Unusual Optical Properties of the Chl a Molecule as a New Photoprotection Mechanism

High-resolution crystal structures (Kurisu et al. 2003; Stroebel et al. 2003) of the cyt b6f complex unequivocally confirm the earlier findings through biochemical analysis (Huang et al. 1994; Pierre et al. 1995) of a Chl a molecule as an intrinsic component of the cyt b6f complex and raises intriguing questions about the role of the Chl a. The function of the cyt b6f complex does not require light-harvesting, and the Chl a molecule is not part of the electron transfer (ET) chain, which are the usual functions of Chl molecule in photosynthetic complexes. Moreover, the functionally similar cyt bc1 complex of the respiratory chain does not contain a Chl a molecule (Xia et al. 1997; Iwata et al. 1998).
     Regardless of the role of the Chl a in the cyt b6f complex, the introduction of the Chl a molecule into the structure may pose a serious threat to the stability of the complex. The triplet excited state of Chl a molecule is known to transfer its energy with high efficiency to oxygen, generating singlet oxygen that is extremely toxic to the pigment-protein complex (Fujimori and Livingston 1957; Krinsky 1979). Under illumination, the triplet excited state of monomeric Chl a in solution forms with high quantum yield (~0.64) through intersystem crossing from the Chl a singlet excited state (Bowers and Porter 1967).

The problem: Excitation of Chl a may lead to creation of harmful singlet oxygen!
To prevent singlet oxygen formation in Chl-containing proteins, carotenoid (Car) is typically positioned close (~4 Å) to the Chl a molecule, effectively quenching the triplet excited state of the Chl a due to rapid triplet-triplet energy transfer to Car (Foote 1976; Siefermann-Harms 1987). It was widely expected that a similar protection mechanism would exist in the b6f complex. In fact, along with the Chl a, a β-carotene was found to be stoichiometrically bound in the b6f complex (Zhang et al. 1999). It was also reported that, qualitatively, the Chl a photodegradation rate depended inversely on the Car concentration (Zhang et al. 1999).
     However, the structures of the cyt b6f complex (Kurisu et al. 2003; Stroebel et al. 2003) show that the β-carotene is too far removed from the Chl a for effective direct quenching of the Chl a triplet excited state. Triplet-triplet energy transfer from Chl to Car occurs via Dexter exchange mechanism and requires the interacting cofactors to form a collision complex, restricting triplet-triplet energy transfer to rather short distances (Renger 1992; van Grondelle et al. 1994).
In this study, we report that effective protection against singlet oxygen formation in the cyt b6f complex is, at least in part, realized through an alternative mechanism that has not been reported previously. Optical ultrafast pump-probe experiments reveal that the Chl a in the enzymatically active and ultrapure cyt b6f complex exhibits an unusually short excited state lifetime of ~200 ps that is a factor of ~25 times shorter than the fluorescence lifetime of monomeric Chl a in solution. We found that, due to the short lifetime of the Chl singlet excited state, the formation of the triplet excited state of the Chl a is dramatically reduced, which would result in a decrease of the consequent production of singlet oxygen in the b6f complex. It was inferred that excitation-induced ET interaction with nearby aromatic amino acid residue(s) is the most likely explanation of the observed effect. Based on the new structure data, it is proposed that the local structure around the Chl a facilitates rapid quenching of the Chl singlet excited state and thereby minimizes the formation of singlet oxygen.
Conserved geometry of the Chl a binding: According to the high-resolution crystallographic structures (Kurisu et al. 2003; Stroebel et al. 2003) of the cyt b6f complex, the monomeric Chl a molecule is bound between helices F and G of subunit IV, with the phytyl tail threaded through the portal connecting the electropositive Qp quinone binding niche, and the central quinone-exchange cavity (Fig. 1). The chlorin ring of the Chl a is parallel to the plane of the heme bn, which is separated by approximately 16 Å center-to-center from the Chl a, whereas the β-carotene molecule near the center of the transmembrane region is approximately 14 Å (the closest distance) from the Chl a. These geometrical configurations of the Chl a binding are essentially identical in both crystal structures from two species, whose appearance in evolution is separated by 10 billion years.
