PUFA-synthase-specific PPTase enhanced the polyunsaturated fatty acid biosynthesis via the polyketide synthase pathway in Aurantiochytrium

Background Phosphopantetheinyl transferase (PPTase) can change the acyl-carrier protein (ACP) from an inactive apo-ACP to an active holo-ACP that plays a key role in fatty acids biosynthesis. Currently, the PPTase has been proved to be involved in the biosynthesis of polyunsaturated fatty acids (PUFAs) via a polyketide synthase (PKS) pathway in Thraustochytrids, while its characteristics are not clarified. Results Here, the heterologous PPTase gene (pfaE) from bacteria was first co-expressed with the PKS system (orfA–orfC) from Thraustochytrid Aurantiochytrium. Then, a new endogenous PPTase (ppt_a) in Aurantiochytrium was identified by homologous alignment and its function was verified in E. coli. Moreover, the endogenous ppt_a was then overexpressed in Aurantiochytrium, and results showed that the production and proportion of PUFAs, especially docosahexaenoic acid (DHA), in the transformant SD116::PPT_A were increased by 35.5% and 17.6%, respectively. Finally, higher DHA and PUFA proportion (53.9% and 64.5% of TFA, respectively) were obtained in SD116::PPT_A using a cerulenin feeding strategy. Conclusions This study has illustrated a PUFAs-synthase-specific PPTase in PKS system and provided a new strategy to improve the PUFA production in Thraustochytrids.


Background
Polyunsaturated fatty acids (PUFAs), especially docosahexaenoic acid (DHA, 22:6 ω3) and eicosapentaenoic acid (EPA, 20:5 ω3), are rapidly gaining attention, due to their beneficial effects in the cognitive development of infants and their use to reduce the risk of hypertension, cardiovascular diseases, inflammation, and certain cancers [1][2][3][4]. Currently, the major commercial source of PUFAs is fish oil; however, several factors such as the reduction of marine fish source, the increasing environmental pollution, and undesirable fishy flavor limit the supply of high-quality PUFAs [5,6]. Therefore, alternative sources of high-quality PUFAs, especially DHA and EPA, have drawn increasing amounts of attention. Heterotrophic thraustochytrids, are capable of accumulating large of lipids, have been increasing embraced by the market [7,8].
As it is known, PUFAs are synthesized by two pathways, the aerobic desaturase/elongase pathway and the

Open Access
Biotechnology for Biofuels Full list of author information is available at the end of the article anaerobic polyketide synthase (PKS) pathway [9][10][11]. Compared with the desaturase/elongase pathway, PKS pathway directly synthesizes the PUFAs from acetyl-CoA and malonyl-CoA substrates with a few intermediates. The PKS pathway is catalyzed by an enzyme complex, and the gene codings for the members of this complex are composed of three or four subunits and possess similar domain structures including β-ketoacyl synthase (KS), chain length factor (CLF), malonyl-CoA transacylase (MAT), acyl-carrier protein (ACP), ketoacyl reductase (KR), acyltransferase (AT), enoyl reductase (ER), and dehydratase (DH) domains ( Fig. 1) [8,[12][13][14]. In addition to the above domains, a phosphopantetheine transferase (PPTase) is also reported to be essential for PUFA synthesis [15]. ACP domain in PKS needs a posttranslational modification to become enzymatically active. PPTase transfers the pantetheine moiety from Coenzyme A to a conserved serine residue of an inactive carrier protein to produce its active form [16,17]. The gene encoding that PPTase (PfaE) has been identified in several bacterial PKS systems, such as Shewanella oneidensis [18] and Moritella marina [19], and has been proved to be essential for PUFA synthesis. The putative PPTase encoding gene was also identified in the Thraustochytrid species, such as Aurantiochytrium limacinum [20] and Hondaea fermentalgiana [7]; however, its characteristics are not clearly clarified.
PPTases can be subdivided into three groups according to their primary sequence and substrate specificity [19,21,22]. The name-giving prototype of the first group is AcpS from Escherichia coli. The AcpS-type PPTases are 120-140 amino acids in size, act as homotrimers, are associated with primary metabolism in prokaryotes, and accept only ACPs from FAS and ACPs of type II PKS. The second group of PPTases, which can be found in eukaryotes, represents integrated domains of their cognate FAS.
PPTases of the third, Sfp-type PPTases, are about twice the size of AcpS, have an extraordinarily broad substrate spectrum, and are active as monomers. Sfp has proven to recognize every carrier protein (CP) tested so far, including PCPs of NRPSs and ACPs of FASs and PKSs [16,19].
Heterotrophic thraustochytrid, Aurantiochytrium sp., has a strong ability to produce biomass and lipid. Particularly, it can produce large amounts of DHA, which comprises nearly 50% of the total fatty acid (TFA) proportion [23][24][25]. Therefore, Aurantiochytrium has become one of the major alternative sources for commercial DHA production.
It is well known that there are two fatty acid pathways in Aurantiochytrium [4,26]. The conventional fatty acid synthase (FAS) pathway is mainly responsible for the synthesis of saturated fatty acids (SFAs), while PUFAs including DHA and docosapentaenoic acid (DPA) in Aurantiochytrium are synthesized via the PKS pathway [4,27]. This PKS system has been confirmed to contain three genes, called orfA, orfB, and orfC ( Fig. 1). However, no PPTase gene or PPTase function domain was contained in this PKS system [13,28].
In the present work, we first confirmed that PPTase was involved in and essential for the PKS system in Aurantiochytrium and that the efficiency of DHA synthesis was influenced by overexpression of PPTase. Then, a new endogenous PPTase was identified and overexpressed in Aurantiochytrium sp. SD116 to strengthen the PKS system resulting in the improvement of DHA production. The proportion of PUFAs, especially the DHA proportion, had significantly increased in the recombinant SD116::PPT_A strain. The PUFAs proportion was further enhanced by depressing the FAS pathway. Based on this study, a PUFAs-synthase-specific PPTase in PKS system was identified and a new strategy for high PUFA production in Aurantiochytrium was provided.

