During desiccation, fluorescence emission and stable light-dependent charge separation in the reaction centers (RCs) of photosystem II (PSII) declined strongly in three different lichens: in Parmelia sulcata with an alga as the photobiont, in Peltigera neckeri with a cyanobacterium and in the tripartite lichen Lobaria pulmonaria. Most of the decline of fluorescence was caused by a decrease in the quantum efficiency of fluorescence emission. It indicated the activation of photoprotective thermal energy dissipation. Photochemical activity of the RCs was retained even after complete desiccation. It led to light-dependent absorption changes and found expression in reversible increases in fluorescence or in fluorescence quenching. Lowering the temperature changed the direction of fluorescence responses in P. sulcata. The observations are interpreted to show that reversible light-induced increases in fluorescence emission in desiccated lichens indicate the functionality of the RCs of PSII. Photoprotection is achieved by the drainage of light energy to dissipating centers outside the RCs before stable charge separation can take place. Reversible quenching of fluorescence by strong illumination is suggested to indicate the conversion of the RCs from energy conserving to energy dissipating units. This permits them to avoid photoinactivation. On hydration, re-conversion occurs to energy-conserving RCs.
minimum modulated chlorophyll fluorescence, QA in the RC of PSII oxidized
minimum modulated fluorescence in the absence of water
maximum modulated fluorescence, QA reduced
Fv = Fm− Fo
pulse-induced variable fluorescence
light-harvesting chlorophyll complex
non-photochemical fluorescence quenching, usually calculated as (Fm/Fm′− 1), but (Fm/Fo′− 1) after desiccation of lichens
photosynthetically active photon flux density
primary quinone acceptor in the RC of PSII
secondary quinone acceptor of the RC
In photosynthesis, light is harvested by the pigment system of the photosynthetic apparatus. Absorbed light energy migrates to specialized reaction centers (RCs), where endergonic redox reactions initiate the complex process of photosynthesis. Chlorophyll molecules incorporated into the RCs are, in contrast to the light-harvesting chlorophyll in the pigment–protein complexes of the thylakoid membranes, chemically reactive. Light-dependent charge separation in the RC of photosytem II (PSII) creates, within a few picoseconds, an oxidized chlorophyll dimer termed P680+ and a reduced pheophytin, Pheo− (Holzwarth et al. 2006). The oxidant is capable of oxidizing water and the reductant reduces CO2 with the help of photosystem I (PSI). Carbohydrates are the predominant products of photosynthesis of hydrated photosynthetic organisms.
Very few higher plants, but a large number of mosses and most lichens tolerate complete desiccation. Many of them remain green even in strong light when desiccated. According to the first law of thermodynamics, the energy of light can either support photosynthesis, or is converted into heat, or absorbed light is re-emitted as fluorescence. The high quantum efficiency of photosynthesis of hydrated plants in low light is testimony to the competitive strength of photosynthesis vs thermal energy dissipation and fluorescence emission. However, while desiccation-tolerant photosynthetic organisms dry out, not only photosynthesis declines but fluorescence also decreases. Loss of fluorescence during desiccation is caused either by decreased light absorption by photosynthetic pigments or by a reduction in the quantum yield of fluorescence emission or by a combination of both. In fact, desiccation has been shown to decrease the light absorption of lichens (Dietz et al. 2000, Gauslaa and Solhaug 2001, Heber et al. 2007). Nevertheless, light still reaching the photosynthetic apparatus will be absorbed by photosynthetic pigments, whether an organism is hydrated or desiccated. Insofar, as loss of fluorescence during desiccation cannot be traced back to reduced light absorption, it indicates a reduced quantum efficiency of fluorescence emission (Veerman et al. 2007). Owing to the competitive relationship between the different forms of light use, decreased quantum efficiency of fluorescence emission indicates increased thermal energy dissipation because light use for photosynthesis is no longer possible after desiccation (Heber et al. 2007, Kopecky et al. 2005, Soni and Strasser 2008). By dissipating light energy in the form of heat faster than it can be used for photosynthetic charge separation in the RCs or for fluorescence emission, the desiccation-induced mechanism of energy dissipation is central for photoprotection (Heber 2008, Komura et al. 2010, Veerman et al. 2007). It differs from the well-investigated zeaxanthin-dependent mechanism of energy dissipation and related mechanisms (Avenson et al. 2008, Björkman and Demmig-Adams 1994, Demmig-Adams 1990, Finazzi et al. 2004, Gilmore and Govindjee 1999, Holt et al. 2005, Li et al. 2004, Mozzo et al. 2008, Niyogi 1999, Pascal et al. 2005, Ruban et al. 2007 and others) in that desiccation, neither light nor a protonation reaction triggers the activation of energy dissipation.
