Electron-transfer oxidation properties of DNA bases and DNA oligomers

Kinetics for the thermal and photoinduced electron-transfer oxidation of a series of DNA bases with various oxidants having the known one-electron reduction potentials (E(red)) in an aqueous solution at 298 K were examined, and the resulting electron-transfer rate constants (k(et)) were evaluated in light of the free energy relationship of electron transfer to determine the one-electron oxidation potentials (E(ox)) of DNA bases and the intrinsic barrier of the electron transfer. Although the E(ox) value of GMP at pH 7 is the lowest (1.07 V vs SCE) among the four DNA bases, the highest E(ox) value (CMP) is only 0.19 V higher than that of GMP. The selective oxidation of GMP in the thermal electron-transfer oxidation of GMP results from a significant decrease in the pH dependent oxidation potential due to the deprotonation of GMP*+. The one-electron reduced species of the photosensitizer produced by photoinduced electron transfer are observed as the transient absorption spectra when the free energy change of electron transfer is negative. The rate constants of electron-transfer oxidation of the guanine moieties in DNA oligomers with Fe(bpy)3(3+) and Ru(bpy)3(3+) were also determined using DNA oligomers containing different guanine (G) sequences from 1 to 10 G. The rate constants of electron-transfer oxidation of the guanine moieties in single- and double-stranded DNA oligomers with Fe(bpy)3(2+) and Ru(bpy)3(3+) are dependent on the number of sequential guanine molecules as well as on pH.     

DNA vibrational coupling revealed with two-dimensional infrared spectroscopy: insight into why vibrational spectroscopy is sensitive to DNA structure

Two-dimensional infrared (2D IR) spectroscopy was used to study the carbonyl vibrational modes of guanine and cytosine bases in A- and B-form DNA. Located between 1600 and 1700 cm(-1), these modes are often used to monitor DNA secondary structure with traditional infrared spectroscopies such as FTIR, but traditional spectroscopies lack the necessary observables to unravel the coupling mechanisms that make these modes sensitive to secondary structure. By using 2D IR spectroscopy and electronic structure calculations on d(G(5)C(5)) and d(GC)(8) model nucleic acids, we find that hydrogen-bonded guanine/cytosine base pairs are primarily electrostatically coupled and that the coupling between these modes can be modeled with a transition dipole density approach. In comparison, electrostatics is insufficient to model stacked bases because of cooperative charge-sharing effects, but the coupling can be accurately calculated using a finite difference method. We find that the coupling is very strong for both hydrogen-bonded and stacked base geometries, creating vibrational modes that extend both across the base pairs and along the lengths of the helices. Our results provide a physical basis for understanding how strong coupling gives rise to the empirically established relationship between infrared spectroscopy and DNA/RNA secondary structure.     

Site-selective electron transfer from purines to electrocatalysts: voltammetric detection of a biologically relevant deletion in hybridized DNA duplexes.

BACKGROUND: The one-electron oxidation of guanine nucleobases is of interest for understanding the mechanisms of mutagenesis, probing electron-transfer reactions in DNA, and developing sensing schemes for nucleic acids. The electron-transfer rates for oxidation of guanine by exogenous redox catalysts depend on the base paired to the guanine. An important goal in developing the mismatch sensitivity is to identify a means for monitoring the current resulting from electron transfer at a single base in the presence of native oligonucleotides that contain all four bases. RESULTS: The nucleobase 8-oxo-guanine (8G) is selectively oxidized by the redox catalyst Os(bpy)(3)(3+/2+) (bpy = 2,2'-bipyridine) in the presence of native guanine. Cyclic voltammograms of Os(bpy)(3)(2+) show current enhancements indicative of nucleobase oxidation upon addition of oligonucleotides that contain 8G, but not in the presence of native guanine. As expected, similar experiments with Ru(bpy)(3)(2+) show enhancement with both guanine and 8G. The current enhancements for the 8G/Os(III) reaction increase in the order 8G-C approximately 8G.T < 8G.G < 8G.A < 8G, the same order as that observed for guanine/Ru(III). This site-selective mismatch sensitivity can be applied to detection of a TTT deletion, which is important in cystic fibrosis. CONCLUSIONS: The base 8G can be effectively used in conjunction with a low-potential redox catalyst as a probe for selective electron transfer at a single site. Because of the high selectivity for 8G, rate constants can be obtained that reflect the oxidation of only one base. The mismatch sensitivity can be used to detect biologically relevant abnormalities in DNA.     