PPTase is essential for the PKS system
In thraustochytrid Aurantiochytrium, the so-called PKS, can anaerobically synthesize PUFAs de novo. The PKS organization consists of three subunits OrfA, OrfB, and OrfC ( Fig. 1). To verify all components of PKS in Aurantiochytrium, orfA, orfB, and orfC were cloned and successfully co-expressed in E. coli. GC-MS was used to analyze the lipid profiles of E. coli, and the result showed that no DHA was produced by the recombinant strain orfABC (Fig. 2), suggesting that the heterologous expression of PKS system is not intact. The PKS organization in Aurantiochytrium possesses AT, multiple ACPs, MAT, KS, KR, DH, ER, and CLF domains, while no PPTase encoding gene or PPTase domain was identified. The previous studies showed that PPTase is essential for PUFA production in several marine bacteria, and the recombinant product of EPA and DHA in E. coli was achieved using the intact PKS gene cluster which contains the PPTase encoding gene from Shewanella and Moritella marina, respectively [29,30]. Therefore, we deduced that the PPTase may be also essential for the PKS system of Aurantiochytrium. pfaE gene, which encodes a PPTase in Shewanella, was cloned and co-expressed with orfA, orfB, and orfC in E. coli, and the result showed that the recombinant DHA and DPA were produced (Fig. 2). Based on this result, we confirmed that PPTase is involved in and essential for the PKS system in Aurantiochytrium.
As functional expression of the PKS requires posttranslational activation by a PPTase, overexpressing PPTase may be beneficial for PUFA synthesis in Aurantiochytrium. Then, the pfaE gene from Shewanella was overexpressed in Aurantiochytrium to attempt to improve the DHA production. As shown in Additional file 1, the plasmid pWE-PfaE containing the pfaE expression cassette driven by a tubulin promoter (Ptub) and actin transcription terminator (Tactin) was transformed into Aurantiochytrium sp. SD116, and the correct transformants named SD116::PfaE were screened by the selection stress of zeocin and further confirmed by genomic PCR. A 0.9-Kbp DNA band was detected in the transformant, but not in SD116 (Additional file 2), suggesting that pfaE was successfully inserted in SD116.
To demonstrate the effects of pfaE overexpression, the biomass, lipid accumulation, and fatty acid composition were analyzed. As shown in Fig. 3, the final biomass (dry cell weight) of the transformant SD116::PfaE was 21.6% coli. orfABC: orfA, orfB, and orfC were co-expressed in E. coli. orfABC + pfaE: orfA, orfB, orfC and pfaE were co-expressed in E. coli. orfABC + ppt_a: orfA, orfB, orfC, and ppt_a from Aurantiochytrium were co-expressed in E. coli Fig. 3 Fermentation profiles in strains SD116 and SD116::PfaE. Biomass and the total fatty acid production in SD116 and SD116::PfaE (left). DHA production and DHA proportion in SD116 and SD116::PfaE (right). **p < 0.01. *p < 0.05 (p < 0.001) higher than that of SD116 at the stationary phase (96 h). In addition, the total fatty acid (TFA) production and the DHA production of the transformant SD116::PfaE were 33.4% (p < 0.001) and 25.9% (p < 0.05) higher than those in SD116, respectively. However, the DHA proportion (percentage in TFAs) in SD116::PfaE is almost the same to that in SD116 (Fig. 3). Thus, it is evident that overexpression of the PPTase encoding by pfaE in Aurantiochytrium can significantly improve its ability of fatty acid synthesis (Additional file 3). From these results, we concluded that the PfaE, an Sfp-type enzyme from Shewanella, had a relatively broad substrate spectrum which was consistent with the results inferred from its primary structure [17], and it could activate the acylcarrier protein both from the FAS and PKS pathway in Aurantiochytrium.

