Unlinked strands as a topological constraint on chromosomal DNA, plasmid integration, and DNA repair.

It is proposed that circular chromosomal DNA must be constrained such that the two strands are topologically unlinked. This structure can be replicated without strand breakage (nicking) by locally acting enzymes. The recently discovered form V DNA satisfies this topological constraint. The structure is likely to consist of left-handed and right-handed segments of double helix. The constraint of zero linkage has to be preserved by DNA repair, plasmid insertion and by crossing over. The argument presented in this paper is a topological one, following W.F. Pohl, that linkage is a global property that cannot be measured by locally acting enzymes. In contrast to Pohl no argument in favor of a side-by-side structure is presented. A zero linkage constraint would be hereditary and compatible with a multitude of local structures.