The basic DNA transactions involved in NER broadly consist of the following steps: damage surveillance and recognition; damage verification; helix opening and stabilizing of the repair intermediates; dual incision of the DNA in the context of the lesion; repair synthesis and DNA ligation. At each step along this complex transformation and restoration of DNA, protein partners are recruited to the site of repair to perform these DNA transactions. Each of these transactions proceeds via combinations of highly regulated protein-protein and protein-DNA interactions that act upon specific substrate repair intermediates to transform them into a new set of product repair intermediates that can be recognized by subsequent players in the pathway. Here we examine some of the features of damage recognition and verification.
The impairment of Watson-Crick pairing may result in a greater ability of the orphaned base or the lesion to be flipped out. The base flipping transition involves two regimes, a regime in which the penalty of disrupting the Watson-Crick pairing rises quadratically around the mean position of the base in the intrahelical conformation until about a displacement by 25◦ on either side, and a second regime in which the energy rises linearly. The energetic costs of flipping undamaged adenine (A), thymine (T), guanine (G) and cytosine (C) are estimated to lie in the range of 15-20 kcal mol-1. Alteration of base chemistry results in perturbations in its base flipping energy. In contrast, the energetic cost of flipping a cis, syn-cyclobutane pyrimidine dimer (CPD) is estimated to be 6.25 kcal mol-1. Further, this estimated value was found to be sensitive to the sequence context with a lower ∆Gflip for the CPD in the context of A or T sequences by about 0.5-2 kcal mol-1. Similarly, the energetic cost of flipping another NER substrate 14R (+)-trans-anti-dibenzo[a,l]pyrine-N2-dG (14R-dG*) was found to be 10.4 kcal mol-1, about 7.7 kcal mol-1 lower than the corresponding undamaged DNA.
It is important to note that even with these lower penalties of base flipping, neither the damaged nor the orphaned bases are predominantly extrahelical in nature. These values merely indicate that damaged bases have a greater propensity to make extrahelical excursions compared to undamaged bases, in a lesion dependent manner. The destabilization of the damaged and orphaned bases arising from the lower energetic penalty of base flipping of damaged bases may be exploited by damage recognition enzymes to recognize damage with high specificity. Given the significantly lower energetic barrier to flipping damaged bases, a first test for damage might involve the probing of the Watson-Crick base pairing (or lack thereof) between the bases being tested in the search for damage. Structural features on the protein which act as the probe and evict the damaged bases could interact with undamaged bases in the context of the DNA, resulting in the lowering of the barrier for base-flipping so that it would be accessible to thermal fluctuations in the presence of the repair factor. Damage recognition might involve probing the DNA for deformability via the energetically unfavorable transition of base flipping and this may represent a test for the presence of damage. Such a mechanism would confer damage specificity, enabling the recognition of rare damaged bases in a vast majority of undamaged bases upon rapid sampling. Indeed examination of the DNA bound states of various photoproduct recognizing repair factors reveals that base flipping is a common theme for the recognition of UV-induced photoproducts and likely other bulky lesions.
If damage recognition occurs spontaneously, the energetic cost of displacing the lesion must be compensated by enthalpic interactions which release energy upon binding or increase the entropy of the system upon binding. How might these constraints be met?
Two possible solutions involve providing enthalpic stabilization by providing interactions to the displaced lesion or to the orphaned bases. Thus, successful damage recognition would be bipartite: in the first step, the lesion is displaced to an extrahelical conformation and in the second step, the orphaned base or the lesion is stabilized via interactions with the protein.
An examination of crystal structures of damage recognition enzymes reveals that these ideas outlined above are general features of photoproduct recognition and variations on the combinations of these are present in nature.
The substrate affinity, specificity and repertoire for DNA damage recognition enzymes arise from the steric constraints imposed by the lesion recognition pocket. Indeed, from the examples provided here, it is evident that the lesion binding pocket has evolved specifically to recognize certain substrates and these direct the substrate specificity of the enzyme. Highly functionally specialized enzymes such as CPD or 6-4 photolyases demonstrate a high specificity for one substrate over another that arises from sterically selective interactions in the lesion binding pockets. Broader substrate specificity is obtained by more permissive lesion binding pockets. For example, the UVDDB protein exhibits a lesion binding pocket that accommodates 6-4PP and CPD lesions, suggesting that this repair factor has evolved for the specific recognition of these lesions. Other damage recognition factors possess even broader substrate specificity. The global damage recognition factor, Rad4 (orthologue of the human XPC protein) exhibits a binding pocket for the orphaned base while exhibiting no apparent binding pocket for the lesion. This lack of a binding pocket for the lesion explains the broad substrate repertoire of Rad4 and the related XPC protein, while conferring the properties of a sensor of altered DNA dynamics arising from DNA damage.