Figure 1. The nanospace of the monomeric Chl a molecule in the cyt b6f complex. Edge-to-edge distances (Å) to three closest TYR and TRP residues are specified in the context of ET theory (Hanson 1990; Moser et al. 1995; Page et al. 1999).
Ultrafast Kinetics of the Chl a Singlet Excited State
The ultrafast kinetics of the singlet excited state of the monomeric Chl a in the cyt b6f complexes were probed by femtosecond time-resolved pump-probe spectroscopy. The samples were excited at 660 nm and absorbance difference kinetics were recorded at 5 nm intervals at multiple probe wavelengths covering the entire Qy absorption band of the Chl a between 665 nm and 695 nm.
Figure 2. All time-resolved absorption difference profiles for the ultra-pure cyt b6f-crystal complex probed at several wavelengths between 665 nm and 695 nm. Positive going amplitudes correspond to photobleaching/stimulated emission and/or excited state absorption rise components. Absorbance units are absolute under the experimental conditions.
Unusually short excited state lifetime: Global analysis of the data obtained in ultrafast time-resolved pump-probe experiments reveals that the Chl a excited state population dynamics can be described by a single lifetime of 200 ps. This lifetime is ~25 times shorter than the excited state lifetime of the monomeric Chl a in solution.
Figure 3. Time-resolved anisotropy of absorption difference signal porbed at 680 nm for the ultra-pure cyt b6f-crystal complex. Corresponding anisotropic profiles are shown in the inset.
Ultrafast time-resolved anisotropy: Stroebel et al. (2003) suggested that the Chl a could facilitate interaction with other components of the photosynthetic apparatus. The plane of the Chl a may be, for example, “twisted” in response to absorption of light and cause structural response of the protein that could propagate through the cyt b6f complex and signal proteins, e. g., LHC kinase, on the stromal side of the membrane. To test this hypothesis, we monitored the orientation of the Chl a in the excited state by measuring the anisotropy r(t) of the absorption difference signal probed by polarized light at wavelength ranging from 665 nm to 695 nm after excitation of the complex at 660 nm. The anisotropy dynamics showed negligible dependence on the probe wavelength. The anisotropy is close to the theoretical maximum of 0.4 for a completely rigid molecule with parallel absorption and emission transition dipole moments. It is almost constant during the lifetime of the Chl a excited state indicating that, during this time (~200 ps), Chl a excitation does not lead to detectable reorientation of the molecule (Fig. 3).
The Excited State Quenching Mechanism
There are 3 possible mechanisms that may cause the observed rapid quenching of the Chl a singlet excited state: (1) increased rate of intersystem crossing, (2) interaction with the heme, and (3) excitation induced ET between the Chl a and nearby aromatic amino acid residue(s).
Increased rate of intersystem crossing: To account for the observed ~200 ps excited state lifetime of the Chl a in the b6f complex, the local protein environment must increase the intersystem crossing rate 40–50 fold, which would result in Chl a triplet formation with 98% efficiency. Since the triplet excited state lifetime is ~200 ns under aerobic conditions (Fujimori and Livingston 1957), this will also significantly delay the recovery of the Chl a ground state. Our absorbance difference measurements, however, show that the recovery of the Chl a ground state in the b6f complex occurs within ~200 ps and rules out the possibility that any significant amount of Chl a triplet state can be formed. Thus, we can rule out intersystem crossing as a possible mechanism responsible for the unusually short singlet excited state lifetime of the Chl a.
Interaction with the nearby heme: According to high-resolution crystal structures of the cyt b6f complex (Kurisu et al. 2003; Stroebel et al. 2003), the heme bn of cyt b6 is parallel to the chlorin ring of the Chl a, from which it is separated by 16 Å (Fe-Mg center-to-center distance). However, dithionite reduction of the initially oxidized heme revealed no dependence of the Chl singlet excited state lifetime on the redox state of the nearby hemes, ruling out the involvement of the heme in the quenching process. Similar results were obtained by Peterman et al. (1998) for the enzymatically inactive b6f complex.
