Establishing an enzyme cascade for one-pot production of α-olefins from low-cost triglycerides and oils without exogenous H2O2 addition

Background Biological α-olefins can be used as both biofuels and high value-added chemical precursors to lubricants, polymers, and detergents. The prototypic CYP152 peroxygenase family member OleTJE from Jeotgalicoccus sp. ATCC 8456 catalyzes a single-step decarboxylation of free fatty acids (FFAs) to form α-olefins using H2O2 as a cofactor, thus attracting much attention since its discovery. To improve the productivity of α-olefins, significant efforts on protein engineering, electron donor engineering, and metabolic engineering of OleTJE have been made. However, little success has been achieved in obtaining α-olefin high-producer microorganisms due to multiple reasons such as the tight regulation of FFA biosynthesis, the difficulty of manipulating multi-enzyme metabolic network, and the poor catalytic performance of OleTJE. Results In this study, a novel enzyme cascade was developed for one-pot production of α-olefins from low-cost triacylglycerols (TAGs) and natural oils without exogenous H2O2 addition. This artificial biocatalytic route consists of a lipase (CRL, AOL or Lip2) for TAG hydrolysis to produce glycerol and free fatty acids (FFAs), an alditol oxidase (AldO) for H2O2 generation upon glycerol oxidation, and the P450 fatty acid decarboxylase OleTJE for FFA decarboxylation using H2O2 generated in situ. The multi-enzyme system was systematically optimized leading to the production of α-olefins with the conversion rates ranging from 37.2 to 68.5%. Furthermore, a reaction using lyophilized CRL/OleTJE/AldO enzymes at an optimized ratio (5 U/6 μM/30 μM) gave a promising α-olefin yield of 0.53 g/L from 1500 μM (~1 g/L) coconut oil. Conclusions The one-pot enzyme cascade was successfully established and applied to prepare high value-added α-olefins from low-cost and renewable TAGs/natural oils. This system is independent of exogenous addition of H2O2, thus not only circumventing the detrimental effect of H2O2 on the stability and activity of involved enzymes, but also lower the overall costs on the TAG-to-olefin transformation. It is anticipated that this biotransformation system will become industrially relevant in the future upon more engineering efforts based on this proof-of-concept work.


