Euglena gracilis Klebs (strain Z) were grown in 100 ml of Cramer-Myers (CM) medium (Cramer and Myers, 1952) containing 0.1% (v/v) ethanol as a carbon source, at an initial cell concentration of 3 × 10³ cells ml⁻¹. The mixture was kept in a 300 ml conical flask under continuous light conditions (40 μmolm⁻² s⁻¹) using fluorescent lamps FL40SEX-N-HG (NEC lighting, Tokyo, Japan) at 25°C with rotary shaking (90 rpm) using a Double Shaker NR-30 (Taitec, Aichi, Japan). The light intensity was measured using a 3664 optical power meter (Hioki, Nagano, Japan).

Total RNA was isolated from E. gracilis using the Cica geneus RNA Prep Kit (for Plant) (Kanto Chemical, Tokyo, Japan), in accordance with the manufacturer’s instructions. First-strand cDNA was synthesized using a PrimeScript RT reagent kit with gDNA Eraser (Takara Bio, Shiga, Japan), in accordance with the manufacturer’s instructions. The open reading frame of EgLCY was amplified from first-strand cDNAs using the EgLCY-F and EgLCY-R primer set. All primers used in this study are listed in Supplementary Table 1. The amplified DNA fragments were ligated into the pMD20-T vector (Takara Bio) to check for PCR errors. EgLCY DNA fragments for the In-Fusion reaction were amplified from the obtained EgLCY clone using the EgLCY-InFusion-F and EgLCY-InFusion-R primer set. The expression vector pETDuet-1 DNA (Merck, NJ, United States) was digested with NcoI and reacted with amplified DNA fragments using the In-Fusion HD Cloning Kit (Takara Bio). The constructed plasmid was designated as pET-EgLCY. The pET-EgLCY was co-transformed into E. coli BL21(DE3) (New England Biolabs, MA, United States) with pACCRT-EIB (Misawa et al., 1995), which facilitates the accumulation of lycopene by heterologous expression of lycopene biosynthetic genes crtE, crtB, and phytoene desaturase crtI from Pantoea ananatis in host cells.

The lycopene-accumulating E. coli cells with pETDuet-1 empty vector or pET-EgLCY were grown in 3 ml of LB medium containing 50 μgml⁻¹ ampicillin and 30 μgml⁻¹ chloramphenicol. After overnight culture at 37°C, the cultures were inoculated into 2 × YT medium (1.6% tryptone, 1% yeast extract, and 0.5% sodium chloride) with the same concentrations of antibiotics and grown to an OD₆₀₀ of 0.5. IPTG was added to a concentration of 0.5 mM, and the cultures were incubated for 48 h at 21°C. The cultured medium was centrifuged, and E. coli cell precipitate was stored at −60°C.

Under dim light, carotenoids were extracted from E. coli cells with 1 ml of acetone. After centrifugation, the extracts were dried using a rotary evaporator. Carotenoid extracts were dissolved in 1 ml of ethyl acetate and filtered using a Columngard-LCR₁₃ (Merck) prior to HPLC analysis. Fifteen microliters of extracts were analyzed in an HPLC system equipped with a PEGASIL ODS SP100 column (6 mM × 150 mM, 5 μm particles, Senshu Scientific, Tokyo, Japan). The mobile phase was acetonitrile/methanol/tetrahydrofuran (58:35:7, v/v/v) at a flow rate of 0.8 mlmin⁻¹. Absorbance spectra (250–700 nm, 1.2-nm resolution) and retention times were recorded for 50 min using a Photodiode Array Detector SPD-M20A (Shimadzu, Kyoto, Japan).

