Plants and algae have a love/hate relationship with light. As oxygenic photoautotrophic organisms, they require light for life; however, too much light can lead to increased production of damaging reactive oxygen species as byproducts of photosynthesis. In extreme cases, photooxidative damage can cause pigment bleaching and death, a phenomenon all too familiar to anyone who has tried to move a houseplant outdoors into full sunlight. The quantity of the light in natural environments can vary over several orders of magnitude and on a time scale that ranges from seconds to seasons. Because light is such an important environmental parameter, plants have evolved numerous biochemical and developmental responses to light that help to optimize photosynthesis and growth. For example, plants rely on photoreceptors such as phytochrome for shade avoidance responses. Some plants are able to adjust their capacity for harvesting sunlight through leaf and chloroplast movements. During long-term acclimation to changes in light intensity many algae and plants regulate the size of their light-harvesting pigment antennae through changes in gene expression and/or proteolysis. Large antennae are necessary for efficient light capture in limiting light, but they can be a liability when light is abundant or excessive. On a daily as well as seasonal basis most plants receive more sunlight than they can actually use for photosynthesis. Under these circumstances, regulation of light harvesting is necessary to balance the absorption and utilization of light energy, thereby minimizing the potential for photooxidative damage. Besides adjusting light absorption, algae and plants have ways of getting rid of excess light energy that has already been absorbed. This update will focus on protective non-photochemical mechanisms that quench singlet-excited chlorophylls (Chl) and harmlessly dissipate excess excitation energy as heat. These non-photochemical quenching (NPQ) processes occur in almost all photosynthetic eukaryotes, and they help to regulate and protect photosynthesis in environments in which light energy absorption exceeds the capacity for light utilization. We will summarize progress in understanding NPQ that has been made since the last Updatearticle on NPQ appeared in this journal (Horton et al., 1994). Possible fates of excited Chl. When Chl absorbs light it is excited from its ground state to its singlet excited state,1Chl*. From there it has several ways to relax back to the ground state. It can relax by emitting light, seen as fluorescence (1). Its excitation can be used to fuel photosynthetic reactions (2), or it can de-excite by dissipating heat (3); both of these mechanisms reduce the amount of fluorescence. They are therefore referred to as qP and NPQ of Chl fluorescence. Last,1Chl* can, by intersystem crossing, produce3Chl* (4), which in turn is able to produce1O2*, a very reactive oxygen species. Summary of Chl fluorescence quenching processes Summary of Chl fluorescence quenching processes Chl fluorescence measurement from an Arabidopsis leaf. In the presence of only weak measuring light the minimal fluorescence (F o) is seen. When a saturating light pulse is given, the photosynthetic light reactions are saturated and fluorescence reaches a maximum level (F m′). Upon continuous illumination with moderately excess light (750 μmol photons m−2sec1; growth light was 130 μmol photons m−2 sec−1), a combination of qP and NPQ lowers the fluorescence yield. NPQ (qE + qT + qI) can be seen as the difference between F m and the measured maximal fluorescence after a saturating light pulse during illumination (F m′). After switching off the light, recovery of F m′ within a few minutes reflects relaxation of the qE component of NPQ. NPQ can be divided into at least three different components according to their relaxation kinetics in darkness following a period of illumination, as well as their response to different inhibitors (Fig. 2; Horton and Hague, 1988). The major and most rapid component in most algae and plants is the pH- or energy-dependent component, qE. A second component, qT, relaxes within minutes and is more important in algae, but rather negligible in most plants during exposure to excess light. This component is due to the phenomenon of state transition, the uncoupling of