Caged O 2 Reaction of cytochrome bo 3 oxidase with photochemically released dioxygen from a cobalt peroxo complex Claudia Ludovici, Roland Fro¨ hlich*, Karsten Vogtt, Bjo¨ rn Mamat† and Mathias Lu¨ bben Lehrstuhl fu ¨ r Biophysik, Ruhr-Universita ¨ t Bochum, Germany We developed the synthesis of the caged oxygen donor (l-peroxo)(l-hydroxo)bis[bis(bipyridyl)cobalt(III)] complex (HPBC) as nitrate salt, which has, compared with the perchlorate-form described previously [MacArthur, R., Sucheta, A., Chong, F.F. & Einarsdottir, O ¨ . (1995) Proc. Natl Acad. Sci. USA, 92, 8105–8109], greatly enhanced solubility. Now, the quantum efficiency of the photolytical release of dioxygen was determined to be 0.4 per photon at a laser wavelength of 308 nm, which was used to observe biological reactions. The X-ray structure of HPBC has been solved, and the molecular interactions of photochemically generated oxygen with cytochrome oxidase were investi- gated with optical and FT-IR spectroscopy: it acts as acceptor of electrons transferred from prereduced cyto- chrome bo 3 , the heme-copper oxidase from Escherichia coli. FT-IR spectra revealed typical absorbance difference chan- ges in the carbonyl region of cytochrome bo 3 , supported by bandshifts due to solvent isotope exchange and by assign- ment using site-directed mutants. IR difference spectra of the photooxidation reaction using the caged oxygen compound, and of the photoreduction reaction using the caged electron donor FMN, have inverted shapes. The spectroscopic sig- nals of carboxyl groups are thus equivalent in both reactions: the use of chemically produced oxygen allows the observa- tion of the ongoing molecular changes of cytochrome bo 3 oxidase under quasi-physiological conditions. Keywords: cytochrome oxidase; caged compound; FT-IR spectroscopy; oxygen, l-peroxo cobalt complex. Cytochrome oxidases are hetero-oligomeric integral mem- brane proteins that belong to the superfamily of heme- copper oxidases [1,2]. They are terminal parts of the aerobic respiratory chains of bacteria and mitochondria, and their common characteristic is the transfer of electrons from cytochrome c or ubiquinol to the acceptor substrate, molecular dioxygen [3]. Cytochrome bo 3 oxidase of Escherichia coli is a ubiquinol oxidase. It transfers electrons fromthemembranesiteviahemeb to the binuclear reaction center, which consists of a heme o plus a Cu B as redox carriers. The reaction center provides the binding site of molecular oxygen, which receives electrons and protons necessary for water formation. The electronic energy is sufficient to drive transmembrane proton transport, which is tightly coupled to the processes of oxygen reduction and of water formation [4,5]. X-ray structure data of ubiquinol oxidase from Escheri- chia coli have been recently published. The resolution of 3.5 A ˚ allows the reconstruction of the backbone but not of the amino-acid side chain conformations [6]. Detailed molecularstructuresofthecytochromec oxidases from Paracoccus denitrificans and beef heart mitochondria [6–10] have been determined. Due to their extensive sequence similarities these structures could serve as models for the ubiquinol oxidase. They allow the prediction of two different proton-translocating channels, called the K- and D-channels. The D-channel contains an array of charged or polar amino acids, and is located within two different hydrogen-bonded networks above and below the central Glu286 (numbering according to the subunits I and II of the E. coli oxidase), which interacts with the binuclear center [11]. Molecular dynamics calculations [12,13] have predicted a special role of the central Glu286, which could provide the contact between both partial networks. FT-IR difference spectroscopy, using either an electrochemical cell [14,15] or photoreduction techniques [16,17] provides information about the orientation of amino-acid side chains and about molecular interactions. The photoreduction experiments are designed in such a way, that pre-equilibrated molecules become activated by light to undergo redox changes. Out of the many functional groups present in the oxidase, only those affected by the redox transition become visible in FT-IR difference spectra. In a previous report, the band signature at 1745 cm )1 and at 1735 cm )1 occurring in redox FT-IR difference spectra of different heme-copper oxidases has been assigned to Glu286 [17]. In order to study the oxidase reaction with the natural substrate dioxygen at the molecular level with FT-IR spectroscopy, we established a caged dioxygen system that allows O 2 release via photolysis. Photoactivation of (l-peroxo)(l-hydroxo)bis[bis(bipyridyl)cobalt(III)] complex Correspondence to M. Lu ¨ bben, Lehrstuhl fu ¨ r Biophysik, Ruhr-Universita ¨ t Bochum, Universita ¨ tsstr. 150, D-44780 Bochum, Germany Fax: + 49 234 32 14626, Tel.: + 49 234 32 24465, E-mail: luebben@bph.ruhr-uni-bochum.de Abbreviations: HPBC, (l-peroxo)(l-hydroxo)bis[bis(bipyridyl)- cobalt(III)]; BC, bis(2,2¢-bipyridyl)cobalt(II). *Present address: Organisch-chemisches Institut, Universita ¨ tMu ¨ nster, Correnstraße 40, D-48149 Mu ¨ nster, Germany. Present address: Max-Planck-Institut fu ¨ rBiophysik, Heinrich-Hoffmann-Str. 7, D-60528 Frankfurt/Main, Germany. (Received 18 January 2002, revised 15 April 2002, accepted 19 April 2002) Eur. J. Biochem. 