Amino Acid Radicals Research Articles

Tryptophan-Based Radical in the Catalytic Mechanism of Versatile Peroxidase from Bjerkandera adusta

Versatile peroxidase (VP) from Bjerkandera adusta is a structural hybrid between lignin (LiP) and manganese (MnP) peroxidase. This hybrid combines the catalytic properties of the two above peroxidases, being able to oxidize typical LiP and MnP substrates. The catalytic mechanism is that of classical peroxidases, where the substrate oxidation is carried out by a two-electron multistep reaction at the expense of hydrogen peroxide. Elucidation of the structures of intermediates in this process is crucial for understanding the mechanism of substrate oxidation. In this work, the reaction of H(2)O(2) with the enzyme in the absence of substrate has been investigated with electron paramagnetic resonance (EPR) spectroscopy. The results reveal an EPR signal with partially resolved hyperfine structure typical of an organic radical. The yield of this radical is approximately 30%. Progressive microwave power saturation measurements indicate that the radical is weakly coupled to a paramagnetic metal ion, suggesting an amino acid radical in moderate distance from the ferryl heme. A tryptophan radical was identified as a protein-based radical formed during the catalytic mechanism of VP from Bjerkandera adusta through X-band and high-field EPR measurements at 94 GHz, aided by computer simulations for both frequency bands. A close analysis of the theoretical model of the VP from Bjerkandera sp. shows the presence of a tryptophan residue near to the heme prosthetic group, which is solvent-exposed as in the case of LiP and other VPs. The catalytic role of this residue in a long-range electron-transfer pathway is discussed.

EPR properties of a g=2 broad signal trapped in an S 1 state in Ca 2+-depleted Photosystem II

We investigated a new EPR signal that gives a broad line shape around g=2 in Ca 2+-depleted Photosystem (PS) II. The signal was trapped by illumination at 243 K in parallel with the formation of Y Z . The ratio of the intensities between the g=2 broad signal and the Y Z signal was 1:3, assuming a Gaussian line shape for the former. The g=2 broad signal and the Y Z signal decayed together in parallel with the appearance of the S 2 state multiline at 243 K. The g=2 broad signal was assigned to be an intermediate S 1 X state in the transition from the S 1 to the S 2 state, where X represents an amino acid radical nearby manganese cluster, such as D1-His337. The signal is in thermal equilibrium with Y Z . Possible reactions in the S state transitions in Ca 2+-depleted PS II were discussed.

Theg-values and hyperfine coupling of amino acid radicals in proteins: comparison of experimental measurements withab initio calculations

Electron paramagnetic resonance (EPR) spectroscopy has been extensively used to identify and characterize protein-based redox active amino acid radicals based on their g-values and hyperfine couplings. To better understand how these parameters depend on the electronic structure and environment of the radical, the theoretical g-values and proton hyperfine tensors of three models corresponding to the tyrosyl, tryptophanyl and glycyl radicals were calculated using Gaussian 03. The g-values were determined using the B3LYP/6-31+G(D,P) combination of density functional and basis set, while the hyperfine tensors were determined using the B3LYP/EPR-III and PBE0/EPR-III combinations. Comparisons are made to measured values. It was found that by appropriately accounting for hydrogen bonds and the dielectric constant of the environment, good agreement could be achieved between the calculated and measured g-values. In all three cases, the g-anisotropy arose from significant spin density on a nitrogen or oxygen atom. The calculated hyperfine tensors for the three radicals did not differ significantly from previous calculations. In the case of the tyrosyl radical, it is shown for the first time that the para-position substituent that is opposite of the C-O group can break the symmetry of the phenyl ring, leading to different hyperfine tensors for the two large ortho proton couplings. For the tyrosyl and tryptophanyl models, the calculated hyperfine couplings to hydrogen-bonding protons were in very good agreement with measured values for both the tyrosyl and tryptophanyl models.