Background
Volatile geopolitical factors, depletion of petroleumbased fuels, and serious environmental concerns have been spurring the development of alternative, sustainable, and cost-effective biofuels from renewable feedstocks [1,2]. Biofuels are considered as the most promising green alternatives to petroleum-based fuels because their combustion could be carbon neutral or even negative (when produced by photosynthetic microalgae) and of near-zero air pollution [3]. Among different biofuel types, aliphatic hydrocarbons such as fatty alkanes and alkenes are regarded as ideal biofuels due to their high energy content, low hygroscopicity, and compatibility with the existing engine and distribution systems [4,5].
Chemically, a number of thermochemical methods including gasification, pyrolysis, and liquefaction, and various metal-based catalytic reactions can be utilized to convert biomass into hydrocarbon-based fuels [6,7]. However, these chemical approaches often lead to poor hydrocarbon yields, high energy consumption, and considerable side products due to simultaneous occurrence of several types of undesired reactions at high temperatures. Thus, these problems together with significant environmental concerns on hazardous chemical wastes have prompted continuous searches for green, robust, and economic biocatalytic methods to produce aliphatic hydrocarbons [8,9].
Compared to fatty alkanes, fatty alkenes especially terminal olefins (i.e., α-olefins) are more valuable products since they can be used as both biofuels and important precursors to lubricants, polymers, and detergents [10]. Nature has evolved various α-olefin-producing enzymes that use either free fatty acids (FFAs) or fatty acyl-acyl carrier proteins (acyl-ACPs) as starting materials [8]. Among them, P450 fatty acid decarboxylases have attracted the most attention in recent years because this P450 enzyme family efficiently catalyzes a single-step decarboxylation of FFAs to form α-olefins by consuming H 2 O 2 (as sole oxygen and electron donor) stoichiometrically, with varying amounts of hydroxylated fatty acids as side products.
As the first identified P450 fatty acid decarboxylase, OleT JE from Jeotgalicoccus sp. ATCC 8456 [11] has been intensively studied to understand its unique catalytic mechanism and to harness its valuable decarboxylation capacity [12][13][14][15][16][17]. Furthermore, significant efforts on protein engineering, electron donor engineering, pathway engineering, and metabolic engineering have been made for both titer improvement and product profile tuning [14,18,19]. Nonetheless, the reported highest total alkene titers of different engineered microorganisms have only reached the level of several 100 mg/L, which are far from the high cost requirement of commercial production of α-olefins. The major reasons accounting for the low α-olefin yields include the tightly regulated FFA biosynthesis from glucose, the difficulty of manipulating multi-enzyme metabolic network, and the poor catalytic behaviors of OleT JE likely owing to poor stability, low activity, substrate accessibility, and the availability of H 2 O 2 cofactor in vivo [18,20,21].
Thus, the multi-enzyme in vitro transformation of renewable oil-based feedstock into α-olefins has appeared to be a promising alternative strategy. For example, in our previous study, a two-enzyme system consisting of the lipase Tll for hydrolysis of triglycerides (TAGs) to generate FFAs and OleT JE for decarboxylation of the resulting FFAs to produce α-olefins in the presence of exogenously added H 2 O 2 was engineered; and the overall TAG-to-olefin yields reached 6.7-46.0% [19]. Matthews et al. recently engineered an OleT JE -AldO fusion enzyme, in which the alditol oxidase AldO from Streptomyces coelicolor [22] was responsible for oxidizing glycerol to glyceraldehyde and glyceric acid sequentially and generating H 2 O 2 as a co-product to drive the following OleT JE -mediated FFA decarboxylation.
To further improve the efficiency of TAG-to-olefin transformation and lower the overall costs, a number of key factors including the catalytic efficiencies, the mismatch between the fatty acyl chain length specificity of lipase and P450 fatty acid decarboxylase, the cofactor supply, and the low stability of proteins (especially for P450 enzymes) in presence of H 2 O 2 must be addressed. Thus, in this work, a novel enzyme cascade was designed and assembled for efficient one-pot production of α-olefins from low-cost triglycerides and natural oils by integrating the activities of lipase, P450 fatty acid decarboxylase, and alditol oxidase. Conceptually, a lipase hydrolyzes one molecule of TAG to release three molecules of FFAs and one molecule of glycerol; an alditol oxidase (AldO) [22] oxidizes glycerol to generate two equivalents of H 2 O 2 in situ as the cofactor of OleT JE to drive FFA decarboxylation yielding α-olefins; the shortage of one equivalent of H 2 O 2 can be resolved by exogenous addition of glycerol (Fig. 1).
In specific, three well-characterized lipases including CRL from Candida rugose [23], AOL from Aspergillus oryzae [24], and Lip2 from Yarrowia lipolytica [25] were examined for their cooperativity with OleT JE and AldO Keywords: Lipase, P450 fatty acid decarboxylase, Alditol oxidase, Bio-catalysis, Triglycerides, Natural oils, α-Olefins toward different substrates, including pure TAGs, and several kinds of natural oils. Moreover, the enzyme cascade comprising CRL, OleT JE , and AldO was optimized for the preferable coconut oil using lyophilized enzymes.

Lipase selection in terms of the TAG hydrolytic activity
The P450 fatty acid decarboxylase OleT JE favors the FFAs with the acyl chain length between C 10 and C 18 [18,26]. To identify an appropriate lipase to efficiently supply the favorable FFA substrates for OleT JE , the hydrolytic activities of three well-characterized lipases toward the pure TAGs (500 μM) with the fatty acyl chains of C 10 -C 18 were evaluated, including the two commercial lipases, namely, CRL from Candida rugose and AOL from Aspergillus oryzae, and the laboratory-prepared Lip2 from Yarrowia lipolytica (Additional file 1: Figure S1) [23][24][25]. As results, CRL released 1429.9 μM capric acid (C 10 ), 1450.8 μM lauric acid (C 12 ), 1293.6 μM myristic acid (C 14 ), 1016.8 μM palmitic acid (C 16 ), and 109.6 μM stearic acid (C 18 ) from their corresponding TAGs, corresponding to the conversion rates of 95.3%, 96.7%, 86.2%, 67.8%, and 7.3%, respectively (Fig. 2). AOL exhibited analogous hydrolytic activities and a similar substrate preference profile to CRL with tricaprin (C 10 ) as the optimal substrate (92.3% conversion rate). However, Fig. 1 The schematic enzyme cascade for one-pot production of α-olefins from low-cost triglycerides Fig. 2 Released FFAs from 500 μM of different pure TAGs by CRL, AOL, and Lip2. In a standard assay, 5 U lipase and 500 μM TAG were co-incubated in a 200-μL reaction system at 30 °C for 6 h. Error bars represent standard deviations derived from at least two independent experiments. Statistical analysis was performed using a Student's t test (one-tailed; *P < 0.05, **P < 0.01, ns: P > 0.05, no significant; two-sample unequal variance) (same below) the laboratory-prepared Lip2 showed significantly lower activities than both AOL and CRL, with the highest yield of 58.6% against tricaprin (C 10 ). Considering the overall hydrolytic activity and the factor that OleT JE prefers C 10 -C 14 FFAs [26], CRL and AOL were selected for the following enzyme cascade setup.

