This work is a valuable review of experiments that probe the interactions of small molecules and DNA under tension. However, I disagree with the review's premise (Sections 1.3 and 2.1) that overstretching of bare, torsionally unconstrained DNA consists solely of dehybridization of the molecule. Understanding the nature of overstretching is a prerequisite for understanding the interactions of stretched DNA and small molecules.
Two physical pictures have been advanced in an attempt to explain overstretching. The force-induced melting picture assumes that the stable overstretched state near 65 pN is a form in which both strands bear tension but do not interact -I will call this 'molten' (M) DNA -punctuated by regions of B-DNA. It is crucial to distinguish M-DNA from the unhybridized state in which only one strand bears tension; following Ref.
[5] I will call the latter 'unpeeled' (U) DNA. U-DNA is argued to emerge if the molecule dehybridizes at nicks or free ends . The S-DNA picture, by contrast, assumes that there exists an elongated, hybridized phase of DNA (S-DNA) . I wish to address two claims made in the review regarding this picture. The first challenges the need for such an assumption (Section 1.3: "However, it is not clear that additional fitting parameters [i.e. those corresponding to S-DNA] are needed to explain DNA stretching experiments"). The second claim is that recent visualization of dehybridization during overstretching is incompatible with the S-DNA picture (Sections 1.3 and 2.1), and therefore that overstretching is force-induced melting.
In response to the first claim I argue that there are indeed several experimental results that do not sit easily with the notion that DNA can only dehybridize upon stretching. Consider the following four sets of observations. First, overstretched DNA is under some conditions stiffer than both forms of unhybridized DNA and B-DNA (if one extrapolates its force-extension profile beyond 65 pN) . This suggests the existence of a phase mechanically distinct from M-, U-, or B-DNA. Second, under standard conditions, poly(dA-dT) overstretches at about 35 pN, but λ-DNA (about 50%