Proton-coupled electron transfer (PCET) reactions have received much attention over the past 10 years, from an experimental as well as from a theoretical point of view. At the heart of many chemical and biological processes, such reactions are of particular interest in energy conversion and enzymatic processes. Among the numerous examples of PCET reactions, photosynthesis and particularly reactions inside the Photosystem II (PSII) subunit, involving a global four electrons and four protons process to perform water oxidation and respiration, is the most emblematic one. This review focuses on the photochemical approaches of PCET reactions involving phenolic molecules. Indeed, a significant part of photochemical PCET studies were conducted on tyrosine or phenol relevant to PSII and charge transport in enzymes. The mechanisms of these reactions, sequential or concerted, with particular emphasis on the influence of pH, temperature, solvent nature and H-bonding pattern are presented based on photochemical techniques and related theoretical analysis.
The coupling of proton and electron transfer is at the core of numerous natural and artificial processes in chemistry and biology (1–7). With the exponentially increasing interest for energy conversion, depollution processes and design of new and performing catalytic molecular machineries, oxidation–reduction reactions that involve electron and proton transfers require a better understanding and thorough description. Indeed, unlike hydrogen atom transfers (HAT), proton-coupled electron transfer (PCET) reactions involve different centers for the electron and the proton transfers, located, or not, on the same molecule and being in the same direction, or not.
PCET reactions can occur through stepwise mechanisms, with the electron being transferred first, followed by the proton transfer (electron–proton transfer, EPT, Fig. 1). A symmetric reaction pathway is given by the transfer of the proton first, followed by the electron transfer (proton–electron transfer, PET). These two stepwise mechanisms are thermodynamically unfavorable due to the formation of high-energy intermediates, a radical ion or high reducing or oxidizing species, respectively (8). Another possibility is a concerted pathway, in which the electron and the proton are transferred in the same elementary step (concerted proton–electron transfer, CPET). This concerted pathway has the advantage of bypassing the high-energy intermediates, but the counterpart is that this apparent energetic gain is usually counterbalanced by a kinetic cost.
The goal of actual PCET studies is two-fold. First, analysis of experimental results is faced with the challenge to distinguish between those three pathways, which usually compete with one another. The different parameters (driving force of the reaction, hydrogen bonding or reorganization energy upon reaction) that control the mechanism have to be identified in order to further establish quantitative models describing PCET reactions. Second, in the actual growing wave of renewable energy, the study of biomimicking PCET reactions at the molecular scale aim at designing molecular devices able to efficiently convert energy or undertake renewable and green chemistry (9–12).
Several experimental methods are employed to study PCET reactions, mainly electrochemistry, e.g. by cyclic voltammetry, and photochemistry, most often using flash-photolysis methods coupled to transient absorption and/or EPR. The present review will focus on the latter through fundamental examples all containing phenol-type molecules. We present first the oxidation of tyrosine in Photosystem II (PSII) where water splitting occurs, then the description of PCET in water for the model molecule of phenol is detailed and finally, PCET examples in biological systems such as peptides, enzymes and DNA are discussed.
Several theoretical descriptions of PCET mechanisms were progressively developed and refined thanks to experimental data, mainly by Cukier and coworkers, Hammes-Schiffer and coworkers and Savéant and coworkers. A selection of articles and reviews is given for the benefit of interested readers (1,2,5,6,13–22). A simplified model has been developed for concerted proton-coupled electron transfer (CPET) based on a double adiabatic approximation with proton and electron being treated as light particles, so that the transition state is obtained by the reorganization of the solvent molecules and heavy reactant atoms (Fig. 2, ). At the transition state, both reactants and products have the same configuration, and proton transfer is electronically adiabatic since the system may be described by two electronic states, each of them resulting from a pair of proton diabatic states (Fig. 2). The potential energy surfaces of diabatic states may be approximated using harmonic approximation, like in the Marcus-Hush model for outersphere electron transfer, thus leading to a quadratic term for the activation free energy:
ΔG0 is the standard free energy, λ is the total reorganization energy for the reaction (involving reorganization of both the reactant and the surrounding solvent molecules) and ΔZPE is the difference between the transferring proton zero point energies at the transition state and at the reactant state. The rate constant of the CPET reaction is then expressed by , where the pre-exponential factor Z is a combined measure of the formation of the precursor complexes over a range of significant reacting distances and of the efficiency of proton tunneling.
