724 ANTIGENIC PROPERTIES OF THE BACTERIAL CELL 



ciples of manipulation are roughly the same. The method has consisted in extracting 

 with slightly alkaline salt solution, usually about N/ioo NaOH. Such extracts, pro- 

 duced in various ways, can then be centrifugated to remove all suspended particles, 

 or filtered through a Berkefeld filter, though the colloidal nature of the nucleoprotein 

 mixtures leads to a considerable quantitative loss from filtration. The extract is then 

 precipitated by the addition of acetic acid in the cold, care being taken to avoid an ex- 

 cess of acid. The flocculent precipitate, which is usually most profuse at a hydrogen- 

 ion concentration in the vicinity of pH 3-4, can then be separated by centrifugation. 

 The sediment may be redissolved in alkali and reprecipitated for purposes of purifica- 

 tion, but it is questionable whether one can count upon much chemical purification 

 by a repetition of the same method, and such re-solution and re-precipitation always 

 leads to considerable loss of material, the sediment never entirely redissolving. We 

 believe that this may be due to a gradual alteration of some of the materials by the 

 repeated and prolonged treatment with acid and alkali. This opinion is consistent 

 with Wells's' statement that the loss of substance is probably due to denaturation of 

 the albumins present in the original precipitate. 



It is quite impossible at the present time to make any specific statements about 

 the exact chemical constitution of the so-called "bacterial nucleoproteins" or to sepa- 

 rate the constituents into those which are antigenic and those which are not. Probably 

 there are many entirely non-antigenic admixtures which are immunologically neg- 

 ligible. The antigenic constituents of these nucleoproteins, however, act very much 

 like casein, in that very large amounts and persistent injections are necessary for the 

 production of any considerable concentration of antibodies in animals. We have been 

 forced to immunize for months with large amounts of nucleoprotein produced in the 

 foregoing manner from tubercle bacilli and other organisms in order to obtain anti- 

 body titres of 1-1,000, or even less. This indicates that we are dealing with a rather 

 poor antigen, feeble perhaps because of the injury done by the alkali and acid treat- 

 ment. Indeed, we believe this likely because, for reasons which will become apparent 

 in our section on immunological correlations, the antigen obtained by the simple solu- 

 tions of pneumococci with bile are qualitatively similar to the nucleoproteins obtained 

 by other methods from the same organisms, but are quantitatively much more potent 

 as antigens. The solution with bile is obviously a much less violent method of dis- 

 rupting the bacterial body than drying, grinding, and extracting with alkali and pre- 

 cipitating with acid. 



THE SPECIFIC SOLUBLE (rESIDUe) SUBSTANCE 



The possible importance of carbohydrates in immune phenomena was not recog- 

 nized until quite recently, though the fact that polysaccharides of the nature of gums 

 were present in certain bacteria has been known since the time of Pasteur. A review 

 of the earlier work dealing with this subject has been published by Buchanan.^ In 

 the more recent literature the bacterial carbohydrates have been more particularly 

 associated with the capsular structure,' but in none of this work did it become apparen 1 

 that (he carljohydrate materials had any immunological significance. 



' Wells, H. (i.: op. cil., p. 3S. 1925. 



* Buchanan, R. E.: Centriilbl. f. Bakleiiol., \h\. II, 22, 371. igoq. 



^ Toenniessen, E.: ibid., Abt. I, Orii;., 85, 225. 1920-21; Kramar, E.: ibid., 87, 401. 1922. 



