446 ROLLIN D. HOTCHKISS 



essentially a deoxypentose polynucleotide containing the usual four bases 

 and lacking uracil. 



Physically, the transforming preparations appear to consist largely of 

 asymmetric particles giving, like thymus DNA, highly viscous aqueous 

 solutions, and behaving similarly to this material in the ultracentrifuge. 

 Their biological activity is destroyed, and viscosity reduced, by deoxyribo- 

 nuclease.^® Preparations of this enzyme from streptococci, and crystallized 

 pancreatic preparations, which do not have any other recognized en- 

 zymic activity, in very small amounts rapidly inactivate the transforming 

 agent. ®^' ^^ Significant also is the fact that of a number of tissue, serum, 

 and heated serum preparations, only those which contained active deoxy- 

 ribonuclease were able to bring about this inactivation.^^ Ribonuclease, 

 trypsin, and chymotrypsin had no effect upon activity. Certain parallels 

 have also been drawn between the effects of heat and pH change upon 

 viscosity and upon biological activity.^^ 



(2) Evidence that Other Substances Are Not Present. These results demon- 

 strated that the transforming preparation was essentially a DNA prepara- 

 tion and that its biological activity was dependent upon the intactness of 

 high-molecular DNA which it contained. Although such extracts have 

 detectable transforming activity at concentrations as low as a few thou- 

 sandths of a microgram per milliliter, these findings do not point un- 

 equivocally to identity of the DNA and the transforming factor. One 

 might suppose that (1) only a part of the DNA is biologically functional, 

 or possibly that (2) a small, still more highly active fraction of the material 

 consists of carbohydrate*^ or protein,*^ combined with, or stabilized by, 

 DNA. The serological tests originally applied to the material should have 

 given a barely recognizable positive reaction if there had been around 5 % 

 of a serologically active protein, or 0.05% of type-specific carbohydrate 

 present.*^ 



The generally accepted — though still largely unexplained — biological 

 specificities of protein made it seem most essential to demonstrate, if 

 possible, that the biologically specific transforming activity was not at- 

 tributable to protein, e.g., to a trace of nucleoprotein. The evidence against 

 a major content of protein includes the following facts: (1) the bulk of 

 the pneumococcal protein, at least, is removed during purification without 

 loss of the transforming activity,*^ ■^'' (2) as mentioned, the product does not 



66 M. McCarty and O. T. Avery, J. Exptl. Med. 83, 89 (1946). 



6' S. Zamenhof, in "Phosphorus Metabolism" (McElroy and Glass, eds.), Vol. 2, 



p. 301. Johns Hopkins Press, Baltimore, 1952. 

 68 M. Stacey, in "The Nature of the Bacterial Surface" (Miles and Pirie, eds.), 



p. 29. Blackwell Scientific Publications, O.xford, 1949. 

 6' A. E. Mirsky and A. W. Pollister, J .Gen. Physiol. 30, 117 (1946). 

 '" S. Zamenhof, G. Leidy, H. E. Alexander, P. L. FitzGerald, and E. Chargaff, Arch. 



Biochem. and Biophys. 40, 50 (1952). 



