Does different orientation of the methoxy groups of ubiquinone-10 in the reaction centre of Rhodobacter sphaeroides cause different binding at Q A and Q B ? Andre ´ Remy 1 , Rutger B. Boers 2 , Tatiana Egorova-Zachernyuk 2 , Peter Gast 3 , Johan Lugtenburg 2 and Klaus Gerwert 1 1 Lehrstuhl fu ¨ r Biophysik, Ruhr-Universita ¨ t Bochum, Germany; 2 Department of Chemistry, Gorlaeus Laboratories, Leiden University, the Netherlands; 3 Department of Biophysics, Huygens Laboratory, Leiden University, the Netherlands The different roles of ubiquinone-10 (UQ 10 ) at the primary and secondary quinone (Q A and Q B ) binding sites of Rho- dobacter sphaeroides R26 reaction centres are governed by the protein microenvironment. The 4C¼O carbonyl group of Q A is unusually strongly hydrogen-bonded, in contrast to Q B . This asymmetric binding seems to determine their dif- ferent functions. The asymmetric hydrogen-bonding at Q A can be caused intrinsically by distortion of the methoxy groups or extrinsically by binding to specific amino-acid side groups. Different X-ray-based structural models show con- tradictory orientations of the methoxy groups and do not provide a clear picture. To elucidate if distortion of the methoxy groups induces this hydrogen-bonding, their (ring-)C-O vibrations were assigned by use of site-specifically labelled [5- 13 C]UQ 10 and [6- 13 C]UQ 10 reconstituted at either the Q A or the Q B binding site. Two infrared bands at 1288 cm )1 and 1264 cm )1 were assigned to the methoxy vibrations. They did not shift in frequency at either the Q A or Q B binding sites, as compared with unbound UQ 10 .Asthe frequencies of these vibrations and their coupling are sensi- tive to the conformations of the methoxy groups, different conformations of the C(5) and C(6) methoxy groups at the Q A and Q B binding sites can now be excluded. Both methoxy groups are oriented out of plane at Q A and Q B . Therefore, hydrogen-bonding to His M219 combined with electrostatic interactions with the Fe 2+ ion seems to determine the strong asymmetric binding of Q A . Keywords: electron transfer; Fourier-transform infrared spectroscopy; isotopic labelling; photosynthetic reaction centre; ubiquinone. The photosynthetic reaction centre (RC) of the purple nonsulphur bacterium Rhodobacter sphaeroides is a trans- membrane pigment–protein complex, the structure of which has been determined with up to 2.2 A ˚ resolution [1–4]. Upon light excitation, an electron is transferred from the primary donor P (bacteriochlorophyll a dimer) via a monomeric bacteriochlorophyll a and a bacteriopheo- phytin a molecule to the primary quinone Q A and finally to the secondary quinone Q B . Although ubiquinone-10 (UQ 10 ) is found at Q A and Q B , the two molecules differ in function: Q A is tightly bound to the RC. By accepting one electron, a semiquinone anion radical Q A –• is created which quickly transfers the electron to Q B .Q B is less tightly bound. After the formation of a nonprotonated semiqui- none anion radical Q B –• , a second electron and two protons are accepted here to form a hydroquinone (Q B H 2 ), which is finally released from the RC; for a recent review see [5]. To elucidate the protein–cofactor interactions that deter- mine the different functions of UQ 10 at Q A and Q B , Fourier- transform infrared (FTIR) difference spectroscopy has been applied [6–9]. By the use of UQ 10 specifically 13 C-labelled at the ring positions 1, 2, 3, and 4, the 1C¼Oand4C¼Oand 2/3C¼C stretching vibrations of UQ 10 in the RC have been assigned in the Q A – ) Q A and Q B – ) Q B difference spectra [10–13]. At the Q A site, the mode dominated by the 4C¼O vibration is dramatically downshifted compared with unbound UQ 10 , indicating unusually strong hydrogen- bonding to the protein environment [10,11]. In contrast, the 1C¼O group is only weakly bound to the protein. This asymmetric binding is conserved in the charge-separated state [10,11]. At the Q B site, two fractions of UQ 10 are found. The minor fraction is loosely bound and almost unaffected by the protein. In the major fraction, both C¼O vibrations show symmetric hydrogen-bonding, but weaker than the hydrogen bond of 4C¼OattheQ A site [12,13]. These results for the charge-separated state are supported by EPR [14] and NMR spectroscopy [15]. It is proposed that this difference in binding governs the different roles of UQ 10 at the Q A and Q B sites. However, the molecular origin of the strong binding of the 4C¼O group is not clear. The conformation of the C(5) and C(6) methoxy substituents of UQ 10 may differ at both binding sites as Correspondence to K. Gerwert, Lehrstuhl fu ¨ r Biophysik, Ruhr- Universita ¨ t Bochum, Postfach 102148, 44780 Bochum, Germany. Fax: + 49 234 321 4626, Tel.: + 49 234 322 4461, E-mail: gerwert@bph.ruhr-uni-bochum.de Abbreviations: FTIR, Fourier-transform infrared; IR, infrared; LDAO, lauryldimethylamine N-oxide; Q A , primary acceptor quinone; Q B , secondary acceptor quinone; Rb., Rhodobacter; RC, reaction centre; UQ 10 , ubiquinone-10. (Received 14 April 2003, revised 25 June 2003, accepted 8 July 2003) Eur. J. Biochem. 270, 3603–3609 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03746.x predicted by theoretical studies [16–19], and the different conformations would then lead to a shift in electron density towards the 4C¼O group, which weakens the 4C¼O bond order. We present Q A – ) Q A and Q B – ) Q B difference spectra and IR spectra of unlabelled and site-specifically 13 C-labelled UQ 10 at the C5 and C6 positions. Thereby, the correspond- ing (ring-)C-O vibrations are clearly assigned. The implica- tions for the C(5) and C(6) methoxy conformations at Q A and Q B will be discussed. Materials and methods UQ 10 , selectively 13 C-labelled at positions C5 and C6, was synthesized [20]. RC protein was purified from Rb. sphaero- ides strain R26 [21]. Either the native UQ 10 was removed from the Q B site or the native UQ 10 was removed from the Q A and Q B sites [22]. The Q A and Q B contents were determined by fitting the recombination kinetics at 865 nm after photobleaching of the primary donor to a sum of two exponentials: A ¼ A 1 exp(–k 1 t 1 )+A 2 exp(–k 2 t 2 ). Upon normalization of the amplitudes A 1 (fast Q A – decay) and A 2 (slow Q B – decay) (A 1 + A 2 ¼ 100%), the fraction of func- tionally bound secondary quinone was obtained. In the case of Q A reconstitution, the occupancy of the Q A site was analysed by measuring the photobleaching at 865 nm before and after addition of a 100-fold excess of UQ 0 (¼ 100% activity) and comparing the amplitudes. The Q A and Q B contents after reconstitution were always better than 85%. Samples were prepared for the IR measurements as previously described [10]. A 45 lL portion of 40 l M RCs, dissolved in buffer [10 m M Tris/HCl, 1 m M EDTA, 0.025% (w/v) lauryldimethylamine N-oxide (LDAO), pH 8] was pipetted on a CaF 2 window,  10-fold concentrated under a gentle stream of nitrogen, and mixed with 5 lL10m M sodium ascorbate/20 m M diaminodurene dissolved in the same buffer as the RCs. After further careful drying to a final volume of  1 lL, the sample was sealed with another CaF 2 window and thermostabilized at 283 K or 295 K (for Q A – ) Q A or Q B – ) Q B difference spectra, respectively) in the FTIR apparatus. The ubiquinones were dissolved in n-pentane and deposited on a CaF 2 window. After evapor- ation of n-pentane, the remaining UQ 10 film was measured in the IR. IR spectra of unbound ubiquinones, Q A – ) Q A difference spectra and Q B – ) Q B difference spectra were recorded as reported [10,23,24]. Spectral resolution was 4 cm )1 . Double difference spectra were computed as described [10]. The difference spectra with unlabelled and 13 C-labelled UQ 10 at the Q A or Q B sites were normalized on the 1800– 1700 cm )1 region, which was unaffected by the labelling. The IR spectra of the unbound UQ 10 were normalized on the 1500–1350 cm )1 region, which was unaffected by the labelling. Results and discussion To investigate the influence of the protein microenviron- ment on UQ 10 at the Q A and Q B binding sites, first the vibrational modes of the unbound UQ 10 were determined. FTIR spectra of pure UQ 10 areshowninFig.1inthe spectral range in which methoxy vibrations are expected. The spectrum of unlabelled UQ 10 (Fig. 1a) agrees with the one published [25]. The absorption spectrum