Sustainable production of FAEE biodiesel using the oleaginous yeast Yarrowia lipolytica

Abstract Fatty acid ethyl esters (FAEEs) are fatty acid‐derived molecules and serve as an important form of biodiesel. The oleaginous yeast Yarrowia lipolytica is considered an ideal host platform for the production of fatty acid‐derived products due to its excellent lipid accumulation capacity. In this proof‐of‐principle study, several metabolic engineering strategies were applied for the overproduction of FAEE biodiesel in Y. lipolytica. Here, chromosome‐based co‐overexpression of two heterologous genes, namely, PDC1 (encoding pyruvate decarboxylase) and ADH1 (encoding alcohol dehydrogenase) from Saccharomyces cerevisiae, and the endogenous GAPDH (encoding glyceraldehyde‐3‐phosphate dehydrogenase) gene of Y. lipolytica resulted in successful biosynthesis of ethanol at 70.8 mg/L in Y. lipolytica. The engineered Y. lipolytica strain expressing the ethanol synthetic pathway together with a heterologous wax ester synthase (MhWS) exhibited the highest FAEE titer of 360.8 mg/L, which is 3.8‐fold higher than that of the control strain when 2% exogenous ethanol was added to the culture medium of Y. lipolytica. Furthermore, a synthetic microbial consortium comprising an engineered Y. lipolytica strain that heterologously expressed MhWS and a S. cerevisiae strain that could provide ethanol as a substrate for the production of the final product in the final engineered Y. lipolytica strain was created in this study. Finally, this synthetic consortium produced FAEE biodiesel at a titer of 4.8 mg/L under the optimum coculture conditions.


| INTRODUC TI ON
Currently, the demand for renewable and sustainable energy is rapidly increasing. Biodiesel is an important sustainable energy source, and fatty acid ethyl esters (FAEEs), popularly known as "biodiesel," can serve as important alternative industrial diesel fuels.
In the current global biodiesel market, most commercially available FAEEs are produced via a transesterification reaction between ethanol and various lipid feedstocks, such as plant oils or animal fats, in the presence of a catalyst (Dunn, Ngo, & Haas, 2015;Santana, Tortola, Reis, Silva, & Taranto, 2016;Suppalakpanya, Ratanawilai, & Tongurai, 2010). However, the major problems associated with this chemical synthesis method include the restricted availability of lipid sources and the risk of environmental pollution. In recent years, the use of genetically engineered microorganisms has provided a sustainable and environmentally friendly bioroute for the production of value-added products, including biofuels, biochemicals, and other bioactive compounds (Cheon, Kim, Gustavsson, & Lee, 2016;Lee, Chou, Ham, Lee, & Keasling, 2008;Mao, Liu, Sun, & Lee, 2017;Marienhagen & Bott, 2013;Yu, Pratomo Juwono, Leong, & Chang, 2014).
Escherichia coli was the first genetically engineered microorganism used for FAEE production, which was achieved by coexpression of the pyruvate decarboxylase gene and alcohol dehydrogenase gene from Zymomonas mobilis and the atfA gene encoding WS/DGAT (wax ester synthase/acyl-coenzyme A: diacylglycerol acyltransferase) from Acinetobacter baylyi ADP1 (Kalscheuer, Stölting, & Steinbüchel, 2006 showed the greatest potential for the heterologous production of wax esters and FAEEs in microbes (Barney, Wahlen, Garner, Wei, & Seefeldt, 2012). Moreover, the production of FAEEs by engineered E. coli cell factories was further enhanced upon adopting several metabolic engineering strategies and synthetic biology tools (Steen et al., 2010;Wierzbicki, Niraula, Yarrabothula, Layton, & Trinh, 2016).
However, in S. cerevisiae, the reported FAEE yield by heterologous expression of a wax ester synthase remains rather low because the production of FAEEs is greatly limited by the small pool of fatty acyl-CoA and/or free fatty acids Yu et al., 2012). Compared to the previously used E. coli and S. cerevisiae systems, the oleaginous yeast Y. lipolytica has outstanding lipid accumulation capacity. Thus, abundant intracellular fatty acyl-CoA (or free fatty acids) is available for the production of FAEEs and other fatty acid-derived bioproducts through this host's metabolic system (Abghari & Chen, 2014Mlíčková et al., 2004;Tai & Stephanopoulos, 2013). Xu, Qiao, Ahn, and Stephanopoulos (2016) demonstrated the construction of an engineered Y. lipolytica strain for FAEE production through the expression of A. baylyi ADP1 targeted to the endoplasmic reticulum and peroxisome. To our knowledge, this was the first report of FAEE production in Y. lipolytica using metabolic engineering strategies. Gao et al. (2018) further optimized FAEE production in Y. lipolytica through extensive metabolic engineering, and increased FAEE titer was achieved in Y. lipolytica with the addition of exogenous ethanol. Very recently, Ng et al. (2020) demonstrated the production of FAEEs from vegetable cooking oil as a model food waste in the engineered Y. lipolytica. In this small number of studies conducted to date, the FAEE titers obtained with engineered Y. lipolytica strains were low. In this work, we aimed to test the usefulness of new strategies by optimizing FAEE production from endogenously produced ethanol in Y. lipolytica and by cocultivation of an ethanol producer with Y. lipolytica. Finally, engineered Y. lipolytica strain was shown to be capable of producing FAEEs using endogenously produced ethanol. As shown in Figure 1, FAEEs can be generated from ethanol and fatty acyl-CoA by expressing heterologous WS in our engineered yeast strains. Besides, we demonstrated that a coculture system consisting of the yeasts Y. lipolytica and S. cerevisiae has potential applications in the sustainable production of FAEEs.

