Enzyme activity highlights the importance of the oxidative pentose phosphate pathway in lipid accumulation and growth of Phaeodactylum tricornutum under CO2 concentration

Background Rising CO2 concentration was reported to increase phytoplankton growth rate as well as lipid productivity. This has raised questions regarding the NADPH supply for high lipid synthesis as well as rapid growth of algal cells. Results In this study, growth, lipid content, photosynthetic performance, the activity, and expression of key enzymes in Calvin cycle and oxidative pentose phosphate pathway (OPPP) were analyzed in the marine diatom Phaeodactylum tricornutum under three different CO2 concentrations (low CO2 (0.015 %), mid CO2 (atmospheric, 0.035 %) and high CO2 (0.15 %)). Both the growth rate and lipid content of P. tricornutum increased significantly under the high CO2 concentration. Enzyme activity and mRNA expression of three Calvin cycle-related enzymes (Rubisco, 3-phosphoglyceric phosphokinase (PGK), phosphoribulokinase (PRK)) were also increased under high CO2 cultivation, which suggested the enhancement of Calvin cycle activity. This may account for the observed rapid growth rate. In addition, high activity and mRNA expression of G6PDH and 6PGDH, which produce NADPH through OPPP, were observed in high CO2 cultured cells. These results indicate OPPP was enhanced and might play an important role in lipid synthesis under high CO2 concentration. Conclusions The oxidative pentose phosphate pathway may participate in the lipid accumulation in rapid-growth P. tricornutum cells in high CO2 concentration.


Background
As a result of increased industrialization and human activities, global carbon dioxide (CO 2 ) emission has increased dramatically and induces seawater acidification, which affects the growth and photosynthesis of marine phytoplankton [1]. Photosynthetic CO 2 fixation by phytoplankton (e.g., eukaryotic microalga) mainly depends upon Calvin cycle, and ribulose-1, 5-bisphosphate carboxylase/ oxygenase (Rubisco) catalyzes the initial step of CO 2 fixing. The response of photosynthesis, metabolite including the Calvin cycle enzymes (especially Rubisco), and growth to elevated CO 2 have been well studied in higher plants. It was shown that short-term CO 2 elevating generally accelerates carbon fixation and leads to an increase of Rubisco activity followed by an enhancement of growth in C 3 plants [2][3][4][5][6]. Yet, in the long-term, the photosynthesis decreased with increasing CO 2 , which is typically accompanied by a decline in the amount and activity of Rubisco and other enzymes in the Calvin cycle and a decrease of growth rates. Compared to high plants, the impacts of elevated CO 2 concentration on microalgae and their response to CO 2 levels, especially Calvin cycle enzyme (including Rubisco) activity and amount, have been learned much less extent and attentions mainly paid to photosynthesis and algal growth. Only in the study of Euglena gracilis Z, Nakano et al. focused on Rubisco and found Rubisco activity was higher in high CO 2 conditions [7]. Whereas, elevated CO 2 concentration increased the efficiency of photosynthetic carbon fixation and growth of phytoplankton was gradually known as general phenomenon [8,9]. Kim et al. [10] reported that the growth of Skeletonema costatum enhanced at higher concentrations of CO 2 . Tortell et al. [11] also found that rising CO 2 can enhance Chaetoceros spp. growth. In addition, various studies have shown that rising CO 2 concentration increases lipid productivity as well as phytoplankton growth rate, such as in Phaeodactylum tricornutum [12], Nannochloropsis oculata [13], and Chlorella vulgaris [14], high levels of CO 2 concentration enhanced both biomass production and lipid content, thus shedding light on the potential for biodiesel production from microalgae. To select microalgae for obtaining a higher lipid productivity, even higher concentrations of CO 2 (10 % CO 2 and flue gas) were used to cultivate Botryococcus braunii and Scenedesmus sp. [15]. At present, about 60 species of microalgae have been well domesticated with high concentration of CO 2 for producing large biomass and achieving high biofuel yields [16]. However, most previous studies have only focused on microalgal growth rate, lipid content, and tolerance to high levels of CO 2 [17][18][19][20]. Little effort has been directed toward the analysis of the mechanism involved in lipid accumulation in microalgae and their simultaneous rapid growth rate.
Microalgal lipids, as a source of biofuel, are usually derived from long-chain fatty acids, which require NADPH for synthesis [21,22]. For example, to produce an 18carbon fatty acid, 16 NADPH molecules are required as electron donors. Therefore, enhanced lipid accumulation will surely increase metabolic demand in microalgae for NADPH. Microalgae have been demonstrated to grow rapidly in high CO 2 concentrations. This suggests that quantity of NADPH, which is supplied by the light reaction, is required for photosynthetic carbon fixation which supplies substrates and energy for the synthesis of major constituents (proteins, nucleic acids, and carbohydrates) essential for algal growth. For effective CO 2 fixation, ATP and NADPH produced by photosynthetic light reactions must be maintained at a molar ratio of 3:2 [23]. Once a large number of NADPH molecules are consumed, the ratio will be disrupted leading to a reduction in carbon fixation activity. An important question therefore remains about how NADPH is supplied for high fatty acid synthesis as well as rapid growth of algal cells cultured under high CO 2 concentration. It is more likely that another pathway may contribute to providing this reductant.
In the present study, we evaluated the lipid content in the diatom P. tricornutum which was cultivated in three different CO 2 conditions (0.015 %, atmospheric, and 0.15 %). Furthermore, we measured the activity of seven key enzymes and mRNA expression in P. tricornutum to explore the mechanism of rapid growth and the simultaneous increase in lipid accumulation in high-CO 2 cultured algal cells. Our research showed that the pentose phosphate pathway may be incorporated in maintaining the NADPH supply under high CO 2 concentrations.

