Recent Advances in Natural Product Biosynthesis and Yield Improvement Strategies Using Yarrowia lipolytica

Abstract

Microorganisms are increasingly being used for the industrial production of raw materials for food, chemical products and pharmaceuticals. The unconventional yeast Yarrowia lipolytica has a rising profile as a platform for industrial biotechnology. It has attractive physiological and metabolic properties, including high terpene and lipid production, high tolerance to complex environments, and amenability to genetic modification. Y. lipolytica naturally produces sufficient levels of cytosolic acetyl-CoA and malonyl-CoA to achieve lipid accumulation. Engineering biology methods allow transformation of these native metabolites into synthetic precursors for high-value compounds such as terpenes and flavonoids. Gene-editing, expression, and regulation tools have been developed for Y. lipolytica, facilitating improvement in bio-manufacturing yields for this chassis. This review summarizes natural product yields in Y. lipolytica and strategies for improving productivity. We highlight morphological engineering, metabolic engineering, and adaptive laboratory evolution as key strategies that can be used to improve the future yield, productivity and controllability of target molecules for Y. lipolytica engineering.

1. Introduction

Microorganisms are increasingly replacing traditional chemical synthesis for industrial production, driven by sustainability goals, cost efficiency, and raw-material flexibility [1,2]. To compete with traditional chemical synthesis, platform microorganisms must demonstrate rapid growth, easy cultivation, scalability, metabolic optimization, and regulatory and safety compliance [3]. Such microorganisms can be engineered to provide a broad factory platform for the synthesis of a wide range of chemical substances. Selection of the appropriate platform microorganism is a key criterion for the industrial success of microbial bioprocesses. Common platform chassis for the synthesis of bulk chemicals include laboratory Escherichia coli, Bacillus subtilis, Corynebacterium glutamicum, and Saccharomyces cerevisiae. Key example products include l-threonine, d-lactic acid, and l-lysine [4,5,6] (E. coli); hyaluronic acid, chondroitin sulfate, riboflavin, and N-acetylglucosamine (B. subtilis) [7,8,9]; and l-glutamic acid and l-lysine [10,11] (C. glutamicum). These microorganisms have a long history of safe use in biosynthesis.

Interest in the use of the unconventional yeast Yarrowia lipolytica for the production of natural products is growing because of its unique metabolic features and high productivity [12]. Y. lipolytica, a yeast in the order Saccharomycetales [13], is mainly found in lipid- and protein-rich environments. Y. lipolytica has GRAS (generally regarded as safe) status, making its products fit for human and animal consumption [14]. The whole genome sequence and mitochondrial genome sequence of Y. lipolytica are available [15]. Y. lipolytica has six chromosomes, with a genome size of about 20.5 Mb. Compared with other members of Saccharomycotina, such as Saccharomyces cerevisiae, Nakaseomyces glabrata, and Debaryomyces hansenii, Y. lipolytica exhibits a relatively lower gene density [16]. The molecular manipulation of Y. lipolytica is similar to that of S. cerevisiae, with well-established plasmid expression and gene recombination methods [17,18].

An important feature of Y. lipolytica is its dimorphism [19]. Y. lipolytica generally appears round or oval (yeast morphologies), as a short rod, or in a filamentous form (hyphal morphologies) when examined under a microscope (Figure 1). Colonies are white and smooth or wrinkled. Fine surface villi are often observed [20]. Factors including the composition of the culture medium, growth conditions, the strain chosen, and genetic background affect the ratio of yeast to hyphal cells [21,22,23,24,25]. Starvation [26], heat shock [27], pH [24], carbon/nitrogen ratios [28], and immune stress [29] affect morphology. Current research suggests that morphological transformation of Y. lipolytica is an adaptive strategy with which to cope with environmental changes or threats to survival [30].

Y. lipolytica produces a rich enzyme complement, including esterases, proteases, and lipases, providing strong substrate metabolic ability [31,32]. It can metabolize both common hydrophilic substrates, such as sucrose, glucose, fructose, polyols, and organic acids, and hydrophobic substrates such as lipids, alkanes, and alkenes [31,33,34]. Additionally, Y. lipolytica metabolism produces a large amount of acetyl-CoA [35,36]. Y. lipolytica therefore efficiently synthesizes molecules for which acetyl-CoA serves as a common precursor. These include lipids and terpenoids [37,38]. Y. lipolytica has a high metabolic flux through the pentose phosphate pathway (PPP), providing a large NADPH pool for lipid and terpenoid biosynthesis [38,39]. To store excess hydrophobic oils, Y. lipolytica contains abundant subcellular structures of lipid droplets. These provide storage sites for hydrophobic products such as carotenoids [40]. Y. lipolytica resists stresses effectively and can tolerate moderately high-salt, low-temperature, and acidic or alkaline culture environments [41,42,43]. These properties make it a versatile organism for metabolic engineering, with lower risks of contamination by organisms that cannot tolerate such conditions.

Y. lipolytica is therefore an attractive industrial chassis strain for producing a broad spectrum of chemicals. Developments in synthetic biology and metabolic pathway regulation strategies from other organisms have been deployed to improve the fermentation yield of Y. lipolytica [44,45]. This review summarizes progress in the heterologous synthesis of natural products using Y. lipolytica and highlights strategies for improving yields (the chemical structures of key products synthesized by Y. lipolytica are provided in the Supplementary Information). Current methods for improving metabolic yields focus mainly on the regulation of microbial metabolic flux networks [46] with little consideration of cell morphology. Given the impact of dimorphic transformation on the characteristics of Y. lipolytica, this review aims to introduce morphological engineering [47] as a route for regulating yield. We hope to inspire new ideas for regulating metabolic yield in Y. lipolytica.

2. Progress and Current Strategies in Natural Product Synthesis in Y. lipolytica

2.1. Lipids and Polymers

As a natural oil-producing yeast with high acetyl-CoA production flux capacity, Y. lipolytica is a very suitable chassis for the production of oils and fatty acids [48,49]. The high concentration of intracellular lipids provides an attractive platform for the synthesis and storage of other hydrophobic natural products. These include some of the important human polyunsaturated fatty acids (PUFAs), which are common dietary supplements. Typical PUFAs include arachidonic acid, eicosapentaenoic acid (EPA), and linolenic acids (Figure 2A) [50]. PUFA production in Y. lipolytica provides higher yields than extraction from animal and plant sources. For example, arachidonic acid production was successfully upregulated by fusing the enzymes Δ-9 elongase and Δ-8 desaturase to increase pathway flux [51].

Another common approach is to exploit Y. lipolytica’s rich substrate utilization range to optimize production of desired chemicals, often at a reduced cost. When engineered Y. lipolytica was used to produce poly-3-hydroxybutyrate, acetate provided a higher yield than glucose. The effective accumulation accounted for 3.84% and 1.50% of the dry cell weight, respectively [52]. Xu et al. established a semicontinuous process for efficient and complete upgrading of low-strength acetic acid into lipids in Y. lipolytica, with a final lipid titer of 115 g/L [53]. With overexpression of appropriate endogenous genes, Y. lipolytica can catabolize xylose to produce lipid precursors and use sugarcane bagasse hydrolysate as a growth medium [54].