| Strains, media, and culture conditions
The E. coli strain TOP10 was used as the host in this study for the cloning and propagation of plasmids. E. coli strains carrying recombinant plasmids were routinely cultured at 37°C in Luria-Bertani (LB) broth containing 1% tryptone, 0.5% yeast extract, and 1% sodium chloride or on LB agar plates supplemented with 100 μg/ ml ampicillin. The Y. lipolytica strain Po1g, a leucine-auxotrophic derivative of the wild-type strain W29 (ATCC 20460), was chosen as a host organism for heterologous gene expression in this study. A yeast coculture of the S. cerevisiae strain S288C and the engineered Y. lipolytica strain was designed and developed in this study for the direct production of FAEEs. Routine cultivation of the Y. lipolytica strains, S. cerevisiae strains, and yeast coculture was carried out at 30°C in yeast extract-peptone-dextrose (YPD) liquid medium containing 1% yeast extract, 2% peptone, and 2% dextrose or on YPD agar plates. Synthetic complete medium lacking leucine (YNBleu) and containing 2% glucose and 0.67% yeast nitrogen base w/o amino acids was used for the selection of Y.
lipolytica Leu + transformants. The strains and plasmids used in this study are listed in Table 1.

| DNA manipulation
The plasmid pYLEX1 (Yeastern Biotech, Taipei, Taiwan) containing a strong hybrid promoter (hp4d) was used for gene expression in the Y. lipolytica host strain Po1g. Recombinant plasmids containing different gene expression cassettes were constructed by the following procedures and are graphically depicted in Appendix 1, Figures A1-A8. The PDC1 and ADH1 genes from S. cerevisiae S288C were ligated into the Pml I/BamH I sites of pYLEX1 to yield the plasmids pYLEX-1PDC1 and pYLEX1ADH1, respectively (Appendix 1, Figures A1 and   A2). The gene expression cassette of ADH1 was amplified by PCR from pYLEX1ADH1 using a forward primer containing Cla I, Nhe I, and Xba I sites and a reverse primer containing Cla I, Xma I, and Mlu I sites, and then ligated into the Cla I site of pYLEX1PDC1 to yield the plasmid pYLPDC1ADH1 (Appendix 1, Figure A3). The native GAPDH gene encoding glyceraldehyde-3-phosphate dehydrogenase from Y.
lipolytica Po1g was ligated into the Pml I/BamH I sites of pYLEX1 to yield the plasmid pYLEX1GAPDH (Appendix 1, Figure A4). The gene expression cassette of GAPDH was amplified by PCR from pYLEX-1GAPDH using a forward primer containing a Mlu I site and a reverse F I G U R E 1 Schematic diagram summarizing metabolic engineering strategies for fatty acid ethyl esters (FAEE) production in engineered Yarrowia lipolytica using monoculture and coculture cultivation. The reconstructed biosynthesis pathway for endogenous ethanol production was constructed in Y. lipolytica Po1g via co-overexpression of two heterologous enzymes pyruvate decarboxylase (Pdc1p) and alcohol dehydrogenase (Adh1p) from Saccharomyces cerevisiae S288C, one native glyceraldehyde-3-phosphate dehydrogenase (GAPDH) from Y. lipolytica Po1g for NADH regeneration. Wax ester synthases from Marinobacter aquaeolei VT8 (MaWS1) and Marinobacter