Results
Algal growth, pH changes, and photosynthetic performance under low, mid, and high CO 2 concentrations When grown under low, mid, and high CO 2 concentrations, the growth rate of P. tricornutum showed significant differences among the groups. The highest growth rate occurred under high CO 2 concentration and was 0.20 at the beginning of treatment and 0.70 on the seventh day at the optical density (OD) of 730 nm, followed by mid CO 2 concentration, and was 0.20 at the beginning of treatment and 0.41 on the seventh day at OD 730 nm . No significant growth was observed in the low-CO 2 cultured alga (Fig. 1a). This indicated that when nutrients were sufficient, the carbon source performs as the limiting factor which affected algal growth. Figure 1b shows that the medium pH had a significant increase on day 1 under both mid-CO 2 (from 8.83 to 9.41) and low-CO 2 (from 8.83 to 9.45) conditions, and from day 2 to 7, the pH of mid-CO 2 and low-CO 2 cultures gradually increased, whereas the pH of algal culture bubbled with high level of CO 2 strongly decreased on day 1 (from 8.83 to 7.07) and then gradually increased from 7.07 to 8.32. Faster growth and greater carbon fixation were observed in high-CO 2 cultured P. tricornutum under the low pH and high CO 2 condition. The consumption of NO 3 − during algal rapid growth may cause the increase of culture pH from day 2 to 7 under high CO 2 concentration, and this increase of pH might correlate with algal growth rate and carbon fixation efficiency. Compared with high-CO 2 culture, both mid-and low-CO 2 cultures had an increase of culture pH values associated with a relative slow growth after bubbled with mid and low levels of CO 2 from day 2 to 7.
As shown in Fig. 1c, d, both high-CO 2 and mid-CO 2 cells obtained high values of maximal photosynthesis system II (PSII) quantum yield (F v /F m ) and effective PSII quantum yield (YII) during cultivation, and they all firstly represented a significant increase of F v /F m and Y (II) values and then gradually decreased. The high-CO 2 cells consistently had the highest F v /F m and Y (II) values, followed by mid carbon cultured alga. Whereas, low CO 2 cultivated alga had the lowest F v /F m and Y (II) values, which obviously decreased during cultivation. This suggested that the response to CO 2 concentration in algal cells was also reflected in the changes in algal photosynthesis.
Lipid, fatty acid composition, total water-soluble proteins, chlorophyll a + c content and RNA concentration in algal cells measured under different CO 2 concentrations As described above, the concentration of CO 2 had significant effects on P. tricornutum growth. To investigate the influence of CO 2 concentrations on intracellular substances, we determined the lipid, total water-soluble proteins, and chlorophyll a + c content in P. tricornutum treated with different CO 2 concentrations. Table 1 shows that when grown under the high CO 2 condition, P. tricornutum had the highest lipid content (53.71 ± 2.41 %, w/w, % of dry cell weight (DCW)), followed by mid CO 2 cultivated cells of 35.87 ± 1.72 %, and low CO 2 cultured P. tricornutum had the lowest lipid content of 33.13 ± 1.21 %. As shown in Table 2, in low-and mid-CO 2 cultured cells, 7.41 % (% of DCW) and 8.72 % (% of DCW) are longer-chain fatty acids (≥20C), respectively. Whereas, in high-CO 2 cultured cells, the longer-chain fatty acids (≥20C) account in 16.41 % of dry cell weight. It was represented that a significant increase of longer-chain fatty acids content in high-CO 2 cultured cells, compared to low-and mid-CO 2 cultured cells. These results suggested  that the high-CO 2 cultured cells need much more cytosolic-generated reductant for lipids or longer-chain fatty acid synthesis. Table 1 exhibits the maximum water-soluble protein