Involvement of proton transfer in the reductive repair of DNA guanyl radicals by aniline derivatives

The most easily oxidized sites in DNA are the guanine bases, and major intermediates produced by the direct effect of ionizing radiation (ionization of the DNA itself) are electron deficient guanine species. By means of a radiation chemical method (gamma-irradiation of aqueous thiocyanate), we are able to produce these guanyl radicals in dilute aqueous solutions of plasmid DNA where the direct effect would otherwise be negligible. Stable modified guanine products are formed from these radicals. They can be detected in the plasmid conversion to strand breaks after a post-irradiation incubation with a DNA base excision endonuclease enzyme. If aniline compounds are also present, the yield of modified guanines is strongly attenuated. The mechanism responsible for this effect is electron donation from the aniline compound to the guanyl radical, and it is possible to derive rate constants for this reaction. Aniline compounds bearing electron withdrawing groups (e.g., 4-CF3) were found to be less reactive than those bearing electron donating groups (e.g., 4-CH3). At physiological pH values, the reduction of a guanyl radical involves the transfer of a proton as well as of an electron. The mild dependence of the rate constant on the driving force suggests that the electron is not transferred before the proton. Although the source of the proton is unclear, our observations emphasize the importance of an accompanying proton transfer in the reductive repair of oxidative damage to guanine bases which are located in a biologically active double stranded plasmid DNA substrate.     

Direct chemical evidence for charge transfer between photoexcited 2-aminopurine and guanine in duplex DNA

Photoexcited 2-aminopurine (Ap*) is extensively exploited as a fluorescent base analogue in the study of DNA structure and dynamics. Quenching of Ap* in DNA is often attributed to stacking interactions between Ap* and DNA bases, despite compelling evidence indicating that charge transfer (CT) between Ap* and DNA bases contributes to quenching. Here we present direct chemical evidence that Ap* undergoes CT with guanine residues in duplex DNA, generating oxidative damage at a distance. Irradiation of Ap in DNA containing the modified guanine, cyclopropylguanosine (CPG), initiates hole transfer from Ap* followed by rapid ring opening of the CPG radical cation. Ring opening accelerates hole trapping to a much shorter time regime than for guanine radicals in DNA; consequently, trapping effectively competes with back electron transfer (BET) leading to permanent CT chemistry. Significantly, BET remains competitive, even with this much faster trapping reaction, consistent with measured kinetics of DNA-mediated CT. The distance dependence of BET is sharper than that of forward CT, leading to an inverted dependence of product yield on distance; at short distances product yield is inhibited by BET, while at longer distances trapping dominates, leading to permanent products. The distance dependence of product yield is distinct from forward CT, or charge injection. As with photoinduced charge transfer in other chemical and biological systems, rapid kinetics for charge injection into DNA need not be associated with a high yield of DNA damage products.     

Acidity and proton affinity of hypoxanthine in the gas phase versus in solution: intrinsic reactivity and biological implications

Hypoxanthine is a mutagenic purine base that most commonly arises from the oxidative deamination of adenine. Damaged bases such as hypoxanthine are associated with carcinogenesis and cell death. This inevitable damage is counteracted by glycosylase enzymes, which cleave damaged bases from DNA. Alkyladenine DNA glycosylase (AAG) is the enzyme responsible for excising hypoxanthine from DNA in humans. In an effort to understand the intrinsic properties of hypoxanthine, we examined the gas-phase acidity and proton affinity using quantum mechanical calculations and gas-phase mass spectrometric experimental methods. In this work, we establish that the most acidic site of hypoxanthine has a gas-phase acidity of 332 +/- 2 kcal mol-1, which is more acidic than hydrochloric acid. We also bracket a less acidic site of hypoxanthine at 368 +/- 3 kcal mol-1. We measure the proton affinity of the most basic site of hypoxanthine to be 222 +/- 3 kcal mol-1. DFT calculations of these values are consistent with the experimental data. We also use calculations to compare the acidic and basic properties of hypoxanthine with those of the normal bases adenine and guanine. We find that the N9-H of hypoxanthine is more acidic than that of adenine and guanine, pointing to a way that AAG could discriminate damaged bases from normal bases. We hypothesize that AAG may cleave certain damaged nucleobases as anions and that the active site may take advantage of a nonpolar environment to favor deprotonated hypoxanthine as a leaving group versus deprotonated adenine or guanine. We also show that an alternate mechanism involving preprotonation of hypoxanthine is energetically less attractive, because the proton affinity of hypoxanthine is less than that of adenine and guanine. Last, we compare the acidity in the gas phase versus that in solution and find that a nonpolar environment enhances the differences in acidity among hypoxanthine, adenine, and guanine.     

