|Home | About | Journals | Submit | Contact Us | Français|
Lincomycin-treated pumpkin leaves were illuminated with either continuous light or saturating single-turnover xenon flashes to study the dependence of photoinactivation of photosystem II (PSII) on the mode of delivery of light. The flash energy and the time interval between the flashes were varied between the experiments, and photoinactivation was measured with oxygen evolution and the ratio of variable to maximum fluorescence (Fv/Fm). The photoinhibitory efficiency of saturating xenon flashes was found to be directly proportional to flash energy and independent of the time interval between the flashes. These findings indicate that a low-light-specific mechanism, based on charge recombination between PSII electron acceptors and the oxygen-evolving complex, is not the main cause of photoinactivation caused by short flashes in vivo. Furthermore, the relationship between the rate constant of photoinactivation and photon flux density was similar for flashes and continuous light when Fv/Fm was used to quantify photoinactivation, suggesting that continuous-light photoinactivation has a mechanism in which the quantum yield does not depend on the mode of delivery of light. A similar quantum yield of photoinhibition for flashes and continuous light is compatible with the manganese-based photoinhibition mechanism and with mechanisms in which singlet oxygen, produced via a direct photosensitization reaction, is the agent of damage. However, the classical acceptor-side and donor-side mechanisms do not predict a similar quantum yield for flashes and continuous light.
Light is needed for photosynthesis, but light also damages the photosynthetic machinery. Light-induced inhibition of the activity of photosystem II (PSII) has been termed photoinhibition (for reviews, see Aro et al., 1993; Melis, 1999; Tyystjärvi, 2008; Vass and Aro, 2008). A characteristic feature of photoinhibition is that degradation and de novo synthesis of the D1 protein are needed for restoration of the activity of PSII. All light intensities cause inactivation of PSII with a rate constant directly proportional to the light intensity (Tyystjärvi and Aro, 1996), and therefore plants need to repair the inactive PSII units continuously. The rate of inactivation of PSII can be measured by illuminating leaves in the presence of an inhibitor such as lincomycin that prevents the concurrent synthesis of the D1 protein.
Photoinhibition is usually caused by continuous light, but strong, short single-turnover flashes also inactivate PSII (Keren et al., 1995, 1997). Single-turnover flashes do not occur in nature, but short flashes are valuable tools in studies of the mechanism of photoinactivation of PSII (see, for example, Keren et al., 1995, 1997, 2000; Szilard et al., 2005; Tyystjärvi et al., 2008).
Photoinactivation of PSII has been explained by several hypothetical mechanisms. In the acceptor-side hypothesis (Vass et al., 1992), PSII is inactivated when the primary quinone electron acceptor of PSII, already in state QA–, becomes double reduced to QA2– or stabilized by protonation. Simulations imply that double reduction would not occur under low light (Tyystjärvi et al., 2005) or during illumination with single-turnover flashes. However, if QA remains reduced from flash to flash, then double reduction might occur even during illumination with single-turnover flashes. It has also been suggested that photoinactivation is caused by chlorophyll (Chl) triplets produced in the presence of unstabilized QA– (Vass and Aro, 2008).
In the ‘statistical’ donor-side hypothesis (Anderson et al., 1998), PSII is inactivated when the oxygen-evolving complex is occasionally unable to donate an electron, that is a miss occurs in the oxygen-evolving complex (Kok, 1970; Antal et al., 2009). After a miss, the oxidized primary donor P680+ lives for 100–200μs and then recombines with QA– (Renger and Wolff, 1976; Renger and Holzwarth, 2005) or damages PSII by oxidizing an inappropriate target. Saturating single-turnover flashes would saturate the formation of long-lived P680+, and therefore donor-side photoinhibition would be saturated at the same flash energy that saturates PSII electron transfer.
Generation of singlet oxygen by uncoupled Chls (Santabarbara et al., 2001) or by iron–sulphur centres and cytochromes (Jung and Kim, 1990) only requires the formation of the triplet state of the photoreceptor and would therefore have the same quantum yield in continuous light and short flashes. Also manganese-dependent inactivation of PSII (Hakala et al., 2005; Ohnishi et al., 2005) would have the same quantum yield in continuous and flash light because manganese ions have low absorbance in the visible range, and therefore xenon flashes would not saturate absorption.