Excitation induced ET between the Chl a and nearby aromatic amino acid residues: It has been demonstrated by several groups (Karen et al. 1983; Visser et al. 1987; Mataga et al. 2000; Zhong and Zewail 2001) that in flavin-binding proteins the electronic excited state of a chromophore can be efficiently quenched via ET exchange with a nearby aromatic amino acid residue. The possible involvement of aromatic residues in fluorescence quenching of bacteriochlorophyll and chlorophyll molecules was also discussed by Li et al. (1997) and Peterman et al. (1998).
     The sequence of ET events that may lead to quenching of the singlet excited state of the Chl a is depicted in the redox potential diagram shown in Fig. 4A. When a Chl a molecule is excited its electron donating (oxidation) potential becomes more negative, while its electron accepting (reduction) potential becomes more positive by a value approximately equal to the singlet excitation energy (Watanabe and Kobayashi 1990; Jones and Fox 1994; Oda et al. 2001). The energy of the Qy transition of the Chl a is ~1.85 eV and using the Chl reduction potential of -0.88 V (Watanabe and Kobayashi 1990) the electron accepting potential becomes anodic enough to initiate the first electron transfer step from nearby Tyr residue since the Tyr electron donating potential is about + 0.93 V (Harriman 1987; DeFelippis et al. 1989; Jovanovic et al. 1991). Once an electron is transferred from the Tyr to the Chl singlet excited state, the singlet excited state of the Chl a is transformed into a non-fluorescent reduced state. The Chl electron accepting potential is, however, significantly more negative than the Tyr electron donating potential, forcing electron transfer from the reduced Chl back to the Tyr. As the result, both reactants return to their neutral ground states (Fig. 4A).
Figure 4. (A) The proposed quenching mechanism of the Chl singlet excited state a by excitation induced ET process. Absorption of a photon promotes the Chl a into its singlet excited state and rises the oxidation potential from –0.88 V to +0.97 V (wavy arrow). In the following quenching process, the electron is first donated by a nearby Tyr to Chl, transforming the Chl singlet excited state into unexcited Chl reduced state (arrow 1). In the second ET step (arrow 2), the reduced Chl donates electron to oxidized Tyr, resulting in neutral Chl and Tyr and completing the quenching process. (B) Similar scenario proposed for fluorescence quenching of the riboflavin (RF) in riboflavin-binding protein with the experimentally measured lifetimes of ET (Mataga et al. 2000; Zhong and Zewail 2001).
This model of the quenching kinetics of the singlet excited state of the Chl a can be described by the following sequential ET scheme:
where k1 and k2 are the intrinsic rates of the respective ET processes denoted by arrows “1” and “2” in the redox potential diagram (Fig. 4A). To reproduce the essentially single-exponential absorption difference kinetics measured at ~670 nm in the proposed scenario, the rate of the first ET should be close to ~1/(200 ps), while the lifetime of the second ET step should be faster than ~1/(150 ps) (Fig. 4A).
Calculation of the Electron Transfer Rates
The mechanism of ET and underlying theory have been described (see, for example, Gray and Winkler 2003; Page et al. 2003). In the following, we used the Moser-Dutton semi-empirical relationship (Page et al. 1999) to estimate the rates of exothermic (downhill) ET for steps 1 and 2 at room temperature:
where R is the edge-to-edge distance between donor and acceptor, ΔG° is the Gibbs free energy change for the ET, and λ is the reorganization energy. By comparing the values of R and ΔG° for all nearest aromatic residues we inferred that Tyr105 ( R=6 Å) is the most likely residue responsible for electron-transfer mediated quenching of the Chl a excited state. Using a value of λ=0.7 eV for the reorganization energy (Page et al. 2003), the equation predicts a rate of k1=1/(234 ps) for the first ET step, which is in good agreement with the experimental value 1/(200 ps). However, this equation results in the rate constant k2=1/(0.8 ms) for the second ET, which is 4 orders of magnitude slower than the observed recovery rate of the Chl a ground state. This dramatic discrepancy may stem from uncertainties in reorganization energy and/or redox potential values. For example, using λ=1.05 eV (still within the range of 0.9±0.2 eV, cited by Page et al. 2003), the relationship results in a rate of 1/(140 ps) for the second electron transfer, which is not inconsistent with the experimental results.