Phenolic compounds are worth investigating for several reasons. In addition to PSII (24–26) and mimicking strategies developed for energy issues, the tyrosyl radical appears to be the most prominent amino acid radical involved in enzymatic PCET mechanisms, such as those encountered in ribonucleotide reductase (RNR) (27–30), prostaglandin H synthase (31–34), galactose oxidase (35–37), adenosylcobalamin-dependent enzymes (38,39), cytochrome c oxidase (CcO) (40–42) or amine synthase (43). It is indeed now known or in some cases suspected that these radicals play a crucial role in DNA damage and repair (see for example [44–46]). The interest for phenolic compounds is also related to their antioxidant properties (47–50).
The Photosystem II
Numbers of biological processes undergo PCET reactions, the most emblematic one being oxygenic photosynthesis by which sunlight is used to release dioxygen into the atmosphere from water. This key-for-life reaction occurs inside the PSII unit, a complex and efficient machine. The detailed mechanism for water oxidation still challenges researchers.
PSII is the water oxidizing enzyme by which plants, algae and cyanobacteria produce dioxygen from water (51). It is a protein complex located in the thylakoid membrane of photosynthetic organisms and is constituted by several subunits. It reduces plastoquinone to plastoquinol in a two-electrons two-protons reaction at one side of the membrane (reaction ) and oxidizes water in a four-electron four-proton process at the other membrane side (reaction ).
In this process, water acts as an electron donor in the photochemical generation of dioxygen. The resultant reducing equivalents are transferred into the Photosystem I unit and are used to drive CO2 reduction into higher and useful carbohydrates (reaction ):
A first subpart of PSII efficiently absorbs solar energy thanks to pigments and chlorophyll molecules and then transfers it to the reaction center. A second subpart contains the manganese catalytic center (the oxygen evolving complex, OEC), in which the oxidizing equivalents are accumulated through the increase in the valence of this manganese cluster. The cluster also comprises calcium and chloride ions as cofactors. The “special pair,” consisting of two Chlorophyll a molecules, absorbs light at an optimal wavelength of 680 nm and is named P680. The reaction chain into PSII is generally described as follows: (1) after the absorption of a red photon (P680→*P680), the excited state of P680 rapidly transfers an electron to a vicinal pheophytine a, forming a radical pair (P680•+–PheoA•−); (2) a proximal bound plastoquinone is then reduced leading to a redox separated pair (P680•+–PQA•−); (3) a second plastoquinone (QB−) is then reduced by PQA•− in a further step; (4) the highly oxidizing chlorophyll cation radical rapidly accepts an electron from one of the two (52) active tyrosine close to the OEC, tyrosine Z (TyrZ) (53,54), leading to a proton release and the formation of a tyrosyl radical (TyrOZ•); (5) the latter oxidizes the manganese ions of the OEC cluster, its precise geometrical arrangement being still unclear (see for example [55–57] and references therein). Thanks to its four manganese ions, each of it possessing multiple stable redox states (from Mn+II to Mn+V), the OEC cluster is able to store the energy needed to oxidize water. The cluster’s valence increases during the four successive photochemical steps and this increase is compensated by proton loss. The whole sequence is known as the Kok cycle (58), describing the transition between the four states labeled S0–S4 (corresponding to the number of electrons transferred).
A key role is played by the oxidation of tyrosine residues (59). Initially described as an H-atom abstraction (HAT) (60–64), the reaction centers for electron and proton transfers are in fact different, indicating a PCET reaction. It is widely admitted that electron donation from the OEC cluster to P680•+ is relayed by the redox-active tyrosine TyrZ (Fig. 3), its phenolic oxygen hydrogen-bonded to the imidazole ring of a vicinal histidine residue (His190) (65). The oxidation of tyrosine induced a drastic pKa change from ca 10 to −2 (66,67), meaning that a proton loss can occur readily. Different studies were conducted on the formation of the tyrosyl radical and on the OEC oxidation to establish the exact mechanism (EPT/PET or CPET pathways, see for example [25,68–73] and references therein). Tyrosine always plays an important role as an intermediate for an efficient reduction of the oxidized photosensitizer. This intermediate would prevent fast recombination between *P680 and the manganese cluster (Mn(II) being also involved in the fully reduced S0 state of the OEC) and/or degradation reactions between the excited photosensitizer and the OEC. It has been also suggested that direct quenching of the excited photosensitizer by OEC would lead to lower efficiency of the process (due to high activation energy for the electron transfer reactions). Finally, water oxidation being a multiple electron reaction, the proximity of a photosensitizer which may be a good reductant would interfere through redox reactions with storage of chemical energy.
