Engineering Yarrowia lipolytica for nervonic acid production by organelle compartmentalization and dynamic promoter coordination

TL;DR

This study engineered the yeast Yarrowia lipolytica to produce nervonic acid, a compound important for nerve repair, using metabolic engineering techniques.

Contribution

The study introduces an integrated strategy combining organelle compartmentalization and dynamic promoter coordination to enhance nervonic acid production.

Findings

The engineered strain achieved 451.80 mg/L of nervonic acid in shake-flask culture.

Fed-batch fermentation in a 5-L bioreactor yielded 6.55 g/L, with nervonic acid comprising 29.81% of total fatty acids.

Optimized β-ketoacyl-CoA synthases and precursor availability improved nervonic acid accumulation.

Abstract

Nervonic acid (NA) is of great significance in repairing damaged nerve fibers and promoting the regeneration of neural cells. This study aimed to construct a Yarrowia lipolytica strain capable of producing NA. The β-oxidation pathway was first disrupted and evaluated β-ketoacyl-CoA synthases (KCSs) with different substrate specificities. The substrate specificity of KCS was further optimized through semi-rational design. Oleic acid biosynthesis and the desaturation of tetracosanoic acid (C24:0) were enhanced to increase precursor availability. Furthermore, organelle compartmentalization and dynamic promoter coordination were implemented to synergistically improve NA accumulation. Finally, an orthogonal malonyl-CoA biosynthetic route was introduced together with an NADPH regeneration system to boost reducing power supply, leading to a substantial increase in production. The engineered strain achieved a titer of 451.80 mg/L in shake-flask culture, and reached 6.55 g/L during fed-batch fermentation in a 5-L bioreactor, where NA accounted for 29.81% of TFAs. This study presents an integrated metabolic engineering strategy providing novel insights into the microbial biosynthesis of NA.

Figures

Fig. 2Evaluation of the effect of single or combined KCS expression on NA production based on the substrate specificity of KCS.(A) Screening of KCS variants. (B) Schematic diagram of gene knockout. (C) Combined expression of KCS genes and knockout of degradation-related genes.Fig. 2

Full text

Introduction

1

Nervonic acid (NA, ω-9 C24:1) is an ω-9 very-long-chain monounsaturated fatty acid first discovered in the nervous tissues of humans and other mammals [1]. It is an essential nutrient for the growth, development, and maintenance of neural cells—particularly in the brain and peripheral nervous system. Thus, it has been widely recognized as a key natural compound with dual functions in repairing damaged nerve fibers and promoting neuronal regeneration [2,3]. NA can improve cognitive impairment caused by neurological disorders and prevent neurodegenerative diseases, thereby highlighting its potential as a valuable bioactive fatty acid [4,5]. Briefly, NA has attracted increasing attention because of its potential benefits for neural development and repair, and its applications in pharmaceuticals, nutraceuticals, and functional foods. As a result, the industrial demand for high-purity NA has been steadily growing. Currently, the commercial production of NA mainly relies on plant extraction and chemical synthesis. However, plants rich in NA are constrained by specific growth requirements, low propagation efficiency, and long cultivation cycles, which collectively hinder large-scale production [[6], [7], [8]]. In contrast, chemical synthesis typically uses erucic acid as the starting material, but it often generates NA with undesired trans-isomers and multiple by-products, resulting in poor selectivity and low overall yield [9]. Therefore, both approaches are insufficient to meet increasing market demand and sustainability requirements, limiting broader industrial application of NA. With advances in synthetic biology, microbial production has emerged as an alternative route. Engineered microbial cell factories provide advantages such as product consistency, process controllability, and improved sustainability, offering a scalable platform for NA production [10].