269, 2630–2637 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02944.x (HPBC) has been described previously, but the reported chemical (a perchlorate salt) had rather low solubility and the photochemical conditions were very unfavorable [18,19]. Due to the strong IR absorbance of water, it is desirable that the FT-IR samples consist of thin and highly concentrated protein films. Hence a highly soluble and stable HPBC complex had to be used in order to release enough dioxygen to circumvent the possible problem of substrate limitation. In this study, we describe the synthesis of a highly soluble HPBC salt and its (photo)chemical characterization, and we demonstrate the validity of the caged oxygen complex (HPBC-nitrate salt) as a suitable probe for FT-IR spectroscopic studies of cytochrome bo 3 . MATERIALS AND METHODS Synthesis of the HPBC-perchlorate salt Solutions were prepared of 2.33 g of Co(NO 3 ) 2 Æ6H 2 Oin 20 mL of water and of 2.5 g of 2,2¢-bipyridine in 20 mL ethanol. The pH was adjusted with 1 M NaOH to 9.2. The following reaction steps were carried out in the dark. Under stirring, oxygen was streamed into the reaction vessel for 10 min. The pH of the mixture was readjusted to 9.6, and oxygen streaming was continued for another 20 min. For crystallization, 2.2 mL of a 6 M sodium perchlorate solution in 50% (v/v) of aqueous ethanol was added and the mixture was kept for 16 at 10 °C. Black crystals were filtered with suction, washed with ice-cold ethanol and dried under vacuum for at least 3 h. The molar yield of solvent-free salt was 73%. IR(KBr): 1088 cm )1 and 625 cm )1 (perchlorate), 855 cm )1 m(O-O). UV/vis: k max : 460 nm, 395 nm (7100 M )1 Æcm )1 ), 314 nm (43 300 M )1 cm )1 ), 304 nm and 212 nm. Synthesis of the HPBC-nitrate salt Co(NO 3 ) 2 Æ6H 2 O (2.33 g) and of 2,2¢-bipyridine (2.5 g) were both dissolved separately in 30 mL ethanol. The following steps where performed in the dark. The two solutions were combined and 170 mg solid NaOH, dissolved in 215 lLof water, was added. A stream of oxygen was bubbled into the liquid for 30 s. The reaction mixture was shaken at 37 °C for each 5 min at 100, 50 and 25 revolutions per min on a rotary platform; for crystallization it was kept at 30 °C for at least 12 h in the dark. Crystals were collected on a sintered glass funnel, washed with 20 mL of ethanol and dried for 6 h by desiccation. The product (yield: 70%) was stored at )20 °C. IR(KBr): 1380 cm )1 (nitrate), 855 cm )1 m(O-O). UV/vis: k max : 460 nm, 395 nm (7100 M )1 cm )1 ), 314 nm (43 300 M )1 Æcm )1 ), 304 nm and 212 nm. Synthesis of the bis(2,2¢ -bipyridyl)cobalt(II) (BC)-perchlorate salt Co(NO 3 ) 2 Æ6H 2 O(2mLof0.2 M in water) and 2,2¢-bipyri- dine (2 mL of 0.4 M in ethanol) were mixed and 6 sodium perchlorate solution in 50% (v/v) of aqueous ethanol was added to a final concentration of 2 and the mixture was kept for 16 h at 10 °C. The yellow hexagonal crystals formed were collected as described above. UV/vis: k max :293nm. Determination of the molar yield of photolytical oxygen release HPBC-nitrate salt (20 mg) and EDTA (100 mg) were placed into a stoppered glass vessel of a total volume of about 120 mL. It was filled to the edge with bidistilled water, 1 mL of a 3 M NaI solution in 40% (w/v) NaOH and 1 mL of 40% (w/v) MnCl 2 solution were added. Another reaction vessel without added HPBC-nitrate salt served as reference. To attain completion of the photolytic reaction, the compound (which was kept in the non-UV-transmitting glass bottle) had to be irradiated with visible light emitted by a workshop-made lamp arrangement for several hours. Measurement of dissolved plus photolytically liberated oxygen was chemically determined in sample and reference mixtures according to the Winkler titration technique [20]. Crystallization and X-ray structure determination For crystal structure determination, data sets were collected with a Nonius KappaCCD diffractometer, equipped with a rotating anode generator Nonius FR591. The following computer programs were used for different steps data recording and evaluation: COLLECT for data collection (Nonius BV), DENZO - SMN for data reduction [21], SORTAV for absorption correction [22,23], SHELXS -7 for structure solution [24], SHELXL -7 for structure refinement (G. M. Sheldrick, Universita ¨ tGo ¨ ttingen, Germany), DIAMOND for the graphic display of structures (K. Brandenburg, Univer- sita ¨ t Freiburg, Germany). Determination of quantum yield of oxygen release The quantum yield of the photorelease of molecular oxygen from HPBC was determined after quantification of the photon flux emitted by a Xe lamp at different wavelengths by means of the chemical actinometer compound Aber- chrome 540 [25]. The samples were placed in 1-cm stirred cuvettes and were irradiated with monochromatic light for defined time intervals to correct for wavelength-dependent emission intensities. The