We introduced an entire EPA biosynthesis pathway into Yarrowia lipolytica Po1f, which is capable of guiding the conversion from the C18:2 to EPA step by step. In addition, many strategies were applied to enhance the EPA content, including inhibiting β-oxidation pathway, screening for different sources of elongases/desaturates, increasing the copy numbers of target genes, and replacement of promoter. All BioBricks were registered and tested, and most of them work well to perform their functions as expected.

Design

Because Y. lipolytica Po1f is lack of EPA biosynthesis-related enzymes, we attempted to create an EPA biosynthesis pathway in Y. lipolytica Po1f by integrating heterogenous genes from Euglena gracilis and Pythium aphanidermatum. Here, three genes encoding ∆-9 elongase (EgElo9), ∆-8 desaturase (EgDes8) and ∆-5 desaturase (EgDes5), respectively, from E. gracilis, and a ∆-17 desaturase gene (PaDes17) from P. aphanidermatum were selected as candidate genes. In order to express the heterogenous genes in Y. lipolytica Po1f, the coding sequences (CDS) of selected genes were optimized based on the codon usage bias of Y. lipolytica Po1f (http://www.kazusa.or.jp/codon/). The native strong constitutive promoter PTEF was used to express these genes.

Build

In order for Y. lipolytica Po1f to produce EPA, ∆-9 elongase, ∆-8 desaturase and ∆-5 desaturase needed to be expressed in Y. lipolytica. Firstly, the CDS of EgElo9 BBa_K4343095, EgDes8 BBa_K4343096, EgDes5 BBa_K4343097 and PaDes17 BBa_K4343098 were optimized, and strong constitutive promoter (PTEF) fragments were subcloned into the linear pUC-HUH together by the enzyme digestion and ligation, yielding the desired plasmids pUC-HUH-EgElo9 BBa_K4343066, pUC-HUH-EgDes8 BBa_K4343067, pUC-HUH-EgDes5 BBa_K4343068, pUC-HUH-PaDes17 BBa_K4343069. Take the process of constructing pUC-HUH-EgElo9 for instance, i) the DNA fragment of EgElo9 was synthesized with the restriction enzyme cutting site BsaI, and PCR-amplified was used to plus the restriction enzyme cutting site BsaI on the vector pUC-HUH; ii) then, T4 ligase was used to connect the DNA fragment of EgElo9 and the vector pUC-HUH. iii) next, the reaction mixture was transformed to E. coli DH5α; iv) corrected colonies were verified by PCR and sequenced; v) the correctly sequenced plasmid was digested by the nuclease SwaI to obtain a linearized fragment, which was transferred into the Y. lipolytica Po1f. A complete overview of the detailed processes visit our Part.

Fig. 1 (A) Construction of target plasmid. (B) Constructing expression vectors of ∆-9 elongase, ∆-8 desaturase, ∆-5 desaturase and ∆-17 desaturase. (C) The PCR validation diagram of pUC-HUH-EgElo9.

Test

The percentages of different fatty acids in TFAs in the recombinant strains were detected by gas chromatography (GC) analysis. Results showed that the Po1f-1 strain that has been integrated with ∆-9 elongase and ∆-8 desaturase genes produced eicosadienoic acid (EDA; C20: 2n-6) and DGLA at 20.8% and 8.5% of the TFAs, respectively, indicating the two enzymes were successfully expressed in the Y. lipolytica Po1f. Moreover, transforming all the four heterogenous genes into Y. lipolytica Po1f led to EPA production at 2.4% of the TFAs.

Fig. 2 (A) Fatty acid profile of the Y . lipolytica wild-type strain Po1f determined by Gas Chromatography analysis. (B) Different fatty acid percentage of TFAs.

Learn

In this cycle, we learned that codon optimization is helpful strategy to improve the success rate of heterogenous gene expression. We were able to assemble the two DNA fragments (CDS and promoter) with the backbone to obtain a full construct. Then the plasmid was linearized and transformed into the receptor state of Y. lipolytica Po1f. Besides, the linearized fragments were randomly integrated into the genome by non-homologous End Joining (NHEJ). The recombinant strains were endowed ability to successfully produce EPA, suggesting our design idea is flexible.

Design

Fatty acid β-oxidation is the process of breaking down a long-chain acyl-CoA molecule to acetyl-CoA molecules. Therefore, blockade of β-oxidation pathway contribute to accumulation of PUFAs. In order to disrupt this pathway, the homologous recombination strategy will be utilized to delete the key gene PEX10 of β-oxidation pathway in Po1f-2 stain. However, the number of selection markers that can be used for gene editing is limited. In this study, we employed a marker recycling system that allows the recycle of the selection marker URA3 in successive transformations in Y. lipolytica Po1f. Our URA3 sequence was flanked by the same 1000-bp hisG sequence. URA3 can be removed from the genome by homologous recombination when the cell is grown in the presence of 5-fluoroorotic acid (5-FOA), a chemical that is converted into a toxin by URA3 activity.

