UvA-DARE ( Digital Academic Repository ) In vivo manipulation of the xanthophyll cycle and the role of zeaxanthin in the protection against photodamage in the green alga Chlorella pyrenoidosa

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

protection mechanism against photodamage was first suggested by Krinsky (1971). Chl triplet states can relax via triplet energy transfer to zeaxanthin followed by dissipation of the excited triplet via the trans-cis isomerization of zeaxanthin, the latter reaction is exothermic. More recently, another energy dissipating process (non-photochemical quenching, qNP) connected with the xanthophyll cycle was introduced by Demmig- Adams (1990). Contrary to the relaxation of Chl triplet states in the former process the qNP has been proposed to be a singletsinglet exchange between chlorophyll and carotenoids (Owens et al., 1992).
The scheme of reactions that take place in the light involves two de-epoxidation steps through which violaxanthin via the intermediate antheraxanthin becomes zeaxanthin (Hager and Stransky, 1970). This way the latter compound accumulates in the light. In darkness the reactions are reversed to violaxanthin. All reaction steps have been well characterized, except for the epoxidizing step from zeaxanthin to antheraxanthin in which a "mixed-function oxygenase" (Hager, 1981) was suggested to be involved. The different pH ranges at which the respective enzymes operate give rise to a scheme in which the steady-state concentrations of the components of the xanthophyll cycle are determined by the pH of the lumen. The deepoxidation reactions yielding the final product zeaxanthin rely on enzymes that become activated at a thylakoid lumen pH of 5.2 and thus operate in the light. The backreactions involve enzymatically catalyzed epoxidation steps that rely on a higher pH of the thylakoid lumen and by consequence operate in darkness (Pfundel and Dilley, 1993;Gilmore and Yamamoto, 1993). Thus, according to these observations, pH transitions between light and dark effect the differences in the presence of violaxanthin and zeaxanthin relative to one another.
Dithiothreitol has been applied as a successful inhibitor of the violaxanthin de-epoxidation steps (Yamamoto and Kamite, 1972). However, additional effects of dithiothreitol under in vivo conditions on several other thioredoxin-regulated reactions, such as carbon metabolism enzymes (Rowell et al., 1986) or the ATP synthase (Mills, 1986) may obscure the answer to the question whether in addition to the decreased availability of zeaxanthin other inhibitory effects of dithiothreitol are responsible for the observed increased sensitivity to photodamage in the presence of dithiothreitol. In addition, the use of a n inhibitor in the study of a cyclic process, excludes the possibility to retrieve information about the dynamic properties of such a cycle.
The aim of the present study was to evaluate the photoprotective potential of the xanthophyll cycle with different steadystate contents of violaxanthin and zeaxanthin generated in vivo without disturbance of the cellular metabolism by external additions other than light. The data presented indicate that epoxidation of zeaxanthin also proceeds in the light as a result of the photoprotective (excited oxygen quencher activity) proc-

In Vivo Manipulation
of the Xanthophyll Cycle essing of zeaxanthin. This observation reveals that the dynamic function of the xanthophyll cycle in vivo is larger than would be predictable from existing data. Our approach to assess photodamage in a constant background of photoprotection, established by introducing continuous background illumination, may be useful in other areas of photosynthesis research.

MATERIALS AND METHODS
Cu1ture"Two types of steady-state continuous cultures of Chlorella pyrenoidosa were used, both were grown in 2-liter chemostats in BG-11 medium (Rippka et aZ., 1979) at 20 "C. One was grown at 30 pE.m-2.s-1 (low light, LL) the other one at 240 pE.m-2.s-1 (high light, HL). Circular fluorescent tubes (Philips TLE 32W/33) were used for continuous illumination. The set up of the culture system was as in Van Liere and Mur (1978). Aeration at 60 litedmin provided adequate mixing and COP supply. The cultures were maintained at an A,,, of 0.18-0.20.
Preaclaptation and Hash Experiments-Samples from the HL and LL cultures were preadapted during 30 min at 20 "C in either darkness or in the presence of actinic (background) light. The actinic light intensities for the LL and HL samples were 430 and 600 pE.m-2.s-1, respectively. These light conditions were arrived at to be saturating from the photosynthesis uersus irradiance (Pn) curves (see Fig. 1). Preadaptation proceeded directly in the 12-ml oxygen electrode measuring chamber. This device has been described elsewhere (Dubinsky et al., 1987). The samples were bubbled with air to ensure a constant partial oxygen pressure. Next, while maintaining the conditions of preadaptation (i.e. background light or darkness), one group of samples was exposed to one thousand supersaturating flashes (see below) in order to incite photodamage. The other group was not exposed to flashes and remained in the preincubation conditions during this time.