yield 16.29 ± 0.02 mg g −1 FW in P. tricornutum also occurred in the high CO 2 cultivated cells, while the mid-CO 2, and low-CO 2 cells had relatively lower soluble protein contents of 12.76 ± 0.04 and 7.63 ± 0.03 mg g −1 fresh weight (FW), respectively. Although high-CO 2 P. tricornutum cells had high lipid and water-soluble protein contents, the concentration of chlorophyll a + c and RNA were lower than those in mid CO 2 cultured cells (Table 1). Low-CO 2 cultured cells had the lowest yield of intracellular substances, which may have been due to the limiting of carbon source.
Assays of the change in enzyme activities P. tricornutum maintained in high CO 2 concentration had the highest growth rate among the three experimental groups (Fig. 1). To investigate the mechanism underlying the different growth rates due to CO 2 concentration, the activities of key enzymes in the Calvin cycle, including Rubisco, 3-phosphoglyceric phosphokinase (PGK), and phosphoribulokinase (PRK) were analyzed. As shown in Fig. 2, these enzymes showed maximum activity in P. tricornutum under the high CO 2 concentration, followed by mid-CO 2 cultured cells, which were slightly higher than low-CO 2 cultured cells, but no significant difference was observed in the activity of PRK between mid-CO 2 and low-CO 2 cultured cells (P > 0.05). These results suggest that the activity of the Calvin cycle was enhanced under high CO 2 cultivation, which was well matched with the high growth rate and photosynthetic performance of P. tricornutum in the high-CO 2 concentration.
In addition, alga cells grown under the high CO 2 concentration had a maximum yield of lipids (Table 1). This prompted us to investigate probable pathways which supplied NADPH for lipid synthesis, as NADPH is essential for elongation of fatty acids. To the best of our knowledge, in the photosynthetic cells of plants, de novo biosynthesis of 16-carbon and 18-carbon fatty acids occurs in the chloroplast, where NADPH is produced via the light reactions of photosynthesis. The 16-carbon fatty acids (e.g., palmitate) are elongated to yield the 20carbon and even longer-chain fatty acids in the cytosol, so these elongation steps are the only ones that require cytosolic-generated reductant [24].  The oxidative pentose phosphate pathway (OPPP) is the major resource of NADPH in the cytoplasm. To better understand the carbon flux and reductant source for lipid synthesis in P. tricornutum cells under three different CO 2 concentrations, the activities of G6PDH and 6PGDH, two key enzymes in the OPPP, were evaluated. Figure 3 shows the activities of G6PDH and 6PGDH, and these results showed that G6PDH and 6PGDH activities were highest under high-CO 2 cultivation, while algal cells maintained in mid-and low-CO 2 showed lower and lowest activity, respectively. These findings indicate that the activity of the OPPP was enhanced in cells under high-CO 2 cultivation and may have supplied NADPH for lengthening the fatty acids chain.
As the OPPP is not the sole producer of NADPH, to verify whether the other sources for NADPH could contribute to cytosolic fatty acids elongation, two tricarboxylic acid (TCA) cycle enzymes activity were detected. Figure 4 shows that no significant differences were observed in the activity of MDH of three different CO 2 concentration cultured P. tricornutum cells. Whereas, the activity of α-ketoglutarate dehydrogenase (α-KGDH) was peaked in high-CO 2 cultivated P. tricornutum cells, followed by mid-CO 2 cells, and the low-CO 2 cultivated cells had the lowest α-KGDH activity. As α-KGDH is the rate-determining enzyme of the TCA cycle [25], the activity of α-KGDH was positively correlated with TCA cycle activity. High α-KGDH activity indicated a general enhancement of the TCA cycle under high CO 2 concentration.