The low-light photoinhibition hypothesis (Keren et al., 1995, 1997, 2000; Szilard et al., 2005) applies to inactivation of PSII with short pulses of light or with low light. According to this hypothesis, generation of the triplet excited state of the primary donor (3P680*) by recombination reactions (e.g. S2QB–→S1QB) triggers the inactivation of PSII. The hypothesis is supported by the finding that nanosecond laser pulses fired with long dark intervals are more photoinhibitory than pulses fired with short intervals (Keren et al., 1997, 2000). This flash interval dependence has been explained by the fact that the flash yield of charge recombination increases with the length of the dark interval between the flashes. The low-light hypothesis is also supported by data showing that in the green alga Chlamydomonas, less degradation of the D1 protein per flash is caused by pairs of microsecond xenon flashes than by single flashes (Keren et al., 1995). In vitro experiments of inactivation of PSII with xenon flashes have produced contrasting results. The measurements of Szilard et al. (2005) suggest that the photoinhibitory efficiency of xenon flashes depends on the length of the interval between flashes, but the data of Hakala et al. (2005) did not show any flash interval dependence. The low-light hypothesis predicts that flash-induced inactivation of PSII becomes saturated at the same flash energy that saturates PSII electron transport.
The aim of the present study was to probe the mechanism of photoinactivation of PSII by comparing the effects of short flashes and continuous light in intact leaves. For this, photoinhibition from lincomycin-treated pumpkin (Cucurbita maxima) leaves was measured under continuous white light and under illumination with 3μs single-turnover xenon flashes, varying the energy of the flashes and the time interval between the flashes.
Pumpkin (C. maxima) plants were grown in a research greenhouse under the rhythm of 18h light/6h dark until leaves were fully expanded. Photosynthetic photon flux density (PPFD) was 150μmol m−2 s−1, relative air humidity was 70%, and temperature was 24°C.
Prior to illumination, leaves were incubated overnight with their petioles in 2.3mM lincomycin. The leaf petioles were also kept in 2.3mM lincomycin throughout the photoinhibition treatments during which the leaf blades were gently pressed on moist paper. A 400W Philips HPI-T Plus lamp was used for continuous-light photoinhibition experiments at a PPFD of 0.025, 3, 30, or 50μmol m−2 s−1 (duration 45h) and 300μmol m−2 s−1 or 500μmol m−2 s−1 (duration 4h). A 10cm water filter and a cooling fan were used to remove heat. For photoinhibition experiments at a PPFD of 900, 1500, 1800, or 2300μmol m−2 s−1 (duration 6–12h), a 300W high-pressure ozone-free xenon lamp (Oriel Instruments, Stratford, CT, USA) equipped with a 9cm water filter and a Schott GG400 UV-blocking filter was used. Neutral density filters (Lee Filters, Andover, UK) were used to fine-tune the PPFD, which was measured with a LI-189 quantum sensor (LiCor, Lincoln, NE, USA).
Photoinhibition experiments with xenon flashes were done by illuminating lincomycin-treated pumpkin leaves with 3μs flashes from an FX-200 xenon flash lamp (EG&G, Gaithersburg, MD, USA) through a Schott GG400 filter. The upper surface of the filter was cooled with an air flow. Capacitors were used to adjust the flash energy, and the time interval between the flashes was adjusted with a computer. The durations and flash energies of the flash treatments are listed in Table 1. Each flash treatment was done with 1800 flashes and the duration of the treatment was varied in accordance with the time interval between the flashes, as indicated in Table 1. The average PPFD of each flash treatment was determined as previously described (Hakala et al., 2005).
Prior to photoinhibition treatments, two leaf disks (2.8cm2) were always cut off from each leaf and analysed as control samples. In the middle and at the end of each continuous-light illumination treatment, two leaf disks were cut off from the illuminated area and used for measurements of fluorescence and oxygen evolution. One time point was used in the flash photoinhibition treatments. Pearson correlation coefficients and P-values reflecting their statistical significance were calculated with SPSS software (SPSS Inc., Chicago, IL, USA).
The dependence of the amount of oxygen evolved per flash on the energy of the xenon flashes was measured from pumpkin leaf disks (9.6cm2) with a Clark-type oxygen electrode (Hansatech, King's Lynn, UK) at 23°C in ~5% carbon dioxide, essentially as in Lee et al. (1999). The FX-200 xenon lamp was put in place of the standard lamp of the device, 5.7cm above the leaf, and the flash energy was adjusted to 0.1–1.35J. During the measurements, xenon flashes were applied for 4min, followed by 4min of darkness. As high a repetition rate of the flashes was used as technically possible in order to obtain a reasonable signal to noise ratio of the oxygen measurements, and therefore flashes of 0.1–0.4J were applied at 30 flashes per second (fps), flashes of 0.5–0.6J at 20fps, and flashes of 0.7–1.35J at 10fps. Several cycles of flashing (each with different energy, in random order) and darkness were measured from each leaf disk, and the measurements were repeated with several leaf disks.