The mechanisms of lipid accumulation and degradation in Y. lipolytica have been characterized extensively. Intracellular lipid accumulates from both de novo lipid biosynthesis and ex novo lipid biosynthesis [55,56]. The degradation of lipids in Y. lipolytica mainly commences via the β-oxidation pathway in peroxisomes. Metabolic engineering of these processes increases lipid yields. GPD1 overexpression, GUT2 inactivation, or both mutations together result in 1.5-, 2.9-, and 5.6-fold increases, respectively, in the level of glycerol-3-phosphate. This led to an increase in triacylglyceride (TAG) accumulation [57]. Alternatively, reducing peroxisome biogenesis via inactivation of PEX10 increased EPA production [58]. The engineered yeast lipid comprises 56.6% EPA and less than 5% saturated fatty acids by weight [59]. The production of EPA by Y. lipolytica strains has been commercialized [60]. These examples illustrate how metabolic engineering of Y. lipolytica provides an effective platform for the overproduction of lipids.

2.2. Sugar Alcohols

Sugar alcohols are highly desirable industrial products that are used as feedstocks for biological metabolism, and they serve as building blocks for drug synthesis and environmental protection products [61,62]. Y. lipolytica produces essentially no ethanol during fermentation, and so ethanol production does not compete with other alcohols [63]. However, it naturally produces erythritol in response to osmotic stress [64,65]. Consequently, this yeast has been applied in alcohol biosynthesis. In Y. lipolytica and other osmotolerant yeasts, the synthesis of erythritol starts from erythrose-4-phosphate, an intermediate of the pentose phosphate pathway (PPP) (Figure 2B). Overexpression of erythrose reductase (YALI0F18590g) enhanced erythritol production by 20% compared to a control [66]. As seen with lipids, fermentation methods and culture media components have a significant impact on erythritol production. Addition of sorbitan monolaurate to fed-batch cultures with glycerol as a carbon source increased erythritol production by 15% compared with a control [67]. Threitol (TOL), a diastereoisomer of erythritol, was synthesized by overexpressing the gene xylitol dehydrogenase (Ss-XDH) from Schefferomyces stipites CBS6054 in Y. lipolytica. A maximum titer of 112 g/L was achieved [68].

Mannitol (MTL) is a six-carbon sugar alcohol that plays an important role in microbial stress tolerance. The accumulation of intracellular mannitol helps cells resist metabolically generated reactive oxygen species under adverse conditions [69]. Mannitol production by Y. lipolytica is often accompanied by erythritol as a by-product. In one case, addition of NaCl to the culture medium improved erythritol biosynthesis while inhibiting mannitol formation [70]. Zhang et al. engineered a strain adapted to high temperatures by overexpressing the hsp90 gene. This allowed Y. lipolytica to produce mannitol at 34–35 °C, greatly reducing the cooling cost of fermentation [71]. Using low-cost culture substrates for Y. lipolytica is attractive as a route to commercial chemical production. For example, using olive mill wastewater as a substrate for Y. lipolytica alcohol production did not affect production value, providing mannitol yields between 5.3 g/L and 13.1 g/L [72].

The five-carbon sugar arabitol has important applications in the food industry as a sweetener and adipose tissue reducer. The best-demonstrated yield of arabitol by Y. lipolytica was 118.5 g/L in 264 h in a fed-batch system controlling osmotic pressure with glycerol [72,73]. This outcome is nearly three times the productivity found in other microbial cells tested for arabitol production [74]. Together, these examples highlight that common approaches can be deployed individually or in combination to optimize Y. lipolytica productivity.

2.3. Terpenoids

Terpenoids are industrially important chemicals widely applied in the fragrance, pharmaceutical, agricultural, and chemical industries [75]. Terpenoids also have potential applications in environmental protection (e.g., limonene derivatives) [76]. Y. lipolytica generates isoprenoids through the mevalonate pathway (MVA pathway) [77]. This converts intracellular acetyl-CoA into isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP). IPP and DMAPP are condensed to form geranyl pyrophosphate (GPP), farnesyl pyrophosphate (FPP), and geranylgeranyl pyrophosphate (GGPP) under the action of farnesyl pyrophosphate synthase (FPPS)/geranylgeranyl pyrophosphate synthase (GGPPS), respectively [78,79,80]. The GPP, FPP, and GGPP endogenously produced by Y. lipolytica are important precursors for the biosynthesis of various terpenoids (Figure 3). The greater natural acetyl-CoA flux of Y. lipolytica provides greater capacity for terpenoid synthesis compared to E. coli and S. cerevisiae.

Carotenoids are an important class of natural pigments. They are the main source of vitamin A in the body and have antioxidant, immunomodulatory, anti-cancer, and anti-aging effects [81]. Lycopene is a carotenoid with a relatively simple structure, and it is an important precursor for the synthesis of most other carotenoid compounds. Y. lipolytica synthesizes carotenoids from GPP, FPP and GGPP [82,83] precursors using a cyclase or dehydrogenase. AMP deaminase has an important role in lycopene synthesis: in one case, optimizing expression of this gene raised lycopene yields to 46–60 mg/g dry cell weight (DCW), the highest lycopene content reported in Y. lipolytica [84]. β-carotene production was increased 100-fold by using strong promoters and overexpressing 11 genes producing β-carotene and its precursors. Promoter optimization in this fashion is a common strategy. Combined with fed-batch fermentation using optimized media, the final β-carotene titer was 4 g/L [85]. Golden Gate assembly is an effective strategy for screening promoter–gene pairs for optimal expression [86]. Using this approach, β-carotene production was further raised to 6.5 g/L [87]. Another common strategy is to regulate the flux of pathways by providing metabolic precursors. Two such approaches further increased β-carotene yields. Protein engineering of lycopene cyclase relieved the substrate inhibition that conventionally limits product levels without reducing enzyme activity [88]. Conversely, oversupply of the substrate GGPP can cause substrate inhibition. Replacing the natural GGPP synthase with a lower-rate enzyme regulated metabolic flux, overcoming lycopene synthase substrate inhibition. A strain capable of producing 39.5 g/L β-carotene at 0.165 g/L/h in bioreactor fermentation was established; this is the highest recorded yield of β-carotene synthesized by microorganisms [88]. As a high-value antioxidant xanthophyll carotenoid, astaxanthin is widely used in pharmaceuticals and nutraceuticals. Through metabolic pathway engineering, enzyme complex construction, and fermentation optimization, its production capacity in Y. lipolytica was improved to 2820 mg/L [89].

Limonene, a cyclic monoterpene, has anti-inflammatory, anti-allergic, antioxidant, and antibacterial effects [90,91]. Limonene is formed via the cyclization of geranyl pyrophosphate. Limonene shows cytotoxicity at higher concentrations, making its industrial biosynthesis challenging. Expression of neryl diphosphate synthase 1 (NDPS1) and limonene synthase (LS) results in limonene production in Y. lipolytica. By optimizing the substrate pyruvic acid and dodecane concentrations in a flask culture, a maximum limonene titer and content of 23.56 mg/L and 1.36 mg/g DCW were achieved using the engineered strain Po1f-LN-051 [92]. Using the common strategy of introducing an additional copy of the limonene synthesis gene, along with choosing glycerol as the carbon source (and citrate as an auxiliary carbon source), improved the limonene yield to 165.3 mg/L in one study [93]. This yield remains lower than the highest yield obtained using E. coli (1.29 g/L [94]). However, this required far more extensive modification, and Y. lipolytica can likely be further optimized in the same manner as employed for β-carotene production.

