Amino sugars: new inhibitors of zeaxanthin epoxidase, a violaxanthin cycle enzyme

The effect of three sugars and their amino derivatives on violaxanthin cycle enzymes activity was investigated in duckweed (Lemna trisulca), a model water-plant. No effect of sugars and amino sugars on violaxanthin de-epoxidase was observed independent of incubation time; however, epoxidation of zeaxanthin to violaxanthin was inhibited. The minimum amino sugar concentrations causing maximum inhibition of zeaxanthin epoxidation have been estimated. Amino sugars but not sugars caused more than a 50% inhibition of zeaxanthin epoxidation in duckweed after a 24h incubation when applied at a concentration of 0.5%. Incubation with amino sugars under a 6d photoperiod enhanced the inhibitory effect. Zeaxanthin epoxidation was completely inhibited under such conditions, whereas only a minor inhibitory effect was observed in sugar treated plants. The strong amino sugar inhibition of zeaxanthin epoxidase activity represents additional evidence for the creation of an unstable carotenoid carbocation in the molecular mechanism of epoxidation.     

A mathematical model describing kinetics of conversion of violaxanthin to zeaxanthin via intermediate antheraxanthin by the xanthophyll cycle enzyme violaxanthin de-epoxidase

The xanthophyll cycle is one of the mechanisms protecting the photosynthetic apparatus against the light energy excess. Its action is still not well understood on the molecular level.Our model makes it possible to follow independently the kinetics of the two de-epoxidation steps occurring in the xanthophyll cycle: the conversion of violaxanthin into antheraxanthin and the conversion of antheraxanthin into zeaxanthin. Using a simple form of the transition rates of these two conversions, we model the time evolution of the concentration pattern of violaxanthin, antheraxanthin and zeaxanthin during the de-epoxidation process. The model has been applied to describe the reactions of de-epoxidation in a system of liposome membranes composed of phosphatidylcholine and monogalactosyldiacylglycerol. Results obtained within the model fit very well with the experimental data. Values of the transition probabilities of the violaxanthin conversion into antheraxanthin and the antheraxanthin conversion into zeaxanthin calculated by means of the model indicate that the first stage of the de-epoxidation process is much slower than the second one.     

Regulation of violaxanthin de-epoxidase activity by pH and ascorbate concentration

The activity of violaxanthin de-epoxidase has been studied both in isolated thylakoids and after partial purification, as a function of pH and ascorbate concentration. We demonstrate that violaxanthin de-epoxidase has a K_m for ascorbate that is strongly dependent on pH, with values of 10, 2.5, 1.0 and 0.3 mM at pH 6.0, 5.5, 5.0 and 4.5, respectively. These values can be expressed as a single K_m±0.1±0.02 mM for the acid form of ascorbate. Release of the protein from the thylakoids by sonication was also found to be strongly pH dependent with a cooperativity of 4 with respect to protons and with an inflexion point at pH 6.7. These results can explain some of the discrepancies reported in the literature and provide a more consistent view of zeaxanthin formation in vivo.     

Ascorbate deficiency can limit violaxanthin de-epoxidase activity in vivo

As a response to high light, plants have evolved non-photochemical quenching (NPQ), mechanisms that lead to the dissipation of excess absorbed light energy as heat, thereby minimizing the formation of dangerous oxygen radicals. One component of NPQ is pH dependent and involves the formation of zeaxanthin from violaxanthin. The enzyme responsible for the conversion of violaxanthin to zeaxanthin is violaxanthin de-epoxidase, which is located in the thylakoid lumen, is activated by low pH, and has been shown to use ascorbate (vitamin C) as its reductant in vitro. To investigate the effect of low ascorbate levels on NPQ in vivo, we measured the induction of NPQ in a vitamin C-deficient mutant of Arabidopsis, vtc2-2. During exposure to high light (1,500 micromol photons m(-2) s(-1)), vtc2-2 plants initially grown in low light (150 micromol photons m(-2) s(-1)) showed lower NPQ than the wild type, but the same quantum efficiency of photosystem II. Crosses between vtc2-2 and Arabidopsis ecotype Columbia established that the ascorbate deficiency cosegregated with the NPQ phenotype. The conversion of violaxanthin to zeaxanthin induced by high light was slower in vtc2-2, and this conversion showed saturation below the wild-type level. Both the NPQ and the pigment phenotype of the mutant could be rescued by feeding ascorbate to leaves, establishing a direct link between ascorbate, zeaxanthin, and NPQ. These experiments suggest that ascorbate availability can limit violaxanthin de-epoxidase activity in vivo, leading to a lower NPQ. The results also demonstrate the interconnectedness of NPQ and antioxidants, both important protection mechanisms in plants.     