numbers of incident photons and the photolytic turnover were quantified by static UV/vis spectroscopy by measurement of the absorbance changes of Aberchrome 540 dissolved in toluene at 494 nm and of HPBC dissolved in 100 m M KP i pH 7.4 at 293 nm. At high concentrations of HPBC, the absorbance change at 390 nm was also used to quantify the photolytic yield. Preparation of duroquinol-reduced samples for visible spectroscopy and FT-IR spectroscopy and recording of UV/vis spectra Cytochrome bo 3 was expressed using the vector pHCL [17] and purified using Ni-agarose chromatography exactly as described previously [26]. Optical absorbance spectra of cytochrome bo 3 in the presence of caged oxygen were performed using workshop-made CaF 2 cuvettes constructed for FT-IR spectroscopy (see below). Sample preparations were carried out under an Ar atmosphere in a plastic container (Atmosbag, Sigma) equipped with grips for better sample handling. A small volume (2.5 lL) of a 10 m M ethanolic solution of 10 m M duroquinol was pipetted on the center of a CaF 2 window, covered perimetrically with a thin Ó FEBS 2002 Caged oxygen reaction with cytochrome oxidase (Eur. J. Biochem. 269) 2631 layer of grease (Apiezon). After evaporation of the solvent the duroquinol was redissolved with 2–3 lL of a concen- trated cytochrome bo 3 solution (about 0.3 m M )in20m M Tris/HCl, pH 8.0, 0.3% (w/v) b-decylmaltoside. The mix- ture was concentrated in an Ar stream. The dried layer was rehydrated with 3 lL100m M sodium borate, 1 m M EDTA, 0.1% b-decylmaltoside, pH 9.0. Again the mixture was concentrated under Ar and it was finally redissolved by adding 0.5 lLof10m M HPBC in borate buffer. The cuvette was sealed with another CaF 2 plate, and placed into a metallic sample holder. The following cuvette handling was carried out in the aerobic atmosphere. The absorbance spectra of the mixture before and after irradiation with a 150- Xe arc lamp (Oriel) or LPX 240i excimer laser (Lambda Physics, Go ¨ ttingen) were measured with a Hitachi UV/vis spectrometer. Preparation of thiol-reduced samples of cytochrome bo 3 The samples were prepared in a workshop-constructed glass chamber equipped with plate holders for sample and counter CaF 2 plates (technical details will be described elsewhere). Volumes of 2.5–3.5 lL of 0.2–0.3 m M cyto- chrome bo 3 solution (as above) were pipetted on a greased CaF 2 sample plate and mixed with 1 lLofa20m M freshly prepared dithiothreitol solution in 20 m M Tris, pH 8.0, 0.3% (w/v) b-decylmaltoside, and the mixture was spread to a spot of 5-mm diameter. The chamber was assembled and a CaF 2 counter plate, spotted with a 0.5-lLdropof 20 m M HPBC dissolved in glycerol, was placed in position opposite to the sample plate. The chamber was evacuated for 2 min to a residual pressure of 10–50 mbar to allow dehydration of the sample and formation of a thin film, which was re-equilibrated with aqueous vapor from a water reservoir for 30 s. Sample and counter plates were then pressed together, which resulted in efficient mixing of the reduced protein with the caged oxygen compound. The cuvette was sealed airtight and kept at 4 °C until measurement. Recording of FT-IR spectra Static IR spectra were recorded with a Bruker 66V/S spectrometer, evacuated to 8–10 mbar residual pressure. The sample containment, maintained at 4 °C, was purged with dry air to minimize absorbance by water vapor. A water-cooled globar was used as source of radiation, which was measured by a nitrogen-cooled HgCdTe detector, using a low-pass filter which cut off intensity above 1975 cm )1 . The scanner mirror was moved in the single-sided mode to achieve a scan rate of 100 kHz. Spectra were measured at nominal resolution of 2 cm )1 , Mertz phase correction was adjusted and the Blackman–Harris three-term function was used for apodization. If not otherwise indicated, reference spectra of 800 coadded scans was recorded. The sample photolysis was initiated by application of 15 flashes (90– 140 mJ) of light with a pulse length of 20–30 ns at 308 nm from an LPX240i excimer laser (Lambda Physics, Go ¨ ttin- gen), and 800 scans were coadded. To verify that the redox reaction of protein molecules was complete, a second spectrum was recorded as above (without reference meas- urement) after application of another 15 laser flashes. Double difference calculations were carried out using the OPUS software. In order to normalize distinct spectra, the absorbance difference bands of caged oxygen at 1443 and 1451 cm )1 were brought to the same scale. Redox spectra with flavine mononucleotide as caged electron donor Sample preparation and recording of FTIR spectra was carried out as described previously [16]. Enzyme activity test Quinol oxidase activity using duroquinol as artificial substrate of cytochrome bo 3 was performed as described previously [26]. RESULTS Synthesis of the HPBC complex After Skurlatov [27] introduced the dibridged dinuclear complex HPBC, MacArthur used this compound as a very poor photoactivatable donor of dioxygen; the photolytic quantum yield was as low as 0.04 if the irradiation was carried