Identification of the pyruvate transporter-plastid type (PtPTP) gene in P. tricornutum
Based on previous studies on the plastidial pyruvate transporter in plants, we assessed the pyruvate transporter-plastid type gene in P. tricornutum. To identify the gene encoding the pyruvate transporter plastid-type in P. tricornutum, we first performed a sequence similarity search using BLAST analysis in the NCBI and JGI databases for P. tricornutum (https://mycocosm.jgi.doe.gov/Phatr2/Phatr2.home.html). The protein similarity test to the Arabidopsis thaliana plastidial pyruvate transporter revealed that AtBASS2 (At2g26900), the candidate gene of PHATRDRAFT_3046 (GenBank XP_002179421) showed the highest similarity (49% protein identity and 65% protein similarity). However, this is only a partial gene sequence that lacks the start codon. We, therefore, cloned the full-length pyruvate transporter (PtPTP) sequence using 5′- and 3′-RACE analysis from the corresponding cDNA (Fig. 1a). In the genomic sequence of P. tricornutum, PtPTP has four exons, three introns, a 183 bp 5′-UTR, and a 99 bp 3′-UTR. The entire coding sequence (CDS) of PtPTP is 1311 bp. The complete amino acid sequence of PtPTP showed high similarity (62% and 61%, respectively) to that of other BASS2 proteins of the dicotyledonous C3 plant A. thaliana (At2g26900, NP850089) and C4 plant Flaveria trinervia (BAJ16226) (Fig. 1b). PTP-related protein sequences were derived from the NCBI database using BLAST with PtPTP, and a phylogenetic tree was constructed using the neighbor-joining method in the MEGA 10 software [37]. The phylogenetic tree indicated that PtPTP sequences have high similarity with those of the centric marine diatom Thalassiosira pseudonana and generate one cluster with the polar diatom Fragilariopsis cylindrus; they belong to different clades from those of higher plant groups and microorganisms (Fig. 1c). The characterization of the PtPTP sequence implied that the protein is conserved with high similarity and is closely related to those of other diatoms. With this transporter, we attempted to regulate the flux of pyruvate in metabolic pathways and engineer P. tricornutum. Its subcellular localization was further analyzed in order to clarify the function of the gene as a transporter.
Subcellular Localization of PtPTP
To determine the function and structure of the PtPTP protein, we examined the transmembrane helices of the protein using TMHMM Server v. 2.0 (http://cbs.dtu.dk/services/TMHMM/). Examination revealed that PtPTP is a transmembrane protein that has 9 transmembrane helicase regions (Fig. 2a). Based on other prediction programs TargetP and ChloroP, the intracellular location of PtPTP protein was expected to be localized to chloroplasts and contain a signal peptide. TargetP indicated a higher probability (0.471) of the presence of a chloroplast transit peptide (cTP) than others (secretory pathway signal peptide [SP]: 0.446, mitochondria targeting peptide [mTP]: 0.052) and predicted a potential presequence length of 63 amino acids. ChloroP also indicated that PtPTP had a 0.581 probability of the presence of a cTP with a length of 63 amino acids.
To determine the subcellular localization of PtPTP, we generated transformants expressing enhanced yellow fluorescence protein (EYFP) and PtPTP::EYFP (EYFP fused to the C-terminus of PtPTP) in P. tricornutum. PtPTP-EYFP fluorescence was detected in the chloroplast in the transgenic strains using fluorescence microscopy; plastid auto-fluorescence (PFA) indicated subcellular localization to the chloroplast. This corresponded precisely to the PtPTP-EYFP fluorescence. In contrast, in transformants expressing EYFP, the EYFP signal was dispersed in the cytoplasm (Fig. 2b). Based on the results of prediction and observation, the location and topology of PtPTP indicated it to be a plastid-localized transmembrane protein. Therefore, we could regulate the flux of pyruvate in plastids of P. tricornutum by manipulating the gene of pyruvate transporter localized in plastids.