Fig. 3 The recycle of the selection marker ura3 in successive transformations in Y. lipolytica Po1f.

Build

Firstly, we removed the selection marker ura3 of Po1f-2 by reverse selection on the YPD media with 5-FOA. The survival clone was uridine/uracil auxotrophic mutant and served as the recipient strain for PEX10 deletion. In order to knock out the PEX10 gene, the upstream and downstream sequences (both 100bp) flanking PEX10 were PCR-amplified, then they were subcloned into the linear pUC-HUH together by the BsaI enzyme digestion and T4 ligase ligation, yielding the knockout plasmids pUC-HUH-ΔPEX10. The PCR was used to verify the deletion of PEX10 (Validation primer F: TCCGACGAcactcgtcttct; Validation primer R: TATCATTAACCAGCTTTGAT). The obtained vector was sequenced by Sangon Biotech (Shanghai, China). The correctly sequenced plasmid was digested by the nuclease SwaI to obtain a linearized fragment, which was transformed into Po1f-2, generating the strain Po1f-3.

Fig. 4 (A)The deletion process of PEX10 gene. (B) PCR validation diagram for verifying whether the PEX10 was knocked out on the genome.

Test

The percentages of different fatty acids in TFAs in the related strains were analyzed by GC. In comparison to Po1f-2, the PEX10 deletion stain produce more EPA at 4.5% of the TFAs, demonstrating inhibition of β-oxidation pathway indeed contributes to EPA accumulation in Y. lipolytica Po1f.

Fig.5 Different fatty acid percentage of TFAs in Po1f-3.

Learn

In this cycle, we learned how to recycle the ura3 selection marker by the 5-FOA-meidated reverse selection and how to delete the target gene by homologous recombination strategy. In addition, we realized that inhibition of the catabolic pathway of the target product was a very effective strategy in microbial metabolic engineering.

Design

Studies have demonstrated that elongases and desaturases form different species have different catalytic activities and substrate specificities for fatty acid substrates, so the selection of elongases/desaturases with high substrate specificity is a significant strategy to improve the synthesis of EPA.

Build

To screen the optimal elongases/ desaturates, the optimized CDS and strong constitutive promoter fragments were subcloned into the linear pUC-HUH together. Take ∆-12 desaturase for example, pUC-HUH-Intc-FmDes12 BBa_K4343077, pUC-HUH-Intc-CcDes12 BBa_K4343078 and pUC-HUH-Intc-PaDes12 BBa_K4343079 were constructed. i) Firstly, the upstream and downstream squences (both 100bp) of flanking IntC were PCR-amplified. The three fragments, the upstream, the downstream and FmDes12 were subcloned into the linear pUC-HUH by T4 ligase, obtaining the targeted integration plasmid pUC-HUH-Intc-FmDes12 BBa_K4343077. ii) Next, the obtained vector was sequenced by Sangon Biotech (Shanghai, China), iii) the correctly sequenced plasmid was digested by the nuclease SwaI to obtain a linearized fragment, which was transferred into the Y. lipolytica Po1f to obtain Po1f-4. Besides, ∆-9 desaturase was integrated at the Scp2 site of the genome of Po1f-4, yielding the desired plasmid pUC-HUH-Scp2-BgElo9 BBa_K4343080, pUC-HUH-Scp2-EhElo9 BBa_K4343081, pUC-HUH-Scp2-IgASE1 BBa_K4343082, pUC-HUH-Scp2-IgASE2 BBa_K4343083, pUC-HUH-Scp2-PpElo9 BBa_K4343084 and pUC-HUH-Scp2-PsElo9 BBa_K4343085. And C16/18 elongase was also integrated at the Scp2 site of the genome of Y. lipolytica Po1f, yielding the desired plasmid pUC-HUH-Scp2-MaElo2 BBa_K4343086, pUC-HUH-Scp2-MaElo16 BBa_K4343087, pUC-HUH-Scp2-PaElo16 BBa_K4343088 and pUC-HUH-Scp2-rElo2 BBa_K4343089, respectively. The above plasmids were linearized by enzyme cleavage and transformed into the corresponding competent cell of Y. lipolytica.

Fig.6 (A) Construction of plasmids of elongases/ desaturates from different sources. (B) PCR validation diagram for FmDes12 integrated in the genome. (C) Squencing of FmDes12, EhElo9 and MaElo16.

Test

GC analysis indicated that the enzyme activity of FmDes12 was the highest among all the tested ∆-12 desaturases due that the FmDes12-expressing strain displayed the conversion rate (81.1%) of C18:2 to C18:1. Besides, IgASE2 and EhElo9 performed better to promote the synthesis of C20:2 from C18:2. It is worth mentioning that all the four tested C16/18 elongases, and all have catalytic activity for C18:1, however, only MaElo2 and rElo2 can efficiently convert C16:0 to C18:0.

