Cellular and metabolic engineering of oleaginous yeast Yarrowia lipolytica for bioconversion of hydrophobic substrates into high‐value products

Abstract

The non‐conventional oleaginous yeast Yarrowia lipolytica is able to utilize both hydrophilic and hydrophobic carbon sources as substrates and convert them into value‐added bioproducts such as organic acids, extracellular proteins, wax esters, long‐chain diacids, fatty acid ethyl esters, carotenoids and omega‐3 fatty acids. Metabolic pathway analysis and previous research results show that hydrophobic substrates are potentially more preferred by Y. lipolytica than hydrophilic substrates to make high‐value products at higher productivity, titer, rate, and yield. Hence, Y. lipolytica is becoming an efficient and promising biomanufacturing platform due to its capabilities in biosynthesis of extracellular lipases and directly converting the extracellular triacylglycerol oils and fats into high‐value products. It is believed that the cell size and morphology of the Y. lipolytica is related to the cell growth, nutrient uptake, and product formation. Dimorphic Y. lipolytica demonstrates the yeast‐to‐hypha transition in response to the extracellular environments and genetic background. Yeast‐to‐hyphal transition regulating genes, such as YlBEM1, YlMHY1 and YlZNC1 and so forth, have been identified to involve as major transcriptional factors that control morphology transition in Y. lipolytica. The connection of the cell polarization including cell cycle and the dimorphic transition with the cell size and morphology in Y. lipolytica adapting to new growth are reviewed and discussed. This review also summarizes the general and advanced genetic tools that are used to build a Y. lipolytica biomanufacturing platform.

Abbreviations

EPA

eicosapentaenoic acid

FAEEs

Fatty acid ethyl esters

TAG

Triacylglycerol

Y. lipolytica

Yarrowia lipolytica

YH transition

yeast‐to‐hypha transition

1. AN OVERVIEW OF THE NON‐CONVENTIONAL YEAST YARROWIA LIPOLYTICA

The non‐conventional dimorphic oleaginous yeast Yarrowia lipolytica has been developed as a promising host for a diverse range of biotechnological and pharmaceutical applications. The species name “lipolytica” originates from remarkable lipolytic activity in the hydrolysis of lipid 1, 2. Y. lipolytica has been classified as Generally Regarded As Safe (GRAS) by the American Food and Drug Administration (FDA) 3 and is considered as non‐pathogenic yeast due to the maximum growth temperature is below 32°C that is lower than the human average internal temperature. As eukaryote, Y. lipolytica demonstrates not only lower degree of glycosylation in post‐translation modification as compared to the conventional yeast Saccharomyces cerevisiae, but also the high secretion capacity for specific enzymes, lipase, protease and phosphatase. Y. lipolytica also belongs to oleaginous yeast and displays the high tolerance of endogenous lipid content owing to the greater lipid storage in lipid body as triacylglycerols (TAGs) and steryl esters. Both hydrophilic substrates, including glucose, fructose, mannose, galactose, glycerol and ethanol, and hydrophobic substrates, such as free fatty acids, plant oils, and animal fats as low‐cost carbon sources can be utilized by Y. lipolytica and then converted to value‐added bioproducts, which include extracellular enzymes 4, 5, 6, organic acids such as citric/isocitric acids, pyruvate, succinate and alpha‐ketoglutarate 7, 8, 9, 10, 11, heterologous proteins such as laccase and epoxide hydrolyase 12, 13, fatty acids and alkane derived products 4, 14, 15, 16, 17, food and feed supplements such as carotenoid 18 and pharmaceuticals such as omega‐3 fatty acids 19, 20, 21, 22. Due to its great potentials of making various high‐value products, Y. lipolytica has been metabolically engineered to establish an efficient expression platform for microbial biomanufacturing in last two decades.

Considerable efforts have been made and significant progresses have been achieved for producing wide ranges of homologous and heterologous metabolites via the modulation of the growth conditions and engineering the metabolic pathways in Y. lipolytica, but very little attention has been paid to other unique features of the nonconventional yeast. For example, Y. lipolytica exhibits yeast to hyphae morphology change when grows under nutrient limiting or other stressful conditions, which may further relate to the efficiency of the substrate utilization and product formation. This review first summarizes several important high‐value products produced from the metabolic engineered Y. lipolytica. Then, the general genetic engineering strategies in this yeast are discussed. We also include the current understandings of the cell morphology and the key genes controlling the dimorphic feature in Y. lipolytica. The potential benefit of the cell morphology engineering in bioconversion of the hydrophobic substrates will be also highlighted.

2. Y. LIPOLYTICA AS THE BIOMANUFACTURING PLATFORM

To meet the increasing market demands of some important products in both food and pharmaceutical applications, Y. lipolytica is becoming a desirable bioproduction host to boost the current capabilities as compared to the engineered bacteria and mammalian cells. Analysis on the metabolic pathways in Y. lipolytica reveals that the yeast is capable of producing a series of value‐added bioproducts via homologous and heterologous pathway engineering. These products include organic acids (e.g. citric/isocitric, α‐ketoglutaric, succinic, fumaric, malic and pyruvic acids), extracellular proteins (e.g. lipase, esterase, alkaline protease and phosphatase), carotenoids, wax esters, fatty acid ethyl esters (FAEEs), omega‐3 fatty acids, and many other products derived from the metabolic engineering pathways (Figure 1).

Figure 1.

An overview of metabolic engineering of Y. lipolytica for biomanufacturing of citric acid, wax esters, long‐chain diacids, carotenoids, and omega‐3 fatty acids. Abbreviations: PEP, phosphoenolpyruvate; FK, fructose kinase; HK, hexokinase; α‐KG, alpha‐ketoglutarate; OAA, oxaloacetate; GA3P, glyceraldehyde3‐phosphate; GUT1, glycerol‐kinase 1; GLUT2, glucose transporter; ME1, malic enzyme 1; TGLs, TAG lipases; PYC, pyruvate carboxylase; TAG, triacylglycerol; ACL, ATP‐citrate lyase; ACC1, malonyl‐CoA by carboxylase; GPD1, glycerol‐3‐phosphate dehydrogenase; Gxf1, glucose/xylose facilitator; Sut1, sucrose transporter; FAA1, fatty acyl‐CoA synthetase; ACS1, cytosolic acyl‐CoA synthase; THIOs, Acyl‐CoA thioesterases;  TesA, thioesterase I; TCA, tricarboxylic acid cycle; FAS, fatty acid synthesis. Green frame: Fatty acid catabolism in β‐oxidation pathway in peroxisome. Abbreviations: Pox1 to Pox6, acyl‐CoA oxidases 1–6, respectively; MFE2, peroxisomal multi‐functional enzyme; Pex3 and Pex10, peroxisome biogenesis factor 3 and 10, respectively. Yellow box: Metabolic engineering pathway of wax ester. Abbreviations: FAR, fatty acyl CoA reductase; WES, wax ester synthase. Blue box: Pentose phosphate pathway. Abbreviations: 6PGD, 6‐Phosphogluconate dehydrogenase; 6PGL, 6‐Phosphogluconolactonase; GA3P, glyceraldehyde3‐phosphate. Green box: ω‐oxidation pathway. Abbreviations: CYP450, cytochromes P450 enzyme; FAO, fatty alcohol oxidase; ADH, fatty‐alcohol dehydrogenase; FALDH, fatty aldehyde dehydrogenase. Gray and blue box within purple frame: Metabolic engineering of aerobic pathways for ω‐3 and ω‐6 FA biosynthesis in endoplasmic reticulum. Abbreviations: C16E, D9E and C20E are C16/C18, Δ‐9 and C20/C22 elongases, respectively. D4D, D5D, D8D, D9D, D12D, D15D and D17D are Δ‐4, Δ‐5, Δ‐8, Δ‐9, Δ‐12, Δ‐15 and Δ‐17 desaturases, respectively. Gray box: Part of the cytosolic fatty acid synthesis pathway. Abbreviations: LPA, lysophosphatidic acid; PA, phosphatidic acid; DAG, diacylglycerol; DGA1 and DGA2, DAG acyltransferase; PDAT, PL and DAG acyltransferase. Purple box: Metabolic engineering of carotenoid biosynthesis pathway. Abbreviations: DMAPP, dimethylallyl diphosphate, IPP, isopentenyl diphosphate; GPP, geranyl pyrophosphate; FPP, farnesyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; HMG, 3‐hydroxy‐3‐methyl‐glutaryl reductase; CrtE and GGS1, GGPP synthase; CrtB, phytoene synthase; CrtI and CarB, phytoene dehydrogenase; CrtY, lycopene cyclase; CarRP, bifunctional enzyme phytoene synthase/lycopene cyclase; CrtW, β‐carotene ketolase; CrtZ, β‐carotene hydroxylase. Blue ovals: transporters. Dash lines: Putative route, not confirmed