An approximately 500-bp EgLCY partial cDNA template with the T7 RNA polymerase promoter was amplified using the EgLCY-RNAi-F and EgLCY-RNAi-R primer set (Supplementary Table 1). The double-stranded RNA (dsRNA) was synthesized from this PCR product using a MEGAscript RNAi Kit (Thermo Fisher Scientific, MA, United States), in accordance with the manufacturer’s instructions. The electroporatic introduction of EgLCY gene dsRNA into E. gracilis cells was performed following the method described in our previous study (Tamaki et al., 2019). Aliquots of the culture were collected daily, and cell density was determined using a microscope equipped with a plankton counting chamber. The cultured medium was centrifuged, and the E. gracilis cell precipitate was stored at −60°C.

Total RNA was isolated from six-day-old E. gracilis cells, into which dsRNA was introduced, and then, first-strand cDNA was synthesized as described above. Suppression of targeted gene expression was confirmed by semi-quantitative RT-PCR analysis using the EgLCY-sqRT-PCR-F and EgLCY-sqRT-PCR-R primer set (Supplementary Table 1). The E. gracilis Actin (EgActin) gene was chosen as a control, using the EgActin-sqRT-PCR-F and EgActin-sqRT-PCR-R primers (Supplementary Table 1). The EgAPX, EgGR, and EgDHAR genes were amplified using the EgAPX-sqRT-PCR-F and EgAPX-sqRT-PCR-R, EgGR-sqRT-PCR-F and EgGR-sqRT-PCR-R, and EgDHAR-sqRT-PCR-F and EgDHAR-sqRT-PCR-R, respectively. PCR amplification conditions consisted of 22–35cycles at 98°C for 15s, 55°C for 30s, and 68°C for 30s. The PCR products were analyzed by gel electrophoresis on a 1% agarose gel. The amount of each amplified DNA fragment was corrected using the EgActin PCR products.

Chlorophyll and carotenoid extraction and measurements were performed following the method described in our previous study (Tamaki et al., 2019).

Euglena gracilis (2 × 10⁷ cells) was suspended in 500 μl of ice-cold buffer (50 mM potassium phosphate buffer, pH 7.0) and disrupted by sonication. The cell lysate was centrifuged at 100,000 × g at 4°C for 60 min, and the supernatant was used for each enzyme assay. For the APX activity assay, buffer was replaced with 50 mM potassium phosphate, pH 7.0, containing 1 mM EDTA and 1 mM ascorbate.

The APX activity assay was performed following the method described in our previous study (Tamaki et al., 2014). GR activity was measured as the decrease in absorbance at 340 nm due to NADPH oxidation. The reaction mixture contained 1 mM EDTA, 0.2 mM NADPH, and 1 mM oxidized glutathione in 50 mM potassium phosphate buffer, pH 8.2, in a final volume of 1 ml. DHAR activity was measured as the increase in absorbance at 265 nm due to dehydroascorbate reduction. The reaction mixture contained 2.5 mM reduced glutathione and 0.2 mM dehydroascorbate in 50 mM potassium phosphate buffer, pH 7.0, in a final volume of 1 ml. MDAR activity was measured as the decrease in absorbance at 340 nm due to NADPH oxidation. The reaction mixture contained 1 mM ascorbate, 0.2 mM NADPH, and 0.2U ascorbate oxidase in 50 mM potassium phosphate buffer (pH 7.0) in a final volume of 1 ml. SOD activity was measured using a SOD assay kit-WST (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) in accordance with the manufacturer’s instructions.

To determine ascorbate, 5 × 10⁷ E. gracilis cells were suspended in 600 μl of ice-cold 2% metaphosphoric acid and disrupted by sonication. The cell lysate was centrifuged at 15,000 rpm at 4°C for 15 min. Ten microliters of extracts were analyzed in an HPLC system equipped with a Mightysil RP-18 GP analytical column (4.6 mM × 150 mM, 5 μm particles, Kanto Chemical, Tokyo, Japan) and guard column (4.6 mM × 5 mM, 5 μm particles, Kanto Chemical). The mobile phase was 1% metaphosphoric acid at a flow rate of 0.5 mlmin⁻¹. Ascorbate was detected at 245 nm using a Photodiode Array Detector SPD-M20A. The reduced ascorbate concentrations were calculated from the peak area of the HPLC chromatogram using the calibration curve of the authentic ascorbate standard. Total ascorbate was measured after reducing the oxidized form by incubating with 35 mM tris (2-carboxyethyl) phosphine hydrochloride at 4°C for 3 h. The contents of oxidized ascorbate were calculated as the difference between total and reduced ascorbate.