Effect of H 2 O 2 concentration on the OleT JE activity
OleT JE , a CYP152 peroxygenase family member, utilizes H 2 O 2 as a cofactor (or co-substrate) to support the unique oxidative decarboxylation reactions that convert C n (n = 4-22) chain length FFAs into C n-1 α-olefins and release one molecule of CO 2 [26]. In principle, a higher concentration of H 2 O 2 co-substrate presumably enhances the reaction rate kinetically. However, excessive addition of H 2 O 2 would oxidatively inactivate enzymes via radical mechanisms in general [27]. Therefore, it is important to balance the two opposite sides of H 2 O 2 by identifying an optimal working concentration of H 2 O 2 for OleT JE 's activity.
In specific, 0-5000 μM of H 2 O 2 was added into each individual OleT JE reaction mixture containing 1 μM P450 enzyme and 500 μM lauric acid. As expected, the conversion of lauric acid increased in proportion to the amount of H 2 O 2 at the low concentration range (0-550 μM); then, the substrate conversion rates declined with the increase of H 2 O 2 concentrations between 1000 and 5000 μM (Fig. 3). Taken together, direct addition of H 2 O 2 to a high concentration appeared not to be an effective way to maintain a high activity of OleT JE .

Determination of the optimal ratio between OleT JE and AldO
Although the OleT JE -AldO fusion enzyme developed by Munro and co-workers could enable efficient channeling of H 2 O 2 between these two enzyme active sites [28], the fusion nature of OleT JE and AldO fixed the molar ratio of these two functional domains to be 1:1, which might not be an optimal ratio. Moreover, the chimeric protein might cost its high expression level.
Thus, in this study, OleT JE and AldO were separately expressed (Additional file 1: Figure S2), by which their ratio could be conveniently adjusted. Using 500 μM lauric acid as substrate and 0.01% v/v (~ 1000 μM) glycerol loading, a select number of different OleT JE to AldO (OA) ratios were examined. It was revealed that the best OA ratio turned out to be 1:10 (or 1:5), at which the substrate conversion rate and the alkene production ratio were 99.7% (93.6%) and 64.9% (61.5%) (Fig. 4), respectively. The differences between lauric acid (C 12 ) conversion and undecene (C 11 ) production were due to the unquantified side products, namely, different hydroxylated fatty acids. In consideration of both the overall catalytic efficiency and cost-effectiveness, the OA ratio of 1:5 with the lower amount of AldO was selected for the following experiments.