Structure reactivity relationship in the reaction of DNA guanyl radicals with hydroxybenzoates

In DNA, guanine bases are the sites from which electrons are most easily removed. As a result of hole migration to this stable location on guanine, guanyl radicals are major intermediates in DNA damage produced by the direct effect of ionizing radiation (ionization of the DNA itself and not through the intermediacy of water radicals). We have modeled this process by employing gamma irradiation in the presence of thiocyanate ions, a method which also produces single electron oxidized guanyl radicals in plasmid DNA in aqueous solution. The stable products formed in DNA from these radicals are detected as strand breaks after incubation with the FPG protein. When a phenolic compound is present in the solution during gamma irradiation, the formation of guanyl radical species is decreased by electron donation from the phenol to the guanyl radical. We have quantified the rate of this reaction for four different phenolic compounds bearing carboxylate substituents as proton acceptors. A comparison of the rates of these reactions with the redox strengths of the phenolic compounds reveals that salicylate reacts ca. 10-fold faster than its structural analogs. This observation is consistent with a reaction mechanism involving a proton coupled electron transfer, because intra-molecular transfer of a proton from the phenolic hydroxyl group to the carboxylate group is possible only in salicylate, and is favored by the strong 6-membered ring intra-molecular hydrogen bond in this compound.     

Determination of redox potentials for the Watson-Crick base pairs, DNA nucleosides, and relevant nucleoside analogues

Redox potentials for the DNA nucleobases and nucleosides, various relevant nucleoside analogues, Watson-Crick base pairs, and seven organic dyes are presented based on DFT/B3LYP/6-31++G(d,p) and B3YLP/6-311+G(2df,p)//B3LYP/6-31+G* levels of calculations. The values are determined from an experimentally calibrated set of equations that correlate the vertical ionization (electron affinity) energy of 20 organic molecules with their experimental reversible oxidation (reduction) potential. Our results are in good agreement with those estimated experimentally for the DNA nucleosides in acetonitrile solutions (Seidel et al. J. Phys. Chem. 1996, 100, 5541). We have found that nucleosides with anti conformation exhibit lower oxidation potentials than the corresponding syn conformers. The lowering in the oxidation potential is due to the formation of an intramolecular hydrogen bonding interaction between the 5'-OH group of the sugar and the N3 of the purine bases or C2=O of the pyrimidine bases in the syn conformation. Pairing of adenine or guanine with its complementary pyrimidine base decreases its oxidation potential by 0.15 or 0.28 V, respectively. The calculated energy difference between the oxidation potential for the G.C base pair and that of the guanine base is in good agreement with the experimental value estimated recently (0.34 V: Caruso, T.; et al. J. Am. Chem. Soc. 2005, 127, 15040). The complete and consistent set of reversible redox values determined in this work for the DNA constituents is expected to be of considerable value to those studying charge and electronic energy transfer in DNA.     

Base sequence effects on transport in DNA

Given the success of the polaron model based on solvation in accounting for the width of a hole polaron on an all-adenine (A) sequence on DNA, we extend the calculations to other sequences. We find excellent agreement with the free energy differences measured by Lewis et al. (J. Am. Chem. Soc. 2000, 122, 12037-12038) between a guanine (G) cation and a pair of bases, GG, or a triple of bases, GGG, in all cases surrounded by As, by treating AGGA and AGGGA as solvated polarons. There is additional support for hole polaron formation in DNA from experiments in which oxidative damage due to injected holes is investigated in sequences involving Gs and As. Theory and comparison with transport measurements on repeated sequences involving multiple thymines (Ts) or combinations such as ATs or GCs, where C is cytosine, led to the suggestion that the basic sequences in these cases must be polarons whose wave functions have substantial amplitudes on both chains in a duplex. The size of an electron polaron in DNA is predicted to be similar to that of a hole polaron, approximately 4 or 5 bases. Although experiments have shown that polaron hopping is the dominant mode of charge transport in DNA with repeated sequences such as AGGA, further investigations, particularly of temperature dependence of site energies and transfer integrals, are needed to determine to what extent hole transport takes place by polaron hopping for arbitrary DNA sequences.     