To determine glutathione, 5 × 10⁷ E. gracilis cells were suspended in 1 ml of ice-cold 0.2 M HCl and disrupted by sonication. The cell lysate was centrifuged at 15,000 rpm at 4°C for 10 min. One hundred microliters of 0.2 M potassium phosphate buffer, pH 5.6, was added to an aliquot of 0.5 ml of supernatant and vortexed. The pH of the extracts was adjusted to 4–5 by adding small volumes of 0.2 M NaOH. Total glutathione was determined using an enzymatic recycling assay based on GR. The reaction mixture contained 5 mM EDTA, 0.5 mM NADPH, 0.6 mM 5,5′ dithiobis-(2 nitrobenzoic acid), and cell extract in 100 mM potassium phosphate buffer, pH 7.5, in a final volume of 1 ml. The reaction was initiated by the addition of 1 unit of GR (Oriental Yeast, Tokyo, Japan), and the absorbance at 412 nm was monitored for 2.5 min. The total glutathione content was calculated from the increase in absorbance using the calibration curve of the authentic glutathione standard. The oxidized glutathione was selectively determined by assaying samples in which glutathione was masked by pretreatment with 2-vinylpyrimidine. Two microliters of 2-vinylpyrimidine was added to 200 μl of extracts and incubated at room temperature for 30 min after vortexing. After that, it was assayed as described above. The difference between the total glutathione and oxidized glutathione contents was presented as the reduced glutathione content.

Intracellular H₂O₂ levels in E. gracilis cells were detected by staining with the fluorescent reagent BES-H₂O₂-Ac (FUJIFILM Wako Pure Chemical, Osaka, Japan). BES-H₂O₂-Ac was dissolved in dimethyl sulfoxide at a concentration of 1 mM as a stock solution. To begin, 1 × 10⁶ E. gracilis cells were collected by centrifugation, and the medium was completely removed. Cells were resuspended in 500 μl of medium that contained 2% dimethyl sulfoxide and 20 μM BES-H₂O₂-Ac and incubated at room temperature for 30 min in the dark. Cells without BES-H₂O₂-Ac were incubated in medium that contained 2% dimethyl sulfoxide. After incubation, cells were centrifuged and washed twice with 1 ml of fresh medium. Cells were resuspended in 50 μl of fresh medium and immediately observed using a confocal laser microscope. The fluorescent images were obtained under excitation at 488 nm using a TCS SP8X (Leica, Wetzlar, Germany) equipped with a highly flexible pulsed white-light laser (WLL). For imaging in green wavelength regions, a 485 nm laser (21% of WLL) was employed. The BES-H₂O₂-Ac emission spectra was detected at 490–580 nm using a hybrid detector. Chlorophyll autofluorescence was eliminated by time-gated fluorescence imaging (gate-on time: 0.3–12.0nsec; Kodama, 2016). The fluorescence images were acquired at 100 Hz (100 lines s⁻¹) with a 3 × line average.

The cDNA sequence of the putative EgLCY gene (GenBank accession number: LC590220) was obtained via a tblastn search of E. gracilis transcriptome data (GenBank accession number: GDJR00000000.1; Yoshida et al., 2016) using Arabidopsis thaliana LCYB (GenBank accession number: U50739) as the query. The cDNA sequence generated by this search in E. gracilis had a spliced leader sequence, which is a characteristic short sequence that is normally transferred to the 5′ end of pre-mature mRNAs by trans-splicing in E. gracilis (Tessier et al., 1991). This indicates that recovered sequence is a full-length cDNA.