of [5- 13 C] UQ 10 is displayed in Fig. 1b. Isotopic labelling induces a frequency shift of the absorption of the labelled group to lower wave numbers and thereby allows unequivocal band assignment. Apart from this, bands of nearby groups, the vibrational modes of which are coupled to the vibrations of the labelled group, may also be shifted. In fact, various band shifts of C¼CandC¼O vibrations, which are coupled to the (ring-)C-O vibrations of the methoxy groups, occur (spectral range not shown). This is in agreement with previous assignments of C¼CandC¼O vibrations [10,11]. A detailed discussion of the C¼CandC¼O vibrations is beyond the scope of this paper and will be given elsewhere. Here we focus on the methoxy vibrations only. The strong bands at 1447, 1434 (shoulder) and 1380 cm )1 are almost unaffected by the labelling, whereas the bands at 1287 and 1263 cm )1 are downshifted to 1283 and 1254 cm )1 , respectively. This is visualized in the double difference spectrum (Fig. 1d). If the spectra of unlabelled and specifically 13 C-labelled UQ 10 are subtracted as described [10], all unshifted bands should in principle disappear, and only the shifted bands appear as difference signals in the double difference spectrum. The two band shifts described above occur at 1288/1277 cm )1 and 1264/1252 cm )1 . The down- shift from 1264 to 1252 cm )1 is obvious, whereas the Fig. 1. IR absorption spectra of (a) unlabelled UQ 10 ,(b)[5- 13 C]UQ 10 , and (c) [6- 13 C]UQ 10 and (d) difference b ) a and (e) difference c ) a. Inset: structure of site-specifically labelled UQ 10 . 3604 A. Remy et al.(Eur. J. Biochem. 270) Ó FEBS 2003 downshift from 1288 to 1277 cm )1 is just above the resolu- tion. However, even though the band is small, the shift is highly reproducible. The absorption spectrum of [6- 13 C]UQ 10 (Fig. 1c) is similar to that of [5- 13 C]UQ 10 (see Fig. 1b), but the band shifts are slightly different. The band at 1287 cm )1 shifts to 1280 cm )1 [in the double difference spectrum (Fig. 1e) 1288/1274 cm )1 ], and the band at 1264 cm )1 shifts to 1252 cm )1 . From the observed frequency shifts caused by site-specific isotopic labelling, the bands at 1288 and 1263 cm )1 were assigned to C(5) and C(6) methoxy vibrations of UQ 10 . To assign the methoxy vibrations at the Q A binding site, Q A – ) Q A difference spectra of Rb. sphaeroides RCs recon- stituted with unlabelled and site-specifically labelled UQ 10 were measured (Fig. 2). The differences between the charge- separated and the ground state absorption selectively represent the light-induced absorption changes of the RCs. Positive bands belong to the charge-separated state, and negative signals to the ground state. The Q A – ) Q A difference spectrum of unlabelled UQ 10 (Fig. 2a) agrees well with the one published [23]. The Q A – ) Q A difference spectrum of [5- 13 C]UQ 10 is displayed in Fig. 2b. As for unbound UQ 10 , various band shifts of coupled C¼CandC¼OvibrationsofQ A occur, in agreement with previous assignments [10,11] (spectral range not shown). As both ground state and charge- separated state contribute to the Q A – ) Q A difference spectra, the coupled C-O – vibration of Q A – at 1486 cm )1 [10,11] is also affected. This is not obvious in the Q A – ) Q A difference spectra (Fig. 2b,c), but resolved in the double difference spectra (Fig. 2d,e). Unexpectedly, a positive signal occurs in the double difference spectra, whereas a Q A – vibration should cause a negative one. The C-O – vibration of Q A – , however, shows highly coupled behaviour on isotopic labelling, as described and dis- cussed previously [10,11,26]. Moreover, the present study focuses on the methoxy vibrations, and because of the lack of labelling effects in this region in the spectra of unbound UQ 10 , any contributions of methoxy vibrations to this band are most unlikely. Two negative bands at 1287 and 1263 cm )1 are downshifted due to [5- 13 C]UQ 10 , to 1273 and 1254 cm )1 , respectively. The double differ- ence spectrum (Fig. 2d) shows respective downshifts of these bands from 1288 to 1277 cm )1 and from 1263 to 1254 cm )1 . These effects are due to the ground state of Q A . In principle, contributions