THE STRUCTURE OF DNA 



For this to occur the hydrogen bonds Hnking the complementary chains must 

 break and the two chains unwind and separate. It seems likely that the single 

 chain (or the relevant part of it) might itself assume the helical form and serve as 

 a mould onto which free nucleotides (strictly polynucleotide precursors) can 

 attach themselves by forming hydrogen bonds. We propose that polymerization 

 of the precursors to form a new chain only occurs if the resulting chain forms the 

 proposed structure. This is plausible because steric reasons would not allow 

 monomers "crystallized" onto the first chain to approach one another in such a 

 way that they could be joined together in a new chain, unless they were those 

 monomers which could fit into our structure. It is not obvious to us whether a 

 special enzyme would be required to carry out the polymerization or whether the 

 existing single helical chain could act effectively as an enzyme. 



Difficulties in the Replication Scheme 



While this scheme appears intriguing, it nevertheless raises a number of 

 difficulties, none of which, however, do we regard as insuperable. The first 

 difficulty is that our structure does not differentiate between cytosine and 5- 

 methyl cytosine, and therefore during replication the specificity in sequence in- 

 volving these bases would not be perpetuated. The amount of 5-methyl cytosine 

 varies considerably from one species to another, though it is usually rather small 

 or absent. The present experimental results (Wyatt, 1952) suggest that each 

 species has a characteristic amount. They also show that the sum of the two 

 cytosines is more nearly equal to the amount of guanine than is the amount of 

 cytosine by itself. It may well be that the difference between the two cytosines is 

 not functionally significant. This interpretation would be considerably strength- 

 ened if it proved possible to change the amount of 5-methyl cytosine in the DNA 

 of an organism without altering its genetical make-up. 



The occurrence of 5-hydroxy-methyl-cytosine in the T even phages (Wyatt 

 and Cohen, 1952) presents no such difficulty, since it completely replaces cytosine, 

 and its amount in the DNA is close to that of guanine. 



The second main objection to our scheme is that it completely ignores the role 

 of the basic protamines and histones, proteins known to be combined with DNA 

 in most living organisms. This was done for two reasons. Firstly, we can for- 

 mulate a scheme of DNA reproduction involving it alone and so from the view- 

 point of simplicity it seems better to believe (at least at present) that the genetic 

 specificity is never passed through a protein intermediary. Secondly, we know 

 almost nothing about the structural features of protamines and histones. Our 

 only clue is the finding of Astbury (1947) and of Wilkins and Randall (1953) 

 that the X-ray pattern of nucleoprotamine is very similar to that of DNA alone. 

 This suggests that the protein component, or at least some of it, also assumes a 

 helical form and in view of the very open nature of our model, we suspect that 

 protein forms a third helical chain between the pair of polynucleotide chains 

 (see Figure 4). As yet nothing is known about the function of the protein; 

 perhaps it controls the coiling and uncoiling and perhaps it assists in holding 

 the single polynucleotide chains in a helical configuration. 



203 