The EgLCY amino acid sequence shares 30–36% identity and 44–50% similarity with lycopene β-cyclases from A. thaliana, Solanum lycopersicum, Phaeodactylum tricornutum, and Synechococcus elongatus PCC 7942. The amino acid sequence alignment and phylogenetic tree of LCY proteins from E. gracilis and other photosynthetic organisms and bacteria are shown in Supplementary Figures 1, 2, respectively. The FAD/NADPH-binding site and cyclase motifs of LCY proteins are conserved in EgLCY and lycopene β-cyclases from other organisms (Cunningham et al., 1996). In contrast, EgLCY was phylogenetically separated from other lycopene cyclases (LCYB, LCYE, CrtL, CruA, CruP, and CrtY) from photosynthetic organisms and bacteria.

To evaluate the enzyme activity of EgLCY, the EgLCY cDNA was heterologously expressed using a T7 expression vector in E. coli cells. E. coli cells co-transformed with pACCRT-EIB, which expresses the lycopene synthetic genes crtE, crtB, and crtI from Pantoea ananatis, and a pETDuet-1 empty vector was red in color and produced large amounts of lycopene. These lycopene-accumulating E. coli cells became orange in color and accumulated β-carotene as the main carotenoid when the EgLCY gene was co-expressed (Figure 2). This indicates that EgLCY is a lycopene β-cyclase that converts lycopene to β-carotene.

To examine the physiological role of EgLCY in E. gracilis, EgLCY knockdown cells were generated. Knockdown experiments were performed by introducing dsRNA of the EgLCY gene into E. gracilis cells by electroporation. E. gracilis cells electroporated without dsRNA showed the same carotenoid synthesis as wild-type cells (Kato et al., 2017). Semi-quantitative RT-PCR analysis indicated that EgLCY transcript levels were suppressed in EgLCY dsRNA-treated cells grown both mixotrophically (Figure 3A) and autotrophically (Supplementary Figure 3A) for 6 d. In contrast to the cells treated without dsRNA, both the EgLCY dsRNA-treated cells and culture medium were colorless. Moreover, the colorless EgLCY dsRNA-treated cells tended to adopt a hypertrophic morphology (Figure 3B). The cell densities of EgLCY dsRNA-treated cells grown mixotrophically decreased to 29% of cells treated without dsRNA after 6 d and 43% after 10 d (Figure 3C). The cell density of EgLCY dsRNA-treated cells grown autotrophically for 6 d also decreased significantly, whereas there was no significant difference after 10 days of growth (Supplementary Figure 3B). This may be due to the inability of EgLCY dsRNA-treated colorless cells to grow autotrophically and exclusive growth of a few green cells to which EgLCY dsRNA is not successfully introduced. These results suggest that EgLCY is functional in vivo and physiologically important for cell growth, regardless of the presence of a carbon source. Mixotrophic EgLCY dsRNA-treated cells tended to maintain longer phenotypic changes, and thus were used in subsequent analyses.

Next, we examined the chlorophyll and carotenoid content in EgLCY dsRNA-treated cells. The total chlorophyll content of the EgLCY dsRNA-treated cells decreased to 16% of the cells treated without dsRNA, in accordance with their appearance (Figure 4A). Carotenoid content and composition were analyzed using HPLC. In cells treated without dsRNA, diadinoxanthin, diatoxanthin, neoxanthin, and β-carotene were detected, whereas EgLCY dsRNA-treated cells contained additional lycopene, which is a substrate of EgLCY (Figure 5), further supporting the functional activity of EgLCY in E. gracilis cells. The total carotenoid content in EgLCY dsRNA-treated cells decreased to 19% of the cells treated without dsRNA (Figure 4B). β-carotene, neoxanthin, diadinoxanthin, and diatoxanthin were significantly decreased in EgLCY dsRNA-treated cells compared to cells treated without dsRNA, and lycopene accounted for 12% of the total carotenoid content in EgLCY dsRNA-treated cells (Table 1). These results indicate that EgLCY is essential for carotenoid synthesis in E. gracilis, and deficiencies in β-carotene and downstream carotenoid species, which are abundant and physiologically functional, suppress chlorophyll synthesis, resulting in colorless cells and probable loss of photosynthetic activity.

