Article pubs.acs.org/JPCB
Two Different Mechanisms Cooperate In The Desiccation-Induced Excited State Quenching In Parmelia Lichen Chavdar Slavov,† Michael Reus, and Alfred R. Holzwarth* Max Planck Institute for Chemical Energy Conversion, D-45470 Mülheim a.d. Ruhr, Germany S Supporting Information *
ABSTRACT: The highly efficient desiccation-induced quenching in the poikilohydric lichen Parmelia sulcata has been studied by ultrafast fluorescence spectroscopy at room temperature (r.t.) and cryogenic temperatures in order to elucidate the quenching mechanism(s) and kinetic reaction models. Analysis of the r.t. data by kinetic target analysis reveals that two different quenching mechanisms contribute to the protection of photosystem II (PS II). The first mechanism is a direct quenching of the PS II antenna and is related to the characteristic F740 nm fluorescence band. Based on the temperature dependence of its spectra and the kinetics, this mechanism is proposed to reflect the formation of a fluorescent (F740) chlorophyll-chlorophyll charge-transfer state. It is discussed in relation to a similar fluorescence band and quenching mechanism observed in light-induced nonphotochemical quenching in higher plants. The second and more efficient quenching process (providing more than 70% of the total PS II quenching) is shown to involve an efficient spillover (energy transfer) from PS II to PS I which can be prevented by a short glutaraldehyde treatment. Desiccation causes a thylakoid-membrane rearrangement which brings into direct contact the PS II and PS I units. The energy transferred to PS I in the spillover process is then quenched highly efficiently in PS I due to the formation of a long-lived P700+ state in the dried state in the light. As a consequence, both PS II and PS I are protected very efficiently against photodestruction. This dual quenching mechanism is supported by the low temperature kinetics data.
■
state (see18 for a review). Normal photosynthetic activity is lost upon drying. Wetting almost immediately restores the fully active photosynthetic state.18−20 In addition, to understand the mechanism of the desiccation-induced quenching, it is also of interest to clarify whether the mechanism(s) of this type of quenching and photoprotection are perhaps related in some way to the light-induced NPQ of higher plants.21,22 So far, the Chl excited state quenching mechanismsand in particular those of the drying-induced quenchingare not understood well at the molecular level.1,23−26 For poikilohydric organisms like lichens and mosses, there presently exists agreement that the drying-induced quenching is neither related to the activation of special carotenoids in the xanthophyll-cycle nor to the ΔpH-induced activation of PsbS or of a similar molecule.19,20,27−30 Rather the extremely strong drying-induced quenching has been related to the desiccation-induced appearance of a far-red emitting special Chl form in the antenna of photosystem II (PS II) (Chl-720), which has been proposed to act as an efficient quencher of PS II-associated antenna Chls.31 For investigating the molecular mechanism(s) of Chl excited state quenching, ultrafast fluorescence and transient absorption
INTRODUCTION Photosynthetic organisms in natural environments have to cope with largely varying light intensities often exceeding their maximal photosynthetic capacity. Without special mechanisms to convert this excess light harmlessly into thermal energy severe photodamage to the photosystems would occur, eventually leading to death of the whole organism.1,2 For these reasons, all photosynthetic organisms have developed a variety of photoregulation and photoprotection mechanisms that allow them to quench excess light energy mostly at the level of the antenna systems and to also efficiently scavenge already formed harmful reactive oxygen species (ROS).3,4 The most important regulation mechanism for vascular plants and microalgae is the so-called nonphotochemical quenching (NPQ), which is associated with the light- and ΔpH-induced activation of the xanthophyll cycle and the protonation of a small essential protein, PsbS. In combination, these mechanisms deactivate excited Chlorophyll (Chl) states efficiently.5−12 Under cold or water stress, photosynthetic organisms have to cope with even more severe conditions that could damage the photosynthetic apparatus. Thus additional specialized protection mechanisms are required which can compete more efficiently with charge separation in the reaction centers (RCs) to prevent their damage.13−17 Most vascular plants are not desiccation tolerant. However, lichens and many mosses are so-called poikilohydric organisms that can cope with high light intensities in a totally desiccated © 2013 American Chemical Society
