Generating DNA with Superhelical Turns
To understand how we may experimentally vary Lk, and consequently the degree of supercoiling, let us consider the lambda phage DNA. The molecules of this DNA are about 50,000 base pairs long and possess what are called sticky ends; that is, the ends of the DNA duplex are not flush. As shown in Fig. 2.9, the 5’ ends protrude in a single-stranded region of 12 bases. The sequence of the left end is complementary to the sequence of the right end. These sticky ends can be reassociated together to form a circle, which sometimes is called a Hershey circle after its discoverer. The phosphodiester bonds are not contiguous around the Hershey circle; hence its other name, a nicked circle. Circles having a break in only one of their backbones also are called nicked.
Nicks can be covalently sealed with DNA ligase. This enzyme seals the phosphodiester backbone of DNA between nicks that have a 5’-phos-phate and a 3’-hydroxyl. Following ligation which forms circles, Lk cannot be altered without breaking the backbone of one of the two strands. Hence, the sum of Tw, the right-helical turns, and Wr, the number of superhelical turns, is fixed. If under fixed buffer and tem-perature conditions, we were to anneal the ends of the lambda DNA together and then seal with ligase, the number of superhelical turns would be zero and Lk would be about 5,000, about one turn per ten base
Figure 2.9 Association of the self-complementary single-stranded ends oflambda phage DNA to form a nicked circle
pairs. For convenience let us say that the number is exactly 5,000. Furthermore, if we were to introduce distortion or even to wrap this DNA around a protein, the sum of Tw and Wr must still remain 5,000.
Suppose instead, the annealing and sealing had been done in the presence of ethidium bromide. Its intercalation between bases pushes the bases apart and partly untwists the DNA in this region because the phosphodiester backbone of the DNA cannot lengthen (Fig. 2.10). Hence the amount by which one strand wraps around another is decreased by the intercalation of the ethidium bromide. In the common B form of DNA, the bases are twisted about 34° per base, but the intercalation of an ethidium bromide molecule removes 24° of this twist. The number of helical turns in a lambda DNA molecule sealed in the presence of a particular concentration of ethidium bromide might be about 100 less than the number contained in a lambda DNA molecule sealed in the absence of ethidium bromide. Treating with DNA ligase under these conditions would produce a molecule with no writhe, Wr = 0, and with Lk = Tw = 4,900. If the ethidium bromide were then removed byextraction with an organic solvent, Tw would return to near its standard value of 5,000; but because of the requirement that Tw + Wr = 4,900 be a constant, Wr would become -100 and the circular DNA would writhe. It would have 100 negative superhelical turns, or σ or -100/5000 = -0.02.