The biosynthetic pathways of NA in plants and yeasts have been well characterized. Its biosynthesis comprises two modules: the de novo fatty acid synthesis module and the elongation module. Fatty acids generated from the de novo pathway provide precursors for subsequent elongation. During this process, C18:1 Δ9-CoA serves as the initial substrate, malonyl-CoA as the carbon donor, and NADPH as the reducing power. Chain elongation is catalyzed by a multi-enzyme complex comprising β-ketoacyl-CoA synthase (KCS), β-ketoacyl-CoA reductase, β-hydroxyacyl-CoA dehydratase, and enoyl-CoA reductase, which function cyclically. The resulting nervonyl-CoA is subsequently incorporated into triacylglycerols that are stored in lipid droplets (LD) [11]. Among the aforementioned enzymes, KCS plays a pivotal role due to its substrate specificity, which determines the carbon chain length of the synthesized fatty acid. Previous studies demonstrated the feasibility of heterologous NA production in yeast. Taylor [12] and Guo et al. [13] expressed KCS genes derived from Cardamine graeca (CgKCS) and Lunaria annua (LaKCS) in Saccharomyces cerevisiae, yielding trace amounts of NA. Later, Zhang et al. [14] optimized enzyme combinations and expression levels in S. cerevisiae, which increased NA production to 57 mg/L, thereby underscoring the importance of synergistic interactions among elongase enzymes. However, the supply of NADPH and acetyl-CoA becomes a limiting factor in most hosts during high-throughput lipid synthesis, inhibiting elongation reactions [15,16].

The selected chassis organism, Yarrowia lipolytica, is a Generally Recognized As Safe oleaginous yeast capable of providing abundant acetyl-CoA and NADPH for fatty acid biosynthesis. Given its strong lipid-producing metabolism, Y. lipolytica has been successfully engineered to synthesize various polyunsaturated fatty acids, including eicosapentaenoic acid (EPA) [17], docosapentaenoic acid (DPA) [18], and docosahexaenoic acid (DHA). Y. lipolytica naturally accumulates high levels of oleic acid (C18:1), thereby representing an ideal microbial platform for NA biosynthesis [19]. With the rapid advancement of synthetic biology, significant progress has been made in enhancing NA production using this yeast chassis. Wang et al. [20] combined plant-derived elongation pathways (KCS) with nonplant desaturation pathways, achieving an NA titer of 13.56 g/L in a 5-L fermenter. Similarly, Su et al. [1] optimized the expression of the CgKCS gene and applied enzyme fusion strategies, thus successfully constructing the highest-yield Y. lipolytica strain that produced 17.3 g/L of NA in a 50-L fermenter. These advances highlight the potential of yeast cell factories for industrial NA production, yet most strategies primarily focus on KCS expression and process optimization. Despite substantial progress made through strategies such as elongase screening, desaturase engineering, and optimization of lipid metabolic pathways, several challenges still remain [21]. The low substrate specificity of KCS limits its ability to efficiently utilize the key precursors for NA biosynthesis, C20:1 and C22:1, thereby restricting the elongation of very-long-chain fatty acids (VLCFAs). The endoplasmic reticulum (ER), as the central organelle for lipid metabolism, provides the structural framework for fatty acid elongation, with its morphology and organization significantly influencing metabolic efficiency [22]. Moreover, the desaturase responsible for introducing double bonds in this pathway requires iron as a catalytic cofactor and oxygen as an electron acceptor, which are factors still largely unaddressed in current metabolic engineering efforts. Systematic strategies that simultaneously optimize KCS performance, organelle organization, and cofactor/precursor supply in Y. lipolytica are still lacking. These gaps were addressed by (i) rationally engineering the elongation modules, (ii) modulating ER architecture and enzyme localization, and (iii) reinforcing key lipid biosynthetic precursors, thereby constructing a robust Y. lipolytica platform for efficient NA production.

In this study, a Y. lipolytica strain capable of de novo synthesis of NA was successfully constructed. The substrate specificity of the key enzyme KCS was enhanced through semi-rational design, thereby improving its selectivity toward the essential precursors C20:1 Δ9-CoA and C22:1 Δ9-CoA. An oxygen–iron synergistic enhancement system was further established by co-expressing Vitreoscilla hemoglobin (VHb) and the iron transporter Ftr1, which significantly increased intracellular oxygen and iron availability and enhanced the catalytic efficiency of desaturases. Moreover, subcellular engineering combined with dynamic promoter coordination precisely regulated the localization and timing of key enzyme expression within the ER, thereby improving the efficiency of the engineered biosynthetic pathway. Coupling a noncarboxylating malonyl-CoA synthesis pathway with an NADPH regeneration module further strengthened precursor supply and reducing power, leading to a marked increase in NA accumulation. The engineered strain achieved a titer of 451.8 mg/L in shake-flask culture and 6.55 g/L in a 5-L fed-batch bioreactor. Overall, this study provides novel insights into the microbial production of other VLCFAs and their derivatives.

