Vol. 45, 1959 BIOCHEMISTRY: KORNBERG ET AL. 781 



Control of DNA Synthesis at the Level of Deoxy nucleotide Phosphorylation. — 

 It appears from studies with various analogues of the naturally occurring purines 

 and pyrimidines^ that an analogue in the form of a deoxynucleoside triphosphate 

 is incorporated into DNA when its structure has the hydrogen-bonding properties 

 of the base it replaces. For example, uracil and 5-bromouracil are both effective 

 substitutes for thymine. Yet, uracil is never found in DNA^* while bromouracil, 

 when supplied to the cell, readily replaces thymine in the DNA.^^ The reason ap- 

 pears to be the absence of any mechanism for phosphorylating deoxyuridylate as 

 contrasted with the availability of an enzyme for phosphorylating 5-bromodeoxyuri- 

 dylate or 5-methyldeoxyuridylate (thymidylate).^ 



The present studies of phage-infected cells provide additional examples of control 

 at the phosphorylation level. Earlier observations with 5-methyldeoxycytidylate3 

 had suggested the lack of an enzyme for phosphorylation of a 5-substituted cytosine 

 deoxynucleotide, and posed the problem of how 5-hydroxymethylcytosine deoxy- 

 nucleotide becomes a substrate for phage T2 DNA synthesis. This problem seems 

 to be solved by the development of a new enzyme after phage T2 infection of the 

 cell. According to our studies, and those of Flaks and Cohen,^ the synthesis of a 

 compound novel for the cell is carried out by a new enzyme, the formation of which 

 is presumably induced by the phage DNA. Another example of control at the 

 phosphorylation level is provided by the device which appears to be responsible for 

 the absence of cytosine in T2 DNA. For lack of a mechanism to eliminate the 

 dCTP-synthesizing enzyme system, the evolution of an enzyme to destroy the 

 dCTP provides an effective alternative. 



Direct Suhstitvtion on DA^^ .—Genetic studies indicate that the glucose contents 

 of DNA of phages T2, T4, and T6, and strains derived from them by recombination, 

 are a fixed property resembling other phenotypic characters of these phages. ^^ ^° 

 It appears plausible, therefore, that the precise arrangement of these glucosylated 

 HMC residues in the DNA may be a source of genetic information, and insight into 

 the nature of the replication of these types of phage DNA would clearly be of con- 

 siderable value. 



The incorporation of a fixed proportion of non-, mono-, and polyglucosylated 

 derivatives of dHMC-TP during the polymerization of the DNA chain is difficult 

 to conceive because it is likely that these derivatives would all behave similarly in 

 hydrogen-bonding to guanine. It becomes even more difficult to conceive the in- 

 corporation of these various HMC residues in a precise sequence in DNA on the 

 basis of selection of the triphosphate derivatives. The existence of an enzymatic 

 system for direct glucosylation of DNA offers an alternative which seems to circum- 

 vent these difficulties. At this stage, the information is still too fragmentary to 

 determine whether this glucosylating enzyme, and perhaps an additional one for 

 polyglucosylation, will be sufficient to explain the replication of various phage 

 DNA's. It is apparent that further studies with T4- and T6-, as well as T2-in- 

 fected cell enzyme systems are essential. 



Superficially analogous to the partial glucosylation of the HMC residues in phage 

 DNA is the partial methylation of the cytosine residues in certain plant and animal 

 DNA's, such as wheat germ and calf thymus. 2" In the light of our findings, it 

 would be interesting to look for an enzymatic mechanism for direct methylation of 

 DNA in these tissues. 



134 



