Re-evaluation of intramolecular long-range electron transfer between tyrosine and tryptophan in lysozymes Evidence for the participation of other residues Marilyne Stuart-Audette 1 , Yves Blouquit 2 , Moshe Faraggi 3 ,Ce ´ cile Sicard-Roselli 1 , Chantal Houe ´ e-Levin 1 and Pierre Jolle ` s 4 1 LCP, Centre Universitaire, Orsay Cedex, France; 2 Institut Curie-Section de Recherche, Centre Universitaire, Orsay Cedex, France; 3 Department of Nuclear Engineering, Ben Gurion University of the Negev, Israel; 4 Laboratoire de Chimie, Muse ´ um National d’Histoire Naturelle, and LNCP, Universite ´ Paris 6, France One-electron oxidation of six different c-type lysozymes from hen egg white, turkey egg white, human milk, horse milk, camel stomach and tortoise was studied by gamma- and pulse-radiolysis. In the first step, one tryptophan side chain is oxidized to indolyl free radical, which is produced quantitatively. As shown already, the indolyl radical subse- quently oxidizes a tyrosine side chain to the phenoxy radical in an intramolecular reaction. However this reaction is not total and its stoichiometry depends on the protein. Rate constants also vary between proteins, from 120Æs )1 to 1000Æs )1 at pH 7.0 and room temperature [extremes are hen and turkey egg white (120Æs )1 ) and human milk (1000Æs )1 )]. In hen and turkey egg white lysozymes we show that another reactive site is the Asn103–Gly104 peptidic bond, which gets broken radiolytically. Tryptic digestion followed by HPLC separation and identification of the peptides was performed for nonirradiated and irradiated hen lysozyme. Fluorescence spectra of the peptides indicate that Trp108 and/or 111 remain oxidized and that Tyr20 and 53 give bityrosine. Tyr23 appears not to be involved in the process. Thus new features of long-range intramolecular electron transfer in proteins appear: it is only partial and other groups are involved which are silent in pulse radiolysis. Keywords: gamma and pulse radiolysis; intramolecular long- range electron transfer; lysozyme; one-electron oxidation. The early suggestions [1,2] of intramolecular long-range electron transfer (LRET) have now been verified by numer- ous observations of LRET between donor/acceptor redox centres with known separation distances in proteins, peptides and other small rigid organic molecules. This property exists in polymers and especially in biopolymers (proteins and DNA). Thus these macromolecules are considered as candidates for nanoelectronics components. In this view, several studies were devoted to a better understanding of the factors leading to modulation of LRET (for a review, see [3] and references therein [4–6]). As part of a program to unravel the mechanistic basis of LRET in proteins, systematic investigations on electron transfer between tyrosine and tryptophan in proteins and peptides have been conducted [7–10]. In this experimental system, developed first by Pru ¨ tz et al. [11], pulse radiolytically generated azide radical and dibromide radical anion preferentially oxidize the side chain of tryptophan (Trp) to the indolyl radical (Trp Æ ) in molecules that also contain tyrosine (Tyr). N Æ 3 þ Trp-X-Tyr ! Trp Æ -X-Tyr+N À 3 þ H þ ð1Þ The initiating oxidation (1) can be effected by other radiolytically generated species such as the dibromide (Br 2 – ) and dichloride (Cl 2 – ) radical anions [12]. The trypto- phan neutral radical, with a midpoint reduction potential that is near 1 V at pH 7.0 [13], rapidly oxidizes the tyrosine side chain to the phenoxy radical (TyrO Æ )intheintra- molecular equilibration: Trp Æ -X-Tyr $ Trp-X-TyrO Æ ð2Þ Both TyrO Æ and Trp Æ absorb strongly in different parts of the visible region. One of the first observations of reaction (2) in proteins was obtained in hen egg white lysozyme (HEWL) [14–16]. However previous pulse radiolysis studies indicated a low yield of tyrosyl radical compared to that of initial oxidant and of tryptophanyl radical [14]. Since the final products coming from this process were never characterized, the stoichiometry of LRET was never evaluated. The aim of the experiments described in this paper was thus to verify whether the