Special Issue: Rienk van Grondelle Festschrift Received: March 24, 2013 Revised: June 16, 2013 Published: July 10, 2013 11326
dx.doi.org/10.1021/jp402881f | J. Phys. Chem. B 2013, 117, 11326−11336
The Journal of Physical Chemistry B
■
kinetics studies can provide the most detailed insight.10,32−38 The first ultrafast Chl fluorescence study on desiccationinduced quenching in lichens has been performed by Veerman et al. on the chlorolichen Parmelia sulcata.39 This study provided important insights into the properties and conditions of desiccation-induced quenching even though the detailed molecular mechanism could not be revealed. The most important result was the proof of the involvement of the farred emitting Chl species (termed F740 based on the more detailed data from the time-resolved studies) in the quenching. It was furthermore shown that F740 is located in the antenna of PS II. A very surprising and particularly interesting result was the finding that the F740 band was also present, at least in part, in the hydrated state at room temperature (r.t.). One might have expected the band to be absent in the hydrated state since the fluorescence emission spectra of all known in vitro as well as in vivo PS II units do not have a pronounced contribution around 740 nm35,40−44 and thus the pronounced F740 nm fluorescence of PS II in Parmelia is highly unusual. One notable exception is, however, the r.t. PS II emission spectrum in intact leaves of the npq4 (i.e., PsbS-lacking) mutant of Arabidopsis which showsunder high-light-induced quenching conditions35a strongly enhanced F740 nm band of PS II as compared to the unquenched situation. Notably, two different quenching mechanisms have been found in higher plants.35,45 Therefore, one could speculate that the strong quenchingassociated F740 band in dried lichen and the quenchingassociated F740 band in light-induced quenching of Arabidopsis are related phenomena even though some additional mechanism of quenching may exist in lichens that would be different from that in higher plants.28 In a recent time-resolved study, Komura et al. detected an ultrafast energy transfer from the peripheral antenna to the F740 band, which was assigned as a signature of the quenching species in the desiccated chlorolichen Physciella melanchla at r.t.46 In addition to this main quenching mechanism operating via the F740 band, recent picosecond fluorescence time-resolved and near-infrared absorption changes studies (800−1000 nm) were interpreted as support of a second quenching mechanism operating both in lichens and in mosses.29,30 It was suggested that the quenching is related to the formation of carotene and Chl radicals within the RCs of PS II.30 Clearly, at present there is a significant interest in understanding of the photoprotection mechanisms operating in desiccation tolerant photoautotrophs and their relation to the mechanisms operating in normal plants. Within the current work we have performed room and cryogenic temperature ultrafast time-resolved fluorescence measurements on the thalli of the poikilohydric lichen Parmelia sulcata to elucidate the desiccation-induced quenching mechanism(s) in this organism. Extensive kinetic modeling revealed the presence of two quenching mechanisms. The less efficient one is similar to antenna NPQ in vascular plants. The more efficient mechanism of the two is a novel mechanism that is based on efficient spillover (energy transfer) from PS II to PS I whose development can be prevented by mild glutaraldehyde treatment prior to desiccation. These results bear important implications for understanding of the photoprotection strategies adopted not only by lichens but also by other photosynthetic organisms.
