Results
Screening the phenylacetaldehyde reductases/alcohol dehydrogenases in Y. lipolytica and constructing their expression vectors (please see the part of engineering success).
Identifying the optimal phenylacetaldehyde reductase PAR4 by performing the whole-cell biocatalyst (please see the part of engineering success).
To obtain an optimized alcohol dehydrogenase or phenylacetaldehyde reductase, we firstly screened genome of Y. lipolytica by using amino sequences of the rose phenylacetaldehyde reductase PARL, and got eight putative phenylacetaldehyde reductase, which encoding genes include YALI0D08844g (PAR1, BBa_K4297009), YALI0F09097g (PAR2, BBa_K4297010), YALI0F24937g (PAR3, BBa_K4297011), YALI0D07062g (PAR4, BBa_K4297012), YALI0D12386g (PAR5, BBa_K4297013), YALI0C20251g (PAR6, BBa_K4297014), YALI0D11616g (PAR7, BBa_K4297015), and YALI0D08778g (PAR8, BBa_K4297016). We also synthesized the rose phenylacetaldehyde reductase gene PARL with the codon optimization. These candidates were overexpressed by plasmid pYLXP’, respectively, and transformed into Y. lipolytica po1g, obtaining strains PAR1, PAR2, PAR3, PAR4, PAR5, PAR6, PAR7, PAR8, and PARL. Whole-cell biocatalytic conversion of 1 g/L phenylacetaldehyde by these strains gave 2-PE titers of 762.71, 703.62, 759.37, 786.58, 435.61, 437.33, 399.11, 501.10, and 379.28 mg/L, which the highest 2-PE titer was produced by PAR4.
Additionally, we also investigate the performances of alcohol dehydrogenases in Y. lipolytica. All alcohol dehydrogenases annotated by Genbank (https://www.ncbi.nlm.nih.gov/genbank/) in Y. lipolytica were respectively cloned into plasmid pYLXP’, and yield plasmids pYLXP’-ADH1 (YALI0D25630g, BBa_K4297034), pYLXP’-ADH2 (YALI0E17787g, BBa_K4297035), pYLXP’-ADH3 (YALI0A16379g, BBa_K4297036), pYLXP’-ADH4 (YALI0A15147g, BBa_K4297037), pYLXP’-ADH5 (YALI0E07766g, BBa_K4297038), pYLXP’-ADH6 (YALI0E15818g, BBa_K4297039), pYLXP’-ADH7 (YALI0D02167g, BBa_K4297040), pYLXP’-ADH8 (YALI1C17782g, BBa_K4297041) and pYLXP’-ADH9 (SFA1, BBa_K4297062). Each plasmid was then introduced into po1g, and obtained stains (ADH1, ADH2, ADH3, ADH4, ADH5, ADH6, ADH7, ADH8 and ADH9) were performed the whole-cell biocatalytic conversion. Unexpectedly, suboptimal results of 2-PE titers were detected in these strains compared to that of overexpressing putative phenylacetaldehydes (Fig. 4d), and the highest titer of 2-PE just reached 723.74 mg/L produced by strain ADH2.
Fig. 1 Identifying the optimal phenylacetaldehyde reductase/alcohol dehydrogenase a) The morphology analysis of engineering strains; b) Performing the whole-cell biocatalytic conversion; c) The 2-PE titer of the whole-cell biocatalytic conversion
Table 1 Plasmids constructed for identifying the optimal phenylacetaldehyde reductase/alcohol dehydrogenase
Constructing the expression vectors, including pYLXP’-ylARO10, pYLXP’-ylARO8, pYLXP’-ylARO9, pYLXP’-ylARO10-ylARO8, pYLXP’-ylARO10-ylARO9, and pYLXP’-ylARO10-ylARO9-PAR4 (please see the part of engineering success).
Obtaining a remarkable increase of 2-PE titer by overexpressing the Ehrlich pathway, which reached 1086.86 mg/L and is the 1.17-fold of the control (please see the part of engineering success).
