Plastid terminal oxidases in Chlamydomonas: connections with astaxanthin and bio-hydrogen production

Background: as a plasto quinol oxidase involved in plastoquinol oxidation in higher plants and microalgae, the plastid terminal oxidase (PTOX) was rst recognized in the tomato mutant GHOST (GH) and Arabidopsis mutant IMMUTANS (IM). Genome sequence analysis revealed that duplication of the PTOX gene occurs in certain eukaryotic microalgae, but not in cyanobacteria and most higher plants. PTOX may also be involved in carotenoid synthesis and play a critical protective role against stress, such as high light, heat shock and hyperosmosis. However, the connections of PTOX with astaxanthin and biohydrogen production and their functional relationship between two PTOX genes in the model green microalga Chlamydomonas is unknown. Results: we successfully knocked down two ptoxs through RNAi in Chlamydomonas, respectively. We demonstrated that expression levels of both PTOXs were increased under stress conditions, and interestingly when one PTOX was silenced the other’s transcriptional level was signicantly raised. Conclusions: this shows a complementary relationship under high light condition. In addition, the astaxanthin accumulation level was up-regulated in silenced ptox2 strain, compared to the wide type strain. What’s more, signicantly increased hydrogen production was observed in silenced ptox1 strain. In conclusion, PTOXs in Chlamydomonas are connected with not only astaxanthin accumulation but also hydrogen production, and their knock-down strains provide new insights in manipulating microalgae for high light stress tolerant strains, carotenoid production and even biofuels.

plays important roles in SA-mediated resistance to water stress during soybean seedling [16]; It was also proposed to have the potential to act as a safety valve for the excess excitation energy in the alpine plant species Ranunculus glacialis caused by rapid light and temperature changes [17,18]. In microalgae, by an enhanced activity of PTOX cells could decrease PSII excitation pressure in a gun4 mutant of Chlamydomonas reinhardtii [19] and PTOX responds to different environmental stress in Haematococcus pluvialis [20].
Carotenoids are accumulated at some development stage and in environmental stimuli in microalgae and higher plants, which is an acclimation strategy of the cells and helps to protect against photo-oxidative stress since carotenoids act as antioxidants [21]. The possible involvement of PTOX in the pathway of carotenoid biosynthesis [22] and chlororespiration [23] were summarized. Data show that PTOX also participates in carotenoid desaturation via a complex regulatory pattern [24] in addition to its role during early chloroplast development [25,26]. The GH terminal oxidase regulates developmental programming in tomato fruit [27] and fruit ripening [28]. Additionally, an increasing pattern of PTOX transcripts leads to carotenoid accumulation during ower bud development in Liriodendron tulipifera [29]. In another green microalga, Haematococcus pluvialis, comparative lipidomic and transcriptomic analyses reveal a concerted action of multiple defensive systems against HL stress [30] and PTOX, together with PSI cyclic electron transport, defensive enzyme and the accumulated astaxanthin, can protect microalgal cells against photoinhibition [31].
Genome sequence analysis revealed that duplication of PTOX gene occurs in certain eukaryotic algae, but not in cyanobacteria and most higher plants. In previous study in Haematococcus, two ptoxs showed differential expression patterns under different stresses [20]. The differential expression of two ptoxs in Chlamydomonas was previously reported under phosphate depletion [32]. The knockout mutant of PTOX2 in Chlamydomonas shows lower tness than wild type when grown under phototrophic conditions [33], and another PTOX2 mutant of C. reinhardtii results in almost complete reduction of the plastoquinone pool in light [34]. But no further research shows the interaction between these two PTOX genes.
Hydrogen production in green mciroalgae requires electrons from the photosynthetic electron transfer chain to reduce H + into H 2 , especially in Chlamydomonas [35]. Since PTOX involves in chlororespiration, it is reasonable to propose that PTOX may also be one of the manipulation targets to improve the hydrogen production in microalgae.
In this study, RNA interference technique was employed to knock down two PTOXs of Chlamydomoans individually, investigating the functional relationships of these two PTOXs under stress condition. Also, the potential correlation of PTOX with astaxanthin and hydrogen bio-production was also studied.