Analysis of enzyme mRNA expression measured by qPCR
The results of real-time PCR showed that under high CO 2 concentration, P. tricornutum had the highest relative expression level of Rubisco, PRK, PGK, G6PDH, and 6PGDH among the three treatments (Figs. 5 and 6). The expressions of Rubisco, PGK, and G6PDH in mid-CO 2 cultivated P. tricornutum cells were higher than those of the low-CO 2 cultured cells, but the expression of PRK and 6PGDH in mid -CO 2 cultivated P. tricornutum was not significantly different to that found in low-CO 2 cultured cells (P > 0.05) The mRNA expression level of these enzymes was consistent with the results of enzyme activity, which indicated that P. tricornutum under high CO 2 cultivation not only had high enzyme activities but also high mRNA content. Higher enzyme activities and mRNA expression levels of Rubisco, PGK, and PRK suggested that, in Calvin cycle, the CO 2 fixation, carbon reduction, and ribulose-1,5-bisphosphate (RuBP) regeneration were enhanced; in other words, the carbon fixation efficiency of Calvin cycle elevated and the production of carbon skeleton for biosynthesis increased. These might positively correlate with rapid growth of P. tricornutum under high CO 2 concentration.

Discussion
As described in Fig. 1, P. tricornutum can tolerate relatively high levels of CO 2 and has a rapid growth rate when cultivated in high CO 2 concentration, which was consistent with the finding that rising CO 2 concentration increased phytoplankton growth rate [1,8,9,12,26]. Moreover, Rubisco activity and mRNA expression were increased (Figs. 2 and 5) under high CO 2 concentration. As the rate-limiting and key regulatory enzyme of the Calvin cycle, the level of Rubisco monitors the carbon flux through this pathway [27]. In addition to high Rubisco activity, the activities of PRK and PGK were high in high CO 2 concentration cultured cells and significantly higher than Rubisco activity. Rubisco acts as a CO 2 receptor and takes part in the carbon fixation stage of the Calvin cycle, and PGK participates in the reduction stage of the Calvin cycle, whereas PRK is incorporated in the regeneration phase in the Calvin cycle; high activities of Rubisco, PGK, and PRK in high CO 2 cultured cells indicated that the Calvin cycle activity was enhanced and was operating effectively in algal cells under high CO 2 cultivation. This may account for the rapid growth rate observed. However, when P. tricornutum was grown under low CO 2 concentration, the activities of the Calvin cycle-related enzymes were weaker, which suggested that substrates and energy required for algal growth were deficient. This might be the major reason why no significant growth under the low CO 2 concentration of algal cells was observed. These results indicated elevated CO 2 concentration provided plenty of carbon resource for photosynthetic carbon fixation and then induced an enhancement of algal photosynthetic carbon fixation and growth rate which were consistent with numerous previous studies that higher levels of CO 2 support higher growth rates.
As expected, a significant increase in lipid content was also found in the high CO 2 cultivated rapid-growth P. tricornutum cells. As we know, microalgal lipid production typically begins at the onset of nitrogen depletion as the C/N ratio (carbon to nitrogen ratio) in algal cells increased, which lead to most of the fixed carbon divert into nitrogen deficient compounds such as lipids [28]. In this research, P. tricornutum were maintained in nitrogen repletion cultures, which might not trigger lipid accumulation. Yet we observed 1.5-fold increase of lipid content in high-CO 2 cultured cells, compared to mid-CO 2 cultured cells. We thought that maybe because of a higher C/N ratio in algal cells that caused by high CO 2 concentration. This is supported by the conclusion that the elemental stoichiometry (carbon to nitrogen ratio) was raised under high CO 2 conditions in both nitrogen limited and nitrogen replete conditions in P. tricornutum according to Li et al. [29].
Lipids, especially triacylglycerol (TAG), primarily serve as a storage form of carbon and energy in algal cells [30], which require numerous NADPH molecules for biosynthesis in the cytoplasm. During rapid growth of algal cells, ATP and NADPH generated by the light reaction are mainly supplied for the maintenance of algal growth. Therefore, under high CO 2 cultivation, the key problem in the rapid growth of algal cells is how the additional NADPH required for lipid synthesis comes from. NADPH can be primarily generated from the OPPP or TCA cycle. The first two steps of the OPPP, catalyzed by G6PDH and 6PGDH respectively, are the major source of NADPH. As shown in Fig. 6, with a high level of CO 2 concentration, enhanced activity and mRNA expression of G6PDH and 6PGDH were observed, except for increased lipid accumulation, which indicated that OPPP activity was enhanced. In a study of Chlorella, Xiong et al. [31] found that NADPH supplied through OPPP increased under nitrogen limited condition when lipid synthesis increased, and they speculated that NADPH from the mitochondrial TCA cycle was difficult to be trapped for lipid metabolism in the cytoplasm. The enhancement of TCA cycle might contribute to two possibilities: (i) to provide ATP for lipid synthesis; (ii) to produce quantity of intermediates (e.g., malic acid) to offer substrate or carbon skeleton for fatty acid synthesis [32]. Thus, G6PDH together with 6PGDH that are of great importance in the synthesis of the reductant NADPH through the OPPP [33], may play an important role in elongation of fatty acids under high CO 2 concentration.