out at 355 nm [18]. However, in our hands the published preparation protocol for the HPBC-perchlorate salt yielded a product that was contaminated with up to 70–80% of the mononuclear Co(II) species, BC-perchlorate. Therefore we established a highly reproducible procedure, in which the pure perchlorate salt could be obtained at > 70% molar yield. The final product could be gained readily by precipitation of the perchlorate salt; this implies that low solubility in water is an inherent property of the HPBC-perchlorate salt preparation and is a major limiting factor for the maximum oxygen concentration attainable by photo-release. For FT-IR difference spectroscopy of cytochrome oxid- ases, it is necessary to adjust high levels of molecular dioxygen; thus a derivative with much higher solubility had to be synthesized. To this purpose we prepared the nitrate salt of the HPBC complex, which is about 10 3 times more soluble in water than the perchlorate complex. Crystallization and X-ray structure determination In order to determine the HPBC-perchlorate and -nitrate structures, crystallization trials were set up by mixing solutions of HPBC-nitrate salt with various different anions such as tetrafluoroborate and perchlorate. By use of the precipitation/ether diffusion technique, well-ordered large monoclinic crystals (space group P2 1 /c) suitable for X-ray diffraction (Fig. 1) were obtained with perchlorate. Both Co centers have octahedral coordination and are connected by l-hydroxy and l-peroxo bridges. The bond distances [Co-l(O) 1.868 (± 0.005) A ˚ and 1.877 (± 0.004) A ˚ , respectively, l(O)-l(O) 1.415 (± 0.006) A ˚ ] of the bridging core are very similar to that of the corresponding ethylendi- amine complex [28] except the l(O)–l(O) distance, which is significantly smaller (at the short end of the usual range for binuclear l-peroxo complexes) [29]. The cation structures of the HPBC-nitrate salt and the -perchlorate salt complexes were identical; the nitrate complex yielded a somewhat 2632 C. Ludovici et al. (Eur. J. Biochem. 269) Ó FEBS 2002 higher R value due to disordering of the nitrate groups and solvent molecules included in the crystal (data not shown). Spectroscopical and photochemical properties of HPBC The perchlorate and nitrate salt of HPBC had identical optical absorbance spectra with maxima at 212, 304, 314, 395 nm and a shoulder at 460 nm (Fig. 2, insert), which indicates that it is the cation which determines the optical properties. In contrast to the published extinction coeffi- cientof1540 M )1 Æcm )1 at 390 nm [30], we measured a value of 7000 M )1 Æcm )1 based on a molecular mass of 977 gÆmol )1 for the trinitrate salt of the HPBC complex. If it is assumed that possible impurities might contribute to the weighted mass, an even higher numerical value of the extinction coefficient is expected. Thus the quantities of molecular oxygen reported to be photoreleased by others [18] must have been overestimated by a factor of at least 4, if calculated with the low published extinction coefficient (1540 M )1 Æcm )1 for 390 nm,  1350 M )1 Æcm )1 for 355 nm [18]). HPBC could be photolyzed efficiently by UV light from different sources, e.g. transilluminator (mercury lamp), Xe lamp or UV laser. In all cases, the end product of the photolytic reaction had absorbance maxima of 230, 293, 304 nm, identified to be the mononuclear BC. The amount of photolytically released oxygen was ascertained to be 100% by UV/vis spectros- copy. As an independent check of molar yield, the production of O 2 was determined iodometrically accord- ing to Winkler [20] to about 80%; these data confirm our revision of the published extinction coefficient as pointed out above. The experiment demonstrates that the photo- release of dioxygen from HPBC virtually has a stoichi- ometry of 1. The uncaging reaction has to be very efficient in order to make HPBC a useful photo-trigger. With the irradiation wavelength of 355 nm a quantum yield of as low as 0.04 was reported previously [18]. We expected better photolytic yields at shorter wavelengths due to the higher extinction coefficient of the compound as it becomes evident from Fig. 2. A quantum yield of 0.5 was obtained, if the oxygen release was activated by irradiation at 314 nm. To investigate the potential for measuring time-resolved reactions of the released oxygen, transient photoactivation at 308 nm was probed with a single flash from an eximer laser source. Figure 3 displays the dependence of single- shot induced product formation on the total concentra- tion of HPBC. The photolysis led to a yield of almost 100% at a concentration of 0.5 m M HPBC. This corres- ponds to the same concentration of liberated dioxygen, which caused gas bubble formation due to the limited solubility of oxygen in aqueous medium. At higher HPBC concentrations the O 2 yields decreased because the high UV absorbance leads to a pronounced Ôinner filter effectÕ of the samples. Even higher oxygen concentrations could be attained by the use of thinner cuvettes, by lowering of temperature and by variation of the solvent composition. Fig. 2. Determination