Materials and methods

2

Strains and medium

2.1

Escherichia coli JM109 is used for constructing and preserving plasmids. The E. coli strains used in this study were cultured in Luria-Bertani medium at 37 °C and 220 rpm. Also, 100 mg/L ampicillin was added when necessary to inhibit the growth of other bacteria, thereby ensuring the specificity and accuracy of the experiments. All Y. lipolytica transformants used in this study were derived from Y. lipolytica Po1f (ATCC MYA-2613) and preserved in our laboratory [23]. Y. lipolytica was cultivated in YPD medium (10 g/L yeast extract, 20 g/L peptone, and 20 g/L glucose) for growth. Additionally, appropriate dropout amino-free yeast nitrogen base (YNB) medium was used, containing 6.74 g/L YNB and 20 g/L glucose. This medium was supplemented with appropriate amino acids as needed for transformant selection. All cultures were conducted at 30 °C and 220 rpm to optimize cell growth and metabolic activity, thereby enhancing the production efficiency of the target products.

Construction of plasmids and strains

2.2

All plasmids and Y. lipolytica strains constructed in this study are listed in the Supplementary Information Tables S2 and S1. Codon-optimized genes, including TnKCS (Teesdalia nudicaulis), MoKCS (Malania oleifera), CgKCS (Cardamine graeca), AtKCS (Arabidopsis thaliana), LaKCS (Lunaria annua), CraKCS (Crambe abyssinica), Maω9 and MaELO2 (Mortierella alpina), ELOVL6 (goat), VHb (Vitreoscilla), and oleosin (Zea mays), were synthesized and cloned into the vectors pYLXP’ and pYLXP’2 using Gibson assembly. Endogenous genes and promoters were amplified from the Y. lipolytica genome. Plasmids were constructed using Gibson assembly unless otherwise specified. DNA fragments, including promoters, the above-mentioned related genes and terminators, were amplified by PCR using the 2 × PhantaMax MasterMix high fidelity enzyme (Vazyme, Nanjing, China), with primers listed in Supporting Information Table S4. The CRISPR–Cas9 system (with a LEU2-marked plasmid) was employed for gene deletion and single-copy targeted integration. Plasmids and linearized expression cassettes were introduced into Y. lipolytica via the Frozen-EZ Yeast Transformation II Kit (Zymo Research, Orange, CA), and transformants were selected on YNB agar plates supplemented with the appropriate amino acids.

Bioreactor fermentation

2.3

In fed-batch fermentation, suitable colonies were first selected and inoculated into 25 mL of YPD medium under the same conditions (220 rpm and 30 °C) for 24 h to obtain the primary seed solution. Next, 10 mL of the primary seed solution was inoculated into 200 mL of YNB and cultured for another 24 h to obtain the secondary seed solution [24]. The fed-batch fermentation was carried out in a 5-L bioreactor. During fermentation, the pH was maintained at 5.5 using 12.5% (v/v) ammonia solution or 10 M NaOH. The cultivation temperature was set at 28 °C, and the airflow rate was maintained at 3.0 vessel volumes per minute. The dissolved oxygen (DO) level was maintained at 20% by adjusting the agitation speed (300–900 rpm), and it was allowed to decrease to below 5% after 24 h. The initial concentration of glucose was set at 80 g/L. After depletion of the initial glucose, 800 g/L glucose and 17.6 g/L (C/N = 63) or 11.1 g/L (C/N = 100) ammonium sulfate were continuously added, thereby providing a stable carbon source for the growth and metabolism of Y. lipolytica.

Fluorescent protein visualization by laser scanning confocal microscopy

2.4

The ER targeting signal KDEL was used to localize the gene to the ER, CRMVGKSKL to target it to the peroxisome, and oleosin to target it to LD. The targeting efficiency of organelle-specific signal peptides was evaluated by cultivating Y. lipolytica strains expressing mCherry, mCherry-KDEL, mCherry-CRMVGKSKL, and Oleosin-mCherry in YNB medium supplemented with the required amino acids for 3 days. The cells were then washed twice with phosphate-buffered saline (Sangon, Shanghai, China). Further, 5 μL of the cell suspension was placed on a glass slide and visualized using a Nikon AX-SIM S laser scanning confocal microscope at an excitation wavelength of 516 nm [25]. Fig. S2 shows that the localization signals could direct mCherry to the ER, peroxisomes, and LD.