Spectroscopic characterization of the iron-oxo intermediate in cytochrome P450

From analogy to chloroperoxidase from Caldariomyces fumago, it is believed that the electronic structure of the intermediate iron-oxo species in the catalytic cycle of cytochrome P450 corresponds to an iron(IV) porphyrin-pi-cation radical (compound I). However, our recent studies on P450cam revealed that after 8 ms a tyrosine radical and iron(IV) were formed in the reaction of ferric P450 with external oxidants in the shunt pathway. The present study on the heme domain of P450BM3 (P450BMP) shows a similar result. In addition to a tyrosine radical, a contribution from a tryptophan radical was found in the electron paramagnetic resonance (EPR) spectra of P450BMP. Here we present comparative multi-frequency EPR (9.6, 94 and 285 GHz) and Mössbauer spectroscopic studies on freeze-quenched intermediates produced using peroxy acetic acid as oxidant for both P450 cytochromes. After 8 ms in both systems, amino acid radicals occurred instead of the proposed iron(IV) porphyrin-pi-cation radical, which may be transiently formed on a much faster time scale. These findings are discussed with respect to other heme thiolate proteins. Our studies demonstrate that intramolecular electron transfer from aromatic amino acids is a common feature in these enzymes. The electron transfer quenches the presumably transiently formed porphyrin-pi-cation radical, which makes it extremely difficult to trap compound I.

Multi-frequency EPR and Mössbauer spectroscopic studies on freeze-quenched reaction intermediates of nitric oxide synthase

It is believed by analogy to chloroperoxidase (CPO) from Caldariomyces fumago that the electronic structure of the intermediate iron-oxo species in the catalytic cycle of nitric oxide synthase (NOS) corresponds to an iron(IV) porphyrin-pi -cation radical. Such species can also be produced by the reaction of ferric NOS with external oxidants within the shunt pathway. We present multi-frequency EPR (9.6, 94, 285 GHz) and Mössbauer spectroscopic studies on freeze-quenched intermediates of the oxygenase domain of nitric oxide synthase which has reacted with peroxy acetic acid within 8-200 ms. The intermediates of the oxygenase domain of both the cytokine inducible NOS (iNOSox) and the neuronal NOS (nNOSox) show an organic radical signal in the 9.6-GHz spectrum overlapping with the spectrum of an unknown species with g-values of 2.24, 2.23 and 1.96. Using 94- and 285-GHz EPR the organic radical signal is assigned to a tyrosine radical on the basis of g-values (i.e. Tyr*562 in nNOSox and Tyr*341 in iNOSox). Mössbauer spectroscopy of (57)Fe-labeled unreacted nNOSox shows a ferric low-spin heme-iron (delta = 0.38 mms(-1), deltaE(Q) = 2.58 mms(-1)). The reaction of nNOSox with peroxy acetic acid for 8 ms leads to the disappearance of the magnetic background characteristic for native nNOSox and a new species with delta = 0.27 mms(-1) and deltaE(Q) = 2.41 mms(-1) is detected at 4.2 K which does not resemble the parameters typical for a Fe(IV) center. It is proposed that this intermediate species corresponds to a ferric low-spin species which magnetically couples to an amino acid radical (presumably Trp*409).

Three Different Oxygen-induced Radical Species in Endothelial Nitric-oxide Synthase Oxygenase Domain under Regulation by L-Arginine and Tetrahydrobiopterin

Endothelial nitric-oxide synthase (eNOS) plays important roles in vascular physiology and homeostasis. Whether eNOS catalyzes nitric oxide biosynthesis or the synthesis of reactive oxygen species such as superoxide, hydrogen peroxide, and peroxynitrite is dictated by the bioavailability of tetrahydrobiopterin (BH(4)) and L-arginine during eNOS catalysis. The effect of BH(4) and L-arginine on oxygen-induced radical intermediates has been investigated by single turnover rapid-freeze quench and EPR spectroscopy using the isolated eNOS oxygenase domain (eNOS(ox)). Three distinct radical intermediates corresponding to >50% of the heme were observed during the reaction between ferrous eNOS(ox) and oxygen. BH(4)-free eNOS(ox) produced the superoxide radical very efficiently in the absence of L-arginine. L-Arginine decreased the formation rate of superoxide by an order of magnitude but not its final level or EPR line shape. For BH(4)-containing eNOS(ox), only a stoichiometric amount of BH(4) radical was produced in the presence of L-arginine, but in its absence a new radical was obtained. This new radical could be either a peroxyl radical of BH(4) or an amino acid radical was in the vicinity of the heme. Formation of this new radical is very rapid, >150 s(-1), and it was subsequently converted to a BH(4) radical. The trapping of the superoxide radical by cytochrome c in the reaction of BH(4)(-) eNOS(ox) exhibited a limiting rate of approximately 15 s(-1), the time for the superoxide radical to leave the heme pocket and reach the protein surface; this reveals a general problem of the regular spin-trapping method in determining radical formation kinetics. Cytochrome c failed to trap the new radical species. Together with other EPR characteristics, our data strongly support the conclusion that this new radical is not a superoxide radical or a mixture of superoxide and biopterin radicals. Our study points out distinct roles of BH(4) and L-arginine in regulating eNOS radical intermediates. BH(4) prevented superoxide formation by chemical conversions of the Fe(II)O(2) intermediate, and l-arginine delayed superoxide formation by electronic interaction with the heme-bound oxygen.