Chemical and mutational modification of histidines in violaxanthin de-epoxidase from Spinacia oleracea

The violaxanthin de-epoxidase (VDE) gene from spinach ( Spinacia oleracea ) was cloned, sequenced (GenBank AJ 250433), and expressed in Escherichia coli. The highest obtained conversion rate of violaxanthin was 86 nmol s⁻¹ per litre of growth medium, corresponding to an amount of active enzyme of 0.4 mg l⁻¹. Sequence comparison between VDE from different species were made and particular interest was focused on four highly conserved histidines (H121,124,167,173) and their possible involvement in enzymatic activity. Chemical modification of the histidines using DEPC or by site-directed mutations resulted in partial or total inactivation of the enzyme. The chemical modification could be reversed by hydroxylamine treatment, regenerating a large percentage of the original activity. The histidine residues, which are located in pairs close to each other, were pairwise substituted for either alanine or arginine. This resulted in one inactive mutant (H121,124R) and three mutants with very different activities and decreased binding of ascorbic acid, as reflected by an up to four-fold increase in K _m. A substitution of all four histidines for either alanine or arginine resulted in inactive enzymes. Based on these results it is suggested that the histidine residues are important for the activity of VDE.     

Overexpression of violaxanthin de-epoxidase: properties of C-terminal deletions on activity and pH-dependent lipid binding

Violaxanthin de-epoxidase (VDE) is localized in the thylakoid lumen and catalyzes the de-epoxidation of violaxanthin to form antheraxanthin and zeaxanthin. VDE is predicted to be a lipocalin protein with a central barrel structure flanked by a cysteine-rich N-terminal domain and a glutamate-rich C-terminal domain. A full-length Arabidopsis thaliana (L.) Heynh. VDE and deletion mutants of the N- and C-terminal regions were expressed in Escherichia coli and tobacco (Nicotiana tabacum L. cv. Xanthi) plants. High expression of VDE in E. coli was achieved after adding the argU gene that encodes the E. coli arginine AGA tRNA. However, the specific activity of VDE expressed in E. coli was low, possibly due to incorrect folding. Removal of just 4 amino acids from the N-terminal region abolished all VDE activity whereas 71 C-terminal amino acids could be removed without affecting activity. The difficulties with expression in E. coli were overcome by expressing the Arabidopsis VDE in tobacco. The transformed tobacco exhibited a 13- to 19-fold increase in VDE specific activity, indicating correct protein folding. These plants also demonstrated an increase in the initial rate of nonphotochemical quenching consistent with an increased initial rate of de-epoxidation. Deletion mutations of the C-terminal region suggest that this region is important for binding of VDE to the thylakoid membrane. Accordingly, in vitro lipid-micelle binding experiments identified a region of 12 amino acids that is potentially part of a membrane-binding domain. The transformed tobacco plants are the first reported example of plants with an increased level of VDE activity.     

Functional and structural characterization of domain truncated violaxanthin de‐epoxidase

Photosynthetic organisms need protection against excessive light. By using non‐photochemical quenching, where the excess light is converted into heat, the organism can survive at higher light intensities. This process is partly initiated by the formation of zeaxanthin, which is achieved by the de‐epoxidation of violaxanthin and antheraxanthin to zeaxanthin. This reaction is catalyzed by violaxanthin de‐epoxidase (VDE). VDE consists of three domains of which the central lipocalin‐like domain has been the most characterized. By truncating the domains surrounding the lipocalin‐like domain, we show that VDE activity is possible without the C‐terminal domain but not without the N‐terminal domain. The N‐terminal domain shows no VDE activity by itself but when separately expressed domains are mixed, VDE activity is possible. This shows that these domains can be folded separately and could therefore be studied separately. An increase of the hydrodynamic radius of wild‐type VDE was observed when pH was lowered toward the pH required for activity, consistent with a pH‐dependent oligomerization. The C‐terminally truncated VDE did not show such an oligomerization, was relatively more active at higher pH but did not alter the K_M for ascorbate. Circular dichroism measurements revealed the presence of α‐helical structure in both the N‐ and C‐terminal domains. By measuring the initial formation of the product, VDE was found to convert a large number of violaxanthin molecules to antheraxanthin before producing any zeaxanthin, favoring a model where violaxanthin is bound non‐symmetrically in VDE.     