Several supramolecular mimicking assemblies were developed to reproduce the active redox triad of PSII. They usually include a photosensitizer, typically a ruthenium complex, a phenolic residue and a manganese complex.
One must note that the second redox-active tyrosine of PSII, tyrosine D (TyrD), also plays an important role in the enzyme reaction. TyrZ and TyrD are symmetrically located on the homologous polypeptide subunits D1 and D2 of PSII, and are thus at equal distance from the two chlorophyll centers PD1 and PD2 (74). However, they exhibit different kinetics and redox potentials and perform completely different functions in the enzyme. TyrZ is directly involved in the catalytic oxidation of water as described above, while TyrD, which is not needed by the photosynthetic organism to survive (75,76), is implicated as an oxidant in the lowest state (S0) of the OEC, participating in the redox tuning of P680+ and possibly providing a thermodynamical boost to TyrZ oxidation (77,78). These differences highlight the role of protein local environment (79) on the chemistry of tyrosine/tyrosyl radical. The oxidized form of TyrD is deprotonated (80–82). Thus, TyrD oxidation also follows a PCET (83–85). Proton is transferred to a vicinal histidine (His 189) (86,87) upon oxidation and is transferred back upon reduction. Depending on the pH, multiple PCET pathways have been proposed, one involving water as the proton donor (85). However, TyrOD• radical is very stable (lifetime of minutes to hours) (52,75), whereas TyrOZ• has a much shorter lifetime in natural conditions (from micro to milliseconds) (88,89), highlighting complex interaction between the tyrosine and its close environment (H-bond pattern, presence of an adjacent base, accessibility of the phenolic proton to the bulk, etc.).
The tyrosine molecule, as well as its simple analog, the phenol molecule, is of fundamental interest for determining the mechanism and the factors that influence the electron and proton transfer in PSII and other biomimicking supramolecular complexes (90,91). This is the scope of the next section.
Phenol and tyrosine: key molecules for understanding the PCET mechanism
Mimicking TyrZ chemistry
Hammarström and coworkers studied different supramolecular complexes, as model system for PSII, in order to identify the parameters that drive the PCET reaction. The experimental technique employed was the flash-quench photolysis. It consists in exciting a photosensitizer, most of the time a Ruthenium(II)-based complex, that is oxidatively quenched by an internal or external electron acceptor (sacrificial or not), thus leading to a Ru(III)-complex that can oxidize a tyrosine residue. Evolution of reactants and products are followed by time-resolved absorption spectroscopy and/or EPR. For that purpose, both Ruthenium–Manganese (Ru–Mn) complexes and a Ruthenium polypyridyl complexes covalently linked to a tyrosine (Ru–Tyr) were synthesized (92–94). By following both the absorption of the Ru complex at 450 nm, the appearance of the methyl viologen (typically used as an external electron acceptor) cation radical at 605 nm and the absorbance of phenoxyl radical around 410 nm, it was established that the reductive internal electron transfer to the Ru was feasible both in Ru–Mn and Ru–Tyr complexes. It was shown on Ru–Tyr assemblies that an electron transfer from the attached tyrosine intramolecularly occurred, thus creating a tyrosyl radical. Further step was achieved by adding a dinucluear manganese complex to the reaction medium, thus forming a mimicking triad photosensitizer–tyrosine–manganese complex of the water oxidizing enzyme (95). The tyrosyl radical formed upon photooxidation was able to oxidize the Mn–Mn complex, from Mn(III)–Mn(III) state to Mn(III)–Mn(IV) state, with redox potentials and reaction rates similar to those of the natural PSII, highlighting the peculiar role of tyrosine as an intermediate. It appears that the reduction rate (10 ms) is too slow to be a simple ET or HAT. From pH and temperature dependence of the reaction rate, it has been concluded that the electron transfer in the model system as well as in PSII is a PCET (96). Among the possible mechanisms, the two stepwise (EPT, PET) and the concerted one (CPET), the experimental observations led to the conclusion that the mechanism is concerted. At low pH, where the tyrosine is protonated, the rate constant is pH dependent and increased with pH, what was attributed to an increase of the driving force of the reaction with pH. However, the concept of a pH-dependent driving force was demonstrated to be incorrect from first principles (97–100). At higher pH, above tyrosine pKa, the rate constant becomes pH independent and faster. H-bonding was identified as an important factor in the PCET process, in line with the crystallographic structure of PSII that shows the presence of a histidine base nearby the TyrZ (101–104). Kinetic isotope effects (KIE) were also measured for the concerted reaction (CPET), being comprised between 1 and 3 (105–114).
