Elucidating the molecular details of qE (energy quenching) induction in higher plants has proven to be a major challenge. Identification of qE mutants has provided initial information on functional elements involved in the qE mechanism; furthermore, investigations on isolated pigment–protein complexes and analysis in vivo and in vitro by sophisticated spectroscopic methods have been used for the elucidation of mechanisms involved. The aim of the present review is to summarize the current knowledge of the phenotype of npq (non-photochemical quenching)-knockout mutants, the role of gene products involved in the qE process and compare the molecular models proposed for this process.
Organization of PS (Photosystem) I and II supercomplexes
Plants use light as the energy source for their metabolism. During the early steps of photosynthesis, solar energy is absorbed efficiently, and excitons are transferred to the photosynthetic reaction centres by a complex array of pigment-binding proteins, the LHCs (light-harvesting antenna complexes), localized at the periphery of each PS [1,2]. LHC proteins not only are involved in light harvesting, but also act in photoprotection by multiple mechanisms, including Chl (chlorophyll) singlet energy dissipation, Chl triplet quenching and scavenging of ROS (reactive oxygen species). The antenna system thus has a dual function: on one hand, it harvests photons and extends the cross-section for light absorption under light-limiting conditions; on the other hand, it limits damage to the photosynthetic apparatus when light is in excess. In higher plants, two classes of proteins compose the antenna system, which can be divided into two moieties. The inner antenna is made by the plastid-encoded proteins, CP43 and CP47, which bind Chl a and β-carotene, whereas the outer moiety is composed of the nuclear-encoded LHC proteins, which bind Chl a, Chl b and xanthophylls. In plants the PSII outer antenna consists of one copy each of three monomeric proteins called CP29 (Lhcb4), CP26 (Lhcb5) and CP24 (Lhcb6) and two to four copies, depending on light conditions during growth, of the major antenna complex, called LHCII. The latter is a heterotrimer of the Lhcb1, Lhcb2 and Lhcb3 subunits in different combinations [3–6]. The D1/D2/cytochrome b559 subunits constitute a ‘core’ complex with the inner antenna moiety and a core dimer, which also includes many small transmembrane subunits , forms a supramolecular complex together with two copies each of the monomeric Lhcb4, Lhcb5 and Lhcb6 subunits and, more peripherally, with several LHCII trimers [2,8,9]. PSI is simpler, with electron-transport domains fused with the inner antenna in a core complex composed of PsaA/PsaB large subunits and many small elements. The outer antenna is smaller than in PSII with only four Lhca subunits organized as a crescent on one side of the monomeric core moiety [10,11].
The need for photoprotection and the role of xanthophylls
Under normal stable light conditions, the efficiency of the photosynthetic machinery is optimal for a given light intensity once plants are acclimatized to it. Nevertheless, in the natural environment, plants often experience rapid fluctuations in light intensity, temperature and water availability that easily lead to overexcitation of photosystems. Therefore, since absorbed light may exceed the capacity to use reducing equivalents for CO2 fixation and primary metabolic roles, such conditions lead to over-reduction of the NADP+ pool and to superoxide (O2−) production in the chloroplast by the Mehler's reaction [12,13]. O2− can be metabolized to hydrogen peroxide (H2O2) or hydroxyl radical (OH•), the latter being a very aggressive ROS [14–16]. Furthermore, incomplete photochemical quenching leads to an increased lifetime of Chl excited state (1Chl*), increasing the probability of Chl a triplet formation (3Chl*) by intersystem crossing. 3Chl* promptly reacts with oxygen (3O2) to form single oxygen (1O2), a harmful ROS. Thus PSII and LHCs, when overexcited, become an important source of 1O2 [17,18]. These harmful events are counteracted by photoprotection mechanisms that either detoxify the ROS produced or prevent their formation. ROS can be deactivated by small chloroplast antioxidant molecules, including α-tocopherol, glutathione and carotenoids, or enzymes, such as the thylakoid-bound superoxide dismutase and ascorbate peroxidase. Carotenoids, and their oxygenated derivatives, xanthophylls, play a key role as photoprotective agents. The biosynthetic pathway of carotenoids is shown in Figure 1. The carotenoid composition of thylakoids is not constant and undergoes modifications during long-term acclimatization of plants to stress conditions, as well as during rapid fluctuations of solar light intensity, owing to the operation of the xanthophyll cycle . The xanthophyll cycle involves the xanthophylls violaxanthin (Viola) and zeaxanthin (Zea), and consists of a light-dependent reversible de-epoxidation of Viola to Zea via the intermediate antheraxanthin; the reaction is catalysed by VDE (violaxanthin de-epoxidase), a luminal enzyme activated by the build-up of a high transmembrane proton gradient, which is a good indicator for the presence of light in excess [20,21].
