CHEMICAL FRACTIONATION 825 



Sievert (1936), Olin and Lindstrom (1934), Olin (1935), and Veazie and Meyer 

 (1936), using the method of quantitative absorption. 



As a means of typing unknown strains, direct agglutination with monospecific 

 serum is one of the simplest and most rapid methods. Like other tests, however, 

 for differentiation of this group, it cannot be relied upon entirely, because, as Wilson 

 (1933) has shown, certain strains, particularly from the South-East of France, may 

 possess the antigenic structure of abortus, while having the biochemical and patho- 

 genic characteristics of melitensis. 



Several workers claim to have established an antigenic affinity between members 

 of the Brucella group and those of the Pasteurella, Proteus, or Pfeifferella groups. 

 Most of these conclusions have been based on the observation that a given serum 

 agglutinated strains of both Brucella and some other group. By itself this is, of 

 course, quite insufficient evidence on which to base any such conclusion, since it 

 takes no account of non-specific agglutination or of the presence of specific 

 agglutinins to two different organisms co-existing in the same serum. Absorption 

 and other experiments have entirely failed to confirm the conclusions of these 

 workers (for references see Priestley 1933, Wilson 1934). 



Chemical Fractionation. — The reported results of chemical fractionation of 

 Brucella strains are to some extent conflicting, especially with regard to the fractions 

 exhibiting antigenic activity. The chemical characterization of many of the 

 protein, nucleo-protein and polysaccharide fractions isolated varies greatly, and 

 it is hard to judge the degree of purification, l)oth chemical and serological, that 

 has been achieved in each fraction. The more recent work has shown that a 

 precipitin titre, by a constant-antibody titration, of the order of 1 : 5 million 

 represents the limit of purification so far achieved in the main antigenic fractions 

 of this group. 



The fractions isolated by the earlier workers had precipitin titres of 1 : 2000 to 1 : 50,000, 

 and were clearly impure (Favilli and Biancalani 1932, 1934, Topping 1934, Gwatlcin 1935, 

 Reiter 1936, Schapira 1936, Higginbotham and Heathman 1936). These substances 

 appeared to be polysaccharide in nature, and when fractions from different species were 

 compared serologically they proved to be mainly group-specific, and displayed only 

 minor degrees of species specificity. Huddleson and his colleagues (Huston, Huddleson 

 and Hershey 1934, Hershey, Huddleson and PenneU 1935) isolated inactive proteins, 

 polysaccharides and lipoids from the three Brucella species, whose proportions varied 

 with the species studied. They also found a substance " S," free from protein and poly- 

 saccharide, precipitating with antisera at 1 : 2,000,000, which was readily extractable 

 from Br. melitensis, but was apparently bound to protein in Br. abortus and Br. suis. 

 The " S " substance appeared to be related to the specific soluble substances isolated 

 by Favilli and Biancalani (1932, 1934). The technique of treatment of the bacterial 

 cell with trichloracetic acid developed by Boivin for the extraction of " complete " antigens 

 from salmoneUse (see Chapter 8) was appUed to Br. melitensis by Lisbonne and Monnier 

 (1936) and to all three species of Brucella by Pop and his colleagues (Pop et al. 1938, 

 Damboviceanu et al. 1938, Vanghelovici et al. 1938). (See also Davoli 1938, Stahl and 

 Hamann 1941). The solutions obtained by dialysing the trichloracetic acid extracts 

 were antigenic, toxic and preciijitated in titres up to 1 : 100,000. The reported nitrogen 

 content varied from 0-4 to 8 per cent. ; lipins 10-17 per cent., and reducing sugars on 

 hydrolysis, 4—20 per cent. PenneU and Huddleson (1937), by digestion of acetone-dried 

 organisms, extracted toxic and antigenic " endo-antigens " with a precipitin titre of 

 1 : 5 mUhon which in many respects resembled the " complete " antigen obtained by the 

 Boivin technique. They found certain chemical differences between the endo-antigens 