Phenol as a simple model system for PCET in water
In order to decipher the main parameters controlling PCET oxidation of tyrosine and phenolic compounds, our group recently combined electrochemical and photochemical techniques to investigate the oxidation rate constant of simple phenol in neat water upon varying the driving force by the use of different electron Ru-based acceptors (115–117). Using flash-photolysis and stopped-flow in concert with redox catalysis experiments, we measured the bimolecular rate constant between the flash-quench-generated RuIII(bpy) and phenol in pure water and in deuterium oxide, as a function of pH or pD. The results revealed a transition between a direct phenol oxidation reaction at low pH, with a pH-independent rate constant, and a stepwise PET reaction at higher pH consisting of a preliminary deprotonation of phenol by OH− ions (no other base being present in solution) with a remarkable unity-slope variation, the latter observation being in disagreement with previously reported studies (96,107,118). Analysis of the kinetics and KIE (2.5–3, observed all along the pH range) allowed us to rule out the EPT stepwise mechanism (electron transfer first followed by proton transfer) at low pH in favor of the concerted CPET pathway, since the electron transfer is, in this case, the rate-determining step, and so no KIE is expected. In the presence of phosphate buffer, with hydrogen phosphate as the proton acceptor, it was demonstrated that a concerted process is followed, with a pseudosecond order rate constant proportional to the hydrogen phosphate concentration, and a significant KIE (119). Further experiments were also conducted taking pyridine as a prototypal example of biologically important nitrogen bases involved in PCET (120). We observed a concerted mechanism also for the oxidation of phenol with pyridine as the proton acceptor.
Temperature and buffer dependence of the rate constant gave further insights (119). It was indeed possible from these studies to obtain separately the reorganization energy and the pre-exponential factor. We were able to conclude, on the basis of converging results obtained by the different techniques, that the concerted pathway, in which water was the proton acceptor, was characterized by the smallest reorganization energy (0.45 eV) while it amounts to 0.86 eV with hydrogen phosphate as the accepting base. From these reorganization energies, the corresponding solvation radius was estimated in each case, since the reorganization energy is mostly made of solvent reorganization. With hydrogen phosphate as the proton acceptor, a radius of ca 3.5 Å is obtained (Fig. 4), corresponding to a single hydrogen phosphate molecule, meaning that the proton remains localized between the donor and the acceptor. With pyridine as the proton acceptor, the reorganization energy (0.53 eV) corresponds to a solvation radius of ca 5.7 Å, suggesting the involvement of a primary shell of water molecules, tightly bound to the pyridinium cation. Finally, when water (in water) is the proton acceptor, the small reorganization energy leads to a solvation radius of ca 6.5 Å, indicating a fairly large delocalization of the proton over several water molecules and a concerted displacement of several protons, in line with the recent spectroscopic data (121) as well as recent findings related to photochemically triggered proton transfer (122). We also noticed that pre-exponential factors for both CPET reactions (with water or hydrogen phosphate as proton acceptor) were lower than for simple outer-sphere electron transfers, an expected result since proton tunneling, implying lower values of the transmission coefficient, is a key feature of these processes. When comparing water and hydrogen phosphate it appears that the pre-exponential factor is higher with water than with phosphate, i.e. water is a much more efficient proton acceptor. The oxidation of phenol with water (in water) as proton acceptor is thus concerted, the electron is transferred concertedly with a Grotthuss-type transport of the proton (123), the charge being delocalized over a large cluster involving several water molecules.