The influence of phase transitions in phosphatidylethanolamine models on the activity of violaxanthin de-epoxidase

In the present study, the influence of the phospholipid phase state on the activity of the xanthophyll cycle enzyme violaxanthin de-epoxidase (VDE) was analyzed using different phosphatidylethanolamine species as model lipids. By using (31)P NMR spectroscopy, differential scanning calorimetry and temperature dependent enzyme assays, VDE activity could directly be related to the lipid structures the protein is associated with. Our results show that the gel (L beta) to liquid-crystalline (L alpha) phase transition in these single lipid component systems strongly enhances both the solubilization of the xanthophyll cycle pigment violaxanthin in the membrane and the activity of the VDE. This phase transition has a significantly stronger impact on VDE activity than the transition from the L alpha to the inverted hexagonal (HII) phase. Especially at higher temperatures we found increased VDE reaction rates in the presence of the L alpha phase compared to those in the presence of HII phase forming lipids. Our data furthermore imply that the HII phase is better suited to maintain high VDE activities at lower temperatures.     

Identification of key residues for pH dependent activation of violaxanthin de-epoxidase from Arabidopsis thaliana

Plants are often exposed to saturating light conditions, which can lead to oxidative stress. The carotenoid zeaxanthin, synthesized from violaxanthin by Violaxanthin De-Epoxidase (VDE) plays a major role in the protection from excess illumination. VDE activation is triggered by a pH reduction in the thylakoids lumen occurring under saturating light. In this work the mechanism of the VDE activation was investigated on a molecular level using multi conformer continuum electrostatic calculations, site directed mutagenesis and molecular dynamics. The pK(a) values of residues of the inactive VDE were determined to identify target residues that could be implicated in the activation. Five such target residues were investigated closer by site directed mutagenesis, whereas variants in four residues (D98, D117, H168 and D206) caused a reduction in enzymatic activity indicating a role in the activation of VDE while D86 mutants did not show any alteration. The analysis of the VDE sequence showed that the four putative activation residues are all conserved in plants but not in diatoms, explaining why VDE in these algae is already activated at higher pH. Molecular dynamics showed that the VDE structure was coherent at pH 7 with a low amount of water penetrating the hydrophobic barrel. Simulations carried out with the candidate residues locked into their protonated state showed instead an increased amount of water penetrating the barrel and the rupture of the H121-Y214 hydrogen bond at the end of the barrel, which is essential for VDE activation. These results suggest that VDE activation relies on a robust and redundant network, in which the four residues identified in this study play a major role.     

Inhibition of zeaxanthin epoxidase activity by cadmium ions in higher plants

The effect of cadmium and zinc ions on violaxanthin cycle enzymes, violaxanthin de-epoxidase and zeaxanthin epoxidase, has been investigated on selected plant species, as well as in vitro. About 50% inhibition of zeaxanthin epoxidase by cadmium ions was found for duckweed (Lemna trisulca) and tomato (Lycopersicon esculentum) leaves but for apricot (Prunus armeniaca) leaves no cadmium inhibition of the epoxidation reaction was observed. The cadmium inhibition of zeaxanthin epoxidase in tomato was abolished by zinc ions. Zinc ions alone did not affect the activity of neither of the enzymes of the violaxanthin cycle. This suggests that mechanism of cadmium inactivation of the enzyme relies on cadmium interaction with a cysteine residue of the protein, important for the enzyme activity. The target cysteine in tomato epoxidase could be the cysteine residue present in the most conservative part of the molecule which is not present in the apricot enzyme sequence. Neither stimulation nor inhibition of violaxanthin de-epoxidase by cadmium ions both in vivo and in vitro studies was detected. It confirms the proposed mechanism of zeaxanthin epoxidation inhibition by cadmium ions because the cysteine residue in the conservative motif of violaxathin de-epoxidase is not present.     

Engineering seed dormancy by the modification of zeaxanthin epoxidase gene expression

Abscisic acid (ABA) is a plant hormone synthesized during seed development that is involved in the induction of seed dormancy. Delayed germination due to seed dormancy allows long-term seed survival in soil but is generally undesirable in crop species. Freshly harvested seeds of wild-type Nicotiana plumbaginifolia plants exhibit a clear primary dormancy that results in delayed germination, the degree of primary dormancy being influenced by environmental culture conditions of the mother plant. In contrast, seeds, obtained either from ABA-deficient mutant aba2-s1 plants directly or aba2-s1 plants grafted onto wild-type plant stocks, exhibited rapid germination under all conditions irrespective of the mother plant culture conditions. The ABA biosynthesis gene ABA2 of N. plumbaginifolia, encoding zeaxanthin epoxidase, was placed under the control of the constitutive 35S promoter. Transgenic plants overexpressing ABA2 mRNA exhibited delayed germination and increased ABA levels in mature seeds. Expression of an antisense ABA2 mRNA, however, resulted in rapid seed germination and in a reduction of ABA abundance in transgenic seeds. It appears possible, therefore, that seed dormancy can be controlled in this Nicotiana model species by the manipulation of ABA levels.     

