In photosynthetic organisms, carotenoids participate in light harvesting within the reaction centers and the light harvesting complexes (LHCs), the protection of the photosynthetic apparatus via the direct quenching of chlorophyll reactive intermediates and/or singlet oxygen molecules, the dissipation of absorbed energy beyond that required for photosynthesis [1,2], and the structural maintenance of photosynthetic complexes [3]. Therefore, carotenoids participate in keeping a balance between capture, use of light and cell protection of excess excitation energy. In particular the xanthophylls zeaxanthin and lutein, whose formation is enhanced under high light conditions, help in the efficient transition of LHCs of PSII to a conformation that favors the thermal dissipation of parts of the excess of excitation energy. This energy dissipation, measured as non-photochemical quenching of the chlorophyll fluorescence (NPQ) prevents over-reduction of the reaction centers and must be strictly controlled to avoid photosynthetic efficiency loss [4,5].

The photo-protective functions of the de-epoxidazed carotenoid zeaxanthin have been intensively studied [6], however, it was not until 1989 when Demming-Adams and co-workers [7] first proposed a now widely accepted role linking zeaxanthin formation with photoprotection within the violaxanthin cycle. The dissipation of excess excitation energy mediated by zeaxanthin in this cycle is a key process in the photosynthetic systems of plants and algae [8,9]. A similar cycle, known as the diadinoxanthin cycle, has also been identified in certain algal groups with cycling occurring between the two xanthophylls diadinoxanthin and diatoxanthin, which only comprises one de-epoxidation step [10,11,12,13]. However, in these algae this cycle is the main one and the role of the violaxanthin cycle is mainly to produce intermediates for the biosynthesis of the diadinoxanthin cycle pigments.

The violaxanthin cycle is found in plants and the green (chlorophyta) and brown (phaeophyceae) algae. In this cycle, when the irradiance is higher than that required for photosynthesis in the chloroplasts of plants and algae, a reversible violaxanthin de-epoxidation reaction occurs to form the intermediate antheraxanthin, subsequently resulting in the accumulation of zeaxanthin in a second de-epoxidation reaction. The extent of the de-epoxidation depends on two factors at least, namely, the pigment pool size and the fraction of the violaxanthin pool that is available for de-epoxidation [14]. When the absorbed irradiance is lower than that required for the saturation of the photosynthetic apparatus, zeaxanthin is converted back into violaxanthin. This reaction occurs on the stromal side of the chloroplast thylakoid membrane, and it is relatively slower than the de-epoxidation reaction [15,16]. The de-epoxidation sequence of the violaxanthin cycle is catalyzed by the enzyme violaxanthin de-epoxidase (VDE), and the epoxidation from zeaxanthin into violaxanthin is carried out by the enzyme zeaxanthin epoxidase (ZEP).

ZEP has a specificity for carotenoids with a 3-hydroxy-β-hexenyl ring and requires the presence of NADPH and FAD as cofactors, ferredoxin and ferredoxin-like reducing agents are also required for its activity [17]. The first isolation of this gene was performed in a transposon-tagged Nicotiana plumbaginifolia ABA-deficient mutant called aba2 [18,19] and many genetic studies on this cycle have been performed in plants, such as Arabidopsis thaliana in which different aba mutants have been characterized, providing an opportunity to study the physiological roles of epoxy-carotenoids in photosynthesis [20], as well as the npq mutants in the green alga Chlamydomonas reinhardtii [21]. This unicellular green alga has been used as a good alternative molecular system to plants in order to study photosynthesis and other metabolic processes [22].