Determination of the optimal glycerol concentration
As the substrate of AldO, glycerol is presumably a key factor in the OleT JE /AldO reaction system. Thus, the optimal glycerol concentration for OleT JE 's activity in the range of 0.01-10% was determined with the exogenous addition of H 2 O 2 as a control. For 500 μM lauric acid, 0.01% glycerol loading resulted in approximately 100% increase in both substrate conversion rate and total turnover number (TTN) compared to the OleT JE reaction supported by 500 μM H 2 O 2 ( Table 1). These results clearly demonstrated the advantage of continuous release of H 2 O 2 over the direct addition of H 2 O 2 at once.
At a higher concentration of lauric acid (1000 μM), the AldO-based in situ H 2 O 2 -generating system showed more significant improvements under all tested glycerol concentrations relative to 1000 μM H 2 O 2 ( Table 1). In the presence of 0.01-10% glycerol, the TTNs ranged  from 273 to 695. When the glycerol concentration was higher, the decarboxylation activity was better. Surprisingly, at the same 0.01% glycerol loading, the reaction with 1000 μM substrate gave a much lower TTN than that with 500 μM substrate. This result suggested that the activity of AldO might be inhibited by the high concentration of hydrophobic products or substrates at a relatively low glycerol concentration.
Evidently, 1 μM OleT JE and 5 μM AldO were not sufficient for complete decarboxylation of 1000 μM lauric acid even in the presence of 10% glycerol. To decarboxylate 1500 μM lauric acid that can be maximally released from 500 μM trilaurin, higher concentrations of OleT JE and AldO were required to be used with an appropriate amount of glycerol. Thus, the reaction system of 3 μM OleT JE , 15 μM AldO, and 1.5% glycerol was proposed and tested, by which we assumed that a triplicated TTN could be achieved by higher enzyme concentrations.

Production of α-olefins from natural oils
Next, the activities of the CRL/OleT JE /AldO and AOL/ OleT JE /AldO systems were examined toward a number of natural oils including coconut oil and palm oil featuring saturated fats, peanut oil and olive oil containing monounsaturated fats, and soybean oil representing  Compared to pure TAGs, these natural oils are more readily available and industrially relevant [29].
In the CRL hydrolytic system, 1304.9, 1108.1, 708.1, 998.8, and 1301.9 μM of total FFAs were produced from 500 μM of coconut oil, palm oil, soybean oil, peanut oil, or olive oil, corresponding to 86.9%, 73.9%, 47.2%, 66.6%, and 86.8% of TAG-to-FFA conversions, respectively. Regarding the released FFA profiles of different natural oils (Additional file 1: Table S2), lauric acid, palmitic acid, and linoleic acid were the main hydrolytic products of coconut oil, palm oil, and soybean oil, respectively, accounting for 47.5%, 41.3%, and 54.7% of total FFAs. For both peanut oil and olive oil, oleic acid was the major released FFA species, reaching 45.6% and 71.9% of total FFAs, respectively. Moreover, the hydrolytic activities (Fig. 6a, b) and FFA profiles of the AOL hydrolytic system (Additional file 1: Table S3) toward the same group of natural oils were highly similar to those of the CRL hydrolytic system. Notably, coconut oil turned out to be the preferred substrate in both hydrolytic systems (Fig. 6a, b).

Optimization of the coconut oil bioconversion system using the lyophilized CRL/OleT JE /AldO enzymes
Coconut oil among the tested natural oils appeared to be the best feedstock for α-olefin production (Fig. 6c,  d), from which 1-undecene derived from lauric acid, the OleT JE -favored substrate, was the predominant product. Thus, the CRL/OleT JE /AldO system using coconut oil as the starting material to produce biological α-olefins was chosen for further optimization. To make the system more practical and cost-effective, lyophilized enzymes were used. Notably, the mixed lyophilized CRL, OleT JE , and AldO demonstrated comparable olefins-producing activity as the freshly purified proteins at the same levels (Additional file 1: Figure S3).
To examine the application potential of the one-pot reaction system consisting of CRL, OleT JE , AldO, coconut oil, and glycerol, the coconut oil concentration was increased from 500 to 1500 μM (~ 1 g/L). Under the higher substrate concentration, we re-optimized the amount of each individual component. As a result, 5 U CRL was shown to be able to hydrolyze 1500 μM coconut oil almost to the theoretical maximum yield of 4500 μM, and further increase of lipase amount had no significant beneficial effect on FFA production (Additional file 1: Figure S4a). However, when 5 U CRL, 3 μM OleT JE , and 15 μM AldO were reacted with 1500 μM coconut oil and 1.5% glycerol for 6 h, the total alkene yield was 14.3% lower than that of the previous 500 μM substrate system, indicating this reaction system needed to be further optimized.
Since 5 U CRL (among 2.5, 5, and 10 U) gave the highest alkene production (Additional file 1: Figure S4b), we subsequently optimized other contents in the TAG-toolefin reaction system containing 5 U CRL. As expected, higher concentrations of OleT JE and AldO resulted in improved alkene production under the same glycerol content (Table 2). Moreover, increase of glycerol content significantly improved the total yield of α-olefins. These results suggested that the H 2 O 2 releasing upon glycerol oxidation might be the major limit of the overall efficiency of the alkene-producing system. Thus, the highest alkene yield achieved by this enzyme cascade system was 52.6% (0.53 g/L) at a 20% glycerol loading in a 200-μL reaction ( Table 2). Of note, in a 10-mL reaction, this system delivered a similar α-olefin yield of 49.1% (0.49 g/L) ( Table 2).