Conclusions
In this study, growth, lipid content, photosynthetic performance, the activity, and expression of key enzymes in Calvin cycle and OPPP were analyzed in the marine diatom P. tricornutum under three different CO 2 concentrations. Both the growth rate and lipid content of P. tricornutum increased significantly under the high CO 2 concentration. Enzyme activity and mRNA expression of three Calvin cycle-related enzymes (Rubisco, PGK, PRK) were also increased under high CO 2 cultivation, which suggested the enhancement of the Calvin cycle activity. This may account for the observed rapid growth rate. In addition, high activity and mRNA expression of G6PDH and 6PGDH, which produce NADPH through OPPP, were observed in high CO 2 cultured cells. These results indicate OPPP was enhanced and might play an important role in elongation of fatty acids or lipid synthesis under high CO 2 concentration.

Methods
Alga strain cultivation P. tricornutum was screened from the East China Sea and confirmed unialgal as previously described by Wu et al. [34]. The alga strain was cultivated in sterilized carbon source deprived artificial seawater enriched with f/2-Si medium [35] and grown at 20 ± 1°C under a constant light intensity of 100 μmol m −2 s −1 with a 14:10 h light-dark (L/D) cycle.

CO 2 treatment
The alga culture was maintained with aeration and cultivated in 2-L flasks containing 1.5-L medium. Three different CO 2 concentrations in the bubbling aeration system were tested: low CO 2 (0.015 %), mid CO 2 (atmospheric, 0.035 %), and high CO 2 (0.15 %). The handheld indoor air quality testing detector Telaire 7001 CO 2 Monitor (USA), which can be able to display CO 2 readings in less than 30 s, was used to measure CO 2 concentration in this study. Each treatment included three replicates. The pH of algal culture bubbled with different CO 2 concentration was daily measured using a Switzerland Mettler Toledo Delta 320 pH meter.

Growth and photosynthetic determination
Algal growth was measured daily by cell number and the OD value at the optical wavelength for P. tricornutum (730 nm) using a spectrophotometer (UV-1800). A hemocytometer was used to count algal cells.
To investigate the photosynthetic performance of P. tricornutum in response to CO 2 concentration, the chlorophyll fluorescence of PSII was measured by a Dual-PAM-100 fluorometer (Heinz Walz, Effeltrich, Germany) which was conducted with a PC using the WinControl software (Heinz Walz). Two parameters, the maximum quantum yield of PSII (F v /F m ) and the effective PSII quantum yield (YII), were evaluated using the pulse-amplitude modulated method as described by Lin et al. [36] with some modifications. To ensure the consistency of cell density for PAM fluorescence measurement, the growing algal samples were adjusted to 0.6 at OD 730 nm before detected. The minimum fluorescence (F 0 ) and maximum fluorescence (F m ) were determined after the samples were incubated in darkness for 5 min.
Variable fluorescence (F v ) was calculated according to formula F v = F m −F 0 . The maximal PSII quantum yield was calculated as F v /F m . The effective quantum yield of PSII (Y (II)) was calculated as (F m ' −F)/F m ' , where F m ' was the maximal fluorescence detected under photosynthetic active radiation (PAR) of 58 μmol m −2 s −1 and F is the real-time fluorescence yield under illumination.