LHCs from PSII. qT will not be considered further here because it does not seem to be important for photoprotection (Niyogi, 1999). The third component of NPQ shows the slowest relaxation and is the least defined. It is related to photoinhibition of photosynthesis and is therefore called qI. Absorption of sunlight that exceeds a plant's capacity for CO2 fixation results in a buildup of the thylakoid ΔpH that is generated by photosynthetic electron transport. The decrease in pH within the thylakoid lumen is an immediate signal of excessive light that triggers the feedback regulation of light harvesting by qE. The control by lumen pH allows induction or reversal of qE within seconds of a change in light intensity (see Fig. 2), which is fast enough to cope with natural fluctuations in light intensity that are due to, for example, passing clouds on a partly sunny day. The requirement for low lumen pH is evidenced by the inhibition of qE by uncouplers of the ΔpH such as nigericin. Screening for mutants with lower qE levels has uncovered several mutants that are defective in generation of the ΔpH due to defects in photosynthetic electron transport. In these mutants qP is also affected (Shikanai et al., 1999). However, low lumen pH that induces qE does not have to be generated by light-dependent reactions. Using isolated thylakoids it is possible to induce qE in darkness by simply lowering the pH of the buffer or by generating a ΔpH via ATP hydrolysis and reverse proton pumping by the ATP synthase (Gilmore and Yamamoto, 1992;Krieger et al., 1992). Intensive research during the past several years has led to a concept of the role of the ΔpH in qE. A decrease in lumen pH induces qE through protonation of PSII proteins and activation of xanthophyll synthesis via a xanthophyll cycle. Together, binding of protons and xanthophylls to specific sites in the PSII antenna causes a conformational change that switches a PSII unit into a quenched state with a short 1Chl* lifetime and a low fluorescence yield (Gilmore, 1997). In the following sections we will describe further the mechanistic details of this pH-dependent switch and the physiological significance of qE. Xanthophyll cycles. A, The violaxanthin cycle consists of the de-epoxidation of violaxanthin in high light to first antheraxanthin and then zeaxanthin, catalyzed by VDE; ZE catalyzes the reverse reaction. B, The diadinoxanthin cycle consists of the conversion of diadinoxanthin to diatoxanthin by diadinoxanthin de-epoxidase in high light and the reverse reaction in low light. C, In the recently discovered lutein-5,6-epoxide cycle, conversion of lutein-5, 6-epoxide to lutein might also be catalyzed by VDE. In plants the de-epoxidation reaction is catalyzed by violaxanthin de-epoxidase (VDE). VDE is a 43-kD nucleus-encoded protein that is localized in the thylakoid lumen (Bugos and Yamamoto, 1996). The purified VDE enzyme is activated by low pH (Eskling et al., 1997), and cloning of the VDE gene revealed that the enzyme has a Glu-rich region that may be protonated at low pH (Bugos and Yamamoto, 1996). Upon acidification of the lumen, VDE associates with the thylakoid membrane (Hager and Holocher, 1994) where it can interact with its substrate violaxanthin. VDE uses ascorbic acid (vitamin C) to reduce the epoxide group, and it has aK m for ascorbic acid that is strongly dependent on pH, probably because ascorbic acid rather than ascorbate is the actual cosubstrate (Bratt et al., 1995). A different enzyme, zeaxanthin epoxidase (ZE), catalyzes the epoxidation reactions that complete the violaxanthin cycle. ZE is a flavin adenine dinucleotide-containing, O2-dependent mono-oxygenase that uses reduced ferredoxin to epoxidize first zeaxanthin and then antheraxanthin (Bouvier et al., 1996). Because of its pH optimum of 8, ZE is thought to be located on the stromal side of the thylakoid membrane and to be constitutively active. The level of zeaxanthin is therefore determined by the activity of VDE compared with ZE, with rapid accumulation of zeaxanthin occurring upon activation of VDE in excessive light. ZE and VDE are the first known plant members of the lipocalin family, a diverse group of proteins that bind small lipophilic molecules and share a conserved tertiary structure of eight β-strands in a barrel configuration (Bugos et