The present communication extends earlier work (Heber 2008, Heber et al. 2010). We wished to know to which extent decreased light absorption can explain observed losses of fluorescence emission during desiccation of lichens. We also wished to explore the relationship between losses of fluorescence emission and decreased photochemical activity of PSII RCs. Measurements of chlorophyll fluorescence give us information on thermal energy dissipation because fluorescence is inversely related to thermal energy dissipation when photochemical light use is negligible. Magnitude and rates of energy dissipation are central for protection against photooxidative damage. Three different lichen species were selected for experimentation. All three activate thermal energy dissipation upon desiccation but their antenna composition is different. Two of them contain algae as photobionts and possess the light-harvesting chlorophyll complex (LHCII), which has been proposed to have a central function in thermal energy dissipation (Horton et al. 1996, Pascal et al. 2005, Ruban et al. 2007; but see Barros et al. 2009 for counter arguments). They can accumulate zeaxanthin in the light using the so-called xanthophyll cycle of zeaxanthin synthesis. The third species, associated with cyanobacteria, lacks LHCII and the xanthophyll cycle. It has phycobilisomes as the antenna. Chlorophyll fluorescence of all three lichens originates mainly from PSII. It is quenched by desiccation.
The data obtained in the present investigation not only suggest that decreased light absorption during desiccation of lichens contributes little to the photoprotection of light-sensitive RCs in the desiccated state, but also that a main mechanism of photoprotection, activation of fast thermal energy dissipation in the antenna outside the RCs, is supplanted, when insufficient for full protection, by energy dissipation within the RCs themselves.
Materials and methods
The foliose lichen Parmelia sulcata Ach. (family Parmeliaceae) was obtained from the bark of trees in the Botanical Garden of the University of Würzburg, Germany. The sun-exposed thalli of the foliose tripartite lichen Lobaria pulmonaria (L.) Hoffm. (Lobariaceae) were collected from the bark of a free-standing tree (Acer pseudoplatanus) at the Ramsau, Bavaria, not far from Berchtesgaden. Although this lichen is associated not only with a green alga but also with a cyanobacterium, the alga is the dominant symbiotic partner. The thalli of L. pulmonaria grown in the deep shade were from the bark of a willow shrub (Salix spec.), also from the Ramsau, Hintersee. The foliose cyanolichen Peltigera neckeri Hepp ex Müll. Arg. (Peltigeraceae) was from a shaded site under the trees in the Botanical Garden of the University of Würzburg. When not collected in the desiccated state, the thalli were slowly dried at room temperature. Desiccated thalli were sent by ordinary mail from the sites of collection to Würzburg (L. pulmonaria) or to Geneva (Parmelia, Peltigera). When storage was necessary, it was over silica gel at 5°C in a desiccator.
Modulated fluorescence was measured after excitation at about 650 nm at wavelengths above 700 nm with the pulse amplitude modulation fluorometer 101 (PAM) of Walz, Effeltrich, Germany (Schreiber et al. 1986). The measuring beam had an averaged photosynthetically active photon flux density (PPFD) of 0.04 µmol m−2 s−1 or 2 µmol m−2 s−1 when it was not important, whether the measuring beam caused some charge separation. To probe for stable light-dependent charge separation, strong or weak short pulses (1 s; PPFD = 10000or only 3µmol m−2 s−1) were given usually every 500 s. For heat protection of the thalli, the strong pulses of white light were filtered through Calflex c and DT-Cyan of Balzers, Liechtenstein. The temperature dependence of fluorescence responses of the desiccated thalli was measured using a temperature-controlled device equipped with a Peltier element. The temperature of the thalli was monitored by a thermocouple.
For the kinetic experiments shown in Figs 3 and 4, the fluorescence of the thalli was measured in a leaf clip of a HandyPEA and PocketPEA fluorimeter of Hansatech Instruments, King's Lynn, UK (Soni and Strasser 2008). The leaf clip was modified so as to facilitate the addition of water from the back side even during multiple light pulse measurements.
Absorption of light by desiccated thalli and by thalli after prolonged hydration was measured in a large custom-built Ulbricht sphere (diameter 24 cm) using light provided by a halogen lamp. The light was passed through 479, 651, 675 or 751 nm interference filters from Balzers, Liechtenstein. The filters had a half band width of transmission close to 12 nm. The foliose lichens were attached to a strip of white filter paper in the center of the sphere. The monochromatic light was guided either to the white paper (to measure zero absorption) or to the lichen (to measure absorption). The measuring beam entered the sphere through a round perspex rod (diameter 0.5 cm) coated with an aluminum foil. The area of illumination of the probe could be changed by changing the distance between the end of the rod and the probe. Visual control of the size of the illuminated area was possible through a small hole in the sphere. A sensitive photodiode was used as a detector. Only indirect light reached the detector. Zero absorption was indicated by the 100% reading of the recording instrument when the white paper was in the light path. The difference in the reading taken when the lichen was in the light path indicated percentage absorption by the lichen. The reading taken with the lichen in the light path is ascribed to the sum of light reflection and light scattering.
Glutaraldehyde was obtained from Sigma-Aldrich/ Fluka (Seelze, Germany).