A delay between the flashes of 3 s was chosen in the samples without background light. At this frequency, controls demonstrated that the oxygen consumption rate (dark respiration) remained identical, i.e. no oxygen production was revealed, with or without flashes. The light preadapted samples (which receive the flashes in the continued presence of saturating background light), already perform photosynthesis at a maximal rate. This allowed a faster flashing regime with 300-ms intervals. A General Electric FT 230 flash tube was used at a discharge voltage of 1.3 kV, which provided flashes of 5 ps half-width with an energy output of 2 J/flash in the forward direction. Calculated by the surface of the incubation chamber this amounts to the supersaturating photonflux of approximately 10,000 pE.m-Vflash. The flash tube was connected directly to the incubation chamber (Le. the one used for the oxygen and fluorescence measurements, cf. below). During the flashes aeration was continued. The number of flashes was selected to yield appreciable photodamage (as judged from changes in the pigment content and physiological activity presented), while avoiding lethality. All flash-treated sampies used for the photosynthesis activity assays were allowed recovery during 15 min in darkness to equalize the metabolic conditions of the various samples. Although the pigment composition changes during this 15-min period (especially zea-and antheraxanthin disappear, whereas violaxanthin increases, data not shown), the overall losses of pigment are not replenished in this short period. Samples for pigment analysis were taken immediately after the incubation period (with or without flashes) but before the relaxation time introduced in the other assays.
PII Curves and Fluorescence Measurements-After the preadaptation, flash and recovery periods were terminated, the cuvette was closed, and PII curves were recorded according to standard procedures (Dubinsky et al., 1987). Fluorescence measurements included two types of experiments. A, photochemical quenching was monitored with a pulse-amplitude-modulated chlorophyll fluorescence measuring system (Walz, Germany) as described by Schreiber et at. (1986). During the measurements the fiber optic light guide was directly placed against one side of the oxygen measuring chamber. This way, oxygen production and photochemical quenching (qP) could be estimated simultaneously. qP was estimated every 120 s by firing saturating pulses of 500-ms duration (Schott KL-1500 light source, 12,000 pE.m-2.s-1).
E, relative energy transfer efficiency from carotenoids to Chl was measured by comparing the fluorescence yield after excitation with broad blue and orange light. The former excites both carotenoids and chlorophyll, the latter chlorophyll only. Data were normalized on the emission resulting from the chlorophyll excitation in the orange. These measurements were done with a Perkin-Elmer 1000 spectrofluorometer, emission wavelength 685 nm, slit width "M." Before the measurements the samples were preadapted to light or dark conditions as described above. Samples were treated with 3-(3,4-dichlorophenyl)-l,l-dimethylurea (10 p~) during 1 min, either in the dark or light analogous the preadaptation conditions, immediately before the assay. Excitation was done with orange light through a 628-nm interference filter (Schott) and with blue light through a 2-mm BG28 cut-off filter (Schott).
Data shown are the average of three separate experiments. Differences between comparable data points in the three experiments were below 10% of each of the numeric values given.
PS I and PS II Actiuity Measurements-Cells were harvested and resuspended in a buffer containing: 0.33 M sorbitol, 2 m~ EDTA, 1 m~ Mg&, 1 m~ MnCI,, 50 m~ Hepes-KOH, pH 7.6. Cells were broken by one passage through a French press (Aminco) at 8,000 megapascals. Oxygen uptake was measured with a Clark oxygen electrode (Yellow Spring Instruments) in a thermostated 1-ml laboratory built oxygen electrode cuvette. Saturating white light was supplied with a Schott 1500 light source, equipped with a light guide. Full chain (PS I1 and PS I) electron transfer capacity was estimated in the presence of 100 p~ methylviologen and 1 m~ sodium azide, following the recommendations ofAllen and Holmes (1986a). PS I electron transfer was measured with 2 m~ ascorbate and 50 p~ dichlorophenolindophenol as electron donor system and 100 p~ methylviologen as electron acceptor. The PS I assay was done in the presence of 3-(3,4-dichlorophenyl)-1,l-dimethylurea (10 p~) and 1 m~ sodium azide (Allen and Holmes, 1986b).
