Mechanism of Energy Transfer from Carotenoids to

Department of Chemistry, 215 Glenbrook Road, University of Connecticut, Storrs, Connecticut 06269-4060, Gorlaeus Laboratories, Leiden University, 2300...
1 downloads 0 Views 186KB Size
J. Phys. Chem. B 1998, 102, 8151-8162

8151

Mechanism of Energy Transfer from Carotenoids to Bacteriochlorophyll: Light-Harvesting by Carotenoids Having Different Extents of π-Electron Conjugation Incorporated into the B850 Antenna Complex from the Carotenoidless Bacterium Rhodobacter sphaeroides R-26.1 Ruel Z. B. Desamero,† Veeradej Chynwat,† Ineke van der Hoef,‡ Frans Jos Jansen,‡ Johan Lugtenburg,‡ David Gosztola,§ Michael R. Wasielewski,§,⊥ Agnes Cua,| David F. Bocian,| and Harry A. Frank*,† Department of Chemistry, 215 Glenbrook Road, UniVersity of Connecticut, Storrs, Connecticut 06269-4060, Gorlaeus Laboratories, Leiden UniVersity, 2300 RA Leiden, The Netherlands, Chemistry DiVision, Argonne National Laboratories, Argonne, Illinois 60439, Department of Chemistry, Northwestern UniVersity, EVanston, Illinois 60208, and Department of Chemistry, UniVersity of California, RiVerside, California 92521 ReceiVed: January 28, 1998; In Final Form: April 14, 1998

Spheroidene and a series of spheroidene analogues with extents of π-electron conjugation ranging from 7 to 13 carbon-carbon double bonds were incorporated into the B850 light-harvesting complex of Rhodobacter sphaeroides R-26.1. The structures and spectroscopic properties of the carotenoids and the dynamics of energy transfer from the carotenoid to bacteriochlorophyll (BChl) in the B850 complex were studied by using steady-state absorption, fluorescence, fluorescence excitation, resonance Raman, and time-resolved absorption spectroscopy. The spheroidene analogues used in this study were 5′,6′-dihydro-7′,8′-didehydrospheroidene, 7′,8′-didehydrospheroidene, and 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene. These data, taken together with results from 3,4,7,8-tetrahydrospheroidene, 3,4,5,6-tetrahydrospheroidene, 3,4-dihydrospheroidene, and spheroidene already published (Frank, H. A.; Farhoosh, R.; Aldema, M. L.; DeCoster, B.; Christensen, R. L.; Gebhard, R.; Lugtenburg, J. Photochem. Photobiol. 1993, 57, 49. Farhoosh, R.; Chynwat, V.; Gebhard, R.; Lugtenburg, J.; Frank, H. A. Photosynth. Res. 1994, 42, 157), provide a systematic series of molecules for understanding the molecular features that determine the mechanism of energy transfer from carotenoids to BChl in photosynthetic bacterial light-harvesting complexes. The data support the hypothesis that only carotenoids having 10 or less carbon-carbon double bonds transfer energy via their 21Ag (S1) states to BChl to any significant degree. Energy transfer via the 11Bu (S2) state of the carotenoid becomes more important than the S1 route as the number of conjugated carbon-carbon double bonds increases. The results also suggest that the S2 state associated with the Qx transition of the B850 BChl is the most likely acceptor state for energy transfer originating from both the 21Ag (S1) and 11Bu (S2) states of all carotenoids.

Introduction Carotenoids are well-known light-harvesting agents.1 They absorb light in the 400-500 nm region of the visible spectrum where chlorophylls are not very good absorbers and then transfer the energy to chlorophyll for use in driving the process of primary electron transfer in the reaction center. However, the mechanism of energy transfer from carotenoids to chlorophylls is not well understood.2 It is uncertain which excited electronic states of carotenoids participate in energy transfer, and the optimal structures, geometries, and dynamics of carotenoids that would lead to efficient energy transfer are unknown. In purple non-sulfur photosynthetic bacteria, discrete antenna complexes exist and consist of carotenoids and bacteriochlorophylls (BChls) with specific pigment stoichiometries.3,4 One light-harvesting complex, denoted B800-850 owing to its * To whom correspondence should be addressed: Harry A. Frank, Department of Chemistry, 215 Glenbrook Road, University of Connecticut, Storrs, CT 06269-4060. Fax: 860-486-3772. E-mail: frank@ uconnvm.uconn.edu. † University of Connecticut. ‡ Leiden University. § Argonne National Laboratories. ⊥ Northwestern University. | University of California, Riverside.

absorption maxima at 800 and 850 nm, has been purified in high yield from several strains of photosynthetic bacteria and used extensively in spectroscopic and structural studies to elucidate the controlling factors of energy transfer.5-8 The B800-850 complex, also referred to as LH2, has a minimal unit consisting of three BChl molecules and two carotenoids noncovalently bound to two low-molecular-weight hydrophobic apoproteins labeled R and β.9 Both of these apoproteins have been sequenced.9 Two of the BChls are bound in close proximity to each other in the protein and are responsible for the absorption at 850 nm. The remaining monomeric BChl is responsible for the absorption at 800 nm. Using X-ray crystallographic methods, two groups have solved the structures of LH2 complexes from two different bacteria, Rhodopseudomonas acidophila strain 1005010 and Rhodospirillum molischianum.11 The structures reveal that the active light-harvesting assembly consists of two concentric cylinders of helical proteins forming a ring-shaped complex from either eight or nine independent minimal units of R and β subunit pairs. In analyzing the effect of the carotenoids on the spectroscopic properties of the antenna complexes, it has proven useful to compare complexes isolated from carotenoid-containing bacteria with those obtained from carotenoidless mutants. The caro-

S1089-5647(98)00911-0 CCC: $15.00 © 1998 American Chemical Society Published on Web 06/05/1998

8152 J. Phys. Chem. B, Vol. 102, No. 42, 1998 tenoidless mutant, Rhodobacter (Rb.) sphaeroides R-26, is an example of a bacterium that has been used extensively in these types of studies.12 The light-harvesting assembly prepared from Rb. sphaeroides R-26 when it was first described in 196313 had a peak at 870 nm and was shown to lack the LH2 pigmentprotein complex. Over the years, however, it was noticed that the main absorption attributable to the antenna complex in Rb. sphaeroides R-26 had migrated between 5 and 10 nm to shorter wavelength. A detailed analysis14,15 revealed a partial revertant of the original R-26 strain had resulted in the bacterium regaining the LH2 protein. The revertant was denoted R-26.114 and the R and β subunits of the LH2 protein complex sequenced and compared with the standard form of the LH2 complex from Rb. sphaeroides wild-type strain 2.4.1.15 The result was a single replacement of a phenylalanine in the R26.1 LH2 complex for a valine at position 24 in the R-subunit of the wild-type complex.15 In addition, the N-terminal formyl extension of methionine, which is the putative ligand to the Mg of the 800 nm absorbing BChl,10 is absent in the R26.1 LH2 R- subunit. This is apparently sufficient to prevent the binding of the 800 nm BChl molecule to the LH2 complex of the R26.1 strain.15 Thus, the LH2 complex of Rb. sphaeroides R-26.1 is a B800850 type of protein, having high sequence homology with the LH2 complex from Rb. sphaeroides wild-type strain 2.4.1, but lacking the 800 nm absorbing BChl. For this reason, the LH2 complex from Rb. sphaeroides R-26.1 is sometimes referred to as the B850 complex. Further work on this complex demonstrated that various carotenoids could be incorporated into the protein and function in the light-harvesting and photoprotection roles.16-18 The incorporated carotenoids were shown to transfer singlet energy to BChl and to quench BChl triplet states.16,19,20 From extensive spectroscopic experimentation it is known that carotenoids possess two low-lying excited electronic states, denoted S1 and S2, that play important roles in energy transfer in photosynthesis.21-23 The ground state, S0, and the first excited, S1, of polyenes possess Ag symmetry in the idealized C2h point group. Electronic transitions between these states, i.e., S0 f S1 (11Ag f 21Ag) absorption or S1 f S0 (21Ag f 11Ag) fluorescence, are symmetry forbidden. In contrast, electronic transitions to and from S0 and the second excited state, S2, which has Bu symmetry are allowed; i.e., the transitions S0 f S2 (11Ag f 11Bu) absorption and S2 f S0 (11Bu f 11Ag) fluorescence are allowed. The S0 f S2 (11Ag f 11Bu) transition is responsible for the characteristic strong absorption associated with all polyenes in the visible region. Carotenoids do not adhere strictly to C2h symmetry but still possess many of the spectral characteristics of the parent polyenes from which they were derived. Hence, the symmetry designations used for polyene electronic states are also used when referring to carotenoids. Recent studies have shown that the S1 states of carotenoids are much lower than previously thought.24-26 Qualitative and quantitative extrapolations of the S1 energies of short carotenoids and polyenes determined from fluorescence studies suggest that the S1 energies of carotenoids having chain lengths greater than 10 conjugated double bonds may lie below the S1 state of BChl.27,28 The implication of this result is that if longer carotenoids are to transfer singlet state energy, the mechanism may involve the carotenoid S2 state. This was suggested from fast-transient experiments on the LH2 complex from Rb. sphaeroides wild-type strain 2.4.1.29-31 The present work makes use of the facility of binding of exogenous carotenoids to the B850 complex to study systematically the effect on energy transfer of changing the extent of π-electron conjugation of the carotenoid from 7 to 13 conjugated

Desamero et al.