The acidity and proton affinity of the damaged base 1,N6-ethenoadenine in the gas phase versus in solution: intrinsic reactivity and biological implications

1,N(6)-ethenoadenine (epsilonA) is a highly mutagenic lesion that is excised from human DNA by the enzyme alkyladenine DNA glycosylase (AAG). In an effort to understand the intrinsic properties of 1,N(6)-ethenoadenine, we examined its gas phase acidity and proton affinity using quantum mechanical calculations and mass spectrometric experimental methods. We measure two acidities for epsilonA: a more acidic site (DeltaH(acid) = 332 kcal mol(-1); DeltaG(acid) = 325 kcal mol(-1)) and a less acidic site (DeltaH(acid) = 367 kcal mol(-1); DeltaG(acid) = 360 kcal mol(-1)). We also find that the proton affinity of the most basic site of 1,N(6)-ethenoadenine is 232-233 kcal mol(-1) (GB = 224 kcal mol(-1)). These measurements, when compared to calculations, establish that, under our experimental conditions, we have only the canonical tautomer of 1,N(6)-ethenoadenine present. We also compare the gas phase acidic properties of epsilonA with that of the normal bases adenine and guanine and find that epsilonA is the most acidic. This supports the theory that AAG and other related enzymes may cleave damaged bases as anions. Furthermore, comparison of the gas phase and aqueous acidities indicates that the nonpolar environment of the enzyme enhances the acidity differences of epsilonA versus adenine and guanine.     

Theoretical determination of one-electron redox potentials for DNA bases, base pairs, and stacks

Electron affinities, ionization potentials, and redox potentials for DNA bases, base pairs, and N-methylated derivatives are computed at the DFT/M06-2X/6-31++G(d,p) level of theory. Redox properties of a guanine-guanine stack model are explored as well. Reduction and oxidation potentials are in good agreement with the experimental ones. Electron affinities of base pairs were found to be negative. Methylation of canonical bases affects the ionization potentials the most. Base pair formation and base stacking lower ionization potentials by 0.3 eV. Pairing of guanine with the 5-methylcytosine does not seem to influence the redox properties of this base pair much.     

Mechanisms of oxidation of guanine in DNA by carbonate radical anion, a decomposition product of nitrosoperoxycarbonate

Peroxynitrite is produced during inflammation and combines rapidly with carbon dioxide to yield the unstable nitrosoperoxycarbonate, which decomposes (in part) to CO(3) (.-) and (.)NO(2) radicals. The CO(3) (.-) radicals oxidize guanine bases in DNA through a one-electron transfer reaction process that ultimately results in the formation of stable guanine oxidation products. Here we have explored these mechanisms, starting with a spectroscopic study of the kinetics of electron transfer from 20-22mer double-stranded oligonucleotides to CO(3) (.-) radicals, together with the effects of base sequence on the formation of the end-products in runs of one, two, or three contiguous guanines. The distributions of these alkali-labile lesions were determined by gel electrophoresis methods. The cascade of events was initiated through the use of 308 nm XeCl excimer laser pulses to generate CO(3) (.-) radicals by an established method based on the photodissociation of persulfate to sulfate radicals and the oxidation of bicarbonate. Although the Saito model (Saito et al., J. Am. Chem. Soc. 1995, 117, 6406-6407) predicts relative ease of one-electron oxidations in DNA, following the trend 5'-GGG > 5'-GG > 5'-G, we found that the rate constants for CO(3) (.-)-mediated oxidation of guanines in these sequence contexts (k(5)) showed only small variation within a narrow range [(1.5-3.0)x10(7) M(-1) s(-1)]. In contrast, the distributions of the end-products are dependent on the base sequence context and are higher at the 5'-G in 5'-GG sequences and at the first two 5'-guanines in the 5'-GGG sequences. These effects are attributed to a combination of initial hole distributions among the contiguous guanines and the subsequent differences in chemical reaction yields at each guanine. The lack of dependence of k(5) on sequence context indicates that the one-electron oxidation of guanine in DNA by CO(3) (.-) radicals occurs by an inner-sphere mechanism.     