     Similar high electron back-transfer rates to amino acid residues have been measured for the ET mediated fluorescence quenching of the riboflavin (RF) in riboflavin-binding protein (Mataga et al. 2000; Zhong and Zewail 2001) (Fig. 4B). The RF reduction potential (–0.8 eV) is very close to that of the Chl (–0.88 eV), which makes the comparison between these two cases especially relevant. Using transient absorption and fluorescence spectroscopy, Zong and Zewail (2001) determined that the first and second electron transfer steps occur with lifetimes ~100 fs and ~8 ps, respectively and proposed that ET exchange with the nearby Trp residue was responsible for the quenching of the RF excited state (Fig. 4B ). The Trp oxidation potential is 1.03 eV (Harriman 1987; DeFelippis et al. 1989; DeFelippis et al. 1991), R=3.7 Å and, using λ=0.7 eV, the equation yields lifetimes 160 fs and 52 ns for the first and second ET steps, respectively. As in the case of the Chl a and Tyr105 in the cyt b6f complex, the kinetics of the first ET step in riboflavin-binding protein is described very well by the equation, but the rate calculated for the second ET step is 4 orders of magnitude slower than the measured value. Agreement could, however, be attained if a reorganization energy λ=1.04 eV is used for the second ET step. This value of the reorganization energy is consistent with a value of λ=1.05 eV for the Chl to Tyr ET step required to reproduce the experimental data by the proposed kinetic model. It was concluded that the electron transfer mediated quenching is the most plausible mechanism responsible for the unusually short lifetime of the singlet excited state of the Chl a in the cyt b6f complex.
Photochemical degradation of the Chl a in the cyt b6f complex and protection against singlet oxygen formation: Using triplet-triplet energy transfer theory described in (Dexter 1953) and data published elsewhere (Schödel et al. 1998; Bodunov and Berberan-Santos 2004), we estimated that triplet-triplet energy transfer from the Chl to Car in the b6f complex should occur in ~0.3 ms, which is much too slow to compete with singlet oxygen formation. Thus, the conventional mechanism of singlet oxygen protection by the direct triplet-triplet energy transfer process does not apply to the cyt b6f complex.
Conclusion: While the β-carotene is too far from the Chl a for direct protection, our photodegradation experiments demonstrate that the Chl a in the b6f complex is 130–140 times more stable than in solution. We propose that protection, at least in part, is realized through specific arrangement of the local protein environment of the Chl a to ensure rapid quenching of the Chl singlet excited state. Shortening of the Chl singlet excited state lifetime from 5–6 ns to 200 ps causes a 25–30 fold decrease in the quantum yield of the Chl triplet excited state formation, and thus reduces the rate of singlet oxygen formation. To the best of our knowledge, this mechanism of Chl protection against singlet oxygen formation has not been yet reported. The unusually short singlet excited state lifetime of the Chl a in the cyt b6f complex can account only for 25–30 fold protection, while our experiments reveal that Chl a in the complex is 130–140 times more stable than monomeric Chl a in solution. This implies that one or more additional unconventional protection mechanism(s) exist in the cyt b6f complex.

Related Publications

Dashdorj N, Zhang H, Kim H, Yan J, Cramer WA, and Savikhin S. The Single Chlorophyll a Molecule in the Cytochrome b6f Complex: Unusual Optical Properties Protect the Complex against Singlet Oxygen, Biophysical Journal 88: 4178-4187 (2005). | PDF
Dashdorj N, Zhang H, Kim H, Yan J, Cramer WA, and Savikhin S. Unusual Optical Properties of the Monomeric Chlorophyll a in the Cytochrome b6f Complex of Oxygenic Photosynthesis, In Photosynthesis: Fundamental Aspects to Global Perspectives. Art van der Est and Doug Bruce, Editors, ACG Publishing, Lawrence, KS, U.S.A. 451-453 (2005). | PDF

Bibliography

Bodunov EN and Berberan-Santos MN. Chem. Phys. 301: 9-14 (2004).