of the semiquinone state Q A – may also occur in the difference spectra, but their frequencies are probably below 1000 cm )1 ,sotheyare not observed in this study. The Q A – ) Q A difference spectrum of [6- 13 C]UQ 10 is displayed in Fig. 2c. This spectrum is similar to that of [5- 13 C]UQ 10 at the Q A site (see Fig. 2b). As in the case of [5- 13 C]UQ 10 , the same band shifts down to 1273 and 1254 cm )1 occur. In the double difference spectrum (Fig. 2e), the bands at 1287 and 1263 cm )1 are downshifted to 1274 and 1254 cm )1 , respectively. Therefore, the bands at 1287/88 and 1263 cm )1 are assigned to C(5) and C(6) methoxy vibrations of UQ 10 at the Q A binding site. This assignment agrees with the methoxy vibrations of unbound UQ 10 . The Q B – ) Q B difference spectrum of Rb. sphaeroides RCs reconstituted with unlabelled UQ 10 is displayed in Fig. 3a. It agrees well with the one published [24]. The Q B – ) Q B difference spectrum of [5- 13 C]UQ 10 is shown in Fig. 3b. As for Q A – ) Q A , band shifts of coupled C¼C, C¼OandC-O – (1479 cm )1 ) vibrations occur in agreement with former assignments [12,13] (spectral range partially not shown). As for Q A – ) Q A ,intheQ B – ) Q B difference spectrum also two negative bands are down- shifted due to [5- 13 C]UQ 10 from 1290 to 1277 cm )1 and from 1264 to 1253 cm )1 . This is better visualized in the double difference spectrum below (Fig. 3d). The double difference spectrum shows downshifts from 1289 to 1277 cm )1 and from 1265 to 1252 cm )1 . The Q B – ) Q B difference spectrum of [6- 13 C]UQ 10 is displayed in Fig. 3c. This spectrum is similar to that of [5- 13 C]UQ 10 at the Q B site (see Fig. 3b), and the same band shifts to 1277 and to 1253 cm )1 are seen. Also in the double difference spectrum (Fig. 3e) the bands at 1288 and 1265 cm )1 are downshifted to 1277 and 1252 cm )1 , respectively. Therefore, the bands at 1288/89 and 1265 cm )1 are assigned to C(5) and C(6) methoxy vibrations of UQ 10 at the Q B binding site. This is in agreement with the assignment of the methoxy vibrations of UQ 10 at the Q A site and of the unbound UQ 10 . Fig. 2. Q A – ) Q A difference spectra of Rb. sphaeroides RCs reconstitu- ted with (a) unlabelled UQ 10 ,(b)[5- 13 C]UQ 10 , and (c) [6- 13 C]UQ 10 at the Q A site and (d) double difference b ) a and (e) double difference c ) a. Inset: structure of site-specifically labelled UQ 10 . Ó FEBS 2003 Assignment of methoxy vibrations of ubiquinone-10 (Eur. J. Biochem. 270) 3605 Conclusion Assignment of methoxy vibrations In the spectra presented, the bands at 1288 and 1264 cm )1 show frequency shifts due to labelling at the C5 or C6 position, whereas all the other ring carbon vibrations show smaller shifts or do not shift at all [10–13,26]. The bands at 1288 and 1264 cm )1 can now unambiguously be assigned to (ring-)C-O vibrations of the C(5) and C(6) methoxy groups. The vibration at 1264 cm )1 has been assigned to a C-C-ring vibration by normal mode analysis [27,28]. In contrast, Breton et al.proposedthisbandtobe a combined (ring-)C-C vibration and C-O vibration of the methoxy groups [29]. The latter proposal agrees with our results. We did not observe an isolated vibration, but both methoxy vibrations were coupled to various (ring-)C-C vibrations. As the band at 1264 cm )1 shows larger shifts due to isotopic labelling of the other ring carbons than the band at 1288 cm )1 [10–13,26], we conclude that the vibration at 1264 cm )1 is more strongly coupled than the vibration at 1288 cm )1 . Interestingly, on labelling one of the methoxy- bearing carbons, both bands shift. This indicates that the vibrations of both methoxy groups are strongly coupled and cannot be distinguished. That these two bands do not shift on exchanging the methoxy substituents into one or two ethoxy groups [30] excludes a significant contribution of the O-CH 3 vibrations and thus favours the assignment to the (ring-)C-O stretch- ing mode as the dominant mode at 1288 and 1264 cm )1 . The C-O-C bending and O-C-H bending vibrations may also contribute to these bands. However, the clear shifts show that the (ring-)C-O vibration is the dominating mode, as expected by normal mode analysis (M. Nonella, P. Tavan, personal communication, referring to [31]). In this normal mode analysis work [31], only the C¼Cand C¼O vibrations of the quinones in the RC are reported, but the calculations include the methoxy vibrations of the quinones (M. Nonella, P. Tavan, personal communica- tion), which are useful for the conclusions drawn in this work. Therefore reference [31] is quoted in combination with the cross reference to the personal communication to make clear that our conclusions are not only based on the published data [31], but also on the information commu- nicated by M. Nonella and P. Tavan which complements the published calculations [31]. The two bands at 1450 and 1436 cm )1 have been proposed to arise from CH 3 and CH 2 deformation vibra- tions of the isoprenoid chain [32]. As they disappear in duroquinone, which is lacking the methoxy groups, they have tentatively been assigned to the O-CH 3 vibration of the methoxy groups [29]. Duroquinone, however, lacks not only the methoxy groups, but also the whole isoprenoid chain. FTIR difference spectroscopy using specifically labelled UQ 10 revealed that only the isoprenoid chain is responsible for both bands at 1450 and 1436 cm )1 [33]. In addition, our measurements of [5- 13 C]UQ 10 and [6- 13 C]UQ 10 do not show any shift of these bands due to isotopic labelling and thus support the latter assignment. Implications for the binding of UQ 10 at the Q A and Q B binding sites How can chemically identical molecules take over different functions in the RC? FTIR difference spectroscopy identi- fied a large downshift of the 4C¼O stretching vibration of Q A by  60 cm )1 [10,11], indicating strong asymmetric binding of UQ 10 at the Q A site, in contrast to symmetric, weaker binding of UQ 10 at the Q B site [12,13]. To explain the difference in binding, it has been proposed that the conformation of the two methoxy substituents is sterically hindered at the Q A site, and therefore electrostatic and/or steric interactions between one methoxy group and the oxygen at C4 lower the binding order of the carbon C4 and strengthen the downshift of the 4C¼O mode [10,11]. In principle, X-ray-based structural models of the reaction centre of Rb. sphaeroides should provide the methoxy orientations of UQ 10 at the Q A and Q B binding sites. However, there are a large variety of contradicting confor- mations in the different structural models: in refs [1,2] one methoxy group is shown within the plane of the quinone ring and the other out of the plane, as shown in Fig. 4 (upper part), whereas refs [3,34–37] show both methoxy substituents in out-of-plane conformations (Fig. 4, lower part). However, even within one type there are several variations. The downwards or upwards orientation within these two classes, in-plane/out-of-plane and out-of-plane/out-of-plane, Fig. 3. Q B – ) Q B difference spectra of Rb. sphaeroides RCs reconstitu- ted with (a) unlabelled UQ 10 ,(b)[5- 13 C]UQ 10 , and (c) [6- 13 C]UQ 10 at the Q A site and (d) double difference b ) a and (e) double difference c ) a. Inset: structure of site-specifically labelled UQ 10 . 3606 A. Remy et al.(Eur. J. Biochem. 270) Ó FEBS 2003 respectively, is found in all permutations within the different structural models. From the contradictory picture of the different structural models with regard to the methoxy orientations, we conclude that the resolution of the X-ray- based structural models of the RC is too low to discriminate between the different orientations of the methoxy groups. We used FTIR spectroscopy to determine these orienta- tions. The methoxy vibrations assigned surprisingly appear at almost the same frequencies in the unbound UQ 10 (1287 cm )1 , 1264 cm )1 ) and in the protein-bound UQ 10 at the Q A (1287/88 cm )1 , 1263 cm )1 )andQ B (1288/89 cm )1 , 1265 cm )1 ) binding sites. If the conformation of one or both methoxy groups were greaty affected in the protein-bound case by steric hindrance or electrostatic interactions, one would expect a clear shift in frequency compared with the unbound UQ 10 . Vibrations of proteins and of their cofac- tors are very sensitive to changes in conformation [6–9]. There have been no experimental