Discussion
With continuous discovery and characterization of novel hydrocarbon biosynthetic systems, a growing number of α-olefin-producing enzymes have been reported, including P450 fatty acid decarboxylases such as OleT JE and CYP-Sm46 [11,30], the non-heme iron-dependent decarboxylase UndA [31], and the membrane-bound desaturase-like UndB [32]. Although these α-olefin synthases can use different chain length FFAs as substrates, when they were coupled to the fatty acid biosynthetic systems of engineered E. coli or Saccharomyces cerevisiae strains, the total alkene titers turned out to be very low with the highest reported yield (97.6 mg/L) only accounting for about 1% of theoretical maximum yield (3% glucose) [18,20,32,33]. Mechanistically, the complexity and tight regulation of fatty acid metabolic network in vivo could be one of the major reasons for the cost-prohibitive α-olefin yields [9].
To overcome this challenge, we previously engineered an in vitro tandem reaction system, in which the lipase Tll and OleT JE cooperate to transform TAGs to α-olefins with exogenous addition of H 2 O 2 [19]. Subsequently, Li et al. developed a genetically encoded synthetic selfassembled Tll/OleT JE complex for bioproduction of fatty alkenes, wherein three non-catalytic modules including cohesion, dockerin, and CBM were used to control enzyme ratio, position, reusability, and stability [34]. However, the α-olefin yields of both approaches were low likely due to the detrimental effect of H 2 O 2 for enzymes.
Although CYP152 peroxygenases are more H 2 O 2 -tolerant than most of P450 monooxygenases, a high concentration of H 2 O 2 could still result in the loss of catalytic activity of these P450 peroxygenases, which was observed in the process of myristic acid hydroxylation catalyzed by the peroxygenase P450 CLA from Clostridium actetobutylicum [28,35]. Despite the high activity (up to 200 min −1 ), substantial inactivation of P450 CLA occurred within 2-4 min when 200 μM H 2 O 2 was added at once. Our results also showed that OleT JE was plagued by the H 2 O 2 concentrations greater than 1000 μM (Fig. 3). To resolve this problem, a growing number of approaches have been developed to elegantly control the H 2 O 2 supply for reduction of the P450 inactivation [18,[36][37][38]. For example, a light-driven in situ H 2 O 2 -generating system employed flavin adenine mononucleotide (FMN) as a photocatalyst to reduce O 2 to H 2 O 2 with ethylenediaminetetraacetic acid (EDTA) as electron donor, which well supported OleT JE to react with stearic acid. However, the conversion ratios of lauric acid and myristic acid were not satisfactory [38]. Moreover, redox partner engineering by making OleT JE -reductase fusion protein or constructing alternative OleT JE reaction systems using separate redox partners has been proven as an effective method to reconstitute the decarboxylation activity of OleT JE toward C 4 -C 22 FFAs in the presence of NADPH regeneration system (e.g., formate/formate dehydrogenase, glucose/glucose oxidase) [18,36,37]. However, the requirement of expensive materials (FMN or NADPH) in both strategies and the complexity in the latter system consisting of four redox proteins would hinder their further application.
In this study, a novel enzyme cascade system was established to convert TAGs and natural oils into α-olefins in vitro (Fig. 1). Using the low-cost and renewable TAG/ oil feedstock, the FFAs hydrolyzed off by lipases were well accepted by the downstream P450 fatty acid decarboxylase OleT JE . The required H 2 O 2 cofactor of OleT JE was continuously supplied in situ by the alditol oxidase AldO using the glycerol substrate derived from TAG hydrolysis, through which the H 2 O 2 -induced enzyme inactivation was significantly attenuated and the highest yield of α-olefins reached 0.53 g/L from 1500 μM coconut oil.
Although the enzyme cascade system provides a new paradigm for catalytically efficient and cost-effective biotransformation from TAGs/oils to α-olefins, the highest yield is still far from the stringent cost requirement of industrial production of α-olefins. There remain a number of significant problems to be overcome such as enzyme stability and solubility of hydrophobic substrates/products. It is anticipated that more enzyme engineering, process optimization, and other interdisciplinary approaches are required to make this enzyme cascade more industrially relevant.