Pigment Analysis and Zeaxanthin Conversion-HPLC analysis was done as in Mantoura and Llewellyn (1983), as specified in Van der Staay et al. (1992). Detection wavelengths were chosen at 440 and 480 nm. This way, zeaxanthin and lutein contents (which have very similar retention times in the HPLC procedure used) could be estimated separately. The specific extinction coefficients were taken from Mantoura and Llewellyn (1983). At the end of the flash periods samples were immediately removed from the reaction chamber and processed for pigment content estimation. This involved immediate centrifugation during precisely 1 min and mixing of the pellet with ice-cold acetone. Before the actual HPLC analysis, the samples were stored at -18 "C until used. The zeaxanthin samples used in the in vitro degradation studies (cf, Fig. 2) were from Chlorella and were purified by HPLC.
A, , and Chl Estimation-A,,o was measured on a Pharmacia Novaspec I1 photometer. Chlorophyll was measured in acetone extracts (Jeffrey and Humphrey 1975).

RESULTS
Changes in photosynthetic activity (02 production) and photochemical quenching (qP) after exposure of C. pyrenoidosa cells to control or photodamaging conditions are shown in Fig.  1. Control samples of LL and HL Chlorella cells behave differently. The LL cells have a lower maximal photosynthesis activityXh1 than the HL ones. The LL cells show a stronger qP decrease than the HL cells. The rate of O2 evolution decreases at higher irradiance of the LL cells. Using preadaptation in the light or darkness did not induce major changes, both HL and LL cells retain comparable activities. Following exposure to the photodamaging flash treatment in the continued presence of actinic background light gave rise to relatively minor losses of activity through photodamage, both in the LL and HL cells. In contrast, clear photodamage is obvious in the samples that were kept in darkness during preadaptation and while being exposed to the photodamaging flashes. Especially the LL cells show a n appreciable loss of oxygen evolution and qP at increasing actinic light intensities over the course of the P/I curve determination.
The observed differences of the photosynthetic activities were related to changes in the pigment composition of the samples. Table I depicts the pigment analysis of the HL and LL cultured cells. The data reflect that in the LL cells the total Chl to carotenoid ratio is at least twice that of the HL cells, the Chl to summed xanthophyll cycle components ratio is 3-fold higher. Differences produced by the dark or light preadaptation conditions are mainly restricted to the three xanthophyll cycle pigments. The violaxanthin content decreases in the light and the zeaxanthin content increases. This way, variable pool sizes of the xanthophyll cycle components were established before exposure to potentially photodamaging conditions.
The    Table I, the overall picture depicts photodamage of most pigments, including Chl a, lutein, and p-carotene, with the marked exception of the antheraxanthin content in the light preadapted samples. The neoxanthin content decreases in the LL cells only. In general, the damage is small in the light-preadapted HL cells and somewhat more pronounced in the dark-preadapted HL cells. Noticeable damage is induced in the dark preadapted cells of the LL culture. As opposed to the HL grown cells in which the total of xanthophyll cycle components becomes reduced by 17% in the dark-flashed group, the loss in the analogous LL experiment amounts to 66% (Table 11).
Zeaxanthin is the predominantly disappearing compound in the dark-flashed HL cells with reference to the just dark-incubated HL control cells. In the absence of zeaxanthin, p-carotene is a target for breakdown, as can be seen most clearly in a comparison of the LL dark-adapted and dark-flashed samples. Lutein appears to be relatively little involved in the protection. To define the site where the actual photodamaging process occurs and especially to locate the site a t which the xanthophyll cycle provides protection against photodamage, the electron transfer capacity of the total electron transfer chain (PS I1 and PS 1) was compared to the capacity of PS I alone (Table 111).
