ANTIGENIC STRUCTURE 823 



The early work of Sergent, Gillot, and Lemaire (1908), and Negre and Raynaud 

 (1912a, h), demonstrated the existence of strains morphologically and culturally 

 resembling Br. melitensis but failing to agglutinate to more than a fraction of the 

 titre with an anti-melitensis serum. These irregular strains were given the name of 

 faramelitensis. Later on, so-called para-abortus strains were encountered, and these 

 were believed to represent merely a special antigenic type of Br. abortus. Further 

 study, however, by such workers as Fa villi (1926a, b), Ross (1927a), Valenti (1927), 

 Vidal and Abella (1928), de Antoni (1929), Zdrodowski et al. (1930), Pampana (1931) 

 and Pandit and Wilson (1932), showed that paramelitensis and para-abortus strains 

 were agglutinable by non-specific agents, particularly acid, salt, peptone, and certain 

 aniline dyes, whereas freshly isolated strains of Br. melitensis and Br. abortus were 

 unaffected by these agents under similar conditions. Moreover, it was found that 

 continued cultivation in broth, or better still in broth containing immune serum, 

 led to a transition of melitensis and abortus strains into paramelitensis and para- 

 aborius respectively. The transition, it may be noted, occurs much more readily 

 with melitensis than with abortus strains, and evidence of its commencement is often 

 noticeable within a very short time of isolation, even when the organisms are kept 

 on solid medial There seems to be little doubt that this change, which may be 

 accompanied by alterations in colonial appearance and by a decrease in virulence 

 for laboratory animals, is essentially a manifestation of the S — > R variation. 

 The antigenic change concerned is not yet clearly understood, but there is evidence 

 that it involves a loss of the specific smooth antigen. Strains of different degrees 

 of roughness are encoimtered, varying from those that agglutinate to titre with a 

 smooth serum but are slightly susceptible to non-specific agglutination to those 

 that are unaffected by a smooth serum and are incapable of remaining 

 homogeneously distributed even in cold saline. Moreover the degree of roughness 

 of a given strain appears to vary from one culture to another. Once roughness 

 has appeared, it persists, or recurs after an intervening period of apparent smooth- 

 ness. Roughness may be tested for by boiling in saline for 2 hours (thermo- 

 agglutination test), by incubation at 37° C. with 1/500 or 1/1,000 acriflavine 

 (Alessandrini and Sabatucci 1931, Pampana 1931), by agglutination with an 

 antiserum prepared against a completely rough strain of the corresponding type, 

 or by the ease of phagocytosis by normal leucocytes (Munger and Huddleson 1938). 

 Little work has so far been done on the antigenic structure of the rough types. 

 Our own incomplete observations suggest that, though there may be a common 

 antigen to paramelitensis, para-abortus, and parasuis types, there are certain 

 differences between them. One important practical point is that partly rough 

 strains are liable to be agglutinated non-specifically by the sera of normal persons, 

 and particularly of those suffering from certain febrile diseases (see Mohr 1935), 

 and are therefore liable to lead to an erroneous diagnosis of undulant fever in 

 routine serological work. 



Turning now to the differentiation of Br. melitensis, Br. abortus, and Br. suis 

 on the basis of antigenic structure, we are faced with a mass of conflicting reports 

 most of which may be summarized by saying, either that no difference was found 

 between the three organisms, or that they fell serologically into a number of different 

 types (see Feusier and Meyer 1920, Burnet 1925, Evans 1925a, b, Ross 19276, 

 Cerruti 1927, Kristensen and Holm 1929, Bieling 1930, Kristensen 1931, Francis 

 1931, Plastridge and McAlpine 1932). The reason for this confusion is probably 

 due to the failure of most of these workers to realize the disturbance caused by 



