220 LINUS PAULING 



The careful search for errors in the synthesis of protein molecules should lead 

 to information that would either support or discredit the template theory of 

 biological specificity. One careful study of this sort has already been made, 

 D. Allen & W. A. Schroeder [unpubhshed obser^'ation] have analysed haemo- 

 globin from normal adult human beings and from patients with phenylpyruvic 

 oligophrenia, who have in their plasma and cerebral-spinal fluid a concentration 

 of phenylalanine 25 or 50 times greater than that in normal individuals. They 

 found no difference in the phenylalanine content of the two haemoglobins, to 

 within their experimental error, which was less than 3% (one residue per haemo- 

 globin molecule — the molecule contains about 32 residues of phenylalanine). 

 There are about twelve residues of tyrosine per molecule in that haemoglobin, 

 and this result shows that the probability that a phenylalanine residue will be 

 introduced in place of tyrosine (this is the most likely sort of error involving 

 phenylalanine) as a result of the increased concentration of phenylalanine in the 

 body fluids of the patients with phenylpyruvic oligophrenia is smaller than 8%. 

 at is possible, of course, that the introduction of the phenylalanine residue at 

 I tyrosine locus would change the properties of the haemoglobin molecule 

 enough to cause it to be rejected by the red cell. 



The extent to which the properties of a molecule are determined by the folding 

 of the polypeptide chains is suggested by the available information about the 

 difference in structure of normal adult human haemoglobin (haemoglobin A) 

 and sickle-cell-anaemia haemoglobin (haemoglobin S) [14]. The difference in 

 electrophoretic mobilities of these two forms of haemoglobin corresponds to a 

 difference in electric charge per molecule of about three electronic imits. The 

 analyses that have been made of amino acid composition show no difference in 

 composition within experimental error; in particular, there is no difference in 

 acidic groups or basic groups great enough to explain the difference in charge 

 of three electronic units. Moreover, the change in electrophoretic mobihty of 

 the globins obtained from these haemoglobins on slight denaturation, which 

 causes the molecules to acquire essentially the same electrophoretic mobility 

 shows clearly that the electric charge is determined in considerable part by the 

 way in which the polypeptide chains arc folded. 



There is now available a great deal of information about possible configurations 

 of polypeptide chains. Precise determinations of the structure of crystals of 

 amino acids, simple peptides, and other substances closely related to proteins 

 have been made by Professor Robert B. Corey, Dr E. W. Hughes, and their 

 collaborators, as well as, in recent years, by other investigators. It has been found 

 that the interatomic distances and bond angles are essentially constant in this 

 group of substances. The distances found are aC — C = 1*53 Â, aC — N = 1-47 Â, 

 C— N=i-32Â, C— 0=1-24 Â, angles N— aC— C = 110°, aC— C— N 

 = 114°, C— N— aC= 123°, O— C— N= 125°. The principal degrees of 

 freedom of the polypeptide chain, not determined by these parameters, are the 

 azimuthal angles about the two single bonds aC — C and aC — N. The stable 

 configurations of polypeptide chains that have been reliably recognized so far 

 (the a helix, the parallel-chain pleated sheet, the antiparallel-chain pleated 

 sheet, and the polyglycine-II structure) all involve azimuthal orientations 