Turning on ribonucleotide reductase by light-initiated amino acid radical generation.

Ribonucleotide reductases (RNRs) catalyze the conversion of nucleotides to deoxynucleotides in all organisms, providing the monomeric precursors required for DNA replication and repair. The class I RNRs are composed of two subunits; the R1 subunit contains the active site for nucleotide reduction and allosteric effector binding sites, whereas the R2 subunit houses the essential diirontyrosyl (Y.) radical cofactor. A major unresolved issue is the mechanism by which the tyrosyl radical on R2 (Y122, Escherichia coli numbering) reversibly generates the transient thiyl radical (S.) on R1 that initiates nucleotide reduction. This intersubunit radical initiation is postulated to occur through a defined pathway involving conserved aromatic amino acids (R2: Y122, W48, Y356; R1: Y731, Y730) over a long distance of 35 A. A 20-mer peptide identical to the C-terminal tail of R2 (356-375) and containing Y356 is a competitive inhibitor with respect to R2, and it effectively blocks nucleotide reduction. We now report that a 21-mer peptide, in which a tryptophan has been incorporated at the N terminus of the 20th mer, can replace the R2 subunit and initiate nucleotide reduction by photoinitiated radical generation. The deoxynucleotide generated depends on the presence of allosteric effector and is pathway-dependent. Replacement of Y731 of R2 with phenylalanine prevents deoxynucleotide formation. These results provide direct evidence for the chemical competence of aromatic amino acid radicals and the importance of Y356 in R2 in the radical initiation process of the class I RNRs.

Generation of the R2 Subunit of Ribonucleotide Reductase by Intein Chemistry: Insertion of 3-Nitrotyrosine at Residue 356 as a Probe of the Radical Initiation Process

Escherichia coli ribonucleotide reductase (RNR) catalyzes the conversion of nucleoside diphosphates to deoxynucleoside diphosphates. The enzyme is composed of two subunits: R1 and R2. R1 contains the active site for nucleotide reduction and the allosteric effector sites that regulate the specificity and turnover rate. R2 contains the diferric-tyrosyl (Y(*)) radical cofactor that initiates nucleotide reduction by a putative long-range proton-coupled electron transfer (PCET) pathway over 35 A. This pathway is thought to involve specific amino acid radical intermediates (Y122 to W48 to Y356 within R2 to Y731 to Y730 to C439 within R1). In an effort to study radical initiation, R2 (375 residues) has been synthesized semisynthetically. R2 (residues 1-353), attached to an intein and a chitin binding domain, was constructed, and the protein was expressed (construct 1). This construct was then incubated with Fe(2+) and O(2) to generate the diferric-Y(*) cofactor, and the resulting protein was purified using a chitin affinity column. Incubation of construct 1 with 2-mercaptoethanesulfonic acid (MESNA) resulted in the MESNA thioester of R2 (1-353) (construct 2). A peptide containing residues 354-375 of R2 was generated using solid-phase peptide synthesis where 354, a serine in the wild-type (wt) R2, was replaced by a cysteine. Construct 2 and this peptide were ligated, and the resulting full-length R2 was separated from truncated R2 by anion-exchange chromatography. The purified protein had a specific activity of 350 nmol min(-1) mg(-1), identical to the same protein generated by site-directed mutagenesis when normalized for Y(*). As a first step in studying the radical initiation by PCET, R2 was synthesized with Y356 replaced by 3-nitrotyrosine (NO(2)Y). The protein is inactive (specific activity 1 x 10(-4) that of wt-R2), which permitted a determination of the pK(a) of the NO(2)Y in the R1/R2 complex in the presence of substrate and effectors. Under all conditions, the pK(a) was minimally perturbed. This has important mechanistic implications for the radical initiation process.