Experiments were also done to determine the extent of dark inactivation of PSII activity during the maximum time span of the experiments (45h). These assays were done in the same set-up as was used for photoinhibition treatments, except that the leaves were covered with several layers of aluminium foil.
Chl a fluorescence of leaves was measured with a PAM-101 fluorometer (Heinz Walz GmbH, Effeltrich, Germany) and the FIP fluorescence program (Tyystjärvi and Karunen, 1990) was used to control the fluorometer. For measurements of Fv/Fm, unilluminated and illuminated pumpkin leaf disks were dark-incubated for 20min. The initial fluorescence level (F0) was measured under a weak measuring beam, and a saturating white light pulse (2s; 5800μmol m−2 s−1) was used to induce maximum fluorescence (Fm). For measurements of non-photochemical quenching (NPQ), pumpkin leaf disks were dark-incubated for at least 30min and NPQ was induced by 20min actinic illumination, either with continuous light (175, 300, 930, or 1350μmol m−2 s−1) or with xenon flashes of 1.35, 9.3, or 13.5J fired at 1fps. The respective PPFD values are 130, 930, and 1350μmol m−2 s−1. NPQ was calculated as (Fm–Fm′)/Fm′ (Bilger and Björkman, 1990).
The effect of the xenon flashes on the openness of PSII centers was tested with the Walz PAM-2000 fluorometer by flashing pumpkin leaf disks from 1cm distance through a GG400 filter under the light guide. After 2min of flashing with a xenon lamp, a saturating pulse of 0.8s (4500μmol m−2 s−1) was fired to measure Fm′, and then flashing was switched off and a far red light-emitting diode (LED) was switched on for 2s to measure F0′. These measurements were done using the flash rates of 1, 10, 20, or 30fps, and the respective flash energies of 13.5, 1.35, 0.6, and 0.1J.
Thylakoids were isolated from illuminated and unilluminated leaf disks as described by Sarvikas et al. (2010). Chl concentration was determined according to Porra et al. (1989) and adjusted to 10μg ml−1 with PSII measuring buffer (40mM HEPES-KOH pH 7.6; 0.33M sorbitol; 5mM MgCl2; 5mM NaCl; 1M glycine betaine; 1mM KH2PO4; 5mM NH2Cl). Oxygen evolution activity was measured with an oxygen electrode (Hansatech, King's Lynn, UK) in saturating red light with 0.125mM 2,6-dichlorobenzoquinone as electron acceptor. Chl concentrations of pumpkin leaves were determined in N,N-dimethylformamide (Inskeep and Bloom, 1985) before and after the illumination treatments.
Pumpkin leaves were first kept at a PPFD of 20μmol m−2 s−1 for 30min. Leaf petioles were then immersed either in control solution containing water and 2.5% dimethylsulphoxide (DMSO) or in solution containing 5μM cytochalasin D (cytD), in 2.5% (v/v) DMSO, and the leaves were kept at a PPFD of 20μmol m−2 s−1 until the transmittance was measured the following morning. Leaf disks (4.5cm2) were then placed on wet paper on a temperature-controlled surface and illuminated from the adaxial side at a PPFD of 1350μmol m−2 s−1 with either a 300W high-pressure ozone-free xenon lamp (continuous light) or an FX-200 flash lamp (xenon flashes, 13.5J, 1fps). Before the illumination treatment and every 10min after the beginning of illumination, the transmittance of the leaf was measured in a separate set-up in which the leaf was illuminated from the adaxial side with continuous light (PPFD 1350μmol m−2 s−1) and transmission was measured from the abaxial side with an LI-189 quantum sensor. The positions of the lamp and the quantum sensor were fixed and the leaf was always placed in exactly the same position in the transmission set-up. Measurement of transmission of light through a leaf is a standard method for measuring chloroplast movements (see, for example, Berg et al., 2006).
Laser scanning confocal microscopy was performed with an inverted confocal laser scanning microscope (Zeiss LSM510 Meta) with a ×20/0.50 water objective. Chloroplasts were imaged by Chl autofluorescence excited at 488nm with an argon diode laser and detected through a 650nm emission filter. Maximal projections of the sequential confocal images were created with Zeiss LSM Image Browser.