Carotenoid biosynthetic pathway in higher plants
Plants have evolved several mechanisms in order to minimize ROS production. Chloroplasts are able to relocate their positions in a cell in response to the incident light intensity, moving to the side wall of the cell to avoid strong light, but gathering at the front face under low light in order to maximize light harvesting . Within the chloroplast, both PSII antenna size and composition are adjusted through changes in gene expression and protein turnover [23–25] or through covalent modification of LHCII proteins by phosphorylation, which modulates the interaction of LHCII with PSII and PSI [26–28] thus balancing the energy pressure between the two PSs (state transition). Mechanisms acting as alternative sinks for excess electrons, such as photorespiration [29,30], water–water cycle  or cyclic electron transport , appear as photochemical strategies to relieve overexcitation.
Excess energy dissipation into heat
Since 3Chl* production is a constitutive intrinsic property of Chls, the capacity to control its formation is essential for plant survival. This is obtained by preventing the formation of the 3Chl*, derived from excess 1Chl*, through a set of inducible mechanisms, collectively referred to as NPQ (non-photochemical quenching). NPQ allows for the harmless thermal dissipation of excess absorbed photons by PSII  and can be monitored as a light-dependent quenching of leaf Chl fluorescence , whereas thermal dissipation can be evaluated by measuring the heat production by photo-acoustic spectroscopy . The predominant NPQ component is triggered by the pH change across the thylakoid membrane and rapidly reversible. It is thus dependent on the ‘energization’ of the chloroplast and defined as qE (energy quenching) [36–38]. In addition to qE, a slowly relaxing component of the NPQ process is known as qI (inhibitory quenching), which relaxes with a half-time of approx. 30 min. This component has been attributed to photoinhibition and its relaxation is suggested to be dependent on D1 re-synthesis . Although this attribution is justified when NPQ is induced by very HL (high light), qI is also observed under intermediate light conditions, where it depends on the accumulation of Zea . A third quenching component (qT), relaxing within minutes, has been reported . Its attribution to the establishment of state 1–state 2 transitions appears unlikely on the basis of an unchanged NPQ in the stn7 mutant lacking the LHCII kinase. Moreover, LHCII phosphorylation is saturated at 200 μEm−2·s−1 , a light intensity rather inefficient for NPQ induction.
Overall, the ability of plants to modulate light-utilization efficiency to the fluctuating light intensity, through a feedback mechanism coupled to pH changes in the chloroplast lumen, is determinant for plant fitness in natural environments [42,43].
In the following sections, we summarize the current knowledge regarding the photoprotective mechanism of excess energy dissipation (NPQ) in higher plants and discuss the different models proposed.
Genetic analysis of qE
A number of studies have used genetic analysis in order to identify mutants in NPQ. qE mutants can be classified into three major groups: (i) mutants with altered capacity of build-up of the ΔpH gradient or that detect luminal pH; (ii) mutants affected in pigment composition; and (iii) knockout mutants lacking external antenna protein subunits of PSII (Lhcbs).