Article
MATERIALS AND METHODS
Parmelia Lichen Material. Thalli of Parmelia sulcata lichen were collected in the hydrated state during rainy days from the bark of trees at a location near Leinach, 25 km from Würzburg, Bavaria, Germany. They were kept hydrated for about 36 h in the dark before being dried in darkness and then kept in a desiccator. These dark-dried thalli were used for the experiments. For comparison the dried thalli were hydrated by short immersion into water followed by packing the lichen into moist tissue paper for at least 6 h (hydrated lichen). The fluorescence induction response immediately after wetting was followed at different time delays using a HandyPea instrument (Hansatech, UK) (see Figure S1). All spectral changes observed upon drying and wetting are in full agreement with those reported earlier.31,39 For measuring the fluorescence kinetics with closed PS II RCs in hydrated lichen at r.t. the dry lichen was immersed for several hours into an aqueous solution containing 100 μM DCMU in the dark. During measurement of the time-resolved data at r.t., very dim (1−2 μE/m2/sec) background light was provided to completely close the RCs. Glutaraldehyde treatment was carried out by transferring hydrated Parmelia thalli for 1 h to 0.2% aqueous glutaraldehyde and then drying them at room temperature at low humidity. Glutaraldehyde was obtained from Sigma-Aldrich/Fluka, Germany. Fluorescence Experiments. The r.t. stationary fluorescence spectra and the slow fluorescence kinetics were recorded on a CCD fluorescence spectrometer (Ocean Optics) with the thalli held between two glass plates. Ultrafast fluorescence kinetics at r.t. were recorded using a previously described single-photon timing (SPT) setup.47,48 The setup consists of a synchronously pumped, cavity-dumped, mode-locked dye laser with DCM as the laser dye. The laser system generates ∼10 ps full width at half-maximum (fwhm) pulses at 800 kHz repetition frequency. The whole response of the system is about 30 ps fwhm, which after deconvolution results in a time resolution of 1−2 ps. For r.t. measurements, the dry or wet lichen thalli were placed in a rotation cuvette (10 cm diameter), which was both rotated and shifted sideways in order to keep the average intensity very low to avoid changes in the sample condition during measurements. The laser excitation occurred in front face mode. Low temperature measurements were performed using a cryostat (Oxford Instruments CF100, UK) with a small (0.5 × 0.5 cm) piece of the lichen thalli held between two glass plates. The laser light of 663 nm and ∼100− 200 μW power was focused on the sample to a spot of ∼0.8 mm diameter. The chosen excitation wavelength excited the bulk and peripheral antenna Chls about evenly, which simplifies the overall kinetics since it reduces the presence of intraantenna energy transfer processes except for the far-red components of PS I. The fluorescence decays at different wavelengths were selected by a double monochromator (spectral bandwidth 4 nm), which completely blocked the scattered laser excitation light at ca. 12 nm separation from the excitation wavelength. All stationary and time-resolved fluorescence spectra are corrected for the wavelength-dependent sensitivity of the detection system. Data Analysis. Fluorescence decays were generally analyzed first by means of global analyses as previously described in detail.49,50 Global analysis is a mathematical fitting of the decay curves at different wavelengths simultaneously in a single fitting procedure. The analysis results in lifetimes and decay-associated spectra (DAS), describing the whole set of original data.49 11327
dx.doi.org/10.1021/jp402881f | J. Phys. Chem. B 2013, 117, 11326−11336
The Journal of Physical Chemistry B
Article
Positive DAS amplitude is associated with decay of a fluorescence component and negative DAS amplitude with the rise of a fluorescence component. The presence of a positive-negative feature within a DAS of a single lifetime component is a clear indicator of energy transfer, since the fluorescence signal simultaneously decays in one wavelength range while rising in another. Kinetic modeling was performed by global target analysis49,50 on the original data sets testing various different kinetic reaction schemes (models). This analysis results in the rate constants constituting the kinetic model and the fluorescence spectra of the associated species or intermediates (species-associated spectra, SAS). The first albeit insufficientcriterion is a good fit quality for the kinetics. Further criteria are the internal consistency and whether the obtained rate constants and SAS are physically reasonable and consistent with other known data of similar processes and species, respectively. The tested kinetic models were derived based on the results of our previous ultrafast kinetics studies on isolated PS II, PS I, and light-harvesting complexes II (LHC II) samples in various forms35−37,40 and on data resulting from the analysis of the kinetics of intact algal cells and higher plant leaves (for further details see35 and references discussed therein).