Figure 1. Molecular structures of spheroidene and analogues.

carbon-carbon double bonds. The carotenoids shown in Figure 1 have been incorporated into the B850 complex of Rb. sphaeroides R-26.1 and studied by using steady-state absorption, fluorescence, fluorescence excitation, resonance Raman, and time-resolved absorption spectroscopy. The data taken together with results from 3,4,7,8-tetrahydrospheroidene, 3,4,5,6-tetrahydrospheroidene, 3,4-dihydrospheroidene, and spheroidene previously published19,20 show that the mechanism of energy transfer from carotenoids to BChl depends strongly on the extent of π-electron conjugation in the carotenoid. Materials and Methods Preparation and Purification of the Samples. The B850 light-harvesting complex was obtained from Rb. sphaeroides cells grown anaerobically in modified Hunter’s media.32 The cells were harvested by centrifugation, washed, and suspended with 15 mM Tris buffer, pH 8.0, and disrupted by passing through a French pressure cell at 20 000 psi. The broken cells were resuspended in 15 mM Tris buffer and then centrifuged at 27000g and 4-15 °C for 10 min by using a Sorvall SS34 rotor. To isolate the chromatophores, the pellet-free supernatant was ultracentrifuged in a 55.2 Ti rotor at 250000g and 4 °C for 90 min. The pellets containing the chromatophore were resuspended and incubated in 15 mM Tris buffer, pH 8.0, 150 mM NaCl, and 0.6% lauryldimethylamine-N-oxide (LDAO) for 30 min in the dark at room temperature. The B850 lightharvesting complex was obtained as a pellet after ultracentrifugation at 250000g and 4 °C for 90 min. The resuspended pellet containing the B850 complex was then applied to a discontinuous sucrose density gradient, consisting of 0.3, 0.6, and 1.2 M sucrose in 15 mM Tris buffer, pH 8.0, with 0.1% lithium dodecyl sulfate (LDS) and spun at 150000g and 4 °C for 18 h to further purify the B850 light-harvesting complex. Spheroidene was obtained from anaerobically grown Rb. sphaeroides wild type strain 2.4.1 cells by extraction with

Energy Transfer from Carotenoids to Bacteriochlorophyll acetone and partitioning with pentane. After evaporating the pentane and redissolving in acetone, the sample was loaded onto a DEAE sephacel column that was equilibrated with acetone. A mixture of carotenoids, primarily spheroidene, and spheroidenone was then eluted from the column using acetone. Spheroidene was separated from the other carotenoids by using an alumina column with 0.25, 0.5, and 1% ethyl acetate in petroleum ether solutions as stepwise eluants. The synthesis and purification of 3,4,7,8-tetrahydrospheroidene, 3,4,5,6-tetrahydrospheroidene, and 3,4-dihydrospheroidene have been previously described.33 The synthesis and purification of 5′,6′dihydro-7′,8′-didehydrospheroidene, 7′,8′-didehydrospheroidene, and 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene will be reported in detail elsewhere.34 The method for reconstituting the carotenoids into the lightharvesting complex was done by using the procedure outlined by Noguchi et al.18 with some modifications introduced by Frank et al.19 Before the carotenoids were incorporated into the B850 light-harvesting complex, the detergent in the solution was exchanged from 0.1% LDS to 2% deoxycholate (3R,12Rdihydroxy-5β-cholan-24-oic acid). This was done by dialyzing the purified B850 light-harvesting solution overnight against 15 mM Tris buffer, pH 8.0, containing 2% deoxycholate. The carotenoid dissolved in petroleum ether was layered on the surface of the B850 light-harvesting complex in a molar ratio of 15:1 carotenoid to BChl. A stream of nitrogen gas was then passed over the surface of the solution until most of the petroleum ether had evaporated and the carotenoid was deposited as a thin film on the side of the vial. The mixture was then sonicated 30-45 min at 4 °C in the dark. After which an additional 15-fold molar excess of carotenoid in petroleum ether was added. Again, the petroleum ether was evaporated by using the stream of nitrogen gas and the mixture was sonicated in the dark. Excess carotenoids were removed by application of the solution to a discontinuous sucrose density gradient, consisting of 0.75, 1.5, and 2 M sucrose solutions, and subsequent ultracentrifugation at 150000g and 4 °C for 18 h. The purified carotenoid-reconstituted B850 light-harvesting complex was then dialyzed overnight against 15 mM Tris buffer, pH 8.0, with 0.02% deoxycholate to remove the sucrose from the solution. Steady-State Spectroscopy. A Milton Roy SLM Aminco Spectronic 3000 single-beam diode array spectrophotometer was employed to obtain the absorption spectra of the pigmentprotein complexes. Fluorescence and fluorescence excitation spectra of the samples were taken by using an SLM Instruments, Inc., 8000C spectrofluorimeter. A 450 W ozone-free xenon arc lamp and a 1500 grooves per mm grating monochromator comprised the variable-wavelength excitation light system. The samples were held in 2 mm × 10 mm quartz cuvettes and excited along the short dimension. Fluorescence was detected 90° to the excitation beam. Sample emission passed through a 10-mm Glan-Thompson calcite prism type polarizer, through an appropriate glass cutoff filter, and into a detection monochromator. The excitation system intensity profile was monitored on a Hamamatsu model R-928 photomultiplier tube and the emitted light was detected by an RCA 7102 photomultiplier tube. Contributions to the emission spectra resulting from the Raman scattering bands of the solvent were corrected by subtracting a solvent blank taken under identical conditions. The fluorescence spectra were also corrected for the wavelength dependencies of the optical components by using a correction factor generated with a standard lamp. Fluorescence excitation spectra were obtained in ratio mode with Rhodamine 800 in ethylene glycol as a reference. The optical density (OD) of

J. Phys. Chem. B, Vol. 102, No. 42, 1998 8153 the samples for the fluorescence experiments was between 0.75 and 1.0 in a 1-cm path length measured at 850 nm. Circular dichroism (CD) spectra of the samples were taken with a Jasco-710 spectropolarimeter. The OD of the samples for the CD experiments was between 1.0 and 1.2 at 850 nm in 1-cm quartz cuvettes. All the above spectroscopic determinations were done at room temperature. Resonance Raman scattering was induced by the 514.5-nm line of a Coherent Innova 400-15UV Ar+ laser and collected by using a 55-mm f/1.4 Canon camera lens positioned 90° to the incident beam. The scattering passed through a Spex Industries polarization scrambler placed in front of the entrance slit of the spectrometer and was dispersed by a Spex Industries 1877 0.6 m triple spectrograph equipped with a 600/mm gratings blazed at 750 nm. The detection system consisted of a liquid nitrogen cooled 512 × 512 pixel back-illuminated Princeton Instruments LN/charge-coupled device (CCD) with a Tektronix chip. Laser power was at 10 mW and spectral resolution was 2 cm-1. The spectral data were calibrated by using the known frequency of fenchone. Resonance Raman measurements were obtained at 4 °C on samples placed in a 1-mm-i.d. capillary tubes. The OD of the samples for the resonance Raman measurements was approximately 1.0 in a 1-cm path length measured at 514 nm. Transient Absorption Measurements. Nanosecond Apparatus. The nanosecond optical spectrometer used measuring light from a 150-W Xe arc lamp focused through a 12.5-mm2 pinhole aperture onto a 2-mm path sample cuvette after passing through a 470-nm cutoff filter. Light transmitted through the sample was detected by an Instrument SA model LH20 1200 grooves per mm monochromator employing a Products for Research 4840 photomultiplier tube. The output was amplified by using an Evans Associates model 4163 amplifier and fed to a Tektronix oscilloscope model TDS 620A for signal averaging. Each transient profile was recorded as an average of 2000 scans. The power of the measuring beam was kept at ∼25 mW, which was below the saturation level for the transient signals. To prevent the photooxidation of the light-harvesting complex during measurement, the samples were rendered anaerobic by adding sodium dithionite to a final concentration of ∼3 mM. The flash-induced photochemistry was initiated by using a Quanta-Ray DCR-3/PDL-2 Nd:YAG-pumped dye laser having a pulse duration of 7 ns at 580 nm using Rhodamine 610 in spectroscopic-grade methanol. Pulse repetition rate was 20 Hz. The laser beam was focused through a lens and onto the sample at a right angle to the probe beam. Flash laser intensity was kept in the region where the amplitudes of the transients increased linearly with the flash intensity. Femtosecond Apparatus. The transient femtosecond absorption apparatus has been previously described.35,36 During the experiment the samples were placed in 2-mm path length cuvettes and stirred continuously. Variations in the transmission of the measuring probe light through the sample and changes in the reference probe beam were monitored by photodiodes, the output of which were integrated, digitized, and recorded by a personal computer. Kinetic parameters were obtained by iterative deconvolution using the Levenberg-Marquardt algorithm. The instrumental time response was 100 fs at 400 nm. Results Steady-State Spectra. Absorption and Fluorescence Spectra. The extents of incorporation of the spheroidene analogues into the B850 complex were obtained by comparing the steady-state absorption spectra of the complexes with carotenoid incorporated