stoichiometry in LRET (equality of tyrosyl vs. tryptophanyl radicals) was respected, to identify some of the aromatic amino acids responsible of LRET, and possible other processes involved. Correspondence to C. Houe ´ e-Levin, LCP, UMR 8000 baˆ t. 350, Centre Universitaire, 91405 Orsay Cedex, France. Fax:+33169153053,Tel.:+33169155549, E-mail: chantal.houee-levin@lcp.u-psud.fr, Web site: http://www.lcp.u-psud.fr Abbreviations: HEWL, hen egg white lysozyme; TEWL, turkey egg white lysozyme; Trp Æ , the indolyl radical of tryptophan; TrpH Æ+ ,the protonated indolyl radical of tryptophan; TyrO Æ , the phenoxy radical of tyrosine; LRET, long-range electron transfer. (Received 7 April 2003, revised 6 June 2003, accepted 7 July 2003) Eur. J. Biochem. 270, 3565–3571 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03741.x The c-type lysozymes (129–130 amino acids in most cases) having similar sequences (for the sequences see [17]), three-dimensional structures (obtained by crystallography and two-dimensional NMR in solution) and biochemical activity [18–22] provide an excellent model system for exploring the effect of amino acid substitution on the intramolecular electron transfer. This approach complements other studies, directed at understanding the electron transfer process, which involves changing the protein sequence by site-directed mutagenesis and the use of chemically modified proteins. However in this work we show that another process i.e. a peptide bond fragmentation, is initiated by LRET, which means that there might be some ÔleakageÕ in electron transfer. These points are crucial for understanding of redox processes in biology and also for an eventual development of nano- electronics. Experimental procedures Chemicals (sodium azide, potassium phosphate, sodium hydroxide) were of the highest quality available (Prolabo Normatom or Merck Suprapure). Nitrous oxide was delivered by ALPHA GAZ; its purity was higher than 99.99%. HEWL, EC 3.2.1.17 (crystallized three times, dialyzed and lyophilized), and turkey egg white lysozyme (TEWL) were supplied by Sigma. Other lysozymes were prepared using known procedures [23]. The purity of the protein preparations was tested by SDS/PAGE according to Sha ¨ gger and von Jagow [24] and by reversed-phase HPLC. All proteins were used after dialysis against buffer. The solution concentration was verified by the protein absorbance at 278 nm [25]. Water was obtained from an Elga Maxima device (resistivity ¼ 18.2 MX). In gamma radiolysis, protein solutions (1.4 · 10 )4 M )in 10 m M potassium phosphate buffer (pH 7.0) and 0.1 M sodium azide (NaN 3 ) in order to produce the azide radical, were saturated with N 2 O. Protein solutions were prepared shortly prior to irradiation. The samples were submitted to c irradiation in a panoramic IL60PL (Cis Bio International) 60 Co source at a dose rate, determined by Fricke dosimeter, of approximately 1.0 GyÆs )1 . For pulse radiolysis, radicals were generated by the introduction into an aqueous solution of a 200-ns pulse of high-energy electrons (% 4 MeV) from a linear accelerator located at the Curie Institute, Orsay France [26]. In both cases, the predominant primary radical products, e aq – and OH Æ , formed by ionizing particle (high energy electrons or photons) interactions with water [27], were converted to the azide radical, N 3 , by reactions with N 2 O at saturation and 0.1 molÆL )1 azide anion [27]. The pH of the solution was adjusted with 5–20 mmolÆL )1 phosphate buffer. In order to prevent foam formation (denaturation of the protein) saturation with N 2 O was carried out by introducing the gas on top of the stirred cold (4 °C) solution for 60 min. The total radical concentration was % 2 lmolÆL )1 determined by methyl viologen dosimetry [28]. These low concentrations were used to minimize possible second-order decays of TyrO Æ and Trp Æ . The protein was present at 0.14 mmolÆL )1 (2 mgÆmL )1 ), except when we determined the concentration dependence of the rate constant for reaction (1). For this experiment the protein concentration was varied between 0.05 and 0.5 mmolÆL )1 . All reactions, monitored by changes in the