al., 1998). The amount of zeaxanthin synthesized via the violaxanthin cycle is highly correlated with the level of qE in a large number of plants under a variety of conditions (Demmig-Adams, 1990). In a similar manner, a correlation has been shown between qE and the conversion of diadinoxanthin to diatoxanthin in diatoms (Arsalane et al., 1994). Isolated thylakoids that are devoid of zeaxanthin have been observed to exhibit high levels of qE (Rees et al., 1992), but only at lumen pH values that are lower than those normally occurring in vivo. The requirement for the xanthophylls in qE has been investigated in vivo by using inhibitors and mutants. Dithiothreitol has been used extensively as a remarkably specific inhibitor of VDE and diadinoxanthin de-epoxidase. Blocking zeaxanthin synthesis in leaves with dithiothreitol results in inhibition of qE, the extent of inhibition depending on the plant species examined (Horton et al., 1994). Mutants that are unable to convert violaxanthin to antheraxanthin and zeaxanthin have been isolated from Arabidopsis andChlamydomonas (for review, see Niyogi, 1999). As in the inhibitor studies, lower levels of qE accompanied the lack of zeaxanthin in these npq1 mutants. Recent studies using antisense VDE in tobacco have confirmed the results obtained from mutant analyses (Chang et al., 2000). The alga Mantoniella squamata has an incomplete xanthophyll cycle, only leading to antheraxanthin in vivo, but still exhibits qE (Goss et al., 1998). This implies that the role of zeaxanthin in qE can be replaced by antheraxanthin in this alga. The involvement of antheraxanthin in plants had been proposed in earlier studies where zeaxanthin-independent qE could be explained by taking the amounts of antheraxanthin into account (Gilmore and Yamamoto, 1993). Therefore, it has become a common practice to calculate the level of de-epoxidation of a given organism as the amount of antheraxanthin and zeaxanthin in comparison with the total amount of antheraxanthin, zeaxanthin, and violaxanthin. In addition to antheraxanthin and zeaxanthin, a third xanthophyll, lutein, has also been implicated in qE. These suggestions were supported by studies on the Chlamydomonas lor1mutant, which only lacks lutein and loroxanthin, but shows lower qE than the wild type (Niyogi et al., 1997). A similar mutant in Arabidopsis, lut2, which is defective in the lycopene ε-cyclase and therefore lacks lutein, also has less qE (Pogson et al., 1998). Double mutants of Chlamydomonas or Arabidopsis that lack lutein and zeaxanthin are totally devoid of any qE and are very sensitive to high light (Niyogi et al., 1997, 2001). Furthermore, plants that overexpress the lycopene ε-cyclase have an increased lutein content and show a slight, but significant increase in the rate of qE induction, even though their xanthophyll cycle pool size is reduced in comparison with the wild type (Pogson and Rissler, 2000). It is interesting that a third kind of xanthophyll cycle involving lutein-5,6-epoxide has been found in a parasitic plant, Cuscuta reflexa (Fig. 3). In this plant neoxanthin is missing and is replaced by lutein-5, 6-epoxide. In high light, lutein-5,6-epoxide is de-epoxidized to lutein, presumably by VDE, which was previously shown to use lutein-5,6-epoxide as substrate (Bungard et al., 1999). The existence of this new cycle is consistent with the idea that lutein has a photoprotective function that the epoxide lacks. Although zeaxanthin is generally necessary for maximal qE in vivo, it is not sufficient. In mutants that accumulate zeaxanthin constitutively, qE must still be induced by a low pH (Niyogi, 1999). This demonstrates that the low pH has an additional role in qE, besides activation of the xanthophyll cycle. Lowering the pH in the thylakoid lumen not only activates the de-epoxidation of violaxanthin to zeaxanthin, but it is also necessary for a conformational change in the thylakoid membrane that can be monitored by absorbance changes. Two high light-induced absorbance changes in leaves or isolated thylakoids are associated with qE. One absorbance change occurs at 505 nm and is due to the conversion of violaxanthin to zeaxanthin. The second one at 535 nm (ΔA535) depends on both zeaxanthin and low pH and is thought to be due to a conformational change in the thylakoid membrane (Krause, 1973; Bilger and Björkman, 1994). qE is always accompanied by the ΔA535. Conformational changes have also been inferred from measurements of Chl fluorescence lifetime distributions, which depend on the molecular environment of the excited Chl (Gilmore, 1997). In these experiments the presence of a ΔpH alone causes a lifetime shift from approximately 2 to 1.6 ns. This shift likely reflects a protonation-dependent conformational change that is independent of zeaxanthin. When both ΔpH and zeaxanthin are present, a fluorescence lifetime component at 0.4 ns appears at the expense of the 1.6 ns component. The amount of the 0.4 ns component is proportional to qE. Together, the absorbance and Chl fluorescence lifetime results suggest that a conformational change due to binding of protons and xanthophylls (maybe zeaxanthin) is necessary for qE. Several LHC proteins associated with PSII have been implicated in qE. In particular the minor LHC proteins CP29 and CP26 were suggested to be involved in qE based on the relative enrichment of associated xanthophyll cycle pigments (Bassi et al., 1997) and binding ofN,N′-dicyclohexylcarbodiimide (Walters et al., 1994), an inhibitor of qE that reacts with proton active residues. However, these proteins have not been found in some organisms that exhibit qE such as Mantoniella and diatoms. To identify proteins involved in qE Arabidopsis mutants have been isolated that are defective in qE, but have normal xanthophyll levels (Li et al., 2000; Peterson and Havir, 2000). Characterization of one of these mutants, npq4-1, revealed that a PSII protein, PsbS, is essential for qE (Li et al., 2000). PsbS belongs to the LHC protein superfamily, but it has four transmembrane helices instead of three and different pigment-binding characteristics (Funk et al., 1995). Despite the absence of the PsbS protein, light harvesting is not impaired in the npq4-1 mutant and the levels of the other LHC proteins are normal. However, in addition to lacking qE, npq4-1 also lacks the conformational change monitored by ΔA535 (Li et al., 2000). The pH- and xanthophyll-dependent conformational change and the PsbS protein are necessary for qE, but the actual biophysical mechanism of 1Chl* de-excitation is still unknown. The central and long-standing question is whether the involvement of xanthophylls is direct or indirect. The xanthophylls may act indirectly as allosteric regulators of the LHCs that cause a switch from light harvesting to energy dissipation (qE; Horton et al., 2000). In this case the conformational change must somehow facilitate1Chl* de-excitation, which may occur by internal conversion of Chl itself to the ground state, releasing excitation energy as heat. Isolated, detergent-solubilized LHCs can exhibit pH-dependent Chl fluorescence quenching, which is inhibited by violaxanthin and promoted by zeaxanthin (Ruban et al., 1997). On the other hand, xanthophylls may directly de-excite1Chl*. This is theoretically possible according to recent spectroscopic experiments, which showed that isolated xanthophyll cycle pigments possess a lowest singlet excited state that is below that of 1Chl* (Polı́vka et al., 1999; Frank et al., 2000). However, because both violaxanthin and zeaxanthin could potentially accept energy from1Chl*, these results do not explain why the xanthophyll cycle is necessary for qE. In vivo, other factors such as different binding sites for zeaxanthin and violaxanthin, as well as distance and orientation of these xanthophylls relative to Chl may be important in energy Therefore, rather than between zeaxanthin and violaxanthin are likely to be For example, the xanthophylls that are important for qE antheraxanthin, lutein, and all have a de-epoxidized (Fig. in to violaxanthin, and other A for qE. A, In limiting light or darkness quenching B, In high light a likely several bind to PsbS and LHC a shift in Chl fluorescence lifetime to 1.6 ns. C, A quenching with a different as is when zeaxanthin and protons are the Chl fluorescence lifetime to 0.4 ns. of violaxanthin to zeaxanthin occurs more than might bind to the as violaxanthin or a different When the light has the PSII proteins are the epoxidation of zeaxanthin to violaxanthin is In the of from A is not possible in vivo, but can be induced by the pH in the thylakoid The is based