Impairment of starch biosynthesis results in elevated oxidative stress and autophagy activity in Chlamydomonas reinhardtii

Autophagy is a self-degradation system wherein cellular materials are recycled. Although autophagy has been extensively studied in yeast and mammalian systems, integrated stress responses in microalgae remain poorly understood. Accordingly, we carried out a comparative study on the oxidative stress responses of Chlamydomonas reinhardtii wild-type and a starchless (sta6) mutant previously shown to accumulate high lipid content under adverse conditions. To our surprise, the sta6 mutant exhibited significantly higher levels of lipid peroxidation in the same growth conditions compared to controls. The sta6 mutant was more sensitive to oxidative stress induced by H2O2, whereas the wild-type was relatively more resistant. In addition, significantly up-regulated autophagy-related factors including ATG1, ATG101, and ATG8 were maintained in the sta6 mutant regardless of nitrogen availability. Also, the sta6 mutant exhibited relatively higher ATG8 protein level compared to wild-type under non-stress condition, and quickly reached a saturation point of autophagy when H2O2 was applied. Our results indicate that, in addition to the impact of carbon allocation, the increased lipid phenotype of the sta6 mutant may result from alterations in the cellular oxidative state, which in turn activates autophagy to clean up oxidatively damaged components and fuel lipid production.

With the growing demand for environmentally friendly sustainable energy alongside an increasing world population, the subject of algal biomass production has attracted much attention since microalgae can convert atmospheric CO 2 and water pollutants into diverse value-added products including biofuel and other bioactive compounds 1,2 . Among various compounds, algal lipid, which is mainly composed of triacylglycerol (TAG), is considered a promising resource for biodiesel production 2 . The extracted algal lipid contents can be converted into biodiesel and glycerol via a transesterification reaction with alcohols and catalysts 2 . Furthermore, the overexploitation of fish oil to produce bioactive fatty acids including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) can be solved by using microalgae as alternative resources 1 . The advantages of lipid production using microalgae over terrestrial plants have already been well discussed in previous studies. Microalgae can be grown in wastewater and are able to accumulate higher lipid content with a faster growth rate than common terrestrial plants. In addition, the cultivation of microalgae does not require large areas of land, and they do not compete with general human food resources 2,3 . However, cost-effective lipid production from microalgae remains in its infancy owing to poorly developed mass cultivation techniques and the low lipid productivity of target microalgal species. In order to promote lipid production from microalgae, many technological approaches including genetic engineering, stress regulation, and modification of cultivation apparatus have been proposed [4][5][6][7] . One of these approaches, the construction of more efficient algal mutants using genetic engineering, is considered a promising and important subject in the microalgae-to-bioproducts process 8 .
Changes made to cellular metabolic pathways via genetic modification in microalgae can enhance or reduce specific metabolite production, thereby artificially promoting lipid production. Among several strategies, enhancing algal lipid production by altering the starch biosynthesis pathway has been the most widely used technique in diverse microalgal candidates. According to de Jaeger et al., genomic DNA of the freshwater microalga Scenedesmus obliquus was randomly modified by UV light irradiation to obtain a starchless mutant, which exhibited enhanced lipid accumulation compared to the wild-type under nitrogen-depleted conditions 9 . In addition, Li et al. also reported that the starchless C. reinhardtii mutant BAFJ5 (referred to as sta6), which is defective in ADP-glucose pyrophosphorylase, exhibited approximately 10-fold higher TAG accumulation than the wild-type 10 . Considerable efforts have been made to address the allocation of carbon toward lipids in mutants by modifying their starch biosynthesis metabolism 9,10 . However, little attention has been paid to the regulation of autophagy and the stress response in these genetically modified algal strains.