Mutants altered in ΔpH building or luminal pH sensing
The excess proton concentration in the thylakoid lumen is perceived as a signal that the absorbed light exceeds the capacity of electron transport to NADP+ and/or the capacity of ATPase to use the proton gradient for ATP synthesis, thus requiring thermal energy dissipation. Treatment with the ionophore nigericin collapses ΔpH and prevents the activation of NPQ otherwise activated within a few seconds of exposure to HL . Thus mutations that reduce ΔpH building capacity also affect qE amplitude and kinetics. LET (linear electron transport) and CET (cyclic electron transport) are both coupled to the generation of a trans-thylakoid ΔpH that drives ATP synthesis , whereas only LET produces NADPH. It has long been discussed whether CET around PSI is involved in regulating photosynthesis via lumen acidification , and, in the last few years, isolation of several mutants defective in CET and with altered qE phenotype provided further clues. In vascular plants, PSI CET consists of two partially redundant pathways: the PGR (protein gradient regulation) 5-dependent pathway, which is inhibited by AA (antimycin A), and the AA-resistant NAD(P)H dehydrogenase-dependent pathway . Arabidopsis mutants pgr5 and pgr1 were identified owing to their low qE phenotype by Chl fluorescence imaging [47,48]. Deletion of subunits involved in the PGR5-dependent pathway revealed that it is essential for qE induction , whereas the contribution of the NAD(P)H dehydrogenase-dependent pathway to ΔpH building is not significant for qE activation .
Signal transduction of lumen over-acidification involves the PSII subunit PsbS that is essential for qE induction (Figure 2), as demonstrated by the phenotype of the npq4 mutant [50,51]. PsbS belongs to the LHC protein superfamily, but differs from other members for having four transmembrane helices rather than the three generally found in most LHC proteins  and for not binding pigments as shown from the non-conservation of Chl-binding residues . Typical of this protein is the presence of two lumen-exposed glutamate residues, Glu122 and Glu226, that bind DCCD (N,N′-dicyclohexylcarbodi-imide), a protein-modifying agent that covalently binds to protonatable residues in hydrophobic environments . In Arabidopsis, mutation of each glutamate to a non-protonatable residue, e.g. E122Q and E226Q, decreased by 50% both qE and DCCD-binding capacity, whereas the double mutant has a qE-null phenotype like npq4 .
NPQ phenotype of WT and
npq Arabidopsis mutants
Altogether, phenotypes of qE-triggering mutant (npq4) and CET mutant (pgr) confirmed the tight dependence of NPQ on the low luminal pH and the detection of acidification through PsbS.
Xanthophyll biosynthesis mutants
Single and multiple mutations targeting genes encoding enzymes in the carotenoid biosynthesis pathway have been produced by several laboratories, yielding mutants with altered xanthophyll composition. Most of these mutations affect both qE amplitude and kinetics, thus suggesting that xanthophylls have important roles in qE.
The npq1 mutant lacks VDE activity and is thus unable to convert Viola into Zea upon exposure to HL ; qE in npq1 has approx. 30–40% amplitude with respect to WT (wild-type), showing that Zea synthesis is needed for full expression of qE in Arabidopsis. Moreover, quenching in npq1 is saturated within 1 min, whereas WT, in addition, develops a slower phase of quenching, proceeding for up to 30 min , which is related to Zea accumulation (Figure 2). Upon its synthesis, Zea binds to specific binding sites in Lhcb proteins, namely the site L2 in monomeric Lhcb4–Lhcb6 [57,58] and the peripheral V1 site of LHCII [59,60]: the slow kinetics of the Zea-dependent qE component reflect the replacing of Viola bound to sites L2 and/or V1 in Lhcb proteins with Zea, a process limited by diffusion in the lipid phase. This suggestion is consistent with results of experiments in which DTT (dithiothreitol) is used as VDE activity inhibitor: DTT blocks the slow phase of NPQ rise [61,62], while retaining the fast, initial, phase of quenching. Thus a component of qE is activated independently of the presence of a functional xanthophyll cycle  and is due to the action of a xanthophyll species already bound to Lhcb proteins in the dark, before NPQ induction. This xanthophyll appears