Loss of fluorescence during desiccation as caused by decreased light absorption
Desiccation of lichens is known to result in strongly decreased emission of chlorophyll fluorescence (Lange et al. 1989). It is also known to decrease light transmission through the hyphal cortex of lichens above the algal layer (Dietz et al. 2000, Gauslaa and Solhaug 2001). Hydration increases fluorescence, often by factors ranging between 4 and more than 7. The amplitude of chlorophyll fluorescence is a product of light absorption and the quantum yield of fluorescence. Therefore, loss of fluorescence on the desiccation of lichens can result either from reduced light absorption or from lowered quantum efficiency of fluorescence emission, or from a combination of both. Reduced light absorption by the algae is photoprotective because it decreases the danger of photoinactivation. The same is true for a reduced quantum efficiency of fluorescence emission because the corresponding loss of fluorescence results from (and is actually indicative of) the activation of thermal energy dissipation which drains light energy from potentially photo-sensitive RCs before charge separation can take place (Heber 2008). We have compared light absorption of hydrated and desiccated thalli using the three-dimensional (3D) geometry of the Ulbricht sphere. Light not absorbed by a probe is measured as brightness of the interior of the sphere. Table 1 shows reduced light absorption by desiccated thalli of three different lichen species. Differences between hydrated and desiccated thalli were small for P. sulcata and L. pulmonaria when the thalli of the latter species were collected from a sun-exposed site. They were larger when the thalli were from a shaded site. Parmelia thalli were green, whether or not they were desiccated. Desiccated Lobaria thalli from exposed habitats were brown because a melanin pigment in the cortex shaded photosynthetic pigments of the algal layer (Gauslaa and Solhaug 2004, McEvoy et al. 2007, Solhaug et al. 2003). They were green or grayish when collected from the shade. Hydration did not much change their appearance. The thalli of the cyanolichen P. neckeri exhibited the largest differences in light absorption between the thalli before and after hydration. They were gray before and almost blue shortly after hydration.
Table 1. Absorption of 651 nm light by desiccated and hydrated lichen thalli in percentage of incident light. Measurements were taken in the Ulbricht sphere.
Species and color of the thalli
Desiccated thalli (A)
Hydrated thalli (B)
Parmelia sulcata, green
82.14 ± 3.17 (n = 12)
83.78 ± 2.30 (n = 14)
Lobaria pulmonaria, fully sun-exposed, brown
83.24 ± 2.06 (n = 12)
85.92 ± 2.45 (n = 13)
Lobaria pulmonaria, shade-adapted, gray to green
70.35 ± 3.10 (n = 17)
78.96 ± 7.10 (n = 20)
Peltigera neckeri, dry gray, hydrated, dark blue to green
76.81 ± 3.92 (n = 8)
89.95 ± 0.87 (n = 9)
Decreased light absorption in the desiccated state as shown in Table 1 for 651 nm light was also observed at 675 and 479 nm (data not shown). Increased light scattering within and escape of light from the cortex above the algal or bacterial layer are held responsible for decreased absorption of photosynthetically competent light after desiccation.
In agreement with the data of Veerman et al. (2007), absorption measurements as shown in Table 1 suggest that decreased light absorption during desiccation is not large. It does not contribute much to the very large loss of desiccation-induced fluorescence emission of lichens.
Inhibition of desiccation-induced loss of fluorescence emission by glutaraldehyde
Glutaraldehyde has been shown to inhibit desiccation-induced loss of fluorescence emission of lichens (Heber et al. 2007, Heber 2008) by reacting with proteins (Coughlan and Schreiber 1984). Figure 1A shows the fluorescence emission of a thallus of P. sulcata after hydration in 0.25% glutaraldehyde during slow desiccation. Very weak 1-s light pulses were given every 500 s to probe for stable light-dependent charge separation in PSII RCs which is shown as a pulse-dependent fluorescence increase. Four very strong light pulses given at different times during the experiment served to indicate maximum charge separation. Desiccation decreased, but did not fully inhibit the pulse-induced fluorescence responses. It failed to decrease steady-state fluorescence emission. This is in strong contrast to the control experiment of Fig. 1B with an untreated thallus, where steady-state fluorescence emission was strongly decreased during desiccation. Strong light pulses given every 500 s caused large reversible increases in fluorescence emission while the untreated thallus was hydrated. Desiccation decreased and finally completely inhibited these responses. The ratio of maximum modulated fluorescence to minimum modulated fluorescence in the absence of water (Fm/Fo′) was about 10. Non-photochemical fluorescence quenching (NPQ) was 9.35, and the Fo/Fo′ ratio was close to 5, where Fo is the minimum modulated chlorophyll fluorescence.
Pulse-induced fluorescence responses persisted after treatment with glutaraldehyde even after full desiccation (Fig. 1A). Such responses were progressively lost during desiccation in the control experiment without glutaraldehyde (Fig. 1B), where the Fv/Fm ratio (where Fv is the pulse-induced variable fluorescence) was close to 0.5 during hydration. It decreased toward zero during desiccation. In Fig. 1A, Fv/Fm was 0.26 during hydration indicating that the RCs had not been left unaffected by glutaraldehyde. After complete desiccation, it was still 0.06.
The cyanolichen Peltigera required a higher concentration of glutaraldehyde for inhibition of desiccation-induced fluorescence loss than Parmelia. Figure 2A shows the effects of 2% glutaraldehyde on fluorescence emission in comparison with an untreated thallus which served as a control (Fig. 2B). Strong light pulses spaced 500 s apart did not increase fluorescence in Fig 2A as they had done in Fig. 1A, but decreased fluorescence transiently suggesting damage to the RCs. These effects persisted during desiccation. Slow desiccation decreased minimum fluorescence. The Fo/Fo′ ratio was 1.5 after desiccation.