It has been previously reported that most microalgal species can accumulate increased lipid content under stress conditions, such as nitrogen starvation, oxidative stress and high salinity 11 . For instance, Yilancioglu et al. showed a halophilic microalga Dunaliella salina accumulated cellular lipid content and exhibited extensive lipid peroxidation under nitrogen deficiency condition, indicating high level of oxidative stress 12 . The study also confirmed that hydrogen peroxide, an oxidative stress inducer, significantly induced algal lipid accumulation 12 . Additionally, H 2 O 2 treatment and UV irradiation has been shown to enhance lipid production in Scenedesmus sp. 13 . These studies suggest that lipid accumulation is highly influenced by oxidative stress in microalgae. Although the exact mechanism of this widespread phenomenon has not yet been elucidated, lipid accumulation has been linked to an increase in reactive oxygen species (ROS) along with the degradation of cellular protein components in various algal species 11,12,14 . On the other hand, ROS-mediated stress responses and protein turnover are closely related to the autophagy mechanism [15][16][17] . Autophagy is a highly conserved catabolic process that regulates cellular component degradation and recycling in a wide range of eukaryotic organisms 16 . Since it is considered that autophagy plays a significant role in the stress response of eukaryotic cells, the study of the relationship between ROS-mediated stress and autophagy induction is essential 18,19 . Recently, Couso et al. demonstrated that maintaining normal autophagic flux is essential for lipid production in C. reinhardtii 15 . Based on previous studies, it is hypothesized that cellular stress levels and the subsequent regulation of autophagy may also be correlated with an enhanced TAG accumulation in starchless mutants. Since this has not yet been concretely investigated, in the present study, we examined the nitrogen starvation-induced oxidative stress response and autophagy regulation in the starchless sta6 mutant of C. reinhardtii.

Results and Discussion
Confirmation of experimental algal strains. In order to confirm the starch phenotype of the experimental strains, wild-type (WT), complemented STA6 mutant (STA6-C6) and starchless mutant (sta6) were spotted on plates with or without a nitrogen source, and subsequently stained with iodine vapour as described in the Materials and Methods section. As shown in Fig. 1a, we could observe a dark-brown colour in WT and STA6-C6 colonies under N-depleted conditions, whereas sta6 colonies exhibited an opaque yellowish green colour indicating that starch could not be produced in this mutant. Additionally, the levels of lipid droplet accumulation in WT, STA6-C6 and sta6 cells under N-depletion was also compared (Fig. 1b,c) to verify the ability to accumulate high amount of TAG of the sta6 mutant as described previously 10,20 . As shown in Fig. 1b, the number of Nile Redstained lipid droplets in the sta6 mutant was more pronounced compared to WT and STA6-C6 mutant at 24 h after nitrogen starvation. In addition, fluorescence-based microplate assays revealed an approximate 5-fold and 8-fold increase in Nile Red fluorescent intensity in sta6 cells at 24 h and 48 h post-starvation, respectively. Neutral lipid content of the control cells was also changed in response to nitrogen starvation but less significant compared to that observed in the sta6 mutant.
Change of daily growth and stress responses. After confirmation of the sta6 mutant, daily growth tests were performed in TAP medium. As shown in Fig. 2a, the growth curves based on algal cell concentration of WT, STA6-C6 and sta6 mutant were exponential up to day 3 before reaching a stationary phase that was maintained with a cell density of approximately 2 × 10 7 cells/ml. The total nitrogen concentration was rapidly undermined at day 3 of cultivation, then slightly reduced thereafter as the cells entered the stationary growth phase (Fig. 2b). Based on these results, we considered that limiting nitrogen in the medium might induce algal cells to enter stationary growth phase after day 3.