Ginsenosides are triterpenoid compounds and the main active ingredient in ginseng-based medicinal materials [95,96,97]. Ginseng is an important component in traditional Chinese medicine [98]. Ginsenosides have similar basic structures and are divided into the oleananes and protopanaxatriols/protopanaxadiols according to their core skeletons (Figure 4) [99]. The specific ginsenosides Rg1, Ro, and CK are associated with fatigue relief [100], inflammation reduction [101], and anti-tumor activity [102], respectively. Ginsenosides show no adverse effects in animals except at very high doses; treatment of humans with 114 μg/kg of ginseng for 12 weeks showed no adverse effects [103]. The traditional chemical extraction method used for ginsenosides has a long cycle and low yields, leading to a high price for ginsenoside products [104,105]. Microbial production of ginsenosides requires introduction of the final biosynthetic genes. Strategies for increasing yield include fusion of these biosynthetic genes (e.g., fusing the required cytochrome P450 to its NADPH-P450 reductase to optimize oleanolic acid production [106]). Alternatively, xylose reductase and xylitol dehydrogenase can be introduced to allow use of xylose as the sole carbon source, thus nearly doubling the yield of protopanaxadiol [107].

Two other terpenoids of industrial importance are linalool and farnesenes [108]. Like ginsenosides, the necessary biosynthetic genes have been added to Y. lipolytica to exploit the terpenoid precursor pools [109]. Improving the GPP/FPP pool by overexpressing genes in the MVA pathway is a common strategy, providing 70- and 20-fold increases in the production of linalool and α-farnesene, respectively [109,110]. Using a library approach to identify the most effective genes to overexpress further increased the yield of α-farnesene to 25.6 g/L [111], the highest microbial titer reported. Disruption of diacylglycerol kinase (DGK1) to reduce fatty acid production also proved effective, leading to a 15-fold linalool yield increase compared to overexpressing MVA pathway genes [112]. Similarly to other products, feedstock optimization is an effective approach. Bao et al. applied nitrogen-limited conditions to maximize carbon flux towards β-farnesene synthesis. The optimal condition led to a C/N ratio of 100 and increased the β-farnesene titer by 227% [113]. Carbon flux has been regulated by removing the native 6-phosphofructokinase and introducing an exogenous non-oxidative glycolysis pathway. This connects the pentose phosphate pathway and the mevalonate pathway, leading to a β-farnesene titer of 28.9 g/L in a 2 L bioreactor [114]. The acyclic triterpene squalene can spontaneously accumulate in Y. lipolytica and be used for the synthesis of downstream metabolites. However, natural accumulation is low, limiting the yield of downstream products. By augmenting the acetyl-CoA supply in peroxisomes and the cytoplasm, Ning et al. [115] obtained 51.2 g/L of squalene in a 5 L bioreactor. This suggests that industrial synthesis of terpenes and squalene derivatives in Y. lipolytica is feasible. Lipids have also been used as fermentation substrates, and they greatly reduce by-product accumulation [116]. Future research is likely to combine these complementary approaches that exploit the common core terpenoid biosynthetic pathway for application to new terpenoids.

2.4. Flavonoids

Naringenin, scutellarin and p-coumaric acid are three important flavonoids with significant antioxidant, anti-inflammatory and antimicrobial properties [117,118,119,120]. Flavonoids are biosynthesized through a common pathway from phenylalanine, with p-coumaric acid and naringenin serving as key intermediates (Figure 5A) [121,122]. Y. lipolytica naturally produces p-coumaric acid [123], and insertion of the relevant biosynthetic genes allows production of further flavonoids. Flavonoid biosynthesis has been optimized using approaches similar to those used for other biomolecules. Optimization of p-coumaric acid yields was achieved using a pull–push approach. The push was achieved by overexpressing bottleneck enzymes in the pentose phosphate, shikimate (from which tyrosine is derived), and flavonoid pathways, whilst pyruvate kinase was deleted to eliminate a competing pathway. With optimization of the feedstock, 593 ± 29 mg/L of p-coumaric acid was produced [124]. Naringenin synthesis was similarly achieved by adding biosynthetic genes. Overexpression of bottleneck genes, controlling pH, and optimizing the C/N ratio incrementally increased bioproduction more than 10-fold, rising to 252.4 mg/L [125]. An alternative approach involving the expression of xylose utilization genes (as used with terpenoids) resulted in nearly three-fold-greater production [126]. Compared with traditional naringenin extract, naringenin extracts produced via Y. lipolytica show significantly fewer DPPH· and ABTS· radicals [127]. A combination of enzyme engineering, precursor supply enhancement, multicopy pathway integration, and fed-batch fermentation optimization increased the yield of (2S)-naringenin to 8.65 g/L. This is the highest yield of this compound achieved for Y. lipolytica [128]. Wang et al. constructed a heterologous synthesis route for the de novo production of the flavonoid scutellarin in Y. lipolytica. Production was optimized and improved 23-fold, increasing to 346 mg/L, by screening libraries of biosynthetic enzymes from different species and using a fed-batch bioreactor. This is the highest yield obtained using a microorganism [129].

2.5. Organic Acids

Organic acids are biosynthetic organic compounds containing one or more low-molecular-weight acidic groups (such as carboxyl or sulfonic-acid groups). Itaconic acid, acetic acid, α-ketoglutaric acid, succinic acid, and crotonic acid are widely used in chemical, food additive, and pharmaceutical production (Figure 5B) [131]. Compared with bacteria and filamentous fungi, Y. lipolytica generally shows stronger low pH tolerance, making it an attractive host for organic-acid production. Producing acids at low pH allows isolation of pure acid, whilst at neutral pH, salts are formed, which require expensive separation [132]. Y. lipolytica robustly produces α-ketoglutarate (α-KG). Optimization of culture conditions allowed production of 47 g/L of α-KG using n-paraffin as a substrate [133]. Further substrate optimization has used glycerol, cellulose [134], and rapeseed oil [135], with glycerol producing a titer of 138 g/L. To improve yields, metabolic engineering approaches have focused on overexpressing biosynthetic genes from other fungi [136,137]. Expression of heterologous NADP⁺-dependent isocitrate dehydrogenase and pyruvate carboxylase increased α-KG production to 186 g/L [137].

Succinate can be prepared chemically from α-KG [138]. However, with optimization of culture conditions (particularly lowering the pH to 3.65), succinate can be directly produced by Y. lipolytica [139]. Again, a genetic approach improved yields. In this case, as succinate is turned over in vivo as part of the citric acid cycle, a natural mutant in succinate dehydrogenase provided greater succinate accumulation. This mutant required extensive in vitro evolution to allow it to grow efficiently on cost-effective substrates and achieved a 60% improvement over the previously reported conditions [140].