An endogenous PPTase was identified in Aurantiochytrium
The previous study found that substitution of the PPTase gene from the heterologous species did not influence the function of the PKS pathway producing DHA, but led to different DHA yields [17], suggesting that PPTases are also required specifically to recognize the structure of substrates. There are four repeated ACP domains (for S. oneidensis.) or nine (Aurantiochytrium sp. SD116) are integrated in the large multifunctional PfaA (or OrfA) product as substrates of PPTases (Fig. 1). Therefore, although different PPTases could lead to PUFA synthesis, they may have a very strict specificity for efficiently activating their substrates. A PPTase which can specifically recognize the nine repeated ACP domains may be better for PUFAs synthesis in Aurantiochytrium.
The recombinant DHA was produced in E. coli after the pfaE was co-expressed with the PKS system of Aurantiochytrium, so we deduced that there might be a PPTase in Aurantiochytrium involved in the synthesis of PUFAs. A blast search using S. oneidensis MR-1 PfaE as probe from the Aurantiochytrium sp. SD116 genome yielded one contig encoding a putative PPTase. The putative PPTase (named as PPT_A) contains 271aa (30.5 kDa) with a pI of 5.22, and it exhibits a relatively low similarity (only 24.19%) to the PfaE. To confirm that the PPT_A has PPTase activity, ppt_a was cloned and co-expressed with orfA, orfB, and orfC in E. coli. As shown in Fig. 2 4). These conserved amino acid domains were defined previous [19,21]. P2 and P3 are domains participating in Mg 2+ binding, and P1 (both P1a and P1b) and P3 are involved in substrate (coenzyme A) binding and catalysis [31,32]. PfaE for EPA and DHA had another conserved sequence of L/VRxL/ VLS (P0) [19]. The P0 domain and P1a and P1b domains were predicted to be associated with recognition of the specific tertiary structure of the substrates carrying repeated ACP domains of the PfaA product. However, no similar P0 sequence was identified in PPT_A. Peng et al. alignment of four representative PfaE protein sequences revealed that the P0 domain is not conserved among these four PfaE sequences [17]. The substrate of PPT_A contains nine repeated ACP domains, while the PfaE from S. oneidensis and M. marina are required to recognize four and five repeated ACP domains, respectively. Therefore, we deduced that the P0 sequences might be conserved for the structure of substrates.