investigations of how different methoxy group conformations influence the frequency of the methoxy vibrations. However, the effect of different conformations of the UQ 10 methoxy groups on the IR frequency have been studied in model compounds [16,17]. The calculations show that the frequencies and the coupling of the (ring-)C-O methoxy vibrations are sensitive to their orientations. As the same frequency and coupling are observed in unbound UQ 10 ,atQ A and at Q B , the methoxy groups must have the same orientation. Therefore, different orientations of UQ 10 at Q A and Q B can be excluded. If one methoxy substituent is in plane and the other is in an out-of-plane conformation relative to the quinone ring (conformation A in [2], Fig. 4), the C(5) and C(6) (ring-)C-O modes occur at different frequencies and they are not coupled (M. Nonella, P. Tavan, personal communication, referring to [31], see above). Therefore, isotopic labelling at either the C(5) or the C(6) methoxy group would lead to different shifts depending on the labelling position. This is not observed. In contrast, when both methoxy substituents are in out- of-plane positions (conformation B in [3], Fig. 4), the C(5) and C(6) (ring-)C-O vibrations of the methoxy groups are coupled and at the same position (M. Nonella, P. Tavan, personal communication, referring to [31], see above). Isotopic labelling should lead to identical shifts independent of the labelled group. This is experimentally observed in unbound UQ 10 as well as at the Q A and the Q B binding sites. Therefore, we conclude that the same conformation is present in unbound UQ 10 andinprotein-boundUQ 10 at the Q A and Q B binding sites. The agreement with the above calculations indicates that both methoxy substituents are in an out-of-plane conformation as in [3]. Furthermore, these theoretical studies confirm our suggestion that the (ring-)C-O vibration mainly contributes to the assigned methoxy vibrations. This is an example of how IR spectroscopy can give detailed local structural information which complements the data obtained by X-ray crystallography. A further FTIR approach proposes Ile M265 to be constitutive for the electrostatic interaction with UQ 10 at Q A [38], as mutation of this site to Thr or Ser leads to an upshift of about 4–5 cm )1 of the 4C¼Ovibration. It is not the methoxy group orientation, but strong binding to His M219 at Q A (Fig. 5), and His L190 at Q B combined with electrostatic interactions with the Fe 2+ ion and with further amino-acid side chains in the Q A binding niche (e.g. Ile M265) that may explain the strong binding of UQ 10 at the Q A site. Site-directed mutagenesis of these groups should provide a clear answer. Acknowledgements Drs M. Nonella and P. Tavan are acknowledged for providing unpublished information about normal mode analysis studies on ubiquinones. This work was financially supported by the Deutsche Forschungsgemeinschaft (SFB 480-C3). References 1.Feher,G.,Allen,J.P.,Okamura,M.Y.&Rees,D.C.(1989) Structure and function of bacterial photosynthetic reaction cen- tres. Nature 339, 111–116. 2. Ermler, U., Fritzsch, G., Buchanan, S. & Michel, H. (1994) Structure of the photosynthetic reaction centre from Rhodobacter sphaeroides at 2.65 A ˚ resolution: cofactors and protein–cofactor interactions. Structure 2, 925–936. Fig. 5. Orientation of UQ 10 and its methoxy groups at the Q A and Q B binding sites [3] including the connecting His L190–Fe 2+ –His M219 complex. Fig. 4. UQ 10 (isoprenoid chain cut) of two structural models. In the upper model, the C(5) methoxy group is out of plane, whereas the C(6) methoxy group is orientated in the ring plane [2]. In reference [3] the methoxy groups point upwards. The IR results show that the methoxy groups are orientated as proposed by reference [3] in the ground state. Ó FEBS 2003 Assignment of methoxy vibrations of ubiquinone-10 (Eur. J. Biochem. 270) 3607 3. Stowell, M.H.B., McPhilipps, T.M., Rees, D.C., Soltis, S.M., Abresch, E. & Feher, G. (1997) Light-induced structural changes in photosynthetic reaction center: implications for mechanism of electron-proton transfer. Science 276, 812–816. 