Itaconic acid (IA) is a small-molecule metabolite used as a petroleum-replacement monomer in plastics and rubbers [141,142]. IA can be produced in Y. lipolytica through heterologous expression of itaconate synthase to an initial titer of 33 mg/L. In further genetic engineering, a pull approach was employed by overexpressing precursor production genes and inhibiting a competing pathway. Media optimization strategies (especially optimizing the C/N ratio) further improved yields, leading to a 140-fold increase in titer [143]. Overexpression of the mitochondrial cis-aconitate transporter further improved yields 5-fold, raising levels to 22 g/L IA, the highest reported IA yield in a low-pH environment [144].

The unsaturated fatty acid crotonic acid is used in fungicides, resins and pharmaceutical intermediates [145]. Crotonate production in Y. lipolytica was achieved by introducing biosynthetic genes. Overexpressing orthologous genes to produce precursors for the pathway led to a 3.5-fold increase in production, with this value increasing to 220 ± 8 mg/L [130].

2.6. Others

The metabolic flexibility of Y. lipolytica has made it an attractive production chassis for a diverse range of chemicals, which exploit the yeast’s natural metabolism and genetic tractability. Adding genes to convert Y. lipolytica metabolites from the extensive lipid pool into alkanes has successfully delivered high yields. Initial approaches involved the introduction of soybean lipoxygenase to decompose linoleic acid into pentane [146]. Although overexpression of lipoxygenase more than tripled the yield, the production of by-products limited the potential of this approach. Greater success has been achieved by decarboxylating fatty acids. The natural metabolism of peroxisomes was initially exploited by directing fatty acid reductases and fatty aldehyde dehydrogenases/decarbonylases to this compartment [147]. This approach was improved by using fatty acid photodecarboxylase (FAP) from Chlorella variabilis instead. With overexpression of the (slow) FAP enzyme and optimization of the C/N ratio, a titer of 1.47 g/L from glucose was achieved [148].

The nucleoside antibiotic cordycepin, a fungal natural product, is a derivative of the nucleoside adenosine that is being explored as an anti-tumor agent, among other biological roles [149,150]. Harvesting cordycepin from its natural source is insufficient to meet market demand [151]. Y. lipolytica can synthesize cordycepin upon introduction of the biosynthetic pathway genes [152]. Strategies similar to those presented earlier have optimized production. Promoter optimization, fusion of biosynthetic genes, overexpression of enzymes supplying precursors and cofactors, and determining optimal feedstocks [152,153] have allowed progressive yield improvements. The highest yield recorded is 4.36 g/L [153,154].

Methods similar to those discussed above have been exploited to optimize Y. lipolytica-based production of a range of other chemicals for which green synthesis is desirable. Good examples are the glycoside arbutin [155] (used in the skincare industry [156]), violacein [157] (an antimicrobial [158]), and triacetic acid lactone [159,160] (used as a plasticizer and in adhesive manufacturing [159]). To summarize, the most common strategies used for Y. lipolytica to date have been promoter optimization or including copies of orthologous biosynthetic genes, fusing enzymes to increase flux, overexpressing pathways providing precursors, controlling the C/N ratio, and optimizing the feedstock (

Table 1. Summary of industrially relevant chemicals produced by Yarrowia lipolytica and important engineering strategies. Where different approaches are highlighted, these are indicated by #1, #2, etc.

Product (by Class)	Maximum Yields Achieved	Metabolic Engineering Strategy	Reference(s)
Lipids
Arachidonic acid	118.1 mg/L	Fuse Δ-9 elongase and Δ-8 desaturase	[51]
Eicosapentaenoic acid	EPA at 15% of dry cell weight	Inactivate PEX10 to reduce peroxisome biogenesis	[59]
poly-3-hydroxybutyrate	7.35 g/L	Utilize acetate as substrate; use fed-batch fermentation	[52]
Sugar alcohols
Erythritol #1	44.5 g/L; 20% yield increase compared to control	Overexpress erythrose reductase	[66]
Erythritol #2	142 g/L; 15% yield increase compared to control	Add sorbitan monolaurate to fed-batch cultures with glycerol as a carbon source	[67]
Threitol	112 g/L	Overexpress xylitol dehydrogenase	[68]
Mannitol	5.3 g/L ~ 13.1 g/L	Overexpress Hsp90; add olive mill wastewater to media	[72]
Arabitol	118.5 g/L	Control osmolarity and optimize C/N ratio	[72,73]
Terpenoids
β-carotene #1	4 g/L	Overexpress 11 biosynthetic genes; use fed-batch fermentation	[85]
β-carotene #2	6.5 g/L	Optimize promoter–gene pairs for expression	[87]
β-carotene #3	39.5 g/L	Regulate metabolic flux	[88]
Lycopene	46–60 mg/g DCW (dry cell weight)	Optimize and overexpress related genes	[84]
Astaxanthin	2820 mg/L	Employ metabolic pathway engineering and enzyme complex construction	[89]
Limonene #1	23.56 mg/L	Optimize the substrate pyruvic acid and dodecane concentrations in flask culture	[92]
Limonene #2	165.3 mg/L	Introduce an additional copy of limonene synthesis gene	[93]
Oleanolic acid	540.7 mg/L	Fuse cytochrome P450 (CYP716A12) to NADPH-P450 reductase	[106]
Protopanaxadiol	300 mg/L	Introduce xylose reductase (XR) and xylitol dehydrogenase (XDH); utilize xylose as sole carbon source	[107]
Linalool #1	7.0 ± 0.3 mg/L	Overexpress genes in the MVA pathway	[109]
Linalool #2	110 mg/L	Disrupt diacylglycerol kinase (DGK1)	[112]
β-Farnesene #1	22 g/L	Apply nitrogen-limited conditions to maximize carbon flux	[113]
β-Farnesene #2	28.9 g/L	Regulate carbon flux and optimize metabolic Pathways	[114]
α-Farnesene #1	57 ± 1 mg/L	Overexpress genes in the MVA pathway	[110]
α-Farnesene #2	25.6 g/L	Overexpress genes in the MVA pathway highlighted as bottlenecks; employ strain selection	[111]
Squalene	51.2 g/L	Increase the supply of acetyl-CoA in peroxisomes and cytoplasm	[115]
Flavonoids
p-Coumaric acid	593.53 ± 28.75 mg/L	Overexpress bottleneck genes; remove competing pathways	[124]
Naringenin #1	252.4 mg/L	Overexpress bottleneck genes; control pH and C/N ratio	[125]
Naringenin #2	715.3 ± 12.8 mg/L	Introduce xylose reductase (XR) and xylitol dehydrogenase (XDH); utilize xylose as a carbon source	[126]
(2S)-Naringenin	8.65 g/L	Employ enzyme engineering, precursor supply enhancement, and multicopy pathway integration	[128]
Scutellarin	346 mg/L	Express multiple copies of biosynthetic genes and use a fed-batch bioreactor	[129]
Organic acids
α-Ketoglutarate	186 g/L	Co-express orthologous biosynthetic genes	[137]
Succinic acid	87.9 g/L	Evolve a natural mutant in vitro; employ batch fermentation with food-waste hydrolysate	[140]
Itaconic acid #1	4.6 g/L	Overexpress precursor pathway; inhibit competing pathway; optimize C/N ratio	[143]
Itaconic acid #2	22 g/L	Overexpress the mitochondrial cis-aconitate transporter MTT	[144]
Crotonic acid	220 ± 8 mg/L	Overexpress orthologous genes to produce precursors	[130]
Others
Pentane	4.98 mg/L	Overexpress lipoxygenase	[146]
Diverse alkanes and alcohols	142.5 mg/L of FAEEs, 23.3 mg/L of fatty alkanes, 2.15 g/L of fatty alcohols, 9.67 g/L of free fatty acids, and 66.4 g/L of triacylglyceride	Activate free fatty acids by conjugation to CoA; overexpress processing enzymes; decouple nitrogen starvation from lipogenesis	[147]
Alkanes	1.47 g/L	Overexpress fatty acid photodecarboxylase; optimize C/N ratio	[148]
Cordycepin	4.36 g/L	Optimize promoters; overexpress genes supporting metabolic precursors; optimize feedstock	[153]
Arbutin	8.6 ± 0.7 g/L	Optimize promoters	[155]
Violacein	366.30 ± 28.99 mg/L	Optimize promoters; overexpress genes in the shikimate, pentose phosphate, and glycolytic pathways	[124]
Triacetic acid lactone	4.76 g/L	Overexpress genes in precursor pathways; utilize glacial acetic acid as a substrate; optimize C/N ratio; inhibit fatty acid synthesis	[36]

