. On the whole, her results agreed with those of Griffith, but some 

 discrepancies occurred. Further observations (Lancefield 1940, 1943, Swift et al. 



1943, Lancefield and Stewart 1944, Watson and Lancefield 1944, Stewart et al. 



1944, Zittle 1942, Elliott 1943, Krumwiede 1943) have shown that two separate 

 antigens are concerned in the type-specificity of Group A strains. One, referred 

 to as M, is a protein : the other, referred to as T, is of undetermined nature. The 

 M antigen is present only in mucoid or matt colonies (see p. 595) and is most 

 abundant in freshly isolated pathogenic strains ; the T antigen is found not only 

 in mucoid and matt, but also in the degraded glossy colonies containing avirulent 

 organisms. Antibody to M seems to be responsible for the M precipitin reaction, 

 for type-specific protection, and as a rule for part of the type-specific agglutination 

 of matt variants. Antibody to T appears to be solely responsible for type-specific 

 agglutination of glossy variants, and mainly responsible for type-specific agglutina- 

 tion of matt variants, but to be unconcerned in protection. The M antigen can 

 be destroyed by peptic or tryptic digestion ; heat-killed enzyme-treated organisms 

 stimulate the production of T antibodies alone, so that pure T antibody can be 

 obtained and the distribution of T antigen determined. In most types M and 

 T antigens occur together, but in a few types one or other is missing or is shared 

 with some other type — hence the occasional discrepancies between the results of 

 agglutination and precipitation tests. 



For example, Types 10 and 12 contain serologically identical M antigens, but different 

 T antigens. Closely related T antigens occur in Types 15, 17, 19, 23, and 30, and another 

 set of related T antigens occurs in Tjrpes 4, 24, 26, 28, 29, and 46. In some types, strains 

 are encountered that contain no T antigen at all. Moreover, the amount of M and T 

 antigen that is formed, or at least that can be detected by agglutination methods, depends 

 on the temperature of incubation. There is evidence to suggest that the M substance 

 is formed best at 37° C, but is Uable to inactivation at this temperatiu-e — possibly through 

 enzyme action (EUiott 1944). 



An additional complication to the precipitin reaction is that extracts may 

 contain variable amounts of non-specific nucleo-proteins, which are common to 

 hsemolytic streptococci, pneumococci, and streptococci of the viridans group. It 

 may also be noted that there is at least one other non-type-specific protein (" Y ") 

 about which little is as yet known. It will therefore be clear that the type deter- 

 mination of Group A streptococci is far from straightforward ; nevertheless their 

 general antigenic constitution is now fairly well understood. 



P.B. V 