Analytical methods

2.5

The OD_600_ was measured using a UV-visible spectrophotometer (Jinghua Instruments, Shanghai, China). The dry cell weight (DCW) was determined by centrifuging 2 mL of the fermentation broth at 8000g for 3 min. The resulting cell pellet was freeze-dried for 48 h and subsequently weighed to obtain the DCW.

Glucose was quantified by centrifuging 2 mL of culture broth at 13,500g for 3 min. The supernatant was filtered through a 0.22-μm membrane. Glucose concentration was measured using a biochemical sensor analyzer (Siemens, Shenzhen, China).

The fatty acid composition was analyzed by mixing about 2 mg of freeze-dried biomass with heptadecanoic acid methyl ester as an internal standard. The sample was suspended in 500 μL of 1 M NaOH–methanol solution and shaken at 1200 rpm for 6–8 h. Subsequently, 40 μL of 98% H_2_SO_4_ and 500 μL of n-hexane were sequentially added, followed by shaking at 1200 rpm for 30–40 min to extract fatty acid methyl esters. The mixture was then centrifuged at 8000g for 2 min. The upper organic phase was collected and transferred into glass vials for gas chromatography analysis [16]. Gas chromatographic analysis was performed using an RTX-5MS column (30 m × 0.25 mm × 0.25 μm). The injector temperature was set at 260 °C. High-purity helium was used as the carrier gas at a flow rate of 1.0 mL min^−1^. The injection volume was 1.0 μL with a split ratio of 1:10. The oven temperature program was as follows: initial temperature 140 °C, held for 5 min; then increased at 4 °C·min^−1^ to 240 °C and held at 240 °C for 15 min.

The composition of total fatty acids (TFAs) was determined based on the following components: palmitic acid (C16:0), palmitoleic acid (C16:1 Δ9), stearic acid (C18:0), oleic acid (C18:1 Δ9), linoleic acid (C18:2 Δ9Δ12), arachidic acid (C20:0), cis-11-eicosenoic acid (C20:1 Δ11), behenic acid (C22:0), cis-13-docosenoic acid (C22:1 Δ13), lignoceric acid (C24:0), and NA (C24:1 Δ15).

Three independent parallel experiments were conducted to ensure the reliability and reproducibility of the results. A significance level of 0.05 was set in the analysis to determine the statistical significance of the results. All statistical tests were executed as two-tailed tests. In figures, NS stands for Not Significant, indicating p > 0.05; ∗ denotes 0.01 < p < 0.05, ∗∗ denotes 0.001 < p < 0.01, and ∗∗∗ denotes p < 0.001.

Results

3

Construction of the NA biosynthetic pathway

3.1

In the de novo biosynthetic pathway of NA (Fig. 1), KCS functions as the key rate-limiting enzyme within the fatty acid elongase complex, using substrates of different chain lengths. Modulating the expression of specific KCS isoforms directly influences the chain length and composition of very-long-chain fatty acids (VLCFAs) [26]. The fatty acid profile of wild-type Y. lipolytica primarily consists of C16:0, C16:1 Δ9, C18:0, C18:1 Δ9, and C18:2 Δ9Δ12. Therefore, NA synthesis requires the introduction of heterologous KCS enzymes. The background strain polf-Δku70 was further engineered by deleting tgl4 and pex10, which are associated with fatty acid degradation, to improve the accumulation of fatty acids [17]. Subsequently, six KCS genes from different sources were evaluated. Among these, LaKCS enabled the production of 1.2% NA, AtKCS generated 16.3% eicosenoic acid, and CraKCS yielded 4.3% erucic acid (Fig. 2A). Given the abundance of oleic acid in Y. lipolytica, the use of a single KCS enzyme was insufficient to drive efficient stepwise elongation of acyl-CoAs toward NA. This limitation was addressed by combining AtKCS (preferentially using oleoyl-CoA) [27], CraKCS (preferring eicosenoyl-CoA), and LaKCS (favoring erucoyl-CoA), increasing the NA content to 3.65%. Additionally, the activity of the mfe1 gene, which encodes a core multifunctional enzyme in the peroxisomal β-oxidation pathway, was disrupted to inhibit fatty acid degradation. This modification increased NA content to 4.16% of TFAs (Fig. 2C).Fig. 1Biosynthesis of NA from glucose. The biosynthesis of NA involves elongation, storage, and degradation modules.Ac-CoA, Acetyl-CoA; DAG, diacylglycerol; DHAP, dihydroxyacetone phosphate; FBP, fructose-1,6-bisphosphate; FFA, free fatty acid; GAP, glyceraldehyde 3-phosphate; G3P, glycerol-3-phosphate; LPA, lysophosphatidic acid; Mal-CoA, malonyl-CoA; mfe1, multifunctional enzyme; PA, phosphatidic acid; pex10, peroxisome biogenic gene; Pyr, pyruvate; TAG, triacylglycerol; tgl4, TAG lipase encoding gene.Fig. 1. Fig. 2Evaluation of the effect of single or combined KCS expression on NA production based on the substrate specificity of KCS.(A) Screening of KCS variants. (B) Schematic diagram of gene knockout. (C) Combined expression of KCS genes and knockout of degradation-related genes.Fig. 2