). These approaches have delivered impressive yields but fail to meet the requirements for profitability in many cases.

3. Upcoming Yield Improvement Strategies

3.1. Establishment of DNA Modular Assembly Platforms

Synthetic biology approaches have resulted in a step change in tools for engineering organisms and predicting the outcomes of complex genetic manipulations. This is applicable to the bioproduction of complex chemicals [161,162]. This is particularly the case for DNA manipulations, previously a critical bottleneck [163]. Advances initially made for model organisms can often be widely applied to industrial chassis. For Y. lipolytica, provision of genetic toolkits allows more rapid, iterative, and automatable modular cloning options [86,164,165].

Based on the Golden Gate assembly method, Celińska et al. developed a versatile DNA assembly platform for Y. lipolytica (Figure 6) [86]. The DNA module is assembled on a pre-designed scaffold with four nucleotide overhangs, covering three transcription units (each with a promoter, gene, and terminator), a selectable marker gene, and a genome integration targeting sequence, making up a total of 13 genetic elements. This toolkit facilitates modular assembly of expression cassettes for up to three transcription units. A key advantage is that it allows one to rapidly test alternative promoters for each gene and so optimize throughput and limit cellular burden. Larroude et al. subsequently expanded the toolkit to include six different markers (three auxotrophic markers, two antibiotic resistance markers, and one metabolic marker) [165]. The 64 modules (including promoters and terminators) were validated and characterized using three different fluorescent reporter proteins, providing high confidence in the applicability of the toolkit. To test the utility of this approach, Larroude et al. overexpressed a three-gene xylose utilization pathway. The whole process, from cloning to phenotype screening, took less than 10 days, faster than could be achieved with sequential plasmid transformation in Y. lipolytica [165].

Wong et al. engineered a set of modular cloning vectors compatible with BioBrick standards, called YaliBricks, to allow for rapid assembly of multigene pathways with customized genetic control elements (promoters, intronic sequences, and terminators) in Y. lipolytica [164]. This study characterized twelve promoters with different strengths. The YaliBricks vectors exploit compatible restriction enzyme sites to enable modular pathway engineering, as demonstrated by the construction of a five-gene pathway in a single plasmid. These toolkits have been made publicly available through Addgene (www.addgene.org). Integration of these toolkits with automation developed for model organisms will likely facilitate rapid screening of combinations of genes and promoters in Y. lipolytica.

Convenient and easy-to-use CRISPR toolkits have been developed to enable rapid iterative metabolic engineering of Y. lipolytica. Larroude et al. [166] developed a set of CRISPR/Cas9 vectors based on Golden Gate assembly. These vectors contain different selection markers, and the guide RNA can be rapidly inserted and replaced using Golden Gate assembly, allowing for rapid editing of Y. lipolytica. Holkenbrink et al. developed a CRISPR/Cas9 toolkit called EasyCloneYALI that allows efficient genome editing and construct insertion into Y. lipolytica. This toolkit supports marker-free integration of gene expression vectors into identified genomic sites and can achieve marker-free gene deletion using CRISPR/Cas9 technology. When transformation protocols with non-replicating DNA repair fragments (such as DNA oligos) are used, genome-editing efficiency can reach over 80% [167]. These CRISPR/Cas9 toolkits, specifically designed for the Y. lipolytica platform, are undoubtedly very convenient for researchers and industrial modification, significantly improving the efficiency of modifying this strain.

3.2. Acetyl-CoA and NADPH Supply Optimization

Y. lipolytica’s high lipid content is supported by its enhanced production of acetyl-CoA [168]. This is the central precursor for the biosynthesis of both fatty acids and high-value natural products such as terpenes, flavonoids and polyketides. Ensuring there is an adequate supply of acetyl-CoA not only contributes to the accumulation of target metabolites but also reduces the impact on the metabolic network of the microbial host. In Y. lipolytica, the synthesis of cytosolic acetyl-CoA is driven by ATP-citrate lyase (ACL). The acetyl-CoA produced predominantly enters the lipid anabolic pathway.

Strategies for increasing the production of metabolites have focused on adapting Y. lipolytica metabolism to produce more acetyl-CoA. An important innovation is the discovery of a pyruvate bypass pathway that increases accumulation of acetyl-CoA [169]. The applicability of this approach was demonstrated by increasing the titer of the polyketide TAL from 2.1 g/L to 35.9 g/L. In an alternative approach, Y. lipolytica was engineered to direct acetate in the growth media to acetyl-CoA, yielding 4.76 g/L of TAL [36]. Huang et al. used a genome-scale metabolic network analysis to identify “pull”, “push”, and “block” strategies to redirect flux towards acetyl-CoA. These increased the accumulation of acetyl-CoA by 50% and squalene production 16.4 times [170].

The supply of cofactors is an important factor in maintaining cell growth and improving biochemical metabolism efficiency [171]. NADPH is the most important biosynthetic reducing equivalent in cells and the main rate-limiting factor for fatty acid synthesis [172]. Enhancing the expression of corresponding dependent dehydrogenases is a commonly used NADPH enhancement method. Enhancing succinic semialdehyde dehydrogenase, 6-phosphogluconate dehydrogenase, malic enzyme, and mannitol 2-dehydrogenase in Y. lipolytica has enhanced the yields of terpenoids, polyketides, and lipids [36,71,173,174]. An alternative strategy is the conversion of NADH into NADPH through overexpression or introduction of heterologous kinases. For example, Qiao et al. engineered 13 strains of Y. lipolytica with synthetic pathways converting glycolytic NADH into NADPH, thereby increasing the strain’s lipid levels by 25% [38]. These approaches are generally applicable and could be used to improve the efficiency of high-value chemical production in Y. lipolytica.

3.3. Adaptive Laboratory Evolution (ALE)

Adaptive laboratory evolution (ALE) [175] is a set of approaches that leverage natural selection of microorganisms to generate improved or novel phenotypes, usually for bioproduction. ALE methods originated from controlled evolution experiments conducted on bacteria [176]. This set of techniques has become a powerful tool for quickly and efficiently constructing microbial engineering strains and can be readily applied to different microorganisms [177].