Full chain electron transfer rates in the samples that had received the strong flashes in the presence of background light appeared to remain nearly unaltered. The samples that were exposed to the flashes in the absence of background light displayed more than 20% photodamage (both HL and LL), comparable to the data given in Fig. 1. As opposed to the full chain data, PS I capacity appeared to diminish even when the strong flashes were administered in the presence of background light. The inhibition was stronger in the LL samples. However, in the dark-flashed samples and in comparison to the full chain, the damage to PS I appeared relatively low. Compared to the full chain electron transfer rate numbers, the PS I change in the light-flashed samples is already big. The increased damage observed for the full chain rates in the dark-flashed sample does not correspond to a similar decrease in the PS I sample. The protective function of the xanthophyll cycle therefore ap-

The effect of phofodamaging conditions on the electron transfer capacity of PS II plus PS I and of PS I measured separately
Results are given in m o l of oxygen per mg of Chl a and hour. Assay conditions are given under "Materials and Methods." The control cells were preadapted in light, otherwise sample preparation and exposure to photodamaging conditions were as in Fig. 1 pears to be predominantly effective for PS 11. The observed changes in the relative abundance of the carotenoids may exert effects on the light energy transfer efficiency of PS 11. If so, a lower light energy conductance would give rise to a lesser fluorescence output from PS I1 in the presence of 3-(3,4-dichlorophenyl)-l,l-dimethylurea. To eliminate effects of sample geometry, fluorescence excitation was done with broad blue as well as 628-nm orange light. The latter excites chlorophylls only, the former both chlorophylls and carotenoids. Normalizing on fluorescence yield in the orange by using ratio's equalizes any changes in fluorescence yield of Chl related, for example, to qNP. In HL and LL cultures, the lower fluorescence ratio observed in the dark-preadapted samples indicates that the light energy transfer efficiency from carotenoids to chlorophyll remains higher in darkness than following preadaptation in the light ( Table IV). The difference in the efficiency of energy transfer carotenoids + chlorophyll between the dark and light adapted samples is more pronounced in the HL cells.
The results presented in Table I1 indicate that the formation of the monoepoxide antheraxanthin can only in part be accounted for by conversion of violaxanthin, in addition, the disappearance of zeaxanthin appears to add to antheraxanthin formation as well. The apparent two reactions by which antheraxanthin can be formed addresses the question on the nature of the molecular conversions of zeaxanthin that happen as a part of the protective function. In order to investigate the involvement of nonenzymatic processes, the breakdown of zeaxanthin under in vitro conditions was examined. The HPLC chromatograms shown in Fig. 2 indicate that when isolated zeaxanthin (retention time 13.4 min) was exposed to damaging conditions (10,000 pE.m-2.s-1 white light, 50 "C, air) degradation occurred. With oxygen present, formation of violaxanthin (retention time of 8.6 min) became evident. The identity of the other "breakdown products" with retention times between 16 and 17 min has not yet been extensively determined. The cis peak in the UV region in the absorbance spectra (data not shown) indicated that these may be the different cis-isomers of zeaxanthin. A similar experiment with zeaxanthin was performed in the presence of the singlet oxygen-generating agent eosine, in just room light and at room temperature. Violaxanthin was formed, other zeaxanthin conversion products were nearly absent (data not shown). The in vitro conversion reactions of zeaxanthin support our view on the actual process of singlet oxygen quenching as part of a dynamic xanthophyll cycle in the light: zeaxanthin is recycled into violaxanthin in the light. Zn vivo, antheraxanthin is formed this way as well, either by monoepoxidation of zeaxanthin or by the normal viola-to anther-TMLE IV Light energy transfer eficiency HL and LL cells were preadapted in light or in darkness. DCMU (10 w) was added 1 min prior the fluorescence measurements. Maximal fluorescence emission at 685 nm was measured after excitation with broad blue or orange light (cf. "Materials and Methods"). The Table  depicts  light (10,000 pE.m-2.s-1) during 20 min at 50 "C in the presence of air; B , as A but in the presence of nitrogen gas instead of air, and C , rechromatograped after storage on ice in the dark in the presence of air. Other details are given under "Materials and Methods." axanthin enzymatically catalyzed reactions of the xanthophyll cycle in the light.

DISCUSSION
The two different types of cultures (i.e. LL and HL grown) allowed assays with different contents of xanthophyll cycle pigments present, i.e. relatively abundant in HL cells and low in LL cells as in Thayer and Bjorkman (1990). Applying or omitting actinic background light appeared to be a useful approach to allow or avoid the conversion of violaxanthin to zeaxanthin (Blass et al. (1959) and Yamamoto et al. (1962)).
