166 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES 



in other substrates. The first enzyme of this type was isolated by Dubos 

 and Avery (195, 199, 202) from certain soil bacteria. These enzymes 

 are highly specific, some being able to act only upon one type of pneu- 

 mococci. As a result of their action, the pneumococcus cell is rendered 

 susceptible to destruction by phagocytosis (819,821,822). This enzyme 

 was produced by the soil bacteria under selective conditions of culture, 

 that is, when the capsular polysaccharide of the pneumococcus was pres- 

 ent in the medium j the only other substance that could be used for its 

 production was aldobionic acid, derived from the above polysaccharide. 

 Yields of the enzyme were increased by increasing the concentrations of 

 the specific substrate in the medium from o.Oi to o.i per cent. Above 

 0.1 per cent, the yields decreased, 0.3 to 0.4 per cent inhibiting the 

 growth of the bacterium. The addition of o. i per cent yeast extract fa- 

 vored the production of the enzyme j proper aeration was essential, the 

 bacterium making the best growth in shallow layers of medium. The 

 enzyme was concentrated by distillation in vacuo and by ultrafiltration. 

 Toxic substances accompanying the active preparation could be largely 

 removed by the use of an aluminum gel. The enzyme is associated with 

 a protein which passes through a collodion membrane with an average 

 pore size of 10.6 Mj but is held back by pores having a diameter of 8.2 m- 

 After filtration, the enzyme can be recovered in solution by immersing 

 the membrane in distilled water or in physiological salt solution (30, 

 195,293,337). 



Dubos (188) believed that it is possible to develop "adaptive" bac- 

 terial enzymes against many organic substances. These enzymes exhibit 

 a great degree of specificity, as in the case of the enzyme that hydrolyzes 

 the capsular polysaccharide of the pneumococcus. The cell of this or- 

 ganism contains an enzyme that changes the cell from the gram-positive 

 to the gram-negative state, but is ineffective against streptococci or 

 staphylococci. 



Active preparations of the enzyme protected mice against infection 

 with as many as i ,000,000 lethal doses of the specific pneumococcus. The 

 enzyme retained its activity for 24 to 48 hours after its injection into 

 normal mice ; it also exerted a favorable influence on the outcome of an 

 infection already established at the time of treatment. A definite rela- 