In 1997, Niyogi and co-workers [21] generated a series of violaxanthin cycle mutants in an Arg^- auxotroph Chlamydomonas reinhardtii strain. These mutants were obtained by insertional mutagenesis using a linearized vector containing the wild type argininosuccinate lyase gene (arg7). These mutants exhibited an aberrant NPQ, being defective in two different NPQ phases, either being defective in the quick-turn-over first phase (npq4), or in the slower second phase (npq1) or in both of them (npq2 and npq11). Npq2 had lost the zeaxanthin epoxidase activity, since it exhibited a violaxanthin cycle pool of entirely zeaxanthin even after dark adaptation. After examining the co-segregation of the insertional mutagen (ARG7 DNA) and the npq aberrant phenotype in the npq2, it was concluded that this mutation was closely linked to a single insertion locus of the ARG7 vector.

The chlorophyta Chlorella zofingiensis is considered as a model organism to study the regulation of the carotenoid biosynthetic pathway since it produces both the primary carotenoid lutein as well as the secondary carotenoid astaxanthin [23]. Several carotenogenic genes have been isolated and characterized in this microalga, such as β-carotene oxygenase (bkt) [24], phytoene desaturase (pds) [25], carotene β-hydroxylase (chyB) [26], lycopene β-cyclase (lcyB) [27] and phytoene synthase (psy) [28]; however, genes involved in the violaxanthin cycle have not been isolated yet.

In this paper, we report the isolation and characterization of the zeaxanthin epoxidase gene involved in the violaxanthin cycle in Chlorella zofingiensis. The transcriptional regulation of this gene in response to light and nitrogen, as well as its functionality by heterologous transformation of Chlamydomonas reinhardtii npq2 mutant, which lacks ZEP activity, has also been performed.

Although previous work concerning zeaxanthin epoxidase has essentially focused on its role in ABA biosynthesis during drought stress and seed development in plants [19,20,30], its relationship with photosynthesis and the carotenoids synthesis pathway using mutants lacking the zeaxanthin epoxidase activity has also been reported both in plants and algae [21,31,32,33,34]. The zep gene cDNA has been isolated from higher plants such as Nicotiana plumbaginifolia (tobacco), Capsicum annuum (pepper), Lycopersicon esculentum (tomato) and Arabidopsis thaliana [15,18,34] and from some algae such as the well-known Chlamydomonas reinhardtii [35]. In this work, we describe the isolation and characterization of the zep gene from the green alga Chlorella zofingiensis. This new gene showed six exons and five introns, and the ORF of the cDNA encoded a hypothetical protein of 596 amino acids (Figure 1). The length of the putative protein was similar to those described for microalgae, being slightly shorter than plant enzymes, according to alignments performed with the Blocks program (data not shown). Southern blot analysis has demonstrated that in C. zofingiensis, as in other green algae such as C. reinhardtii [35] and Chlorella variabilis [36], as well as in angiosperms, only a single gene encoding ZEP is present. However, it has been recently described that other algae classes such as the diatoms P. tricornutum and T. pseudonana contain three and two copies, respectively, of the zep gene, which could be a reflection of the presence of both cycles, violaxanthin cycle and diadinoxanthin cycle in these organisms. The diadinoxanthin cycle is generally found in heterokons but it is absent in plants and chlorophytes [37]. On the other hand, the zep gene has not been found in the genomes of cyanobacteria or the red algae, such as C. merolae and G. sulphuraria [38].

The first zep gene isolated in Nicotiana plumbaginifolia [18], called aba2, encoded a chloroplast-imported flavin monooxygenase (FMO) containing a FAD-binding domain. The amino acid sequences of the pepper [15], tomato [34] and tobacco ZEP were similar to that of a flavoprotein monooxygenase (salicylate-1-monooxygenase) and those of a number of other bacterial hydroxylases [39]. Therefore, ZEP enzymes are considered FAD-dependent monooxygenases. The amino acid sequence analysis of the predicted protein encoded by the new gene isolated from C. zofingiensis revealed two monooxygenase FAD-binding domains [37,40]. In addition, the predicted CzZEP sequence showed four putative transmembrane domains that could play a role in the localization of this enzyme on the stromal side of the chloroplast thylakoid membrane, and a transit peptide which would allow CzZEP to travel from the cytoplasm, where the polypeptide is synthesized, to its final location in the chloroplast as in the case of other carotenogenic enzymes in this alga which are encoded by nuclear genes [23,27].

