Biological Significance of Superhelical Turns

Biological Significance of Superhelical Turns

As explained above, supercoiling results from a linking number deficit in covalently closed, double-stranded circles. If such DNA did not form supercoils, it would possess a twist of less than one per 10.5 base pairs. It might be one twist per 11 base pairs, for example. Since DNA can wrap upon itself globally as it tries to attain a local twisting of once per 10.5 base pairs, it forms supercoils. Of course, the DNA resists the introduction of too many superhelical turns. Therefore, not all the linking number deficit is taken up by supercoiling. The deficit is parti - tioned between supercoiling and reducing the local twist of the DNA. The greater the linking number deficit, the greater the supercoiling and the greater the untwisting of the DNA. Untwisting DNA helps separate its strands. Therefore, negative supercoiling assists the formation of melted sections of DNA. This is the situation in vitro with pure DNA possessing a linking number deficit. What about in vivo?

Does the same DNA in vivo feel such a torsion, or are there unmelted regions of the DNA here and there, perhaps formed by bound proteins, that generate the overall linking number deficit? When these proteins are removed, we would find the linking number deficit. Despite its topological linking number deficit, such DNA in vivo would not feel the torsion described above.

The considerations discussed above can also apply to linear DNA molecules if its ends are prevented from free rotation. DNA may be constrained from free rotation because it is attached to a cellular structure or because free rotation is hindered due to its great length or the bulkiness of proteins that may be bound to it.

Several experiments suggest that DNA in bacteria not only possesses superhelical turns but also is under a superhelical torsion. The in vitro integration reaction of lambda phage, in which a special set of enzymes catalyzes the insertion of covalently closed lambda DNA circles into the chromosome, proceeds only when the lambda DNA possesses negative superhelical turns. In fact, tracking down what permitted the in vitro reaction to work led to the discovery of DNA gyrase. Presumably the enzymology of the in vitro and in vivo integration reactions is the same, and the supercoiling requirement means that in vivo the chromosome possesses superhelical turns.

A second experiment also suggests that DNA in normally growing E. coli contains superhelical torsion. Adding inhibitors of the DNA gyrase,such as nalidixic acid or oxolinic acid, which block activity of the A subunit of the enzyme, or novobiocin or coumermycin, which inhibit the B subunit, alters the rates of expression of different genes. The activities of some genes increase while the activities of others decrease. This shows that the drug effects are not a general physiological response and that the DNA must be supercoiled in vivo. Yet another indication of the importance of supercoiling to cells is shown by the behavior of DNA topoisomerase I mutants. Such mutants grow slowly, and faster-growing mutants frequently arise. These are found to possess mutations that compensate for the absence of the topoisomerase I by a second mutation that reduces the activity of gyrase, topoisomerase II. A third line of experimentation also suggests that the DNA in bacteria, but not eukaryotic cells, experiences an unwinding torsion from the linking number deficit. This is the rate at which an intercalating drug called psoralen intercalates and reacts with DNA when irradiated with UV light. The reaction rate is torsion-dependent. Altogether, it seems rea-sonable to conclude that the DNA in bacterial cells not only is super-coiled but also is under a supercoiling torsion.