726 ANTIGENIC PROPERTIES OF THE BACTERIAL CELL 



and chemical investigation was carried out on larger quantities, it was found that 

 practically all traces of protein recognizable by available methods could be eliminated 

 and that the material bearing the specific properties consisted largely, or entirely, of 

 polysaccharide. When relatively pure, these carbohydrates in many cases will give 

 specific precipitation and complement fixation reactions in dilution of the dried sub- 

 stance in from 1-1,000,000 to 1-6,000,000, but fail entirely to produce antibody for- 

 mation in rabbits, even when immunization is continued for months with large quan- 

 tities. 



The organisms in which the soluble specific or residue substance have been partic- 

 ularly studied are the pneumococcus, the tubercle bacillus, the Friedlander bacillus, 

 the streptococcus, and yeast cells, and they have been determined and to some ex- 

 tent studied immunologically in the staphylococci, typhoid and influenza bacilli, and 

 meningococcus. 



In the case of the pneumococci, Heidelberger and Avery' in 1923 described care- 

 fully the substance obtained by them from a type II pneumococcus. 



They used eight-day broth cultures in large quantities and separated the specific ma- 

 terial by a method of fractional precipitation with alcohol, acetone, ammonium sulphate, 

 and dialysis. The yield was about i gm. from 75 liters of broth culture. It proved to be main- 

 ly a polysaccharide gum, yielding 79 per cent reducing sugar on hydrolysis, at least a part 

 of which was identified as glucose. Specific precipitates were given at a dilution of 1-3,000,- 

 000, and while no claims as to the purity of the material were made, since it stiU contained 

 1.2 per cent nitrogen and a trace of phosphorus, it was shown that the activity of the prepara- 

 tion increased progressively with the removal of nitrogenous materials. More recently, com- 

 pletely nitrogen-free preparations have been made. 



Since that time types I and III pneumococci have also been investigated.^ Substances 

 of widely differing chemical properties but also carbohydrate in nature have been obtained. 

 In the case of type I the gum appears to be made up of monosaccharide units which contain 

 nitrogen as a part of the molecule. The gum itself contains 5 per cent N, half of which is 

 liberated by HNO2, which causes a loss of specific properties. An examination of the products 

 of hydrolysis indicates that it may be composed of an amino-sugar derivative linked to 

 galacturonic acid. It behaves as a strong acid and a weak base, with an iso-electric point at 

 about pH 4, at which reaction it is fairly insoluble. The type III compound is produced in 

 larger quantities than the others, and is relatively easy to prepare, since it appears to be a 

 soluble salt of an insoluble acid. It may be precipitated simply by acidification of aqueous 

 solutions of its salts and is readily purified. It is hydrolyzed almost quantitatively to a 

 disaccharide acid which in turn splits with some difficulty into glucose and perhaps glucuronic 

 acid. All three of these specific substances possess acid properties to a greater or lesser 

 degree. Their optical properties vary widely as might be expected. For a more detailed 

 consideration of these substances see Heidelberger.^ 



Yeast has been shown by MueUer and Tomcziki to produce a carboh}'drate with specific 

 properties similar in all respects to those shown by bacterial gums. The abundant supply 

 of this material in the form of bread yeast offered a possibility of finding a method of purifica- 



' Heidelberger, M., and Avery, O. T.: ibiJ., 38, 73. 1923. 



^ Heidelberger, M., and Avery, O. T.: ibid., 40, 301. 1924. 



3 Heidelberger, M.: Physiol. Rev., 7, 107. 1927; Chem. Rev., 3, 403. 1927. 



■> Mueller, J. H., and Tomcsik, J.: /. Exper. Med., 40, 343. 1924. 