A larger decrease in minimum fluorescence was observed in the control experiment of Fig. 2B, where Fo/Fo′ was finally 2.3. Fv/Fm was about 0.5 during hydration and zero after desiccation. Fm/Fo′ was 4.75 and NPQ was 3.75.
The decreased Fo/Fo′ ratio in Fig. 2A is seen as a result of inhibition of desiccation-induced fluorescence quenching by glutaraldehyde. The loss of fluorescence still suffered during desiccation is then attributable either to incomplete inhibition of fluorescence quenching by glutaraldehyde, or to loss of light absorption during desiccation (see Table 1 for Peltigera), or to a combination of both.
Glutaraldehyde experiments very similar to those shown in Figs 1A and 2A were also performed in L. pulmonaria (not shown). Desiccation-induced loss of fluorescence was suppressed in this lichen already by 0.25% glutaraldehyde. The extent of suppression was comparable in sun-exposed and in shade-adapted thalli of Lobaria. Fo/Fo′ was close to 1.3 in both the cases. As in the experiment of Fig. 2A, it was not clear whether the remaining level of fluorescence after desiccation had to be attributed to incomplete inhibition of fluorescence quenching or to the loss of light absorption during desiccation (see desiccation-induced loss of light absorption of shade-adapted Lobaria in Table 1). In several untreated controls, Fo/Fo′ ratios were between 2.5 and 3.7.
The kinetics of the hydration experiments of Figs 1B and 2B which served as controls for the glutaraldehyde experiments differed in that hydration increased fluorescence in Fig. 2B initially more than in the experiment of Fig. 1B. This was a typical difference between Parmelia and Peltigera also in different hydration experiments. The larger initial increase in the Peltigera experiment was reversed after less than 1 h, long before extensive loss of water caused appreciable fluorescence quenching. Optically, Parmelia remained green on hydration. In contrast, Peltigera turned rapidly from gray to violet. This change in color was reversed to gray, while fluorescence returned from the initial peak to a lower steady-state level. For this reason, part of the rapid fluorescence rise in Peltigera could be attributed to increased light absorption in the cortex of the thallus (Table 1). The change in color from violet to gray was caused by the slow loss of water from the cortex. This explains the differences in the kinetics of the hydration curves of Parmelia and Peltigera in Figs 1B and 2B.
Changes in the quantum efficiency of fluorescence emission during desiccation and rehydration
Figure 3 shows the responses of chlorophyll fluorescence of P. sulcata to a sequence of strong light pulses. The full duration of the experiment was 200 s and the number of pulses were 20 (not all of them shown). Before addition of water to the desiccated thallus, little fluorescence was emitted (Fo′ traces in Figs 3 and 4). Its intensity did not change during the first light pulse. Only 10 s after the addition of water, fluorescence had increased, and a small further increase occurred during the light pulse (sequence of light pulses was from bottom to top). The subsequent light pulses increased fluorescence further, and the variable part of the fluorescence rise also increased.
Transients of fluorescence as seen after hydration of the thallus have been analyzed before and are known as OJIP kinetics (Strasser et al. 1995, Strasser et al. 2004, Tsimilli-Michael and Strasser 2001). They indicate electron transport within the RCs of PSII. The time course of variable fluorescence reflects changes in the redox state of the couple QA−/QA, where QA is the primary quinone acceptor in the RC of PSII. Different waves of the fluorescence transient are indirectly linked to the redox state of all the electron carriers of the electron transport chain from the water splitting side to, finally, the reduction of NADP. At the beginning of the first wave in Fig. 3, the primary quinone acceptor QA of the RC is oxidized. As the wave progresses, QA becomes partially reduced. A second wave, when present, shows additional reduction of QA because of the following reduction of the plastoquinone pool. This wave is barely visible in Fig. 3. Full absence of variable fluorescence in the desiccated thallus indicates the inactivity of the RC.
A very similar experiment was performed with Peltigera. The kinetics of the RC reactivation after hydration of a desiccated thallus was similar in principle to the fluorescence kinetics shown for Parmelia in Fig. 3, but fluorescence increased much faster than restoration of charge separation in PSII RCs. This is also seen in Fig. 2B. After the awakening of fluorescence waves, the first wave indicated partial reduction of QA. The second wave denoting QB (secondary quinone acceptor of the RC) reduction and electron flow to plastoquinone was, as in Parmelia (Fig. 3), almost absent in short-time experiments (data now shown).
In the experiment of Fig. 4, slow desiccation is shown for hydrated Parmelia (starting from top toward bottom). Light pulses of 300 ms were given every 15 min during drying. Initially, two large waves are seen. The first one is indicative of partial QA reduction as in Fig. 3, the second one of further reduction of QA because of the accumulation of reduced QB. A third wave, when present, would have shown full reduction of QA because of the complete reduction of all electron carriers of the electron transport chain. It is absent in Fig. 4 because the hydration time had been brief. The waves are affected by the loss of water. As desiccation neared completion, fluorescence emission approached the lowest level as shown in Fig. 4. Absence of kinetics shows inactivity of the RCs.
