Studies on the Mechanism of Protein Synthesis 163 



molecule may pass through the cytoplasmic ribonucleoprotein 

 particle. 



In parallel with these in vivo studies, we have investigated 

 the incorporation of labelled amino acids into proteins in cell- 

 free fractions of rat liver. One has little hesitancy in calling 

 the in vivo work a study on the mechanism of protein synthesis. 

 In the cell-free preparations, however, there is no net protein 

 synthesis as measured by the usual procedures, and the term 

 "incorporation" encompasses this degree of uncertainty. 

 In our context, however, this term "incorporation" does have 

 a rather precise meaning, which begins with the choice of 

 labelled amino acid. On the basis of past experiments in our 

 own laboratory (Zamecnik and Frantz, 1949) and of others 

 (cf. Tarver, 1954), we consider it safest to employ labelled 

 amino acids which are least likely to engage in other known 

 metabolic processes of the cell. For example, glutamic acid, 

 aspartic acid and alanine are located at the gateway to the 

 citric acid cycle ; glycine is a precursor of the purine molecule, 

 and is readily converted to serine and phosphatidyl serine; 

 lysine may become bonded to protein in side-chain linkage 

 by its £-amino group (Schweet, 1955); and the sulphur- 

 containing amino acids can attach to protein by disulphide 

 bonding (Tarver, 1954). The possibility that glutathione is 

 being synthesized in a cell-free system is further reason for 

 avoiding when possible glutamic acid, cysteine, and glycine as 

 amino acid labels for experiments on incorporation into pro- 

 teins. No doubt in time a particular pitfall will be found for 

 the use of every amino acid as a labelled precursor for protein. 



These considerations, however, have led us to the general 

 use of L-leucine and L-valine in our recent incorporation experi- 

 ments (Loftfield and Harris, 1956). Partial hydrolysis of 

 labelled protein at the end of typical in vitro experiments has 

 indicated the presence of a variety of labelled peptides 

 (Zamecnik and Keller, 1954; Keller and Zamecnik, 1956; 

 Zamecnik et al., 1956), and provides evidence of alpha- 

 peptide bonding. 



We have several types of evidence bearing on the question 