PTOX knockdown and growth curves
Firstly, we veri ed the successful construction of RNAi vectors using restriction digestions and DNA sequencing (Fig. S1, S2). Then more than 20 positive transformants derived from each ptox vector were screened based on their gene expressions under high light by qRT-PCR. The RNAi backbone pH124 is a high light induced vector, thus we measured expression levels of different genes under both normal and HL conditions. We selected the strains with the lowest ptox mRNA levels under HL, 4-6 (~ 10% of control only) and 5 − 3 (~ 29%), as PTOX1-RNAi (PTOX1i) and PTOX2-RNAi (PTOX2i) inferred strains for further investigations, respectively (Fig. S3).
Then we are interested in the potential correlation of PTOX knockdown and cell growth. Overall, HL inhibits cell growth in all 3 strains, PTOX1i, PTOX2i and CC849. However, RNAi mutants showed signi cant cell growth inhibition under HL, while there was some insigni cant difference compared with control under normal culture condition. Under normal culture condition, PTOX1i cells grow slowly after day 3 at the log phase but reached similar cell density to the control and PTOX2i showed higher cell density after 5 days culture till the platform stage, but not statistically signi cant (Fig. 1).
With HL treatment, cell growth in all strains were signi cantly inhibited, more than 30% and 56-60% inhibition in C. reinhardtii CC849 PTOX1i and PTOX2i compared to those under normal culture condition at the day 8. Though no obvious difference was observed between two mutants under HL, the RNA interferences of both ptoxs in C. reinhardtii signi cantly inhibited cell growth compared to the C. reinhardtii CC849. This indicates that PTOXs may play an important role in cell growth and high light stress response. ptox1 and ptox2 differential expression in normal and HL conditions Signi cantly different responses to cell growth were detected in two ptox genes in Chlamydomonas. Under normal condition, ptox2 expression showed no change compared with the CC849, while ptox1 mRNA level was being increased continuously and reached the highest level at 72 h, about 20 fold changes compared to that of 0 h in both C. reinhardtii CC849 and PTOX1i strain. When cells entered the log phase, ptox1 started to down-regulated and reached relatively low level at 120 h. In such normal culture condition, similar patterns were observed in both WT and PTOX1i. Combining with the cell growth curves, ptox1 expression pattern showed highly correlation, with increasing levels at log phase and reached the maximum level at day 3 (Fig. 2a).
Interestingly, different ptox gene expression patterns were counted in different strains under HL stress (Fig. 2b, c). In this condition, the ptox1 mRNA levels were up-regulated with more than 2 fold changes in the control cells, and its up-regulation was signi cantly inhibited in RNAi mutant PTOX1i (Fig. 2b). In contrast, ptox1 was signi cantly increased in the PTOX2i mutant cells, with the highest up-regulation (-10 fold) observed at 48 h of HL treatment.
Gene ptox2 seems to sensitively respond to HL in Chlamydomonas, with nearly a maximum 20 fold increases at 48 h in the control CC849 cells, and a similar high level at 96 h was also detected in PTOX1i mutant cells. The RNAi inhibition of ptox2 was found after 12 h and relatively low expression level were recorded after 48 h till 120 h (Fig. 2c).
The interaction of ptox1 and ptox2 was also noted at the transcriptional level. Based on qRT-PCR analyses, the inhibition of ptox2 in PTOX2i under HL resulted in the signi cant increase of ptox1 with the maximum expression peak at 48 h with ~ 10 fold increase in PTOX2i compared with ~ 1.5 fold in PTOX1i; moreover, inhibition of ptox1 caused a delayed up-regulation of ptox2 in PTOX1i with a peak at 94 h instead of 48 h in CC849.