These fundamental features highlight on one hand the role of the reorganization factor λ that is small in CPET reaction, and most importantly the key role of preorganization of the system that allows for efficient proton and electron tunneling in the transition state. This latter factor appears as crucial in favoring the concerted pathway over the sequential ones. In this connection, the strong, well-defined H-bond between TyrZ and histidine 190 in the OEC provides an ideal structure for a concerted proton-coupled oxidation of the phenolic group.
PCET in enzymes containing a phenolic residue
As mentioned in the introduction, PCET reactions are also of fundamental importance in the generation and transport of amino acid radicals which serve as cofactors for substrate activation and as charge transporter in enzymes (124). Whereas these radicals, when free in solution, have short lifetimes (typically from micro to millisecond), enzymes acquired the capability of managing electron and proton equivalents to use radicals oxidizing power for chemical reactions over these different timescales. As oxidation of amino acids in the physiological pH range often requires a proton loss, many redox mechanisms in enzymes are PCET (15,125,126). In such processes, proton transfer is fundamentally limited to short distances (tunneling of a “heavy” particle) whereas electron transfer can occur over very long distances (127,128). Without the effective aid of relays, i.e. redox-active amino acids, long-range electron transfer in peptides and proteins would be too slow to allow for catalysis or energy conversion. In this context, several unnatural amino acids were synthesized and pKa, redox potential and/or O-H bond strength were tuned to access a wide range of proton- and electron-donor ability and induce spectroscopic changes (e.g. EPR—detectable charges) that can help in clarifying PCET mechanisms (129).
Among the different examples of enzymes involving tyrosine/tyrosyl radical chemistry, class-I RNR is probably the most studied. There are three classes of RNRs, and their common function is to catalyze the conversion of nucleotides to deoxynucleotides and thus they play a key role in DNA replication and repair (130). Class-I RNRs are composed of two types of homodimeric subunits called α2 and β2 with a complex between them that is necessary for enzyme activity and turnover (131). α2 subunits contain binding sites for nucleotide NDP substrates and ATP effectors that govern the specificity and the rate of the nucleotide reduction (132) whereas β2 subunits contain a diferric tyrosyl radical cofactor, Y122•. The enzyme activity is thought to be initiated by the transfer of the oxidizing equivalent from Y122• in β2 to the C439 cysteine residue at the active site of α2, generating a thiyl radical S• (Fig. 5, 133). The question of how this transfer occurs, knowing that this radical transport occurs at very long distance (over 35 Å or so) (134,135), was proposed to be via a PCET amino acid residues pathway involving conserved amino acid residues, a side chain with three tyrosines (Y356, Y731 and Y730), lying between the cofactor and the catalytic cysteine.
Several studies were thus undertaken by Stubbe, Nocera and coworkers to elucidate the detailed mechanism of the radical transport and to identify radical intermediates. For that purpose, photochemical RNRs (photoRNRs) were designed to probe the mechanistic steps associated with radical transport between tyrosine residues from the β2 to the α2 subunit (136–141). These assemblies were based on a α2 subunit bound to the 20-mer, C-terminal peptide tail of β2 (Y-βC19) which contains a bonded photooxidant, being known that the Y-βC19 contain the redox-active Y356 residue in addition to the β2-α2 binding determinant (133,142,143). They showed that the excitation of the appended photooxidant efficiently induced the production of Y356• and that, in the presence of RNR substrate and effector, the latter could be transported into the α2 active site to initiate deoxynucleotide reduction. Several photo-oxidants with tyrosine were tested, containing tryptophan (Trp) (144), benzophenone containing artificial amino acid (145), anthraquinone (138) and finally Re-based inorganic compounds previously synthesized and photophysically described (146,147). The latter construct showed remarkable high turnover yield (139) and was then associated with fluorotyrosine residues, thanks to the ability of these artificial amino acids to lower the phenolic proton pKa enabling photoactivity in the RNR pH range 6–8.5. The combination of these two moieties was demonstrated to be a selective method for the generation of tyrosyl radicals on the peptide bond. Emission quenching measurement of the photoexcited Re-center by the electron transfer from the vicinal tyrosine with pH allowed the probing of the tyrosine phenol pKa in the peptide. These results showed that the pKa is identical to its solution value, showing that the photoRNR construct is truly modeling RNR. Emission characteristics of the polypyridyl complexes highly depend on the polarity of their environment, and the Re-moiety was shown not to perturb the peptide-protein environment, consequently the designed assemblies may serve as probe for peptide conformation and binding to class-I RNR. Emission lifetimes revealed multiple solvation environments of the chromophore, described as protein bound and solvent exposed, highlighting the conformational flexibility within the photoRNR and demonstrating that PCET provides a basis for conformational gating in RNR (140).