8154 J. Phys. Chem. B, Vol. 102, No. 42, 1998

Desamero et al.

TABLE 1: Extents of Incorporation of Spheroidene Analogues and the Carotenoid to BChl Energy Transfer Efficiencies of the Rb. sphaeroides R-26.1 B850 Light-Harvesting Complexes and from Spheroidene in the B800-850 Complex from Rb. sphaeroides Wild Type 2.4.1a sample

analogue

spheroidene in the B800-850 complex from Rb. sphaeroides wild type 2.4.1 carotenoids in the B850 complex from Rb. sphaeroides R-26.1 3,4,7,8-tetrahydrospheroidene 3,4,5,6-tetrahydrospheroidene 3,4-dihydrospheroidene spheroidene 5′,6′-dihydro-7′,8′-didehydrospheroidene 7′,8′-didehydrospheroidene 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene

extent of incorporation/% 100

1 2 3 4 5 6 7

89 ( 5b 48 ( 5b 80 ( 3b 92 ( 5 49 ( 4 55 ( 5 52 ( 3

extent of incorporation/% 95 ( 4 54 ( 2b 66 ( 4b 71 ( 6b 52 ( 5 47 ( 8 26 ( 4 12 ( 5

a The uncertainties represent the standard deviation from the mean values of at least two trials, and in each trial, the efficiencies measured at the two red most peaks of the carotenoid spectra were averaged. The percent incorporation was measured relative to that observed from the wild-type B800-850 complex, which is assigned the value 100%. b From Farhoosh et al.20

to that of the B800-850 light-harvesting complex from Rb. sphaeroides 2.4.1 wild type. The B800-850 complex has a 2:3 carotenoid-to-BChl stoichiometry and a high sequence homology with the B850 complex.14,15,37 The spectra of the B850 complexes with different carotenoids incorporated were normalized to the 850-nm absorbance of BChl in the B800-850 complex. From the ratios of the intensities of the carotenoid peaks in the two different types of complexes, the percent incorporation relative to spheroidene in the B800-850 complex could be obtained. The percents of incorporation of the carotenoids into the B850 complex were determined to be 92 ( 5% for spheroidene (4), 49 ( 4% for 5′,6′-dihydro-7′,8′didehydrospheroidene (5), 55 ( 5% for 7′,8′-didehydrospheroidene (6), and 52 ( 3% for 1′,2′-dihydro-3′,4′,7′,8′tetradehydrospheroidene (7). These values are reported in Table 1. Despite the fact that the carotenoids were treated similarly during the incorporation procedure, the extents of incorporation were somewhat different for each of the carotenoids. This may be attributed to any of three factors: (1) the differences in solubilities the molecules have in the detergent solution, (2) their different propensities to form aggregates, and (3) their disparate chemical stabilities. Figure 2a shows the difference absorption spectra of the B850 light-harvesting complex with carotenoid incorporated minus that of the carotenoidless B850 complex. Each spectrum is proportionally red-shifted as the extent of the π-electron conjugation increases. Similar red-shifting with extent of π-electron chain length is exhibited by the purified carotenoids in n-hexane solution (Figure 2b). Also, each spectrum of a carotenoid in the protein is proportionally red-shifted by approximately 17 nm relative to when it is in the n-hexane solution. Finally, the BChl peak at 850 nm is observed to shift to longer wavelength by 1-2 nm upon incorporation of a carotenoid into the complex (data not shown). The (1 - T) transmittance and fluorescence excitation spectra of the B850 complex with and without the different carotenoids incorporated are shown in Figure 3. The fluorescence excitation spectra were measured by monitoring the BChl fluorescence at 877 nm and scanning the excitation through the carotenoid and BChl Qx absorption spectral regions. The calculation of the efficiency of energy transfer from the carotenoid to BChl was done as follows: First, the fluorescence excitation and (1 - T) transmittance spectra were normalized to the BChl Qx absorption peak of the B850 complex at 595 nm (Figure 3). Then, a difference fluorescence excitation spectrum of the carotenoidincorporated B850 light-harvesting complex minus that of the carotenoidless protein complex was computed (dashed line in Figure 4). A difference (1 - T) transmittance spectra was then

Figure 2. (a) Difference absorption spectra of the B850 light-harvesting complex with carotenoid incorporated minus that of the carotenoidless B850 complex. Each minor tick mark on the y-axis corresponds to an absorption of 0.05. The spectra are offset vertically for clarity. (b) The absorption spectra of spheroidene (4) and analogues 5, 6, and 7 in n-hexane solution. Each minor tick mark on the y-axis corresponds to an absorption of 0.25. The spectra are offset vertically for clarity.

obtained by subtracting the (1 - T) transmittance spectra of the carotenoidless protein complex from that of the B850 complex incorporated with a carotenoid (solid line in Figure 4). Finally, the efficiency of energy transfer was calculated as the ratio of the carotenoid intensity in the difference fluorescence excitation spectrum to that in the difference (1 - T) transmittance spectrum (Figure 4). The percent efficiencies of energy transfer from carotenoid to BChl were determined to be 52 ( 5%, 47 ( 8%, 26 ( 4%, and 12 ( 5% for the B850 complex incorporated with spheroidene (4), 5′,6′-dihydro-7′,8′-didehydrospheroidene (5), 7′,8′-didehydrospheroidene (6), and 1′,2′dihydro-3′,4′,7′,8′-tetradehydrospheroidene (7), respectively (Table 1). The uncertainties represent the standard deviation from the mean values of at least two trials, and in each trial, the efficiencies measured at the two red-most vibronic features of the carotenoid spectra were averaged. Frank et al.19 previously reported the values 54 ( 2% for 3,4,7,8-tetrahydrospheroidene (1), 66 ( 4% for 3,4,5,6-tetrahydrospheroidene (2), 71 ( 6% for 3,4-dihydrosheroidene (3), and 56 ( 3% for spheroidene (4) incorporated into the B850 complex. The last of these values

Energy Transfer from Carotenoids to Bacteriochlorophyll

Figure 3. Room temperature (1 - T) transmittance (solid line) and fluorescence excitation (dashed line) spectra of the B850 complex with spheroidene (a) 4 and analogues (b) 5, (c) 6, and (d) 7 incorporated and (e) without carotenoid. The fluorescence spectra were obtained by monitoring the BChl fluorescence at 877 nm and scanning the excitation through the carotenoid and BChl Qx absorption spectral region. The y-axes are in units of transmittance for the (1 - T) transmittance and fluorescence excitation spectra normalized to the BChl Qx absorption peak of the B850 complex at 595 nm.

was also measured here and is in excellent agreement with the previous published value. Circular Dichroism (CD) Spectra. CD spectra of the B850 complexes incorporated with carotenoids were taken and compared to those of native B850 pigment-protein complex without carotenoid and to the spheroidene-containing B800850 complex from Rb. sphaeroides 2.4.1 wild type (Figure 5). The CD spectra of the B850 complex with carotenoids incorporated showed strong positive peaks in the region 450580 nm where the carotenoids absorb. The CD spectral features of the carotenoids are highly reminiscent of the carotenoid bands in the B800-850 complex from Rb. sphaeroides 2.4.1 wild type (Figure 5f) and are proportionally red-shifted as the extent of conjugation increases (Figure 5a-d). The positive and negative peaks between 300 and 440 nm and between 590 and 800 nm in the CD spectra of the B850 complex with carotenoids incorporated belong to BChl and are reproduced in the CD spectrum of the carotenoidless B850 complex (Figure 5e). Resonance Raman Spectra. The resonance Raman spectrum of purified all-trans spheroidene (4) in methanol solution has three prominent peaks at 1003, 1157, and 1524 cm-1 (Figure 6a). These have been assigned to the methyl in-plane rocking mode, the C-C stretching/H rocking modes, and the CdC stretching mode, respectively.38-40 The small feature at 962 cm-1 has been attributed to the C-H out-of-plane wag coupled with the CdC torsion.38,39 The methyl in-plane rocking mode at 1003 cm-1 is the least sensitive to carotenoid structure38 and

J. Phys. Chem. B, Vol. 102, No. 42, 1998 8155

Figure 4. Difference fluorescence excitation spectra (dashed line) and a difference (1 - T) transmittance spectra (solid line) of the B850 complex with spheroidene (a) 4 and analogues (b) 5, (c) 6, and (d) 7 incorporated. The efficiencies of energy transfer are given in Table 1 and were calculated for each system as the ratio of the carotenoid intensity in the difference fluorescence excitation spectrum to that in the difference (1 - T) transmittance spectrum.