Previous studies have reported that microalgae cultivated under nitrogen limiting conditions increased cellular ROS levels along with enhanced lipid accumulation, and the elevated ROS may inhibit algal growth by causing cellular damage 12,21 . Therefore, we measured the production of malondialdehyde (MDA), which is considered a marker of lipid peroxidation caused by enhanced oxidative stress, in algal samples collected at day 2 and day 4. As shown in Fig. 2c, the MDA content of WT cells was increased at day 4 compared to that at day 2 of cultivation, indicating the occurrence of lipid peroxidation and oxidative stress in stationary-phase cells. Interestingly, the sta6 mutant showed significantly higher MDA levels than WT and STA6-C6 cells at the same growth phase, and a high level of MDA was observed even under optimal growth conditions (day 2 sample). These results suggest that the sta6 mutant was likely exposed to higher oxidative stress levels compared to control cells growing in the same conditions. Thus, as a further experiment, we compared the cellular sensitivity of WT, STA6-C6 and sta6 to H 2 O 2 -induced oxidative stress conditions. Cells were grown to exponential phase and cell concentration was adjusted to 1 × 10 6 cells/ml before treating with H 2 O 2 to induce oxidative stress (Fig. 3). The concentrations of H 2 O 2 higher than 250 μM rapidly triggered cell death in the sta6 mutant within 16 h (Fig. 3a). Cell viability of the sta6 mutant were 77.5 ± 3.5%, 60.0 ± 7.1%, 42.5 ± 10.6% and 22.5 ± 3.5% at 250, 500, 1,000 and 2,000 μM H 2 O 2 , respectively. In contrast, WT and STA6-C6 cells exhibited relatively higher resistance to H 2 O 2 -induced cell death (less than a 25% and 40% decrease in viability at 1,000 and 2,000 μM H 2 O 2 , respectively). These results indicate that the sta6 www.nature.com/scientificreports www.nature.com/scientificreports/ mutant is more sensitive to oxidative stress compared to the WT and complemented STA6 mutant, which is in accordance with the elevated cellular MDA content observed earlier.

Effects of H 2 o 2 on neutral lipid accumulation in Chlamydomonas reinhardtii. Treatment with
H 2 O 2 has been shown to effectively enhance lipid yield in D. salina and Scenedesmus sp. 12,13 . Indeed, the addition of H 2 O 2 to the culture medium caused an increase in intracellular ROS levels in a wide range of organisms [22][23][24] . As cellular signalling molecules, the accumulated ROS may act as potent inducers of multiple metabolic functions including lipid metabolism and autophagy 25 . To determine whether the increased lipid accumulation in the sta6 mutant correlates with its distinct oxidative profile, the cells in Fig. 3a were stained with DCFH-DA and fluorescence-based microplate was used to measure cellular ROS levels. The ROS contents were increased in a dose-dependent manner with H 2 O 2 concentration in all strains (up to 500 μM H 2 O 2 in case of sta6 mutant) (Fig. 3b). At the concentrations of H 2 O 2 higher than 500 μM, the sta6 mutant had no increase in ROS levels but dramatic increase in cell death, indicating that sta6 cells has reached an oxidative saturation point where the level of ROS exceeds cellular defence capacitance. Notably, we observed a correlation between increased neutral lipid content and increased cellular ROS levels in controls and sta6 mutant (except for the 2,000 μM H 2 O 2 condition). As shown in Fig. 3c, with WT cells exposed to increasing amounts of H 2 O 2 , lipid content increased from 271.2 ± 15.5 A.U. (non-treated control) to a maximum of 447.5 ± 23.2 A.U. (at 1,000 μM H 2 O 2 ) (A.U. means arbitrary units). The changes in Nile Red fluorescence intensity for the sta6 mutant were more pronounced compared to that observed in WT and STA6-C6 cells, which showed a positive, yet saturating relationship with the changes in its ROS levels (Fig. 3c). These results suggest that the overproduction of ROS in the sta6 mutant may contribute to its elevated lipid accumulation. Differential autophagic activity in the sta6 mutant. Autophagy is a catabolic process that usually occurs in eukaryotic organisms subjected to stress. The damaged cellular organelles caused by increased ROS are degraded and recycled via autophagic activity to maintain cellular homeostasis 26 . While the degradation of cellular materials is performed by the lysosome in animal cells, the same function is performed by lytic vacuoles in microalgae, fungi and plant cells 27 . As discussed in a previous review 28 , the autophagic machinery of www.nature.com/scientificreports www.nature.com/scientificreports/ Chlamydomonas comprises numerous autophagy-related (ATG) proteins. In the case of macroautophagy, hereafter referred to as autophagy, a double-membrane vesicle called the autophagosome is formed via a complicated mechanism. Autophagosome formation begins by the activation of the ATG1 complex composed of ATG1, ATG13 and probably ATG101 28,29 . ATG101, as previously described in the model terrestrial plant Arabidopsis thaliana, may help link the ATG1 initiation complex to autophagic membranes 30 . The phosphorylation of ATG1 triggers the formation of a phagophore assembly site (PAS) through phosphoinositide 3-kinase (PI3K) nucleation complex and phosphatidylinositol 3-phosphate (PI3P) binding complex. PI3K (VPS34) plays an important role in converting phosphatidylinositol (PI) to PI3P, which is required for the recruitment of other ATG proteins onto the autophagosome membrane. Among ATG proteins, ATG8 is essential for the formation of the autophagosome, while ATG7 is an E1-like enzyme involved in the activation of ATG8 28,30,31 .