UV–visible region, were functionally first order, and the rate constants were obtained by nonlinear least-squares fitting to the absorbance data. The numbering of the amino acid residues in each lysozyme protein follows that of the Protein Data Bank (PDB) of the Brookhaven National Laboratories. Irradiated protein solutions were analysed by SDS/ PAGE and by HPLC. b-mercaptoethanol was added to the irradiated samples to reduce the disulphide bridges before electrophoretic analysis on a SDS-containing gel. Proteins bands were stained with Coomassie blue R-250. Stained gels were scanned using an Ultrascan XL laser microdensitometer (LKB). The proportion of material present in each band was determined using the intensity of the vertical cut corrected for the bandwidth. Dimers were separated from the monomers by HPLC, using a PorosR1 (4 · 100 mm) column. A gradient of 35–60% (trifluoroacetic acid 0.1%/acetonitrile 70%) in trifluoroacetic acid over 30 min (at a flow rate of 1.5 mLÆmin )1 ) was used. Detection was at 214 nm. Evaluation of the fraction of dimerized proteins was achieved by calculation of the portion of the total area under the curve occupied by the dimers peak. The obtained value was divided by two. MALDI-TOF MS was performed at The Institute of Nuclear Physics, Orsay, France. N-terminal sequencing was done at IBCP, Lyon, France. For aminoethylation, proteins were kept 2 h in a solution containing 6 M urea, 1 M Tris pH 8.6, 10 m M EDTA; b-mercaptoethanol (50 m M ) was added to reduce the disulfide bridges. Ethylene imine (Fluka) was added (20 times expected thiol concentration) and the solution was incubated in the dark for 40 min before 20 lLÆmL )1 of 2-mercaptoethanol was added. Salts were removed using reversed-phase HPLC and the samples were dried using a Speed-vac Savant apparatus. Trypsin hydrolysis was performed in ammonium hydro- genocarbonate (100 mgÆmL )1 ) aqueous solution. Protein concentration was 2 mgÆmL )1 , 3% w/w of trypsin (Worth- ington) was added. The solution was kept overnight at 30 °C. Peptides were separated on a column Vydac C4 (4.6 · 250 mm). Solution A was a solution of trifluoroacetic acid (0.1%), solution B was a solution of 0.1% trifluoro- acetic acid and 70% acetonitrile (gradient 0–100% B in 90 min, flow rate 1 mLÆmin )1 ). Detection was at 214 nm. Gel permeation chromatography was performed at 5 °C with buffer 0.2 M Na 2 SO 4 ,0.1 M phosphate pH 6.5, sapo- nine 0.005%. Precipitation of proteins was first achieved with ammonium sulphate. The column (3 · 50 · 90 cm) was filled with gel TSK Toyopearl (HW40S or HW50S) or Ultrogel ACA50. Results Pulse radiolysis Pulse radiolysis studies of amino acids and peptides have shown that below pH 9.0 the tryptophan side chains, compared with other amino acid residues, react rapidly with the azide radical so that one-electron indole oxidation is the predominant reaction in peptides and proteins containing 3566 M. Stuart-Audette et al. (Eur. J. Biochem. 270) Ó FEBS 2003 both tryptophan(s) and tyrosine(s) [7–14]. This initial, rapid increase of absorbance between 350 nm and 600 nm with a maximum at 510 nm is due to Trp formation (Fig. 1). From the linear dependence of the pseudo-first order rate constants on protein concentration an average second-order rate constant of (9.5 ± 0.9) · 10 8 LÆmol )1 Æs )1 (25 °C) for TEWL was calculated. Similar second-order rate constants were obtained for the other lysozymes (Table 1). This initial, rapid increase of absorbance is followed by a slower decrease of the 510 nm band due to Trp Æ reduction. This Trp Æ reduction is accompanied by tyrosine oxidation to TyrO Æ , seen from the simultaneous absorbance increase at 410 nm (Fig. 1). The time-dependent spectral changes during this one-electron transfer from tyrosine to Trp Æ resemble the changes obtained with peptides containing both tryptophan and tyrosine; i.e. absorbance maxima at % 510 and 410 nm, and isosbestic points at % 430 nm and % 370 nm [7,8]. These isosbestic points, together with the fact that the apparent first-order reaction rate constant is the same whether measured at 410 or 510 nm, suggest that electron transfer from tyrosine