To get an efficient Ehrlich pathway, we adopted the stepwise pathway engineering. Firstly, the recognized genes ylARO8, ylARO9 and ylARO10 in Ehrlich pathway were overexpressed in Y. lipolytica po1g under the control of strong constitutive promoter pTEP with intron via plasmid pYLXP’, respectively. Shaking flask results (Fig. 2d, e, f) showed that overexpression of genes ylARO8 and ylARO9 (strain po1gP1 and po1gP2) has no effect on the 2-PE production. However, overexpression of gene ylARO10 (strain po1gP3) has a remarkable increase of 2-PE titer, which reached 922.86 mg/L and is the 1.33-fold of the control. It strongly suggested that the reaction catalyzed by phenylpyruvate decarboxylase ylARO10 is a limiting-step in Ehrlich pathway.
Next, we constructed plasmids pYLXP’-ylARO10-ylARO8 and pYLXP’-ylARO10-ylARO9, and transformed into po1g, obtaining strains po1gP4 and po1gP5. However, shake flask results showed no significant differences in 2-PE production between these two strains, indicating that Y. lipolytica has the obstacle in catalyzing phenylacetaldehyde to 2-PE. Therefore, we further constructed the plasmid pYLXP’-ylARO10-ylARO9-PAR4 and transformed into Y. lipolytica, obtaining strain xx. Shaking flask results showed that overexpression of Ehrlich pathway has an important effect on the 2-PE production, which reached 1086.86 mg/L and is the 1.17-fold of the control.
Table 2 Plasmids constructed for overexpressing the Ehrlich pathway
Fig. 2 Overexpressing the Ehrlich pathway to increase the production of 2-PE a) The 2-PE titer, yield, and cell growth of strains po1g1, po1g2, and po1g3; b) The colony of po1g6; c) The 2-PE titer, yield, and cell growth of strains po1g4, po1g5, and po1g6
Constructing gene knockout vectors, including pYXLP’-loxp-uras-ALD2 and pYXLP’-loxp-uras-ALD3 (please see the part of engineering success).
Deleting gene ALD2 and ALD3 in Y. lipolytica, which increased the 2-PE production (please see the part of engineering success).
To further improving the production of 2-PE, we turned to remove the competitive metabolic pathways. The main byproduct was phenylacetate, which was synthesized by aldehyde dehydrogenases ALD2 and ALD3 with phenylacetaldehyde as the precursor and generation of one mole NADH. A marker-free gene knockout method based on Cre-lox recombination system was used as previously reported (Lv et al., 2019; Lv et al., 2020). Here, we take the process of constructing the gene knockout vectors pYXLP’-loxp-uras-ALD2 as an example. For performing gene knockout, the upstream and downstream sequences (both 1000 bp) flanking ALD2 were PCR-amplified. These two fragments, the loxP-Ura/Hyr-loxP cassette (digested from plasmid pYLXP’-loxP-Ura/Hyr by AvrII and salI), and the residual plasmid backbone of pYLXP’-loxP-Ura/Hyr were joined by Golden Gate assembly, obtaining the knockout plasmids pYXLP’-loxp-uras-ALD2. The obtained vector was sequenced by Sangon Biotech (Shanghai, China).
Table 3 Plasmids constructed for blocking competitive pathways
It should be noted that we orderly knocked out genes ALD2 (YALI0D07942g, BBa_K4297060) and ALD3 (YALI0F04444g, BBa_K4297061) in po1fk, obtaining strain po1fk3, and transformed plasmids pYLXP’-ylARO10-ylARO8-PAR4 (BBa_K4297056), obtaining strain po1fk3. As a result, shaking flask of po1fk3 showed an increased 2-PE titer, which was 1846.49 mg/L, 1.70-fold of the control.
Fig. 3 Enhancing the production of 2-PE by blocking competitive pathways. a) The colony of po1g6; b) The 2-PE titer and cell growth of strain po1fk3
Moreover, as suggested by the mathematical model, 2-PE yield is more sensitive to the supply of akG. We added the 2-oxoglutarate (5g/L) into the fermentation medium. As a result, the 2-PE titer reached to 2015.57 with the yield of 0.575 g/g.
Fig. 4 The 2-PE titer and cell growth of strain po1fk3 under adding the 2-oxoglutarate
Lv, Y., Edwards, H., Zhou, J., Xu, P. 2019. Combining 26s rDNA and the Cre-loxP system for iterative gene integration and efficient marker curation in Yarrowia lipolytica. ACS Synth Biol.
Lv, Y., Gu, Y., Xu, J., Zhou, J., Xu, P. 2020. Coupling metabolic addiction with negative autoregulation to improve strain stability and pathway yield. Metab Eng, 61, 79-88.