of the quantum yield of photolytic reaction of HPBC and concomitant oxygen release. Samples were irradiated with monochromatic light at different times to correct for the wavelength- dependent photon fluxes. The numbers of incident photons were determined with a chemical actinometer compound. Molecular yields of HPBC photolysis were quantified spectrophotometrically, these numbers were equivalent to the amounts of oxygen released. Inset: optical spectra of HPBC before and after photolysis by continuous irradiation at 314 nm with a Xe lamp at low intensity. Fig. 1. X-ray structure of the HPBC complex. Structure analysis of the HPBC-perchlorate salt (deposited under accession no. CCDC 169345 at the Cambridge Crystallographic Data Centre): Formula C 40 H 33 N 8 O 3 CoÆ3ClO 4 ÆH 2 O, m ¼ 1107.97, black crystal with dimen- sions 0.50 · 0.30 · 0.20 mm; a ¼ 23.541(1), b ¼ 16.353(1), c ¼ 11.377(1) A ˚ , b ¼ 96.71(1)°, V ¼ 4349.8(5) A ˚ 3 , q calc ¼ 1.692gcm )3 , l ¼ 10.31 cm )1 , empirical absorption correction via SORTAV (0.627 £ T £ 0.820), Z ¼ 4, monoclinic, space group P2 1 /c no. 14); k ¼ 0.71073 A ˚ , T ¼ 198 K, x and / scans, 31572 reflections collected (± h,±k,±l), [(sinh)/k] ¼ 0.67 A ˚ )1 , 10625 independent (R int ¼ 0.038) and 9456 observed reflections [I ‡ 2 r(I )], 625 refined param- eters, R ¼ 0.097, wR 2 ¼ 0.267. The maximal residual electron density was 1.90 () 1.04) eÆA ˚ )3 in the region of the perchlorate groups; the perchlorate groups are disordered (disorder was not refined). The hydrogen on the bridging oxygen was obtained from difference Fourier calculations, other hydrogens were calculated and refined riding. Ó FEBS 2002 Caged oxygen reaction with cytochrome oxidase (Eur. J. Biochem. 269) 2633 Reaction of HPBC with cytochrome bo 3 : visible spectral region HPBC photochemistry was employed to explore the electron transfer from fully reduced cytochrome bo 3 oxidase to photo-released dioxygen; the reaction was monitored by optical absorbance spectroscopy. Because of the intention to eventually study the interactions by IR spectroscopy (see below), the experiments were carried out in FT-IR spectro- meter-type CaF 2 cuvettes at a sample thickness of equal or less than 5 lm. The visible spectrum of reduced cyto- chrome bo 3 exhibits a broad Soret peak at 425 nm and bands at 530 and 560 nm in the dark. Upon irradiation with a Xe lamp (Fig. 4) or with an excimer laser after 15 laser flashes at 308 nm with an intensity of 90–140 mJ per pulse (data not shown), typical oxidized spectra are found with the Soret peak shifted to 409 nm and absorbance loss at higher wavelengths (Fig. 4). The integrity of the protein sample after irradiation was also checked with SDS/PAGE. If HPBC is irradiated in the absence of protein, no pronounced peaks contributed within the investigated spectral region before or after photolysis. As being a prerequisite for sample stabilization during longer periods in FT-IR experiments, the spectrum of reduced cyto- chrome bo 3 oxidase in presence of HPBC remained unchanged in the dark after incubation for 48 h at 4 °C. It is now possible to observe the oxidation of the fully reduced oxidase in situ through dioxygen release from HPBC after photolysis. Reaction of HPBC with cytochrome bo 3 : IR spectral region FT-IR spectra were recorded to study the molecular details of the reaction of caged oxygen and cytochrome bo 3 .The protein was reduced with the quasi-natural substrate analog duroquinol. The samples reached a stable baseline after 2–4 h at 4 °C, and the completeness of caged oxygen photolysis was checked using bundles of 15 laser flashes. The spectrum (Fig. 5A) shows a composite of difference spectra (light ) dark) from cytochrome bo 3 plus caged oxygen before and after the photoreaction. The initial and final states of these static spectra could be classified as to oxidized cytochrome bo 3 /‘oxygen-free HPBC’ (absorbances deflecting upwards) and to reduced cytochrome bo 3 /‘oxy- gen-bound HPBC’ (absorbances deflecting downwards), as assessed by the optical spectra before and after photo- irradiation of the sample cuvette. The absorbance peaks of HPBC at 1443/1451 cm )1 and at 1600/1612 cm )1 stand out clearly. Sharp difference bands (at 1657 cm )1 , 1678 cm )1 ) Fig. 3. Yield of photolysis after single flash activation by excimer laser at 308 nm, 160 mJ per pulse. The solutions of HPBC were prepared in FT-IR type sample cuvette with 500-lm thickness. Yields were determined from the absorbance changes at 293 nm. Fig. 4. Reaction of cytochrome bo 3 fully reduced by the substrate dur- oquinol with HPBC. Absorbance spectra of reduced cytochrome bo 3 in presence of HPBC in the dark and after photolysis with Xe lamp were measured, as described in Materials and methods. The corresponding spectra of HPBC in the absence of protein are included. Fig. 5. Reaction of duroquinol-reduced cytochrome bo 3 with photore- leased dioxygen, monitored by FT-IR spectroscopy (0.85 lmol cytochrome resuspended in 20 m M Tris/HCl pH 8.0, 50 m M NaCl, 0.3% b-decylmaltoside, reduced with 2.5 nmol duroquinol. After reduction, 4.5 nmol HPBC resolved with 100 m M borate pH 9.0, 0.1% b-decylmaltoside, 1 m M EDTA was added. Photolysis conditions: 110 mJ per pulse, 308 nm, XeCl-excimer laser, spectra taken after 15 flashes. (A) The (light ) dark) difference spectrum is shown. (B) As a control, the same experiment as in (A) was carried out, except that duroquinone instead of duroquinol was added. (C) The double difference spectrum between (A) and (B) was calculated, yielding the effect of the protein reaction alone. 