Different environmental conditions can show different regulations of the carotenoid synthesis pathway. As an example, either low light intensities or darkness stimulate the backward reaction of the violaxanthin cycle resulting in the conversion of zeaxanthin back to violaxanthin in Arabidopsis and C. reinhardtii [41,42]. In photoautotrophically grown C. zofingiensis cultures, it is known that the combination of both high irradiance and nitrogen starvation causes changes in the cellular content of several carotenoids [43,44,45]. Recently, it has been shown that high irradiance in this alga up-regulates the pds, chyB, and bkt genes, significantly enhancing astaxanthin synthesis [25,26]. In addition, the nitrogen deprivation per se up-regulated the lcyB gene, but did not trigger astaxanthin biosynthesis [27]. We have studied the regulation of the Czzep gene by nitrogen and irradiance, determining the mRNA levels as well as the cellular content of violaxanthin and zeaxanthin (Figure 4). The transcript levels of Czzep did not increase until 24 h under low irradiance conditions, coming from a dark period, regardless of the nitrogen availability, without any accumulation of zeaxanthin and with the violaxanthin content remaining virtually constant. Therefore, it is possible that nitrogen deprivation itself could not affect the level of zep transcripts. Under high irradiance and as previously described, the light activated the de-epoxidation of violaxanthin, registering zeaxanthin production after 10 h of exposure and the decrease of the violaxanthin content, the effect being more pronounced under nitrogen deprivation (about two-fold zeaxanthin level of the nitrogen complete culture), showing a possible additive effect of those two stress conditions. However, mRNA levels of Czzep surprisingly increased after 10 h, reaching the highest values under nitrogen starvation conditions. For both high light and high light combined with nitrogen starvation stresses, the increase in zeaxanthin content was lower than the decrease in violaxanthin content. This could be due to the presence of antheraxanthin as well as to the astaxanthin accumulation, since in C. zofingiensis it has been reported that astaxanthin is mainly synthesized from zeaxanthin instead of β-carotene as in H. pluvialis [46].

The high violaxanthin de-epoxidase activity under high irradiance has been described previously as being due to a result of the presence of a transmembrane proton gradient during periods of high irradiance [14,47,48]. Until now, the regulation of the ZEP is still not well understood. Some authors have proposed that the zeaxanthin epoxidase activity seems to be constitutive [49], as this is not regulated by the stromal pH. Since ZEP requires molecular oxygen as a second substrate and NADPH as a cofactor, both factors might principally control or limit ZEP activity [49]. The predicted protein encoded by the new gene isolated from C. zofingiensis showed a high homology degree with ZEP of plants and algae (Figure 3). This could be sufficient to consider this new gene as a zeaxanthin epoxidase, but the functional analysis has confirmed this hypothesis. By complementation in E. coli using the pAC25crtx plasmid, which [50] contains the carotenogenic genes responsible for the formation synthesis of zeaxanthin, together with the Czzep gene cloned into pQE80-L expression vector, no conversion of zeaxanthin into violaxanthin or the intermediate antheraxanthin was detected. Molecular oxygen (O₂), NADPH and reduced ferredoxin are known to be co-substrates for the epoxidation reaction [15,17,51]. The prokaryotic host E. coli cannot provide reduced ferredoxin and this seems to be the reason why some zep genes do not work in this bacterium, as it was also the case of Nicotiana plumbaginifolia [18,52]. In Chlamydomonas reinhardtii, ZEP functionality has been studied by using forward genetics. The Chlamydomonas npq2 mutant was isolated because it exhibited an unusual violaxanthin cycle pool, consisting entirely of zeaxanthin in all conditions. In contrast to other npq mutations, npq2 phenotype co-segregated with arginine producer phenotype, demonstrating that the mutation affects a single gene encoding the zeaxanthin epoxidase enzyme [21]. As the CzZEP functionality could not be performed in E. coli, the Czzep gene was inserted into a Chlamydomonas expression vector to be heterologously expressed in vivo in the Chlamydomonas npq2 mutant. The transformants obtained were able to convert back zeaxanthin into violaxanthin, which confirmed the functionality of the Czzep gene isolated.