Lipid analysis
Algal cultures were harvested after 7-day cultivation by centrifugation at 5000g, and the algal cells were quickly frozen in liquid nitrogen. A freeze dryer was used to dry the frozen algal cells. Lipids were extracted from 50 mg dryness algal powder using a modified chloroformmethanol system as described by Bligh and Dyer [37]. Crude samples were dried with N 2 flow until a constant weight was obtained. Gravimetric means, a conventional lipid quantification method, was used for total lipid analysis. The fatty acid compositions of different CO 2 concentration cultivated P. tricornutum were determined by gas chromatography [38].

Enzyme assays
Harvested algal cells were washed twice with 0.01 M PBS buffer (pH 7.4) and were ground into fine powder using a pre-chilled mortar and pestle with liquid N 2 . Crude enzyme extracts were prepared from the algal powder in three times the volume of ice-cold extraction buffer. Enzyme activities were determined spectrophotometrically using a UV-1800 spectrophotometer by measuring at 340 nm in total volumes of 0.4 mL and in triplicate. The change in absorbance was recorded for 5 min. Results were expressed as μmol NAD (P) H oxidation or NAD (P) + reduction min −1 g −1 AFDW, as we had standardized the correlation between fresh weight and ash-free dry weight of different CO 2 cultured P. tricornutum. The compositions of the extract and assay medium for the respective enzymes are detailed below. All chemicals and enzymes used for enzyme assays were purchased from Sigma-Aldrich (Sigma-Aldrich Co. LLC., USA).
Rubisco (ribulose-1, 5-bisphosphate carboxylase/oxygenase), which controls the initial step in photosynthetic carbon fixation via the Calvin cycle, is one of the most abundant proteins in both higher plants and algae [39]. The carboxylase activity of Rubisco relative to photosynthetic rates has been reported in many previous studies of macroalgae and microalgae [27]. Rubisco (EC 4.1.1.39) activity was measured using a modification of the basic procedure according to Gerard and Driscoll [27]. 300 μL of chilled extraction buffer (40 mM Tris-HC1, 0.25 mM EDTA, 10 mM MgCl 2 , 5 mM glutathione, at pH 7.6) was added per 100-mg fresh algal powder. The mixture was stirred for 2 min to homogeneity, and centrifuged at 13,000g for 10 min. The supernatants were incubated on ice for further enzyme assays. The reaction mixture contained 0.1 M Tris-HCl (pH 7.8), 0.2 mM NaHCO 3 , 12 mM MgCl 2 , 5 mM NADH, 50 mM ATP, 50 mM phosphocreatine, 160 units per milliliter of creatine phosphokinase (EC 2.7.3.2), and PGK (EC 2.7.2.3), and 25 mM RuBP was added to initiate the reaction.
For measurement of PRK (EC 2.7.1.19) and PGK activities in crude extracts, the method described by Rao and Terry [40] was used. Frozen fresh algal cells were extracted in three times the volume of extraction buffer (100 mM Hepes-NaOH (pH 8.0), 10 mM MgCl 2 , 0.4 mM EDTA, 1 % polyvinypyrrolidone, 100 mM Naascorbate, 0.1 % BSA at 0-4°C) and then stirred for 2 min with a vortex at a maximum speed. Crude homogenates were then centrifuged at 13,000g for 10 min. The supernatants were prepared for further assays. The oxidation pentose phosphate pathway is the major resource of cytosolic-generated NADPH, which is catalyzed by G6PDH and 6PGDH. G6PDH, the rate-limiting and key regulatory enzyme of the OPPP, controls the flow of carbon through this pathway and produces reductant to meet the cellular needs for reductive biosynthesis. To determine the activities of G6PDH (EC 1.1.1.49) and 6PGDH (EC 1.1.1.44), the method described by Fahrendorf et al. [41] was used with some modifications. Algal cells prepared as described above were extracted in 50 mM Tris-HC1 pH 8.0, 300 mM NaC1, 0.1 mM benzamidine, and 0.1 mM phenylmethylsulfonyl fluoride (PMSF). G6PDH activity was measured in a continuous assay at 30°C, including 0.1 M Tris-HC1 pH 8.0, 0.4 mM NADP + , and 3 mM glucose-6-phosphate. 6PGDH activity was measured at 30°C in 0.1 M Hepes pH 7.5, 0.4 mM NADP + , and 3 mM 6-phosphogluconate. Reactions were started by the addition of algal extracts. Reduction of NADP + at 340 nm was used to monitor G6PDH and 6PGDH activities.
TCA cycle is another major resource of NADPH, via isocitrate dehydrogenase. To verify whether this process contributes reductant to fatty acid elongation, the activity of two TCA cycle enzymes, α-KGDH and MDH was measured following the method as described by Peng L and Shimizu K [42]. Algal cells were extracted in 100 mM Tris-HCl (pH 7.0) containing 20 mM KCl, 5 mM MnSO4, 2 mM DTT, and 0.1 mM EDTA. The assay conditions were as follows: α-KGDH 0.2 M phosphate buffer (pH 7.2), 1 mM CoASH, 0.1 M cysteine-HCl (pH 7.2), 10 mM NAD+ (pH 7.2), 3 mM a-ketoglutarate. Malate dehydrogenase, 0.1 M Tris-HCl (pH 8.8), 0.1 mM sodium malate, 10 mM NAD+, and reactions were started by the addition of algal extracts.