4. Kuglstatter, A., Ermler, U., Michel, H., Baciou, L. & Fritzsch, G. (2001) X-ray structure analyses of photosynthetic reaction center variants from Rhodobacter sphaeroides: structural changes induced by point mutations at position L209 modulate electron and proton transfer. Biochemistry 40, 4253–4260. 5. Okamura, M.Y., Paddock, M.L., Graige, M.S. & Feher, G. (2000) Proton and electron transfer in bacterial reaction centers. Biochim. Biophys. Acta 1458, 148–163. 6. Gerwert, K. (1993) Molecular reaction mechanisms of proteins as monitored by time-resolved FTIR spectroscopy. Curr. Opin. Struct. Biol. 3, 769–773. 7. Ma ¨ ntele, W. (1996) Infrared and Fourier-transform infrared spectroscopy. In Biophysical Techniques in Photosynthesis (Amesz, J. & Hoff, A.J., eds), pp 137–160. Kluwer Academic Publishers, Dordrecht. 8. Vogel, R. & Siebert, F. (2000) Vibrational spectroscopy as a tool for probing protein function. Curr. Opin. Chem. Biol. 4, 518–523. 9. Barth, A. & Zscherp, C. (2002) What vibrations tell us about proteins. Q. Rev. Biophys. 35, 369–430. 10. Brudler, R., de Groot, H.J.M., van Liemt, W.B.S., Steggerda, W.F.,Esmeijer,R.,Gast,P.,Hoff,A.J.,Lugtenburg,J.&Ger- wert, K. (1994) Asymmetric binding of the 1- and 4-C¼Ogroups of Q A in Rhodobacter sphaeroides R26 reaction centres monitored by Fourier transform infra-red spectroscopy using site-specific isotopically labelled ubiquinone-10. EMBO J. 13, 5523–5530. 11. Breton, J., Boullais, C., Burie, J R., Nabedryk, E. & Mioskowski, C. (1994) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: assignment of the interactions of each carbonyl of Q A in Rhodobacter sphaeroides using site-specific 13 C-labelled ubiquinone. Biochemistry 33, 14378–14386. 12. Brudler, R., de Groot, H.J.M., van Liemt, W.B.S., Gast, P., Hoff, A.J., Lugtenburg, J. & Gerwert, K. (1995) FTIR spectroscopy shows weak symmetric hydrogen bonding of the Q B carbonyl groups in Rhodobacter sphaeroides R26 reaction centres. FEBS Lett. 370, 88–92. 13. Breton, J., Boullais, C., Berger, G., Mioskowski, C. & Nabedryk, E. (1995) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: symmetry of the carbonyl interactions and close equivalence of the Q B vibrations in Rhodobacter sphaeroides and Rhodopseudomonas viridis probed by isotope labeling. Biochem- istry 34, 11606–11616. 14. van den Brink, J.S., Spoyalov, A.P., Gast, P., van Liemt, W.B.S., Raap, J., Lugtenburg, J. & Hoff, A.J. (1994) Asymmetric binding of the primary acceptor quinone in reaction centers of the photo- synthetic bacterium Rhodobacter sphaeroides R26, probed with Q-band (35 GHz) EPR spectroscopy. FEBS Lett. 353, 273–276. 15. van Liemt, W.B.S., Boender, G.J., Gast, P., Hoff, A.J., Lugten- burg, J. & de Groot, H.J.M. (1995) 13 C magic angle spinning NMR characterization of the functionally asymmetric Q A binding in Rhodobacter sphaeroides R26 photosynthetic reaction centers using site-specific 13 C-labelled ubiquinone-10. Biochemistry 34, 10229–10236. 16. Nonella, M. & Bra ¨ ndli, C. (1996) Density functional investigation of methoxy-substituted p-benzoquinones: conformational analysis and harmonic force field of 2-methoxy- and 2,3-dimethoxy-1,4- benzoquinone. J. Phys. Chem. B 100, 14549–14559. 17. Burie, J R., Boullais, C., Nonella, M., Mioskowski, C., Nabe- dryk, E. & Breton, J. (1997) Importance of the conformation of methoxy groups on the vibrational and electrochemical properties of ubiquinones. J. Phys. Chem. B 101, 6607–6617. 18. Boullais, C., Nabedryk, E., Burie, J R., Nonella, M., Mioskow- ski, C. & Breton, J. (1998) Site-specific isotope labeling demon- strates a large mesomeric resonance effect of the methoxy groups on the carbonyl frequency of ubiquinones. Photosynth. Res. 55, 247–252. 19. Nonella, M., Boullais, C., Mioskowski, C., Nabedryk, E. & Breton, J. (1999) Vibrational spectrum and torsional potential of 2-methoxy-3-methyl-1,4-benzoquinone. J. Phys. Chem. B 103, 6363–6370. 20. Boers,R.B.,Gast,P.,Hoff,A.J.,deGroot,H.J.M.&Lugtenburg, J. (2002) Synthesis and spectroscopic characterization of [5- 13 C]- and [6- 13 C]-ubiquinone-10 for studies of bacterial photosynthetic reaction centers. Eur. J. Org. Chem. 2002, 189–202. 