2.1. Citric acid

Organic acid secretion, including the precursors and intermediates of TCA cycle, is one of the characteristics for Y. lipolytica. Among the organic acids, citrate plays a prominent role throughout the yeast energy metabolism that acts as prime carbon source for fatty acid synthesis and participates in the regulation of the glucose metabolism 23. Citric acid is one of the fermentation products with the largest market growth in the world, exceeding two million tons per year, due to its wide applications in flavoring additive and antioxidant in the beverage and food, chemical and pharmaceutical industries. Typically, citric acid production is highly affected by the cultivation conditions, especially the use of the carbon sources and the oxygen availability 24, 25, 26. In a recent report, the titer of 55 g/L citrate was achieved by Y. lipolytica ACA DC 50109 with glucose in mono‐substrate cultivation under the control of the dissolved oxygen concentration 24. The increase of citrate production occurred when the higher fluxes toward peptone phosphate pathway were observed upon oxygen controlled condition 27. Furthermore, optimization of the carbon sources and cultivation conditions are correlated to the level of citric acid production. Current report revealed that high citric acid production (100–140 g/L) was obtained by feeding rapeseed oil, ethanol and raw glycerol in the fermentation with the Y. lipolytica NG40/UV5 strain 28.

2.2. Lipases

Lipases as biocatalysts have occupied a prominent position in the industrial sector that contribute to various application fields such as detergents, cosmetics, textile, biodegradable polymers, oleochemicals, food additive, pharmaceuticals, biodiesel applications, and waste treatments 29, 30. The wild type Y. lipolytica is capable of expressing various extracellular and intracellular lipases 31, 32, 33, 34. An early study showed that a production of 34.6 ± 0.1 U/mL of lipase was achieved by Y. lipolytica DSM 3286 when grown on the olive oil supplemented with yeast extract 35. Citric acid was still the major byproduct during the lipase production. Moreover, the coproduction of erythritol (22.1 g/L) and lipase (12.7 U/mL) by Y. lipolytica M53 strain were achieved by feeding with low‐cost waste cooking oil and maintaining its concentrations at about 30 g/L 36. In addition, keeping high osmotic pressure was found important for lipase production, e.g. 80 g/L sodium chloride was added in the medium 37. Currently, the increased copy number of the optimized mature extracellular lipase genes (YlLIP2), the coproduction of lipase (16420 U/mL) and 151.2 g/L of single cell protein (SCP) were achieved in Y. lipolytica YLY5 strain, which used the cost‐effective agro‐industrial waste sugarcane molasses‐based medium at 10‐liter fermentation 38. Hence, two major issues of feed industry, poor lipase utilization and protein source shortage, could be solved by simultaneous production of extracellular lipases and SCP in this engineered strain. Therefore, the application of oleaginous yeast Y. lipolytica as sustainable lipase‐producing platform provides many advantages in biomanufacturing.

2.3. Long‐chain dicarboxylic acids (LCDA)

Due to the great market need of general performance nylons and rapid growth of three‐dimensional printing materials, there is an increasing demand of the long‐chain dicarboxylic acid (LCDA‐16 and 18), which can be used not only for making the nylons, but also for applications in other medical and coating industries. In fact, most of the short chain dicarboxylic acid (DCA) synthesis could be obtained via chemical process under the oxidative stress in the preceding 20 years 39. However, the production costs rise with the increase of carbon chain length of the DCA because the coproduction of the additional byproducts through chemical synthesis results in the difficulties of impurity isolation during the purification process 40, 41.

To achieve an efficient synthesis of LCDAs in a cost‐effective manner, scientists have been inspired to genetically engineer bacteria or yeast to produce this special chemical from alkanes or fatty acids. Although the Candida tropicalis, as one of the oleaginous yeast, has been proved to demonstrate the most significantly production of the DCA (130–140 g/L) in the means of the combining metabolic and fermentation engineering 41, 42, the characteristic of the pathogenicity contributing to several hydrolytic enzyme secretion and in the favor of the non‐renewable fossil oil‐derived alkanes as carbon source are two main concerns of this strain 43, 44. Hence, development of engineered Y. lipolytica has gained more attention due to the low pathogenic grade (BSL1) and the easy assimilation of renewable plant oil to LCDA 41, 45, 46, 47, 48.

In previous reports, the Y. lipolytica was engineered by (i) β‐oxidation impairment to increase the endogenous fatty acid accumulation, (ii) upregulating the fatty acid ω‐oxidation pathway by overexpression of P450 reductase and fatty alcohol oxidase YlFAO1, and P450 oxygenase to enhance the terminal oxidation of the intracellular fatty acid precursors. The generated strain produced 330 mg/L LCDA monomer in shaking flask experiments 47, 49. Higher titer of 3.49 g/L LCDA was achieved when the process was scaled up to one‐liter bioreactor, in which nitrogen was limited and glycerol was fed as the sole carbon source 50. This engineered strain also produced the citric acid as major byproduct with a titer of 39.2 g/L 47. Overall, this de novo synthesis of high value LCDA in non‐pathogenic yeast Y. lipolytica has only been demonstrated the feasibility, but more work in both metabolic engineering and fermentation engineering need to be done to meet the requirements for potential commercialization 51.

2.4. Wax esters

Wax esters are a type of highly hydrophobic neutral lipids. It is formed by combining one fatty acid and one fatty alcohol and releasing one molecule of water. Wax esters are common in certain bacteria and some higher evolution level plants and animals to provide a protective coating on the surface so that they are resistant to dehydration, UV light and pathogens 52. Wax esters are used commercially to serve as a variety of applications, such as cosmetics, printing inks, lubricant, coatings, pharmaceutical and the food industries. Despite the fact that the wax esters are founded in nature universally, the abundance source of wax esters are still low because only few organisms such as Jojoba plant (Simmondsia chinensis) and Sperm whale (Physeter macrocephalus) are capable of accumulating large quantities of intracellular wax esters. Currently, wax esters are in a short supply due to the hunting ban for sperm whale and the high extraction cost and harsh requirement of agriculture system for Jojoba. The oleaginous Y. lipolytica is considered as an ideal host for further engineering for large‐scale wax ester production due to the yeast's capability of high TAG accumulation. Introduction of heterologous metabolic pathway of wax ester into Y. lipolytica is comparatively simple. Only two critical enzymes, fatty acyl coenzyme A (FAR) from Marinobacter hydrocarbonoclasticus strain VT8 and wax ester synthase/acyl‐CoA–diacylglycerol acyltransferase (WS/DGAT) from Acinetobacter calcoaceticus strain ADP1 were required for wax ester biosynthesis 52, 53. Multiple copies of these two enzymes introducing into Y. lipolytica successfully led to production of 570 mg/L wax esters in shake flask experiments 54.

