Topological considerations in the theory of replication of DNA.

An obvious difficulty of the Watson-Crick model of DNA is that the intertwining of the strands would seem to hinder their separation during replication. The nature of the difficulty is here made precise and is called the alignment problem. It is shown that the swivelase theory, found in current textbooks and thought to overcome the difficulty, does not in fact do so. The various conceivable solutions of the alignment problem are considered and rejected, leading to the conclusion that chromosomal DNA is not double-helical. An alternative model of DNA is discussed. In addition a topological classification of DNA denaturation processes is given, and an alternative interpretation of the gel electrophoresis experiments on circular duplex DNA is suggested.