in on by Horton (Horton et al., The mutants have been for the photoprotective function of qE during high light The Arabidopsis mutants npq1 and are more sensitive to photoinhibition than the wild type in a scale of et al., and Niyogi, that qE normally to protect PSII. The photoprotective of qE may be due to production of and other reactive oxygen species. qE may also the of the electron and the of the lumen, which are known to PSII to After several in high light, the npq1 mutant showed more photooxidative bleaching and and Niyogi, 1999). to antisense tobacco plants that lack VDE showed a significant increase in photoinhibition and a decrease in pigment content when to high light or a combination of light and in a growth et al., 2001). When npq1 lut2, an Arabidopsis mutant missing zeaxanthin and lutein, into high light even more photooxidative bleaching and was (Niyogi et al., 2001). These results that xanthophylls have a function not only in qE, but also in the of the thylakoid membrane photooxidative damage. may directly protect the thylakoid membrane are in that are reactions. within leaves with the in leaves of but not in that the photoprotective role of zeaxanthin is not to the LHCs et al., 2000). may be an important in the thylakoid membrane where it could reactive oxygen species and/or reactions. and lutein the in made from in response to a et al., 1999). the of zeaxanthin and lutein has important not only for but also for of the where these xanthophylls are found could also have a function in the This xanthophyll has been shown to decrease the of the membrane and 1997), and a decrease in could be important by lowering the of reactive oxygen species the In qE PSII high light and fluctuations in light xanthophylls have an additional photoprotective role in high light. Although the light of the mutants of Arabidopsis and the VDE antisense plants of tobacco has been these mutants are remarkably of light. In leaves of the mutants are to high light or et al., 2000; et al., that other important photoprotective mechanisms or other can at least for the lack of qE and/or Under more light qE is replaced by a component of called qI. In to qE, is much less and might be due to a of photoprotection and Chl fluorescence measurements can help to between photoprotective mechanisms and The fluorescence level in the (see Fig. 2), is in direct to the maximal F by the photoprotective quenching qE, photoinhibition normally the F level the F m level (Gilmore et al., 1996). plants show an acclimation to the by the xanthophyll pool as well as by an increased of zeaxanthin and antheraxanthin that is associated with et al., 1999). to that dissipate excess excitation energy (Gilmore and 2000). of Chl fluorescence lifetime of this state have revealed lifetime changes that are similar to those observed during qE, but that are only at (Gilmore and 2000). and reversal of in these leaves may major of in the thylakoid Some of the induced by low is actually as qE because it is an this kind of can be (Gilmore and Björkman, 1995). It may be due to the of the ΔpH by the reverse proton pumping catalyzed by the (Gilmore, 1997). Therefore, in to normal qE it relaxes only in has since the last on this was in this journal (Horton et al., 1994). on physiological studies, the function for qE to be the of PSII from However, when qE is other mechanisms are able to for long-term at least in the absence of additional studies using qE mutants are necessary to the of qE and to other important photoprotective mechanisms that qE. The PsbS protein is an essential component of the mechanism of qE. PsbS itself may be the of qE in the or it may function in with other LHC To PsbS actually in qE, its within the as well as possible and pigment-binding sites within PsbS have to be Characterization of additional mutants or the use of reverse may into the involvement of other new in the has been the that both zeaxanthin and violaxanthin are potential of excitation energy from In the it will be important to experiments that measurement of the energy levels of xanthophylls in their protein environment and to there is a direct energy from Chl to a The of new and diverse from to to will be necessary to qE, a response of photosynthetic to excess light We and for results to and and for of the We to we were not able to due to