Real-time PCR analysis
Total RNA was extracted from frozen algal powder using the Plant RNA kit (OMEGA, Norcross, USA) following the manufacturer's instructions which included the first step of genomic DNA digestion. The concentrations of RNA extracts were measured using a NanoDrop 1000 Spectrophotometer (Thermo, Wilmington, Delaware, USA).
For single strand cDNA synthesis, the PrimeScript RT regent Kit with gDNA Eraser (TaKaRa Biotech Co., Dalian, China) was used to perform the reverse transcription reaction according to the user's manual. In this protocol, further genomic DNA removal was performed to purify the RNA extracts. All tests were performed on ice. PCR was performed to confirm the absence of genomic DNA contamination.
Real-time PCR was performed with the cDNA template from the reverse transcription reaction and five pairs of specific primers (Table 3). Primers were designed based on the alignment of the deduced amino-acid sequences, which were obtained from the National Center for Biotechnology Information (NCBI) web site (http://www.ncbi.nlm.nih.gov/ blast), using the Primer Premier 5.0 software. Levels of specific mRNA transcripts were quantified by the Bio-Rad iQ5 Multicolor Real-Time PCR Reaction system (Bio-Rad, Hercules, CA, USA). The real-time PCR amplifications were performed with the reagents from the Fast Essential DNA Green Master (Roche, Germany). The cycling parameters for real-time PCR were 95°C for 10 min (denaturation), followed by 40 cycles at 94°C for 10 s (denaturation), 58°C for 20 s (primer annealing), and (c) 72°C for 10 s (elongation). To ensure only the single specific DNA fragment was amplified, melting curve analyses were performed on all PCR products. Triplicate qPCRs were performed for each sample. Data derived from the PCR program were analyzed with the Bio-Rad optical system software.

Chlorophyll estimations
For P. tricornutum, the chlorophyll a and c1 + c2 content was estimated by a 90 % acetone extraction system. Fifty-milligram fresh algal powder ground from frozen algal cells in liquid N 2 was added in 5 mL 90 % acetone and thereafter read the fluorescence emission of the centrifuged extracts at 664 and 630 nm which were obtained as described by Jeffrey and Humphrey [43].

Statistical analysis
All data are expressed as the mean ± standard deviation (SD) of three independent experiments. All statistical analyses were conducted using SPSS, and t test was used to identify data of any significance within the treatments at P value ≤ 0.05.