306 jBACTERIAL VARIATION 



to synthesize it to subsequent generations, and they suggested that some enzyme, 

 readily liberated from the pneumococcal cells, destroys some substance that is an 

 essential stimulant of this change. AUoway (1932) was able to change rough 

 variants of Type II pneumococci to smooth Type III or smooth Type I, by growing 

 the former on Berkefeld filtrates of extracts derived from the latter, together with 

 normal pig serum, which contains anti-R agglutinins. In further experiments 

 (Alloway 1933) he substituted, for the Berkefeld filtrate, a preparation obtained by 

 dissolving pneumococci of the required type in a solution of sodium desoxycholate, 

 and precipitating the extract so obtained with alcohol. When rough pneumococci 

 were grown in serum broth to which a saline extract of such a precipitate was 

 added they gave rise to smooth strains of the type from which the extract was 

 derived. 



Recently, Avery, MacLeod and McCarty (1944), in defining both the conditions 

 for effecting the transformation and in making a presumptive identification of the 

 substance inducing the transformation, have made a noteworthy advance in the 

 subject of bacterial variation. To effect the transformation, a reactive variant 

 must be selected from an irreversibly rough (R) strain of Type II pneumococcus. 

 It is grown in broth free from as yet unidentified inhibitors, containing R anti- 

 serum free from enzymes that destroy the transforming principle, and the trans- 

 forming substance itself. This substance, isolated from a Type III pneumococcus, 

 was active in a dilution of 1 : 6 X 10^^, appeared to be homogeneous, with a particle 

 weight of 500,000, and had the gross chemical constitution, the exclusive suscepti- 

 bility to a specific depolymerase, and certain physicochemical properties, of a 

 deiSbxyribonucleic acid. 



'It appears that this particular type of nucleic acid interacts with the R cell 

 to give rise to a series of enzyme reactions culminating in the synthesis of Type III 

 capsular polysaccharide. Once the transformation is established, it is permanent, 

 and the transformed cell continues to produce both polysaccharide and the specific 

 desoxyribonucleate. We have in this phenomenon an outstanding example of 

 an impressed genetic variation ; i.e., a type-specific, heritable mutation induced 

 by a specific chemical agent. The mechanism of its action is unknown, but we 

 may provisionally ascribe it to a direct effect on a bacterial gene. 

 Variations in Virulence or Toxigenicity. 



Variations in the characters on which bacteria depend for the production of 

 disease in man and animals are clearly of particular importance to the medical 

 bacteriologist. It would, however, be altogether impossible in the course of a 

 general survey to give illustrative instances of the innumerable types of variation 

 that are associated with some change in virulence, or in toxigenicity. We may, 

 however, note a few general principles, leaving particular instances to be dealt with 

 in the systematic description of the different bacterial species, or in the chapters 

 devoted to the diseases to which they give rise. 



Taking virulence to mean the capacity for tissue invasion, and toxigenicity to 

 mean the power to produce a soluble toxin (see Chapter 44) we may note that 

 these two characters depend on different factors, so that, where both are present 

 in the same bacterial species, they may vary independently. A hsemolytic strepto- 

 coccus, for instance, may lose its power of invading the tissues of a particular 

 animal host without necessarily losing its power to produce a filtrable hsemolysin. 



We have seen that R — >■ S variation may be imj^ressed on relatively rough 

 strains of infective bacteria by passage through a susceptible animal. The varia- 



