SESSION IV. DISCUSSION 379 



levure sert de lien entre la protéine et l'acide nucléique. Nous avons en effet constaté la 

 formation d'un acide hydroxamique durant la synthèse de la 'Nucleoprotéine'. Nous ne 

 pouvons à l'heure actuelle être sûr cet acide hydroxamique soit relié à la synthèse du 

 polynucleotide mais c'est fort possible. 



K. G. loFFE (U.S.S.R.): 



My contribution concerns the connection between the sequence of amino acids in the 

 polypeptide chains of proteins and that of purine and pyrimidine bases in ribonucleic 

 acid (RNA). It seems to me that there must certainly be such a connection, especially 

 since the work of Fraenkel-Conrat and Schramm. It is clear that a definite sequence of 

 bases in a RNA corresponds with its appropriate sequence of amino acids in a particular 

 protein. 



But where can one get such pure individual RNA and protein which can confidently 

 be compared with one another? 



Although the structure of a few proteins (insulin, the adrenocorticotropic hormone, 

 melanophore-stimulating hormone and glucagon) has been estabHshed, they are not 

 suitable for this purpose. Their amino acid composition is too compUcated and their 

 polypeptide chains too anomalously various for them to be suitable for comparing with 

 the corresponding RNA. Furthermore, there is yet another difficulty in the impossibility 

 of isolating from a cell, which synthesizes many proteins, the very RNA corresponding 

 with a particular protein. 



From what has been said above the following conditions emerge : 



1. The cell should elaborate one protein in a strikingly large quantity compared with 

 the sum of all the other proteins of the cell. 



2. The protein should be easy to isolate and its structure should already be known. 



3. The polypeptide chain of the protein should be monotonous with continual repe- 

 tition of the same amino acids or sequences of amino acids. 



If the first condition is fulfilled and the cell really elaborates, for the most part, only 

 one protein, then the quantity of the corresponding RNA in the cell will also be markedly 

 greater than that of other RNA, and the order of the bases in it will also be monotonous 

 with frequent repetition of the same bases or groups of bases. 



The success of a comparison between the two sequences would then be assured. 



But where can we find such cells containing parallel inscriptions in two languages, in 

 that of RNA and that of protein ? 



Probably such Rosetta stones exist in Nature and are not rare. I shall give you one 

 example now. 



It concerns the fibroin of silk. Its structure has been elucidated to the extent of 50-60% 

 and the part which is known may be expressed: (X. Gly. Ala. Gly. Ala. Gly.)n or (Ala. Gly. 

 Ala. Gly. X. Gly.)« where Ala = alanine, Gly = glycine and X = other amino acids. I 

 have put forward the following structure for the tyrosine-containing part of fibroin : 



Ser. Gly. Ala. Gly. Ala. Gly. Ser. Gly. Ala. Gly. Ala. Gly. Tyr. Gly. Ala. Gly. Ala. Gly. 



Val. Gly. Ala. Gly. Ala. Gly or 

 Ser. Gly. Ala. Gly. Ala. Gly. Tyr. Gly. Ala. Gly. Ala. Gly. Val. Gly. Ala. Gly. Ala. 

 Gly. Ser. Gly. Ala. Gly. Ala. Gly 

 where Ser = serine, Tyr = tyrosine and Val = valine [i]. 



This structure has been confirmed to some extent by Lucas et al. [2]. 



As the peptide Ala. Gly. Ala. Gly. accounts for exactly 2/3 of the amino acid residues of 

 the tyrosine-containing part of fibroin, so also it is repeated almost as often in the whole 

 molecule of fibroin in that this sequence is also repeated many times in the length of the 

 peptide chain. 



Therefore, one must expect that in the RNA which synthesizes the fibroin there will 

 also be frequent repetitions of the same bases along the length of the molecule. 



The fibroin-manufacturing part of the silk gland of the silkworm Bombyx mori is sharply 

 demarcated from the rest. From the walls of the gland one can obtain cells of one sort, 

 free from other cells. The posterior part of the gland produced enormous amounts of 

 pure fibroin without sericin. Therefore the form of RNA which synthesizes fibroin should 



