290 SESSION III. DISCUSSION 



continual evolution of protein molecules. It is, therefore, natural that the structure of the 

 macromoleculcs of proteins should retain the impression of many chance alterations which 

 have taken place during the process of evolution. This is probably the explanation of 

 what seem, at first glance, to be queer, general, statistical regularities, in the distribution 

 of different amino acids in various proteins without any functional resemblance, which 

 have been found by Tristram. This accounts for the similarity (in amino acid composition) 

 of completely dissimilar proteins which was pointed out by Synge. If this concept is 

 correct it will oblige us to take a different attitude to the whole problem of the chemical 

 structure of proteins. For example, deciphering of the complete structural formula of a 

 protein or the amino acid sequence of the whole macromolecule will lose its point. Only 

 the functionally reactive centres are constructed in a definite way corresponding to their 

 functions. The other parts of the molecule may not have any constant structure, being 

 the vestigial relics of their earlier history. 



A. G. Pasynskii (U.S.S.R.): 



Catalytic Protein Structures with reference to the Papers of 

 Hoffmann-Ostenhof and Sorm 



We are now in a position to put forward, tentatively, the idea that there are several 

 stages in the formation of catalytic protein structures. At first relatively short polypeptide 

 chains must have been formed from amino acids; it has, in fact been shown by recent 

 work that even small polypeptide chains (of 25-30 residues or less) may have a catalytic 

 activity. Oriented adsorption of such peptides, or simpler ones on kaolin (as suggested by 

 Bemal) or silica gel etc. might be the next stage of significance for the formation of 

 abiogenic or protobiogenic catalytic structures. However, this sort of adsorption on a 

 carrier could not give the dehcate and precise structure of the active centres which could 

 be given by the folding of a long polypeptide chain, not to speak of the biological necessity 

 of having the catalytic centres attached to polypeptide chains rather than to particles of 

 clay etc. The formation of catalytic centres within the long polypeptide chains could, 

 therefore, have attained an evolutionary predominance over adsorbed catalytic structures. 

 This process could have culminated in the formation of catalytic protein molecules 

 with their strictly determinate chemical structure. It must be mentioned that, as well 

 as having a determinate structure of its active centre, the protein molecule can make 

 extremely delicate modifications in the activity of the catalytic centre by changes in its 

 composition, charge, configtiration etc. The part played by the rest of the protein mole- 

 cule must not, therefore, be considered, as suggested by Prof. Bresler, as being vestigial 

 or atavistic, it is absolutely necessary. It is, however, certain that the active centre can 

 only constitute a small part of the protein molecule and that, accordingly, there was no 

 need for the complete protein molecule to be synthesized for the production of the first 

 examples of enzymic activity based on peptide structures. Besides the evidence put 

 forward by Bresler one may point to the recent work of Rogers & Kalnitsky (Biochim. 

 biophys. Acta, 23, 525, 1957) on the active centre of ribonuclease. At the conference on 

 proteins held in Prague in 1956, I gave a paper showing that if we compare the relative 

 frequency of occurrence of different dipeptides, as calculated statistically from the amino 

 acid composition, with the extent to which they actually occur in an amino acid chain, 

 where it has been deciphered, as in insulin (Sanger), silk fibroin (loffe) and ribonuclease 

 (Hirs, Stein & Moore), then we find that in a non-catalytic protein (fibroin) the calculated 

 and actual results are approximately the same, while, in the specific proteins (insulin and 

 ribonuclease) the actual amoimts of most of the dipeptides agree with the calculated 

 results, though the content of some dipeptides differs from what would be expected on 

 statistical grounds by a factor of 8-10. Thus, it may be that the structure of the active 

 centre is really determined by a different factor from that governing the protein molecule 

 as a whole. The interesting facts given in Sorm's paper also show a predominance of 

 some types of dipeptide linkages. It would be interesting to evaluate on models the 

 magnitude of the kinetic parameters (AF* and AH*) determining the relative rates of 