to be lutein, the most abundant among plant xanthophylls: the lut2npq1 double mutant, lacking both Zea and lutein, has an NPQ-null phenotype . Residual qE measured in lut2 plants has been attributed to Zea. In fact, the lut2 mutant has both a low NPQ and slow kinetics [65–68], whereas DTT treatment of lut2 plants (Figure 2) phenocopy lut2npq1, implying that the slowly developing quenching is produced from newly synthesized Zea. The mutation szl1 (suppressor of zeaxanthinless1), identified on the npq1 genetic background and promoting a higher lutein/β-xanthophyll ratio, is effective in partially releasing the qE restriction owing to the npq1 mutation, suggesting that lutein can replace Zea and perform quenching, binding to site L2 of monomeric Lhcb4 and Lhcb6 as proved using recombinant proteins with lutein in L2 . The above evidence suggests that: (i) lutein, in addition to Zea, has a key role in qE modulation and for the full expression of qE ; and (ii) lutein and Zea modulate distinct kinetic components of qE. Lutein is stably bound to several LHCs, whereas Zea is accumulated upon de-epoxidation of Viola, released by peripheral xanthophyll-binding sites on LHCII. The rapid phase in qE induction kinetics, retained in npq1 plants, together with the effect of szl1 mutation, indicate that the functional component responsible for the retained qE in npq1 is lutein, already bound to photosynthetic complexes before light exposure. This is in accordance with the lack of the rapid component of NPQ induction in lut2 plants. The slower phase of fluorescence quenching measured in lut2 leaves is due to Zea. Indeed, by infiltrating lut2 leaves with DTT, the qE amplitude decreases to levels of lut2npq1 (Figure 2).
The primary role of Zea in qE induction is confirmed by the phenotype of npq2 mutants; in these plants, blocking the Zea epoxidation reaction leads to accumulation of high levels of Zea, even in dark-adapted plants [55,71], which leads to a more rapid saturation of qE maximal amplitude with respect to WT, with kinetics presumably limited only by the build-up of a trans-thylakoid ΔpH (Figure 2).
Altogether, lutein and Zea appear to be determinant for qE amplitude, whereas Viola and neoxanthin (Neo) are not involved . This suggests the presence of two distinct quenching sites, either lutein- or Zea-specific. The specificity is proved by the fact that lutein-only plants  or Zea-only plants  cannot reach the maximal amplitude of NPQ. Indeed, the double mutant lut2npq2, in which all xanthophylls are replaced by Zea , shows NPQ amplitude similar to lut2 (Figure 2).
Knockout mutants lacking single or multiple PSII Lhcbs
The sections above showed that PsbS is essential for triggering of NPQ, whereas both lutein and Zea are needed for the full activation of the mechanism. The npq4 mutation is epistatic over mutations in the xanthophyll biosynthesis pathway [74,75], implying that both lutein and Zea act downstream of the mechanistic step controlled by pH sensing through PsbS. In order for PsbS protonation to yield dissipation of 1Chl* and fluorescence quenching, this event must affect a Chl-binding protein. Such a protein should also bind lutein and Zea as shown above or, at least, should interact tightly with a xanthophyll-binding protein, thus providing a quenching effect. Early works proposed that PsbS might bind both Chls and xanthophylls  or Zea alone , making it a candidate for the role of quencher. Nevertheless, further work pointed to the non-conservation of Chl-binding residues in PsbS , while its properties both in vivo and in vitro are not consistent with binding of xanthophylls . Once PsbS is ruled out, Lhcb proteins appear to be ideal candidates for the role of quenching sites owing to their capacity to bind Chls and xanthophylls, with lutein and Zea binding to distinct sites, which could well catalyse the two distinct and complementary quenching events identified above. Indeed, the ch1 mutant of Arabidopsis that lacks Chl b, thus leading to degradation of LHC proteins , exhibits a strongly reduced capacity of NPQ in the presence of both lutein and Zea, suggesting that LHCs are needed for the lutein/Zea-quenching events. Indeed, the small residual fluorescence quenching measured on ch1 and npq4 mutants has been attributed to a quenching mechanism located within the PSII core complex .