Astaxanthin production
In this study, under normal culture condition, astaxanthin was not obviously produced even after 120 h growth in all three strains. While there was no signi cant astaxanthin induction in CC849 and PTOX1i strain, interfered ptox2 in PTOX2i resulted in a signi cant increase of astaxanthin accumulation (p < 0.05).
HL effectively induces astaxanthin accumulation in all strains, compared to normal culture condition (Fig. 3). After 120 h HL treatment, astaxanthin in CC849, PTOX1i and PTOX2i mutants were increased 4.2, 4.8 and 5.4 fold than that of 0 h HL, respectively. RNAi mutants showed higher levels of astaxanthin accumulation compared to the control under HL, PTOX2i and PTOX1i cells contained almost 2.3 and 1.3 fold astaxanthin after 120 h compared to 0 h HL treatment, accordingly. Similarly, only PTOX2i mutant cells showed obviously higher astaxanthin accumulation than CC849 and PTOX1i cells after 120 h HL treatment.

Bio-hydrogen production
To explore the consequence of PTOX inhibitions under HL, the bio-proudction of hydrogen was determined in all three strains. Microalgae cells were cultured in normal condition until exponential phase and then treated with HL, and then the hydrogen production was determined, using gas chromatography to measure the content of gas in the headspace of WT and RNAi strains. It was observed the signi cantly increased hydrogen production in PTOX1i mutant under HL treatment, whereas there was slightly but not signi cantly increase in the other RNA interfered mutant PTOX1i compared to the control C. reinhardtii CC849 (Fig. 4), indicating HL inhibition of PTOX1 and PTOX2 affected the hydrogen production in Chlamydomonas.