to Trp Æ occurs in a single step. It was also established that this transfer is in fact first order by demonstrating that the apparent rate constant is independent of protein concentration from 5 · 10 )5 to 5 · 10 )4 molÆL )1 . Thus, electron transfer between a tyro- sine residue and a tryptophanyl radical in all the lysozymes studied appears to be an intramolecular process. The monomolecular rate constants at pH 7.0 and 25 °Care summarized in Table 1. In the seconds time range there is a final slow radical decay. This reaction was always second- order representing a radical–radical recombination reaction (e.g. dimerization of TyrO Æ radical). Thus, the reaction (2) equilibrium is reached before appreciable radical loss occurs via a slow decay [under the condition of our experiments the slowest intramolecular electron transfer half-life for HEWL and TEWL was 5 ms at pH 7.0). The determination of the intramolecular first-order rate constant is not compromised by a reaction overlap with radical decay. We can estimate the Trp Æ and TyrO Æ yields in the lysozymes from the magnitudes of the absorbance changes at 510 and 410 nm and the assumption that the extinction coefficients of the protein bound radicals and those measured from amino acids peptides are the same: for Trp Æ , these are 1800 and 300 LÆmol )1 cm )1 at 510 and 410 nm, respectively; for TyrO Æ , 70 and 2600 LÆmol )1 Æcm )1 at, respectively, 510 and 410 nm [29,30]. The estimated Trp Æ yields calculated for the initial oxidation of the proteins above pH 4.5 by the inorganic radicals were quantitative. The estimated overall TyrO Æ yields after reaction (2) varied with the protein and could be as low as 40% compared to the 90% loss of Trp after the completion of this reaction. We confirm that the process is nonstoichiometric and that the amount of LRET depends on the primary structure of the protein. Final products The final products were detected and quantified after pulse and steady-state gamma radiolysis, by SDS/PAGE and by HPLC (typical electrophoresis gels are shown on Fig. 2). Localization of dimers For all lysozymes studied, dimers are found and their quantity increases with the dose. The G-values of dimeri- zation measured by HPLC or by SDS/PAGE do not depend much on the origin of the protein [31]. The presence of bityrosine could be detected by absorp- tion [32] and fluorescence, with an excitation wavelength of 325 nm; its emission signal is centred at 410 nm. Using gel permeation liquid chromatography, lysozyme dimers were partially separated from the monomers. It was observed that the fluorescence signal for bityrosine comigrates with the fractions containing the dimerized proteins. Fig. 1. Pulse radiolysis study of TEWL oxidation by azide radicals. Protein 1.4 · 10 )4 ÆmolÆL )1 , phosphate buffer 10 m M pH 7.0, [N 3 – ] ¼ 10 )2 ÆmolÆL )1 ,N 2 O atmosphere (% 1 atm), dose 4 Gy. Optical path 2cm. Table 1. Rate constants of LRET, number and location of Tyr and Trp residues, for the various lysozymes. Lysozyme source (No. of amino acids) Number of Trp/ molecule (position) Number of Tyr/ molecule (position) k Lys+N3 mol )1 Ædm 3 Æs )1 k LRET s )1 Hen egg white (129) 6 (28, 62, 63, 108, 111, 123) 3 (20, 23, 53) (7.9 ± 0.8) · 10 8 120 ± 10 Turkey egg white (129) 6 (28, 62, 63, 108, 111, 123) 4 (3, 20, 23, 53) (9.5 ± 0.9) · 10 8 120 ± 10 Horse milk (129) 5 (28, 62, 63, 108, 111) 4 (23, 34, 54, 123) (7.0 ± 0.7) · 10 8 230 ± 20 Camel stomach (130) 6 (3, 28, 34, 64, 109, 112) 6 (20, 38, 45, 54, 63, 124) (6.3 ± 0.7) · 10 8 500 ± 50 Tortoise (130) 5 (28, 64, 109, 112, 124) 6 (3, 20, 23, 45, 54, 63) (7.8 ± 0.8) · 10 8 300 ± 30 Human milk 5 (28, 34, 64, 109, 112) 6 (20, 38, 45, 54, 63, 124) (4.7 ± 1.0) · 10 8 1000 ± 100 Ó FEBS 2003 Pulse radiolysis oxidation of c-type lysozymes (Eur. J. Biochem. 