Semi-rational engineering of key enzymes CraKCS and LaKCS

3.2

The aforementioned results showed that the key enzymes CraKCS and LaKCS exhibited broad substrate specificities, participating in multiple elongation reactions involving oleic acid, eicosenoic acid, and erucic acid, thereby influencing the overall efficiency of NA biosynthesis. To favor NA formation, semi-rational design was performed on both enzymes, with the aim of improving the utilization of C20:1 Δ11-CoA by CraKCS and C22:1 Δ13-CoA by LaKCS. According to the molecular docking model (Fig. 3A and C), alanine scanning was conducted to identify residues located within 5 Å of the substrate-binding pocket that were predicted to influence substrate recognition. The E130A mutation in LaKCS increased the NA content from 4.16% to 5.17% (Fig. 3B). When combined with mutations in CraKCS, the E294A variant of CraKCS further elevated NA contents to 6.10%, respectively (Fig. 3D). Although detailed kinetic parameters could not be determined in this study, these in vivo production data indicate that the engineered KCS variants contribute effectively to NA biosynthesis.Fig. 3Modification of substrate specificity of LaKCS and CraKCS.(A) Structural analysis of LaKCS-binding domains with erucoyl-CoA and cofactors. (B) Alanine mutagenesis of LaKCS. (C) Structural analysis of CraKCS-binding domains with eicosenoyl-CoA and cofactors. (D) Alanine mutagenesis of CraKCS.Fig. 3

Optimization of the precursor fatty acid supply module enhanced the biosynthesis efficiency of NA

3.3

The key elongases LaKCS and CraKCS were semi-rationally engineered to improve the biosynthetic efficiency of NA. However, the titer of the resulting strain remained relatively low, suggesting precursor substrate availability as a limiting factor. In this study, the fatty acid desaturase gene fad2, which catalyzes the conversion of C18:1 Δ9 into C18:1 Δ9Δ12, was deleted to increase C18:1 Δ9 content. In the engineered strain NA26, the C18:1 Δ9Δ12 content decreased to zero, whereas the C18:1 Δ9 content increased by 4.39%. In addition, the level of C20:1 Δ11 increased significantly, reaching 16.05% (Fig. 4B). Moreover, C16:0 accumulated in NA26, indicating a bottleneck in fatty acid chain elongation. Elongases from various sources were introduced to further promote the production of longer-chain fatty acids. Among these, the strain expressing MaELO2 (NA27) efficiently converted C16:0 into C18:0. Despite no significant increase in the C18:1 Δ9 content, the C20:1 Δ11 content reached 20.1% and the NA content increased to 6.47% (Fig. 4C).Fig. 4Enhancement of oleic acid supply and promotion of C16:0-CoA conversion.(A) Strategy for increasing oleic acid availability and strengthening the desaturase. (B) Changes in NA content following fad2 knockout. (C) Effect of introducing elongase genes on NA accumulation.Fig. 4