Lhcb proteins fall into two groups with respect to the sites where they bind lutein and Viola/Zea: the first group includes Lhcb1, Lhcb2 and Lhcb3, the components of the major trimeric LHCII, binding lutein at sites L1 and L2, whereas Viola/Zea bind at site V1 [60,80,81]. The second group includes monomeric Lhcb4, Lhcb5 and Lhcb6 which bind lutein at site L1, with Viola/Zea binding at site L2 [82–84]. These proteins do not have a V1 site and exchange Viola with Zea in site L2 [85,86], whereas Lhcb1–Lhcb3 do not. This feature is consistent with the slow kinetics of the Zea-dependent quenching component of NPQ and with the need for de novo synthesis of Zea for quenching (Figure 2). Zea binding to Lhcb4–Lhcb6 results in a conformational change  and in a decrease in the fluorescence lifetime [73,87–89]. The presence of a Zea-binding site effective in providing enhancement of NPQ is not a property of a single LHC protein since Zea-dependent enhancement of the mechanism has been observed in plants depleted in Lhcb6, Lhcb4 and LHCII .
The role of individual LHCs has been investigated using reverse genetics. Down-regulation of Lhcb1+Lhbc2 and knockout of Lhcb3 did not significantly decrease NPQ amplitude or slow down its kinetics [91,92]. Targeting monomeric Lhcbs yielded different results: Lhcb5-knockout plants retained qE [93,94], whereas the qI component of NPQ was down-regulated . qE was affected in Lhcb6- and Lhcb4-knockout plants [56,93–95]. In summary, depletion of a single monomeric Lhcb protein could not completely abolish NPQ, implying redundancy within the subfamily members. The making of a mutant lacking all three monomeric Lhcb4–Lhcb6 proteins is awaited in order to verify whether NPQ can be sustained in the absence of these gene products. Nevertheless, the work accomplished so far highlights the role of Lhcb4–Lhcb6 gene products on NPQ, whereas Lhcb1–Lhcb3 antisense or Lhcb3-knockout had little effect on it.
Molecular mechanisms of excess energy dissipation
Elucidating the molecular details of qE induction in higher plants has proven to be a major challenge. Reverse genetics has focused the attention on Lhcb proteins; nevertheless, systems simpler than the whole plant have been used in order to perform spectroscopy, aiming to identify the underlying mechanism(s) that catalyse energy dissipation into heat. At present, investigations by several groups worldwide have contributed to the formulation of three mechanistic models for excess energy dissipation.
(i) Aggregation-dependent LHCII quenching. According to this model, qE occurs upon aggregation of the major trimeric LHCII complex of PSII. This produces a conformational change within the protein and promotes energy transfer from Chl a to a low-lying S1 excited state of lutein bound to site L1 of LHCII [96–98].
(ii) CT (charge-transfer) quenching model. Energy absorbed in excess induces formation of a carotenoid radical cation by charge separation within a Chl a–Zea heterodimer followed by charge recombination at the ground state . The process is located in monomeric Lhcb4–Lhcb6 proteins and does not occur in LHCII [63,100]. Lutein can also be active in this process .
(iii) Zea radical cation in Lhcb–PsbS oligomers. This model agrees with the second model as for the involvement of carotenoid radical cations in quenching, but proposes that the interaction between Chl a and Zea is promoted by the binding of a Zea–PsbS complex with either LHCII or a monomeric Lhcb subunit [101,102].