Figure 5. Circular dichroism (CD) spectra of the B850 complex without carotenoid (e) and with analogues (a) 4, (b) 5, (c) 6, and (d) 7 incorporated. The CD spectrum of the B800-850 complex from Rb. sphaeroides wild type 2.4.1 (f) is also shown for comparison.

the CdC stretching mode decreases with extent of π-electron conjugation.39 These observations are evident in the present

8156 J. Phys. Chem. B, Vol. 102, No. 42, 1998

Desamero et al.

Figure 6. Resonance Raman spectra (600-1900-cm-1 region) of purified all-trans-spheroidene (a) 4 and (b) analogue 7 in methanol. The excitation wavelength for both determinations was 514.5 nm.

Figure 8. Kinetic traces representing the decay of the T1 f Tn transient absorption of spheroidene (a) 4 and analogues (b) 5, (c) 6, and (d) 7 incorporated into the B850 complex and monitored at 525, 540, 565, and 576 nm, respectively. The photochemistry was induced by a laser pulse with a repetition rate of 20 Hz and a duration of 7 ns at the excitation wavelength 580 nm. The smooth curve represents the calculated values for the best fit to a single-exponential rate expression.

Figure 7. Resonance Raman spectra (600-1900-cm-1 region) of B850 complex with spheroidene (a) 4 and (b) analogue 7 incorporated. Also shown for comparison is the resonance Raman spectra (600-1900cm-1 region) of spheroidene in the B800-850 complex from Rb. sphaeroides wild type 2.4.1 (c). Excitation wavelength was 514.5 nm in all cases.

work. For example, the methyl in-plane rocking mode for purified 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene (7) dissolved in methanol which has 13 conjugated carbon-carbon double bonds was found to be 1003 cm-1. Also, the CdC stretching mode for this same analogue was found to be 1514 cm-1. This corresponds to a 10 cm-1 downshift versus spheroidene. Resonance Raman spectra of the B850 complex incorporated with spheroidene and with analogue 7 were taken and compared to that of the spheroidene-containing B800-850 complex from Rb. sphaeroides 2.4.1 wild type (Figure 7). Except for (2 cm-1 shifts in the frequencies of the major features at 1003, 1154, and 1518 cm-1 and slight differences in the intensities of minor peaks at 1193, 1284, and 1450 cm-1, the resonance Raman spectra for B850 complex with spheroidene incorporated and the B800-850 complex from Rb. sphaeroides 2.4.1 wild type

were virtually identical (Figure 7a,c). More pronounced differences were observed when the resonance Raman spectrum for B850 complex incorporated with analogue 7 (Figure 7b) was compared with that of the B800-850 complex. The resonance Raman line due to the CdC stretching mode at 1509 cm-1 for the B850 incorporated with analogue 7 was found to be 9 cm-1 lower than that for the B800-850 complex (Figure 7b,c) consistent with the trend described above for the long analogue 7 molecule in solution. Also, the resonance Raman line at 956 cm-1 in the B850 complex due to the C-H out-ofplane wag was found to be 12 cm-1 lower compared to spheroidene in the B800-850 complex. Time-Resolved Absorption Spectra. Nanosecond FlashInduced Transmittance Spectra. Figure 8 shows the kinetic traces of the nanosecond time-resolved flash-induced transmittance changes for the B850 complex incorporated with spheroidene (4) and analogues 5, 6, and 7. Plots of the transient amplitudes versus wavelength are given in Figure 9. These reveal that the peaks of the transient spectra proportionally redshift as the extent of π-electron conjugation of the carotenoid increases (Table 2). The kinetic traces given in Figure 8 represent the decay of the T1 f Tn transient absorption spectra (Figure 9) of spheroidene (4) and analogues 5, 6, and 7 incorporated into the B850 complex and monitored at 525, 540, 565, and 576 nm, respectively. The triplet state lifetimes (τT) obtained from the traces are 5.6 ( 0.7 µs for the B850 complex incorporated with spheroidene (4), 4.1 ( 0.5 µs for the complex containing 5′,6′-dihydro-7′,8′-didehydrospheroidene (5), 3.5 ( 0.3 µs for the complex containing 7′,8′-didehydrospheroidene (6), and 3.1 ( 0.5 µs for the complex containing 1′,2′-dihydro-

Energy Transfer from Carotenoids to Bacteriochlorophyll

J. Phys. Chem. B, Vol. 102, No. 42, 1998 8157 or both. Populations from the S1 and S2 states may also decay radiatively. However, because the quantum yields of fluorescence1,36,41 from both S1 and S2 are low (∼ 10-4 to 10-5), their contribution to the present model can be neglected. The relevant equations are

Figure 9. Triplet spectra of spheroidene 4 (b) and analogues 5 (4), 6 ([), and 7 (0) incorporated into the B850 complex. The plots show the changes in the initial amplitude of the kinetic traces shown in Figure 8 at the probe wavelengths in the region 480-600 nm. The corresponding molecule number is indicated directly above the wavelength of maximum T1 f Tn transient signal, λTmax, for each of the molecules.

3′,4′,7′,8′-tetradehydrospheroidene (7). The transient profiles were fit best by single-exponential rate expressions. These data are summarized in Table 2. Femtosecond Flash-Transient Absorption Spectra. The decay of the S1 f Sn transient absorption bands of the carotenoids in the B850 complex were monitored at a single wavelength between 500 and 620 nm and plotted in Figure 10. Plots of the transient amplitudes versus wavelength show that the spectra shift to the red proportionally as the number of carbon-carbon double bonds of the carotenoid increases (Figure 11 and Table 3). Transient decay profiles of the B850 complex incorporated with 3,4-dihydrospheroidene (3), spheroidene (4), 5′,6′-dihydro7′,8′-didehydrospheroidene (5), 7′,8′-didehydrospheroidene (6), and 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene (7) were best fit by single-exponential rate expressions. The decay times, denoted, τS1, of the transients were 8.7 ( 0.3, 7.9 ( 0.3, 4.6 ( 0.3, 2.5 ( 0.3, and 1.6 ( 0.3 ps, respectively. The B850 complex incorporated with 3,4,5,6-tetrahydrospheroidene (2) exhibited multiphasic decay kinetics with primary components of 7.7 ( 1.0 ps (40%) and 46 ( 10 ps (52%). The spectrum of the transient amplitude for analogue 2 reveals significant intensity in the 525-575-nm region not evident in the traces from the other molecules (Figure 11), and the fit of the transient decay profile of the B850 complex without carotenoid yielded multiphasic dynamics with the primary component being 7.7 ( 0.4 ps (75%). This indicates that the 7.7-ps component is associated with a BChl decay process independent of the carotenoid. The decays of the S1 f Sn transient absorption bands for the B850 complex incorporated with spheroidene analogues 2 and 3 were faster in the protein compared to these same molecules in n-hexane solution. The lifetimes for analogues 2 and 3 in the protein (Figure 12) were found to be 46 ( 10 and 8.7 ( 0.3 ps, compared to 85 ( 5 and 25.4 ( 0.9 ps in solution, respectively.36 For analogues 4-7, the transients showed no significant difference for the molecules being in the protein or in n-hexane solution (Figure 12). Kinetic Model for the Mechanism of Carotenoid-toBacteriochlorophyll Energy Transfer. Figure 13 shows the routes by which carotenoid excitation may be transferred to BChl. Upon absorption of light, the carotenoid is promoted to its S2 state. The S2 population, [S2], may then relax back to the ground state, S0, via internal conversion with rate constant kic3 and/or transfer energy to BChl with rate constant kET2. The S2 population may also decay to the S1 excited state with rate constant kic2. If the S1 energy level is populated, it could either relax to the ground state, S0, via internal conversion with rate constant kic1 or transfer energy to BChl with rate constant kET1

d[S2]/dt ) -(kET2 + kic2 + kic3)[S2]