Cytochrome c Oxidase
CcO is an enzyme located in bacterial or mitochondrial membrane of aerobic organisms and is the terminal enzyme of the respiration chain that catalyzes the reduction of dioxygen to water (148). This reduction is coupled to a proton transfer through the membrane (149), leading to the establishment of an electrochemical proton gradient (the protonmotive force) across the membrane. This force is a key element in primary biological energy transduction, e.g. the proton gradient is employed for the synthesis of ATP by ATPase (42,150). The active site of the enzyme contains a copper complex and a heme where dioxygen initially binds during its activation. The catalysis of oxygen reduction in CcO requires four electrons from the electron carrier protein (CcO) located at the positively charged side (P-side) of the membrane. Electron transfers in CcO are known to be tightly coupled to proton transfers (151): each electron transfer from CcO to the catalytic site is coupled to the uptake of one proton from the negative side (N-side) of the membrane. Moreover, each electron transfer to the catalytic site is linked to pumping of one proton across the membrane. As a consequence, each reduction of a single O2 into H2O involves eight protons taken up from the N-side and four protons released to the P-side. The precise description and location of the pump site is still unclear (152).
It has been shown that in CcO, a histidine (His240) was covalently linked to tyrosine residue (Tyr244, bovine heart numbering) the latter likely being a redox center which would hold an oxidation equivalent in the P intermediate state of the enzymatic reaction. Regarding the global four electrons and four protons required for the O2 reduction process, Tyr244 would serve as a proton donor but also as an electron donor via a mechanism not yet fully established (40,153–158) as its oxidized state was concluded to be neutral (159). Of course, as for similar systems, a dependence of these processes with the Tyr protonation state was observed, in relation with possible H-bonding with His (160,161). It was also shown that water molecules play a critical role in proton transfer in CcO, both on kinetics and on pKa of residues (162).
To our knowledge, few photochemical studies were conducted on CcO (see for example [163–167]) and none of them explicitly treated the PCET mechanism. The complete elucidation of the CcO mechanism on a molecular scale is still a challenge in view of the system and the possibility of several mechanistic options for the coupled electron and proton transfers (151,168,169).
PCET in DNA
Intracellular oxidations, when occurring aberrantly and under chronic conditions, are at the origin of diseases such as cancer, diabetes or atherosclerosis. Oxidizing attacks on DNA lead to damage and so to a variety of (stable or not) products that can generate nucleobase lesions, single- and double-strand breaks, DNA–DNA and/or DNA–protein cross-links. The latter is predominant when DNA is exposed to ionizing radiation (170), but little is known about the molecular mechanisms involved.
When DNA is exposed to high-energy radiation, its components (bases, sugars, surrounding water molecules) are excited or ionized, generating holes and subsequently ion radicals (171). As already mentioned earlier in this review, one-electron oxidation or reduction of a compound drastically affects its acid–base properties. For DNA bases, an electron loss results in an increase in acidity, whereas an electron gain results in a basicity increase. Therefore, there is little doubt that redox reactions of DNA bases involve PCET. Guanine (G) presents the lowest ionization potential of the four bases (with thymine [T], cytosine [C] and adenine [A] [172,173]) and was shown to be responsible for hole trapping in DNA (174), with formation of the radical cation G•+. The latter is mobile and can migrate away from the initial site of damage and it has been established that reactions between G•+ and amino acids can lead to DNA–protein cross-linking (see for example ). More generally, it is known that nucleobase damage does not necessarily occur at the initial oxidation site and that radical cations in DNA are able to migrate over long distances by a reversible hopping process before irreversible trapping by H2O or O2 (176,177).