In earlier studies dithiothreitol was used to study the role of zeaxanthin in the prevention of photodamage. Those experiments precluded the possibility of studies on a dynamically operating cycle. Our approach involved preadaptation of the cells in either darkness or light to install a stable pH in the thylakoid lumen. To this end, the light intensity was chosen to just reach the P,, condition (Fig. l), while avoiding the occurrence of appreciable photodamage (Table I). The lumen pH has been associated with the equilibria of the xanthophyll cycle (Rees et al., 1989 Pfundel andDilley, 1993). By the preadaptation step, the ratio of the xanthophyll cycle pigments was fmed in a given status before the photodamaging flashes were given. The flashes were administered at a low frequency in order to prevent the built up of a proton gradient for the darkpreadapted cells as much as possible. Obviously, if there had been a substantial acidification of the thylakoid lumen analogy with the samples prepared in the presence of actinic background light would have given a diminished breakdown of pigments through the installment of the xanthophyll cycle in the protective mode.
Regardless of the growth conditions and the preincubations, the flashed light induces general photodamage of nearly all FIG. 3. Ascheme depicting the interactions between the various singlet and triplet states of chlorophyll, oxygen, and carotenoids and the role of the xanthophyll cycle in these processes. Reactions indicated with numbers are: 1 , excitation of ground state chlorophyll; 2, direct singlet ground state relaxation of excited chlorophyll by photosynthesis, radiationless transfer, fluorescence, heat release or singlet quenching via zeaxanthin (Demmig- Adams, 1990Owens et al., 1992; 3, chlorophyll triplet quenching by ground state oxygen which produces (via spin reversal) singlet excited oxygen or by ground state carotenoid which produces triplet excited carotenoid; 4 , reaction of singlet excited oxygen with ground state non-or monoepoxy carotenoids resulting in the epoxidated compounds; 5, singlet energy transfer of violaxanthin-absorbed light to chlorophyll; 6 and 7, enzymatic conversions operating in the xanthophyll cycle. Further details are presented in the text. pigments, be it to different extents. A clear exception is the increase for antheraxanthin in the samples that were flashed in the presence of actinic background light. This increase is of great interest for the understanding of the physiological function of the xanthophyll cycle. Comparison of the pigment distribution in between HL with background light only (Table I) and HL flashed with background light present (Table II), shows that the decrease of the violaxanthin content is less than the actual increase of antheraxanthin. The only feasible explanation for this observation is epoxidation of zeaxanthin. We conclude that epoxidation of zeaxanthin under photodamaging conditions in the light also contributes to antheraxanthin formation. Interestingly, earlier work (Hager, 1981;Pfiindel and Dilley, 1993) established the regulatory function of the lightdependent proton gradient formation for the xanthophyll cycle. From that work can be concluded that epoxidation occurs only after relaxation of the proton gradient, i.e. in darkness. Given the conditions in our experiment, changes of the content of antheraxanthin, other than at the expense of violaxanthin, would not be expected (see above). It is concluded that in addition to the "mixed oxidase" function operating in high lumen pH, i.e. darkness (Hager, 1981), a nonenzymatic epoxidation reaction occurs in the light as well, in accordance with Fig. 2.
Control experiments in which purified zeaxanthin was treated with light plus heat in the presence of air indeed gave rise to the formation of the (di-)epoxy compound violaxanthin. This is comparable to the earlier report on the oxidative degradation of antheraxanthin for which in vitro treatment with heat and oxygen has been shown to facilitate the formation of violaxanthin (Thomas and Goodwin, 1965). A recent report describes that oxidative degradation of p-carotene yields monoand diepoxides (Liebler and Kennedy, 1992). This explains our observation that, regardless of the continued presence of a stable proton gradient, formation of antheraxanthin in the light is possible via a nonenzymatic epoxidation of zeaxanthin, The nonenzymatic epoxidation of zeaxanthin results from its function as a photoprotective pigment, i.e. in quenching of singlet oxygen in this particular case. Thus in the light a complete cycle is active. This includes enzymatic reutilization of nonenzymatically epoxidized zeaxanthin (i.e. recycled violaxanthin).
Our work shows that the quenching of excited oxygen by zeaxanthin involves an epoxidation reaction which effectively results in recycling to antheraxanthin and probably violaxanthin as made likely in the in vitro assay. This means that after reaction of zeaxanthin with singlet oxygen, the zeaxanthin is not lost from the cycle but is actually converted into the epoxy compounds antheraxanthin and violaxanthin, through which in the presence of the appropriate acidification of the lumen in the light zeaxanthin can be made again. Table I11 showed that the xanthophyll cycle was most effective in relation to PS 11, the site at which singlet oxygen generation is most likely to occur.