2.5. Carotenoids

Carotenoids have antioxidant properties that can be used as drugs against cancer and age‐related degeneration 55. Typical carotenoids via biosynthesis include lycopene, β‐carotene, canthaxanthin, and astaxanthin 55. The current sources of carotenoids are mainly from the extracted products of natural organisms or from chemical synthesis 21. However, nature sources of carotenoid could not meet the market needs due to the small amount of the carotenoid within red‐colored fruits, high extraction cost and long growth period of plants. Moreover, chemical synthetic carotenoid is not approved for human consumption because of the variations in product quality, which produce the toxic and carcinogenic intermediates during the chemical process 56.

In the past decades, Escherichia coli, S. cerevisiae, and Y. lipolytica have been engineered to produce carotenoids to meet the increased market need 18, 22, 55, 57, 58, 59. Among these microbial hosts, Y. lipolytica has attracted more and more attention for carotenoid production due to its ability to generate high amounts of the acetyl‐CoA and other precursors in the mevalonate pathway 60. Biosynthesis of carotenoid in Y. lipolytica requires to (i) promote the high carbon fluxes into mevalonate pathway, i.e. disruption of β‐oxidation or integration multiple‐copy of the rate‐limiting genes such as HMG reductase gene (HMG1) and geranylgeranyl diphosphate synthase gene (GGS1), and (ii) introduce the carotenoid pathway, which engineering converts from two C5 precursors (e.g. IPP and DMAPP), through C10‐C20 intermediates to carotenoid‐related metabolites (C40).

Previous study has demonstrated the heterologous biosynthesis of β‐carotene in Y. lipolytica by introducing the intrinsic high carbon flux of Acetyl‐CoA into the pathway via 12 steps after manipulating eleven homogenous and heterogeneous genes 61. High levels of β‐carotene (4 g/L) stored in lipid droplets within engineered Y. lipolytica was achieved under optimized fed‐batch fermentation conditions 18. To further improve the gene expression levels, promoter shuffling strategy via the Golden gate approach was explored to find the best assembly of the promoters and maximize the β‐carotene production 53. The pTEF promoter demonstrated the strongest expression of the downstream target gene as compared to the pGPM and pGAPDH promoters 60, 62. Additionally, the engineered Y. lipolytica strain with an extra three heterologous β‐carotene expression cassettes under the control of constitutive pTEF promoter, with each cassette containing the native geranylgeranyl diphosphate synthase (GGS1) gene and two pytoene synthase/lycopene cyclase (CarRP) and phytoene dehydrogenase (CarB) genes from Mucor circinelloides, produced 6.5 g/L β‐carotene in the fed‐batch fermentation 60, 63. Moreover, the coproduction of the omega3‐fatty acids and carotenoids by engineered Y. lipolytica was developed via the integration of the heterologous codon‐optimized carotenoid synthesis genes (i.e. CrtE, CerB, CrtI and CrtY) into the omega‐3 fatty acid‐producing strain. Total 8.9 mg/g carotenoid in biomass (nearly 6.4% belonging to β‐carotene) was obtained by this engineered strain 64. Interestingly, the further conversion of β‐carotene to astaxanthin requires to introduce two enzymes, β‐carotene ketolase (CrtW from Paracoccus sp. N81106 strain) and hydrolases (CrtZ from Pantoea ananatis), into β‐carotene‐producing strain. Production of 10.4 mg/L astaxanthin with 5.7 mg/L canthaxanthin could be achieved. Furthermore, through the optimization the multiple copy number of the CrtW and CrtZ into the genome, the higher astaxanthin production (54.6 mg/L) was further improved 56.

2.6. Fatty acid ethyl esters (FAEEs)

Biofuel production is another growing interest from the yeast due to the limitation of the fossil fuels. While most previous biofuel researches were focused on production of bio‐ethanol and bio‐butanol, now biodiesel has drawn more and more attention. Fatty acid ethyl esters (FAEEs), one of the general formats of biodiesel, are seen as promising biofuel via development of engineered microorganisms 22, 65. Due to the large quantity of the precursor lipids, such as plant‐based oils and animal fats requirement in the traditional biodiesel synthesis via the chemical transesterification 66, the de novo biosynthesis of FAEEs in the engineered host becomes attractive. Recently, several studies were demonstrated that the increase carbon fluxes toward the fatty acid synthesis pathway through the overexpression of the malonyl‐CoA by carboxylase (ACC1) or other fatty acid synthesis‐related regulators promoted the FAEEs production (5.4 mg/L) directly converting from glucose in engineered S. cerevisiae 67, 68. In addition, impeding the progress of the β‐oxidation or preventing the lipid body accumulation could elevate the FAEEs titer to 17 mg/L 67, 69. By combination of the two strategies of the metabolic engineering (i.e. disruption of β‐oxidation and TAGs synthesis together with introduction of codon‐optimized acyltransferase genes) and fermentation engineering (i.e. nitrogen limitation), a significantly higher FAEEs titer (25 mg/L) was obtained in S. cerevisiae 70.

Due to the capability of the plant oil utilization and lipase secretion, the oleaginous Y. lipolytica has considered as an excellent host for ex novo and de novo FAEEs biosynthesis compared to S. cerevisiae 48, 59, 71, 72, 73, 74. Currently, it was reported that engineered Y. lipolytica with introduction of the heterologous FAEEs synthesis pathway and interruption of β‐oxidation was successfully demonstrated up to 1.18 g/L FAEEs in flask‐scale culture 75. Hence, development of Y. lipolytica biorefinery platform to convert low‐value hydrophobic carbon substrates to high‐value biofuels and oleochemicals could be anticipated in the near future.

2.7. Omega‐3 fatty acids

Omega‐3 fatty acids, which mainly includes cis‐5, 8, 11, 14, 17‐eicosapentaenoic acid (C20:5; EPA) and cis‐4, 7, 10, 13, 16, 19‐docosahexaenoic acid (C22:6; DHA), have significant health benefits in improving human's heart health, immune function, mental health, and infant cognitive development 61. Indeed, EPA and DHA de novo biosynthesis pathway only existing in certain marine microorganisms or the phytoplankton, which accumulated in the marine fishes via the food chain. Due to the shortage of nature source omega‐3 fatty acids, the engineered oleaginous yeast was found to be an alternative source for the sustainable production of EPA and DHA 22, 61, 76. DuPont company have successfully engineered Y. lipolytica strain for omega‐3 EPA production through integration of 30 copies of homogenous and heterogeneous genes that related to omega‐3 biosynthetic and supporting pathways along with β‐oxidation disruption 22, 61. A two‐stage fed‐batch fermentation process was developed to achieve high omega‐3 production: In growth phase, nitrogen is mainly provided by ammonium hydroxide for pH control to accumulate biomass; In the production phase, nitrogen is limited by switching the ammonium hydroxide to potassium hydroxide to produce lipid with high EPA content 19, 21. The EPA production reached its maximum at 25% of dry cell weight and more than 50% of the total fatty acid content by weight 19. Recently, it was also reported that the volumetric productivity of both biomass and omega‐3 fatty acids were almost doubled by a newly developed two‐stage continuous fermentation process 77. Hence, integrating both genetic engineering and fermentation process development led to the large‐scale production of omega‐3 EPA.