Theoretical study of excitation energy transfer in DNA photolyase

Photolyase (PL) is a DNA repair enzyme which splits UV light-induced thymine dimers on DNA by an electron transfer reaction occurring between the photoactivated FADH(-) cofactor and the DNA dimer in the DNA/PL complex. The crystal structure of the DNA/photolyase complex from Anacystis nidulans has been solved. Here, using the experimental crystal structure, we re-examine the details of the repair electron transfer reaction and address the question of energy transfer from the antenna HDF to the redox active FADH(-) cofactor. The photoactivation of FADH(-) immediately preceding the electron transfer is a key step in the repair mechanism that is largely left unexamined theoretically. An important butterfly thermal motion of flavin is identified in ab initio calculations; we propose its role in the back electron transfer from DNA to photolyase. Molecular dynamics simulation of the whole protein/DNA complex is carried out to obtain relevant cofactor conformations for ZINDO/S spectroscopic absorption and fluorescence calculations. We find that significant thermal broadening of the spectral lines, due to protein dynamics, as well as the alignment of the donor HDF and the acceptor FADH(-) transition dipole moments both contribute to the efficiency of energy transfer. The geometric factor of F?rster's dipolar coupling is calculated to be 1.82, a large increase from the experimentally estimated 0.67. Using F?rster's mechanism, we find that the energy transfer occurs with remarkable efficiency, comparable with the experimentally determined value of 98%.     

Electron Transfer through DNA in the Course of Radical-Induced Strand Cleavage

No benefit from base stacking is observed for rates of electron transfer in DNA. This conclusion was drawn from experiments with a new DNA assay in which a radical cationic site, generated by strand cleavage, can be reduced by the guanine bases in the same DNA (the electron transfer is indicated by arrows in the diagram). The distance dependence of this electron transfer step is determined by the chemical yield of the reduction product.     

Studies on the mechanism of the photo-induced DNA damage in the presence of acridizinium salts-involvement of singlet oxygen and an unusual source for hydroxyl radicals

Mechanistic investigations of the photoinduced DNA damage by acridizinium salts (4a-azonia-anthracene derivatives) are presented. Irradiation of 9-bromoacridizinium in the presence of defined double- and single-stranded DNA oligomers under aerobic conditions leads to both frank strand breaks and alkali-labile sites as determined by polyacrylamide gel electrophoresis (PAGE). The extent of the DNA damage increases significantly in D(2)O and occurs selectively at guanosine residues. These observations reveal the formation of singlet oxygen ((1)O(2)) as reactive species, which oxidizes the DNA bases, above all the guanine bases. Further evidence for (1)O(2) formation was obtained from laser-flash spectroscopic investigations, which show intersystem crossing (S(1) to T(1)) of the excited states of the parent acridizinium and of the 9-bromo- and 9-amino-substituted derivatives. The resulting triplet state is efficiently quenched by oxygen (k(q) > 10(9) s(-)(1)M(-)(1)) to yield (1)O(2). Under anaerobic conditions, no significant alkali-labile lesions are observed, but frank strand breaks are induced; however, to lesser extent than under aerobic conditions. The DNA damage is suppressed in the presence of a radical scavenger, namely t-BuOH, and hydroxyl radicals are shown to be the reactive intermediates by trapping experiments with terephthalic acid. Moreover, the intercalated acridizinium molecules are not involved in the DNA damage reactions. The intercalated acridizinium salt leads to a primary PET reaction with the DNA bases; however, a fast BET transfer is proposed that regains the dye and the DNA, so that the excited intercalated dye does not contribute significantly to the overall DNA damage.     