Properties of the flashes. In the photoinhibition experiments, flash energies above the saturation of PSII electron transfer were used. Such flashes cause one turnover of PSII irrespective of the flash energy. To ensure saturation, the flash energy response of PSII oxygen evolution was measured. In these measurements, the lamp to leaf distance was five times longer than in the photoinhibition set-up, and therefore the flash energy response of oxygen evolution provides an upper limit for the flash energy required to saturate PSII electron transport in the photoinhibition set-up. The results indicate that saturation of oxygen evolution of intact leaves occurred at 0.7–0.9J (Fig. 1). Pumpkin leaves contain ~600μmol Chl m−2, and therefore the maximum rate of oxygen evolution (1.2μmol O2 per four flashes) corresponded to 88% of reaction centres, assuming one PSII per 440 Chl molecules. Separate fluorescence measurements showed that PSII reaction centres remain open between the flashes when pumpkin leaves are flashed at 10–30fps as in the flash energy response measurements (Supplementary Table S1 available at JXB online).
For comparison of flashes and continuous light, average PPFD values for flash treatments were calculated, using the known relationship between the flash energy and flash photon content of this lamp (Hakala et al., 2005). In the original flash photon content measurements, the number of visible-light photons per flash was first measured at a narrow wavelength range (475–610nm) by illuminating meso-diphenylhelianthrene, a chemical actinometer substance (Brauer et al., 1983), and the flash photon content in the 400–700nm range was obtained by correcting these data with the emission spectrum of the flash lamp (Supplementary Fig. S2 at JXB online). Average PPFD values, used to compare flashes with continuous light, were obtained by dividing the flash photon content by the target area and multiplying by the number of flashes in unit time.
Experiments were also carried out to measure how efficiently the flashes induce NPQ, as compared with continuous light. For this, pumpkin leaves were illuminated for 20min either with continuous light of different intensities up to a PPFD of 1350μmol m−2 s−1 or with xenon flashes of 1.35, 9.3, or 13.5J fired at 1fps, producing an average PPFD range similar to continuous light (Table 1). Measurements of NPQ at the end of the illumination showed a linear increase of NPQ with PPFD of continuous light (Fig. 2). Illumination with xenon flashes at 1fps at the same PPFD range caused much less NPQ than continuous light, and flashes of 9.3J and 13.5J caused the same NPQ response (Fig. 2). Flashes fired at 90s intervals did not cause NPQ, and the NPQ response caused by flashes fired at 9s intervals was not statistically significant (data not shown). The coefficient of photochemical quenching qP remained at 0.98 during flashing of pumpkin leaf disks at 1fps with the flash energy of 13.5J (Supplementary Table S1 at JXB online).
Dependence of photoinactivation of PSII on flash energy. To induce photoinhibition with flashes, a lincomycin-treated pumpkin leaf was placed under the xenon flash lamp and flashing was repeated until 1800flashes had been fired. Two alternative methods were used to measure photoinhibition: light-induced decrease in Fv/Fm, measured from leaves, and light-induced decrease in the light-saturated rate of PSII oxygen evolution, measured from thylakoids isolated from treated leaves. The flash number-based rate constant of photoinactivation (kPI) was calculated by fitting the original data to the first-order reaction equation using the formula kPI=(1/1800) flash−1×ln(100/Act), where Act is the percentage of PSII activity (Fv/Fm or oxygen evolution, as indicated) left after treatment with 1800 flashes. Slow loss of PSII activity in the dark was measured separately and taken into account by subtracting the rate constant of dark inactivation from the raw kPI. The Chl content of the leaves remained at 80–100% of the control for 45h.
Comparison of photoinhibition induced with saturating xenon flashes of different energies indicated that the kPI of flash photoinactivation was directly proportional to the flash energy, regardless of whether kPI was determined from oxygen evolution or from Fv/Fm (Fig. 3). The correlation between kPI and flash energy was statistically highly significant whether the correlation was calculated from all data or from subsets with the same flash interval (Supplementary Table S2 at JXB online). No saturation of photoinactivation could be detected. Larger kPI values were obtained from oxygen evolution measurements than from fluorescence measurements (Fig. 3).
Dependence of photoinactivation on flash interval. To measure the effect of the flash to flash interval on inactivation of PSII in vivo, xenon flashes with 1, 9, and 90s intervals were applied on pumpkin leaves. The photoinhibitory efficiency of the flashes did not show a consistent dependence on the flash interval when Fv/Fm was used to measure PSII activity, and independence or a slight increase of photoinhibitory efficiency with flash interval was seen in the oxygen evolution data (Fig. 4). However, no statistically significant correlations between kPI values and flash interval could be found (Supplementary Table S2 at JXB online). Thus, when kPI values are plotted as a function of flash energy, all experiments were found to fall on the same line independently of the flash interval (Fig. 3).