The typical ALE strategy consists of cultivating microbial strains under specific environmental conditions to promote the adaptive natural evolution of the strains towards desired characteristics [176]. Target strains evolve through serial batch transfers, and strains with beneficial mutations are identified through screening. This technique is currently well-established and primarily applied in the model species E. coli and S. cerevisiae [178,179]. However, with the growing interest in using non-conventional yeasts as production chassis, ALE is being adapted to a broader range of microbial hosts.

For Y. lipolytica, two key challenges have been addressed by ALE. For metabolic engineering, there is generally a trade-off between product titers and yields, where the oil production rate is inversely proportional to the cell growth rate. ALE has long been considered a complement to traditional metabolic engineering approaches [180]. To address this, the mutagen ethyl methane sulfonate was used to conduct random mutagenesis on wild strains [181]. High-fat mutants were subsequently enriched by serial transfer of floating cells (thereby providing selection at each transfer). Individual colonies were screened on Nile Red plates to detect intracellular fat content and cell growth. After five serial transfers (105 generations), the screened population accumulated up to 44% (w/w) lipids, 30% higher than the starting strain, without compromising overall biomass [181].

ALE has addressed Y. lipolytica’s tolerance to toxic fermentation products. Limonene is toxic to the fungus, reducing limonene titers in fermentation [92]. A short-term ALE strategy demonstrated that limonene-resistant strains can be evolved. Three different ALE strategies were tested (Figure 7). Stepwise adaptation (Methods A–D): These methods focused on gradual escalation of selection pressure. Strains were exposed to increasing concentrations of exogenous limonene across successive passages. Direct adaptation (Methods E and F): These strategies involved sustained exposure to constant inhibitory concentrations of limonene from the onset. Strains were cultivated under a persistent stress of 500 mg/L (Method E) or 1500 mg/L (Method F), aiming to select for mutants with immediate high-level robust growth. Pulse adaptation (Methods G and H): In these regimes, high-intensity, short-duration “shocks” were employed. Strains were intermittently exposed to different concentrations—500 mg/L to 1000 mg/L, 500 mg/L, and 1000 mg/L (Method G) and 500 mg/L to 1000 mg/L, 500 mg/L, and 1500 mg/L (Method H). Methods C, E, and G successfully cultivated strains resistant to high concentrations of limonene and successfully increased the titer by 46%, 52%, and 41%, respectively [182]. Although this ALE strategy successfully achieved the goal of selecting tolerant strains, different screening methods (some with only subtle differences) resulted in different strain results. Choosing an appropriate and effective induction mode is key to successful ALE.

3.4. Regulation and Design of Metabolic Pathways

Classic metabolic engineering strategies include increasing the supply of precursors (push), enhancing the activity of the rate-limiting step (pull), and blocking by-product synthesis (block). These can be used to increase the yield of the target product [183]. The push–pull–block strategy is widely used in metabolic engineering research on unconventional yeasts. For instance, enhancing malonyl-CoA synthesis can effectively promote the synthesis of fatty acids, polyketides, and other products. The supply of malonyl-CoA can be enhanced by overexpressing acetyl-CoA carboxylase (Acc), which has significantly increased product yields in a variety of unconventional yeasts [36,184,185]. Lipid synthesis can be promoted through attenuation of the β-oxidation pathway, which degrades fatty acids. Deletion of triacylglycerol lipase or peroxisomal biogenesis factors has been reported to inhibit degradation and increase the accumulation of lipids [186,187]. For terpenoid synthesis, for which there are many bypass pathways, the block strategy can enhance main-pathway flux by weakening metabolic bypass pathways. For example, β-farnesene yields were increased over 40-fold by deleting the β-carotene synthesis pathway [188]. Conversely, β-carotene yields were improved by 75% by downregulating the competing squalene synthase [189].

Enzyme colocalization is a complementary metabolic engineering strategy for controlling flux and improving pathway efficiency for product synthesis [190]. Common colocalization strategies include enzyme fusion, scaffolding, organelle targeting, and use of synthetic organelles. Enzyme fusion and organelle targeting are most widely used strategies applied to Y. lipolytica. Enzyme fusion engineering is regarded as an efficient strategy for enhancing metabolic flux and increasing target-product titers [191]. For example, fusion of p-coumaryl-CoA ligase and resveratrol synthase in Y. lipolytica significantly increased the conversion of p-coumaric acid to resveratrol by more than 35% [192]. The organelle-targeting strategy can alleviate the inhibition of cell growth caused by product accumulation and exploit more abundant precursor pools. Peroxisomes supply acetyl-CoA by catalyzing the β-oxidation of fatty acids, providing a ready pool of this precursor. This has been exploited for the synthesis of natural products such as carotenoids and farnesene precursors [193,194,195]. Production of the carotenoid astaxanthin was increased approximately six-fold by fusing two biosynthetic enzymes (β-carotene ketolase and hydroxylase) and directing them to the surfaces of liposomes, endoplasmic reticulum, and peroxisomes [195]. Similarly, Yang et al. directed lipase-dependent pathways to the lipid bodies (LB). This resulted in a 10-fold-higher fatty acid methyl ester titer compared to cytosolically expressed enzymes [196]. Notably, this was the first time that the LB targeting pathway had been successfully explored in Y. lipolytica. There remains significant potential for further development of the organelle-targeting strategy in Y. lipolytica.

Yeast organelles provide distinct physicochemical niches. Enzymes, metabolites, and cofactors are distributed heterogeneously between organelles and the cytoplasm, providing specialized environments tailored to diverse metabolic requirements [197]. Spatial compartmentalization of metabolic pathways within these subcellular compartments has emerged as a focal point in yeast metabolic engineering [198]. However, most investigations on yeast metabolic pathway compartmentalization have been conducted using the model organism S. cerevisiae. Key studies have used mitochondria and peroxiosomes to produce fuel, chemicals, and pharmaceuticals [199,200,201]. Despite these advantages, organelle engineering’s full potential in Y. lipolytica remains largely untapped. The expansive intracellular lipid droplets in this yeast provide an ideal venue for pathway compartmentalization [40]. Leveraging these hydrophobic reservoirs enables the spatial organization of biosynthetic enzymes, optimizing the production of lipid-soluble metabolites. Importantly, alterations in cell morphology frequently remodel the abundance, spatial organization, and intracellular trafficking dynamics of organelles. Consequently, the integration of organelle-targeting strategies with morphological engineering (Section 3.6) may represent a promising avenue for coordinating intracellular metabolic pathways with cellular architecture, ultimately enhancing the biosynthetic performance of Y. lipolytica.

Machine learning and statistical methods offer a complementary approach to targeted optimization of metabolic pathways. Machine learning methods require substantial datasets. Although some studies have provided interesting insights into S. cerevisiae [202], the data available for Y. lipolytica currently limit the insight gained [203]. Statistical methods such as design of experiments [204] provide insight into optimal production conditions. Improved production of extracellular enzymes [205], gamma-decalactone [206], and PUFAs [207] has been demonstrated. In the longer term, these data may lead to more powerful, integrated machine learning approaches.