To investigate the possible relationship between carotenoid synthesis and the ascorbate-glutathione cycle, the enzyme activities of APX, GR, DHAR, MDAR, and SOD in EgLCY dsRNA-treated cells were measured. The APX and GR activities of EgLCY dsRNA-treated cells were decreased to 48 and 71% of cells treated without dsRNA, respectively (Figures 6A,B). On the other hand, activity of DHAR and MDAR in EgLCY dsRNA-treated cells increased by 77 and 45%, respectively, compared to cells treated without dsRNA (Figures 6C,D). Moreover, the SOD activity of EgLCY dsRNA-treated cells increased 7.9 times compared to cells treated without dsRNA (Figure 6E). These results suggest the physiological importance of carotenoid synthesis for the upregulation of APX, GR, and SOD activities in E. gracilis.

To examine the effects on the gene expression of the ascorbate-glutathione cycle enzymes, semi-quantitative RT-PCR analysis was performed. Since the cDNA sequence of the E. gracilis MDAR gene is unknown (Ishikawa et al., 2017), EgMDAR was excluded from this analysis. As shown in Supplementary Figure 4, there was no significant difference in the transcript levels of EgAPX, EgGR, and EgDHAR genes between each treatment. This suggests that the ascorbate-glutathione cycle genes are post-transcriptionally regulated when β-carotene synthesis is inhibited.

It is presumed that significant changes in the activities of ascorbate-glutathione cycle enzymes are caused by the cellular levels of ascorbate and glutathione; therefore, the cellular content of these antioxidants was measured. The total ascorbate content of EgLCY dsRNA-treated cells was 55% higher than the cells treated without dsRNA (Figure 7A), while the total glutathione content of EgLCY dsRNA-treated cells was 62% less than the cells treated without dsRNA (Figure 7B). There was no significant difference in the redox ratios of both antioxidants between each treatment. This result clearly indicates that the size of the glutathione pool in EgLCY dsRNA-treated cells decreases with decreased GR activity.

To examine whether the decreases in APX activity and the size of the glutathione pool and increase in SOD activity due to EgLCY suppression affect oxidative stress status, the level of cellular H₂O₂ was visualized by staining with an H₂O₂-specific fluorescent reagent (BES-H₂O₂-Ac). As shown in Figure 8A, no fluorescence was observed in both treatments without BES-H₂O₂-Ac. When stained with BES-H₂O₂-Ac, fluorescence was detected in very few cells treated without dsRNA, whereas it was more apparent in EgLCY dsRNA-treated cells. A criterion for staining was defined based on the fluorescence intensity of each cell, as shown in Supplementary Figure 5, and this was used to calculate staining rates of each treatment. The staining rate of EgLCY dsRNA-treated cells was 27% and higher than that of cells treated without dsRNA (4%; Figure 8B), which suggests that EgLCY suppression causes a reduction in H₂O₂ scavenging capacity and results in the accumulation of oxidative stress in E. gracilis.

E. gracilis is a euglenophyte species that has evolved independently of other photosynthetic organisms, although its strict phylogeny has yet to be fully elucidated. Genetic analysis suggests that two endosymbiotic events occurred; in the first, the euglenophyte acquired red lineage genes from Chromalveolata-like prey algae; in the second, green lineage genes from prasinophyte-like green algae were acquired (Maruyama et al., 2011). Since diadinoxanthin, the major carotenoid species of E. gracilis, is present in algae classified in Chromalveolata, such as diatoms and haptophytes, and not in green algae (reviewed in Tamaki et al., 2021), E. gracilis is believed to have acquired a series of carotenoid synthetic genes, including LCY, in the first endosymbiotic event. However, our analysis indicates that EgLCY is phylogenetically distinct from both diatom and green algae LCYs (Supplementary Figure 2). This may be due to the independent evolution of E. gracilis, as the Euglenozoa phylum diverged in the early stages of eukaryotic evolution (Ahmadinejad et al., 2007). The results of the phylogenetic analysis of β-carotene hydroxylase (CYP97H1; Tamaki et al., 2019) and ζ-carotene isomerase (Z-ISO; Sugiyama et al., 2020) correspond to those of EgLCY, suggesting that independent evolution of carotenoid synthetic genes in E. gracilis may be common.