Generation of transformants overexpressing PtPTP driven by a constitutive promoter
We previously constructed the pPhatT-EF2 expression vector harboring the elongation factor 2 (EF2) promoter from P. tricornutum for the generation of an overexpressing transgenic target gene [21]. The EF2 promoter triggers constitutive gene expression regardless of light/dark cycles [21]. To construct a transgenic transformant that causes continuous PTP gene overexpression using the EF2 promoter, a transformation vector (pPhaT-EF2-PtPTP) capable of expressing an exogenous PtPTP coding sequence was constructed between the EF2 promoter and FcpA terminator of the pPhaT-EF vector (Fig. 3a). With this transformation vector, wild-type P. tricornutum cells were transformed using the particle delivery system and grown on selective media containing Zeocin. Next, 11 colonies were collected from a single transformation event using 1 μg of the linearized pPhaT-EF2-PtPTP digested with restriction enzyme PsiI and KpnI. Transformed cells were then screened using genomic PCR (Fig. 3b) to confirm the successful integration of the EF2::PtPTP expression cassette. For the positive control, the transformation vector (VC) was used as the PCR template. Amplification of genomic DNA by PCR using the primers EF2-iFw and PTP-iRv, flanking the EF2 promoter region and PtPTP, revealed a 0.51 kb PCR product only in the transgenic strains (PtPTP-OE1 and OE2) (Fig. 3b). To identify the presence of both native endogenous PtPTP and the integrated EF2::PtPTP cassette in the genome, PCR using genomic DNA was performed with a specific primer set, PTP-iFw and PTP-iRv (Fig. 3a), which binds to the first and third exon regions containing two intron regions. The wild type and transgenic strains presented large PCR bands (0.53 kb) containing 0.26 kb intron portions (Fig. 1a). However, the small PCR bands (0.26 kb) containing only the exon parts were shown only in transgenic strains harboring the exogenous coding sequence of PtPTP. When PCR was performed using primers of an internal transcribed spacer (ITS) gene as an internal control, the presence of the 0.98-kb PCR product was confirmed in both the wild-type strains and the transformants. These genomic DNA PCR results indicated that the exogenous PtPTP expression cassette of a transformation vector was successfully integrated into the genome of the transgenic strains.
Gene expression levels of the PtPTP gene were assessed by real-time qRT-PCR in the wild-type cells and two transformant cells cultured for 5 d after subculture at the transit point between the exponential to stationary phase. The relative transcript abundances of the transformants, PtPTP-OE1 and PtPTP-OE2, were about 2.2 and 1.4 times higher than that of the wild type cells, respectively (Fig. 3c). This result demonstrated that two transgenic strains exhibited overexpressed levels of the PtPTP transcript compared to the wild type.
To confirm whether the increase in the expression level of the PtPTP transcript affects the amount of PtPTP protein, we assessed the relative amount of the PtPTP protein by immunoblot analysis using an antiserum against two specific synthetic peptides of PtPTP. (Fig. 3d). The relative amount of the PtPTP protein in the two transformants to that in the wild-type was examined using the anti-PtPTP antibody. The antibody reacted with a 40-kDa band corresponding to the predicted weight of PtPTP, with the deletion of 63 amino acids of the chloroplast transit peptide. The relative amounts of PtPTP protein increased in both PtPTP-OE1 (254%) and PtPTP-OE2 (294%), compared to that in the wild type (Fig. 3d). Finally, we successfully generated two transformants overexpressing the plastid-type pyruvate transporter. We expected that overexpressing plastidial pyruvate transporter encourages the influx of pyruvate into plastids and pyruvate metabolism in plastids. Thus, we further analyzed the phenotypes related to biomass and lipid production in the two transformants.
The effects of overexpressing PtPTP on cell growth and their biomass production
The effects of overexpressing PtPTP in P. tricornutum were investigated by the analysis of cell growth and biomass production (Fig. 4). After inoculation with wild type and transgenic strains at the same cell density of ~ 0.8 × 106 cells/mL (OD750 ~0.05), we measured cell growth as a function of cell density. After subculture, their growth pattern showed exponential growth up to 5 days of subculture. Subsequently, the cells entered the stationary phase. Here, the number of wild type cells reached 8.28 ± 0.06 × 106 cells/mL and those of transgenic PtPTP-OE1 and PtPTP-OE2 were 8.58 ± 0.11 and 8.64 ± 0.10 × 106 cells/mL, respectively, which increased by 3.6% and 4.3%. After 3 days of subculture, the transgenic strains began to show a difference from the wild type in cell growth. However, the growth rate and cell density showed some differences compared to those of the wild type from the subculture. Independently repeated experiments (n=3) of cell density showed the significantly increased cell density of PtPTP-OE1 from 5.5% (5 days, p < 0.01) to 3.3% (7 days, p < 0.01) and PtPTP-OE2 from 8.0% (4 days, p < 0.05) to 4.3% (7 days, p < 0.01) compared to that of the wild type (Fig. 4a).
The effects of PtPTP overexpression on biomass production were analyzed. During cell cultures in the flask, the biomass was estimated based on dry cell weight per culture volume at the exponential growth phase of 3 days, the declining growth phase of 5 days, and the late stationary phase of 7 days subculture (Fig. 4b). Compared to the wild type strain, transgenic PtPTP-OE1 and PtPTP-OE2 produced significantly more biomass at 5 and 7 days after inoculation. In the declining growth phase (5 days after inoculation), PtPTP-OE1 and PtPTP-OE2 produced 21.9 ± 0.01% and 20.8 ± 0.03% more biomass, respectively. At 7 days of the stationary phase, PtPTP-OE1 exhibited a 16.6 ± 0.04% increase in biomass and PtPTP-OE2 exhibited a 13.6 ± 0.03% increase in biomass. As a result, PtPTP-overexpressing strains showed higher cell density and more biomass production compared to that of their wild-type counterparts. Therefore, these engineered microalgae could serve as strains for efficient biomass production.