■
RESULTS To investigate the photoprotection mechanisms operating in lichens we have recorded the ultrafast time-resolved fluoresce kinetics from thalii of Parmelia sulcata under room and cryogenic temperatures. Additionally, we have measured the fluorescence kinetics after mild glutaraldehyde fixation before desiccation. Glutaraldehyde is a protein cross-linking agent that may be envisaged to hamper or prevent the reorganization of the photosynthetic complexes within the thylakoid membrane. It is thus suitable to check for mechanisms that require the relative reorganization of protein complexes in the membrane. Figure 1 shows the temperature dependence of the steady state spectra for dried and hydrated samples in the range between r.t. and 4 K. The low temperature spectra are dominated by the emission bands at 720 nm whereas at r.t. a broad 740 nm band dominates. For the hydrated sample the F720 band at low temperature is almost exclusively due to the well-known PS I emission band.51 However, for the dried sample the 720 nm band at low temperatures is much broader (fwhm up to 40 nm) than for the hydrated sample (fwhm ∼25 nm), indicating that additional bands appear in that region in the quenched state. These additional emission bands seem to appear both on the short and on the long wavelength side of the 720 nm (PS I) band. In addition, the dried sample shows pronounced bands at 690 and 705 nm at the lowest temperatures (5 and 10 K), with the latter band being specific for the dried quenched state and absent in the hydrated state. These bands are very likely due to PS II, which indicates that a specific new emission that is strongly temperature-dependent accompanies the drying-induced quenching reaction. The 680− 690 nm band, which derives almost entirely from PS II, is significantly lower at the low temperatures for the desiccated sample, thus demonstrating the strong quenching of this PS II emission even at cryogenic temperatures. At intermediate temperatures (77 K) the hydrated sample shows bands at 685 and 695 nm, which are reminiscent of some characteristic low temperature bands reported for PS II.52 These bands are strongly quenched in the dried state.
Figure 1. Temperature dependence of fluorescence spectra of Parmelia hydrated (top, A); desiccated (middle, B). The r.t. fluorescence spectra for hydrated and desiccated samples are shown in the bottom figure (C). All spectra are normalized to the far-red peak.
The hydration induced reactivation of the photochemistry and the corresponding deactivation of the photoprotection mechanisms operating in lichens can be monitored by their fluorescence induction kinetics at different delays after wetting and is typical for the recovery of desiccation-induced quenching in Parmelia and similar lichen (Figure S1; Note that figures with S-numbers refer to the Supporting Information). The recovery of the photoactivity of the dried lichens occurs quickly after the hydration initiation. All original fluorescence decays for hydrated, desiccated, as well as glutaraldehyde treated samples are collected in Figure S2. Figure S3 shows the detailed fluorescence lifetime comparison based on the global lifetime analysis results for desiccated and hydrated samples at r.t. The strong quenching in the dried sample is best exemplified by the drastic dropby a factor of approximately 20of the average lifetime from ca. 1.5 ns (at 680 nm) in the hydrated sample to ca. 70 ps in the dried sample [Note that the average lifetime is essentially proportional to the fluorescence yield; since the average lifetime is an absolute value that does not depend 11328
dx.doi.org/10.1021/jp402881f | J. Phys. Chem. B 2013, 117, 11326−11336
The Journal of Physical Chemistry B
Article
lifetime components ranging from 7 ps to 3.2 ns in the hydrated sample, from 6 ps to 2.8 ns for the dried sample, and from 40 ps to 3.5 ns in the glutaraldehyde-treated dried sample. The 7 ps and the 50 ps components in the hydrated sample have dominant negative amplitude and are most likely due to energy transfer within the PS I antenna from short-wavelength emitting antenna Chls to the far-red emitting antennae Chls of PS I (negative pre-exponential amplitudes are associated with rise of fluorescence).53 However, in the dried sample, the negative amplitude 50 ps component has disappeared and, most notably, the negative amplitude of the 7 ps component has increased strongly in relation to the other DAS components (by about a factor of 2.5). The other lifetime components show either two peaks (240 ps component) or pronounced shoulders at the short wavelength side. Both of these features indicate that these lifetime components are not pure components reflecting the kinetics of a single antenna complex, but probably mixtures of lifetime components. These features are even more pronounced for the dried sample, where all the intermediate lifetime components show two pronounced peaks in their DAS. The 705 nm