3.5. Substrate Selection and Optimization

Substrate choice directly affects strain growth and target production [208], and this accounts for a significant proportion of the input cost [209], making it particularly important to select the appropriate substrate based on the target product. The commonly used carbon sources for yeast include glucose and glycerol [210]. Each yeast has specific advantages in terms of utilizing carbon sources based on its own metabolic characteristics. As an important precursor for the synthesis of natural products such as polyketides, flavonoids, terpenes, and fatty acids, the cytoplasmic acetyl-CoA supply has been the focus of research on the optimization of metabolic pathways. Choosing a precursor carbon source that can quickly accumulate acetyl-CoA in the cytoplasm facilitates synthesis of these products. For example, as acetate can be converted into acetyl-CoA in one step, it has been used in the production of triacetic acid lactone and lipids in Y. lipolytica [36,53]. Some carbon sources can indirectly increase yields by negating the inhibitory effect of intermediates or products on cell growth. For example, replacing glucose with ethanol as the fermentation substrate for producing baicalein alleviated the toxic effects of the intermediate, cinnamic acid, on cells [211].

Many recent studies have explored valorizing industrial, agricultural, and forestry waste as carbon sources to synthesize high-value products [212,213]. The industrial waste streams explored as feedstocks for Y. lipolytica fermentation include waste cooking oils [214], olive mill wastewater [215], and hydrocarbon-rich effluents [216]. Use of waste streams from other bio-based production contributes to the concept of a circular bioeconomy [217]. Most of the products discussed here are priority products for the circular bioeconomy (the food, pharmaceutical/nutraceutical, and chemical industries [218]). The culture conditions required for the growth of Y. lipolytica are straightforward. This facilitates integration into industrial symbiosis sites [219], as fewer additional inputs are required. Y. lipolytica has strong environmental adaptability and can utilize inexpensive raw materials such as waste industrial oils as substrates. The carbon source utilization profile of Y. lipolytica can be further expanded through metabolic engineering. Xylose, glycerol, and cellulose have been used as carbon sources for Y. lipolytica, demonstrating strong potential for applications in the production of oils and other high-value products [220,221,222,223]. A strong example is the use of crude glycerol as a feedstock [220]. Crude glycerol is a by-product of many industries. It contains multiple impurities and must be pre-treated or fortified for use with many other species [224,225,226]. Engineered Y. lipolytica degraded cellulose (the major component of lignocellulosic biomass [227]) for synthesis of the food additive p-coumaric acid [223]. Utilization of waste substrates can significantly reduce fermentation feedstock costs, which constitute a major fraction of total bioprocess expenditure [228]. Techno-economic and environmental analysis has shown that Y. lipolytica-based valorization of municipal solid waste for succinate production is sustainable [229]. Techno-economic and life-cycle analyses are essential when considering valorization in the circular bioeconomy to ensure the long-term sustainability of approaches [230,231].

3.6. Morphological Engineering

The concept of morphological engineering was first proposed by Jiang et al. in 2016 [47]. Conceptually, this is the use of genetic engineering to modulate and control the morphology of microorganisms, thereby optimizing microbial fermentation. This approach has been widely used in bacterial fermentation. Current regulatory strategies include increasing cell volume, improving cell permeability, and increasing cell growth rates [232]. Down-regulating cell wall biosynthesis enzymes using CRISPRi in E. coli changes cells from being rod-shaped to irregularly shaped [233]. This increases cell volume and allows greater accumulation of products in cells [233]. Conversely, overexpression of the cell division protein FtsZ in E. coli increases the frequency of cell division and produces many minicells [234]. These minicells help achieve high-density fermentation of engineered bacteria, wherein cell density is the driving factor in product biosynthesis.

The developments in the morphological engineering of bacteria (especially E. coli and B. subtilis) provide strong conceptual and technical support for building efficient yeast cell factories for industrial bio-based chemicals. This review proposes that morphological engineering technology is a promising avenue of future Y. lipolytica research. This fungus exhibits a typical dimorphism under different environmental conditions [19]. In industrial fermentation processes, stress conditions such as temperature fluctuations, nutrient deficiencies, or changes in osmotic pressure often occur. These environmental alterations can easily induce morphological changes in Y. lipolytica. Morphological switching in Y. lipolytica is associated with wide-ranging changes in gene expression and metabolomic profile. This can significantly affect the titers of fermentation products (Figure 8) [235]. For Y. lipolytica, the morphological transformation seems not to be limited to cell wall proteins; it also modulates the cell cycle and metabolic intensity, amongst other things [21]. For example, when fermentation pH is decreased from 7 to 5, Y. lipolytica mainly adopts the yeast form, and the formation of hyphae is inhibited. The rate at which Y. lipolytica produced citric acid increased three-fold under these conditions [236]. This suggests that the yield of target products could be increased by regulating cell morphological changes. However, there remain many important questions about how morphological engineering can be applied to Y. lipolytica.

The dimorphic transition in Y. lipolytica has consequences across five interconnected cellular physiology elements (secretion, nutrient uptake, bioreactor performance, cell-cycle regulation, and metabolic intensity). In many fungi, extracellular protein secretion is spatially organized and frequently associated with regions of polarized growth, where vesicle trafficking and exocytosis are concentrated [237,238]. The secretory pathway of Y. lipolytica is characterized by significantly higher complexity than canonical yeast. This places it closer to the filamentous fungi-like systems [239]. Protein secretion pathways are closely linked to morphological development in Y. lipolytica. Mutations in the genes SEC238, SRP54, PEX1, PEX2, PEX6, and PEX9 affect protein secretion, prevent the exit of the precursor form of alkaline extracellular protease from the endoplasmic reticulum, and compromise peroxisome biogenesis. With the exception of PEX1, these mutants are also deficient in the dimorphic transition from the yeast to mycelial form. Only export of plasma membrane and cell-wall-associated proteins specific for the mycelial form are affected, highlighting the essential role of secretory trafficking in fungal morphogenesis [240]. By characterizing the gene SEC14pYL in Y. lipolytica, Lopez et al. established a mechanistic link between the secretory pathway and morphological plasticity. As a phosphatidylinositol transfer protein involved in Golgi-associated lipid trafficking, SEC14pYL is non-essential for viability or general secretion. Its deletion specifically negates the yeast-to-hyphal transition. This distinct phenotype suggests that Golgi-driven lipid transport serves as a regulatory hub for cellular differentiation rather than merely acting as a conduit for bulk protein secretion [241]. These morphological engineering strategies not only change cell shape but also reshape the secretory patterns of cells. For biotechnological applications that target secreted enzymes or metabolites, this distinction is directly related to yield and should guide the selection of morphological engineering targets.

The morphological state of a cell exerts a direct physical effect on how efficiently the cell can access and intake nutrient substrates from the environment. This connection is most clearly articulated for hydrophobic substrates. Kappeli et al. [242] observed hairy protrusions on the cell surface of hexadecane-grown C. tropicalis. While lipid bodies were observed on the cell surface of Y. lipolytica strain H222 grown on methyl ricinoleate, no such protrusions could be observed on the cell surface of odc− mutant strains grown on hexadecane [35]. The hyphal state facilitates spatial foraging and nutrient sequestration in solid substrates through polarized extension. In contrast, the yeast state maintains a superior surface-area-to-volume (SA/V) ratio [243] in homogenous liquid environments due to its smaller cell size. This high SA/V ratio significantly alleviates diffusion resistance during transmembrane transport, fueling rapid metabolic turnover and exponential proliferation under high nutrient concentrations [244]. These findings demonstrate that distinct cellular geometries significantly influence substrate utilization efficiency.