Bowers PG and Porter G. Proc. R. Soc. A 296: 435-441 (1967).
DeFelippis MR et al. J. Phys. Chem. 95: 3416 - 3419 (1991).
DeFelippis MR, Murthy CP, Faraggi M, and Klapper MH. Biochemistry 28: 4847-4853 (1989).
Dexter DL. J. Chem. Phys. 21: 836-850 (1953).
Foote CS. 1976. In Free Radicals in Biology.Pryor WA, editor, Academic Press, New York. 85-133.
Fujimori E and Livingston R. Nature 180: 1036-1038 (1957).
Gray HB and Winkler JR. Q. Rev. Biophys. 36: 341-372 (2003).
Hanson LK. 1990. In Chlorophylls.Scheer H, editor, CRC Press, Boca Raton. 993-1014.
Harriman A. J. Phys. Chem. 91: 6102-6104 (1987).
Huang D et al. Biochemistry 33: 4401-4409 (1994).
Iwata S et al. Science 281: 64-71 (1998).
Jones WE and Fox MA. J. Phys. Chem. 98: 5095-5099 (1994).
Jovanovic SV, Steenken S, and Simic MG. J. Phys. Chem. 95: 684-687 (1991).
Karen A, Ikeda N, Mataga N, and Tanaka F. Photochem. Photobiol. 37: 495-502 (1983).
Krinsky NI. Pure Appl. Chem. 51: 649-660 (1979).
Kurisu G, Zhang H, Smith JL, and Cramer WA. Science 302: 1009-1014 (2003).
Li Y-F, Zhou W, Blankenship RE, and Allen JP. J. Mol. Biol. 271: 456-471 (1997).
Mataga N et al. J. Phys. Chem. B 104: 10667-10677 (2000).
Moser CC, Page CC, Farid R, and Dutton PL. J. Bioenerg. Biomembr. 27: 263-274 (1995).
Oda N, Tsuji K, and Ichimura A. Anal. Science 17: i375-i378 (2001).
Page CC, Moser CC, Chen X, and Dutton PL. Nature 402: 47-52 (1999).
Page CC, Moser CC, and Dutton PL. Curr. Opin. Chem. Biol. 7: 551-556 (2003).
Peterman EJG et al. Biophys. J. 75: 389-398 (1998).
Pierre Y, Breyton C, Kramer D, and Popot J-L. J. Biol. Chem. 270: 29342-29349 (1995).
Renger G. 1992. In The Photosystems: Structure, Function and Molecular Biology. Barber J, Editor, Elsevier, Amsterdam, The Netherlands. 45-99.
Schödel R, Irrgang K-D, Voigt J, and Renger G. Biophys. J. 75: 3143-3153 (1998).
Siefermann-Harms D. Physiol. Plantarum 69: 561-568 (1987).
Stroebel D, Choquet Y, Popot J-L, and Picot D. Nature 426: 413-418 (2003).
van Grondelle RJ, Dekker JP, Gillbro T, and Sundstrom V. Biochim. Biophys. Acta 1187: 1-65 (1994).
Visser AJWG, van Hoek A, Kulinski T, and Gall JL. FEBS Lett. 224: 406-410 (1987).
Watanabe T and Kobayashi M. 1990. In Chlorophylls.Scheer H, Editor, CRC Press, Boca Raton. 287-315.
Xia D et al. Science 277: 60-66 (1997).
Zhang H, Huang D, and Cramer WA. J. Biol. Chem. 274: 1581-1587 (1999).
Zhong D and Zewail AH. Proc. Natl. Acad. Sci. U.S.A 98: 11867-11872 (2001).