The circular dichroism of supercoiled DNA has been calculated as a function of superhelix density using the exciton and free-electron models. The two models agree with each other, and with experiment, at large negative superhelix densities, but disagrees at lower superhelix densities. The reasons are discussed in terms of the role of chiral torsion.
In chromatin, the right-handed DNA double helix is itself helically coiled to give a left-handed superhelix, the tightness of which is believed to act as a "switch" to control gene expression through its effect on transcription [1][2][3][4][5][6]. The unusually tight supercoiling in cancer cells [4,7] is believed to promote uncontrolled transcription, and could find application in clinical diagnosis [7,8].
Further elucidation of the role of supercoiling in the control of gene expression and in carcinogenesis requires accurate and non-destructive means of measuring superhelix density. Traditional methods such as sedimentation and X-ray or electron diffraction can be destructive: the DNA is easily nicked leading to re/ease of the supercoiling. The band-counting gel electrophoresis method, although elegant and accurate, is time-consuming, requires expensive nickingclosing enzymes, and can consume a large amount of the sample if several attempts are necessary to obtain a perfect "ladder" of bands.
Circular dichroism (CD), however, could offer a rapid and non-destructive method, since it neither consumes the sample nor destroys the supercoiling, and is known to be sensitive to superhelix density: in chromatin, the positive CD band at 280 nm is strongly depressed, to about 70% of its height in nonsupercoiled B DNA [9,10]. Some authors have attributed this to secondary structural change [ 11,12], while other suggest tertiary interactions between adjacent turns of the superhelix [13,14]. In an attempt to identify the cause of the change in CD on super-