Oxygen–iron synergistic enhancement of desaturase activity

3.4

The fermentation analysis revealed that strain NA27 accumulated a large amount of saturated tetracosanoic acid. The desaturation step converting tetracosanoic acid into NA represents the key rate-limiting reaction in the NA biosynthetic pathway. This conversion was facilitated by introducing a Δ15 desaturase to catalyze the desaturation of C24:0-CoA to C24:1 Δ15-CoA. The engineered strain NA30 expressing Maω9 efficiently catalyzed the conversion of tetracosanoic acid into NA, increasing the proportion of NA to 13.20% of TFAs (Fig. 5A). Because these desaturation reactions require sufficient oxygen and Fe^2+^, the next step was to improve cellular oxygen and iron availability. This was achieved by co-expressing Vitreoscilla hemoglobin (VHb) and a high-affinity iron transporter (Ftr1). As a result, the engineered strain NA31 exhibited a further increase in NA content, which reached 14.16% of TFAs (Fig. 5B). Also, NA31 exhibited elevated intracellular iron levels and improved oxygen availability compared with the control strain (Fig. S1). These results suggest that enhancing oxygen and iron can facilitate desaturase steps in NA biosynthesis, thereby contributing to the observed increase in NA accumulation.Fig. 5Oxygen–iron synergistic enhancement of desaturase activity.(A) Effect of expressing Δ15 desaturase on NA content. (B) Synergistic effect of VHb and Ftr1 on NA production.Fig. 5

Efficient synthesis of NA through subcellular engineering and dynamic promoter coordination

3.5

The biosynthesis of NA involves the coordinated activity of multiple cellular organelles. The catalytic environment of the key elongases was optimized by combinatorially expressing AtKCS, CraKCS, and LaKCS and individually targeting them to LD, the ER, and peroxisomes to examine the effects of subcellular localization on product formation. The strain NA33, in which the enzymes were localized to the ER, performed the best, achieving an NA content of 15.88% (Fig. 6A). These results indicated that the ER served as the primary site for NA biosynthesis. An ER expansion strategy was employed to strengthen ER-associated metabolic capacity. Accordingly, opi1 was deleted in the NA33 background strain, resulting in a slight increase in NA content (15.92%) but a notable increase in NA titer from 214.38 to 262.65 mg/L (Fig. 6B). Subsequent overexpression of INO2 further enhanced NA accumulation, with the content increasing to 16.45%, suggesting that ER expansion was beneficial for lipid biosynthesis (Fig. 6B). Expansion of the ER membrane was observed using TEM (Fig. S3).Fig. 6Organelle engineering to enhance NA production.(A) Comparison of NA content when key enzymes were targeted to different organelles. (B) Expansion of the ER to improve lipid synthesis capacity. (C) Schematic representation of promoter regulation. (D) Dynamic regulation of NA biosynthesis by pGPD1 and pLDP1.Fig. 6

Furthermore, temporal optimization of key enzyme expression was achieved by constructing a dynamic promoter regulation system based on promoter engineering. The lipid-accumulation-phase-inducible promoter pLDP1 and the robust constitutive promoter pGPD1 were used to control the expression of LaKCS. The strain NA39, expressing LaKCS under pGPD1, achieved the highest NA content of 17.13%, surpassing the pLDP1-driven counterpart (Fig. 6D). This result suggested that continuous, high-level expression of key elongases was more favorable for sustained NA biosynthesis in Y. lipolytica.

Enhancing malonyl-CoA and NADPH supply for improved NA biosynthesis

3.6

The de novo synthesis of NA requires malonyl-CoA as the two-carbon donor and NADPH as the reducing power. The enzyme acetyl-CoA carboxylase (ACC1) catalyzes the direct conversion of acetyl-CoA into malonyl-CoA. In addition, a nonnatural biosynthetic route for malonyl-CoA formation (MCR–C–BauA) has been shown to enhance intracellular malonyl-CoA levels in Y. lipolytica [28]. The pentose phosphate pathway serves as the major source of NADPH in Y. lipolytica [29]. Introducing the NCM pathway in strain NA41 markedly strengthened its ability to synthesize NA, achieving a content of 19.57% (Fig. 7B). Building upon NA41, strains NA42–NA50 were individually engineered to overexpress nine genes related to NADPH regeneration: ZWF1, GND1, IDP2, POS5, TKL1, TKL2, TAL, MAE, and PYC. Among these, GND1 overexpression significantly increased NA accumulation to 21.82% (Fig. S4). The NADPH supply was further enhanced by co-expressing GND1 with other NADPH-related genes, including IDP2, POS5, ZWF1, and TKL1. The strain NA51, co-expressing GND1 and IDP2, performed the best titer, producing 451.80 mg/L of NA, corresponding to 22.33% of TFA (Fig. 7C).Fig. 7Enhancement of cofactor supply for improved NA biosynthesis.(A) Engineering of cofactor supply. (B) Effect of elevating intracellular malonyl-CoA levels on NA accumulation. (C) Influence of enhanced NADPH regeneration on NA production.Fig. 7