The kPI values used to compare continuous-light-induced and flash-induced photoinactivation were obtained by fitting the data to the first-order reaction equation Act=100×exp(–kPI×t), where Act is the percentage activity remaining at time t. The kPI values of the flash and continuous-light treatments showed an essentially direct proportionality to PPFD irrespective of the measurement method (Fv/Fm or oxygen evolution) (Fig. 5), although some deviation from direct proportionality could be found both in very low light and at the high PPFD of 2300μmol m−2 s−1. The slope of the plot describing dependence of kPI on PPFD was the same for flash and continuous-light treatments when Fv/Fm was used to measure photoinhibition (Fig. 5A). Essentially the same slope was also obtained for continuous-light-induced photoinhibition measured with oxygen evolution (Fig. 5B, circles), whereas single-turnover flashes had a stronger inhibitory effect on oxygen evolution (Fig. 5B).
The oxygen evolution measurements represent all cell layers, whereas Chl fluorescence is emitted by the cell layers close to the leaf surface. Therefore, the difference between Fv/Fm and oxygen evolution in flash-photoinhibited leaves (Fig. 5) prompted measurement of chloroplast movements, as the avoidance response in which chloroplasts stack on each other can obviously not protect the topmost chloroplasts of the leaf. During 30min illumination with continuous light, PPFD 1350μmol m−2 s−1, the transmittance of pumpkin leaves increased from 7.5% to 10.4% (Fig. 6A). Transmission of blue light increased more than transmission of red light, and that of green light was intermediate (Fig. 6B).
The increase in transmission was apparently caused by chloroplast movements, as 5μM cytD, an actin-depolymerizing agent, partially inhibited the increase. For comparison, changes in the transmittance of Arabidopsis leaves were also studied, and it was found that 30min illumination induced an increase in transmission from 13% to 25%, twice as much as in pumpkin, and confocal microscopy showed light-induced movement of chloroplasts to the anticlinal cell walls in Arabidopsis (Supplementary Fig. S1 at JXB online). However, both before and after illumination, pumpkin chloroplasts appeared to stay in the anticlinal position in the mesophyll cells accessible with the confocal microscope (Supplementary Fig. S1).
When pumpkin leaves were illuminated for 30 min with xenon flashes (13.5J, 1fps, 1350μmol m−2 s−1), leaf transmission increased only to 8.4% from the initial value of 7.5% (Fig. 6A).
The aim of the present study was to gain insight into the mechanism of photoinactivation of PSII by illuminating intact lincomycin-treated pumpkin leaves with saturating 3μs single-turnover xenon flashes. The single-turnover nature of the flashes was ensured by showing that flashing does not cause accumulation of reduced QA (Supplementary Table S1 at JXB online) and that flashing causes very little NPQ even at the highest flash rate of 1fps (Fig. 2). Measurements of both oxygen evolution and Fv/Fm show that the photoinhibitory efficiency of single-turnover flashes is directly proportional to flash energy (Fig. 3). Furthermore, the fluorescence data showed that the photoinhibitory efficiency of photons of single-turnover flashes is similar to the photoinhibitory efficiency of photons of continuous light (Fig. 5). Direct proportionality between light intensity and kPI is a key feature of continuous-light photoinhibition (Jones and Kok, 1966; Park et al., 1995; Tyystjärvi and Aro, 1996; Allakhverdiev and Murata, 2004). Thus, the finding that the same proportionality constant applies to photoinhibition under flashing light and under continuous light (Fig. 5) suggests that the mechanism of photoinactivation is the same in both conditions.
Similar photoinhibitory efficiency of continuous light and flash light can be used as a criterion for evaluation of photoinhibition hypotheses. Full opening of PSII reaction centres between the flashes (Supplementary Table S1 at JXB online) allows the possibility that the acceptor-side mechanism (Vass et al., 1992) causes flash photoinhibition to be excluded, as neither double reduction of QA nor production of singlet oxygen by charge recombination due to stabilized QA– can happen unless QA is singly reduced when a single-turnover flash is fired. Proportionality between kPI and the energy of the saturating flashes also indicates that the ‘statistical’ donor-side mechanism (Anderson et al., 1998) cannot be responsible for flash photoinhibition, as the amount of long-lived P680+ cannot depend on how oversaturating a flash is.
