Watching DNA repair in real time.

I n 1949, Albert Kelner described a miraculous recovery of cells damaged by UV radiation when cells were exposed to soft sunlight (1). It was found later that an enzyme photolyase (PL) is involved in photoreactivation (2). With a paper in PNAS by Dongping Zhong, Aziz Sancar, and their colleagues (3) describing the real-time progression of the photoreactivation repair reaction observed by femtosecond time-resolved spectroscopy, the story of PL, originating 62 y ago, has come to be as complete as any research subject can be. They traced all the steps of the repair reaction, which involves electron transfer between the enzyme and DNA, the splitting of the thymine dimer on DNA, and the return of the electron back to the enzyme redox cofactor following the mechanism predicted earlier. The whole repair reaction takes less than a nanosecond to complete all the steps (Fig. 1). PL was identified over 50 y ago by Claud Rupert (2), and much of the subsequent work was done by Aziz Sancar (4) and his coworkers. Sancar was a student of Rupert in the early 1970s. By the end of the decade, he cloned the gene of the enzyme, purified it, and identified two chromophores of the enzyme: the light-absorbing MTHF (or sometimes HDF) and the redox cofactor FADH. MTHF absorbs light and transfers the excitation to FADH, which transfers an electron to the dimer, putting it on the antibonding orbital; the antibonding electron splits the dimer, after which the electron returns back to FADH. This scheme emerged by the end of the 1980s. However, neither the structure of the enzyme nor the molecular details of the hypothetical mechanism were known at that time. The breakthrough into the molecular mechanism came in 1995, when the structure of an isolated enzyme was solved by Deisenhofer, Sancar, and coworkers (5). The structure brought the discussion of the mechanism to the atomistic level and allowed performance of the first computational studies, which provided further insights into the nature of the repair reaction. The structure revealed, first of all, the position of two cofactors responsible for the repair reaction in the enzyme. Another feature discovered was that on the positively charged surface of the enzyme that obviously binds the DNA helix, there is an opening leading to a pocket in which the redox cofactor is located. The conformation of the FADH in the binding pocket of the protein was found to be unusual in that the molecule is folded in such a way that the adenine ring and that of the flavin are both exposed to the opening of the binding pocket, which extends to the surface of the protein and can communicate with the redox substrate (i.e., the thymine dimer). However, the dimer in the intact DNA structure was found to be well-folded into the interior of the helix, posing a puzzle as to how the redox cofactor can reach the substrate for an efficient electron transfer reaction. Because the structure of the complex of PL bound to the DNA was not known at that time, the field was open for hypotheses and speculations. The structure immediately prompted molecular dynamics simulations to understand the missing pieces of the repair reaction mechanism. One theory group developed a docking algorithm that probed the rate of electron transfer between PL and DNA in complex models (6, 7). The reasoning was that the long-distance electron transfer is sensitive to the structure of the intervening medium (8), and the time scale of the electron transfer reaction was known to be in the range of 100 ps. It was found that that the only way to understand the high rate of electron transfer was to consider structures with the dimer flipped out of the DNA helix. (Earlier, a similar hypothesis was put forward by Sancar and coworkers.) The simulations resulted in a well-defined structure in which the dimer was flipped out, sitting deep inside the binding pocket, and in hydrogen-bonded contact between the carbonyl oxygens of the dimer and the adenine moiety of FADH. The found structure of the PL/T<>T dimer complex was intellectually appealing because it explained the presence of the deep cavity leading to FADH cofactor, so perfectly matching the dimensions and properties of the dimer itself. Indeed, it would be surprising if nature had created such a special feature of the enzyme and not used it. The well-defined structure determined in simulations suggested another intriguing feature of the electron transfer reaction by which the repair electron is passed from FADH to the thymine dimer on DNA. The dimer was found to be most strongly coupled to the adenine moiety of FADH. As a matter of fact, the structure of the complex suggested the FADH cofactor is bent in the binding pocket in such a way as if to extend a hand (i.e., adenine moiety) to a possible binding partner. The problem, however, is that the initial location of the electron to be sent for the repair mission, even after excitation of FADH, is still on the flavin part of the molecule. The simulation of the repair reaction suggested that the adenine can be used as a virtual stepping stone for the superexchange coupling between the dimer and the flavin. Superexchange, also known as tunneling, is a typical mechanism that nature uses to transfer electrons over long distances (8). Thus, by the end of the 1990s, theory suggested much needed refinement of the picture, offering detailed predictions of the structure and mechanism of electron transfer. In addition, theory suggested some details of the dimer-splitting reaction itself (9). The ball was now in the experimental court. The next major breakthrough came almost 5 y later, at the end of 2004. The group of Carell, Essen and coworkers (10) in Germany solved the structure of PL bound a stretch of DNA containing thymine dimer. The DNA is clearly seen as locally melted; the dimer is flipped out, sitting deep in the pocket of the redox cofactor, forming hydrogen bonds between carbonyl oxygens of the dimer and the adenine moiety of FADH, precisely as theory predicted (6, 7), matching with Fig. 1. Complete photocycle of cyclobutane pyrimidine dimer repair by DNA PL, with all timeresolved elementary steps of the elucidated molecular mechanism (3).