shoulder observed in the dried sample (also seen very clearly in the steady state spectra in Figure 1) is caused primarily by the 66 ps and the 196 ps lifetime components (Figure 2). Since both of these lifetime components also show secondary peaks around 740 nm, it is highly likely that the pronounced broadening of the steady state fluorescence band peaking at 720 nm (Figure 1) is caused primarily by these two lifetime components in the dried state. Their amplitudes either increase strongly with drying or are only present in the dried state. The most pronounced difference between the normally dried sample (Figure 2B) and the glutaraldehyde-treated sample (Figure 2C) is in the 7 ps component of very large amplitude (dried sample), which is absent in the glutaraldehyde-treated dried sample. Overall, the glutaraldehyde pretreatment strongly reduced or even abolished the desiccationinduced quenching as can be seen directly from the increased average lifetimes versus the normally desiccated state (cf. Figure 2B−C, at 680 nm). The two longest (ns) lifetimes have similar DAS and similar lifetimes in the three states. In the glutaraldehyde-treated sample, the amplitude ratio of the ns components differs from the normally dried sample. Both the lifetimes and their spectral features indicate that the nanosecond components in all three samples derive from PS I, partially from the PS I core, but mostly from a small amount of so-called “far-red” Chls in the peripheral PS I antennae.48,51,54,55 Overall, the kinetics at low temperatures in the three samples is much more complex than at r.t. This is expected since additional local energy trapping processes and their associated lifetime components are expected to appear in both PS II and in PS I upon lowering the temperature.52,56,57 At the lowest temperatures, these processes become dominant and complicate the kinetics severely versus the r.t. case. While in principle these local energy trapping and energy transfer components can still be determined well from the time-resolved data, a quantitative and detailed kinetic modeling becomes extremely difficult or even impossible at present for the low temperature kinetics. For this reason we will only present a detailed kinetic modeling for the r.t. data, while the low temperature kinetic data will be discussed on the basis of the lifetimes and DAS. Target Kinetic Analysis of r.t. Data. A global target kinetic analysis was performed on the r.t. kinetic data. In global target analysis, various quantitative kinetic reaction models are
substantially on reflection, scattering, self-absorption, etc. parameters that largely influence the measured fluorescence intensityit is a better measure of fluorescence intensity than the direct measurement of stationary fluorescence intensity itself.] This effect is caused by pronounced changes in both lifetimes as well as amplitudes of the DAS spectral components upon desiccation. Global Analysis of 77 K Data. The results of the global analysis of the 77 K data are shown as DAS49 in Figure 2. In all three samples the kinetics is highly complex, requiring at least 6
Figure 2. DAS from global analysis of 77 K fluorescence kinetics of Parmelia hydrated (top, A), desiccated (middle, B) and desiccated after a short glutaraldehyde treatment (bottom, C). The average lifetimes τavg are also shown (right-hand side scale). The τavg values are the best possible measure of the undistorted steady state intensity or yield since they are not influenced substantially by special factors like self-absorption, scattering, or reflection. Note that the DAS amplitude of the 7 ps component in the dried sample (middle) is multiplied by a factor of 0.2, i.e., the actual amplitudes are 5 times larger in relation to the other DAS amplitudes than shown in the figure. 11329
dx.doi.org/10.1021/jp402881f | J. Phys. Chem. B 2013, 117, 11326−11336
The Journal of Physical Chemistry B
Article
Figure 3. Results from global target analysis of r.t. fluorescence kinetics of Parmelia. Top: Kinetic models with rate constants (units of ns‑1). Bottom: Corresponding SAS. No spillover is required for a good fit of the data in the hydrated case (left) but is required for fitting the kinetics in the desiccated state. The model in the center shows the “normal PS II model” where quenching is described by kD and the CT state formation is contained only indirectly. The model on the right-hand side is a model where the CT state formation is included explicitly and the radical pair (RP) formation in the RC is ignored. As a consequence, kD is not increased versus the value in the hydrated state. RPx, radical pairs in the RC. CTS: ChlChl-CT state; Q: final quencher state. The wavy lines on the left side of the models indicate excitation (hν) and the corresponding numbers are the rel. excitation probabilities for the two photosystems. The red straight arrows indicate the spillover between PS II and PS I and the corresponding rates.