When a thallus of Peltigera was slowly desiccated in an experiment similar to that shown for Parmelia in Fig. 4, the two main waves of rising fluorescence which were produced during the first short light pulses decreased later on as shown already for Parmelia in Fig. 4. The absence of kinetics in response to the light pulses indicated inactivation of the RCs. There were no principal differences between Parmelia and Peltigera in their response to slow desiccation (Figs 1B and 2B).
Lobaria pulmonaria is another foliose lichen which in contrast to many lichens is capable of adapting to both deep shade and full exposure to sunshine (Gauslaa and Solhaug 1999). Figure 5 shows a recording of modulated fluorescence of a sun-adapted thallus of L. pulmonaria. Before addition of water to a desiccated thallus, fluorescence emission was very low and strong light pulses had no apparent effect on fluorescence. After addition of water, fluorescence rose rapidly. Strong 1-s light pulses given 500 s apart produced transient fluorescence spikes Fm which became larger with increasing time of hydration. They continued to increase even after the rising fluorescence Fo level had reached a plateau. This indicated that the reactivation of stable charge separation in PSII RCs was slower than the transition of the dark-adapted fluorescence response Fo′ to Fo which is initiated by hydration. The Fo/Fo′ ratio in Fig. 5 was 7. NPQ was about 15 after desiccation.
However, recovery of stable charge separation as indicated by pulse-induced fluorescence spikes did not always lag behind the recovery of Fo as shown in Fig. 5. Particularly in experiments with desiccated thalli of P. sulcata (Fig. 1B) and also with other chlorolichens such as Hypogymnia physodes, addition of water usually increased Fo and Fm in a very similar fashion. Occasionally, Fm even appeared to recover faster than Fo. Such differences in observation possess considerable significance. Figure 6 relates the maximum quantum yield of primary photochemistry in PSII RCs as measured by Fv/Fm (Paillotin 1976, Genty et al. 1989) to the ratio Fo/Fo′, while the thalli of Parmelia, Lobaria and Peltigera were slowly desiccated and subsequently rapidly hydrated. The fluorescence level Fo of hydrated photosynthetic organisms denotes a situation in which all PSII RCs are open and QA is oxidized. Fo′ is the final level of fluorescence emitted after complete desiccation. The ratio Fo/Fo′ therefore decreases from higher values toward unity. At Fo/Fo′ = 1, stable charge separation in PSII RCs as indicated by the ratios of variable to maximal fluorescence was no longer observed. Hydration, by increasing the Fo fluorescence, not only increased the Fo/Fo′ ratio, but also activated PSII photochemistry.
In the Parmelia experiment of Fig 6A, the maximum Fo/Fo′ ratio was about 7 indicating that much fluorescence was lost during desiccation. As the lichen was slowly desiccated, the loss of RC activity in relation to decreasing Fo/Fo′ ratios was biphasic. Hydration restored RC activity even faster than fluorescence. Comparable restoration of RC activity and of Fo suggests that RCs had remained functional after desiccation.
However, the relationship between energy dissipation and loss of activity of PSII RCs is not generally as simple as it appears in the Parmelia experiment of Fig. 6A. Figure 6B shows the relationship between the photochemistry of PSII RCs and the Fo/Fo′ ratio during slow desiccation and subsequent hydration of a thallus of sun-adapted Lobaria. As in the Parmelia experiment of Fig. 6A, slow desiccation decreased both photochemistry and the fluorescence ratio in a biphasic fashion. But in contrast to the Parmelia experiment, rehydration revealed hysteresis for the restoration of photochemistry. Reactivation was clearly slower than the reversal of desiccation-induced fluorescence quenching. Observations very similar to those shown for sun-adapted Lobaria in Fig. 6B were also made with Lobaria which had been grown in deep shade and was much more sensitive to photoinactivation than sun-adapted Lobaria (Gauslaa and Solhaug 1999). The Fo/Fo′ ratios were lower in shade-adapted Lobaria than in the thalli of sun-expsoded Lobaria.
Figure 6C summarizes the relationship between photochemistry in PSII RCs and loss of fluorescence for a Peltigera experiment. Here, the hysteresis effect shown already in Fig. 6B is strongly expressed. Only part of this effect is caused by increased light absorption on hydration of Peltigera as shown in Table 1.
Photochemical responses of desiccated thalli as revealed in light-dependent fluorescence changes
Even though stable charge separation in PSII RCs, which decreases together with fluorescence during desiccation, appears, at first sight, to be absent in dry lichens (Figs 1B, 2B and 5), sensitive measurements reveal reversible photochemistry in desiccated lichens (Heber 2008, Heber et al. 2010). The quantum efficiency of these reactions was very low. Little reactivity was seen at light intensities up to the photon fluxes of sunlight. It increased essentially linearly with photon flux. Figure 7 shows the responses of chlorophyll fluorescence of desiccated lichens to very strong illumination. In the case of the chlorolichen P. sulcata, which had lost much fluorescence during dessication (Fig. 1B), residual fluorescence increased in a way reminiscent of the known increase caused by the reduction of QA in PSII RCs (Fig. 7A). In the chlorolichen L. pulmonaria, fluorescence responses depended on whether the thalli were collected from a sun-exposed habitat or in the deep shade. In the sun-exposed thalli which are less sensitive to photooxidative damage than shade-adapted thalli, illumination first quenched and then increased fluorescence (Fig. 7B). Darkening reversed the initial quenching and then the secondary rise in fluorescence. In the shade-adapted Lobaria thalli, the secondary rise in fluorescence was absent. Only reversible quenching was observed on illumination (data not shown). In the cyanolichen Peltigera, illumination caused only a reversible quenching response (Fig. 7C). The reversible fluorescence responses of desiccated lichens shown in Fig. 7 were accompanied by a reversible increase in light absorption around 800 nm (data not shown, but see Heber et al. 2010).