In this study, we measured the expression levels of ATG1, PI3K, ATG101, ATG7 and ATG8 genes to evaluate autophagy regulation in the sta6 mutant (Fig. 4). Compared to the expression levels of WT (marked as dash lines), all the autophagy-related genes in the sta6 mutant showed higher values at both day 2 and day 4. Furthermore, the ATG1 and ATG101 genes, which regulate the early steps of the autophagic process, were significantly up-regulated in sta6 cells after 4 days of cultivation, whereas ATG7 and PI3K has shown no statistically significant changes. Moreover, the expression levels of the ATG8 gene were significantly promoted in samples collected at D4 (when the cells reached stationary phase). Previous studies have demonstrated that increased cellular ROS may lead to the activation of ATG1 and subsequent autophagy by either targeting rapamycin (TOR)-dependent or TOR-independent signalling 16 . Thus, the enhanced expression levels of those autophagy-related genes suggested that the sta6 mutant activated autophagy mechanism to cope with its elevated oxidative stress levels.
In addition, Western blot analysis was performed to examine the expression pattern of ATG8 protein in the sta6 mutant. ATG8 appeared as a faint band on the blot of WT and STA6-C6 cells grown under optimal conditions (Fig. 5a, 0 μM H 2 O 2 ). The abundance of ATG8 increased upon exposure to H 2 O 2 -induced autophagy in both controls and the sta6 mutant (Fig. 5a). The quantification of protein band intensity was performed using ImageJ, claiming a significant increase of ATG8 protein in the sta6 cells compared with WT or STA6-C6 mutant subjected to the same amount of stressor (Fig. 5b). Moreover, the dose-dependent expression pattern of ATG8 protein, observed in WT and STA6-C6 cells, was less significant in the sta6 mutant (Fig. 5a,b). It seems that oxidative stress induces autophagy more strongly in the sta6 mutant compared with WT and STA6-C6 mutant. Based on these results, we propose that the high lipid accumulation in the sta6 mutant could be explained from the perspective of stress responses and autophagy regulation, alongside carbon allocation as described previously. Although the underlying mechanism remains to be clarified, we speculate that the absence of starch biosynthesis likely led to increased levels of ROS and oxidative stress in the sta6 mutant. Subsequently, the increased cellular ROS levels could act as autophagy inducers, leading to an enhanced recycling of cellular materials and ROS-damaged components to fuel lipid production in this mutant.