sample are within the characteristic range for open and normally functioning PS I particles.48,55,58,64 By contrast, the PS II emission spectrum (SAS Figure 3A) looks very unusual as compared to other “typical” PS II particles. It shows a peak at 685 nm and a strong second broad band peaking at 740 nm. The PS II particles of all other known organisms show only a 682−685 nm band and at most a very weak second band around 725 nm.41,65 We can exclude that this effect is caused primarily by self-absorption in the 680 nm emission region. While measured fluorescence intensity can be affected by selfabsorption, the average lifetime is not. The average lifetime (Figure S3) is in good agreement with the observed changes in the total fluorescence spectra as well as with the recalculated DAS of the lifetime components from the target analysis (Figure S4). Essentially, the same conclusion has already been reached by Veerman et al.39 The rate constants and lifetimes of the DAS (Figure 3 and S4) for PS II are characteristic for unquenched closed PS II particles, showing a dominant longlived lifetime component (2.14 ns) along with two loweramplitude short- and intermediate-lived components.59 The rate constant kD for deactivation of the PS II antenna is 0.35 ns−1, i.e., it is, if at all, only slightly increased from the value typical for an unquenched PS II system (kD for unquenched PS II antenna is expected to be in the range of 0.25−0.3 ns−1 for higher plants and cyanobacteria).35 Desiccated Sample. The results of the target analysis for the normally desiccated sample are shown in Figures 3B (and for a slightly different kinetic model in Figure 3C) and S4B. The first noticeable difference to the hydrated sample was that a kinetic model without spillover only yielded an extremely poor fit with very large and unacceptable deviations between the data and the predicted model kinetics (c.f. residuals comparison in
tested for their compatibility with the data. The models yield the rate constants of the involved processes (e.g., for energy transfer, quenching, charge separation etc.) as well as the SAS.49 The PS II and the PS I kinetic models that were tested were essentially the same as the ones demonstrated previously to be well suited to describe the respective kinetics on isolated photosystems.40,58−60 Those models have been used successfully to describe also the kinetics and quenching reactions in other intact photosynthetic organisms.35,36,61 The results (SAS and rate constants) are shown in Figure 3 and Figure S4. Hydrated Sample. For the hydrated sample (Figure 3A), the overall kinetics was well described as the sum of the separate kinetics of PS II and PS I. No spillover linking the two systems together was required to describe the data as is indicated by the fact that a kinetic model allowing for spillover did not improve the fitting quality (spillover describes a direct energy transfer from PS II to PS I;62,63 this requires direct contact between the two photosystems). The resulting reaction scheme is thus essentially the same as had been previously demonstrated for describing the fluorescence kinetics of dark-adapted diatoms and also for dark-adapted Arabidopsis leaves,35,36,61 except for the simpler PS I kinetics. The reduced complexity of the PS I kinetics in algae can be understood easily from the fact that green algae (the symbiotic photosynthetic organism in Parmelia) as well as diatoms do not possess pronounced farred emitting antenna pigments in contrast to higher plants and cyanobacteria.35,36,51,53,64 Accordingly, the PS I spectrum in the hydrated sample shows a main emission band at 710 nm and a small shoulder at 750 nm (SAS, Figure 3). Note also that at low temperatures the PS I emission maxima are only slightly shifted to 720−725 nm (Figures 1 and 2) in contrast to PS I of higher plants. The rate constants in the PS I part for the hydrated 11330
