Report

Carotenoid Cation Formation and the Regulation of Photosynthetic Light Harvesting

See allHide authors and affiliations

Science  21 Jan 2005:
Vol. 307, Issue 5708, pp. 433-436
DOI: 10.1126/science.1105833

Abstract

Photosynthetic light harvesting in excess light is regulated by a process known as feedback deexcitation. Femtosecond transient absorption measurements on thylakoid membranes show selective formation of a carotenoid radical cation upon excitation of chlorophyll under conditions of maximum, steady-state feedback deexcitation. Studies on transgenic Arabidopsis thaliana plants confirmed that this carotenoid radical cation formation is correlated with feedback deexcitation and requires the presence of zeaxanthin, the specific carotenoid synthesized during high light exposure. These results indicate that energy transfer from chlorophyll molecules to a chlorophyllzeaxanthin heterodimer, which then undergoes charge separation, is the mechanism for excess energy dissipation during feedback deexcitation.

The regulation of photosynthetic light harvesting through a feedback deexcitation quenching mechanism (qE) is one physiologically important strategy used by plants to minimize the deleterious effects of short-term high light exposure (13). qE involves harmless thermal dissipation of excess energy in the chlorophyll (Chl) singlet excited states (1Chl*) in photosystem II (PSII) of green plants and algae so as to minimize alternative reaction pathways that generate toxic photo-oxidative intermediates (46). Elucidation of the biophysical mechanism of this vital regulatory process is fundamental for understanding photosynthesis on a molecular scale. In addition, because qE has been shown to be important for plant fitness (3), it is a requirement for engineering natural and artificial photosynthetic systems to be more robust when exposed to fluctuations in light intensity.

In excess light, a low thylakoid lumen pH (7, 8) has two effects: It activates formation of the carotenoid (Car) zeaxanthin (Zea) from violaxanthin via the xanthophyll cycle (9) (Fig. 1A), and it drives protonation of PsbS (10, 11), a PSII subunit that is necessary for qE in vivo (12). Determining why Zea is necessary for complete qE induction is central to an understanding of the mechanism(s) of 1Chl* deactivation during qE. Transient absorption (TA) kinetics recorded previously in the spectral region from 530 to 580 nm upon selective excitation of Chl indicated direct Car excitation during qE (13).

Fig. 1.

(A) Xanthophyll cycle. (B) Chlorophyll fluorescence in isolated spinach thylakoid membranes without (black bar above the graph) and with (white bar) continuous high light illumination. F0 is the fluorescence with open reaction centers in the absence of high light; Fm and Fm′ denote fluorescence with closed reaction centers (achieved with a saturating pulse, ∼2200 μmol photons m–2 s–1, 1 s) in the absence and presence of continuous high light, respectively. (C) TA data for spinach thylakoids upon excitation at 664 nm and detection at 1000 nm under quenched (circles) and unquenched (triangles) conditions and their corresponding fits (solid lines). (D) Difference between quenched and unquenched TA curves detected at 1000 nm upon excitation at 664 nm (diamonds) and the corresponding fit (solid line). OD, optical density; a.u., arbitrary units.

We performed ultrafast TA measurements upon excitation of the first excited singlet state of Chl (Qy) at 664 nm under conditions where no qE was present (unquenched) or where maximum, steady-state qE (quenched) was induced and maintained with constant illumination from the high light source (14), as determined by the Chl fluorescence (Fig. 1B). We recorded the kinetics in the region from 900 to 1080 nm where the carotenoid radical cation (Car) absorbs (1517). The traces for spinach thylakoids upon excitation at 664 nm and detection at 1000 nm (the probe wavelength where the largest amplitude differences were observed) showed an additional rise-and-decay component in the quenched case relative to the unquenched case (Fig. 1C). Subtraction of the quenched trace from the unquenched trace and a corresponding fit of the data produced a single exponential rise and decay of ∼11 ps and ∼150 ps, respectively (Fig. 1D). The reconstructed kinetic difference spectrum from kinetic traces measured in the 900- to 1080-nm region upon excitation at 664 nm and a delay time of 20 ps, which corresponds to the maximal difference between the quenched and unquenched traces at 1000 nm, is shown in Fig. 2.

Fig. 2.

Near-IR reconstructed quenched minus unquenched difference spectrum (solid line with circles) for the kinetics measured upon excitation at 664 nm at a time delay of 20 ps in isolated spinach thylakoid membranes. The estimated error for all spectral points is ±30%. The spectrum of β-Car (dashed line) from O2-evolving Synechocystis PSII core complexes at 20 K is also shown (16).

Additional experiments were performed on thylakoids from two Arabidopsis thaliana mutants in addition to the wild type (WT): (i) WT + psbS, which overexpresses PsbS and thus has ∼2.5 times the qE of the wild type (2), and (ii) npq4-1, which lacks the psbS gene and is therefore completely qE-deficient (12). Despite the lack of qE capacity, npq4-1 carries out light-induced photochemistry at the same rate as the wild type (12). The TA measurements for the three genotypes upon excitation at 664 nm and detection at 1000 nm showed kinetic differences with time constants similar to those observed in spinach only for the wild type and WT + psbS, with larger amplitude differences in the latter (Fig. 3). Hence, the kinetic differences we observe are correlated with qE.

Fig. 3.