It has been reported that corn, which contains high levels of amylose, one of the two components of starch, has a good antioxidant capacity 32 . Also, largemouth bass Micropterus salmoides fed with pea starch and high-amylose maize starch showed a lower MDA content and reduced oxidative stress levels 33 . Similarly, oxidative stress suppression by resistant starch diet was also reported by Vaziri et al. 34 . In addition, starch biosynthesis has been shown to be an essential protection mechanism against photooxidative damage in C. reinhardtii 35 . The continuous generation of the reductant NADPH as an end product of the light reactions of photosynthesis leads to cellular over-reduction and the formation of ROS by transferring one electron from NADPH to oxygen 36 . In photosynthetic cells, ATP and NADPH produced by the light reactions are used to form sugars from carbon dioxide (CO 2 ) via the Calvin cycle, and these sugars are further condensed into insoluble starch which has a lower osmotic pressure 37 . By consuming excess NADPH, the production of starch allows the algae to maximize light harvesting while regenerating NADP + from NADPH which can accept electrons and thus reduce the generation of ROS 35 . Since lipid biosynthesis is the second NADPH-consuming process, an increased rate of lipid production in the sta6 mutant would allow the cells to lower the NADPH pool 38,39 . In fact, the levels of oxidative damage in the sta6 mutant were still significantly higher than that in the WT cells despite its elevated lipid accumulation. Juergens et al. highlighted that in C. reinhardtii starch is primarily produced from CO 2 via photosynthesis, whereas the synthesis of fatty acids mostly relies on exogenous acetate and the turnover of membrane lipids 40 . Hence, this indicated that starch plays a more important role in the protection of the photosynthetic machinery from photooxidative damage than other bioproducts, including lipids and proteins 35,40 . Taken together, this suggests that one possible reason for the increase in lipid peroxidation in the sta6 mutant may be the unbalanced redox state of the cells, leading to an elevated generation of ROS due to the lack of the starch biosynthesis pathway.

Conclusion
Superior lipid accumulation in starchless mutants has been generally linked to increased carbon allocation towards the lipid metabolism in previous studies. This work has expanded on those findings from a different perspective, specifically focusing on the stress response and autophagy regulation. The increased lipid peroxidation observed in sta6 cells growing under the same conditions as WT suggests that algal cells with impaired starch www.nature.com/scientificreports www.nature.com/scientificreports/ metabolism need to cope with elevated oxidative stress levels. The increased cellular ROS levels in sta6 cells, therefore, could play a dual role as signals to trigger lipid biosynthesis and as autophagy inducers to activate the recycling of cellular components to fuel lipid production in this mutant. More research is needed to understand the antioxidant capability of purified algal starch molecules, and their roles in the regulation of related cellular pathways.

Materials and Methods
Microalgal growth conditions. Chlamydomonas reinhardtii CC-124 wild-type, complemented STA6 mutant (referred to as STA6-C6) and starchless mutant BAFJ5 (referred to as sta6) were obtained from the Chlamydomonas Resource Center located at the University of Minnesota. In the sta6 mutant, the small subunit of ADP-glucose pyrophosphorylase was deficient and the starch content was not detectable 41 .
Cells were grown in Tris-acetate phosphate (TAP) medium 42 under continuous illumination of 50 ± 10 µmol/ m 2 /sec at 25 °C, with constant shaking at 90 rpm. For growth rate comparison, cells were initially inoculated at a density of approximately 8 × 10 4 cells/ml, and cell density was monitored daily using a haemocytometer.
For the nitrogen depletion studies, cells in exponential phase (at a density of approximately 1 × 10 6 cells/ml) were harvested by centrifugation (2,000 × g for 5 min). The cell pellet was washed twice in nitrogen-free TAP medium before resuspension in the same fresh medium at the same cell density. . Expression levels of genes belonging to autophagy pathways. Transcript levels of autophagyrelated (ATG) genes were measured by real-time qPCR in algal samples at day 2 (exponential phase) and day 4 (stationary phase). Relative expression levels of ATG1, PI3K, ATG101, ATG8 and ATG7 genes are shown. All results were normalized to CBLP (house-keeping gene). The bars represent the fold-change in expression of the target gene in STA6-C6 and sta6 relative to wild-type. The dashed line indicates the expression level of corresponding genes in wild-type. Mean values of two biological replicates were obtained using the medians of three technical replicates. Error bars represent 95% confidence intervals. The overlapping confidence intervals indicate no significant difference. www.nature.com/scientificreports www.nature.com/scientificreports/ Determination of algal starch content using iodine vapor staining. Pure elemental iodine crystals (Junsei Chemical Co., Ltd., Japan) were placed on a watch glass and heated slightly to generate iodine vapor inside a fume hood. To control the staining time, WT, STA6-C6 and sta6 cells were spotted on the same plate. Agar plates containing algal colonies were placed on top of the watch glass for 30 s. Images were captured immediately after staining.