dx.doi.org/10.1021/jp402881f | J. Phys. Chem. B 2013, 117, 11326−11336
The Journal of Physical Chemistry B
Article
F740) in the desiccated sample. In fact, our fluorescence kinetics are very similar to the r.t. data of both Parmelia39 and Physciella.46 This agreement provides a solid basis for the search of the molecular mechanisms based on target kinetic analysis performed in the current work. Our target model at the level of the individual photosystems is in fact similar to that used by ref 46 for low temperature modeling except that the excitation in our experiments occurred directly into the Chl a Qy band at 663 nm, which makes the inclusion of a relaxation term from the Soret band to the Qy band obsolete. It had been recognized in the two previous works39,46 that the 740 nm band is derived from PS II already in the hydrated state and is further increased drastically in the desiccationinduced quenching. These findings are fully consistent with our data but have not actually been explained within a mechanistic model. It is of note that the F740 band is at variance with the PS II emission spectra of all other well-studied organisms and thus must have a special reason that needs to be explained in a rationale model of the quenching. Interestingly, the F740 band is very similar to the F740 nm band of a quenching component appearing in the high-light-induced quenching in higher plants and diatoms at r.t.,35,36 where it has been attributed to the quenched oligomeric LHC II.35 The F740 far-red emission band has been observed also in vitro in highly quenched oligomeric LHC II,37 and it has been shown that it derives from a Chl−Chl charge transfer (CT) state whose formation provides an efficient quenching mechanism.37,38 We thus tested whether a similar Chl−Chl CT state could be responsible for the strong 740 nm fluorescence PS II band in the desiccated Parmelia. At low temperature, the F740 band is shifted hypsochromically to 705−710 nm in oligomeric LHC II,37 and it has been observed in intact leaves of quenched Arabidopsis as a newly appearing low temperature fluorescence component in the same wavelength range.45 Thus the observation of pronounced short-lived emission around 710 nm in the dried lichen at low temperatures is highly reminiscent of one of the components of light-induced NPQ in plants. As pointed out above (c.f. Figure 2), two short-lived lifetime components with bands in that range are responsible for the increased fluorescence intensity (Figure 1) at the blue edge of the 720 nm band at 77 K and for the pronounced band in that range (705−710 nm) at 10 K for the desiccated sample, as compared to the hydrated sample. In the dried lichen, the F740 nm band shifts in a similar manner as in aggregated and highly quenched LHC II in vitro to shorter wavelength upon lowering the temperature (Figure 1, 710 nm, and Figure 2, 66 ps, 196 ps components, and in part also the 630 ps component). All of these observations are hard to explain in a consistent manner except if one suggests that the direct PS II quenching mechanism in desiccated lichen is also a Chl−Chl CT mechanism that leads to pronounced CT fluorescence both at r.t. (the F740 band, Figure 3B) as well as at low temperatures (the 710 nm bands as well as the ca. 740 nm bands in the spectra (Figure 1) and in the low temperature kinetics (Figure 2)). Direct proof for the assignment of the farred fluorescence to a Chl−Chl CT state in the case of the quenched LHC II complex from higher plant has been obtained recently by Stark fluorescence.67 From the amplitude (or area under the SAS) relative to the other spectra one can estimate that the oscillator strength of the CT emission is