The influence of phase transitions in phosphatidylethanolamine models on the activity of violaxanthin de-epoxidase

In the present study, the influence of the phospholipid phase state on the activity of the xanthophyll cycle enzyme violaxanthin de-epoxidase (VDE) was analyzed using different phosphatidylethanolamine species as model lipids. By using ³¹P NMR spectroscopy, differential scanning calorimetry and temperature dependent enzyme assays, VDE activity could directly be related to the lipid structures the protein is associated with. Our results show that the gel (L_β) to liquid-crystalline (L_α) phase transition in these single lipid component systems strongly enhances both the solubilization of the xanthophyll cycle pigment violaxanthin in the membrane and the activity of the VDE. This phase transition has a significantly stronger impact on VDE activity than the transition from the L_α to the inverted hexagonal (H_II) phase. Especially at higher temperatures we found increased VDE reaction rates in the presence of the L_α phase compared to those in the presence of H_II phase forming lipids. Our data furthermore imply that the H_II phase is better suited to maintain high VDE activities at lower temperatures.     

Characterization of violaxanthin de-epoxidase purified in the presence of Tween 20: effects of dithiothreitol and pepstatin A

Violaxanthin de-epoxidase (VDE) was purified from thylakoid membranes of spinach by conventional column chromatography in the presence of Tween 20. The neutral detergent was necessary to prevent non-specific interaction of VDE with column resins. In anion-exchange chromatography on Mono Q, VDE appeared in two peaks. Both peaks exhibited a polypeptide of 41 kDa when fully reduced with 5 mM dithiothreitol. Re-chromatography of either peak gave rise to both peaks, suggesting that the two forms of VDE are interconvertible. VDE characteristically changed its electrophoretic mobility depending on the concentration of dithiothreitol with which the protein was treated. When non-reduced, it showed two polypeptides of 43 and 42 kDa. These polypeptides moved down to the position of 40 kDa, and then up to the position of 41 kDa, along with the increase in the dithiothreitol concentration from 0 to 2 mM. These findings suggest that VDE has more than one disulfide bond and takes multiple forms depending on the extent of the reduction. Studies with various types of protein-modifying reagent revealed that VDE is sensitive to pepstatin A, a specific inhibitor of aspartic protease. This finding suggests that the reaction center of VDE contains a reactive aspartic acid residue(s).     

Membrane barriers and Mehler-peroxidase reaction limit the ascorbate available for violaxanthin de-epoxidase activity in intact chloroplasts

The presence of an acidic lumen and the xanthophylls, zeaxanthin and antheraxanthin, are minimal requirements for induction of non-radiative dissipation of energy in the pigment bed of Photosystem II. We recently reported that ascorbate, which is required for formation for these xanthophylls, also can mediate the needed lumen acidity through the Mehler-peroxidase reaction [Neubauer and Yamamoto (1992) Plant Physiol 99: 1354–1361]. It is demonstrated that in non-CO₂-fixing intact chloroplasts and thylakoids of Lactuca sativa, L. c.v. Romaine, the ascorbate available to support de-epoxidase activity is influenced by membrane barriers and the ascorbate-consuming Mehler-peroxidase reaction. In intact chloroplasts, this results in biphasic kinetic behavior for light-induced de-epoxidation. The initial relatively high activity is due to ascorbate preloaded into the thylakoid before light-induction and the terminal low activity due to limiting ascorbate from the effects of chloroplast membranes barriers and a light-dependent process. A five-fold difference between the initial and final activities was observed for light-induced de-epoxidation in chloroplasts pre-incubated with 120 mM ascorbate for 40 min. The light-dependent activity is ascribed to the competitive use of ascorbic acid by ascorbate peroxidase in the Mehler-peroxidase reaction. Thus, stimulating ascorbic peroxidase with H₂O₂ transiently inhibited de-epoxidase activity and concomitantly increased photochemical quenching. Also, the effects inhibiting ascorbate peroxidase with KCN, and the K_M values for ascorbate peroxidase and violaxanthin de-epoxidase of 0.36 and 3.1 mM, respectively, support this conclusion. These results indicate that regulation of xanthophyll-dependent non-radiative energy dissipation in the pigment bed of Photosystem II is modulated not only by lumen acidification but also by ascorbate availability.Abbreviations APO ascorbate peroxidase - MP Mehler ascorbate-peroxidase - NIG nigericin - NPQ non-photochemical quenching - F_o dark fluorescence - F fluorescence at any time - F_M maximal fluorescence of the (dark) non-energized state - F_M maximal fluorescence of the energized state - q_P coefficient for photochemical fluorescence quenching - VDE violaxanthin de-epoxidase - k first-order rate constant for violaxanthin de-epoxidase activity     

Kinetics of violaxanthin de-epoxidation by violaxanthin de-epoxidase, a xanthophyll cycle enzyme, is regulated by membrane fluidity in model lipid bilayers.