Strategic use of amino acid N-substituents to limit α-carbon-centered radical formation and consequent loss of stereochemical integrity

Ab initio calculations have been used to investigate the effect of N-substituents on the stability of α-carbon-centered amino acid radicals. Optimized structures of glycine derivatives and related substituted methanes, and the corresponding radicals, were determined with B3-LYP/6-31G(d). Single-point RMP2/6-31G(d) calculations on these structures were then used to obtain radical stabilization energies, which were compared with the relative rates of formation of the same or closely similar radicals in reactions with N-bromosuccinimde. These studies show that N-acylation and sulfonation decrease both the stability and the ease of formation of the α-carbon-centered radicals. Greater effects are seen with fluoroacyl, fluoroalkylsulfonyl and imido groups. The extent of the effect of the imido and fluoroalkylsulfonyl groups is such that N-phthaloyl- and trifluoromethanesulfonyl-protected amino acids react by hydrogen-atom abstraction from the side chain, thereby avoiding reaction at the chiral α-center and preserving its stereochemical integrity. The origins of these substituent effects are examined.

2,3-difluorotyrosine at position 356 of ribonucleotide reductase R2: a probe of long-range proton-coupled electron transfer.

Escherichia coli class I ribonucleotide reductase catalyzes the conversion of ribonucleotides to deoxyribonucleotides and consists of two subunits: R1 and R2. R1 possesses the active site, while R2 harbors the essential diferric-tyrosyl radical (Y*) cofactor. The Y* on R2 is proposed to generate a transient thiyl radical on R1, 35 A distant, through amino acid radical intermediates. To study the putative long-range proton-coupled electron transfer (PCET), R2 (375 residues) was prepared semisynthetically using intein technology. Y356, a putative intermediate in the pathway, was replaced with 2,3-difluorotyrosine (F2Y, pKa = 7.8). pH rate profiles (pH 6.5-9.0) of wild-type and F2Y-R2 were very similar. Thus, a proton can be lost from the putative PCET pathway without affecting nucleotide reduction. The current model involving H* transfer is thus unlikely.

Radicals with a controlled lifestyle.

JoAnne Stubbe discusses the importance of stable and transient amino acid radicals in primary metabolism.

Influence of the Protein Environment on the Properties of a Tyrosyl Radical in Reaction Centers from Rhodobacter sphaeroides

The influence of the local environment on the formation of a tyrosyl radical was investigated in modified photosynthetic reaction centers from Rhodobacter sphaeroides. The reaction centers contain a tyrosine residue placed approximately 10 A from a highly oxidizing bacteriochlorophyll dimer. Measurements by both optical and electron paramagnetic resonance spectroscopy revealed spectral features that are assigned as arising primarily from an oxidized bacteriochlorophyll dimer at low pH values and from a tyrosyl radical at high pH values, with a well-defined transition that occurred with a pK(a) of 6.9. A model based on the wild-type structure indicated that the Tyr at M164 is likely to form a hydrogen bond with His M193 and to interact weakly with Glu M173. Substitution of Tyr or Glu for His at M193 increased the pK(a) for the transition from 6.9 to 8.9, while substitution of Gln for His M193 resulted in a higher pK(a) value. Substitution of Glu M173 with Gln resulted in loss of the partial formation of the tyrosyl that occurs in the other mutants at low pH values. The results are interpreted in terms of the ability of the residues to act as proton acceptors for the oxidized tyrosine, with the pK(a) values reflecting those of either the putative proton acceptor or the tyrosine, in accord with general models of amino acid radicals.

Repair of amino acid radicals of apolipoprotein B100 of low-density lipoproteins by flavonoids. A pulse radiolysis study with quercetin and rutin.

Selective oxidative damage to apolipoprotein B in LDL can be effected radiolytically by (*)Br(2)(-) radicals. Twenty-seven Trp residues constitute major primary sites of oxidation, but two-thirds of oxidized Trps ((*)Trp) that are formed are repaired by intramolecular electron transfer from Tyr residues with formation of phenoxyl radicals (TyrO(*)). Analysis of (*)Trp and TyrO(*) transient absorbance changes suggests that other apolipoprotein B residues, probably Cys, are oxidized. LDL-bound quercetin can efficiently repair this damage. Absorption studies show that a LDL particle has the capacity to bind approximately 10 quercetin molecules through interaction with apolipoprotein B. The repair occurs by intramolecular electron transfer characterized by a rate constant of 2 x 10(3) s(-)(1). In contrast, rutin, a related flavonoid which does not bind to LDL, cannot repair oxidized apolipoprotein B. Urate is a hydrophilic plasma antioxidant which displays synergistic antioxidant properties with flavonoids. Urate radicals produced by (*)Br(2)(-) can also be repaired by LDL-bound quercetin. This repair occurs with a reaction rate constant of 6.8 x 10(7) M(-)(1) s(-)(1). Comparison with previous studies conducted with human serum albumin-bound quercetin suggests that quercetin analogues tailored to be carried preferentially by lipoproteins might be more powerful plasma antioxidants than natural quercetin carried by serum albumin.