2634 C. Ludovici et al. (Eur. J. Biochem. 269) Ó FEBS 2002 can be seen also in the amide I region, indicating conform- ational alterations elicited by the redox transition. In the carbonyl region one can clearly distinguish positive bands at 1745 and at 1696 cm )1 . Oxidized cytochrome bo 3 equili- brated with duroquinone and HPBC was photolyzed in a control experiment (Fig. 5B). The net reaction by the caged compound itself could be measured; the difference spectra looked similar to that obtained by the pure HPBC complex itself. The prominent difference bands at 1443/1451 cm )1 and 1600/1612 cm )1 were used to scale the spectra for better comparison. In order to obtain the redox difference spectra of the protein itself, one has to subtract the background from the composite spectrum. Figure 5C displays the double difference spectrum (A minus B): above 1690 cm )1 it is dominated solely by the spectral response of the protein. Dithiothreitol can be used as an artificial reductant of cytochrome bo 3 . Figure 6 (top) shows a redox difference spectrum resulting from the reaction of dithiothreitol- reduced cytochrome bo 3 with caged oxygen. Absorbance patterns in the spectral region below 1670 cm )1 are variable to some extent, because of the high absorption in the amide regions due to variable protein concentration and to the residual amounts of water. The carbonyl region of the spectrum is mostly unaffected by the HPBC difference bands (Fig. 5B); the uncorrected data yielded the same difference band pattern if the spectral transition was recorded with either reductant. The use of dithiothreitol- reduced samples was found to be more practical, because it allowed the preparation of samples in presence of atmo- spheric oxygen. It is important to adjust for low water content for two reasons: (a) water is a strong absorber of IR radiation and (b) a considerable amount of solvent is mobilized by oxygen bubble formation upon photolysis of HPBC at higher concentrations, which critically affect the sample stability. FT-IR spectra of photoreduction and photooxidation In order to validate the redox difference FT-IR signals obtained from the oxidation reaction with caged dioxygen, spectra were compared with those obtained by the reverse reaction triggered by Ôcaged electronÕ FMN [16,17]. The difference FT-IR spectrum obtained in Fig. 6 (bottom) shows the typical redox pattern expected for the carbonyl region of the oxidase; it has the strong negative band at 1696 cm )1 and the prominent carbonyl feature at 1745/ 1735 cm )1 , which had been assigned to Glu286 in previous spectra [17]. The difference spectra in the carbonyl region, generated by either FMN (reduced ) oxidized) or HPBC (oxidized ) reduced) appear to be reciprocal, which dem- onstrates the equivalence of informational content from both experiments. Time resolution of the reaction of photochemically released oxygen with cytochrome bo 3 An efficient caged compound has to provide oxygen very quickly. Time-resolved measurements of cytochrome oxid- ase kinetics have been successfully carried out with the flow- flash method, by observation of heme absorbance [31–35]. HPBC-released dioxygen has been used to measure the kinetics of optical heme absorbance with reduced cyto- chrome c oxidase [19]. In preliminary experiments, the photoirradiation of the caged compound with a single laser flash led to formation of a stable absorbance line after 1 ls, which is the time resolution limit of the apparatus used. It may be assumed that oxygen is liberated in parallel to the absorbance change of the caged compound itself. The relevant lower time-limit could be estimated by reaction of photo-released oxygen with different heme proteins: if 5 l M duroquinol-reduced cytochrome bo 3 is exposed to low concentrations of caged oxygen, a flash-induced transient of absorbance decrease at 430 nm is observed, which is indicative of heme oxidation. The kinetic traces are complex and exhibit half-lives of about 1 ms, which was evaluated without application of spectral deconvolution analysis. Using the oxygenation of myoglo- bin as a different indicator, significant flash-induced absorbance changes were recordable even after only 100 ls. This demonstrates that the chemical formation of oxygen from its precursor was definitely not a rate-limiting step in the reaction of cytochrome bo 3 . DISCUSSION Absorbance changes in the IR region provide information about individual steps of the partial reactions of cytochrome oxidase (and presumably also of other oxygen-binding proteins) at the molecular level. The strong absorbance of the solvent water is a problem inherent to this spectroscopic