Fed-batch fermentation

3.7

The engineered strain NA51 was fermented in a 5-L bioreactor under fed-batch fermentation conditions to maximize NA production. In the initial fermentation phase, dissolved oxygen (DO) was maintained at 20% by adjusting agitation speed and aeration rate. As fermentation progressed, the pH gradually decreased. In the bioreactor fermentation shown in Fig. 8A, the pH was controlled with 10 M NaOH, whereas in Fig. 8B it was adjusted using aqueous ammonia. Glucose was fed at a C/N ratio of 63 in Fig. 8A and at a C/N ratio of 100 in Fig. 8B, maintaining the residual glucose concentration at approximately 10 g/L in the bioreactor. As shown in Fig. 8, NA accumulated significantly after 40 h of cultivation and continued to increase throughout the lipid accumulation phase, ultimately reaching a maximum titer of 6.55 g/L in 140 h under a C/N ratio of 63, with NA accounting for 29.81% of TFAs. This result demonstrated that the engineered metabolic network in NA51 could sustain efficient NA biosynthesis under controlled fermentation conditions, highlighting the importance of process optimization in achieving high product titers.Fig. 8**De novo biosynthesis of NA in a 5-L fermenter by strain NA.**Fermentation of strain NA51 in a 5-L bioreactor (A) under C/N ratios of 63 and (B) under C/N ratios of 100.Fig. 8

Discussion

4

NA plays an indispensable role in neural development, neuroprotection, infant nutrition, and mental health, and holds immense potential for applications in functional foods and biomedicine [30]. However, NA biosynthesis is restricted by the low substrate specificity and affinity of 3-KCSs and the limited metabolic efficiency in the lipid accumulation phase [11,26]. In this study, a de novo NA biosynthetic pathway was established in Y. lipolytica through the screening and combinatorial expression of KCSs from diverse sources. The combination of enzymes with distinct substrate preferences enabled continuous elongation from C18:1 to C24:1. The semi-rational design of CraKCS and LaKCS improved enzyme specificity, whereas the enhanced accumulation of C18:1 Δ9 and facilitated the synthesis of NA. In addition, improved oxygen utilization and iron transport strengthened desaturase performance to promote the conversion of C24:0-CoA.

KCSs from various organisms possess distinct chain-length preferences toward acyl-CoA substrates, hampering the efficient use of long-chain unsaturated fatty acids (C20:1 and C22:1) required for NA synthesis. Liu et al. [31] employed KCSs with varying substrate specificities to successfully catalyze the stepwise elongation of oleic acid into NA through a cascade reaction. However, their catalytic activity was not restricted to specific substrates, which also represents a key factor influencing elongation efficiency [26]. Ghanevati et al. [32] elucidated the active-site residues of KCS, thus providing a structural basis for enzyme engineering and functional modification. Coraline et al. [33] improved medium-chain fatty acid production by rationally redesigning the substrate recognition sites. Chen et al. [34] enhanced the overall substrate preference of PtKCS through a double-mutant strategy. Inspired by these studies, semi-rational engineering of CraKCS and LaKCS was performed to modulate their chain-length utilization, and obtained variants that increased the NA content in our engineered strain to 6.10% of TFAs. In this study, the impact of KCS engineering was primarily assessed at the strain level, focusing on NA production and VLCFA product profiles instead of detailed enzyme kinetic analysis. A robust in vitro system for CraKCS and LaKCS enabling reliable determination of Km and kcat values with the relevant VLCFA-CoA substrates could not be established, the absence of kinetic parameters constitutes a limitation of the present work. Future studies will focus on purifying CraKCS and LaKCS and developing quantitative in vitro assay systems to determine Km and kcat values for the wild-type and mutant enzymes. Such data will provide the most direct mechanistic evidence for the effects of the LaKCS^E130A^ and CraKCS^E294A^ mutations.