Analysis of total nitrogen. Algal culture broth was filtered through a 0.2 µm pore-size membrane filter (Minisart, Sartorius Stedim Biotech, Germany) before measurement. The concentration of total nitrogen (TN) was determined using Nitrogen (total) Cell Test kit (Merck, Germany) in accordance with the manufacturer's protocol.
Determination of lipid peroxidation. Lipid peroxidation products are widely used to represent the levels of oxidative stress in cells. In this study, the thiobarbituric acid reactive substance (TBARS) assay was used to quantify the level of MDA content in microalgal cells (modified from 43 ). Briefly, 1 ml samples with known number of cells were harvested by centrifugation at 2,000 × g for 5 min. Cell pellets were washed twice with cold phosphate buffered saline (PBS, pH 7.4) and resuspended in lysis buffer at a density of approximately 1 × 10 6 cells/ml. Lysis buffer was prepared by mixing 1 µl of 10 mM butylated hydroxytoluene (BHT) solution per 100 µl of radio-immunoprecipitation assay (RIPA) buffer. The cell mixture was bead beaten at 4,500 × g for 1 min. Acid treatment was performed by adding 100 µl of 0.6 N trichloroacetic acid (TCA) solution per 100 µl of cell lysate and incubating at room temperature for 15 min. After centrifugation at 10,000 × g for 5 min, the supernatants were immediately assayed. The TBARS assay was performed by mixing 150 µl of each sample or MDA standard with 75 µl of 0.5% (w/v) thiobarbituric acid reagent. The optical density (OD) was determined at 532 nm immediately (pre-read OD 532 ) and after incubating for 2 h at 45-50 °C. Pre-read OD 532 was subtracted from the final read to obtain data, and MDA contents were calculated based on the MDA standard curve. The results are expressed as µM MDA/10 6 cells.
Western blot analysis. Cell lysates were prepared using RIPA buffer supplemented with Protease Inhibitor Cocktail (Sigma-Aldrich, United States). Then, 30 µg of total soluble protein per sample were separated on 15% SDS-PAGE gels and transferred onto a polyvinylidene difluoride membrane (Bio-Rad, United States) using a wet transfer system (Bio-Rad, United States). Western blot analysis was performed with anti-CrATG8 (Abcam, United Kingdom) and anti-α-tubulin (Sigma-Aldrich, United Kingdom) antibodies (1:2,000). The Clarity TM Western ECL Substrate (Bio-Rad, United States) was used to detect ATG8 and α-tubulin with anti-rabbit (Abcam, United Kingdom) and anti-mouse (Abcam, United Kingdom) secondary antibodies (1:10,000), respectively. Algal lipid droplet observation by fluorescence microscope. Chlamydomonas cells were subjected to nitrogen starvation and microscopic observation was performed at 24 h and 48 h post-treatment. Lipid droplets were stained with Nile Red (Thermo Fisher, United States) and observed using a fluorescence microscope (Eclipse 80i, Nikon, Japan) at excitation 450-500 nm (blue laser) and emission 528 nm wavelength. Chlorophyll a (Chl a) auto-fluorescence was observed using the same filter. statistical analysis. All data are presented as means ± (1.96 × standard error (SE)) for a 95% confidence interval (unless otherwise indicated). The overlapping error bars indicate no significant difference. Pairwise comparisons of data from different experimental groups were performed using a t-test (Excel 2016 program, Microsoft). Statistically significant differences are represented by different letters (Fig. 2).