Cell morphology can significantly influence bioreactor performance by affecting the rheology and oxygen transfer of the bioreactor [245,246]. Filamentous growth often leads to cell aggregation or the formation of mycelial balls or hyphal networks, thereby increasing broth viscosity and reducing mixing efficiency [246,247,248]. Y. lipolytica cell morphology has been shown to affect the rheological properties of fermentation broth and interact with dissolved oxygen levels [249]. Therefore, maintaining the dispersed morphology of yeast is generally advantageous in industrial fermentation because it can improve oxygen transfer and reduce mass-transfer limitations.

The morphological state is intimately connected with cell-cycle regulation. In dimorphic fungi, the transition between the yeast and filamentous states is frequently accompanied by alterations in the regulation of cell-cycle progression and the coordination between growth and division [250]. In Y. lipolytica, forward genetic screening and whole-genome sequencing were used to identify mutants that reliably remain in the yeast phase [251]. All the “smooth” mutants exhibited independent loss of DNA at a repetitive locus made up of interspersed ribosomal DNA and short 10- to 40-mer telomere-like repeats. The loss of this repetitive DNA is associated with downregulation of genes with stress response elements (5′-CCCCT-3′) and upregulation of genes with cell-cycle box (5′-ACGCG-3′) motifs in their promoter regions. The cell-cycle box is bound by the Mbp1p/Swi6p complex in S. cerevisiae to regulate G1-to-S-phase progression. Overexpression of either Ylmbp1 or Ylswi6 decreases hyphal growth, whilst deletion of either of these genes promotes hyphal growth in smooth strains [251]. In parallel, genes such as YlBEM1 [252], Yltec1 [253], and Ylznc1 [254] have been identified as major transcriptional factors controlling morphological transition in Y. lipolytica, with their activity being closely connected to cell cycle progression.

Morphology modulates yield through metabolic intensity—specifically via the partitioning of central carbon flux under nutrient regimes optimized for secondary metabolite biosynthesis [255]. In microbial cell factories, the distribution and intensity of metabolic fluxes through central carbon metabolism pathways largely determine the availability of substrates, reducing equivalents, and energy required for biosynthetic pathways [256]. Alterations in cell physiology can significantly alter carbon flux, shifting it towards biomass formation or product biosynthesis. This affects overall production efficiency [255,257]. The morphological transition of dimorphic yeast alters metabolic priorities and resource allocation. The hyphal state is inherently resource-intensive, requiring sustained polarized extension, continuous vesicular flux, and localized cell wall deposition at the apex [258,259]. This developmental program imposes a significant metabolic burden, as substantial pools of carbon and ATP are siphoned into cytoskeletal remodeling and membrane turnover. This is often at the expense of secondary-metabolite biosynthesis [260,261]. Conversely, the more streamlined, symmetric growth of the yeast state allows for a more efficient allocation of cellular precursors toward central metabolism and targeted biosynthetic pathways. Collectively, these observations highlight the pleiotropic effects of morphology on the suitability of a yeast as a cellular factory. Morphological engineering exploits the cellular adaptations and physical architecture of the yeast and hyphal forms.

Y. lipolytica’s switch to the filamentous state is known to be promoted by the mitogen-activated protein kinase (MAPK) pathway and pH response pathways [262,263]. Conversely, the PKA and high-osmolarity glycerol response pathways repress the filamentous state. Transcription factors are known to be important, with Mhy1 (particularly) and Hoy1 promoting the filamentous morphology [264], whilst Tec1, Zcn1, Fts1-2, and the co-repressor Tup1/ssn6 complex negatively regulate the switch to a filamentous state [265]. An important Y. lipolytica signaling protein is Cla4, which activates the MAPK pathway but not the PKA pathway, thus promoting the filamentous state [266,267,268,269].

Liu et al. successfully inhibited hyphal formation of Y. lipolytica by knocking out CLA4 and MHY1. The CLA4/MHY1 mutant successfully increased the yield of the target product β-carotene by 139%, and the engineered strain produced 7.6 g/L and 159 mg/g DCW of β-carotene [40]. In addition, with in-depth informatics analysis of the morphological transformation process of Y. lipolytica, it can be expected that more signaling pathways or genes that affect morphological transformation will be elucidated [13,253,263,270,271,272]. Histone acetylation and deacetylation are important mechanisms of chromatin remodeling, regulating cell differentiation processes such as dimorphism. These reactions are catalyzed by histone acetyltransferases (HATs) and histone deacetylases (HDACs), respectively. HDACs constitute an evolutionarily conserved superfamily found across diverse biological kingdoms [273,274]. In fungi, HDACs have been shown to be crucial for morphology, growth, sporulation, degradation of complex polysaccharides, sensitivity to antimicrobial compounds, sensitivity to different types of stress, secondary metabolism, dimorphism, and virulence [275,276,277,278,279,280]. Two HDAC inhibitors, sodium butyrate (SB) and valproic acid (VPA), have been shown to strongly inhibit hyphal formation. This suggests that histone acetylation may play a key role in the regulation of morphological transformation [281]. These findings will provide a valuable reference for the further application of morphological engineering strategies for Y. lipolytica. Adding Y. lipolytica morphological engineering strategies to the toolkit of strategies for bioproduction for this yeast should help further optimize the production of target chemicals.

4. Conclusions and Prospects

This review summarizes the approaches that are currently being exploited to synthesize high-value organic compounds using Y. lipolytica. Key approaches to optimizing production and fermentation yields are promoter optimization or including copies of orthologous biosynthetic genes, the use of metabolic engineering to increase flux and provide precursors, and optimization of the carbon source and C/N ratio. Emerging methods include the use of toolkits for promoter optimization and ALE strategies. We further propose using morphological engineering as a new regulatory strategy for product optimization of Y. lipolytica with morphological transformation characteristics. Metabolic engineering primarily focuses on redirecting intracellular carbon flux toward target biosynthetic pathways. In contrast, morphological engineering regulates cell structure and population morphology, which can influence oxygen transfer, broth rheology, nutrient uptake, and overall fermentation efficiency. Morphological engineering can complement metabolic engineering and bioprocess optimization by simultaneously improving intracellular metabolic performance and extracellular fermentation conditions. With further in-depth transcriptomic analysis and informatics analysis of the morphological transformation process in Y. lipolytica, more target genes or signaling pathways related to morphological transformation will be discovered. Systematic multi-omics analyses should be conducted to identify key regulators controlling the yeast–hypha transition. This will allow targeted genome editing and promoter engineering to achieve stable and controllable morphological phenotypes. Integrating morphological engineering with other strain improvement strategies—such as ALE, dynamic gene regulation systems, and computational strain design—may enable the development of robust strains capable of maintaining optimal metabolic activity and morphology under industrial fermentation conditions. Future developments in the use of Y. lipolytica are likely to come from combining these approaches and developing methods for efficiently searching for the best solution. Furthermore, future studies should also address the economic viability and scale-up challenges associated with industrial bioproduction. The integration of multiple strategies will be essential for improving fermentation stability, reducing downstream processing costs, and achieving commercially competitive production levels.