DNA-mediated electron transfers can result in reactions over long distances, depending on distances between reactants, protonation states and base pair stacking, reactions that lead to damage but also to repair of DNA (178,179). Aromatic amino acids were shown to involve interactions with nucleic acid bases and may play a crucial role in oxidative damage through radical migration. Moreover, DNA damage resulting in the oxidation of a guanyl base can be repaired by the refilling of electronic vacancies in it via electron transfer from the surrounding environment, namely proteins, with particular emphasis from histone proteins. Tryptophan and tyrosine amino acids were found to be the most efficient reducing agents for guanyl radicals formed in plasmid DNA. The resulting reductive electron transfer from protein residues to guanyl radicals may involve proton transfer, but no direct evidence exists concerning the proton source and the kinetics of these reactions.
Yurkovskaya and coworkers conducted several studies on the electron transfer between guanosine radical and amino acids in aqueous solutions, to model chemical DNA repair, as well as on intramolecular electron transfer in dipeptides, with tyrosine involved (45,180,181). They used the time-resolved chemically induced dynamic nuclear polarization (TR-CIDNP) technique that allows the fast and fine detection of radical pairs. In these studies, the guanosyl radicals were photochemically generated from guanosine monophosphate in the quenching of the triplet state of a dye (2,2′-dipyridyl). Their subsequent reductions were achieved by an electron transfer from N-acetyl-tyrosine or -tryptophan, forming a photocycle. Since both the electron removal from guanine and the repair of DNA guanyl radicals at physiological conditions were found, in particular by Milligan and coworkers (44,182–184), to be coupled to proton transfer (185), they conducted measurement as a function of pH. They observed that the triplet dipyridyl quenching was dependent on the protonation states of the reactants, i.e. tyrosine essentially. Three different regions were observed, at acidic (pH < 5), moderately basic (6 < pH < 9.5) and more strongly basic pH (>10.5). For the latter region, an acceleration of the deprotonation by HO˙ was also observed. Conclusions were that pH induced a change in the primary photochemical step, from an electron transfer in basic conditions to a HAT at neutral to moderately basic pH, then back to an electron transfer in acidic conditions.
Even if this review focused on systems containing a phenol-type molecule, we must add that similar questions and concepts as those described in the preceding sections also raised in nonphenolic DNA residues and that a large literature exists on this topic (see  and references therein).
For example, the flash-quench technique was employed by Barton, Zewail and coworkers to study the electron transfer chemistry of DNA by using intercalators such as ruthenium complexes. They showed that DNA-mediated electron transfers, and in particular DNA damage through guanine oxidation, were sensitive to π-stacking and heliticity of DNA, can occur very rapidly and over very long distances (177,186–190). Phosphate, present in the DNA backbone, as well as water molecules and ions are likely to play a role in gating charge transport (191) and possibly in the PCET reactions that may occur in DNA. As mentioned earlier, guanine radical cation received much attention as a single-oxidized base as well as a base pair with cytosine (G-C) since its reversible deprotonation is of key importance for hole transfer in DNA (Fig. 6). Shafirovich and coworkers conducted extensive photochemical studies on the oxidation of guanine in DNA or oligonucleotides by aminopurines and pyrenyl radical cations (192–203). Based on their observation of the formation of the guanine radicals and measurements of KIE in the range 1.3–1.7, they proposed that the oxidation of guanine by aminopurines occurred by PCET. Thorp and coworkers also investigated guanine oxidation by electrochemistry and stopped-flow spectrophotometry and suggested a concerted mechanism (204,205). Multisite electron and proton transfer mechanism with solvent as the proton acceptor was later proposed (4) to explain the results obtained by Shafirovich and coworkers as with the G-C base pair (206). To date, the preferred site for the guanine radical cation deprotonation as well as the precise mechanism is still unclear and actively investigated (see for example [44,207–209]). It is worth noting that electrochemical oxidation of a single guanine–cytosine pair in chloroform was recently demonstrated to occur along a stepwise pathway with electron loss being the rate determining step and with little influence of pairing on both the kinetics and the thermodynamics of the oxidation (210). The oxidation process is not concerted in this case. It remains to be explored if these conclusions may apply or not to DNA and/or oligonucleotide oxidation.