Electronic structure contributions to electron-transfer reactivity in iron-sulfur active sites: 3. Kinetics of electron transfer

The kinetics of electron transfer for rubredoxins are examined using density functional methods to determine the electronic structure characteristics that influence and allow for fast electron self-exchange in these electron-transport proteins. Potential energy surfaces for [FeX(4)](2-,1-) models confirm that the inner-sphere reorganization energy is inherently small for tetrathiolates ( approximately 0.1 eV), as evidenced by the only small changes in the equilibrium Fe-S bond distance during redox (Deltar(redox) approximately 0.05 A). It is concluded that electronic relaxation and covalency in the reduced state allow for this small in this case relative to other redox couples, such as the tetrachloride. Using a large computational model to include the protein medium surrounding the [Fe(SCys)(4)](2-,1-) active site in Desulfovibrio vulgaris Rubredoxin, the electronic coupling matrix element for electron self-exchange is defined for direct active-site contact (H0(DA)). Simple Beratan-Onuchic model is used to extend coupling over the complete surface of the protein to provide an understanding of probable electron-transfer pathways. Regions of similar coupling properties are grouped together to define a surface coupling map, which reveals that very efficient self-exchange occurs only within 4 sigma-bonds of the active site. Longer-range electron transfer cannot support the fast rates of electron self-exchange observed experimentally. Pathways directly through the two surface cysteinate ligands dominate, but surface-accessible amides hydrogen-bonded to the cysteinates also contribute significantly to the rate of electron self-exchange.     

Reactivity of DNA guanyl radicals with phenolate anions

Guanine bases are the most easily oxidized sites in DNA. Electron-deficient guanine species are major intermediates produced in DNA by the direct effect of ionizing radiation (ionization of the DNA itself) because of preferential hole migration within DNA to guanine bases. By using thiocyanate ions to modify the indirect effect (ionization of the solvent), we are able to produce these single-electron-oxidized guanine radical species in dilute aqueous solutions of plasmid DNA where the direct effect is negligible. The guanyl radical species produce stable modified guanine products. They can be detected in the plasmid by converting them to strand breaks after incubation with a DNA repair enzyme. If a phenol is present during irradiation, the yield of modified guanines is decreased. The mechanism is reduction of the guanine radical species by the phenol. It is possible to derive a rate constant for the reaction of the phenol with the guanyl radical. The pH dependence shows that phenolate anions are more reactive than their conjugate acids, although the difference for guanyl radicals is smaller than with other single-electron-oxidizing agents. At physiological pH values, the reduction of a guanyl radical entails the transfer of a proton in addition to the electron. The relatively small dependence of the rate constant on the driving force implies that the electron cannot be transferred before the proton. These results emphasize the potential importance of acidic tyrosine residues and the intimate involvement of protons in DNA repair.     

Role of the electronic properties of azurin active site in the electron‐transfer process

Electron transfer proteins, such as azurin (a blue copper protein), are promising candidates for the implementation of biomolecular nanoelectronic devices. To understand the details of electron transfer in redox active azurin molecules, we performed plane‐wave pseudo‐potential density functional theory (DFT) calculations of the protein active site in the two possible oxidation states Cu(I) and Cu(II). The ab initio results are used to discuss how the electronic spectrum and wavefunctions may mediate the shuttling of electrons through the copper ion. We find that the Cu‐ligand hybridization is very similar in the two charge states of the metal center, but the energy spectrum changes substantially. This result might indicate important effects of electronic correlations in the redox activity and consequent electron transfer through the Cu site. © 2004 Wiley Periodicals, Inc. Int J Quantum Chem, 2005     

Ultrafast dynamics and anionic active states of the flavin cofactor in cryptochrome and photolyase

We report here our systematic studies of the dynamics of four redox states of the flavin cofactor in both photolyases and insect type 1 cryptochromes. With femtosecond resolution, we observed ultrafast photoreduction of oxidized state flavin adenine dinucleotide (FAD) in subpicosecond and of neutral radical semiquinone (FADH(*)) in tens of picoseconds through intraprotein electron transfer mainly with a neighboring conserved tryptophan triad. Such ultrafast dynamics make these forms of flavin unlikely to be the functional states of the photolyase/cryptochrome family. In contrast, we find that upon excitation the anionic semiquinone (FAD(*-)) and hydroquinone (FADH(-)) have longer lifetimes that are compatible with high-efficiency intermolecular electron transfer reactions. In photolyases, the excited active state (FADH(-)*) has a long (nanosecond) lifetime optimal for DNA-repair function. In insect type 1 cryptochromes known to be blue-light photoreceptors the excited active form (FAD(*-)*) has complex deactivation dynamics on the time scale from a few to hundreds of picoseconds, which is believed to occur through conical intersection(s) with a flexible bending motion to modulate the functional channel. These unique properties of anionic flavins suggest a universal mechanism of electron transfer for the initial functional steps of the photolyase/cryptochrome blue-light photoreceptor family.