TEE ANTIGENIC RELATIONSHIPS OF THE HEMOPHILIC BACILLI 799 



The Antigenic Relationships of the HsemophUic Bacilli. 



Taking first H. influenzce, as a species, the peculiarity that has emerged from 

 Jiiost of the recorded attempts at serological analysis is its extreme antigenic 

 heterogeneity. 



By direct agglutination with 20 antisera, followed by absorption tests where necessary, 

 Park, WilHams and Cooper (1918) could find only four identical pairs among 160 strains. 

 Valentine and Cooper (1919) record a similar experience. Among 10 strains isolated at 

 autopsy, tested against the 10 homologous antisera, no two were identical. Among 73 

 miscellaneous strains tested against 18 antisera no two were identical. Among 54 strains 

 isolated from a group of marines, and tested against 18 antisera, 2 strains from different 

 individuals were identical. It was noted in this group that strains isolated from the same 

 individual on different days were usually, but not always, identical. Among 28 strains 

 isolated from the inmates of an orphan asylum, there was one pair of identical strains. 

 Of 6 strains isolated from the members of a single family, all of whom had contracted 

 influenza at about the same time, no two strains were identical. 



This extreme heterogeneity, as judged by agglutination tests, lias been amply con- 

 firmed by numerous workers (Rivers and Kohn 1921, Yabe 1921, Anderson and Schultz 

 1921, Cooper et al. 1921, Povitsky and Denny 1921, Kristensen 1922, Knorr 1924, lizuka 

 1938, and others). lizuka, for example, records more than 50 different agglutinating 

 types among 249 strains isolated from sick and healthy persons. 



The actual significance of these earlier observations has been rendered very doubtful 

 by the observations of Pittman (1931). 



Among 97 strains of influenza bacilli isolated from various sources, she noted 15 that 

 produced colonies of a characteristic " smooth " type (see above). All these 15 strains 

 were isolated from sources, or under conditions, which indicated that they were playing 

 a pathogenic role. In addition to forming a characteristic colony, the bacilli of these 

 " smooth " strains were found to be capsulated. When tested by agglutination at 37° C. 

 these 15 strains were found to fall into two antigenic types A and B, one containing 12 

 strains, the other 3. This specificity was not apparent if the reactions were carried out 

 at a higher temperature, a possible reason bemg the loss of the bacterial capsules. It 

 was also found possible to separate from these 15 strains a soluble specific substance, 

 apparently carbohydrate in nature, and presumably associated with the capsule. Pre- 

 cipitin tests carried out with this material gave the same antigenic groupmg as the agglu- 

 tination tests carried out at 37° C. These smooth strains, in artificial culture, readily gave 

 rise to non- capsulated rough variants, usually with a bacillary or filamentous morphology. 

 The rough variants no longer produced the soluble specific substance, nor did they conform 

 to the antigenic grouping of the smooth parent strams. These findings were confirmed, 

 in their essential points at least, by several subsequent workers (see Dochez et al. 1932, 

 Wright and Ward 1932, Piatt 1937). Later work has shown that characteristic " smooth " 

 strains are commonly found among baciUi isolated from mfections of the meninges. Among 

 respiratory strams, either from infected or healthy persons, the " smooth " types are 

 less common. Mulder (1937) for example, foimd 7 in 90 sputum strains, and Piatt (1937) 

 16 in 86 nasopharyngeal strains. Respiratory strains tend to be serologically hetero- 

 geneous, whereas " smooth " meningeal strains tend to be homogeneous (Fothergill and 

 Chandler 1936, Wilkes-Weiss 1937, Piatt 1937). The homogeneity is not complete. 

 Pittman has divided these capsulated smooth strains into six serological types, a, b, c, d, e 

 and f. Type b occurs with the greatest frequency, but aU are rare in the normal upper 

 respiratory tract (see, e.g., Silverthorne et al. 1943). 



As in the pneumococci, type specificity depends on a polysaccharide component in 

 the capsule of the organism (Piatt 1937, Alexander and Heidelberger 1940). If we follow 

 Chandler, Fothergill and Dingle (1939) and designate capsulated forms as M or mucoid 

 forms, then both the S and R forms, which have no type-specific capsular substances, 

 display extreme antigenic heterogeneity. According to Piatt (1939) individual non- 