The Parmelia experiment of Fig. 7A was subjected to a temperature analysis in Fig. 8. Light-dependent fluorescence changes are shown as a function of temperature for two desiccated thalli. Both were from the same location on the bark of a tree. In both the cases, fluorescence increased reversibly under strong illumination as the temperature was increased from room temperature to high but sub-lethal temperatues. This effect has been observed before (Heber and Shuvalov 2005, Heber et al. 2010). As the temperature was lowered, fluorescence no longer increased under strong illumination. Rather, it was quenched under strong illumination. The crossover point where fluorescence changed direction was different in different experiments but was always above 0°C. When, after the temperature had been lowered to −30°C, it was raised again to +40°C, fluorescence increased once again under strong illumination. Reversibility of the temperature effect was observed. It is important to note that the opposite fluorescence responses observed above room temperature and at subzero temperatures reflect different reactions. Kinetics of these reactions are shown as insets in Fig. 8. Light-induced quenching observed at about −20°C relaxed faster than the light-induced increase in fluorescence above +40°C. This is particularly remarkable as thermal reactions are slowed down as temperatures are decreased. It appears that the relaxation in the dark of the opposite fluorescence responses shown in the insets of Fig. 8 reflects recombination reactions after different light-dependent charge transfer reactions.
Fluorescence emissions as shown in Fig. 8 are accompanied by reversible absorption changes at 800 nm (Heber et al. 2010). The latter may be interpreted as revealing the light-dependent formation of cation radicals of chlorophyll. The decay kinetics of fluorescence quenching as shown in the inset of Fig. 8 and of the light-dependent increase in 800-nm absorption were very similar. Still, attempts to identify the quencher as a chlorophyll radical were not successful because the temperature dependence of fluorescence quenching as shown in Fig. 8 and of 800 nm absorption changes was different (data not shown).
In contrast to the Parmelia experiment of Fig. 7A where strong illumination increased fluorescence emission, fluorescence quenching was observed in the Peltigera experiment of Fig. 7C. A temperature analysis of this effect revealed that the reversible loss of fluorescence decreased essentially linearly with increasing temperature within a temperature range between −25°C and +50°C. The rate of relaxation of quenching upon darkening was not much dependent on temperature (data not shown).
Susceptibility of RCs to photodamage
By virtue of their chemical reactivity, chlorophylls of the RC of PSII differ from other chlorophylls. Within the pigment bed of the photosynthetic apparatus, absorbed light energy migrates to and is trapped by the RCs where charge separation converts it into chemical energy when water is available. When it is absent, charge separation threatens to damage the photosynthetic apparatus by producing a strong oxidant which can cause unspecific oxidation reactions. Recombination of the oxidant with the reductant produced during charge separation can lead, by spin reversal, to a long-living triplet chlorophyll which in a reaction with oxygen produces highly oxidative singlet oxygen (Asada 2006, Krieger-Liszkay 2005, Krieger-Liszkay et al. 2008). In this situation, protection of RCs becomes essential for survival of desiccation-tolerant photoautotrophs when water is lost under strong illumination.
Photoprotection was activated during slow drying of hydrated thalli. The activity of PSII RCs as shown by pulse-induced fluorescence responses decreased progressively until it was no longer noticeable in Figs 1B and 2B. Simultaneously, Fo fluorescence decreased. Glutaraldehyde prevented the loss of fluorescence and the full inactivation of RCs in the Parmelia experiment of Fig. 1A. In this situation, the RCs were still capable of reversible photochemical activity even after complete desiccation (Fig. 7A). This made them sensitive to photoinactivation as shown by the loss of photochemical activity after rehydration (data not shown, but see Heber et al. 2007).
Loss of light absorption owing to desiccation
When, in the absence of water, photochemistry is largely silent, only fluorescence and thermal energy dissipation compete for absorbed light energy. Therefore, lowered fluorescence indicates increased energy dissipation. Alternatively, loss of fluorescence can also be a simple consequence of decreased light absorption because the amplitude of fluorescence is the product of light absorption and the quantum efficiency of fluorescence emission. Decreased light absorption as a result of desiccation has been reported by different authors. Cortical transmission measured by fiber optics ranged between 45 and 88% in nine different lichen species (Dietz et al. 2000). Desiccation has been reported to decrease the transmission of light through the cortex of Lobaria by factors of two to three (Gauslaa and Solhaug 2001). Absorption spectra measured in the reflectance mode by Heber et al. (2007) also suggested desiccation-induced reduction of light absorption of more than 50% for Peltigera and 20% for Parmelia. Such data are in contrast to small effects of desiccation on absorption as recorded in Table 1. We suggest that earlier observations require re-interpretation because unidirectional measurements had been used which cannot take the 3D complexity of light scattering within the lichen thalli into full account (Butler 1962). Using the 3D geometry of the Ulbricht sphere, we obtain full information on scattered light escaping from the thalli. In addition, we have complemented the use of the Ulbricht sphere (Table 1) by the glutaraldehyde experiments of Figs 1 and 2 which confirmed that the earlier spectra in Heber et al. (2007) had overestimated the loss of light absorption on desiccation. Similar conclusions apply to the data of Dietz et al. (2000) and Gauslaa and Solhaug (2001). A comparison of the B/A ratios in Table 1 (between 1 and 1.2) with the maximum Fo/Fo′ ratios obtained by desiccation in Fig. 6 (between 4 and 7) shows that the loss of light absorption contributes only little to the loss of fluorescence caused by desiccation. This agrees with conclusions drawn by Veerman et al. (2007) for the lichen P. sulcata.