Conclusions
In this study, we established a new enzyme cascade independent of exogenous addition of H 2 O 2 , which is capable of efficiently converting a range of TAGs/natural oils into α-olefins. This one-pot biocatalytic system consisting of CRL (for TAG hydrolysis to provide FFAs and glycerol), AldO (for in situ H 2 O 2 generation upon glycerol oxidation), and OleT JE (for FFA decarboxylation using H 2 O 2 as cofactor) was able to achieve a 68.5% total alkene yield from 500 μM coconut oil. Using the lyophilized enzymes, ~ 0.5 g/L of α-olefins were produced from the favorable feedstock coconut oil (1500 μM) upon some reaction optimization. Altogether, the three-enzyme cascade provides a new strategy for producing high value-added α-olefins from low-cost and renewable oils, demonstrating promising application potential.

Materials
The strains of Escherichia coli DH5α and BL21(DE3) and the plasmid pET28(b) were preserved by our laboratory.
All antibiotics and chemicals including TAGs, FFAs, and α-olefins were obtained from Tokyo Chemical Industry (TCI) (Shanghai, China), Solarbio (Beijing, China), Sigma Aldrich (St. Louis, MO, USA), and Thermo Scientific (Shanghai, China). Soybean oil, peanut oil, and olive oil were purchased from local market. Coconut oil and palm oil were obtained from Orifera (Malaysia) and Pythonbio (Guangzhou, China), respectively. The 10 × QuickRun ™ Fast Running Buffer and FlexiRun ™ premixed gel solution for SDS-PAGE analysis were obtained from MDBio (Xinbei, China). Purification of DNA fragments was performed using a MonPure ™ Gel & PCR Clean Kit from Monad (Wuhan, China). Ni-NTA resin used for protein purification was purchased from Sangon Biotech (Shanghai, China). PD-10 desalting columns were supplied by GE Healthcare (Piscataway, NJ, USA). Millipore Amicon Ultra centrifugal filters were bought from Millipore (Billerica, MA, USA).