An in vivo enzyme assay demonstrated that EgLCY catalyzes the cyclization of lycopene to β-carotene (Figure 2). RNAi-mediated suppression of EgLCY in E. gracilis caused bleaching, cell growth inhibition, and carotenoid and chlorophyll deficiencies (Figures 3, 4). Knockdown of other carotenoid synthetic genes, namely, EgcrtB and EgCYP97H1, also showed similar phenotypic changes (Kato et al., 2017; Tamaki et al., 2019), suggesting that EgLCY is the sole lycopene cyclase for β-carotene synthesis and for subsequent compounds, such as neoxanthin, diadinoxanthin, and diatoxanthin. EgLCY dsRNA-treated cells accumulated a small amount of lycopene, which is the substrate for EgLCY, and total carotenoid content in these cells was markedly decreased (Figures 4B, 5B and Table 1). As reported in our study of EgCYP97H1-suppressed cells, E. gracilis appears to regulate the whole of carotenoid synthesis in a manner that depends on the abundance of downstream carotenoid products, a regulatory system which differs from that of plants (Kim et al., 2009; Tamaki et al., 2019).

Plants and green algae have lycopene ε-cyclase, which catalyzes formation of the ε-ring of α-carotene and is necessary for the synthesis of abundant compound lutein (Supplementary Figure 2) (Cunningham et al., 1996). The lycopene ε-cyclase gene in E. gracilis was not identified by a homology search. This result supports the observation that E. gracilis lacks α-carotene and its derivatives (Figure 5) and that EgLCY produces β-carotene (Figure 2). The absence of lycopene ε-cyclase is consistent with algae classified in Chromalveolata, including diatoms that produce diadinoxanthin (Dambek et al., 2012).

Carotenoids play an important role in photoprotection by receiving excess energy from photoexcited chlorophyll and scavenging singlet oxygen (reviewed in Tamaki et al., 2021). The inhibited synthesis of β-carotene and downstream carotenoid species by EgLCY suppression would accumulate excess photoexcited chlorophyll and subsequent singlet oxygen and superoxide radical, and cause damage to photosynthetic apparatus and chloroplast dedifferentiation (Cazzaniga et al., 2012; Santabarbara et al., 2013). Carotenoid deficiency-induced impairment of chloroplast development and inhibited chlorophyll synthesis was agreed with previous study using E. gracilis treated with norflurazon (Madhusudhan et al., 2003) and knocked down cells of carotenoid synthetic genes EgcrtB (Kato et al., 2017) and EgCYP97H1 (Tamaki et al., 2019). In the EgLCY dsRNA-treated cells, increase in the SOD activity rapidly converts superoxide radical to H₂O₂ (Figure 6E), which then diffuses into the cytosol from chloroplasts (Ishikawa et al., 1993). E. gracilis APX enzyme is localized in the cytosol and is post-transcriptionally photo-induced by retrograde signal from chloroplasts (Madhusudhan et al., 2003; Ishikawa et al., 2010). Chloroplast dedifferentiation in EgLCY dsRNA-treated cells would stagnates retrograde signaling and inhibits APX photoinduction. The glutathione pool size is speculated to be regulated by a similar mechanism based on the fact that the glutathione pool size is decreased in chloroplast-dedifferentiated SM-ZK strain of E. gracilis in which retrograde signals from plastids are suppressed (Shigeoka et al., 1987b). As a consequence of them, it can be explained that EgLCY suppression causes downregulation of APX activity and decreased glutathione pool size (Figures 6A,B, 7B). In E. gracilis lacking catalase, since APX is a major H₂O₂ scavenging enzyme, downregulation of APX activity due to the inhibited synthesis of β-carotene and downstream carotenoid species excessively increases cellular H₂O₂ level (Figure 8). Although EgLCY suppression increased ascorbate pool size (Figures 6C,D, 7A), it is speculated that excess H₂O₂ accumulation cannot be suppressed because the ROS scavenging capacity of ascorbate itself is extremely lower than that of APX (Smirnoff and Arnaud, 2019).