Aggregation-dependent LHCII quenching
This hypothesis was proposed based on experimental evidence that aggregation of isolated LHCII trimers or monomers, induced in solution by low detergent concentration and/or low pH, causes a decrease in Chl fluorescence lifetime [103–105]. This early model has been refined over the years and presently incorporates a detailed description of the conformational changes induced by aggregation within the LHCII molecule , yielding a tight interaction between lutein bound into the L1 site and Chl a chromophores [98,107]. Evidence that the process as observed in purified LHCII is related to NPQ in vivo relies on the similar spectral changes, attributed to Chl a and Neo, detected in leaves during establishment of NPQ and aggregation of LHCII [97,98,103]. Recently, support for this model was provided by the observation of a red-shifted fluorescence lifetime component in both aggregated LHCII trimers binding Zea and quenched leaves . Zea bound at site V1 of LHCII, acts as an allosteric modulator of lutein-dependent quenching [108,109], whereas aggregation in vitro has been propose to entrain an intrinsic conformational transition in LHCII, in turn responsible for establishment of the quenching reaction . Car S1–Chl excited state coupling was recently measured in isolated LHCII and correlated with the NPQ amplitude in vivo in different mutants such as npq1, npq2, lut2 and PsbS-over-accumulating lines .
Criticisms of this model have been raised on the basis that: (i) the effect of down-regulating the components of LHCII in vivo, namely Lhcb1+Lhcb2  or Lhcb3  is, at best, very small; (ii) quenching and other spectral changes attributed to LHCII occur in monomeric Lhcb4 and Lhcb5 proteins as well, even more rapidly than in LHCII [104,112]; (iii) lutein cannot be the only quencher during NPQ, since the lut2npq2 mutant having Zea as the only xanthophyll is active in NPQ as well as the lutein-less mutant lut2 (Figure 2).
CT quenching in minor complexes
The proposal that NPQ is based on the formation of a CT state between Chl a and Zea has been proposed based on quantum chemical calculations  and ultra-fast pump-probe experiments on isolated thylakoid membranes . The CT mechanism, which accounts for a large fraction of qE , involves energy transfer from bulk Chl molecules to a Chl–Zea heterodimer that undergoes charge separation followed by recombination, thereby transiently producing a Zea radical cation (Zea•+) with a very short relaxation time (50–200 ps), as expected for an efficient quencher. Formation of Zea•+ in thylakoids depends on the three components needed for qE in vivo: lumen acidification, PsbS activation and Zea production [99,100]. The signal from Zea•+ formation has been found in isolated monomeric Lhcbs, but not in LHCII [63,100,114–116], consistent with reverse genetic data showing that no single LHC subunit is essential for qE [93,95,100]; furthermore, the localization of these complexes between LHCII and the reaction centre makes a perfect setting for regulating excitation energy transfer from the peripheral LHCII to the core subunits. Mutation analysis of Chl-binding sites in Lhcb4  showed that a Chl pair (Chl A5 and Chl B5) rather than a single Chl a chromophore, is critical for CT quenching. The involvement of a Chl pair is reasonable since charge delocalization over the two Chl would stabilize the CT state. Chl A5–B5 are located in close proximity to the L2 domain, whereas Zea binding to this site induces a conformational change, bringing Chl A5 into excitonic interaction with Chl B5, switching the protein to a dissipative state by Zea•+ formation. Also, Lhcb6 antenna complexes show Zea•+ formation, whereas in the Lhcb5 complex, two distinct CT quenching sites are detected, involving Zea and lutein radical cation species respectively, depending on Zea binding to the L2 binding site. Thus Zea in the L2 site acts both as a quencher and as an allosteric modulator of lutein CT efficiency into site L1 . Lutein•+ was also recently detected in Lhcb6 and Lhcb4 complexes binding lutein as the only xanthophyll . This result, combined with the rescuing of the NPQ phenotype of npq1 in the double mutant slz/npq1, exhibiting increased lutein content, implies that not only Zea, but also lutein, is involved in CT quenching and accounts for the presence of partial NPQ in Zea-less mutants.
Unanswered problems by the CT model include: (i) a double mutant lacking both Lhcb5 and Lhcb6 and reduced in Lhcb4 still retains most NPQ activity; (ii) a low level (~1%) in vitro of minor complexes undergoing CT quenching compared with in vivo (30%), implying the presence of factors, possibly PsbS, ΔpH or interactions with protein partners, which stabilize the dissipative conformation  in vivo; (iii) the LHC protein conformational change induced by interaction with PsbS has not been reproduced in vitro, so far; and (iv) the relationship between CT quenching and the S1 population, as a consequence of charge recombination upon carotenoid radical cation formation, is still under debate.