Molecular cloning and protein purification
The gene encoding AldO from Streptomyces coelicolor A3(2) (GenBank accession number: NC_003888.3) was codon-optimized and synthesized by Qinglan (Yixing, China), and then cloned into the vector pET28b via the NdeI/XhoI restriction sites for expression of the N-terminal His 6 -tagged recombinant proteins. The sequences of primers used in this study are listed in Additional file 1: Table S1. All cloned sequences were confirmed by DNA sequencing at Sangon Biotech (Shanghai, China), and then used to transform E. coli BL21 (DE3) for protein expression. The plasmid pET28b-oleT JE for recombinant OleT JE expression was constructed by this laboratory previously [18].
The E. coli BL21 (DE3) cells carrying a certain recombinant expression vector were grown at 37 °C for 12 h with shaking at 220 rpm and then used as seed cultures to inoculate (1:100 ratio) a modified Terrific Broth medium containing a rare salt solution [18]. Cells were grown at 37 °C for 3-4 h until the optical density at 600 nm (OD 600 ) reached 0.8 to 1.0, to which 0.2 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) was added. For P450 expression, 0.5 mM δ-aminolevulinic acid (5-ALA) and 1 mM thiamine were supplemented. Afterward, the cultivation continued for another 24 h at 18 °C for better protein folding [18,26]. The cells were harvested (6000 × g, 4 °C, 10 min) and stored at − 80 °C for later use.
Purification of the His-tagged OleT JE was performed by following our previous procedure with some minor modifications [26]. Of note, the elution buffer (pH 7.4, 50 mM NaH 2 PO 4 , 500 mM NaCl) used in this study was glycerol free because glycerol is a reactant in the enzyme cascade reaction. Purification of the His-tagged AldO was carried out as described by Liu et al. with some modifications [18]. In brief, the cell pellets stored at − 80 °C were taken out to thaw at room temperature, then all the following steps were performed at 4 °C. The cells were re-suspended in 50 mL lysis buffer (pH 8.0, 50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole) through vortexing and then disrupted by ultra-sonication (SCIENTZ JY98-IIIDN, Ningbo; on 5 s/off 5 s for 30 min). Cell-free lysate was obtained by high-speed centrifugation (Eppendorf Centrifuge 5810R, Germany) at 10,000 × g for 1 h at 4 °C, to which 2 mL Ni-NTA resin slurry was added and mixed gently at 4 °C for 2-3 h. The mixture was loaded onto an empty column and washed with about 200 mL wash buffer (pH 8.0, 50 mM NaH 2 PO 4 , 300 mM NaCl, 20 mM imidazole) until no protein was detected in flow-through. The His-tagged proteins bound to Ni-NTA resin were eluted with 5-10 mL elution buffer (pH 8.0, 50 mM NaH 2 PO 4 , 500 mM NaCl, 250 mM imidazole). The eluents were concentrated with an Amicon Ultra centrifugal filter (30 kDa cutoff ) and then bufferexchanged into storage buffer (pH 8.0, 50 mM Tris-HCl). The solution containing purified proteins in aliquots were flash-frozen by liquid nitrogen for later use.
For preparation of the lipase Lip2, a single colony of Yarrowia lipolytica strain YLY [39] was inoculated into Yeast Extract Peptone Dextrose (YPD) medium for 12 h at 28 °C, and then used as seed cultures to inoculate (1:40 ratio) a fermentation medium [10 g yeast extract, 20 g tryptone, 10 g sucrose, and 6.7 g yeast nitrogen base (without amino acids) per liter], and shaking cultured at 28 °C, 220 rpm for 72 h. The supernatant containing secreted Lip2 was obtained by removing the cells by centrifugation (4 °C, 6000 × g for 10 min). The supernatant was concentrated with an Amicon Ultra centrifugal filter (30 kDa cutoff ) and then buffer-exchanged into a Tris-HCl buffer (50 mM, pH = 8.0). The commercial lipases CRL and AOL were purchased from Sigma Aldrich (St. Louis, MO, USA).

Enzyme concentration determination
Analysis of the UV-visible spectroscopic properties for OleT JE was carried out as described previously [26]. The P450 protein concentration was calculated based on its reduced CO-bound difference spectrum using the reduced differential extinction coefficient ε 450-490 of 91,000 M −1 cm −1 [18]. The concentration of AldO was determined at 452 nm with the reported extinction coefficient of 12,500 M −1 cm −1 [22]. The hydrolysis activity of lipase (U) was determined by the classical basic titration method using the olive oil emulsion. One hydrolysis activity unit (U) of lipase is defined as the amount of enzyme required to produce 1 μM free fatty acid in 1 min under the specific reaction conditions [40].

In vitro enzymatic assay
Since P450 OleT JE was characterized as a moderate halophilic protein requiring salt solution to maintain its stability and hence activity [26], all the reactions of OleT JE were carried out in a buffer containing 500 mM NaCl. Although the high concentration of NaCl might negatively affect the activities of lipase and alditol oxidase, the compromise was made for the best performance of OleT JE .
All reactions were quenched by adding 20 μL of 10 M HCl, then heptadecanoic acid (C 17 ) was added as internal standard and the mixture was extracted by 150 μL ethyl acetate. The organic phase was analyzed by gas chromatography (GC) as described below. For detection of 1-heptene (C 7 ) product generated from coconut oil in the enzyme cascade reactions, 1.5 mL polytetrafluorethylene (PTFE) septum-sealed glass bottles were used for the 200-μL reactions containing 3 μM OleT JE , 15 μM AldO, and 1.5% glycerol. The reactions were incubated at 30 °C for 6 h with shaking at 200 rpm. Then, the reactions were placed at 4 °C for 12 h to stop reactions prior to heating at 40 °C for 20 min for headspace sampling using a gas-tight Hamilton syringe for GC-MS analysis. Different concentrations of the authentic 1-heptene standard incubated under the same conditions of reactions were analyzed using the same GC-MS method to obtain the standard curve.

Analytical methods
The hydrocarbon and fatty acid samples were analyzed by the methods modified from Guan et al. [42]. The Agilent 7890B gas chromatograph equipped with a capillary