Comparing photosynthetic productivity between transformants and wild type, the photosynthetic productivities of transformants were similar to that of the wild-type cells. This data indicated that an increased influx of pyruvate to plastids does not affect photosynthetic productivity (Additional file 1: Fig. S1). Hence, we purported that enhanced biomass production in transformants may be derived from an increased influx of pyruvate and activated biosynthetic pathways in plastids.
The effects of overexpressing PtPTP on lipid contents and fatty acids composition
The total lipid content per dry cell weight showed considerable accumulation at 7 days of the stationary phase as determined by the weighting method. PtPTP-OE1 and PtPTP-OE2 contained 111±2% and 130±5% of total lipids, respectively, compared to the wild type (Fig. 5a). Furthermore, FAME analysis using GC-MS revealed fatty acid composition in the wild-type and transgenic strains (Fig. 5b). A broad distribution of the fatty acids in the wild-type cells is similar to that of PtPTP-overexpressing transformants; however, there was a significant difference between the sum of polyunsaturated fatty acids (SUM PUFA) in each transformant. PtPTP-OE1 and PtPTP-OE2 (47.44% and 49.51%, respectively) had a higher PUFA content compared to the wild type organism (43.39%). The high polyunsaturated fatty acid content in microalgae provides an advantage for the utilization of microalgal biomass for nutraceutical applications, pharmaceutical applications, and biodiesel production [38,39].
Transcriptional expression of genes related to pyruvate metabolism in plastids
Based on the intracellular distribution of metabolic enzymes in P. tricornutum, several enzymes that utilize pyruvate as a substrate participate in pyruvate metabolism in plastids, such as pyruvate carboxylase 2 (PYC2), pyruvate-phosphate dikinase (PPDK) and pyruvate dehydrogenase complex (PDC) [40,41]. PYC2 and PPDK could be involved in amino acids and lipids biosynthesis, producing oxaloacetate (OAA) and phosphoenolpyruvate (PEP), respectively [40,42,43]. PDC is an essential enzyme for de novo fatty acid biosynthesis in plastids [44]. PDC converts pyruvate to acetyl-CoA, generating NADH and CO2, and acetyl-CoA is a significant substrate for de novo fatty acid biosynthesis. Comparing the expression of their genes between wild type and transformants could help to estimate the pool and fate of pyruvate in plastids. Using quantitative real-time PCR, we compared the expression level of PYC2 (GenBank XM_002183870), PPDK (GenBank XM_002182336) and PDC (pyruvate dehydrogenase complex; subunit E1, GenBank XM_002180298) between wild type and transformants at 3, 5, and 7 days after subculture (Fig. 6). At the 3 days after subculture, relative transcription levels of PDC did not show any difference between WT and transformants. PPDK transcript level in PtPTP-OE2 was up-regulated (1.63-fold), and its PYC2 transcript level was down-regulated (-1.33-fold). At the 5 days after subculture, PDC transcript levels still did not show significant difference in transcription levels between WT and transformants (Fig. 6, 5 days of subculture). Transcription levels of PYC2 were down-regulated in PtPTP-OE1 (-1.49-fold) and PtPTP-OE2 (-1.66-fold), and those of PPDK were up-regulated in PtPTP-OE1 (2.93-fold) and PtPTP-OE2 (1.50-fold) (Fig. 6, 5 days of subculture). In the stationary phase (7 days after subculture), transcriptional expression levels of PDC were up-regulated in PtPTP-OE1 (2.54-fold) and PtPTP-OE2 (1.54-fold) (Fig. 6, 7 days of subculture). Transcriptional expression levels of PPDK in WT is similar with those in transformants, and PYC2 transcript levels are up-regulated in transformants compared to WT (PtPTP-OE1 1.46-fold, -OE2 1.39-fold) (Fig. 6, 7 days of subculture). As a result, increased biomass production in transformants might come from activated PPDK-related amino acids and lipid biosynthesis [30,42], instead of PYC2-related biosynthesis until stationary phase. At the stationary phase, enhanced biomass and lipid production in transformants could be explained with up-regulated PDC in de novo fatty acid biosynthesis, and PYC-related biosynthesis might also contribute their increased biomass production in stationary phase. These transcriptional analysis data is correlated with enhanced biomass production and fatty acid biosynthesis in PtPTP-overexpressing strains.
Based on these results, transformants showed increased pyruvate availability and its metabolism by overexpression of pyruvate transporter in plastids. Finally, we determined that an increased influx of pyruvate into plastids by overexpression of PtPTP could enhance their biomass and lipid production, making this an attractive strategy to engineer microalgae.