Enzymatic synthesis of deoxyribonucleic acid.

Determination of deoxyribonucleotide sequence in a deoxyribonucleic acid molecule is important from both the chemical and genetic points of view. It is also essential for answering the question of whether DNA synthesized in vitro by polymerase (1, 2) is a faithful copy of the nucleotidel sequence of the primer DNA. Although enzymatically synthesized DNA has the same over-all nucleotide composition as the particular primer DNA (3), it could not be inferred that this synthesis is a replication of the nucleotide sequences of the primer. Because of the limitations of present methods, complete sequence studies have never been made. Sinsheimer (4) has been able to identify 16 of the 25 possible dinucleotide pairs resulting from enzymatic digestion of calf thymus DNA. From recent analyses of various DNA’s by Shapiro and Chargaff (5) and by Burton and Peterson (6), there is now information on the frequency of occurrence of a pyrimidine nucleotide flanked on both sides by purine nucleotides and on the frequencies of short runs of pyrimidine nucleotides in the polynucleotide chains. These workers were able to deduce that there is a nonrandom arrangement of pyrimidine nucleotides in all of the DNA’s examined. In the studies to be reported here, we have derived the frequencies of the 16 possible nearest neighbor pairs in a variety of DNA’s by the technique of enzymatic incorporation of 5’-Pazlabeled nucleotides into DNA and then degradation of the DNA into 3’-nucleotides. Briefly, we have found that: (a) each DNA directs the synthesis of a product which has a unique and nonrandom pattern of the 16 nearest neighbor frequencies; (b) the DNA synthesized has the same nearest neighbor frequencies whether the primer is native DNA or enzymatically prepared DNA containing only traces of the original native DNA; and (c) the pattern of nearest neighbor frequencies in every case involves both base-pairing of adenine to thymine and of guanine to cytosine between sister strands of DNA, and opposite “polarity” of the two strands as proposed in the Watson and Crick model (7).