technique, which has to be carried out with highly concentrated samples layered in very thin aqueous films. In the study of oxidases, this need is in conflict with the requirement of efficient mixing of reactants, such as provided by the stopped-flow type apparatus used in flow- flash experiments. The fundamental problem is overcome by delivery of molecular oxygen via photo-triggering of the organometallic oxygen precursor compound HPBC. High concentrations of oxygen could be obtained, because the use of the HPBC nitrate salt as obtained from our preparation does not have the severe solubility problem of the perchlorate derivative. In an extension of earlier experiments with the heme proteins hemoglobin [18] and with cytochrome c oxidase Fig. 6. Comparison of the redox difference FT-IR spectra of cyto- chrome bo 3 in the carbonyl region generated by FMN (bottom spectrum) and by caged dioxygen (top spectrum), using a sample containing 1 nmol dithiothreitol-treated cytochrome bo 3 and 10 nmol HPBC dissolved in glycerol. Ó FEBS 2002 Caged oxygen reaction with cytochrome oxidase (Eur. J. Biochem. 269) 2635 [19] by optical spectroscopy, it was demonstrated in this work that the dioxygen photoreleased by HPBC acts as electron acceptor of cytochrome bo 3 . Pre-reduction of the enzyme with thiol compounds was the most favorable sample preparation method. Electron transfer of the protein has been verified optically; molecular changes induced by the photoreaction were monitored by FT-IR spectroscopy. It was possible to recognize absorbance differences of carboxyl groups, one of which was assigned to the conformational change of the side chain of Glu286 from the catalytic subunit I of cytochrome bo 3 . Comparisons of the spectra obtained in this study with redox spectra measured with the caged electron donor FMN yielded absorbance patterns of inverted shapes. Oxidoreduction of cytochrome oxidase can thus be observed spectroscopically in the forward and backward reaction. Complementary information is gained by the reciprocal experiments of heme oxidation and reduction: The electron flow from heme centers to oxygen arises exactly as expected; it thus seems clear, that the dioxygen produced after uncaging behaves like the natural substrate. According to prelimirary experi- ments it is possible to follow time-resolved absorbance changes at optical wavelengths with cytochrome bo 3 .Inthe future, we will extend the caged oxygen approach to investigate the kinetics of oxidase reaction at the molecular level by IR spectroscopy. ACKNOWLEDGEMENTS This work was supported by the Deutsche Forschungsgemeinschaft grant SFB 394-C6 and the Volkswagen-Stiftung to M. L. We thank Prof Klaus Gerwert for his support in providing the FT-IR equipment. REFERENCES 1. Garcia-Horsman, J.A., Barquera, B., Rumbley, J., Ma, J. & Gennis, R.B. (1994) The superfamily of heme-copper respiratory oxidases. J. Bacteriol. 176, 5587–5600. 2. Pereira, M.M., Santana, M. & Teixeira, M. (2001) A novel scenario for the evolution of haem-copper oxygen reductases. Biochim. Biophys. Acta 1505, 185–208. 3. Wikstro ¨ m, M., Krab, K. & Saraste, M. (1981) Cytochrome Oxidase: A Synthesis. Academic Press, New York. 4. Trumpower, B.L. & Gennis, R.B. (1994) Energy transduction by cytochrome complexes in mitochondrial and bacterial respiration: the enzymology of coupling electron transfer reactions to transmembrane proton translocation. Annu. Rev. Biochem. 63, 675–716. 5. Brzezinski, P. (2000) Proton-transfer reactions in bioenergetics – Introduction. Biochim. Biophys. Acta 1, 1–5. 6. Abramson, J., Riistama, S., Larsson, G., Jasaitis, A., Svensson- Ek, M., Laakkonen, L., Puustinen, A., Iwata, S. & Wikstro ¨ m, M. (2000) The structure of the ubiquinol oxidase from Escherichia coli and its ubiquinone binding site. Nat. Struct. Biol. 7, 910–917. 7. Iwata, S., Ostermeier, C., Ludwig, B. & Michel, H. (1995) Struc- ture at 2.8 A ˚ resolution of cytochrome c oxidase from Paracoccus denitrificans. Nature 376, 660–669. 8. Tsukihara, T., Aoyama, H., Yamashita, E., Tomizaki, T., Yamaguchi, H., Shinzawa-Itoh, K., Nakashima, R., Yaono, R. & Yoshikawa, S. (1995) Structures of metal sites of oxidized bovine heart cytochrome c oxidase at 2.8 A ˚ . Science 269, 1069–1074. 9. Tsukihara, T., Aoyama, H., Yamashita, E., Tomizaki, T., Yamaguchi, H., Shinzawa-Itoh, K., Nakashima, R., Yaono, R. & Yoshikawa, S. (1996) The whole structure of the 13-subunit oxi- dized cytochrome c oxidase at 2.8 A ˚ . Science 272, 1136–1144. 10. Ostermeier, C., Harrenga, A., Ermler, U. & Michel, H. (1997) Structure at 2.7 A ˚ resolution of the Paracoccus denitrificans two- subunit cytochrome c oxidase complexed with an antibody F v fragment. Proc. Natl Acad. Sci. USA 94, 10547–10553. 11. Puustinen, A., Bailey, J.A., Dyer, R.B., Mecklenburg, S.L., Wikstro ¨ m, M. & Woodruff, W.H. (1997) Fourier transform infrared evidence for connectivity between Cu B and glutamic acid 286 in cytochrome bo 3 from Escherichia coli. Biochemistry 36, 13195–13200. 12. Pomes, R., Hummer, G. & Wikstro ¨ m, M. (1998) Structure and dynamics of a proton shuttle in cytochrome c oxidase. Biochim. Biophys. Acta 1365, 255–260. 