In oligonucleotides that do not contain guanine, or containing thymine–thymine mispair, DNA oxidation occurs at thymine. In spite of the general thinking that the low redox potential of guanine governs the reactivity of radical cation in DNA, it appeared more recently that steric effects can affect the accessibility of reactants molecules, such as H2O, to the nucleobase radical cation and modulate reactivity (211). In this connection, Joseph and Schuster recently examined the case of A/T-rich sequences (212). The oxidative damage preferentially occurred at thymine, which has a higher oxidation potential than adenine, in contradiction with the generally accepted argument that saying the lowest redox potential (the one of that of guanine, when present) is the main factor governing the reaction. This particular feature was explained by a PCET mechanism for the formation of the thymine radical, in which an electron is transferred from thymine to adenine radical cation in concert with the transfer of a proton from the thymine methyl group to an H-bonded water molecule. Such a pathway would offer lower activation energy than the sequential pathways (EPT or PET) and would play a dominant role in the formation of thymine oxidation products during the one-electron oxidation of DNA, especially when containing few or even no guanines. So generally speaking, charge transport in DNA is very likely to be coupled to proton transfer since most redox-active residues have acid–base properties, meaning that their pKa depend on their redox state. However, to date, no clear mechanistic scheme has emerged from isotope substitution, leaving the intimate details of the mechanisms unresolved.
Conclusions and perspectives
Photochemical techniques, based on ultrafast “switch on” of the reaction associated with fast and sensitive optical and/or EPR detection of intermediates and products, are useful and powerful methods for probing PCET reactions and processes, leading to mechanistic insights, in particular for the identification of stepwise and concerted processes (CPET) that have the advantage of by-passing high-energy intermediates. These studies may be complemented by other experimental approaches (e.g. electrochemical), opening the possibility of transferring information from one domain to the other so as to obtain a better comprehension of the reaction. Concerted processes during the oxidation of phenol have been shown to be dominant in water with water as proton acceptor, revealing the possibility of transporting protons over large distances along H-bond networks in a Grothuss-type mechanism in concert with electron transfer. These findings may contribute to the understanding of proton transport coupled to charge transfer in natural systems through chains of water molecules and chains of molecules able to give and to accept H-bonds. Purposely designed biomimetic molecules or supramolecular assemblies may help in this connection. PCET processes are involved in the functioning of many enzymes and biomolecules, especially DNA oxidation for which a detailed mechanistic picture is still missing, as well as for the oxidation of tyrosine by guanine radicals (repair process). PCET are also of great importance in the resolution of modern energy challenges, especially those related to the activation of small molecules to convert and store solar energy (e.g. water oxidation, proton and carbon dioxide reduction). In these processes, PCET reactions are also often coupled with bond-breaking (e.g. cleavage of a C-O bond when reducing CO2 to CO) or bond-forming (e.g. O-O bond formation during water oxidation) between heavy atoms. Comprehension at a molecular level of these complex reactions is a key forefront challenge for designing and inventing catalytic systems. In this area also it is anticipated that photochemical approaches will prove to be essential.
[ Julien Bonin ]
Julien Bonin pursued his graduate studies at the Université Paris-Sud XI (France) where he obtained his Ph.D. (2005) in Physical Chemistry working with Prof. Mehran Mostafavi on the femtosecond spectroscopy of solvated electrons. Then he joined the Radiation Laboratory of the University of Notre Dame (IN, USA) for a year, as a Postdoctoral Research Associate, working on the study of hydroxyl radical reactivity in supercritical water by pulse radiolysis with Prof. David M. Bartels. Since 2006, he is an Associate Professor of Chemistry at the Université Paris-Diderot (France), and has been a Visiting Scientist at the Chemistry Department of the Massachusetts Institute of Technology (MA, USA) during the spring of 2010, working with Prof. Daniel G. Nocera. His current research interests include photoinduced electron and proton-coupled electron transfers as well as photoinduced catalysis.
[ Marc Robert ]
Marc Robert was educated at the Ecole Normale Supérieure de Cachan (France) and gained his PhD in 1995 from the Université Paris-Diderot with Profs. Claude Andrieux and Jean-Michel Savéant, working on electron transfer chemistry. After 1 year as a Postdoctoral Fellow at Ohio State University with Prof. Matt Platz, he joined the Université Paris-Diderot as an Associate Professor. He was promoted to Professor in 2004 and became a Junior Fellow of the University Institute of France in 2007. His current interests include electrochemical, photochemical and theoretical approaches to electron transfer reactions and proton-coupled electron transfer in both organic chemistry and biochemistry.