PPT_A is a PUFA-synthase-specific PPTase in the PKS system of Aurantiochytrium
The stated result demonstrated that PPT_A is involved in the PUFA biosynthesis (Fig. 2). However, the transcription level of ppt_a in Aurantiochytrium sp. SD116 is only 4% of that of fas gene at the logarithmic phase of growth, which is obviously lower than other fatty acid synthesis genes (Fig. 4a). Therefore, we deduced that increasing the PPTase activity may improve the lipid and DHA production in Aurantiochytrium. Therefore, the ppt_a gene was overexpressed in vivo. The expression cassette containing ppt_a and zeo R resistance gene was transformed into Aurantiochytrium via electroporation to obtain SD116::PPT_A strain (Additional file 5). The correct transformants were screened by the selection stress of zeocin and further confirmed by genomic PCR. A 0.9-K bp DNA band was detected in the PCR with the SD116::PPT_A, but not with the wild-type SD116 (Additional file 6), suggesting that ppt_a was successfully inserted in SD116.
To further characterize SD116::PPT_A transformant, both the cell biomass and fatty acid profiles were detected. As shown in Fig. 5, the final biomass and TFA production of SD116::PPT_A were 6.1% (p < 0.01) and 15.2% (p < 0.05) higher than that of SD116, respectively. These results suggest that overexpression of PPT_A can improve the cell growth and lipid synthesis in Aurantiochytrium. In addition to the increased biomass and TFA production, the DHA production and DHA proportion of SD116::PPT_A were 35.5% (p < 0.001) and 17.6% (p < 0.001) higher than that of SD116, respectively (Fig. 5a). Interestingly, although the biomass and TFA production were increased, the total saturated fatty acid (SFA) production of SD116::PPT_A is 95.5% (p < 0.05) of that of SD116 (Fig. 5b), suggesting that overexpression of PPT_A has little effect on the SFAs pathway. From these results, we concluded that the newly discovered endogenous PPT_A can specifically activate the PUFA biosynthesis pathway. Due to the successful overexpression of the ppt_a gene, the transcription level of ppt_a in SD116::PPT_A was 1.70-fold of that of fas gene, while the transcription level ratio (l ppt_a vs l fas , the transcription level of fas gene was used as the control, because SFAs' production had little effect by overexpressing the ppt_a in Aurantiochytrium) was 42.5 times that of wild type (Fig. 4b). Moreover, the relative transcription ratios of orfA, orfB, and orfC in SD116::PPT_A were significantly decreased. At the same time, the expression levels of these genes in transformant containing empty plasmid did not change significantly compared to wild-type strain (Additional file 7). Therefore, we deduced that the newly

Enhancement of DHA proportion using the cerulenin feeding strategy
The endogenous PPT_A can specifically activate the PUFA biosynthesis pathway and has little effect on SFA synthesis pathway. Thus, inhibition of the SFA biosynthesis pathway might further improve the DHA proportion and production. Cerulenin is a fungal antibiotic that irreversibly inhibits the activity of KS domain [33]. Previous study showed that lower concentrations of cerulenin inhibited SFA synthesis pathway, and higher concentrations blocked the PUFA synthesis pathway in Schizochytrium [34]. Therefore, 0.1 mg/L cerulenin was used to enhance PKS pathway in Aurantiochytrium. As shown in Fig. 6a, the final biomass and TFA production were decreased by 8.4% (p < 0.001) and 18.4% (p < 0.001) in SD116::PPT_A, respectively, after addition of cerulenin in medium. The DHA production was decreased by 8.0% (p < 0.05) when SD116::PPT_A was grown with cerulenin; however, the DHA and PUFAs proportion were increased by 12.7% (p < 0.001) and 11.3% (p < 0.001), respectively (Fig. 6b). Although cerulenin inhibits the final biomass and TFA production, SD116::PPT_A could produce higher DHA proportion (53.9%) which is beneficial to a high-quality DHA product. As shown in Fig. 6c, the yields of DHA and PUFAs in SD116::PPT_A were 242.5 ± 7.0 mg/g dcw (dry cell weight) and 292.5 ± 3.8 mg/g dcw, respectively. Moreover, these percentages were not significantly changed after addition of cerulenin (243.7 ± 0.9 mg/g dcw and 290.3 ± 1.8 mg/g dcw, respectively), suggesting that cerulenin has little effect on PKS system. The percentage of SFA in SD116::PPT_A decreased by 24.7% (p < 0.001) after addition of cerulenin in medium, suggesting that cerulenin inhibited the SFA synthesis pathway more efficiently. Based on these results, we confirmed that low concentrations of cerulenin could selectively inhibit the SFA synthesis pathway in Aurantiochytrium SD116::PPT_A.
Both PKS and FAS systems use the same substrate to synthesize fatty acids. Here, we found that inhibiting the FAS pathway could improve the DHA proportion in Aurantiochytrium. A high-quality DHA product from SD116::PPT_A was obtained with the cerulenin strategy. However, addition of cerulenin is not conductive to the industrial production of DHA. Further study to downregulate gene expression of fas by genetic engineering technology will provide more advantages for DHA industrial production.