Mechanisms of flavonoid repair reactions with amino acid radicals in models of biological systems: a pulse radiolysis study in micelles and human serum albumin

Neutral tryptophan ( •Trp) and tyrosine (TyrO •) radicals are repaired by certain flavonoids in buffer, in micelles and in human serum albumin (HSA) with corresponding formation of semioxidized flavonoid radicals. In deaerated buffer, •Trp but not TyrO • radicals react with catechin. In micelles, quercetin and rutin repair both •Trp and TyrO • radicals. In addition to amino acid reactivity, microenvironmental factors and nature of the flavonoids govern this repair. Electron transfer efficiencies from quercetin to negatively charged •Trp radicals are 100% in the micellar pseudophases of positively charged cetyltrimethylammonium bromide, (CTAB), and neutral Triton X100 (TX100), but 55% in negatively charged sodium dodecyl sulfate (SDS). In oxygen-saturated CTAB micelles, quercetin also reacts with the superoxide radical anion. When bound to domain IIA of HSA, quercetin repairs, by intra- or intermolecular encounter, less than 20% of oxidative damage to HSA. Quercetin can also repair freely circulating oxidized molecules with repair efficiencies falling to 7% for oxidized Trp, Tyr and α-MSH and to less than 2% for urate radical. This limited effectiveness is attributed both to the inaccessibility of bound quercetin and rutin toward radicals of circulating molecules and to the diffusion-controlled recombination of these radicals.

Recent applications of density functional theory calculations to biomolecules

The purpose of this overview is to highlight the broad scope and utility of current applications of density functional theory (DFT) methods for the study of the properties and reactions of biomolecules. This is illustrated using examples selected from research carried out within our research group and in collaboration with others. The examples include the hyperfine coupling constants of amino acid radicals, the use of an amino acid as a chiral catalyst for the formation of carbon–carbon bonds in the aldol reaction, hydrogen-bond mediated catalysis of an aminolysis reaction, radiation-induced protein–DNA cross-links, and the mechanism by which an antitumor drug cleaves DNA. We demonstrate that DFT-based methods can be applied successfully to a broad range of problems that remain beyond the scope of conventional electron-correlation methods. Furthermore, we show that contemporary computational quantum chemistry complements experiment in the study of biological systems.

Mechanisms of flavonoid repair reactions with amino acid radicals in models of biological systems: a pulse radiolysis study in micelles and human serum albumin

Neutral tryptophan ( •Trp) and tyrosine (TyrO •) radicals are repaired by certain flavonoids in buffer, in micelles and in human serum albumin (HSA) with corresponding formation of semioxidized flavonoid radicals. In deaerated buffer, •Trp but not TyrO • radicals react with catechin. In micelles, quercetin and rutin repair both •Trp and TyrO • radicals. In addition to amino acid reactivity, microenvironmental factors and nature of the flavonoids govern this repair. Electron transfer efficiencies from quercetin to negatively charged •Trp radicals are 100% in the micellar pseudophases of positively charged cetyltrimethylammonium bromide, (CTAB), and neutral Triton X100 (TX100), but 55% in negatively charged sodium dodecyl sulfate (SDS). In oxygen-saturated CTAB micelles, quercetin also reacts with the superoxide radical anion. When bound to domain IIA of HSA, quercetin repairs, by intra- or intermolecular encounter, less than 20% of oxidative damage to HSA. Quercetin can also repair freely circulating oxidized molecules with repair efficiencies falling to 7% for oxidized •Trp, Tyr and α-MSH and to less than 2% for urate radical. This limited effectiveness is attributed both to the inaccessibility of bound quercetin and rutin toward radicals of circulating molecules and to the diffusion-controlled recombination of these radicals.