Taken together, though many current value‐added products are produced by Y. lipolytica from the hydrophilic substrates (e.g. glucose or glycerol), more and more advantages have been found that producing some high‐value products from the hydrophobic substrates (e.g. plant oils). To further improve the capabilities of Y. lipolytica for bioconversion of hydrophobic substrates into high‐value products under fermentation conditions, significant work needs to be done in engineering the yeast. In addition, due to the insolubility of the hydrophobic substrates in the aqueous medium, a novel bioreactor system should also be developed so that the mixing and mass transfer issues can be avoided or minimized to achieve comparable or similar cell growth and product formation under large‐scale fermentor conditions.

3. ASSIMILATIONS OF CARBON SOURCES

Y. lipolytica is a fascinating non‐conventional yeast that can be often found in dairy products such as raw milk, yogurt, cheese and sausage 1, 5. It not only enables to tolerate a fairly wide range of the environment from hypersaline concentration to a wide acidity range from pH 3 to 8 78, 79, but also enables to utilize the various substrates as sole carbon source from sugars, glycerols, alkanes, peptone, fatty acids to other types of lipids 80, 81, 82, 83.

3.1. Hydrophilic substrates

Y. lipolytica can use many hydrophilic substrates such as hexose (e.g. glucose, mannose, galactose, fructose), pentose (e.g. D‐xylose, L‐arabinose), acetate, lignocellulose, ethanol and glycerol. Among the various hydrophilic carbon sources, glucose and glycerol are most studied in lab and different industrial bioprocess scales. In most cases, glucose is widely used as the substrate for Y. lipolytica growth in lab‐scale research. Intriguingly, wild‐type Y. lipolytica is unable to produce alcohol but uses ethanol as carbon source due to the presence of own biosynthesis pathway on several NAD⁺ and NADPH‐dependent alcohol dehydrogenases 84. Crude glycerol is an inexpensive undesirable biodiesel byproduct and is regarded as the desirable substrate in industrial‐scale fermentation 85.

3.2. Hydrophobic substrates

Oleaginous yeast Y. lipolytica favors to grow on hydrophobic substrates such as methyl ester, fatty acids, plant oils (e.g. olive oil and soybean oil), animal fats and fossil oil 83, 86 due to its unique physiological features: (i) Production of the extracellular lipases in Y. lipolytica gives it high efficient utilization of the extracellular lipids as substrate. Lipase functions as triacylglycerol hydrolase catalyzing the ester bond of mono‐/di‐/tri‐glycerides of long‐chain fatty acids into glycerol and free fatty acids, which could be further converted into other metabolites. (ii) The intracellular fatty acids can be accumulated as TAGs or steryl esters in lipid bodies 76. An approximate 40% of dry cell weight as lipid bodies could be observed in wild‐type Y. lipolytica strain and up to 90% of dry biomass could accumulate in engineered strain with β‐oxidation impairment 87.

According to the comparative analysis of genome sequence recently, at least 25 putative extracellular and intracellular lipases are considered to be existed in Y. lipolytica 88, 89. YlLip2p has been characterized as the main extracellular lipase synthesized by Y. lipolytica during the stationary phase 32, 90. YlLip7p and YlLip8p have been characterized as extracellular membrane‐associated lipases, which are demonstrated 86.9% similarity in amino acid sequence with CdLip2p and 89.2% similarity in amino acid sequence with CdLip3p from Candida deformans, respectively 91. The engineered strain with deletions of three YlLip2, YlLip7 and YlLip8 lipase genes exhibited a growth defect on the YNB plate containing tributyrin or triolein as sole carbon source and no extracellular lipase activity was detected by monitoring the hydrolysis of p‐nitrophenyl butyrate to p‐nitrophenol and butyrate 5. This finding indicated that only three YlLip2p, YlLip7p and YlLip8p were able to secrete from wild‐type Y. liolytica cells and the others belong to internal triacylglycerol lipases facilitating the fatty acid metabolism. Furthermore, lipolytic enzyme production and activity have been verified to significantly increase in the present of the inducers such as TAGs, fatty acids and surfactants 4, 14, 92, 93. Conversely, the medium containing glucose, glycerol and inorganic compounds have been reported to repress the lipase synthesis 35, 94, 95.

Due to the lipase secretion, Y. lipolytica has the ability to assimilate lipids as the energy source to support both cell growth and regular maintenance. Hydrophobic substrates become the most effectual carbon sources, which in return promote the production and activity of the lipases 48. Among the hydrophobic carbon sources, plant oils from olive, corn, plam, almond, sunflower, rapeseed and soybean have attracted considerable attention due to their abundance in production as common agriculture commodities 96. Olive and almond oils were assesed to be the best inducer of the extracellular lipase (YlLip2p) in Y. lipolytica due to the high content of the oleic acid 38, 96. Additionally, unsaturated oleic acid (C18:1, 28%) and linoleic acid (C18:2, 51%) represent a significant portion of the accumulated fatty acids in native Y. lipolytica 2, 76, 80, 97. It is believed that the utilization rate by Y. lipolytica depends on the carbon chain length and double bond number of the fatty acids. Previous studies showed that the shorter chain of fatty acids, such as lauric, myristic and palmitic acids (C12:0, C14:0 and C16:0, respectively), were more accessible to degradation via β‐oxidation for supporting cell growth, while longer chains of fatty acids were more susceptible to elongation via ex novo fatty acid biosynthesis for lipid accumulation 98, 99. Notably, Y. lipolytica showed a higher uptake rate of unsaturated fatty acids (e.g. C18:1 and C18:2) than saturated fatty acids (e.g. C16:0 and C18:0) 100. With respect to the carbon source availability and the price fluctuation in the market, plant oils and animal fats have more total production than sugars in the world, thus will provide more flexibilities in both technical routes and economic benefits for development of efficient Y. lipolytica cell factories for making a series of high‐value products.

The utilization of hydrophobic substrates such as alkanes, fatty acids and TAGs in Y. lipolytica for the highly efficient bioconversion involves several metabolic pathways, which take place in different cellular compartment (Figure 1). Alkanes can enter directly into the cell, while the TAGs are hydrolyzed into glycerol and free fatty acids (FFAs) due to the lipase secretion by Y. lipolytica, which are uptaken by the cells via either transporters or diffusion 101. For the cytoplasmic alkanes, the primary oxidation of alkanes in cytoplasmic reticulum requires the NADPH‐dependent cytochromes P450 (CYP450) and alkane monoxygenase system (ALK genes), which are further converted into the FFAs with the catalysis by fatty alcohol oxidases (FAO) and fatty‐aldehyde dehydrogenases (FALDH) 83. When FFAs are used as the substrate, they are first activated into the fatty acyl‐CoA by fatty acyl‐CoA synthase (FAA1 or ACS1). Subsequently the fatty acyl‐CoA can be transported directly into the endoplasmic reticulum for lipogenesis or the peroxisome for β‐oxidation (involved enzymes Pox1‐6, MFE2 and Pex3/10) 102, 103. Fatty acids could be stored as lipid bodies in the format of TAGs or steryl esters. During the TAG synthesis process, DAG is further converted into TAG by diacylglycerol acyltransferases (DGATs: DGA1 and DGA2), which are the major regulator for TAG accumulation and lipid body formation 104, 105. Multicopy of DGAT genes DGA1 or DGA2 overexpression affects the lipid body size and improves lipid accumulation in Y. lipolytica cells. During the β‐oxidation process, six peroxisomal acyl‐CoA oxidases (AOX1‐6) encoded by Pox1‐6 genes are identified as chain‐length‐specific for fatty acid degradation 106, 107, 108, 109. Previous studies demonstrated that the Y. lipolytica strains with Pox1‐6 gene mutations impaired fatty acid degrading ability resulting in the increase of high‐lipid content in biomass 86, 110, 111, 112, 113. The next two steps of the β‐oxidation are catalyzed by the multifunctional enzyme (MFE) 218, 219. MFE knockout may result in a complete stop of β‐oxidation 218, 219. Since the β‐oxidation takes place in the peroxisome, whose function can be controlled by pex genes, therefore disruption of pex genes such as pex3 or pex10 increases lipid accumulation, but it also makes the yeast incapable of using fatty acid 19, 21. Overall, the knowledge on the hydrophobic substrate utilization in Y. lipolytica has been well studied and the efforts of further metabolic engineering strategies could accelerate the research for converting the hydrophobic substrates into high‐value products.