Discussions
We established successfully two knock-down mutants, with 90% and 71% mRNA reduction of ptox1 and ptox2 under HL stress, respectively. Under normal culture condition, single knock-down of either PTOX did not inhibit cell growth, however, both mutants showed the signi cantly reduced growth under HL instead, more than 60% inhibition compared to only 30% in WT.
Current modi cation of PTOX expression, either knock-out or overexpression, shed lights on their multiple cellular functions. Knocking out PTOX in plants or microalgae resulted in severe phenotypes that encompass developmental and growth defects together with increased photosensitivity [36]. Interestingly, down-regulation of PTOX, approximately 3% of WT levels, did not compromise plant growth, under ambient growth conditions in Arabidopsis [37]. While over-expression of C. reinhardtii PTOX1 in plants makes the mutants more sensitive to HL than WT [38,39] and Arabidopsis PTOX in tobacco promotes oxidative stress [40,41]. Similarly, OsPTOX expression in Synechocystis did not affect growth under standard growth conditions (light intensities between 50 and 150 µmol photons m − 2 s − 1 ) [42]. In other stress treatments, over-expression of PTOX from the salt-tolerant brassica species Eutrema salsugineum show faster induction and a greater nal level of PTOX activity once exposed to salt stress [43].
Based on mRNA level comparison between WT and a delta-psbA tobacco plant, up-regulation of the alternative electron transport pathways (NDH complex and PTOX) occurs at the translational or posttranslational levels [44]. This suggests that PTOX is normally in excess, with delicate expression regulations in not only plants but also microalgae.
The differential expression patterns and the complementary relationship of ptoxs under both normal and HL treatments indicate complicate and different potential roles of these two PTOX genes in chlororespiration and stress responses. For instance, the astaxanthin accumulation level was much more in ptox2 silenced strain than ptox1, compared to WT. What's more, signi cantly increased hydrogen production was observed in ptox1 silenced strain.
In Glycine max, differential expression of recently duplicated PTOX genes during plant development and stress conditions were extensively investigated [45]. The majority of plant species contain only a single gene encodes PTOX. Previously, two putative PTOX (PTOX1 and 2) genes were identi ed in Glycine max. In development, PTOX1 was predominant in young tissues, while PTOX2 was more expressed in aged tissues. Under stress conditions, the PTOX transcripts varied according to stress severity, i.e., PTOX1 mRNA was prevalent under mild or moderate stresses while PTOX2 was predominant in drastic stresses.
Overall, the results indicate a functional relevance of this gene duplication in G. max metabolism, whereas PTOX1 could be associated with chloroplast effectiveness and PTOX2 to senescence and/or apoptosis [45].
The differential expression of ptoxs was also observed in Chlamydomonas under phosphate deprivation where ptox2 mRNA level was up-regulated about 13-fold whereas the ptox1 transcripts increased 2.4-fold after 48 h of treatment [32]. In a knockout mutant of PTOX2 in Chlamydomonas the plastoquinone pool is constitutively reduced under dark-aerobic conditions, and the ptox2 mutant shows lower tness than wild type when grown under phototrophic conditions [33]. In another report, the absence PTOX2 and cytochrome b6f complex of C. reinhardtii, results in almost complete reduction of the plastoquinone pool in light [34]. In this study, both ptox genes connected with cell growth inhibition, but with no signi cantly difference. Interestingly, under normal light intensity, ptox1 mRNA levels correlated with the growth in both WT and PTOX1i mutant while ptox2 remained relatively low under both WT and PTOX2i strains in this study. Thus, ptox1 may involve in cell growth and ptox2 plays other metabolic roles in Chlamydomonas.
PTOX is regarded as an enzyme at the crossroads of various metabolic processes, such as regulation of cyclic electron transfer and carotenoid biosynthesis [36]. PTOX is very important for carotenoid biosynthesis, since the phytoene desaturation, a key step in the carotenoid biosynthesis, is blocked in the white sectors of Arabidopsis im mutant [3]. The absence of PTOX in plants usually results in photobleached variegated Arabidopsis leaves [46] and impaired adaptation to environment alteration, and mutant plants will not survive the mediocre light intensity during its early development stage [3]. Although PTOX level and activity has been found to increase under a wide range of stress conditions [9].
Thus, PTOX involves in carotenoid biosynthesis but which ptox gene plays more important role in this process in microalgae is still under investigation.
In other green microalga Haematococcus, Hptox1 and Hptox2 also showed differential expression patterns in response to various oxidative stresses [20]. And the authors regarded Hptox1 as the key PTOX gene for co-regulation of astaxanthin accumulation in Haematococcus [20]. The rice ptox1 mutant accumulated phytoene in white leaf sectors with a corresponding de ciency in beta-carotene, consistent with the expected function of PTOX1 in promoting phytoene desaturase activity. Our results demonstrate that PTOX1 is required for carotenoid synthesis [47].
Similar to heat and drought, HL treatment stimulates chlororespiration in higher plants [48] and microalgae [20], causing the up-regulation of the PTOX and the thylakoidal NADH DH complex [49,50]. The natural astaxanthin mainly derives from the microalgae producer, Haematococcus. The induction of nitrogen starvation and high light intensity is particularly signi cant for boosting astaxanthin production [51]. Under HL, ptox1 mRNA level was up-regulated signi cantly (more than 10 fold compared to WT) in PTOX2i mutant. Together with our study, it suggests that ptox expressions differ temporally or spatially in response to various stressors. We speculated that it is ptox1 rather than ptox2 that is co-regulated, or functionally coupled with carotenoid biosynthesis in Chlamydomonas, similar to Haematococcus.
Hydrogen production in green mciroalgae requires electrons from the photosynthetic electron transfer chain to reduce H + . Sulfur-deprived cultivation of C. reinhardtii [52] was previously regarded as the most e cient technique to enhance photobio-H 2 production in microalgae [53][54][55]. Most recently, modi cation of photosynthetic genes and even non-coding RNAs in Chlamydomonas signi cantly improve the bio-hydrog8 en production [35,56,57].
Under HL, ptox2 responds to the stress with delayed maximum expression level in ptox1 silenced strain-PTOX1i mutant, with a similar but delayed up-regulation level with WT. What's more, signi cantly increased hydrogen production was observed in the PTOX1i mutant. We address the importance of ptox2 gene in hydrogen bioproduction in Chlamydomonas.
The limitations of our study are the absence of evidence based on protein levels, and RNAi knock-down still remains relatively high basal expression of the target genes. Further investigations involving in Western blotting and genome editing (for complete individual or double knock-out, if possible) would be highly required for a better understanding of the diverse functions of PTOX genes in microalgae.