Significance of loss of quantum efficiency of fluorescence emission during desiccation
After light that has been absorbed by the pigment system of the photosynthetic apparatus, its energy must be dissipated thermally, if it cannot be used for photochemistry or give rise to the emission of fluorescence. This is a simple consequence of the first law of thermodynamics. Under very low light intensities, hydrated photosynthetic organisms are known to use almost 100% of absorbed light energy for photosynthesis. Very little fluorescence is emitted. Its low level is termed Fo. Energy dissipation is inactive in very low light. Therefore, energy conservation in photosynthesis and fluorescence emission are in, or close to, equilibrium. Energy capture by open RCs leads to charge separation within a few picoseconds (Holzwarth et al. 2006). It opens the path to photosynthetic carbon reduction.
During slow desiccation of photoautotrophs such as lichens, stable charge separation in the RCs is progressively inhibited. This is shown by decreasing pulse-induced fluorescence spikes in Figs 1B and 2B during drying. Fo fluorescence decreases together with the slow loss of photochemical activity of PSII RCs. After desiccation, it is at a level far below the initial level attained during full hydration. Desiccation has suppressed not only photochemistry but has also decreased fluorescence emission far below the Fo level. Owing to the competitive relationship between energy dissipation, fluorescence and photochemical use of light energy, loss of both photochemical activity and of fluorescence reveals the activation of energy dissipation. The strong loss of Fo fluorescence during desiccation actually shows that the energy dissipation has become faster than charge separation in the RCs (Heber 2008). This protects the RCs because charge transfer in the RCs and ensuing charge recombination between an oxidized chlorophyll and a reduced pheophytin initiate damaging oxidative photochemistry (Krieger-Liszkay et al. 2008).
Activation of thermal energy dissipation during desiccation is a regulated process. This is shown by the effects of glutaraldehyde in the experiments of Figs 1A and 2A. Glutaraldehyde prevented or decreased the loss of fluorescence. It is capable of reacting with proteins. Apparently, a protein is involved in the mechanism of desiccation-induced energy dissipation. Conformational changes of a pigment protein caused by desiccation are thought to activate photoprotective energy dissipation (Heber 2008).
Mechanisms of desiccation-induced energy dissipation
Quenching of the main emission of PSII fluorescence around 685 nm during desiccation of lichens and desiccation-tolerant mosses has been shown to be accompanied by far-red emission. Part of it was attributed to PSII (Heber and Shuvalov 2005). This suggested the involvement of a long-wavelength emitter in the mechanism of desiccation-induced fluorescence quenching. Measurements of fluorescence life times revealed a dramatic shortening of the excitation life time of the far-red emitter (Komura et al. 2010, Veerman et al. 2007). De-excitation of the bulk pool of excited chlorophyll was much faster after desiccation than in the hydrated state of lichens. This drains excitation energy from the RCs protecting them against photoinactivation and permitting them to remain functional even after desiccation. Spillover of energy from PSII to PSI and dissipation within PSI does not contribute much to the desiccation-induced energy dissipation (Bilger et al. 1989, Veerman et al. 2007).
However, several observations suggest that fast migration of excitons from PSII to dissipation centers which are located in a far-red emitting pigment protein complex is not the only pathway of desiccation-induced photoprotection:
1In the kinetic hydration experiment of Fig. 3 with Parmelia, only QA was reduced shortly after hydration as shown by a fast uniphasic rise in fluorescence. The secondary rise which is indicative of QB reduction was absent. Similar observations were made for Peltigera (not shown). QB reduction was seen only after more prolonged hydration (Fig. 4). Apparently, hydration does not restore electron flow from QA to QB as readily as it activates charge separation and reduction of QA. In thermoluminescence experiments, only the Q-band of light emission was observed shortly after hydration of a desiccated lichen thallus, not the B-band (A. Krieger-Liszkay, personal communication). This band appeared only slowly, minutes after hydration of desiccated cyanobacteria (Harel et al. 2004). Light emission from the Q-band is attributed to S2QA− recombination and emission from the B-band to S2,3QB− recombination. According to Krieger-Liszkay et al. (2008), a desiccation-induced shift in the redox potential of QA to positive values could be responsible for the inhibition of electron flow from QA to QB. It could be photoprotective by preventing the re-population of the state P680+Phe− in PSII RCs. Recombination of P680+Phe− is known to give rise to triplet formation and production of highly damaging singlet oxygen.