The position of the steady state of all the processes involved determine the actual distribution of viola-, anthera-, and zeaxanthin in a given sample. This way, the xanthophyll cycle has a real dynamic function in the photoprotective process (Fig. 3).
The net decrease of xanthophyll cycle components over the course of exposure to photodamaging conditions is due to the limited number of times that zeaxanthin, in its function as quencher of excited chlorophyll triplet states, is able to withstand trans-cis-trans transitions. In this, according to Krinsky (1971), zeaxanthin has t o become damaged during the quenching at a statistical rate of 1000 quenching events per degradation.
In addition to the chemical modifications associated with the operation of the xanthophyll cycle, a change of the energy transfer efficiency in the carotenoid absorbance region related to the state of the xanthophyll cycle and the amount of the xanthophyll cycle pigments as well, was observed ( Table IV). The difference in the molecular absorbance coefficient between zeaxanthin and violaxanthin cannot be the only reason for this appreciable change. This points to differences in the transfer efficiency between violaxanthin and zeaxanthin to Chl. An explanation for these differences is the number of conjugated double bonds: 9 in violaxanthin and 11 in zeaxanthin. With an increasing number of conjugated double bonds the energy level of the excited states becomes lower, i.e. the zeaxanthin excited states ('4, 'BU) lies below that of violaxanthin by which the possibility of a n energy transfer to the S1 of Chl a from zeaxanthin becomes increasingly unfavorable (Owens et al., 19921.' Violaxanthin has been shown to act as light-harvesting pigment (Owens et al., 1987). This implies that the energy level of the first excited state of violaxanthin is higher than the one of the final Chl acceptor.
The three ways in which the xanthophyll cycle provides protection against photodamage are qNP (singlet transfer), decreased light harvesting capacity (singlet transfer), and photosensitizer-quenching reactions (triplet related). These processes are cooperative: if a carotenoid has a protective func-* A. Friedman and H. Schubert, unpublished results, tion it also has a shadowing effect in the blue region of Chl absorbance and the possibility to quench excited chlorophylls (both singlet and triplet). This effect is important, not only with reference to the mole % numbers presented in Tables I and 11, but more so because of the about 3.5 times higher molar absorbance coefficient of a carotenoid in comparison to Chl. In other words, in cases of excessive irradiation the shadowing effect is useful, but it should be reversed at less than optimal irradiance, which indeed occurs through the enzymatic epoxidation steps in darkness.
The advantages of the xanthophyll cycle are clear, its dynamically adjustable surdshade function excludes the need for a constantly present shadowing pool of carotenoids, the chemical trans-cis-trans heat release involved in triplet Chl a photosensitizer quenching strongly reduces the need for de nouo synthesis to replace photodamaged molecules. To this, the observation in the present study that singlet oxygen quenching provides a means for recycling of zeaxanthin to violaxanthin in the light further extends the functional role of the xanthophyll cycle. The equilibria of the system can rapidly switch from a protective (shadowing, chl triplet, and singlet quenching (Demmig-Adams, 1990)) to a light harvesting (singlet transfer from violaxanthin to Chl) function.
As stated by Hager (1981), the xanthophyll cycle is present in higher plants and green algae but is absent in phycobilisome containing organisms. This remarkable difference may be related to another way of discarding excess excitation energy in cyanobacteria via decoupling of the phycobilisome antennae (Mullineaux et al., 1990). Otherwise, the spectral region of light harvesting in phycobiliprotein containing organisms is largely shifted outside the carotenoid region. This way, light harvesting in cyanobacteria in the blue spectral region is circumvented through which the capacity losses by shadowing carotenoids are in principle negligible in comparision to Chl a and b containing organisms. Cyanobacteria indeed contain a high carotenoid over Chl ratio, to provide for a shading and a photosensitizer quenching function. Likely, these two functions are confined to the cytoplasmic and thylakoid membranes, respectively. The apparent need for a n appreciably higher poolsize of carotenoids acting as photosensitizer quenching pigments may be explained by the lack of a recycling system.
In conclusion, the xanthophyll cycle provides a dynamic tool for Chl a and b containing organisms and possibly also for brown algae with the diatoxanthiddiadinoxanthin conversion: tailor made photosensitizer quenching without loss of light harvesting efficiency under changing light conditions.