PsbS–Zea complex interaction with Lhcbs
The model proposed from Kuhlbrandt and co-workers is a variant of the above-described CT model [101,102]. They agree that Zea•+ formation has a primary role in modulating qE amplitude and propose that trimeric LHCII or monomeric LHCs are involved in the catalysis of NPQ, despite the fact that LHCII trimers in solution could not form Zea•+ [100–102], by postulating that Zea is brought to LHCII by interaction with PsbS. However, PsbS has not been found as a cross-linking product of LHCII  and cannot bind Zea in vivo or in vitro [52,75].
Triggering of qE
Chloroplast lumen acidification is the original event activating NPQ. The signal transduction into a quenching state appears to be obtained by two independent, although complementary, pathways. The first event is the activation of VDE in the chloroplast lumen. VDE is a monomer soluble in the luminal space at neutral pH, whereas it becomes a dimer bound to the thylakoid membrane at low pH , where it catalyses the synthesis of Zea from pre-existing Viola , bound to the V1 site of LHCII which becomes labile at low pH [60,120]. Newly formed Zea binds to site L2 of monomeric Lhcbs  where it induces the conformational change which allows the interaction between Chl A5 and Chl B5 in Lhcb4 and probably in Lhcb5 and Lhcb6 [63,121], and the Chl a–lutein interaction in the Lhcb5 L1 site .
The second pathway is more direct: low pH protonates two glutamate residues on the luminal loops of PsbS [50,54], each responsible for 50% of the NPQ amplitude, thus implying that PsbS somehow interacts with LHC proteins binding Chl a and xanthophylls responsible for quenching by one or more of the above-described mechanisms. Nevertheless, the mode of interaction is still obscure. In one proposal, PsbS would promote LHCII aggregation , whereas the level of quenching thus induced, is allosterically regulated by the de-epoxidation state of site V1 . An alternative model (by Kuhlbrandt and co-workers [101,102]) postulates that PsbS, a dimer at neutral pH, dissociates at acidic pH into monomers [117,123]. This dimer–monomer transition might uncover the Zea-binding site on the hydrophobic surface of PsbS; PsbS monomers could transiently bind pigments (Zea) under qE conditions and bring them into close proximity with either LHCII or a minor Lhcb, forming a Chl–Zea heterodimer. A third model for the triggering of quenching by PsbS has been advanced on the basis of the observation that any physiological conditions or chemical treatment that prevents dissociation of a pentameric complex, including Lhcb4 and Lhcb6, together with an LHCII trimer also blocks NPQ. Thus the unquenched conformation of Lhcb proteins is stabilized by their inclusion in this large complex, while its dissociation by PsbS would allow transition to the quenching state, also promoted by Zea binding . Indeed, mutations inducing constitutive dissociation of the pentameric complex (designated as ‘B4 complex’ from its order of migration in sucrose gradients) show formation of two-dimensional arrays of C2S2 particles in the centre of grana discs, whereas LHCII is segregated out towards grana margins [94,95]. The two triggering pathways and the reorganization of PSII–LHCII supercomplexes are outlined in Figure 3.
Model of the PsbS-dependent qE mechanism in the PSII supercomplex
Experimental Plant Biology: Why Not?!: 4th Conference of Polish Society of Experimental Plant Biology, an Independent Meeting held at Jagiellonian University, Krakow, Poland, 21–25 September 2009. Organized and Edited by Kazimierz Strzałka (Jagiellonian University, Krakow, Poland).
cyclic electron transport
linear electron transport
light-harvesting antenna complex
protein gradient regulation
reactive oxygen species
This work was supported by the Italian Ministry of Research [grant numbers RBIP06CTBR_006 and 20073YHRLE_003].