hydrocarbonoclasticus DSM 8798 (MhWS) were then introduced into Y. lipolytica, respectively, for FAEE production in the monoculture of Y. lipolytica. Homologous and heterologous enzymes are shown in blue and purple, respectively. Solid lines indicate the single-step reactions of FAEE synthesis in Y. lipolytica, and the multi-step reaction is shown with the dashed line. The FAEE titers in the engineered Y. lipolytica strains could be further improved when cocultured with the yeast S. cerevisiae that could provide ethanol as a substrate for FAEE production primer containing a Xma I site and then ligated into the Pml I/BamH I sites of pYLPDC1ADH1 to yield the plasmid pYLP1A1GA (Appendix 1, Figure A5). Following the construction of pYLP1A1GA, the wax  Figure A8). The primers used for gene cloning and plasmid construction are listed in Appendix 1, Table A1.

| Competent cell preparation and transformation of Y. lipolytica strains
The plasmids pYLEX1, pYLEX1PDC1, pYLEX1ADH1, pYLPDC1ADH1, pYLP1A1GA, pYLEX1MaAtfA, pYLEX1MhAtfA, pYLP1A1GAMa, and pYLP1A1GAMh were first digested with Spe I or Not I, and the resulting fragments were then integrated into the genome of Y. lipolytica Po1g by a chemical transformation process, using a protocol detailed in Appendix

| GC/MS analysis of FAEEs and ethanol produced in engineered Y. lipolytica strains
To measure the production of FAEEs and ethanol in engineered Y.
lipolytica strains, seed cultures were prepared by inoculating 5 ml of YPD medium in 50-mL culture tubes with the corresponding strains.
The cells were incubated overnight with continuous agitation. Next, 250-mL flasks containing 50 ml of YPD medium were inoculated with freshly prepared seed cultures to obtain an OD 600 of 0.05. All cultures were shaken at 225 rpm and 30°C. Samples were then collected at different time points after the start of cultivation, and a 5-ml sample of each culture was centrifuged. The growth of the en-  on ice between each round. Following cell lysis, products were extracted by vortexing for 2 min with 2 ml of n-hexane. The n-hexane extracts were then analyzed by GC/MS using an HP 7890B GC with an Agilent 5977A MSD equipped with an HP-FFAP capillary column (Agilent Technologies, Wilmington, USA). The GC oven temperature was initially held at 50°C for 1 min and then ramped to 210°C at a rate of 10°C/min and held for 5 min. The temperature was subsequently ramped at 5°C/min to 280°C and held for 5 min. Helium was used as the carrier gas, with an inlet pressure of 13.8 psi. The injector was maintained at 280°C, and the ion source temperature was set to 230°C. FAEE levels were quantified by comparing the integrated peak area of the samples with those of the corresponding standards.
Final data analysis was performed using Enhanced Data Analysis software (Agilent, USA).

| Construction and characterization of an ethanol synthetic pathway in Y. lipolytica
In microorganisms, the production of FAEEs requires two substrates: ethanol and fatty acyl-CoA (or free fatty acids). Unlike the conventional yeast S. cerevisiae, the unconventional yeast Y.