At the moment, Chlamydomonas ZEP is the only zeaxanthin epoxidase enzyme whose functionality has been reported in algae [21,53]. On the other hand, the complementation of the mutation that npq2 carried by Czzep, supports the idea that Chlamydomonas can be an efficient host for expressing exogenous genes.

Several studies about over-expression of the zeaxanthin epoxidase gene have been reported in plants [54,55]. Transgenic lines in tomatoes shown altered de-epoxidation kinetics of the violaxanthin cycle especially under stress conditions, but nothing was described regarding the concentration of the rest of carotenoids found in their profile. In the case of the CzZEP transformants of npq2 mutant obtained we found higher levels of β-carotene and lutein reaching between three to six times the levels of the npq2 control cells. Moreover, these increments in carotenoids also correspond to increased Czzep mRNA levels of the transformants analyzed. Thus, the npq2-4 transformant, which exhibited the highest expression pattern, also showed the highest carotenoids levels (Figure 6a) as well as a total conversion from zeaxanthin into violaxanthin (Figure 6b). Chlamydomonas transformants over-expressing the phytoene synthase gene of C. zofingiensis isolated also by us, exhibited a carotenoids content up to 2.2-fold the level of the control cultures [23]. Phytoene synthase catalyzes the first step of the carotenoid biosynthetic pathway and it is considered as a rate-limiting key enzyme in this pathway. Since the Czzep transformants isolated in this work showed higher carotenoids levels than parental ones, the violaxanthin cycle can be postulated as a key step to be considered for biotechnological applications and genetic manipulation.

It has been shown that the violaxanthin cycle pigments are bound to light-harvesting complexes of the two photosystems [56,57]. In fact, ZEP has been proposed to be located close to the light harvesting complex of PSII (LHCII) to enable the binding of violaxanthin and the intermediate antheraxanthin to the antenna complex [58]. The role of zeaxanthin in the chlorophyll fluorescence quenching has also been largely discussed, concluding that Lhcb proteins replace violaxanthin by zeaxanthin in the L2 site which induces the quenching of the chlorophyll fluorescence by increasing the relative amplitude of the short fluorescence lifetime component found in these Lhcb proteins. This correlates with energy dissipation and regulation of the energy transfer to the PSI reaction centers [59,60]. Kalituho and co-workers in 2007 [61] reported a reduced electron transport in Arabidopsis npq2 mutant due to sustained energy dissipation by the high levels of zeaxanthin, and how it might reflect the less efficient utilization of absorbed light energy measured as the maximum quantum yield (Fv/Fm). In the same way, Chlamydomoans npq2 mutant Fv/Fm values were 25% lower than those in the wild type strain. However, the obtained transformants showed restored Fv/Fm values, recovering the levels of the Chlamydomonas wild type cells almost completely (Table 2). These results were also correlated to those obtained from the transformants, some of which showed high expression levels and total conversion from zeaxanthin into violaxanthin. As an example, the npq2-4 transformant exhibited a 100% restored Fv/Fm, in addition to a higher concentration of carotenoids and mRNA levels of Czzep.

Until now, Chlamydomonas ZEP was the only zeaxanthin epoxidase enzyme which functionality was reported in algae. Czzep isolation and characterization was performed in this study. The recovered transformants able to convert back zeaxanthin into violaxanthin confirm the functionality of the Czzep gene isolated and support the idea that Chlamydomonas can be an efficient host for checking the functionality of exogenous genes. Our data on the characterization and regulation of this gene by light and nitrogen could contribute to the understanding of the regulatory mechanism of biosynthesis of carotenoids at the molecular level, which can be very useful for the optimization of the physiological conditions for higher carotenoid production by C. zofingiensis.