The peroxidase activity of cytochrome c‐550 from Paracoccus versutus

Next to their natural electron transport capacities, c-type cytochromes possess low peroxidase and cytochrome P-450 activities in the presence of hydrogen peroxide. These catalytic properties, in combination with their structural robustness and covalently bound cofactor make cytochromes c potentially useful peroxidase mimics. This study reports on the peroxidase activity of cytochrome c-550 from Paracoccus versutus and the loss of this activity in presence of H2O2. The rate-determining step in the peroxidase reaction of cytochrome c-550 is the formation of a reactive intermediate, following binding of peroxide to the haem iron. The reaction rate is very low compared to horseradish peroxidase (approximately one millionth), because of the poor accessibility of the haem iron for H2O2, and the lack of a base catalyst such as the distal His of the peroxidases. This is corroborated by the linear dependence of the reaction rate on the peroxide concentration up to at least 1 M H2O2. Steady-state conversion of a reducing substrate, guaiacol, is preceded by an activation phase, which is ascribed to the build-up of amino-acid radicals on the protein. The inactivation kinetics in the absence of reducing substrate are mono-exponential and shown to be concurrent with haem degradation up to 25 mM H2O2 (pH 8.0). At still higher peroxide concentrations, inactivation kinetics are biphasic, as a result of a remarkable protective effect of H2O2, involving the formation of superoxide and ferrocytochrome c-550.

EPR study of anion radicals of various N-quinonyl amino acids.

Since peptide quinones possess great clinical potential in targeted chemotherapy, several series of novel N-quinonyl amino acids have been synthesized and their first products of reduction were studied by EPR spectroscopy. EPR spectra of the corresponding radical adducts were identified by computer simulation. The dependence between the splitting constants and the chemical structure of the N-quinonyl amino acids anion radicals was examined.

Photochemical Protein Scissors: Role of Aromatic Residues on the Binding Affinity and Photocleavage Efficiency of Pyrenyl Peptides

Photocleavage of proteins by a series of organic probe molecules is examined as a function of probe structure. For example, pyrenyl peptides Py-Gly-X (X=Trp, Tyr, Phe, and His, and Py=4(1-pyrenyl)butyroyl)) are prepared, and their protein binding/photocleavage properties have been examined. The binding constants with bovine serum albumin (BSA) are in the range of 107 to 105 and binding of the probes to the proteins is evident in absorption, fluorescence and circular dichroism experiments. While the fluorescence of Py-Gly-Tyr increases upon binding to BSA, quenching is observed with Py-Gly-Trp or Py-Gly-His. While hyperchromism is observed with Py-Gly-Trp/BSA, hypochromism is the norm for all the other probes with this protein. Binding of all the probes to BSA or lysozyme resulted in major changes in the circular dichroism spectra of the probes. Photoexcitation of the probe/protein complexes, in the presence of an electron acceptor, resulted in protein photocleavage. The phenylalanine and histidine analogs resulted in photocleavage of both BSA and lysozyme while the tyrosine and tryptophan analogs did not yield any fragmentation with either of the two proteins. The photocleavage sites are similar to those reported for Py-Phe. Flash photolysis studies of the probe/protein mixtures indicate that the initially produced pyrene cation radical is strongly quenched by the tyrosine and tryptophan residues and resulted in the corresponding amino acid radicals. Strong modulation of the photoreactivities of the probes by specific residues of the probe provide insight in learning how the photocleavage efficiencies of the probes can be improved in future studies.

Intraprotein radical transfer during photoactivation of DNA photolyase.

Amino-acid radicals play key roles in many enzymatic reactions. Catalysis often involves transfer of a radical character within the protein, as in class I ribonucleotide reductase where radical transfer occurs over 35 A, from a tyrosyl radical to a cysteine. It is currently debated whether this kind of long-range transfer occurs by electron transfer, followed by proton release to create a neutral radical, or by H-atom transfer, that is, simultaneous transfer of electrons and protons. The latter mechanism avoids the energetic cost of charge formation in the low dielectric protein, but it is less robust to structural changes than is electron transfer. Available experimental data do not clearly discriminate between these proposals. We have studied the mechanism of photoactivation (light-induced reduction of the flavin adenine dinucleotide cofactor) of Escherichia coli DNA photolyase using time-resolved absorption spectroscopy. Here we show that the excited flavin adenine dinucleotide radical abstracts an electron from a nearby tryptophan in 30 ps. After subsequent electron transfer along a chain of three tryptophans, the most remote tryptophan (as a cation radical) releases a proton to the solvent in about 300 ns, showing that electron transfer occurs before proton dissociation. A similar process may take place in photolyase-like blue-light receptors.