(1)

d[S1]/dt ) kic2[S2] - (kET1 + kic1)[S1]

(2)

d[S0]/dt ) (kET1 + kic1)[S1] + (kic3 + kET2)[S2]

(3)

From eq 1, [S2] can be determined to be

[S2] ) [S2]0exp(-k2t)

(4)

where k2 ) kET2 + kic2 + kic3. Substituting (4) into (2) and rearranging

d[S1]/dt + k1[S1] ) kic2[S2]0exp(-k2t) and exp(-k2t) - exp(-k1t) (5) [S1] ) [S1]0 + kic2[S2]0 (k1 - k2)

[

]

where k1 ) kET1 + kic1. Substituting (4) and (5) into (3) the ground-state population, [S0], is

[S0] ) [S0]0 + kic1[S1]0t + [S2]0 + kic2 (k2 - k1 - kic2) exp(-k2t) + exp(-k1t) [S2]0 (6) (k1 - k2) (k1 - k2)

[

]

Because the S1 energy state is unpopulated at t ) 0, [S1]0 ) 0,

[S0] ) [S0]0 + (k2 - k1 - kic2) kic2 1+ exp(-k1t) + exp(-k1t) [S2]0 (k1 - k2) (k1 - k2) (7)

[

]

This model predicts that changes in the [So] (eq 7) and [S1] (eq 5) populations are described by double exponential expressions. It should be noted that similar expressions may be obtained for the carotenoids in solution, with the substitutions k2 ) kic2 + kic3 and k1 ) kic1. From this kinetic model, an expression for the efficiency of singlet energy transfer from the carotenoid to BChl may be derived. Assuming kic3 , kic1, kic2, which is justified by the energy gap law,

)

[

][

]

kic2 kET1 kET2 + kET2 + kic2 kET2 + kic2 kET1 + kic1

(8)

Because the population of the S1 state originates from the S2 state, the contribution to the overall efficiency, , of energy transfer from the S1 state is a fraction, [kic2/(kET2 + kic2)], of that obtained from the S2 state. Discussion Structural Integrity of the B850 Complex Incorporated with Carotenoids. For these experiments it is important to demonstrate that the B850 protein is unperturbed by the incorporation of carotenoids and that the series of spheroidene analogues studied here adopt similar configurations in the protein and are anchored in a manner that would lead to singlet energy transfer if it were possible. Several pieces of data make it evident that the carotenoids adopt uniform structures and that the complexes have a high degree of structural integrity:

8158 J. Phys. Chem. B, Vol. 102, No. 42, 1998

Desamero et al.

TABLE 2: Triplet State Lifetimes (τT) and Triplet Absorption Maxima (λTmax) of Spheroidene and Analogue Compounds Incorporated into the B850 Light-Harvesting Complexes from Rb. sphaeroides R-26.1a carotenoid

analogue

τT/10-6 s

λTmax/nm

spheroidene 5′,6′-dihydro-7′,8′-didehydrospheroidene 7′,8′-didehydrospheroidene 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene

4 5 6 7

5.6 ( 0.7 4.1 ( 0.5 3.5 ( 0.3 3.1 ( 0.5

525 ( 2 540 ( 2 565 ( 2 576 ( 4

a Uncertainties in τT represent the standard deviation from the mean of the fit of the kinetic traces to several trials. The uncertainties in λTmax are derived from the reproducibility of the ∆A spectra given in Figure 9.

Figure 10. Decay of the S1 f Sn transient absorbance of molecules (a) 2, (b) 3, (c) 4, (d) 5, (e) 6, and (f) 7 incorporated into the B850 complex. The B850 complex with analogues 2 and 3 incorporated were excited at 482 nm and probed at 508 and 531 nm, respectively. The B850 complex with spheroidene 4 and analogues 5, 6, and 7 incorporated were excited at 506 nm and probed at 548, 569, 602, and 623 nm, respectively. The probe wavelengths correspond to the maxima of the S1 f Sn absorption. The smooth curve represents the theoretical curve for the best fit to a multiphasic decay kinetics expression for trace (a) and single-exponential rate expressions for the kinetic traces (b)-(f). The y-axis denotes the change in absorption, ∆A.

Figure 11. Flash-transient absorption spectra of the excited singlet states of molecules 2-7 incorporated into the B850 complex. The plots represent the changes in the initial amplitudes of the kinetic traces shown in Figure 10 over the probe wavelength range 500-650 nm. The corresponding analogue number is indicated directly above the wavelength of maximum S1 f Sn transient absorption, λSmax, for each of the molecules.

(1) Absorption Spectral Shifts. The carotenoid S0 f S2 transition is highly dependent on solvent polarizability21,42 so that a change in the environment of the carotenoid would be

manifested as a shift in its absorption spectra. The red shift of the carotenoid spectra upon incorporation of the molecules into the B850 complex indicates that the carotenoids are bound in the protein. Furthermore, the 850-nm BChl band shifts by 1-2 nm upon incorporation of the carotenoids, suggesting that the carotenoids are bound in close proximity to the BChls but are not significantly perturbing them in a manner that would diminish their ability to perform light-harvesting. (2) Circular Dichroism. The CD spectra of the carotenoids incorporated into the B850 complex are very similar to each other and to that of spheroidene in the wild-type B800-850 complex (Figure 5), indicating that the carotenoids are bound in the B850 complex in a manner similar to that of the native B800-850 complex. Because no CD spectra are observed for carotenoids in organic solvents or in detergent solution,43-46 the induced optical activity of the series of carotenoids studied here can only be attributed to their interactions with the amino acid residues of the protein. (3) Resonance Raman Spectroscopy. The resonance Raman spectra of the B850 complex incorporated with spheroidene and with analogue 7 show peaks associated with the methyl in-plane rocking mode, the C-C stretching/H rocking modes, and the CdC stretching mode that are consistent with that of the spheroidene-containing B800-850 complex from Rb. sphaeroides 2.4.1 wild type. This indicates that the B850 complex with carotenoid incorporated is adopting a structure similar to that of the wild-type B800-850 complex. The discernible differences in band position and intensities between the spectra of the B850 complex with analogue 7 incorporated and the B800-850 complex are not attributable to changes in the nature of the interaction of the carotenoid with its protein environment. Rather, these differences are due to the different extents of π-electron conjugation of the carotenoids as is apparent from similar variations which are observed in the resonance Raman spectra of spheroidene and analogue 7 in methanol solution (Figure 6). Previous resonance Raman studies by Hayashi et al.17 and Noguchi et al.18 on the light-harvesting complexes from Ch. Vinosum, Rs. rubrum, and Rp. palustris have suggested that some of the bound carotenoids were configurationally twisted. This was based on values of ∼0.5 for the ratios of the intensity of the out-of-plane C-H wagging mode of the polyene chain at 965 cm-1 to the 1003 cm-1 methyl rocking mode band. These authors suggested that a high ratio correlates with a low energy transfer efficiency from the carotenoid to BChl and is due to a distortion of the carotenoid away from planarity. The resonance Raman data presented here do not show a high ratio of the 965 to 1003 cm-1 bands suggestive of configuration distortion, yet the energy transfer efficiencies for B850 complex incorporated with analogue 7 which is 12 ( 5% and with spheroidene which is 52 ( 5% are significantly lower than the ∼ 95% value observed for the spheroidene in the wild-type B800-850 complex.17,18,47,48 The ratios of the 965-cm-1 band to the 1003cm-1 bands are ∼ 0.11 and ∼ 0.25 for the B850 complex

Energy Transfer from Carotenoids to Bacteriochlorophyll

J. Phys. Chem. B, Vol. 102, No. 42, 1998 8159

TABLE 3: Transient Decay Lifetimes (τS1), Rate Constants (kS1), and the S1 f Sn Transient Absorption Maxima (λSmax) of Spheroidene and Analogues Incorporated into the B850 Complex from Rb. sphaeroides R-26.1a analogue

kS1/1011 s-1

τS1/10-12 s

λSmax/nm

3,4,5,6-tetrahydrospheroidene

2 3 4 5 6 7

7.7 ( 1.0 46 ( 10 8.7 ( 0.3 7.9 ( 0.3 4.6 ( 0.3 2.5 ( 0.3 1.6 ( 0.3

508 ( 2

3,4-dihydrospheroidene spheroidene 5′,6′-dihydro-7′,8′-didehydrospheroidene 7′,8′-didehydrospheroidene 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene

1.30 ( 0.17 (40%) 0.22 ( 0.04 (52%) 1.14 ( 0.04 1.26 ( 0.05 2.17 ( 0.14 4.0 ( 0.5 6.2 ( 1.1

carotenoid

530 ( 2 548 ( 2 569 ( 2 602 ( 2 624 ( 2

a Decay profile for the B850 complex containing 3,4,5,6-tetrahydrospheroidene was multiphasic. The uncertainties in k S1 were propogated from the uncertainties of the experimental values of τS1.