270) 3567 The fractions containing the HEWL dimers and samples of irradiated solutions were hydrolysed by trypsin and amino groups were ethylated. Then fragments were separated by HPLC and identified by their amino acid sequences. The relative population of the peptide fractions from the irradiated samples were compared to the corresponding ones obtained with the native lysozyme. The proportions of tyrosine 23- and 53-containing fractions were found to be significantly lower in the dimer-rich preparations. (Fig. 3). A fraction of each peptide was analysed by fluorescence, a very sensitive technique. The bityrosine signal was observed in all peptides containing tyrosine 23 or 53 but not in peptides containing tyrosine 20 alone. According to these results, tyrosines 23 and 53 could be involved in the dimerizations whereas no sign of tyrosine 20 involvement was seen. Oxidized tryptophans An increase in 315 nm absorbance was also observed in nondimerized fractions demonstrating that at least another oxidation product was formed. Tryptophan solutions, irradiated in the presence of N 3 , also show a dose-dependent increase in 315 nm absorbance (M. Stuart-Audette, C. Sicard-Roselli, C. Houe ´ e-Levin, unpublished results). Knowing that the electron transfer between Trp and Tyr is partial, we have been seeking for tryptophan oxidation products. Further fluorescence studies shown the presence of a different emission spectrum upon excitation at 380 nm. The signal is compatible with the presence of N-formyl- kynurenine [33]. The peptides obtained by hydrolysis of irradiated solu- tion, for which the amino acid composition of peptides was determined, were screened for the fluorescence signal obtained when an excitation wavelength of 380 nm is used. The signal was observed in peptides containing tryptophans 108 and 111. Because none of the peptide contains only tryptophan 108 or 111, it was not possible to identify which of the two (if not both) is oxidized and not repaired. The relative proportions of the irradiated fractions, at different doses, were compared to the ones obtained with the native HEWL. Tryptophan 108- and 111-containing peptides decreased rapidly as doses were increased. No indication of tryptophan 28, 62, 63 and 123 modifications was noted. Fragments In addition to dimerization, HEWL and TEWL appear to undergo fragmentation. The N-terminal sequences of the first 10 amino acids of the longest moieties are identical to that of the native proteins. MS of HEWL gave a small peak corresponding to 11 286 Da and another one corresponding to 3047.5 Da (Fig. 4). For TEWL, we find 3055.7 Da for the smaller moiety. This allows us to determine the site of fragmentation at the peptidic bond between Asn103 and Gly104 for both proteins. In addition, N-terminal sequen- cing of the smaller moiety indicated that Gly residue is intact. It is interesting to note that TEWL, which has the same Asn103–Gly104 sequence, gets also fragmented at the same site as HEWL, whereas human milk lysozyme and tortoise lysozyme, which do not get fragmented upon oxidation, do not have the Asn103–Gly104 sequence (Fig. 2). The yields of fragmentation (Table 2) vary for these proteins. Discussion In this work, we performed a comparative study of the one- electron oxidation of six proteins from the c-type family of lysozymes; these proteins differ by several amino acids but their three-dimensional structures are very similar. Pulse radiolysis results show that the rate constants of one-electron oxidation by azide radicals are of the same order of magnitude. For all proteins, tryptophanyl radicals Fig. 2. SDS/PAGE analysis in reducing con- ditions of irradiated lysozymes. (A) Hen egg white, (B) turkey egg white, (C) tortoise egg white, (D) human milk lysozymes. D, Dimer; L,lyoszyme;F,fragment. 3568 M. Stuart-Audette et al. (Eur. J. Biochem. 270) Ó FEBS 2003 are obtained first, and the stoichiometry N 3 Æ /Trp Æ is observed. This oxidation is always followed by a reaction, in which Trp Æ disappears and TyrO Æ is formed. The limiting step is intramolecular, as it was already shown for peptides and proteins. The rate constants differ by a factor of % 8 with the protein (Table 1). For all proteins the LRET is not stoichiometric. In addition the TyrO Æ yield varies with the protein. These apparently low TyrO Æ yields could be due to side reactions in which Trp Æ is lost by some other path. Although Trp Æ is a relatively good oxidant because of its reduction potential, there has been no report of