13. Hofacker, I. & Schulten, K. (1998) Oxygen and proton pathways in cytochrome c oxidase. Proteins 30, 100–107. 14. Moss, D., Nabedryk, E., Breton, J. & Ma ¨ ntele, W. (1990) Redox- linked conformational changes in proteins detected by a combi- nation of infrared spectroscopy and protein electrochemistry. Evaluation of the technique with cytochrome c. Eur. J. Biochem. 187, 565–572. 15. Hellwig, P., Behr, J., Ostermeier, C., Richter, O.M., Pfitzner, U., Odenwald,A.,Ludwig,B.,Michel,H.&Ma ¨ ntele, W. (1998) Involvement of glutamic acid 278 in the redox reaction of the cytochrome c oxidase from Paracoccus denitrificans investigated by FTIR spectroscopy. Biochemistry 37, 7390–7399. 16. Lu ¨ bben, M. & Gerwert, K. (1996) Redox FTIR difference spec- troscopy using caged electrons reveals contributions of carboxyl groups to the catalytic mechanism of haem-copper oxidases. FEBS Lett. 397, 303–307. 17. Lu ¨ bben, M., Prutsch, A., Mamat, B. & Gerwert, K. (1999) Elec- tron transfer induces side-chain conformational changes of glu- tamate-286 from cytochrome bo 3 . Biochemistry 38, 2048–2056. 18. MacArthur, R., Sucheta, A., Chong, F.F. & Einarsdottir, O ¨ . (1995) Photodissociation of a (l-peroxo) (l-hydroxo) bis[bis (bipyridyl)-cobalt(III)] complex: a tool to study fast biological reactions involving O 2 . Proc. Natl Acad. Sci. USA 92, 8105–8109. 19. Van Eps, N., Szundi, I. & Einarsdottir, O ¨ . (2000) A new approach for studying fast biological reactions involving dioxygen: the reaction of fully reduced cytochrome c oxidase with O 2 . Biochemistry 39, 14576–14582. 20. Winkler, L.W. (1888) Die Bestimmung des im Wasser gelo ¨ sten Sauerstoffes. Ber. Dtsch. Chem. Ges. 21, 2843–2855. 21. Otwinowski, Z. & Minor, W. (1997) Processing of X-ray-diffrac- tion data collected in oscillation mode. Methods Enzymol. 276, 307–326. 22. Blessing, R.H. (1995) An Empirical Correction For Absorption Anisotropy. Acta Crystallogr., Sect. A: Found. Crystallogr. 51, 33–38. 23. Blessing, R.H. (1997) Outlier treatment in data merging. J. Appl. Crystallogr. 30, 421–426. 24. Sheldrick, G.M. (1990) Phase annealing in SHELX-90: direct methods for larger structures. Acta Crystallogr. Sect. A: Found. Crystallogr. 46, 467–473. 25. Heller, H.G. & Langan, J.R. (1981) Photochromic heterocyclic fulgides. Part 3. The ese of (E)-a-(2,5-dimethyl-3-furylethylidene) (isopropylidene) succinic anhydride as a simple convenient che- mical actinometer. J.Chem.Soc.PerkinII, 341–343. 26. Prutsch, A., Lohaus, C., Green, B., Meyer, H.E. & Lu ¨ bben, M. (2000) Multiple posttranslational modifications at distinct sites contribute to heterogeneity of the lipoprotein cytochrome bo 3 . Biochemistry 39, 6554–6563. 27. Skurlatov, U.I. & Pourmal, A.P. (1971) Reversible and Irrever- sible Interaction of O 2 with Co 2+ -a,a¢-dipyridyl. Bibl. Haemat, No. 38, 827–830. 28. Fallab, S., Zehnder, M. & Thewalt, U. (1980) Reactions of Oxygenated cobalt-(II) complexes. XIII. Diastereomeric forms of l-peroxo-l-hydroxo-bis [bis(ethylendiamine)cobalt (III)]. 2636 C. Ludovici et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Preparation, X-ray structure determination and reactivity. Helv. Chem. Acta 63, 1491–1498. 29. Vaska, L. (1976) Dioxygen–metal complexes: toward a unified view. Acc. Chem. Res. 9, 175–183. 30. Bogucki, R.F., McLendon, G. & Martell, A.E. (1976) Oxygen complexation by cobaltous chelates of multidentate pyridyl-type ligands. Equilibria, reactions, and electron structure of the com- plexes. J. Am. Chem. Soc. 98, 3202–3205. 31. Babcock, G.T. & Varotsis, C. (1993) Discrete steps in dioxygen activation-the cytochrome oxidase/O 2 reaction. J. Bioenerg. Biomembr. 25, 71–80. 32. Morgan, J.E., Verkhovsky, M.I., Puustinen, A. & Wikstro ¨ m, M. (1993) Intramolecular electron transfer in cytochrome o of Escherichia coli: events following the photolysis of fully and partially reduced CO-bound forms of the bo 3 and oo 3 enzymes. Biochemistry 32, 11413–11418. 33. Sucheta, A., Georgiadis, K.E. & Einarsdottir, O ¨ . (1997) Mechanism of cytochrome c oxidase-catalyzed reduction of dioxygen to water: evidence for peroxy and ferryl intermediates at room temperature. Biochemistry 36, 554–565. 34. Paula, S., Sucheta, A., Szundi, I. & Einarsdottir, O ¨ . (1999) Proton and electron transfer during the reduction of molecular oxygen by fully reduced cytochrome c oxidase: a flow-flash investigation using optical multichannel detection. Biochemistry 38, 3025–3233. 35. A ¨ delroth, P., Karpefors, M., Gilderson, G., Tomson, F.L., Gennis, R.B. & Brzezinski, P. (2000) Proton transfer from glutamate 286 determines the transition rates between oxygen intermediates in cytochrome c oxidase. Biochim. Biophys. Acta 1459, 533–539. Ó FEBS 2002 Caged oxygen reaction with cytochrome oxidase (Eur. J. Biochem. 269) 2637 . Caged O 2 Reaction of cytochrome bo 3 oxidase with photochemically released dioxygen from a cobalt peroxo complex Claudia Ludovici, Roland Fro¨. 1069–1074. 9. Tsukihara, T., Aoyama, H., Yamashita, E., Tomizaki, T., Yamaguchi, H., Shinzawa-Itoh, K., Nakashima, R., Yaono, R. & Yoshikawa, S. (1996) The