21. Feher, G. & Okamura, M.Y. (1978) The Photosynthetic Bacteria (Clayton, R.K. & Sistrom, W.R., eds), pp. 349–386. Plenum, New York. 22. Okamura, M.Y., Isaacson, R.A. & Feher, G. (1975) Primary acceptor in bacterial photosynthesis: obligatory role of ubiquinone in photoactive reaction centers of Rhodopseudomonas spheroides. Proc. Natl. Acad. Sci. USA 72, 3491–3495. 23. Breton, J., Thibodeau, D.L., Berthomieu, C., Ma ¨ ntele, W., Vermeglio, A. & Nabedryk, E. (1991) Probing the primary qui- none environment in photosynthetic bacterial reaction centers by light-induced FTIR difference spectroscopy. FEBS Lett. 278, 257–260. 24. Breton, J., Berthomieu, C., Thibodeau, D.L. & Nabedryk, E. (1991) Probing the secondary quinone (Q B ) environment in photo- synthetic bacterial reaction centers by light-induced FTIR dif- ference spectroscopy. FEBS Lett. 288, 109–113. 25. Keller, R.J. (ed.) (1986) The Sigma Library of FT-IR Spectra.Vol. 1, p. 929. Sigma Chemical Co., St Louis. 26. Remy, A. (1998) FTIR-spektroskopische und molekularbio- logische Untersuchungen zum Elektronen- und Protonentransfer im bakteriellen photosynthetischen Reaktionszentrum von Rhodo- bacter sphaeroides. Thesis, Ruhr-Universita ¨ t Bochum, Germany. 27. Becker,E.D.,Charney,E.&Anno,T.(1965)Molecularvibra- tions of quinones. VI. A vibrational assignment for p-benzoqui- none and six isotopic derivatives. Thermodynamic functions of p-benzoquinone. J. Chem. Phys. 42, 942–944. 28. Chipman, D.M. & Prebenda, M.F. (1986) Structures and fundamental vibrations of p-benzoquinone and p-benzoquinone radical anion from ab initio calculations. J. Phys. Chem. 90, 5557– 5560. 29. Breton, J., Burie, J R., Boullais, C., Berger, G. & Nabedryk, E. (1994) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: binding of chainless symmetrical quinones to the Q A site of Rhodobacter sphaeroides. Biochemistry 33, 12405– 12415. 30. Pennock, J.F. (1965) Spectroscopy of quinones and related sub- stances. II. Infrared absorption spectra and nuclear magnetic resonance spectra. In Biochemistry of Quinones (Morton, R.A., ed.), pp. 67–87. Academic Press, New York. 31. Nonella, M., Mathias, G., Eichinger, M. & Tavan, P. (2003) Structures and vibrational frequencies of the quinones in Rb. sphaeroides derived by a combined density functional/molecular mechanics approach. J. Phys. Chem. B 107, 316–322. 32. Bellamy, L.J. (1980) The infrared spectra of complex molecules. Kluwer Academic Publishers, London. 33. Brudler, R. (1996) Charakterisierung der Chinonbindestellen im photosynthetischen Reaktionszentrum von Rhodobacter sphaeroides mit Hilfe von statischer und zeitaufgelo ¨ ster FTIR-Differenzspektroskopie. PhD Thesis, Ruhr-Universita ¨ t Bochum, Germany. 34. Chang, C H., El-Kabbani, O., Tiede, D., Norris, J. & Schiffer, M. (1991) Structure of the membrane-bound protein photosynthetic 3608 A. Remy et al.(Eur. J. Biochem. 270) Ó FEBS 2003 reaction center from Rhodobacter sphaeroides. Biochemistry 30, 5352–5360. 35. Chirino, A.J., Lous, E.J., Huber, M., Allen, J.P., Schenck, C.C., Paddock, M.L., Feher, G. & Rees, D.C. (1994) Crystallo- graphic analyses of site-directed mutants of the photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 33, 4584–4593. 36. Arnoux, B., Gaucher, J.F., Ducruix, A. & Reiss-Husson, F. (1995) Structure of the photochemical reaction centre of a spheroidene- containing purple bacterium, Rhodobacter sphaeroides Y, at 3 Angstrom resolution. Acta Crystallogr. D 51, 368–379. 37. Camara-Artigas, A., Brune, D. & Allen, J.P. (2002) Interactions between lipids and bacterial reaction centers determined by pro- tein crystallography. Proc.Natl.Acad.Sci.USA99, 11055–11060. 38. Takahashi, E., Wells, T.A. & Wraight, C.A. (2001) Protein control of the redox potential of the primary quinone acceptor in reaction centers from Rhodobacter sphaeroides. Biochemistry 40, 1020– 1028. Ó FEBS 2003 Assignment of methoxy vibrations of ubiquinone-10 (Eur. J. Biochem. 270) 3609 . Does different orientation of the methoxy groups of ubiquinone-10 in the reaction centre of Rhodobacter sphaeroides cause different binding at Q A and. origin of the strong binding of the 4C¼O group is not clear. The conformation of the C(5) and C(6) methoxy substituents of UQ 10 may differ at both binding