4. CELL SIZE AND MORPHOLOGY CONTROL IN Y. LIPOLYTICA

One unique feature for the Y. lipolytica yeast is the cell size and morphology change under different growth and environmental conditions. Cell size is the critical adaptive trait that alters cell physiology entirely. The connection of the cell polarization including cell cycle, growth rate, and the yeast‐to‐hypha transition with the cell size and morphology in Y. lipolytica is in response to the extracellular growth conditions and the gene regulations (Figure 2A). The feedback on the cell size changing also contributes to the cell cycle progression during the mitotic division. Cell increases in size would enhance the cytoskeleton biosynthesis and require more nutrients or other new proteins, enzymes and lipids to maintain its spatial structure and metabolism. On the other side, cell size and morphology determine the specific surface area (cell surface to volume ratio), which may directly affect the chance of lipids/fatty acids attaching to the cell surface, and then further impact the substrate utilization and bioconversion processes.

Figure 2.

Cell Size and Morphology Controlling in Budding Yeast. (A) Cell size and morphology controlling model. Cells size and morphology in budding yeast is controlled by the cell polarization, cell cycle and growth rate in response to the extracellular conditions. (B) Phylogenetic tree of ascomycete yeasts. Phylogeny of Y. lipolytica is relative to the other seven Saccharomycotina species and the Taphrinomycotina S. pombe used as outgroup. (C) Cell division in budding yeast and fission yeast. Upper: In budding yeast, new bud is emerged adjacent to the previous division site as bud site selection in haploid cells and the cells divide asymmetrically during mitosis. Bottom: In fission yeast, S. pombe is rod‐shape and only stable in haploid state. During the mitotic cycle, S. pombe grows in length by tip extension without the change in diameter and then divide into two cells via medial fission. Purple arrow indicates the additional production of new plasma membrane materials, such as the lipids, proteins and the cell wall biosynthesis‐related enzymes. (D) Yeast‐to‐hypha transition in wild type Y. lipolytica ATCC20362 strain

4.1. Cell polarization

Cell polarization is a fundamental mechanism in proliferation, differentiation, migration and morphogenesis, which are involved in cytoskeleton, scaffold proteins, membrane composition and signaling transduction pathways. Actin cytoskeleton distribution is directly related to the cellular structure and morphology. In Y. lipolytica, YlBEM1p, as one of the actin cytoskeleton‐associated protein, plays a prominent role in the spatial backbone of the cell wall structure construction and participates in the actin cytoskeleton organization, cortical actin localization and chitin distribution. YlBEM1 gene is upregulated during the late stage of the cell division and differentiation as high level of the YlBEM1p are found to be accumulated at the site of the growing tips of the Y. lipolytica, especially at the mother‐bud neck and filamentous developmental compartment 114, 115, 116. A strain lacking a functional YlBEM1p exhibited the impairment of the organization of actin and chitin in the cell surface structure 115, indicating a possible role for the YlBEM1p scaffold protein in the maintenance of cell polarity in Y. lipolytica. A good understanding of the polarization mechanisms in budding yeast is needed to reveal the cell morphogenesis. Multiple genes are also involved in the cell cycle progression and the yeast‐to‐hypha transition as well as the cell polarity. The launch of cell polarization process is initial key factor for cell size controlling.

4.2. Cell cycle

The budding yeast Saccharomyces cerevisiae and the fission yeast Schizosaccharomyces pombe are the most studied model organism in the cell cycle control and their cell size (Figure 2B). The size–dependent cell cycle progression is involved in these two species. In S. pombe, the evidence for cell checkpoint in the cell cycle was proven by halting the cell growth via the actin depolymerization that arrested the G2/M transition 117. The size‐sensing mechanism in cell length was emerged during the G2/M progression through the activation of Cdc13/CDk1signaling cascade in fission yeast 118.

By contrast, the size‐threshold is important in the duration of G1 phase in the budding yeast, while the S/G2/M phases are weekly affected by cell size 119, 120. The mother‐daughter cell will separate asymmetry in division size in budding yeast. In general, the size of daughter cell is 20% less than its mother cell 121. Variability in cell size at birth happens, the smaller cells spend longer time in G1 phase during cell cycle for growth as large as initial larger cells to partially compensate the initial variation in size 120, 121. In nutrient‐rich medium, the G1 cyclin gene CLN3 as one of the earliest activated gene promotes the entry into cell cycle in S. cerevisiae 122. Cln3p is growth‐dependent and highly unstable sizer protein which activates the G1/S transcription factor, SBF, by depressing the transcriptional inhibitor, Whi5. This finding suggested that the total amounts of Cln3p correlates with the cell size and is an essential prerequisite to determine the exact timing of the G1/S transition during the cell cycle. Furthermore, deletion of the WHI5 and CLN3 genes in filamentous yeast Ashbya gossypii affected the cell size changes toward smaller and larger size in growth, respectively 123. Although size checkpoint in cell cycle has been little‐studied in Y. lipolytica, the mechanisms in cell growth and cell division compared with S. cerevisiae are shared the similar behaviors as the more closely relationship between both species in evolution belongs to Saccharomycotina rather than the Taphrinomycotina S. pombe (Figure 2C).

The life cycle of the ascomycete Y. lipolytica strain involves haploid and diploid phases that is capable to undergo both sexual and asexual reproduction responding to the different parameters such as genetic background 124, 125, 126 and extracellular environment (e.g. medium composition, cell density, carbon and nitrogen sources, pH value, temperature, aeration and so forth) 1, 127. Meiosis and sporulation phenomena happened to Y. lipolytica is subject to the extremely harsh growth conditions. Most of natural Y. lipolytica isolates show stable haploid form in life cycle via multilateral budding as asexual reproduction mean to form daughter cells that contain identical genetic materials but different sizes at birth. Bipolar bud‐site selection is the event in the early stage of the cell cycle happening prior to the bud appearance. Both YlRSR1 and YlRAS2 genes in Y. lipolytica are considered to participate the determination of budding site, which are activated by YlCdc25p and guanine nucleotide‐exchange factor (GEF), respectively. YlRSR1p and YlRAS2p also take part in the regulation of YlCdc24p/YlCdc42p signaling pathway that function in the control of the cell polarized growth 128, 129, 130.

Haploid cells of Y. lipolytica with opposite mating types A or α will fuse to produce a diploid zygote in the sexual reproduction when encountering the severe environment stresses 131. Previous study has shown that deletion of the YlBEM1 gene demonstrated the loss of the mating ability due to the large‐scale of actin cytoskeleton delocalization, suggesting that the YlBEM1p is essential for the mating process 115. In contrast to Y. lipolytica, the conventional yeast S. cerevisiae has been observed to demonstrate the ability of mating‐type switching 132, 133. The aim of the sexual reproduction would enable to induce the chance of the inherent diversification on the natural selection.