The desaturase is also a rate-limiting enzyme in NA biosynthesis, and enhancing its catalytic activity can significantly increase NA production. C24:0 was successfully converted into NA by expressing a Δ15-desaturase, thereby reducing the excessive accumulation of tetracosanoic acid [2,20]. In this study, the expression of Maω9 enabled the efficient conversion of C24:0 into NA. However, the desaturation reaction depends on both molecular oxygen and ferrous ions; thus, improving oxygen utilization and strengthening iron supply are beneficial for enhancing desaturase catalytic efficiency [35]. Guerreiro et al. [36] achieved excessive lipid accumulation in engineered strains by maintaining elevated dissolved oxygen levels. Zhang et al. [37] introduced VHb into Mucor circinelloides, thereby alleviating oxygen limitation and improving both cell growth and fatty acid synthesis. Cordova et al. [38] demonstrated that supplementing the culture medium with iron ions tripled total lipid content, accompanied by an increase in oleic acid proportion with increasing iron concentration. The high-affinity iron transporter Ftr1 in yeast can efficiently transfer extracellular Fe^2+^ into the cytosol [39]. Inspired by these findings, this study was novel in establishing a synergistic regulation strategy integrating oxygen and iron metabolism to improve desaturase steps in NA biosynthesis, resulting in NA accounting for 14.16% of TFAs.

The spatial distribution and temporal expression of enzymes influence the efficiency of NA biosynthesis. Conventional metabolic engineering usually emphasizes expression strength but overlooks the coordination between subcellular localization and dynamic control. Renne et al. [40] and Tehlivets et al. [41] highlighted the crucial roles of subcellular organelles in fatty acid synthesis and regulation. Zhu et al. [42] and Liu et al. [43] demonstrated that the pLDP1 promoter displayed high transcriptional activity in the lipid accumulation phase and was regarded as a strong promoter in Rhodotorula toruloides. The robust promoter pGPD1 enhances triacylglycerol formation and lipid accumulation by synchronizing gene expression with the onset of fatty acid biosynthesis [44,45]. A combined subcellular engineering and dynamic promoter system was constructed in this study to optimize NA synthesis in both dimensions. The localization of key elongases to the ER improved substrate channeling, whereas the dynamic expression under pGPD1 enabled synchronization with the early fatty acid accumulation stage. This approach provided self-regulating balance across metabolic stages, representing a concise and efficient strategy for lipid-pathway optimization.

In summary, this study achieved de novo NA biosynthesis in Y. lipolytica via a multifaceted metabolic engineering approach. This approach integrated techniques such as enzyme variant screening, precursor pool augmentation, oxygen–iron synergistic enhancement of desaturase activity, subcellular engineering and dynamic promoter regulation, and cofactor balancing. The resulting strain NA51 produced 451.80 mg/L NA in shake-flask culture and 6.55 g/L NA in a 5-L fed-batch bioreactor. Although this value is lower than the highest titer reported so far in Y. lipolytica, namely 17.3 g/L obtained by Su et al. in a 50-L fermenter [1], several factors may contribute to this difference. These include variations in strain background, the specific combinations and expression levels of elongases and desaturases, and the extent of fermentation process optimization. This work improved NA biosynthesis by rationally engineering KCS specificity, tuning desaturase activity, and reinforcing precursor and cofactor supply. Thus, the current strain and pathway design provide an informed chassis that can be further improved. In future work, systematic fermentation optimization—such as implementing optimized fed-batch feeding profiles, adjusting the C/N ratio to favor lipid accumulation, and enhancing oxygen transfer and pH control, may substantially increase NA titers. Such approaches can enable higher yields, improved process stability, and greater tunability in Y. lipolytica–based lipid engineering. Overall, this study establishes a systematic framework for high-level NA biosynthesis and provides novel insights for the industrial production of VLCFAs and other value-added lipids.

CRediT authorship contribution statement

Sen Ye: Writing – review & editing, Writing – original draft, Visualization, Investigation. Wenjing He: Investigation. Niping Yang: Writing – review & editing. Qian He: Funding acquisition. Jingwen Zhou: Writing – review & editing, Supervision. Yu Xia: Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.