The in vivo photoinhibitory efficiency of the flashes did not consistently depend on the time interval between the flashes (Fig. 4). This finding is in agreement with earlier in vitro data (Hakala et al., 2005). The flash interval range of 1–90s was chosen on the basis of the 35s half-time of the charge recombination reaction S2QB– →S1QB, measured with thermoluminescence (Keren et al., 1997). At 90s intervals, 83% of QB– produced by a flash recombines with the S2 state of the oxygen-evolving complex before the next flash comes, while at 1s intervals only 2% recombines and most QB– is converted to QB2– and replaced by oxidized plastoquinone. Thus, flashes fired at 90s intervals cause 40 times as many recombination reactions as flashes fired at 1s intervals. Lack of dependence on flash interval (Fig. 4) indicates that the low-light photoinhibition hypothesis which assumes that photoinactivation starts with charge recombination (Keren et al., 1997, 2000) does not explain flash photoinhibition in vivo. Furthermore, the low-light hypothesis cannot explain why the photoinhibitory efficiency of saturating flashes depends on flash energy, as the same number of recombination reactions is induced with all saturating flashes. The results of the present study are in agreement with the finding that the quantum yield of photoinactivation of PSII is the same from 6μmol m−2 s−1 to 2000μmol m−2 s−1 of continuous light (Tyystjärvi and Aro, 1996). Thus, a specific mechanism for photoinactivation of PSII under low light is improbable.
The manganese mechanism (Hakala et al., 2005) and generation of singlet oxygen by uncoupled Chls (Santabarbara et al., 2001) or by iron–sulphur centres (Jung and Kim, 1990) would have the same quantum yield for continuous light and single-turnover flash illumination, and therefore these mechanisms qualify as explanations for the present in vivo photoinhibition data. The manganese mechanism functions in parallel with photoinactivation mediated by the PSII antenna (Sarvikas et al., 2010). The results of the present study do not define the Chl-dependent mechanism, but generation of singlet oxygen may have a role in it.
In the present study, three flash intervals and five flash energies were tested. The possibility that use of subsaturating flashes or a larger number of flash intervals might provide more insight cannot be excluded.
A similar photoinhibitory efficiency of continuous light in Fv/Fm and oxygen evolution measurements (Fig. 5) is in agreement with earlier data (Krause et al., 1992; Schnettger et al., 1994). Flash-induced and continuous-light-induced photoinactivation had the same quantum yield in Fv/Fm measurements (Fig. 5A), but flashes were more efficient at causing photoinhibition of oxygen evolution (Figs 3, ,5).5). This finding can be explained by assuming that the photoinhibitory efficiency of flashes and continuous light is similar but chloroplasts of deep cell layers are more efficiently illuminated by flashes than by continuous light. The chloroplast avoidance response (Wada et al., 2003) would not protect the top layer of chloroplasts, and the finding that flashes cause very little chloroplast movements may partially explain why the flashes penetrate better in the leaf. The chloroplast avoidance response affects mostly transmission of blue light (Fig. 7), which is the most effective visible-light wavelength range in photoinhibition in vivo (Sarvikas et al., 2006). Chloroplast movements have little effect on leaf reflectance (Park et al., 1996) and therefore transmission measurements estimate changes in leaf absorptance.
The data presented in this paper lead to two important conclusions about the mechanism of photoinhibition in vivo. First, the finding that the photoinhibitory efficiency of saturating flashes depends on flash energy but is independent of flash to flash interval indicates that charge recombination between PSII electron acceptors and the oxygen-evolving complex cannot be the trigger of photoinhibition caused by short flashes in vivo. Secondly, photons of continuous light and photons of short flashes were found to have similar photoinhibitory efficiency, suggesting that continuous-light photoinhibition has a mechanism in which the quantum yield does not depend on the mode of delivery of light.
Supplementary data are available at JXB online.
Figure S1. Confocal microscope images from pumpkin and Arabidopsis leaves.
Figure S2. Emission spectra of the xenon flash lamp and the continuous-light lamps used in this study.
Table S1. Effect of illumination with xenon flashes on photochemical quenching in pumpkin leaves.
Table S2. Pearson correlation coefficients between kPI and flash energy, and between kPI and flash interval.
This work was financially supported by the Academy of Finland (grant number 110409), the Turku University Foundation, the Finnish Cultural Foundation, and the Emil Aaltonen Foundation.