Figure 12. Plot of the S1 transient decay lifetimes of molecules 2-7 incorporated into the B850 complex (τS1, s) and in n-hexane solution (τic1, l). The error bars represent the uncertainties in the τic1 and τS1 values determined from the standard deviations from the mean value of several measurements.

Figure 13. Kinetic diagram showing the routes by which carotenoid excitation may be transferred to BChl (solid lines) or deactivate via internal conversion (broken lines).

incorporated with spheroidene and with analogue 7, respectively, and ∼ 0.11 for the wild-type B800-850 complex. These data suggest that factors other than configurational distortion, which is evident in the preparations studied by Hayashi et al.17 and Noguchi et al.,18 are responsible for modulating the energy transfer efficiency (see below). (4) Triplet State Quenching. Triplet states of carotenoids are not populated to any significant degree by direct excitation. This is because the excited singlet states internally convert by radiationless processes to the ground state so efficiently that the probability of intersystem crossing is essentially negligible.49 The primary mechanism for carotenoid triplet state formation is via triplet energy transfer from BChl. For triplet transfer to be efficient, very close proximity, essentially van der Waals contact, must be achieved between the energy donor and acceptor molecules.50 If spheroidene and its analogues are to

quench the BChl triplet states they must assume positions in close proximity to BChl in the B850 complex. The B850 complexes with carotenoids having nine or more carbon-carbon double bonds exhibit significant triplet energy transfer as indicated by the signals presented in Figures 6 and 7. The redshifting of the transient spectra and the shortening of the triplet lifetimes with increasing π-electron conjugation are consistent with the assignment of the signals to carotenoid triplet absorption as previously reported.20,49,51-53 Clearly, these carotenoids are bound in close proximity to BChl. The carotenoids having less than nine carbon-carbon double bonds do not show triplet absorption because the triplet energy of the carotenoid is too high to be populated from the B850 triplet.20 Despite this, the absorption spectral shifts19 and the CD spectra20 support the notion that analogues 1, 2, and 3 are also bound in the protein in the same manner as spheroidene (4) and analogues 5, 6, and 7. Mechanism of Energy TransfersReconciling the Kinetic Model with the Experimental Dynamics Data. The kinetic model presented above predicts that the depopulation of the S1 state will follow biexponential decay dynamics (eq 5). The data, however, were satisfactorily fit to single-exponential expressions. The first component k2 in eq 5 represents the sum kET2 + kic2 + kic3. This component is probably not resolved from the transient profiles shown in Figure 8 owing to the 100-fs response limitation of the transient spectrometer. k1, which is the sum kET1 + kic1, in eq 5 was obtained as a fit to a single-exponential decay expression. Using kic1 values from a previous fit of the energy gap law to the dynamics for the decay of the S1 states of the carotenoid in solution,36 kET1 values were obtained. These values of kET1, which are given in Table 4, decrease as the extent of the π-electron conjugation of the carotenoids increases. The data indicate that only carotenoids having 10 or less carboncarbon double bonds transfer energy via their S1 states to BChl to any significant degree. For the longer carotenoids, the kET1 values were too small relative to the kic1 values to be determined reliably in the analysis. The values of kET2 were calculated based on eq 8, which is an expression that relates the efficiency of energy transfer, , to the internal conversion rate constants, kic1 and kic2, and the energy transfer rate constants, kET1 and kET2. From the values of  and kET1 determined in this work (Tables 1 and 4) and the values of kic1 and kic2 previously obtained (Table 4),31,36 values for kET2 were calculated. These values of kET2, which are given in Table 4, remain relatively constant for the analogues having 8-11 carbon-carbon double bonds and then decrease slightly as the conjugated chain length is increased further. These data suggest that as the number of conjugated carbon-carbon double bonds increases, the utilization of the S2 state as the route of energy transfer becomes proportionately more important than the S1 route. Ultimately, however, both routes are diminished with increasing conjugated chain length, and the efficiency of

8160 J. Phys. Chem. B, Vol. 102, No. 42, 1998

Desamero et al.

TABLE 4: Internal Conversion Rate Constants (kic1 and kic2), Energy Transfer Times (τET1 and τET2), and Rate Constants (kET1 and kET2) for the Carotenoids Incorporated into the B850 Complex from Rb. sphaeroides R-26.1a carotenoid

analogue

3,4,5,6-tetrahydrospheroidene 3,4-dihydrospheroidene spheroidene 5′,6′-dihydro-7′,8′-didehydrospheroidene 7′,8′-didehydrospheroidene 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene

2 3 4 5 6 7

kic1/1011 s-1

kic2/1012 s-1

0.118 ( 0.007 11 ( 6 0.394 ( 0.023 8 ( 3 1.15 ( 0.07 (6.7 ( 1.2)b 2.56 ( 0.16 6.1 ( 2.6 3.76 ( 0.28 6.3 ( 2.7 9.1 ( 1.6 5.2 ( 2.3

kET1/1011 s-1 τET1/10-12 s kET2/1012 s-1 τET2/10-12 s 1.18 ( 0.17 0.75 ( 0.05 0.11 ( 0.09 nd nd nd

8.5 ( 1.2 13.3 ( 0.7 90 ( 70 nd nd nd

6(4 1.5 ( 0.8 6.0 ( 2.8 5.4 ( 2.4 2.2 ( 0.9 0.7 ( 0.3

0.16 ( 0.11 0.6 ( 0.3 0.16 ( 0.08 0.18 ( 0.08 0.45 ( 0.20 1.4 ( 0.6

a Notation refers to the scheme presented in Figure 11. The values were determined from the model presented in the text in conjunction with the internal conversion rate constants (kic1 and kic2) previously determined,36 the efficiencies of carotenoid to BChl energy transfer (Table 1), and kS1 (Table 3). Because kS1 is equal to k1 in eq 5, the kET1 values were calculated as the differences between kS1 and kic1. The kET2 values were obtained from eq 8. τET1 and τET2 are the reciprocals of kET1 and kET2, respectively. The uncertainties in kET1, kET2, τET1, and τET2 were propagated from the uncertainties in the experimental values from which they were obtained. nd means too small relative to kic1 to be determined. b Experimentally determined value from Ricci et al.31

energy transfer drops to a very low level; viz., 12 ( 5%, for the longest carotenoid studied here, analogue 7. An interesting observation in Table 4 is that the kET1 values fall off much more rapidly with increasing extent of π-electron conjugation than do the kET2 values. The difference may lie in the distinction between the nature of the electronic distributions in the S1 and S2 states and their relative contributions to the mechanism of energy transfer. In general, energy transfer may proceed via (i) the Fo¨rster induced dipole-dipole resonance transfer mechanism,54 (ii) the Coulomb mechanism with strong coupling between closely associated transition dipoles,55 (iii) the Coulomb mechanism involving multipolar interactions,56 or (iv) the Dexter or electron exchange mechanism.57 The low quantum yields of carotenoid fluorescence argue against the Fo¨rster resonance energy mechanism being operative in carotenoid-to-chlorophyll energy transfer. However, the Coulomb mechanism involving either dipolar or higher order multipolar coupling or electron exchange may be important.56 The fundamental equation describing the rate of energy transfer between weakly coupled donor-acceptor pairs is58

kET )

1 2 |T| J cp

(9)

where T is the electronic coupling term and J is the spectral overlap given by

∫0∞Fd(ν)a(ν) dν J) ∞ ∫0 Fd(ν) dν∫0∞a(ν) dν

(10)