2Hydration of desiccated Lobaria and Peltigera restored fluorescence emission faster than stable charge separation (Fig. 6B, C). Evidently, desiccation-induced energy dissipation had been rapidly reversed by hydration, with charge separation in PSII RCs lagging behind. In Parmelia, hydration restored fluorescence emission and charge separation with similar kinetics (Fig. 6A).
3Differences in the state of PSII RCs of desiccated thalli from different lichen species are suggested by the experiments of Fig. 7. Strong illumination increased residual fluorescence at room temperature reversibly in desiccated Parmelia (Fig. 7A). This is indicative of QA reduction in essentially functional RCs. In contrast, readily reversible quenching of fluorescence was observed in the desiccated thalli of Lobaria and Peltigera (Fig. 7B, C).
4The temperature dependence of fluorescence emission from the desiccated thalli of P. sulcata revealed quencher formation at low temperatures. At higher temperatures, increased fluorescence suggested stable charge separation in PSII RCs followed by the reduction of QA (Fig. 8). Relaxation kinetics of the opposing fluorescence responses confirmed that different light-induced reactions had caused the opposite fluorescence responses. Thermal energy dissipation is expected to be decreased at low temperatures. This decreases photoprotection. Excitation energy may now reach the RCs and result in photochemical reactions. The crossover temperature between increased fluorescence emission and fluorescence quenching was variable in experiments with Parmelia thalli of different origin, but was never far from room temperature. Fluorescence emission from PSII responds to energy dissipation not only in dissipation centers outside the RCs but also within the RCs. As only the chlorophylls of the RCs are chemically reactive, both the two contrasting fluorescence responses documented in Fig. 8 are likely to originate from the PSII RCs. The differences then suggest different electron tranfer routes in the RCs which contain six chlorophylls and two β-carotenes.
In a side path reaction, β-carotene can be oxidized in the light. The quantum efficiency of this reaction is very low (Shuvalov and Heber 2003). Oxidized β-carotene can be reduced by a neighboring chlorophyll, ChlZD2, which is a radical quencher when oxidized (Faller et al. 2006). The light-dependent formation of chlorophyll radicals is accompanied by increased absorption around 800 nm. Reversible light-dependent absorption changes in this region were indeed observed but failed to have the temperature profile of the fluorescence quenching shown in Fig. 8. Possibly, the quencher was not revealed because different RC chlorophylls including chlorophylls of PSI RCs are oxidized by strong light.
An alternative or even additional possibility to explain reversible light-dependent fluorescence quenching of desiccated lichens is charge recombination in PSII RCs along a route different from the toxic reversal of the primary charge separation which leads to long-lived triplet chlorophyll and opens the path to photooxidative damage. A non-toxic recombination reaction would be the recombination between oxidized chlorophyll P680+ and QA− as proposed by Krieger-Liszkay et al. (2008). If very fast, it would lead to fluorescence quenching. Although it is not easy to see how reducing the temperature as seen for fluorescence quenching in Fig. 8 could accelerate a P680+QA− recombination reaction, Ohad et al. (2010) have observed faster re-oxidation of QA− in DCMU-poisoned PSII RCs of desiccation-tolerant cyanobacteria when they reduced the temperature. As this organism was also remarkably irradiation-tolerant even when hydrated, they considered this observation in combination with thermoluminescence data as strong evidence for photoprotective non-radiative charge recombination in PSII RCs.
Both the variability of fluorescence responses of desiccated thalli of different lichen species documented in Fig. 7 and the temperature dependence of fluorescence responses of desiccated Parmelia thalli shown in Fig. 8 strongly suggest interconvertibility of PSII RCs. By definition, quenching within PSII RCs is photoprotective because it tends to prevent potentially harmful photochemistry (Ivanov et al. 2008). Hydration re-converts energy-dissipating RCs into energy-conserving RCs. This can explain the delay in the restoration of RC activity observed in Lobaria (Fig. 6B) and Peltigera (Fig. 6C). If energy trapping in dissipation centers formed during desiccation outside the RCs in a far-red emitting pigment protein (Komura et al. 2010, Veerman et al. 2007) is not sufficient to divert excitation energy from the functional RCs, primary charge separation in PSII RCs is suggested to be followed by energy dissipation inside the RC complex. Hydration inactivates both dissipating centers outside the RCs and photoprotective pathways within the RCs thereby restoring fluorescence, charge separation and electron flow in the light.
It thus appears that in desiccation-tolerant photoautotrophs, more than one mechanism of energy dissipation is active in the photoprotection of desiccated thalli. Fast migration of excitation energy along the energy gradient toward highly effective dissipation traps characterized by fluorescence emission in the near far-red protects functional RCs of PSII from photoinactivation. When this is insufficient for full protection, an auxiliary mechanism becomes active within PSII RCs.
We are grateful to Prof. Roman Türk, Department of Organismic Biology, University of Salzburg, for collecting Lobaria pulmonaria in an alpine environment and sending us desiccated thalli. Prof. R. Hedrich, Julius-von-Sachs-Institute, University of Würzburg, provided laboratory space and facilities to one of us (U. H.). Mario Volk built the device used for measuring the temperature dependence of fluorescence emission from desiccated thalli.