4.3. Yeast‐to‐hypha transition

Dimorphic fungus Y. lipolytica has been considered as a suitable model for studies on the yeast‐to‐hypha (YH) transition of the eukaryotic cell differentiation that exhibits various morphological forms, ranging from unicellular yeast form to multicellular filamentous form called pseudohyphae or septate true hyphae 134, 135, 136, 137. Regulation of the YH transition in Y. lipolytica is greatly affected by the environmental and physiological conditions together with genetic characteristics 1, 135, 138.

4.3.1. Controlling cell shape by growth condition of yeast‐to‐hypha transition

Under the nutrient‐rich environment, the cells grow as the yeast‐like form that display the rounded and budding shapes; whereas under the stressful conditions, the cell morphologies tend to be elongated and branched that is in the mycelial state. Filamentous growth in Y. lipolytica could be induced by different effectors, e.g. the present of N‐acetylglucosamine, citrate or serum in the medium 138, 139, the nitrogen source 140, the ratio of the carbon and nitrogen, the pH value 127 or under the osmotic pressure (e.g. hypersaline concentration) and oxidative (e.g. hydrogen peroxide) stress 2, 16. The YH transition process was observed when the wild type Y. lipolytica strain ATCC20362 was cultured in the YPD medium for 72 hours in flask‐scale culture (Figure 2D). Nutrient sources were continuously consumed with the time during the cell growth that induced the cell morphology switching from yeast‐like to filamentous‐form. Therefore, detection extracellular nutrient condition is required for Y. lipolytica to determine its growth rate and the morphology.

4.3.2. Controlling cell shape by gene regulation of yeast‐to‐hypha transition

Gene regulation is able to modify the cell shape through the positive or negative regulators in yeast‐to‐hypha transition in dimorphic Y. lipolytica. Positive gene effectors such as YlMHY1, YlBEM1, YlCLA4, YlRSR1 and YlRAS2 genes induce the pseudohyphal/hyphal growth, while negative gene effectors including YlZNC1 and YlTPK1 regulate the cellular morphology as yeast‐form during YH transition (Table 1) 124, 141, 142, 143. Besides cellular morphologies, wild‐type Y. lipolytica also demonstrated diversity in colony morphologies ranging from smooth and glistening to heavily convoluted and matt with rough‐surface 2, 84.

Table 1.

Genes involved in Dimorphic Y. lipolytica

Yeast‐to‐hypha effectors	Gene	Y. lipolytica ortholog/accession	Function	Reference
Positive effectors: (Promote Dimorphic Transition)	YlMHY1	YALI1B28150g/ AF124404	C2H2‐type Zinc Finger Protein; Transcription Factor; Bind to Putative Stress Response Elements (STREs)	142
	YlBEM1	YALI1F35061g/ AY084035	Cytoskeleton‐associated Protein; Actin Cytoskeleton Organization in Cell Polarization	115
	YlRAS2	YALI1E35305g/ AF321464	GTPase; Regulation Dimorphic Transition	128
	YlHOY1	YALI0A18469g/ Z34956	Transcriptional Regulatory Protein; Function in Hyphal Formation	154
	YlBMH1	YALI1B18930g/ Q876M0	14‐3‐3 Protein; Promotion Filamentous Growth	124, 162
	YlCLA4	AF233061	PAK Protein Kinase in Mitogen‐Activated Protein Kinase (MAPK) Cascade; Maintenance in Cell Polarity; Distribution of Chitin in the Cell Wall; Essential for Filament Formation	124, 167
	YlSTE7	AJ007393	MAPKK Protein Kinase in MAPK Cascade	28, 165
	YlSTE11	AJ577132	MAPKKK Protein Kinase in MAPK Cascade; Induce The Yeast‐to‐hypha Dimorphic Transition; Essential in Mating	28, 165
	YlRKA1	HQ450396	cAMP‐Dependent Protein Kinase Regulatory Subunit; Function in Hyphal Growth	124, 217
Negative effectors: (Inhibit Dimorphic Transition)	YlZNC1	YALI1B06928g/ AJ575099	Zinc Finger Protein; Transcription Factor; Repression Hyphal Cell Formation and Functions as Part of a Complex Network Regulation in Mycelial Growth	124
	YlTPK1	YALI1C11271g/ FM865406	cAMP‐Dependent Protein Kinase A; Regulatory Role of the PKA Pathway in Dimorphism and Mating	28, 141, 165

YlMHY1 encoding protein YlMHY1p is involved in the YH transition in Y. lipolytica. YlMHY1p contains a glutamine‐rich tract at its amino acid terminus and two C₂H₂‐type zinc finger motifs at its carboxyl terminus 142. Zinc finger domain in YlMHY1p shows strong homology to the S. cerevisiae stress‐responsive transcription factors, Msn2p, Msn4p 126, 144 and Crz1p 145, 146, which reveals that MHY1p acting as transcription factor is able to bind to a stress‐response element (STREs). By cis‐acting DNA sequences, the upstream of a number of genes was identified to confer tolerance to a variety of stresses such as carbon source starvation, heat shock, osmotic and oxidative stresses 142, 147, 148, 149. Previous study has demonstrated that during the YH transition in Y. lipolytica response to adverse conditions, the YlMHY1 gene expression dramatically increased 142. Deletion of YlMHY1 gene does not affect cell viability but results in complete inability to undergo mycelial growth, suggesting that YlMHY1 gene expression is the necessary cause for filamentation and acts as positive effectors in dimorphic transition 142. Currently, it was demonstrated that disruption of YlMHY1 in Y. lipolytica increased intracellular oil accumulation due to the upregulation of citrate synthases, which further improved the citrate production and the followed direction of carbon flux towards the lipid biosynthesis rather than amino acid metabolism 150. Moreover, the nitrogen starvation has been widely implemented in oleaginous microorganism for lipid accumulation 151, 152 that also affects the amino acid biosynthesis 153. Hence, the YlMHY1p plays a critical role in several biological pathways that involve in lipid, amino acid and nitrogen metabolisms.

In addition to the YlMHY1 gene, the YlHOY1, YlRAS2, YlBEM1 genes are also reported to involve in the YH transition in Y. lipolytica, which induced the mycelium growth in response to environmental signals 128, 142, 143, 154. For example, mutation in YlHOY1 gene encoding Hoy1p as transcription factor in Y. lipolytica has also confirmed to inhibit YH transition and prevent mycelial growth 154. In addition, Y. lipolytica has three Ras proteins that are small GTP‐binding proteins localized to the plasma membrane and regulate the signal transduction pathways involving cell proliferation, differentiation and survival 155. Particularly, RAS2 has a critical role in cellular morphogenesis 143, 156, 157. The expression of YlRAS2 was increased at the transcription level during the dimorphic transition in Y. lipolytica as YlRAS2 knockout cells exhibited the defect in filamentation 128, 143. Overexpression of YlMHY1 gene in YlRAS2 knockout cells could restore the filamentous growth, whereas overexpression of YlRAS2 gene in YlMHY1 knockout cells were unable to induce the either pseudohyphal or hyphal growth 128, 142. Furthermore, YlBEM1p protein in Y. lipolytica was found strong homology to conserved motifs of ScBEM1p (32.5% identity) involved in the control of cell polarity and differentiation in the S. cerevisiae 115, 158, 159, 160, 161. YlBEM1 transcription was increased during the dimorphic transition and promotes the hyphal growth 115. However, YlBEM1 was non‐essential gene for cellular filamentation that was consistent to the results that overexpression of YlMHY1 in the cells with defective YlBEM1 partially restored the filamentous growth, whereas overexpression of YlBEM1 in the YlMHY1‐deleted cells had no apparent effect on recovering the morphological defect 115, 162. In the absence of functional YlMHY1p would be insufficient to induce any filamentation. In a summary, previous studies revealed that the transcription factor YlMHY1 is a downstream signal transducer of YlRAS2 and YlBEM1 in the regulation of dimorphic transition in Y. lipolytica. Therefore, the regulation of the of YH transition in the dimorphic Y. lipolytica by deletion or overexpression of YlMHY1 can potentially affect the hydrophobic substrate utilization and the further bioconversion processes.