Fd(ν) represents the fluorescence spectrum of the donor, and a corresponds to the absorption of the acceptor on a frequency (ν) scale. With the appropriate electronic coupling term, T, eq 9 is applicable to any type of energy transfer between weakly coupled chromophores, Fo¨rster, Dexter, or otherwise. The dependence of the rate of energy transfer on this interaction energy will be affected by the strength of the molecular coupling compared to the bandwidths of the molecular electronic transitions.59 For transfer from the 21Ag (S1) state of the carotenoid, either the Dexter mechanism or the Coulomb mechanism involving multipolar interactions is appropriate.56 This is because the S0 f S1 (11Ag f 21Ag) transition is symmetry forbidden and therefore has a vanishingly small transition dipole moment. Consequently, energy transfer from the S1 state of the carotenoid via a dipole mechanism is very unlikely. For energy transfer originating from the 11Bu (S2) state of the carotenoid, the Coulomb mechanism is likely to dominate because the strongly allowed S0 f S2 (11Ag f 11Bu) transition is associated with a

Figure 14. (a) Spectral overlap between the absorption of the B850 light-harvesting complex and the hypothetical S1 fluorescence traces from molecules 1-7. The hypothetical traces were derived from the actual fluorescence of analogue 1, shifted to correspond to the spectral origins of the S1 f S0 transition of the other analogues. (b) Spectral overlap between the absorption of the B850 light-harvesting complex and the hypothetical S2 fluorescence traces from molecules 1-7. The hypothetical traces were derived from the actual fluorescence of spheroidene (4), shifted to correspond to the spectral origins of the S2 f S0 transition of the other analogues.

large transition dipole.30 As energy acceptors, both the S1 and S2 states of BChl have been implicated. For example, Scholes et al.58 suggested that energy transfer between the 21Ag (S1) state of the carotenoid and the S1 (Qy) state of the BChl is described by terms proportional to the product of dipole-dipole and polarization interactions. Andersson et al.60 suggested that the major route of energy transfer occurs from the 11Bu (S2) state of the carotenoid to the S2 (Qx) state of the BChl. In an attempt to distinguish between these possibilities for the series of spheroidene analogues studied here, it is interesting to ask whether spectral overlap can account for the trends in the rate constants kET1 and kET2 observed in Table 4. Figure 14a shows the spectral overlap between the B850 Qx and Qy absorption and the hypothetical S1 fluorescence traces from spheroidene analogues 1-7. The hypothetical traces were derived from the actual fluorescence of analogue 1 taken in n-hexane solution, shifted to correspond to the spectral origins of the S1 f S0 transitions of the other analogues. Owing to the small transition dipole of the S0 f S1 absorption, the spectral origins of the transition measured in solution are likely to be

Energy Transfer from Carotenoids to Bacteriochlorophyll

J. Phys. Chem. B, Vol. 102, No. 42, 1998 8161

TABLE 5: Spectral Overlaps of the B850 Soret, Qx and Qy Absorption Bands with the S1 f S0 (21Ag f 11Ag) or S2 f S0 (11Bu f 11Ag) Fluorescence of the Spheroidene Analoguesa spectral overlap carotenoid

analogue

[Car (S1 f S0) and B850 Qy]/10-5

3,4,7,8-tetrahydrospheroidene 3,4,5,6-tetrahydrospheroidene 3,4-dihydrospheroidene spheroidene 5′,6′-dihydro-7′,8′-didehydrospheroidene 7′,8′-didehydrospheroidene 1′,2′-dihydro-3′,4′,7′,8′-tetradehydrospheroidene

1 2 3 4 5 6 7

2.19 12.5 19.9 19.5 14.7 11.3 5.82

a

[Car (S1 f S0) and B850 Qx]/10-5

[Car (S2 f S0) and B850 Qx]/10-5

[Car (S2 f S0) and B850 Soret]/10-5

16.3 5.90 1.80 0.17 0.17 0 0

1.87 3.71 8.53 14.1 18.9 25.7 28.6

11.4 6.14 1.75 0.43 0.11 0.004 0

The spectral overlap was calculated according to eq 10 given in the text.

very similar to those of the carotenoids incorporated into the protein. Figure 14b shows the overlap between the B850 Soret and Qx absorption and the hypothetical S2 fluorescence traces from spheroidene analogues 1-7. The hypothetical traces were derived from the actual fluorescence of analogue 4 in n-hexane solution, shifted to correspond to the spectral origins of the S2 f S0 transitions of the other analogues. This analysis is justified by the fact that the band shapes of the transitions, determined in large part by the vibronic features, are barely affected by changes in the extent of π-electron conjugation.36 The values of spectral overlap between the absorption of the BChls and the fluorescence of the analogues were determined according to eq 10 for both the carotenoid S1 and S2 state cases (see Table 5). The spectral overlap between the hypothetical S2 fluorescence traces from the spheroidene analogues and the B850 Soret (column 4, Table 5) decreases with increasing π-electron conjugation, while the spectral overlap between the hypothetical S2 fluorescence traces from the spheroidene analogues and the B850 Qx transition (column 3, Table 5) increases. These opposing effects offset one another and are consistent with the modest changes in the values of kET2 given in Table 4. In contrast, there is a significant decrease in the values of kET1 with extent of π-electron conjugation increasing from 8 to 10 carbon-carbon double bonds, i.e., for analogues 2-4 (see Table 4). For the analogues having more than 10 carbon-carbon double bonds, the values of kET1 were too small relative to kic1 to be determined. This trend in the data can be explained if one assumes that the primary route of energy transfer for these molecules is between the 21Ag (S1) state of the carotenoid and the S2 (Qx) state of the B850 BChl. The spectral overlap between the 21Ag (S1) f 11Ag (S0) emission profiles of these carotenoids and the Qy absorption of the B850 BChl is reasonably invariant for molecules 2-4 (column 1, Table 5), whereas the overlap between the 21Ag (S1) f 11Ag (S0) emission and the B850 BChl Qx absorption (column 2, Table 5) decreases by more than an order of magnitude over this range of short analogue molecules. This could account for the precipitous decrease in kET1 values evident in the present analysis. Interestingly, the structure of the LH2 complex from Rps. acidophila 10050 shows that the Qx transitions of the 850 nm absorbing BChls are parallel to the direction of the conjugated chains of the carotenoids.37 This would optimize Coulombic interactions between the BChl and carotenoid and lead to efficient light harvesting if the spectral overlap is satisfactory.55 The Qy transition for the B850 BChl is perpendicular to the chain of π-electron conjugation and thus would not be optimized for energy transfer via the Coulomb mechanism. Conclusions Spheroidene and a series of spheroidene analogues with extents of π-electron conjugation ranging from 7 to 13 carbon-

carbon double bonds were incorporated into the B850 lightharvesting complex of Rb. sphaeroides R-26.1 and studied by using steady-state absorption, fluorescence, fluorescence excitation, resonance Raman, and time-resolved absorption spectroscopy. The data indicate that the carotenoids are uniformly bound in the protein in a manner that would lead to efficient singlet state energy transfer, if possible. The data support the hypothesis that only carotenoids having 10 or less carboncarbon double bonds transfer energy via their 21Ag (S1) states to BChl to any significant degree. As the number of conjugated carbon-carbon double bonds increases, energy transfer via the 11Bu (S2) state of the carotenoid becomes more important than the S1 route. The data also suggest that the S2 state associated with the Qx transition of the B850 BChl is the most likely acceptor state for energy transfer originating from both the 21Ag (S1) and 11Bu (S2) states of all carotenoids. Acknowledgment. This work was supported by grants to H.A.F. from the National Institutes of Health (GM-30353), the Human Frontier Science Program, the United States Department of Agriculture (92-37306-7690), and the University of Connecticut Research Foundation and to J.L. from the Human Frontier Science Program and The Netherlands Foundation of Chemical Research (SON), which is financed by The Netherlands Organization for the Advancement of Pure Research (NWO). The work at Argonne National Laboratory was supported by the Office of Basic Energy Sciences, Division of Chemical Sciences, U.S. Department of Energy, under Contract W-31-109-Eng-38. The work at the University of California, Riverside, was supported by grants to D.F.B. from the National Institute of General Medical Sciences (GM-39781). References and Notes (1) Frank, H. A.; Cogdell, R. J. In Carotenoids in Photosynthesis; Young, A., Britton, G., Eds.; Chapman and Hall: London, 1993; Chapter 8, p 252. (2) Frank, H. A.; Christensen, R. L. In Anoxygenic Photosynthetic Bacteria, AdVances in Photosynthesis; Blankenship, R. E., Madigan, M. T., Bauer, C. E., Eds.; Kluwer Academic Publishers: The Netherlands, 1995; p 373. (3) Cogdell, R. J.; Thornber, J. P. FEBS Lett. 1980, 122, 1. (4) Thornber, J. P.; Cogdell, R. J.; Pierson, B. K.; Seftor, R. E. B. J. Cell. Biochem. 1983, 23, 159. (5) Kramer, H. J. M.; van Grondelle, R.; Hunter, C. N.; Westerhuis, W. H. J.; Amesz, J. Biochim. Biophys. Acta 1984, 765, 156. (6) Breton, J.; Nabedryk, E. In The Light Reactions; Barber, J., Ed.; Elsevier Science Publishers B. V.: Amsterdam, 1987; Chapter 4, p 159. (7) van Grondelle, R.; Dekker: J. P.; Gillbro, T.; Sundstro¨m, V. Biochim. Biophys. Acta 1994, 1187, 1. (8) Ma, Y.-Z.; Cogdell, R. J.; Gillbro, T. J. Phys. Chem. B 1997, 101, 1087. (9) Zuber, H.; Brunisholz, R. A. In The Chlorophylls; Scheer, H., Ed.; CRC: Boca Raton, FL, 1993; p 627. (10) McDermott, G.; Prince, S. M.; Freer, A. A.; HawthornthwaiteLawless, A. M.; Papiz, M. Z.; Cogdell, R. J.; Isaacs, N. W. Nature 1995, 374, 517.