This paper describes violaxanthin de-epoxidation in model lipid bilayers. Unilamellar egg yolk phosphatidylcholine (PtdCho) vesicles supplemented with monogalactosyldiacylglycerol were found to be a suitable system for studying this reaction. Such a system resembles more the native thylakoid membrane and offers better possibilities for studying kinetics and factors controlling de-epoxidation of violaxanthin than a system composed only ofmonogalactosyldiacylglycerol and is commonly used in xanthophyll cycle studies. The activity of violaxanthin de-epoxidase (VDE) strongly depended on the ratio of monogalactosyldiacylglycerol to PtdCho in liposomes. The mathematical model of violaxanthin de-epoxidation was applied to calculate the probability of violaxanthin to zeaxanthin conversion at different phases of de-epoxidation reactions. Measurements of deepoxidation rate and EPR-spin label study at different temperatures revealed that dynamic properties of the membrane are important factors that might control conversion of violaxanthin to antheraxanthin. A model of the molecular mechanism of violaxanthin de-epoxidation where the reversed hexagonal structures (mainly created by monogalactosyldiacylglycerol) are assumed to be required for violaxanthin conversion to zeaxanthin is proposed. The presence of monogalactosyldiacylglycerol reversed hexagonal phase was detected in the PtdCho/monogalactosyldiacylglycerol liposomes membrane by 31P-NMR studies. The availability of violaxanthin for de-epoxidation is a diffusion-dependent process controlled by membrane fluidity. The significance of the presented results for understanding themechanism of violaxanthin de-epoxidation in native thylakoid membranes is discussed.     

Antisense suppression of violaxanthin de-epoxidase in tobacco does not affect plant performance in controlled growth conditions

Violaxanthin de-epoxidase (VDE) catalyzes the de-epoxidation of violaxanthin to antheraxanthin and zeaxanthin in the xanthophyll cycle. Tobacco was transformed with an antisense VDE construct under control of the cauliflower mosaic virus 35S promoter to determine the effect of reduced levels of VDE on plant growth. Screening of 40 independent transformants revealed 18 antisense lines with reduced levels of VDE activity with two in particular (TAS32 and TAS39) having greater than 95% reduction in VDE activity. Northern analysis demonstrated that these transformants had greatly suppressed levels of VDE mRNA. De-epoxidation of violaxanthin was inhibited to such an extent that no zeaxanthin and only very low levels of antheraxanthin could be detected after exposure of leaves to high light (2000 μmol m⁻² s⁻¹ for 20 min) with no observable effect on levels of other carotenoids and chlorophyll. Non-photochemical quenching was greatly reduced in the antisense VDE tobacco, demonstrating that a significant level of the non-photochemical quenching in tobacco requires de-epoxidation of violaxanthin. Although the antisense plants demonstrated a greatly impaired de-epoxidation of violaxanthin, no effect on plant growth or photosynthetic rate was found when plants were grown at a photon flux density of 500 or 1000 μmol m⁻² s⁻¹ under controlled growth conditions as compared to wild-type tobacco. This revised version was published online in June 2006 with corrections to the Cover Date.     

pH-dependent reversible inhibition of violaxanthin de-epoxidase by pepstatin related to protonation-induced structural change of the enzyme

The redox enzyme violaxanthin de-epoxidase (VDE) was found to be sensitive to pepstatin, a specific inhibitor of aspartic protease. The inhibition was similar to that of aspartic protease in that it was reversible and accompanied by the protonation of the enzyme. Of the two peaks of VDE appearing on anion exchange chromatography, VDE-I predominated at pH 7.2. On lowering the pH of the chromatography, VDE-I decreased and VDE-II increased. Furthermore, re-chromatography of either peak yielded both peaks. These results suggest that VDE-I and VDE-II are interconvertible depending on pH, and thus, they represent the de-protonated and protonated forms of the enzyme, respectively. Presumably the protonation-induced structural change of the enzyme is responsible for the interaction with pepstatin, and also with substrate.