| FAEE production by cocultivation of the Y. lipolytica
Po1g::pYLP1A1GAMh strain and S. cerevisiae S288C strain As described above, the highest titer of FAEEs achieved in the engineered Y. lipolytica strain was 360.8 mg/L. This titer was much higher than that achieved in the yeast S. cerevisiae harboring the same wax ester synthase from M. hydrocarbonoclasticus DSM 8798 (6.3 mg/L) (Shi et al., 2012). Based on the results obtained here, Y. lipolytica is a more promising candidate yeast species for future applications in  Note: FAEEs produced in shake flasks with YPD media were separated and quantified by GC/MS. All values presented are the mean of three biological replicates. Sample 1 represents the Po1g::pYLP1A1GAMh strain fed with a concentration of 2% ethanol as described in Section 3.2. Sample 2 and Sample 3 represent the microbial coculture under the conditions of the second and the third coculture optimization strategies, respectively, as described in Section 3.3. BD represents "below the detection level." ethanol provided by S. cerevisiae. Many studies have reported the use of microbial coculture systems for direct production of specific target compounds, increasing the production yield, shortening the fermentation time, and/or reducing the process cost and/or realizing some specific function (He, Duan, & Liu, 2014;Hickert, Cunha-Pereira, Souza-Cruz, Rosa, & Ayub, 2013;Minty et al., 2013;Singh, Bajar, & Bishnoi, 2014;Zhou, Qiao, Edgar, & Stephanopoulos, 2015). We therefore sought to investigate whether the use of a microbial coculture system could successfully contribute to the production of FAEEs and potentially enhance FAEE fermentation performance by coculturing the S.
cerevisiae S288C strain with the engineered Y. lipolytica strain Po1g::pYLP1A1GAMh, since S. cerevisiae can metabolize glucose to produce relatively large amounts of ethanol.
To this end, the effect of different coculture designs on FAEE production was investigated using shake-flask experiments. When seed cultures of S. cerevisiae S288C and Y. lipolytica Po1g::pYLP1A1GAMh with the same initial OD 600 of 0.05 were simultaneously inoculated into one flask containing YPD medium (the original coculture strategy), we noted that the production yield of FAEEs in the coculture samples was below the detection limit after three days of cultivation, suggesting that this coculture condition is unsuitable for FAEE accumulation in this original mixed microbial culture. Another cause for this effect could be the unsuitable condition for FAEE accumulation in this original mixed microbial culture.
To validate the hypothesis, FAEE production under different coculture conditions was investigated next. The first coculture optimization strategy was as follows: S. cerevisiae S288C was first cultured separately for 24 hr, and then, a fresh overnight seed culture of Y. Interestingly, the FAEE titer obtained from the two-yeast-strain coculture system is lower than that of the single yeast strain Y. lipolytica Po1g::pYLP1A1GAMh fed with exogenous ethanol. We therefore measured ethanol concentrations in coculture systems.
Results showed that the ethanol concentrations in coculture systems (180. 5, 440.8, 500.4 mg/L for the first, second, and third coculture conditions, respectively) are much lower than 1% (v/v) which was produced in the monoculture of S. cerevisiae. One of the possible reasons for these results may be that the metabolism of two strains changed when they compete for growth resources in one system (Hettich, Sharma, Chourey, & Giannone, 2012;Khan et al., 2018;Zhou et al., 2015). However, this aspect needs to be investigated further in future work. Based on these results, we confirmed that the application of this microbial coculture system could be a viable strategy for the sustainable production of FAEEs. With the increasing knowledge, it is believed that this coculture system could be further optimized to result in a significant improvement in FAEE production.

| CON CLUS IONS
The nonconventional oleaginous yeast Y. lipolytica has previously been shown to be a competitive host organism for the production of fatty acid-derived products owing to several competitive advantages over other microbial species. In this proof-of-principle study, an efficient and eco-friendly catalytic route for the synthesis of FAEE biodiesel was established in the oleaginous yeast Y. lipolytica through metabolic engineering and the coincubation strategy adopted in this study was also shown to be very promising for future FAEE produc-

ACK N OWLED G M ENTS
We are grateful for funding support by the Natural Science

CO N FLI C T O F I NTE R E S T
None declared.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are included in this published article and the appendices.