Trp Æ oxidizing any amino acid other than tyrosine. For example, one could suggest that the disulfide bond could be the electron donor to the indolyl tryptophan radical. However, the reduction potential of the couples TrpH +Æ /Trp (pH < 4.5) and Trp Æ /Trp (neutral pH) are, respectively, 1.15 V and 1.05 V [13], whereas that of the RSSR +Æ /RSSR is % 1.3 V (D. Armstrong, University of Calgary, Canada, personal communication). This difference in reduction potentials makes the oxidation of a disulfide bond by the indolyl radical (protonated or unprotonated) highly improbable. Nevertheless, we have studied this reaction by observing the decay of the indolyl radical in the presence and absence of the disulfide model compound (lipoate anion) and found no effect (data not shown). For the first time in this kind of study, we tried to determine the nature of the final products in HEWL and TEWL. We were expecting mostly dimers, since it is known that the fate of tyrosyl radicals is dimerization. Surprisingly, we found in addition to dimers, oxidized forms (without great change of the molecular mass) and fragments. The site of fragmentation is the peptide bond between Asn103 and Gly104. The fragmentation is highly sequence specific: if Asn–Gly sequence is not present, the protein is not cleaved. The N-terminal amino acid of the shorter fragment could be detected by sequencing, showing that the amino group is intact. Since no hydrolysis is observed without radiolytic oxidation, the fragmentation process appears to be linked to the oxidation. To our knowledge, for the first time, the participation of residues other than aromatic in oxidative processes is demonstrated. Tryptophanyl radicals might reduce the peptide bond which would then undergo fragmentation, in agreement with [34]. Indeed, we found that Trp108 and/or Trp111, which are the closest to the Asn103–Gly104 sequence, are oxidized. We have investi- gated the possibilities of creation of a free radical site, positive or negative, in this sequence by methods of quantum chemistry [35]. The aim was to see if any of this Fig. 4. MALDI-TOF mass spectra of fragments from irradiated lyso- zymes. Lysozymes were reduced by 2-mercaptoethanol and fragments were isolated by HPLC. (A) Hen egg white, (B) turkey egg white lysozymes. Fig. 3. Separation of tryptic fragments of HEWL after aminoethyla- tion. Detection at 214 nm *New fragments formed by irradiation. Nonirradiated; irradiated (100 Gy). Table 2. Yields (G-values) of polypeptide chain fragmentation for hen, turkey and tortoise lysozymes. Lysozyme Fragmentation yield (molÆJ )1 ) Percent of azide radicals (yield 5.5 · 10 )7 ÆmolÆJ )1 ) Hen (1.42 ± 0.03) · 10 )8 2.5 Turkey (0.68 ± 0.03) · 10 )8 1.2 Tortoise 0.0 0 Ó FEBS 2003 Pulse radiolysis oxidation of c-type lysozymes (Eur. J. Biochem. 270) 3569 free radical would lead to peptide bond elongation, a prerequisite for bond breakage. Such a modification was found for the radical anion. This would be in agreement with intramolecular reduction of this sequence. Attempts to identify the position of oxidized residues were performed with HEWL. Tryptic hydrolysis followed by identification of the fragments and fluorescence detec- tion, indicated that: (a) Tyr53 and Tyr23 are oxidized; (b) Trp108 and/or Trp111 are oxidized. Typ53 is very close to Trp62 and Trp63. Bobrowski et al. [36] have shown that residues of HEWL Trp62 are part of the initial process yielding % 50% of the observed LRET. They proposed that these radical residues are reduced by Tyr53, and our results are in agreement with this hypothesis. Based on the first- order rate constants measured in the model peptide compounds with oligoprolines as spacers and using these results as a ruler for the proteins, the value of 120 s )1 found for HEWL and TEWL suggests a separation distance between donor and acceptor of % 1.4 nm [7,8,10] and thus a similar pattern of oxidation. Indeed, the similarity of NOE data, hydrogen-exchange rates, chemical shifts and coupling constants of the two proteins are indicative that structures of HEWL and TEWL are essentially identical [37]. It should be noted that the presence of Tyr