Fig.7 Different fatty acid percentage of TFAs produced by different sources of ∆-12 desaturase, ∆-9 elongase and C16/18 elongase.

Learn

In this cycle, we learned that employing elongases/desaturases with high substrate specificity is an effective strategy to improve EPA synthesis. This cycle provided an effective strategy and a good basis for us to further increase the yield of EPA.

Design

Based on the push-pull strategy in microbial metabolic engineering, the accumulated intermediates (C16:0, C18:1, C18:2) of EPA biosynthesis pathway need to be reduced by genetic modify for elevating EPA yield. We planned to transform the ∆-12 desaturase, ∆-9 elongase and C16/18 elongase from other organisms into the genome of Y. lipolytica Po1f by random integration strategy to reduce the content of intermediates.The resulting strains with the highest EPA yield were analyzed by qPCR to confirm the copy number of target genes.

Build

To obtain engineered strains for efficient EPA production, seven plasmids were successfully constructed, including pUC-HUH-MaElo2 BBa_K4343072, pUC-HUH-FmDes12 BBa_K4343073, pUC-HUH-EhElo9 BBa_K4343074, pUC-HUH-IgASE2 BBa_K4343075, pUC-HUH-EgDes8 BBa_K4343067, pUC-HUH-EgDes5 BBa_K4343068 and pUC-HUH-PaDes17 BBa_K4343069. Then these plasmids were linearized by the nuclease SwaI and transformed into the competent cell of Y. lipolytica to obtain Po1f-17, Po1f-18, Po1f-19 and Po1f-20. All the final plasmids were verified by PCR and commercial sequencing. A complete overview of the detailed processes visit our results.

Fig.8 Corrected colonies were verified by PCR and sequencing.

Test

After GC analysis, the Po1f-18 strain integrated with 3 copies of MaElo2 and 2 copies of FmDes12 can produce at 9.6% of TFAs. Besides, the incorporation of 3 copies of the EhElo9, 3 copies of the IgASE2 and 2 copies of the EgDes8 elevated the EPA content to 16.3%. Finally, employing 2 copies of EgDes5 and 2 copies of PaDes17 further increased the percentage of EPA to 22.4%. All the above results indicated that increasing the copy number of target genes contributed to a significant increase in EPA ratio.

Fig.9 (A) Different fatty acid percentage of TFAs produced by random integration of the best strains of elongases/desaturases. (B) Optimal copy numbers detected by QPCR. (C) Fatty acid profiles of the Po1f-3 and Po1f-20 determined by Gas Chromatography analysis.

Learn

In term of supplementing activity of related enzymes, improving the encoding gene expression level is a routine way. In this cycle, we learned that increasing the copy numbers of gene (MaElo2,FmDes12,IgASE2,EgDes8, EgDes5 and PaDes17) in the host genome by multiple-round transformation can effectively increase gene expression level and product contents. In comparison to the alternative strategy, promoter replacement, the strategy was more convenient and efficient due to omission of promoter search and new plasmid construction.

Design

After communicating with our cooperative partner, a group of university of science and technology of China, we plan to adopt their suggestion to replace the originally used promoter PTEF with the other strong promoter PGAP for further improving the EPA proportion in TFA. In addition, several other strong constitutive promoters (PEXP, PYAT, PFBA in and PTEF in) were selected for probing the variation of different promoters on EPA content.

Build

To test whether different promoter strengths can effectively control EPA production, PTEF was cleaved by the nuclease BsaI. And different strong constitutive promoters (PGAP, PEXP, PYAT, PFBA in and PTEF in) were amplified by PCR. Then they were subcloned into the linear pUC-HUH-Scp2-EhElo9 by T4 ligase and the above plasmids were linearized by the nuclease SwaI. The obtained plasmid was sequenced by Sangon Biotech (Shanghai, China). Next, the correctly sequenced plasmid was digested by the nuclease SwaI to obtain a linearized fragment, which was transformed into Po1f-3.

Fig.10 (A) Constructing expression vectors of ∆-9 elongase under the control of different promoters. (B) Corrected colonies were verified by PCR and sequencing.

Test

As shown in Figure, these engineered strains integrated with new promoters had no showed elevated EPA yield compared with their parent strains, which demonstrates that the selected candidate promoters did not play positive role in our case.

Fig.11 Different fatty acid percentage of TFAs produced by engineered strains transferred of EhElo9 controlled by different promoters.

Learn

Although promoters, PGAP, PEXP, PYAT, PFBA in and PTEF in, were not as superior in Y. lipolytica as the original promoters (PTEF), it does not mean that the promoter replacement strategy does not work in our project. As a next plan, RNA-seq method can be considered to screen the specific strong promoters in the EPA fermentation conditions for continually improving the expression of target genes.