8162 J. Phys. Chem. B, Vol. 102, No. 42, 1998 (11) Koepke, J.; Hu, X.; Muenke, K.; Schulten, K.; Michel, H. Structure 1996, 4, 581. (12) Feher, G.; Okamura, Y. In The Photosynthetic Bacteria; Clayton, R. K., Sistrom, W. R., Eds.; Plenum Press: New York, 1978; Chapter 19, p 349. (13) Crounse, J. B.; Feldman, R. P.; Clayton, R. J. Nature 1963, 198, 1227. (14) Davidson, E.; Cogdell, R. J. FEBS Lett. 1981, 132, 81. (15) Theiler, R.; Suter, F.; Zuber, H.; Cogdell, R. J. FEBS Lett. 1984, 175, 231. (16) Davidson, E.; Cogdell, R. J. Biochim. Biophys. Acta 1981, 635, 295. (17) Hayashi, H.; Noguchi, T.; Tasumi, M. Photochem. Photobiol. 1989, 49, 337. (18) Noguchi, T.; Hayashi, H.; Tasumi, T. Biochim. Biophys. Acta 1990, 1017, 280. (19) Frank, H. A.; Farhoosh, R.; Aldema, M. L.; DeCoster, B.; Christensen, R. L.; Gebhard, R.; Lugtenburg, J. Photochem. Photobiol. 1993, 57, 49. (20) Farhoosh, R.; Chynwat, V.; Gebhard, R.; Lugtenburg, J.; Frank, H. A. Photosynth. Res. 1994, 42, 157. (21) Hudson, B. S.; Kohler, B. E.; Schulten, K. In Excited states; Lim, E. C., Ed.; Academic Press: New York, 1982; Vol. 6, p 2. (22) Hudson, B. S.; Kohler, B. E. Synth. Met. 1984, 9, 241. (23) Cogdell, R. J.; Frank, H. A. Biochim. Biophys. Acta 1987, 895, 63. (24) Snyder, R.; Arvidson, E.; Foote, C.; Harrigan, L.; Christensen, R. L. J. Am. Chem. Soc. 1985, 107, 4117. (25) Cosgrove, S. A.; Guite, M. A.; Burnell, T. B.; Christensen, R. L. J. Phys. Chem. 1990, 94, 8118. (26) DeCoster, B.; Christensen, R. L.; Gebhard, R.; Lugtenburg J.; Farhoosh, R.; Frank, H. A. Biochem. Biophys. Acta 1992, 1102, 107. (27) Owens, T. G.; Shreve, A. P.; Albrecht, A. C. In Research in Photosynthesis; Murata, N., Ed.; Kluwer Academic Publishers: The Netherlands, 1992; Vol. 1, p 179. (28) Frank, H. A.; Chynwat, V. Chem. Phys. 1995, 194, 237. (29) Trautman, J. K.; Shreve, A. P.; Violette, C. A.; Frank, H. A.; Owens, T. G.; Albrecht, A. C. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 215. (30) Shreve, A. P.; Trautman, J. K.; Frank, H. A.; Owens, T. G.; Albrecht, A. C. Biochim. Biophys. Acta 1991, 1058, 280. (31) Ricci, M.; Bradforth, S. E.; Jimenez, R.; Fleming, G. R. Chem. Phys. Lett. 1996, 259, 381. (32) Cohen-Bazire, G.; Sistrom, W. R.; Stanier, R. Y. J. Cell. Comput. Physiol. 1957, 49, 25. (33) Gebhard, R.; van Dijk, J. T. M.; van Ouwerkerk, E.; Boza, M. V. T. J.; Lugtenburg, J. Recl. TraV. Chim. Pays-Bas 1991, 110, 459. (34) van der Hoef, K.; Lugtenburg, J. (to be published). (35) Frank, H. A.; Cua, A.; Chynwat, V.; Young, A.; Gosztola, D.; Wasielewski, M. R. Photosynth. Res. 1994, 41, 389. (36) Frank, H. A.; Desamero, R. Z. B.; Chynwat, V.; Gebhard, R.; van der Hoef, I.; Jansen, F. J.; Lugtenburg, J.; Gostola, D.; Wasielewski, M. R.

Desamero et al. J. Phys. Chem. A 1997, 101, 149. (37) Freer, A.; Prince, S.; Sauer, K.; Papiz, M.; HawthornthwaiteLawless, A.; McDermott, G.; Cogdell, R.; Isaacs, N. W. Structure 1996, 4, 449. (38) Iwata, K.; Hayashi, H.; Tasumi, M. Biochim. Biophys. Acta 1985, 810, 269. (39) Koyama, Y. In Carotenoids; Britton, G., Liaaen-Jensen, S., Pfander, H., Eds.; Birkhauser Verlag: Basel, 1995; Vol. 1B, Chapter 5, p 135. (40) Kok, P.; Kohler, J.; Groenen, E. J. J.; Gebhard R.; van der Hoef, I.; Lugtenburg, J.; Farhoosh, R.; Frank, H. A. Spectrochim. Acta Part A 1997, 53, 381. (41) Cogdell, R. J.; Andersson, P. O.; Gillbro, T. J. Photochem. Photobiol. B 1992, 15, 105. (42) Andersson, P. O.; Gillbro, T.; Ferguson, L.; Cogdell, R. J. Photochem. Photobiol. 1991, 54, 353. (43) Cogdell, R. J.; Kerr, M. A.; Parson, W. W. Biochim. Biophys. Acta 1976, 430, 83. (44) Boucher, F.; van der Rest, M.; Gingras, G. Biochim. Biophys. Acta 1977, 461, 339. (45) Cogdell, R. J.; Crofts, A. R. Biochim. Biophys. Acta 1978, 502, 409. (46) Frank, H. A.; Violette, C. A. Biochim. Biophys. Acta 1989, 976, 222. (47) Cogdell, R. J.; Hipkins, M. F.; MacDonald, W.; Truscott, T. G. Biochim. Biophys. Acta 1981, 634, 191. (48) van Grondelle, R.; Kramer, H. J. M.; Rijgersberg, C. P. Biochim. Biophys. Acta 1982, 682, 208. (49) Borland, C. F.; Cogdell, R. J.; Land, E. J.; Truscott, T. G. J. Photochem. Photobiol. B 1989, 3, 237. (50) Gust D.; Moore, T. A. In AdVances in Photochemistry; Volman, D., Hammond, G., Neckers, D., Eds.; John Wiley & Sons: New York, 1991; Vol. 16, p 1. (51) Mathis, P.; Kleo, J. Photochem. Photobiol. 1973, 18, 343. (52) Truscott, T. G.; Land, E. J.; Sykes, A. Photochem. Photobiol. 1973, 17, 43. (53) Bensasson, R.; Land, E. J.; Maudinas, B. Photochem. Photobiol. 1976, 23, 189. (54) Fo¨rster, Th. Ann. Phys. 1948, 2, 55. (55) Davydov, A. S. Theory of Molecular Excitons; translated by Kasha, M., Oppenheimer, M., Jr.; McGraw-Hill: New York, 1962. (56) Nagae, H.; Kikitani, T.; Katoh, T.; Mimuro, M. J. Chem. Phys. 1993, 98, 8012. (57) Dexter, D. L. J. Chem. Phys. 1953, 21, 836. (58) Scholes, G. D.; Harcourt, R. D.; Fleming, G. R. J. Phys. Chem. 1997, 101, 7302. (59) Kasha, M. Radiat. Res. 1963, 20, 55. (60) Andersson, P. O.; Cogdell, R. J.; Gillbro, T. Chem. Phys. 1996, 210, 195.