TOXIN FORMATION BY DYSENTERY BACILLI 693 



Inman (1919) found that Schmitz's bacillus, Sh. alkalescens, and Sh. dispar were 

 agglutinated, whereas Sh. shigce and Sh. flexneri were not. The test is only a 

 rough one, and it is doubtful whether the information it furnishes justifies its use 

 under ordinary conditions. On the other hand, the susceptibility of different types 

 of dysentery bacilli to the action of certain phages appears to be very much more 

 specific. According to Burnet and McKie (1930), whose article should be consulted 

 for further details, bacteriophages that are active against Flexner bacilli can be 

 divided into four main groups. One of these groups is capable of lysing bacilli only 

 in the smooth phase, while the other three may or may not lyse smooth strains but 

 generally lyse rough strains of all types. Characteristic differences in their sensiti- 

 vity to a series of phages are presented by the V, W, X, Y, and Z types of Sh. 

 flexneri. 



Chemical Fractionation. — Early observations on the chemical structure of the antigens 

 of the dj'sentery bacillus revealed the presence of specific polysaccharides in Sh. shigce 

 (Kurauchi 1929, Meyer 1930, 1931, Morgan 1931), Sh. flexneri, and Sh. sonnei (Kurauchi 

 1929). The more recent work of Boivin and Mesrobeanu (1937a-A, 1938), Mesrobeanu 

 and Boivin (1937), and Haas (1937, 1938rt), carried out by use of the trichloracetic acid 

 technique, shows that the smooth, but not the rough, somatic antigen of the Shiga and 

 Flexner baciUi contains a polysaccharide hapten, which is linked to nitrogenous and Upoid 

 compounds to form a complete antigen. Similar polysaccharides have been demonstrated 

 in Sh. schmiizi (Haas 1938c), Sh. sonnei (Haas 19386), and Sh. alkalescens (see Weil 1943). 

 The still more recent studies of Morgan (1936, 1937) and Morgan and Partridge (1940, 

 1941), who used the less drastic method of extraction of acetone-treated bacterial cells 

 with diethylene-glycol, seem to show that the complete smooth somatic antigen of Shiga's 

 bacillus consists of a specific polysaccharide hapten, a non-antigenic conjugated protein, 

 and a non-antigenic phosphohpm of the cephaUn type. The polysaccharide is strongly 

 dextrorotatory and yields 97 per cent, of reducing sugars on acid hydrolysis. When 

 combined with the conjugated protein, it forms a powerful antigen. It is interesting 

 to note that the conjugated protein extracted from Shiga's bacillus appears to be identical 

 with that found in the somatic antigen of Salmonella typhi. The polysaccharide not 

 only forms a precipitate in the presence of a specific antiseriun, but neutralizes specifically 

 the haemol3/tic action of shigce heterophile antibody (see Chapter 8) on sheep red corpuscles 

 in the presence of complement (Meyer and Morgan 1935). 



Toxin Formation by Dysentery Bacilli. — A large amount of work has been 

 carried out on the formation of toxin by members of the dysentery group, par- 

 ticularly by Sh. shigce. 



In 1903 Conradi prepared an autolysate of dysentery bacilli — probably Shiga's bacillus 

 — which he found to be toxic for rabbits and guinea-pigs. An 18-hours' culture was 

 suspended m saUne, and incubated for 24 to 48 hours at 37-5° C. ; after centrifugalization 

 the yellowish supernatant fluid was removed, diluted with 5 times its volume of saline, 

 and filtered through a Berkefeld candle ; the filtrate was tested for sterihty, and then 

 concentrated to 1/10-1/50 of its bulk at 35° C. This product, when mjected intra- 

 venously into rabbits or intraperitoneally into gumea-pigs in a dose of O-I ml., proved 

 fatal in about 48 hours. In rabbits death was preceded by diarrhoea, collapse, and 

 paralysis of the legs ; in guinea-pigs by a rapid faU of temperatm-e and collapse. At 

 necropsy Conradi found in both animals congestion of the intestine, mucus and blood 

 adhering to the mucosa, and frequently small haemorrhages of the mucous and the serous 

 coats. When a smaller dose was injected into rabbits, the animals hved for 4 to 6 days ; 

 and he found that post mortem the mucosa of the last third of the large intestine was 

 swollen, blackish-red in colour, and ulcerated in several places. 



Neisser and Shiga (1903) confirmed Conradi's results, and noted in addition that the 



