STKUCTURAL AND CHEMICAL ARCHITECTURE OF HOST CELLS 201 



c. Potter and Dounce (1956), have reported on the existence of alkahne 

 stable amino acid-nucleotide complexes in isolated RNA preparations and 

 consider these to imply the presence of amino acids bound in phosphoamide 

 linkage. Domice (1952) has suggested that such compounds are the inter- 

 mediates in polypeptide synthesis. The amide group of one bound amino acid 

 would be transferred to the free carboxyl of a second, adjacent, bound amino 

 acid by a mechanism analogous to that proposed by Borsook for carboxyl- 

 bound amino acids, 



3. The Problem of Polypeptide Order 



The mechanisms of protein sjmthesis outlined above do not accomit for 

 the specificity of the polypeptide cliam and serious thought is beginning to 

 be given to the problem of how a polynucleotide might determine amino 

 acid sequence. Existing evidence appears to suggest that all possible dipep- 

 tide sequences may exist in proteins, although many fewer than the possible 

 400 dipeptides (assuming 20 different L-ammo acids) have so far been found 

 (Gamow et al., 1955; Brenner, 1957). Similar results have been obtained with 

 RNA and DNA, although, as noted earher, in the latter case, 5-methyl 

 cytosine and guanine deoxyribotides are associated as a dinucleotide in a 

 decidedly nonrandom fashion. The lack of restriction to pairing of aU amino 

 acids or of all nucleotides imposes certain requirements on coding systems, 

 while the frequency of pairing in these polymers may help in describing 

 specific properties of both the polymers and the coding systems. It is of 

 interest, in connection with the latter point, that the digestion of tobacco 

 mosiac virus RNA by ribonuclease had led to the production of fragments 

 whose relative contents of purine and pyrimidine residues are essentially 

 similar to those expected if the same numbers of these bases were randomly 

 ordered in a polynucleotide (Hart, 1957). This is only taken to mean that 

 order in this RNA template involves something more complex and more 

 difficult to discern than the repetition of a code of a few residues. 



In nucleoprotamines, the ratio of amino acid to nucleotide approaches 1:1. 

 However, in the more complex nucleohistone, it approaches 2:1. In the 

 residual ribonucleoproteins of the microsomal fraction after treatment with 

 sodium deoxycholate, the ratio also approaches 2 : 1. If it is assumed that in 

 the latter case these amino acids are not part of dangling polypeptides but 

 are as closely associated with RNA, as histone is with DNA (Chargaff et al., 

 1956), such a relationship appears to exclude a simple 1 : 1 relationship 

 whereby a given nucleotide in a sequence determines the fixation of a single 

 amino acid. Such a relation, which does exist in nucleoprotamine, had been 

 widely used as a working hypothesis because of the similarity of the inter- 

 nucleotide distance in DNA to the spacing between amino acid residues in an 

 extended polypeptide chain. 