Conclusion
We tried to verify the different functions of two PTOX genes in a model green microalga, by RNAi technique under normal and high light conditions. We propose that PTOX1 may involve in cell growth, cofunctional with astaxanthin biosynthesis while PTOX2 may respond to stress conditions and a better candidate synthetic target for bio-hydrogen bioproduction.

Materials And Methods
Organism, growth medium, and culture conditions Chlamydomonas reinhardtii cell wall-de cient mutant strain CC849 was obtained from the Chlamydomonas Genetic Centre (Duke University, Durham, USA). Microalgal cells were grown using Tris acetate phosphate (TAP) media with mineral nutrient supplements [58] at 25℃ and under continuous cool-white uorescent lamps (≈ 100 µmol photons/m 2 s).

RNAi vector construction and transformation
Standard gene cloning methods [59] were used to make the gene constructs. Primers were designed (Table 1) to clone the forward and reserve fragments of ptox1 (GenBank ID: 5718064), ptox2 (GenBank ID: 5728910) and their introns. The vector for the RNAi-mediated silencing of the ptox1 and ptox2 gene was constructed as described previously [60]. Transgenic cells were plated on selective media containing 1.5 mM L-tryptophan, 5 µg/mL paromomycin, and 5 µM 5-FI. We synthesized forward and reverse CDS fragments of ptox1 and ptox2 then fused with their introns (Table 1) (Fig. S1) respectively. The products were inserted into plasmid pH124 for PTOXs silence. All these PCR products with correct length were puri ed through Takara Agarose-Gel DNA Puri cation Kit V2.0 and then stored in -20℃. The `hairpin' RNA encoding sequence with plasmid PH124 was constructed by fusing the above available three DNA fragments with SOE-PCR (splicing by overlapping extension-PCR) (Fig. S1).
The pH124-PTOXs-RNAi plasmids were then linearized by Not I and transferred into C. reinhardtii CC849 through the glass bead method [38]. Then qRT-PCR technique was used to screen for the silenced strains with the most signi cant down-regulation of ptoxs and those with lowest ptoxs levels were selected for following experiments.
qRT-PCR TaKaRa  Hydrogen production C. reinhardtii CC-849 and transgenic algal strains (250 mL) were cultured in 500-ml culture bottles sealed with rubber sheet septa until exponential phase in a red light incubator, followed by irradiation with continuous white or blue light to detect hydrogen production. A gas chromatograph was used to detect the concentration of H 2 (Agilent 7890A; Agilent Technologies Inc., USA). H 2 , O 2 , and N 2 in the gas samples were separated by a molecular sieve column (type 5 Å; mesh size 60/80; 6 ft. × 1/8 in. × 2.0 mm), and argon was used as the carrier gas. Data were analyzed using F test to test the homogeneity of variance, and then using t test to determine difference signi cance.

Statistical analysis
All experiments were repeated at least three times independently, and data were recorded as the mean with standard deviation (SD   Relative expression levels (fold change) of ptox1, 2 genes in WT and their respective mutants under normal light density (Fig 2a), ptox1 in WT two mutants under HL culture condition (Fig 2b) and ptox2 in WT and mutants under HL culture condition (Fig 2c).