In most of dimorphic fungi, the yeast‐to‐hyphal transition is also regulated by the RAS/mitogen‐activated protein kinase (MAPK) and the cAMP/protein kinase A (PKA) signal transduction pathways 163, 164, which have opposite actions in regulation of dimorphism. Activation of MAPK signaling pathway regulates the cellular morphology towered to invasive growth. Multiple genes (e.g. YlCLA4, YlSTE7 and YlSTE11) are involved in MAPK pathway and may regulated by YlZNC1p 124. These hyphal‐specific genes are upregulated and encoded by proteins YlCLA4p, YlSTE7p and YlSTE11p that function as PAK, MAPKK and MAPKKK in MAPK signaling 165, 166, 167. YlZNC1p encoded by YlZNC1 gene acts as transcription factor, which contains a Zn(II)₂C₆ fungal‐type zine finger DNA‐binding domain and a leucine zipper domain that play a role for its correcting the subcellular localization to the nucleus. For the gene expression level, YlZNC1 transcription is decreased during the YH transition in Y. lipolytica. Deletion of YlZNC1 gene affects the cell phenotype that increases the hyphal cell formation 124. Both YlBMH1 and YlBMH2 genes encoding proteins YlBMH1p and YlBMH2p, respectively, are also involved in hyphal growth in MAPK cascade, which are also mediated by YlZNC1p in Y. lipolytica 162. Furthermore, the cAMP‐dependent PKA pathway activation controls the cellular morphology as yeast form during the dimorphic transition. Cells with deletion in YlTPK1 gene, which encodes the PKA catalytic subunit, increased the growth in filamentous form. YlRKA1 gene encoding the regulatory subunit of PKA was upregulated in hyphal growth. Both YlTPK1 and YlRKA1 genes are guided by YlZNC1p 141. Based on these previous studies, YlZNC1 as one of critical negative regulators in dimorphic transition is responsible for repressing mycelium formation and involved in part of complicated signal transduction network in Y. lipolytica, particularly via the RAS/MAPK and the cAMP/PKA signaling cascades. Hence, it is expected that the changing in cell morphology via the positive or negative regulators in YH transition may have significant effects on the efficiency of hydrophobic substrate utilization and cell growth of Y. lipolytica.

Based on all discussion above, it is expected that the negative regulation of the YH transition in the Y. lipolytica can be achieved by either mutation of positive effectors such as YlMHY1, YlRAS2, YlCLA4, and YlHOY1or by overexpression of negative effectors such as YlZNC1 and YlTPK1. It is expected that regulation of the YH transition may change the cell size and morphology, which further changes cells’ outer surface area so that the droplets of hydrophobic substrates have more chances of being attached to cells’ surface and then of being further utilized. More future studies are required to reveal the relationship between the cell size and morphology, YH transition, substrate uptake, cell growth, and product formation.

4.3.3. Morphology engineering

Filamentous fungi have been widely used in a board range of industrial applications in many decades. The fungal morphology has been influenced by various process parameters, especially the environmental cultivation conditions including the osmolality, agitation, aeration, pH value, dissolved oxygen levels, mass transfer and mechanical stress 168, 169. Wucherpfennig et al. proposed the following dimensionless morphology number to describe the correlation of the fungal morphology with the productivity 170:

$$Morphologynumber=\frac{2·\sqrt{A}·S}{\sqrt{\pi }·D·E}$$

where A is the projected area of the cell, S is the solidity of the cell (i.e. roughness), D is the maximum diameter of the cell shape, and E is the length to width ratio. By this definition, the morphology number for perfect circle cell is equal to 1, while for the elongated cell including irregular pseudohyphae and hyphae is between 0 to 1. It was found that significantly higher production was achieved by cells with much smaller morphology numbers 112. Both fructofuranosidase and glucomylase production from Aspergillus niger has demonstrated the highly negative correlation with the morphology number 170. Optimal productivity in biotechnological culture process is often strongly correlated 171 with the specific fungal morphological form 169, 172, 173, 174, 175.

There have been several studies that used filamentous fungus Aspergillus niger as a model organism to study the morphology engineering due to its salt‐tolerant characteristic and complex morphology cell cycle ranging from dense spherical pellets to viscous dispersed form consisting of aggregated hyphal structure in response to culture process. In one example, fungus morphology was found affected by the osmolality. At higher osmotic pressure, the smaller spherical pellets were shown 176. The specific productivity of fructofuranosidase as secreted enzyme from A. niger SKAn1015 was strongly enhanced from 0.5 to 9 U/mg/h by providing the additional sodium chloride for high osmolality 170, 176. In addition to the studies on fungi, recent studies on mammalian cells also revealed that the hyperosmotic pressure increases the specific productivity of antibody in Chinese Hamster Ovary (CHO) cells 171, 177, 178, 179.

Taken together, although no work was directly done on the effect of the morphology engineering on the YH‐transition and productivity in Y. lipolyica, the phenomenon of cell morphology highly linking to productivity performance has been verified in other microbial and mammalian cells. Recent research and progresses in cell size and morphology control of the Y. lipolytica cells has built a solid base for us to further improve the bioproduction of the yeast in future by combining both morphology engineering and metabolic engineering.

5. METABOLIC ENGINEERING TOOLS IN Y. LIPOLYTICA

5.1. Traditional cloning methods

Molecular genetic tools have played a pivotal role in fundamental and applied yeast research. Auxotrophic marker genes are often used in molecular genetic engineering in yeast. Genetic modifications require the use of selectable marker genes for efficient detecting and screening out the transformed cells 180. A mutation of certain auxotrophic marker gene used in yeast genetics requires the culture medium to supply specific nutritional complements. Therefore, the auxotrophic yeast strains which lack the functional chromosomal copy of the marker gene can be propagated only in medium containing the appropriate nutrient component for growth or in a complex medium, e.g., yeast extract and/or peptone. Most commonly used auxotrophic marker genes are wild‐type alleles of yeast genes that encode key enzyme in the metabolic pathways towards essential compound biosynthesis 181. A typical example is the URA3 gene, which encodes the protein called orotidine‐5’‐phosphate decarboxylase or OMP decarboxylase 182. OMP decarboxylase is an enzyme necessary for de novo pyrimidine biosynthesis, which is the precursor of DNA and RNA building block. URA3 auxotrophic strains 13 become prototrophic by uracil supplement in the medium due to the function of endogenous uracil phosphoribosyl transferase encoded by the UPP gene in yeast. Other important auxotrophic maker genes include TRP1, ADE2, LEU2, LYS2, XPR2 and HIS3, which encode phosphoribosyl‐anthranilate isomerase Trp1 for de novo biosynthesis of L‐tryptophan 183, 184, adenylosuccinate synthetase for de novo AMP and purine nucleotide 185, 186, 3‐isopropylmalate dehydrogenase for L‐leucine synthesis 187,$\alpha$‐aminoadipate reductase for lysine synthesis 188, 189, alkaline extracellular protease for serine synthesis 190, 191, 192, 193 and imidazoleglycerol‐phosphate dehydratase for L‐histidine synthesis 194, 195, respectively. In addition to the auxotrophic selection, several dominant selection markers including hygromycin, nourseothricin, chlorimuron ethyl and mycophenolic acid resistances were also exploited to increase the applicability for the development of engineering Y. lipolytica strains 196.