at position 3 did not affect the rate constant suggesting that this residue does not participate in the intramolecular process. Nevertheless, dimerization of Tyr23 could come also from Trp62 and/or Trp63. However Trp108 and Trp111 are at a closer distance and since they remain oxidized, we conclude that no intramolecular electron transfer was initiated from these residues. It seems that the argument of distance is not sufficiently explanatory. In conclusion, the long-range intramolecular electron transfer process appears to be much more complicated than was previously thought. Some tryptophan residues do not lead to LRET and no explanation comes from examination of the distances. Some tyrosine residues (such as Tyr20 and Tyr3 in TEWL) are not involved in the process and similarly the argument of distances does not provide explanations. In addition, an unexpected fragmentation site is observed: the Asn103–Gly104 sequence. More careful investigations of LRET are needed, including analysis of the final com- pounds, to reach an understanding of the process. Acknowledgements M.F. thanks the Curie Institute, the Yvette Mayent award and Dr D. Lavalette, director of the INSERM unit 350 for his hospitality. We are indebted to the Curie Institute and especially to Dr V. Favaudon for the use and maintenance of the linear accelerator and for fruitful discussions. References 1. Grossweiner, L.I. (1976) Photochemical inactivation of enzymes. Curr. Top. Radiat. Res. Q. 11, 141–199. 2. Klapper, M.H. & Faraggi, M. (1979) Applications of pulse radio- lysis to protein chemistry. Q. Rev. Biophys. 12, 465–519. 3. Houe ´ e-Levin, C. & Sicard-Roselli, C. (2001) Radiation chemistry of proteins. In Radiation Chemistry: Present Status and Future Prospects (Jonah, C. & Rao, B.M., eds), pp. 553–584. Elsevier, New York. 4. Feng,C.,Kedia,R.V.,Hazzard,J.T.,Hurley,J.K.,Tollin,G.& Enemark, J.H. (2002) Effect of solution viscosity on intra- molecular electron transfer in sulfite oxidase. Biochemistry 41, 5816–5821. 5. Filipe, P., Morliere, P., Patterson, L.K., Hug, G.L., Maziere, J C., Maziere,C.,Freitas,J.P.,Fernandes,A.&Santus,R.(2002) Repair of amino acid radicals of apolipoprotein B100 of low- density lipoproteins by flavonoids. A pulse radiolysis study with quercetin and rutin. Biochemistry 41, 11057–11064. 6. Feng, C., Wilson, H.L., Hurley, J.T., Hazzard, J.K., Tollin, G., Rajakopalan, K.V. & Enemark, J.H. (2003) Role of conserved tyrosine 343 in intramolecular electron transfer in human sulfite oxidase. J. Biol. Chem. 278, 2913–2920. 7. Faraggi, M., DeFelippis, M.R. & Klapper, M.H. (1989) Long- range electron transfer between tyrosine and tryptophan in pep- tides. J. Am. Chem. Soc. 111, 5141–5145. 8. DeFelippis, M.R., Faraggi, M. & Klapper, M.H. (1990) Evidence for through-bond long range electron transfer in peptides. J. Am. Chem. Soc. 112, 5640–5642. 9. Lee, H., Faraggi, M. & Klapper, M.H. (1992) Long range electron transfer along an alpha-helix. Biochim. Biophys. Acta 1159, 286–294. 10. Bobrowski, K., Wierzchowski, K.L., Holcman, J. & Ciurak, M. (1990) Intramolecular electron transfer in peptides containing methionine, tryptophan and tyrosine: a pulse radiolysis study. Int. J. Radiat. Biol. 57, 919–932. 11. Pru ¨ tz, W.A., Butler, J., Land, E.J. & Swallow, A.J. (1980) Direct demonstration of electron transfer between tryptophan and tyro- sine in proteins. Biochem. Biophys. Res. Commun. 96, 513–520. 12. Adams, G.E., Wilson, R.L., Aldrich, J.E. & Cundall, R.B. (1969) On the mechanism of the radiation-induced inactivation of lyso- zyme in dilute aqueous solution. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 16, 333–342. 13. DeFelippis, M.R., Murthy, C.P., Faraggi, M. & Klapper, M.H. (1989) Pulse radiolytic measurement of redox potentials: the tyro- sine and tryptophan radicals. Biochemistry 8, 4847–4853. 14. Butler, J., Land, E.J., Prutz, W.A. & Swallow, A.J. (1982) Charge transfer between tryptophan and tyrosine in proteins. Biochim. Biophys. Acta 705, 150–162. 15. Weinstein, M., Alfassi, Z.B., DeFelippis, M.R., Klapper, M.H. & Faraggi, M. (1991) Long range electron transfer between tyrosine and tryptophan in hen egg-white lysozyme. Biochim. Biophys. Acta 1076, 173–178. 