Escherichia coli photolyase uses blue light to repair cyclobutane pyrimidine dimers which are formed upon irradiation of DNA with ultraviolet (UV) light. E. coli photolyase is a flavoenzyme which contains a flavin adenine dinucleotide (FAD) in its active site and a 5,10-methenyltetrahydrofolate (MTHF) as a light-harvesting pigment. In the isolated enzyme, the FAD cofactor is present as a stable neutral radical semiquinone (FADH • ). In this paper, we investigate the interaction between photolyase and UV-damaged DNA by using resonance Raman and UV-vis spectroscopy. Substrate binding results in intensity changes and frequency shifts of the FADH • vibrations and also induces electrochromic shifts of the FADH • electronic transitions because of the substrate electric dipole moment. The intensity changes in the resonance Raman spectra can be largely explained by changes in the Raman excitation profiles because of the electrochromic shift. The size of the electrochromic shift suggests that the substrate binding geometry is similar to that of oxidized FAD in reconstituted photolyase. The frequency changes are partially a manifestation of the vibrational Stark effect induced by the substrate electric dipole moment but also because of small perturbations of the hydrogen-bonding environment of FADH • upon substrate binding. Furthermore, differences in the resonance Raman spectra of MTHF-containing photolyase and of an MTHF-less mutant suggests that MTHF may play a structural role in stabilizing the active site of photolyase while comparison to other flavoproteins indicates that the FAD cofactor has a strong hydrogen-bonding protein environment. Finally, we show that the electrochromic shift can be used as a direct method to measure photolyase-substrate binding kinetics.

DNA photolyase repairs pyrimidine dimer lesions in DNA through light-induced electron donation to the dimer. During isolation of the enzyme, the flavin cofactor necessary for catalytic activity becomes one-electron-oxidized to a semiquinone radical. In the absence of external reducing agents, the flavin can be cycled through the semiquinone radical to the fully reduced state with light-induced electron transfer from a nearby tryptophan residue. This cycle provides a convenient means of studying the process of electron transfer within the protein by using transient EPR. By studying the excitation wavelength dependence of the time-resolved EPR signals we observe, we show that the spin-polarized EPR signal reported earlier from this laboratory as being initiated by semiquinone photochemistry actually originates from the fully oxidized form of the flavin cofactor. Exciting the semiquinone form of the flavin produces two transient EPR signals: a fast signal that is limited by the time response of the instrument and a slower signal with a lifetime of approximately 6 ms. The fast component appears to correlate with a dismutation reaction occurring with the flavin. The longer lifetime process occurs on a time scale that agrees with transient absorption data published earlier; the magnetic field dependence of the amplitude of this kinetic component is consistent with redox chemistry that involves electron transfer between flavin and tryptophan. We also report a new procedure for the rapid isolation of DNA photolyase.

The reduction potential of the (FADH-/FADH*) couple in DNA photolyase was measured, and the value was found to be significantly higher than the values estimated in the literature. In the absence of substrate, the enzyme has a reduction potential of 16 +/- 6 mV vs NHE. In the presence of excess substrate the reduction potential increases to 81 +/- 8 mV vs NHE. The increase in reduction potential has physiological relevance since it gives the catalytic state greater resistance to oxidation. This is the first measurement of a reduction potential for this class of DNA-repair enzymes and the larger family of blue-light photoreceptors.

Binding of a cis,syn-cyclobutane pyrimidine dimer (CPD) to Escherichia coli DNA photolyase was examined as a function of temperature, enzyme oxidation state, salt, and substrate conformation using isothermal titration calorimetry. While the overall ΔG° of binding was relatively insensitive to most of the conditions examined, the enthalpic and entropic terms that make up the free energy of binding are sensitive to the conditions of the experiment. Substrate binding to DNA photolyase is generally driven by a negative change in enthalpy. Electrostatic interactions and protonation are affected by the oxidation state of the required FAD cofactor and substrate conformation. The fully reduced enzyme appears to bind approximately two additional water molecules as part of substrate binding. More significantly, the experimental change in heat capacity strongly suggests that the CPD lesion must be flipped out of the intrahelical base stacking prior to binding to the protein; the DNA repair enzyme appears to recognize a solvent-exposed CPD as part of its damage recognition mechanism.

No abstract

Escherichia coli DNA photolyase and cryptochrome 1 isolated from Vibrio cholerae, a member of the CRY-DASH family, are directly compared using a variety of experimental methods including UV-vis and Raman spectroscopy, reduction potential measurements, and isothermal titration calorimetry. The semiquinone form of the cryptochrome has an absorption spectrum that is red-shifted from that of the photolyase, but the Raman spectrum indicates that the FAD binding pocket is similar to that of photolyase. The FADH(-)/FADH* reduction potential of the cryptochrome is significantly higher than that of the photolyase at 164 mV vs NHE, but it also increases upon substrate binding (to 195 mV vs NHE), an increase similar to what is observed in photolyase. The FADH(-)/FADH* reduction potential for both proteins was found to be insensitive to ATP binding. Isothermal titration calorimetry found that photolyase binds tighter to substrate (K(A) approximately 10(5) M(-1) for photolyase and approximately 10(4) M(-1) for cryptochrome 1), and the binding constants for both proteins were slightly sensitive to oxidation state. Based upon this work, it appears that this cryptochrome has significant spectroscopic and electrochemical similarities to CPD photolyase. The thermodynamic cycle of the enzymatic repair in the context of this work is discussed.

Transient absorption spectroscopy is used to demonstrate that the electric dipole moment of the substrate cyclobutane thymine dimer affects the charge recombination reaction between fully reduced flavin adenine dinucleotide (FADH-) and the neutral radical tryptophan 306 (Trp306*) in Escherichia coli DNA photolyase. At pH 7.4, the charge recombination is slowed by a factor of 1.75 in the presence of substrate, but not at pH 5.4. Photolyase does bind substrate at pH 5.4, and it seems that this pH effect originates from the conversion of FADH- to FADH2 at lower pH. The free-energy changes calculated from the electric field parameters and from the change in electron transfer rate are in good agreement and support the idea that the substrate electric dipole is responsible for the observed change in electron transfer rate. It is expected that the substrate electric field will also modify the physiologically important from excited 1FADH- to the substrate in the DNA repair reaction.

No abstract