In general, generation of auxotrophic yeast strain is the first step for development of a genetic engineered strain. After an auxotrophic host strain is obtained, transforming a plasmid with the target genes and the corresponding selection marker into the host is one of the general strategies for metabolic engineering, albeit the loss of plasmid copy number is also a concern and may have the issue on genetic instability. To avoid the genetic instability issue caused by plasmid loss, the target genes are often integrated into the chromosome of the production strain 142, 143, 144. It has been demonstrated in many researches that genomic integration provides higher stability of heterologous gene expression over extended culture time and eliminates the constant screening of an auxotrophic genetic marker. Successful transformation with a linear gene cassette containing a gene of interest and a selection marker incorporates into the genome either at a random locus by using non‐homologous end‐joining (NHEJ) DNA repair system or at a specific site by homologous recombination (HR). However, the rate of HR only occurs when bearing long homologous arms (0.5 to 1.0 kb) flank the target gene cassette 197. Indeed, the challenge for multigene engineering in metabolic pathway is the limited number of viable selection markers can be used in non‐conventional yeast 198. Several strategies have been developed by using systems of Cre‐loxp, HisG and lacZ, for marker recovery to overcome this weakness. The selection marker is flanked by HisG, lacZ or loxp gene sequences, which will be removed by HR or Cre recombinase activity after gene cassette incorporated into the chromosome. But, the issue in random insertion or unknown‐site genome integration cannot be completely solved by the methods above. The severe impact is inaccurate recombination that leads to uncertain gain or loss of function, which probably has adverse effect on cell survival, growth rate, or major metabolic pathways.

5.2. Advanced genome editing engineering

Since early 1970s, thousands of restriction‐modification (R‐M) enzymes have been identified and screened from diverse orgnisms, which have been widely used in traditional genetic manuipulations 220. However, discovery of a new R‐M enzyme from a specific organism is very time‐consuming and inefficient, about 20 years ago, hybrid restriction enzymes, Zinc finger fusions to Fok I, were discovered to create “artificial” nucleases to cut DNA near a predetermined site 201. By inserting the desired sequence into the genome, researchers were able to design the the homing endonuclease to recognize the chromosome sequence, e.g. the zinc finger nucleases (ZFNs) 199, 200, 201 and the transcription activator‐like effector nucleases (TALENs) 202, 203, 204. Although ZFNs and TALENs could be used for site‐directed genome editing, the limited number of targets and the difficulty inherient in protein design, synthesis and validation still remain the biggest barrier to be adopted widely for rutine use 205.

In the past a few years, the type II bacterial clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR‐associated protein (CRISPR‐Cas9) system has been exploited as a powerful genome‐editing tool in numerous prokaryotes and eukaryotes that can be programmed to target specific gene and to edit DNA at precise locations via either NHEJ or HR of DNA repair mechanism. CRISPR‐Cas9 system has been excuted in several species of yeast, Schizosaccharomyces pombe 206, Candida albicans 207, S. cerevisiae 208, Kluyveromyces lactis 209, and Y. lipolytica 59, 210, 211. Recent study was exhibited the efficiency of single gene knockout and knockin could be achieved near 100% after cotransformation with CRISPR‐Cas9 in S. cerevisiae 208. A near 92% single gene disurption efficiency was reached when single guide RNAs (sgRNA) were transcribed with syntehtic hybrid promoter that combine native RNA polymerase III promoter with tRNA in Y. lipolytica 212. Over 64% of homologous recombination efficiency was achived after cotransformation of the CRISPR (Cas9 and sgRNA) expressing plasmid with a homologous recombination donor plasmid into the wild‐type Y. lipolytica PO1f strain, whereas the disruption rate increased to 100% via HR when CRISPR and homologous recombination plasmids were cotransformed into the PO1f KU70 knockout strain that was disrupted in NHEJ mechanism 59. Hence, homology‐integrated CRISPR‐Cas9 system becomes a powerful tool for creating the metaboic and morphologic engineered yeast strains currently.

In addition to the CRISPR‐Cas9 system, the novel self‐replicating YaliBricks vectors 213 and piggyBac transposon system 196 have also been develped recently that provide an effiective methods to facilitate the metabolic pathway engineering in Y. lipolytica. Wong et al. 213 have demonstrated the rapid pathway assembly and screening, in which a set of modular assembly of the five‐gene, up to 12 kb, was constructed in one week by using the Yalibricks genetic toolkit to build the biosynthesis pathway for the antiobiotic pigment violacein. The violacein‐producing Y. lipolytica strain could be quickly confirmed by visual screening and approximately 31.5 mg/L violacein was produced. Likewise, the piggyBac TTAA‐specific transposon system as the one of newly widerspread genetic tools in eukaryotic cells, has also showen the increase of foreign DNA integration (up to 100 kb) into the genome, which led to rapid and increased production of the heterologous protein 214, 215. Furthermore, the piggyBac casette with the piggyBac inverted terminal repeats flanking fragments could be excisied precisely from the genome without scar after integration at a TTAA site 216. For all of these reasons, the piggyBac transposon platform in Y. Lipolytica was established by Wagner, Williams and Alper for creating mutant libraries and targeted genome‐editing 196. Indeed, the library of available tools have been expanded and established in Y. lipolytica nowadays that will accelerate the development of engineered strains for sustainable and efficiently production of a wide range of products.

6. CONCLUSION AND PERSPECTIVE

The oleaginous Y. lipolytica yeast has attracted interested due to its ability to convert a range of carbon sources, especially low‐cost hydrophobic substrates, into high‐value products and high capacity for lipid biosynthesis. The recent studies have been highlighted the exploration of the Y. lipolytica in biotechnological applications in various fields such as drug synthesis, chemicals, pharmaceuticals, food and feed industries and so forth. However, compared to the hydrophilic substrates, more work in both cellular engineering and bioreaction engineering of Y. lipolytica should be conducted in future to achieve higher efficiency in oil‐based substance uptake and the bioconversion of the hydrophobic substrates into the high‐value products. It is believed that engineering cell size and morphology has a great potential to significantly improve substrate uptake and the further bioconversion in Y. lipolytica. For the fermentation processes that use hydrophobic substrates such as vegetable oils, filamentous growth cells are capable to attach more small oil droplets around the cell wall, which speed up the hydrophobic substrate uptake form extracellular environment; yeast‐like cells might demonstrate the higher bioconversion of the hydrophobic substrates, which shorten the traveling distance from cell membrane to the organelles. Regulation of cell size and morphology is governed by the cell polarization, cell cycle, growth rate and yeast‐to‐hypha transition in response to the external growth conditions and the gene regulations in budding yeast, which may help increase the chance of the distributed hydrophobic substrate droplets attaching to the cell surface, thus facilitate the substrate consumption and the intracellular bioconversion. It has been demonstrated in some previous research that the fungus morphology was strongly related to the productivity. It is expected that altering the morphology of the Y. lipolytica and improving the metabolic pathways by advanced metabolic engineering approaches or the cultivation process optimization can greatly improve the hydrophobic substrate utilization, titer and productivity of value‐added products in the future.

CONFLICT OF INTEREST

The authors have declared no conflict of interest.

Soong YV, Liu N, Yoon S, Lawton C, Xie D. Cellular and Metabolic Engineering of Oleaginous Yeast Yarrowia lipolytica for Bioconversion of Hydrophobic Substrates into High‐Value Products. Eng Life Sci. 2019;19:423–443. 10.1002/elsc.201800147