16. Bobrowski, K., Holcman, J. & Wierzchowski, K.L. (1989) Temperature dependence of intramolecular electron transfer as a probe for predenaturational changes in lysozyme. Free Rad. Res. Commun. 6, 235–241. 17. Prager, E.M. & Jolle ` s, P. (1996) Animal lysozymes c and g: an overview. In Lysozymes: Model Enzymes in Biochemistry and Biology (Jolle ` s, P. ed.), pp. 9–31. Birkhauser-Verlag, Basel. 18. Smith, L.J., Sutcliffe, M.J., Redfield, C. & Dobson, C.M. (1993) Structure of hen lysozyme in solution. J. Mol. Biol. 229, 930–944. 19. Howell, P.L., Almo, S.C., Parsons, M., Petsko, G.A. & Hajdu, J. (1991) Structure determination of turkey egg-white lysozyme using Laue diffraction data. Acta Crystallogr. Sect. B Struct. Sci. 48, 200–207. 20. Artymiuk, P.J. & Blake, C.C.P. (1981) Refinement of human lysozyme at 1.5 A resolution analysis of non-bonded and hydrogen-bond interactions. J. Mol. Biol. 152, 737–762. 21. Redfield, C. & Dobson, C.M. (1990) 1 H NMR studies of human lysozyme: spectral assignment and comparison with hen lysozyme. Biochemistry 29, 7201–7214. 22. Acharya, K.R., Stuart, D.I., Phillips, D.C., McKenzie, H.A. & Teahan, C.G. (1994) Models of the three-dimensional structures of 3570 M. Stuart-Audette et al. (Eur. J. Biochem. 270) Ó FEBS 2003 echidna, horse, and pigeon lysozymes: calcium-binding lysozymes and their relationship with alpha-lactalbumins. J. Protein Chem. 13, 569–584. 23. Jolle ` s, P. & Jolle ` s, J. (1984) What’s new in lysozyme research? Mol. Cell Biochem. 63, 165–189. 24. Schagger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166,368– 379. 25. Gill, J.C. & von Hippel, P.H. (1989) Calculation of protein extinction coefficients from amino acid sequence data. Anal. Bio- chem. 182, 319–326. 26. Favaudon, V., Tourbez, H., Houe ´ e-Levin, C. & Lhoste, J M. (1990) CO2 .– radical induced cleavage of disulfide bonds in pro- teins. A gamma-ray and pulse radiolysis mechanistic investigation. Biochemistry 29, 10978–10989. 27. Farhataziz, Rodgers, M.A.J. (1987) Radiation Chemistry. Princi- ples and Applications. VCH, Weinheim. 28. Spinks, J.W.T. & Woods, R.J. (1990) Introduction to Radiation Chemistry, 3rd edn. Wiley, New York. 29. Bensasson, R.V., Land, E.J. & Truscott, T.G. (1983) Flash Photolysis and Pulse Radiolysis. Pergamon Press, Oxford. 30. Dudley, F.D., Santus, R. & Grossweiner, L.I. (1975) Laser flash photolysis of aqueous tryptophan. J. Phys. Chem. 79, 2711–2716. 31. Audette, M., Blouquit, Y. & Houe ´ e-Levin, C. (2000) Oxidative dimerization of proteins: role of tyrosine accessibility. Arch. Biochem. Biophys. 76, 673–681. 32. Malencik, Q.A., Sprouse, J.F., Swanson, C.A. & Anderson, S.R. (1996) Dityrosine: preparation, isolation, and analysis. Anal. Biochem. 242, 202–213. 33. Van Wickern, B., Simat, T. & Steinhart, H. (1997) HPLC analysis of c-irradiation-induced products of tryptophan in peptides and lysozyme. Z. Lebensm. Unters Forsch. A. 205, 446–451. 34. Garrison, W.M. (1987) Reaction mechanisms in the radiolysis of peptides, polypeptides, and proteins. Chem. Rev. 87, 381–398. 35. Houe ´ e-Levin, C. & Berge ` s, J. (2002) A DFT study of electron and hole localisation in a peptide containing asparagin. Eur. Phy. J. D 20, 551–555. 36. Bobrowski, K., Holcman, J., Poznanski, J. & Wierzchowski, K.L. (1997) Pulse radiolysis studies of intramolecular electron transfer in model peptides and proteins. 7. Trp fi TyrO radical transfor- mation in hen egg-white lysozyme. Effects of pH, temperature, Trp62 oxidation and inhibitor binding. Biophys. Chem. 63, 153–166. 37. Bartik,K.,Dobson,C.M.&Redfield,C.(1993) 1 H-NMR analysis of turkey egg-white lysozyme and comparison with hen egg-white lysozyme. Eur. J. Biochem. 215, 255–266. Ó FEBS 2003 Pulse radiolysis oxidation of c-type lysozymes (Eur. J. Biochem. 270) 3571 . Re-evaluation of intramolecular long-range electron transfer between tyrosine and tryptophan in lysozymes Evidence for the participation of other residues Marilyne. observed in all peptides containing tyrosine 23 or 53 but not in peptides containing tyrosine 20 alone. According to these results, tyrosines 23 and 53 could