Conclusions
In this study, a new endogenous PPTase (PPT_A) was identified and proved to be essential to synthesize PUFAs in thraustochytrids Aurantiochytrium. Furthermore, the DHA and PUFA proportion in PPT_A overexpressed strain can reach to 53.9% and 64.5% in TFA, respectively. This study has first identified a PUFAs-synthase-specific PPTase in PKS system of Aurantiochytrium and provided

Strains and cultivation
Aurantiochytrium sp. SD116 was cultivated in a seed liquid medium containing 30 g/L glucose, 10 g/L yeast extract, and 10 g/L artificial sea salt. The strain was cultivated at 25 °C and shaken at 200 rpm for 48 h, and then, 2% (v/v) of cultures were transferred into fermentation medium containing 60 g/L glucose, 20 g/L yeast extract, and 15 g/L sea salt. The screening medium, containing 30 g/L glucose, 10 g/L yeast extract, 5 g/L artificial sea salt, 1.5% agar, and 50 ug/mL zeocin, was used for screening of transformants [4,35].

Molecular cloning and plasmid construction
The zeocin expression cassette was amplified from the pGZC-1 [4] with the primers 0426-3F and 0426-3R. The tubulin promoter, the actin terminator, and homologous fragments 18S1 and 18S2 were amplified from Aurantiochytrium genome with primer pair 194F/194R, 196F/196R, 191F/191R and 192F/192R, respectively. pfaE gene was amplified from S. japonica strain KCTC 22435 genome with primers 195F and 195R. The pfaE expression cassette was assembled by the overlap PCR, and then ligated to the zeocin expression cassette and 18S1 and 18S2 by overlap PCR (Additional file 1). To construct the ppt_a expression cassette, the ppt_a gene was amplified from Aurantiochytrium genome with primers 197F and 197R which contains the 2A sequence. The plasmid vector was amplified from pGZC-1 with primers 198F and 198R, and then, ppt_a gene was ligated to plasmid vector by seamless clone method (Additional file 4). The primer pair 191F/192R was used to amplify the line fragments for expression of pfaE and ppt_a, and then electrotransformation of Aurantiochytrium sp. SD116 was performed according to reported protocol [4].

Quantitative real-time PCR (qRT-PCR) analysis
The total RNA was isolated using the Trizol reagent (Thermo Scientific) and miRNeasy Mini Kit (QIAGEN, Germany), and subsequent synthesis of cDNA was achieved using a Revert Aid First Strand cDNA Synthesis Kit (Thermo Scientific). The cDNA was used as the template for qRT-PCR analysis with the primers listed in Additional file 8. PCR was performed using a Fast-Start Universal SYBR Green Master (ROX). 18S rRNA was used as an internal control to normalize the expression levels. The relative transcription level was calculated using the 2 −ΔΔCt method.

Growth and biomass determination
The biomass was expressed as DCW. 5 mL samples were collected and centrifuged at 9000g at 4 °C for 5 min, and then, it was determined by freeze-drying to constant weight at − 50 °C.

Lipid extraction and fatty acid composition analysis
The total lipid was extracted using a combination of chloroform and methanol (2:1, v/v), and weighed according to our previous report [4]. Then, the lipid was dissolved in chloroform and transformed to the fatty acid methyl esters (FAMEs) by using the method of Cui et al. [4].
FAMEs were analyzed by an Agilent 7890B gas chromatograph coupled with 5975C mass spectrometry. Separation was achieved on an HP-INNOWAX (30 m × 0.25 mm i.d., 0.25 μm) capillary column with helium as the carrier gas at a constant flow rate of 1.0 mL/min. The GC temperature programming was set at 100 °C for 1 min, then increased to 250 °C with increments of 10 °C/min, and held for 10 min. The temperature of injection, transfer line, and ion source were 250, 280, and 250 °C, respectively. The mass scan range was 50-800 m/z and selected ion mode (m/z 79.1) for quantitative analysis. FAMEs were identified by comparing mass spectra of the products with those of authentic ones and by utilizing National Institute of Standards and Technology (NIST) mass spectral library.