Both carotenoids and the ascorbate-glutathione cycle are physiologically essential for ROS scavenging in photosynthetic organisms, including A. thaliana and Chlamydomonas reinhardtii; however, the understanding of the possible relationship between carotenoid synthesis and the ascorbate-glutathione cycle components in other photosynthetic organisms is currently limited. Ascorbate-deficient A. thaliana mutants showed decreased ascorbate content and poor growth but no difference in carotenoid and chlorophyll content compared to WT (Plumb et al., 2018). Norflurazon is an herbicide that inhibits phytoene desaturase, resulting in the depletion of carotenoids, and is often used in carotenoid research in photosynthetic organisms. Treatment of A. thaliana and cucumber (Cucumis sativus) with norflurazon had little effect on the activities of APX and GR (Jung et al., 2000; Jung, 2004). Moreover, norflurazon treatment of A. thaliana and rice did not affect ROS (superoxide radical, H₂O₂, and singlet oxygen) accumulation and lipid peroxidation (Kim and Apel, 2013; Park and Jung, 2018). In the green alga C. reinhardtii, APX has been reported to be important for resistance to photooxidative stress, in cooperation with ascorbate recycling by DHAR and MDAR (Kuo et al., 2020). Carotenoid-deficient C. reinhardtii generated by either a phytoene synthase mutation (lts1-204) or norflurazon treatment showed high levels of oxidative stress; however, the effect of carotenoid deficiency on the ascorbate-glutathione cycle enzymes was not examined (Pérez-Pérez et al., 2012). Our study clearly explained that EgLCY-mediated synthesis of β-carotene and downstream carotenoid species caused upregulated APX activity, increased glutathione pool, and decreased cellular H₂O₂ accumulation (Figures 6–8). This suggests a possible relationship between carotenoid synthesis and the ascorbate-glutathione cycle for ROS scavenging in E. gracilis.

In many photosynthetic organisms, ascorbate and glutathione share similar functions and act cooperatively as components of the ascorbate-glutathione cycle. Previous studies reported that ascorbate and glutathione were synthesized in E. gracilis by exposure to blue light (wavelengths from 380 nm to 440 nm) and are not associated with photosynthesis under mixotrophic conditions (Shigeoka et al., 1979, 1987c). Therefore, it is expected that these antioxidants can be synthesized synchronously. In the green alga Dunaliella salina, both cold and high light treatment decreased amounts of β-carotene and chlorophyll, while ascorbate and glutathione increased, suggesting that the oxidative stress caused by pigment deficiencies was alleviated by the induction of both ascorbate and glutathione synthesis (Haghjou et al., 2009). If E. gracilis has a similar regulatory mechanism for ascorbate and glutathione synthesis, the content of these compounds would increase when carotenoid and chlorophyll synthesis are suppressed. However, this study unexpectedly revealed that ascorbate and glutathione synthesis, and their recycling enzyme activities, were different in EgLCY dsRNA-treated cells (Figures 6B–D, 7). Promotion of ascorbate synthesis in EgLCY dsRNA-treated cells is likely to be one of the mechanisms responsible for the toxic effects of H₂O₂ accumulation to avoid stagnation of the entire ascorbate-glutathione cycle. In contrast, we also observed that glutathione synthesis was triggered by carotenoid accumulation, which suggests that ascorbate and glutathione synthesis are differentially regulated in E. gracilis under oxidative stress conditions. E. gracilis may initially synthesize ascorbate rather than glutathione to cope with accumulated oxidative stress.