ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 381 



in electrophoretic mobility to free PNA. The pH of the extracts could be 

 varied from 5.7 to 8.2 and the ionic strength from 0.08 to 0.31, without 

 resulting in the appearance of a band characteristic of free nucleic acid. 

 Szafarz therefore concluded that the nonsedimentable fraction of cyto- 

 plasmic PNA is bound to protein by bonds stable over a considerable 

 range of pH and ionic strength. In contrast, the bonds between simple 

 proteins such as albumin and PNA, which are formed when the two com- 

 ponents are mixed at a pH intermediate between their isoelectric points, 

 are not stable in solutions of an ionic strength equivalent to that used in 

 the experiments with the cytoplasmic nucleoproteins.-* The bond in the 

 cytoplasmic nucleoprotein may owe its strength to the particular proper- 

 ties of either the protein or the PNA, or to a particular steric relation 

 between the two components. An interesting observation by Szafarz'" 

 suggests that proteins of the cytoplasm possess special ability to form 

 stable complexes with PNA. Thus, small amounts of commercial yeast 

 PNA added to cytoplasmic extracts of the flagellate Polyf.omella coeca 

 could not be demonstrated by electrophoresis of the mixture over an 

 extended range of pH and ionic strength. The proteins of the cytoplasm 

 were capable of reacting in vitro with PNA isolated from a different species, 

 and the resulting complex possessed the stability of a cytoplasmic nucleo- 

 protein. The study of such artificial nucleoproteins may lead to the dis- 

 covery of the structural properties of cytoplasmic proteins which enable 

 them to combine with PNA. At present, it would seem most likely that 

 the charged groups on the surface of protein and PNA molecules are re- 

 sponsible for the ease of the interaction, and that the great number of such 

 coulombic bonds between each molecule of protein and PNA accounts for 

 the stability of the nucleoprotein complexes. 



It is evident that free PNA could not be found in tissue extracts in the 

 presence of reactive cytoplasmic protein, even if it did exist in that state 

 in the intact cell. Without further evidence, the nucleoproteins isolated 

 from cytoplasmic extracts cannot be considered to be characteristic com- 

 ponents of the intact cell. 



However, nucleoproteins which differ in properties from the complexes 

 formed by the direct interaction of protein and PNA seem to exist. An 

 interesting example is the crystalline nucleotropomyosin isolated by 

 Hamoir from carp muscle'^ by the method described in detail in the pre- 

 ceding section. The two crystalline proteins, tropomyosin and nucleo- 

 tropomyosin, showed identical behavior in an electric field, although the 

 former was free of nucleic acid while the latter contained 15% PNA. The 

 presence of nucleic acid thus did not affect the charges on the surface of 

 the protein, and it must be assumed that the PNA fits into the pattern of 



28 L. G. Longsworth and D. A. Maclnnes, J. Gen. Physiol. 26, 507 (1942). 