TA data (circles) detected at 1000 nm upon excitation at 664 nm for (A) WT + psbS, (B) wild-type, and (C) npq4-1 A. thaliana plants. Quenched (circles) and unquenched (triangles) kinetic data are shown with their corresponding fits (solid lines). All kinetics were normalized with respect to the maximum ΔOD of the quenched trace.

TA measurements were also performed on wild-type A. thaliana thylakoids and three mutants with distinct Car composition. The wild-type plant has all of the components necessary to form Zea in high light and has the highest amount of qE of the four plants. The npq2 and npq2lut2 mutants lack activity of the Zea epoxidase enzyme, resulting in constitutively high levels of Zea in all light conditions (18, 19). The npq2 mutant contains the xanthophylls Zea and lutein (Lut), whereas the only xanthophyll pigment in npq2lut2 is Zea (19). qE studies on these two mutants have shown that Zea is dominantly responsible for qE in A. thaliana, whereas Lut has a minor role in affecting the rate of qE induction and the net amount of quenching (18). As a result, both npq2 and npq2lut2 have qE levels comparable to those of the wild type. The npq1 mutant lacks Zea in low light and cannot form it in high light, making it severely qE-deficient (6). The TA studies upon excitation at 664 nm and detection at 1000 nm for npq2 and npq2lut2 (20) showed an additional rise-and-decay component in the quenched case relative to the unquenched case, with time constants similar to those obtained for spinach and the wild-type and WT + psbS A. thaliana plants. The kinetics observed for npq1 with and without high light illumination were very similar, displaying only minor differences in the decay component, but neither signal included a rise component (20). The magnitude of the differences varied as follows: wild type ≥ npq2npq2lut2npq1. These findings demonstrate that Zea is necessary to produce the kinetic differences.

On the basis of the similarity in absorption maxima and spectral widths between the kinetic difference spectrum (Fig. 2) and the ground-state absorption spectra of the β-carotene (16) and spheroidene (15) radical cations, we conclude that the changes observed are due to the formation of Car selectively under quenched conditions. Identification of the species as a Car is consistent with the large amplitude of the differences probed in the near-infrared (near-IR) because the absorption cross section of Car is similar to the value for the Car S0 → S2 transition (21), which is substantially larger than the excited-state absorption cross section of Chl, the transition that dominates the signal detected at 1000 nm under unquenched conditions (22). Moreover, the lack of additional strongly allowed transitions from other photosynthetic chromophores in the probe region supports the assignment of the species as a Car. The demonstrated necessity of Zea for the generation of the kinetic changes enables specific assignment of the species observed during qE as a zeaxanthin radical cation (Zea).

The observation of Car formation upon selective excitation of Chl is in agreement with previous experimental findings in model systems comprising Cars and molecules that are structurally and spectroscopically similar to Chl (23, 24), as well as theoretical studies (25). However, our work shows that Zea formation is correlated with qE. The result is in line with calculations that showed Zea to have the lowest ionization potential of the three xanthophyll-cycle Cars (Fig. 1A) (25). A model that is consistent with all our data (13) involves three species: a bulk Chl pool (Chlbulk), a Chl-Zea heterodimer that quenches excited Chlbulk molecules, and a charge-separated ground state consisting of Chl–· and Zea (Fig. 4). The charge-separated state is formed from relaxation of the Chl-Zea excited state, (Chl-Zea)*, which is probably a charge-transfer state (25). In this scheme, the ∼11-ps component corresponds to the net dynamics of the Chlbulk molecules that transfer to Chl-Zea. Simulations of the kinetics, including Chlbulk annihilation dynamics, show that energy transfer from Chlbulk to Chl-Zea occurs in ∼15 to 200 ps and the relaxation time scale of (Chl-Zea)* to form Chl–· and Zea is on the order of 0.1 to 1 ps. Such a fast heterodimer relaxation time scale would ensure efficient quenching because it would prevent the energy that is transferred to the heterodimer from returning to Chlbulk. The Zea signal decays on a time scale of ∼150 ps, which corresponds to charge recombination between Chl–· and Zea. Integration of the photophysical quenching pathways uncovered in this work with the currently emerging spatial and dynamical picture of the entire PSII complex (26), including high-resolution crystal structure information on the major light-harvesting protein LHCII (27), is necessary for precise determination of the time scales that generate efficient regulation of photosynthetic light harvesting.

Fig. 4.

Scheme of the qE quenching mechanism, showing generation of Zea after selective excitation of the Chl Qy band at 664 nm. The numbers correspond to (1) energy transfer from excited bulk Chl (Embedded Image) to Chl-Zea, (2) relaxation of (Chl-Zea)* to Chl–· and Zea, (3) charge recombination of Chl–· and Zea to Chl-Zea, and (4) all dynamics that occur in Embedded Image, with the exception of energy transfer to Chl-Zea.

The detection of selective formation of Zea under conditions of maximum, steady-state qE identifies the key molecular component involved in energy dissipation in PSII. Our results strongly suggest that the mechanism of nonradiative deactivation of 1Chl* during excess light occurs by excitation transfer to a Chl-Zea heterodimer, followed by ultrafast Car formation.

Supporting Online Material

www.sciencemag.org/cgi/content/full/307/5708/433/DC1

Materials and Methods

References

References and Notes

View Abstract

Navigate This Article