Production of Long-Chain α,ω-Dicarboxylic Acids by Engineered

Sep 3, 2015 - Genome Editing Tools for Escherichia coli and Their Application in Metabolic Engineering and Synthetic Biology. Chandran Sathesh-Prabu ...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/JAFC

Production of Long-Chain α,ω-Dicarboxylic Acids by Engineered Escherichia coli from Renewable Fatty Acids and Plant Oils Chandran Sathesh-Prabu† and Sung Kuk Lee*,†,‡ School of Energy and Chemical Engineering, and ‡School of Life Sciences, Ulsan National Institute of Science and Technology (UNIST), Ulsan 689-798, Republic of Korea

J. Agric. Food Chem. Downloaded from pubs.acs.org by CENTRAL MICHIGAN UNIV on 09/14/15. For personal use only.



ABSTRACT: Long-chain α,ω-dicarboxylic acids (LDCAs, ≥C12) are widely used as a raw material for preparing various commodities and polymers. In this study, a CYP450-monooxygenase-mediated ω-oxidation pathway system with high ωregioselectivity was heterologously expressed in Escherichia coli to produce DCAs from fatty acids. The resulting engineered E. coli produced a maximum of 41 mg/L of C12 DCA and 163 mg/L of C14 DCA from fatty acids (1 g/L), following 20 h of whole cell biotransformation. Addition of a heme precursor and the hydroxyl radical scavenger, thiourea, increased product concentration (159 mg/L of C12 DCA and 410 mg/L of C14 DCA) in a shorter culture duration than that of the corresponding controls. DCAs of various chain lengths were synthesized from coconut oil hydrolysate using the engineered E. coli. This novel synthetic biocatalytic system could be applied to produce high value DCAs in a cost-effective manner from renewable plant oils. KEYWORDS: α,ω-dicarboxylic acids, CYP450s, monooxygenase, ω-oxidation pathway, fatty acids



CYP102).9,10 Terminal (ω) oxidation, oxidation of terminal methyl groups of alkanes or fatty acids, is observed in different yeast species, including Candida tropicalis, C. maltosa, and Yarrowia lipolytica.11−13 In the existing platforms for DCA production, the ω-oxidation pathway is overexpressed in these yeast species along with the alteration of the fatty acid degradation pathway (β-oxidation) through deletion of genes involved in fatty acid or DCA catabolism; the imported alkanes and some fatty acids could be oxidized by ω-oxidation, and the inability to oxidize the carboxylic acids by the β-oxidation pathway causes accumulation of DCAs.11,12,14 Therefore, ωDCAs could be effectively produced from fatty acids by the proficient actions of three ω-oxidation pathway genes in a fatty acid degradation-deficient system. Recently, Malca et al.15 identified a CYP450 (CYP153A) from Marinobacter aquaeolei as highly regioselective for fatty acid ω-hydroxylation (>91% at terminal position) with a broad substrate range (saturated C10−C20 and monounsaturated C16:1−C18:1) and high substrate conversion activity (63− 93%). Scheps et al.16 recently established a fusion protein consisting of the monooxygenase CYP153A from M. aquaeolei and the reductase domain of P450 BM3 from Bacillus megaterium to increase the hydroxylation efficiency. This fusion was used for the production of the C12-ω-hydroxy fatty acid from C12 FA with high regioselectivity (>95%) for the terminal (ω) position. The major tailbacks of the CYP450-based whole cell biotransformations include inefficiency of substrate (FAs) uptake5,16 and accumulation of toxic H2O2 that resulted by some unproductive catalytic reactions occurring in the monooxygenase enzyme complex.17 It was found that the outer membrane protein, AlkL, of Pseudomonas putida can

INTRODUCTION Long-chain α,ω-dicarboxylic acids (LDCAs, ≥C12) are widely used as a raw material for preparing various products, such as nylon and other polyamides, polyesters, perfumes, adhesives, and high-quality lubricants.1,2 LDCAs are almost exclusively produced from petrochemical sources by chemical conversion processes that have a number of disadvantages and limitations. These include limitations in the range of carbon chain lengths, use of multistep conversion processes, dependence on nonrenewable petrochemical feedstock, and generation of unwanted and hazardous byproducts.1,3 Biotechnology offers an innovative solution for overcoming these limitations. Fatty acids (FAs) are major constituents of lipids and one of the most abundant renewable resources found in nature. Several studies have previously proposed the production of renewable fuels and industrial chemicals from fatty acids by Escherichia coli,4,5 suggesting that fatty acids could become a sustainable feedstock for industrial production.6 The production of α,ω-dicarboxylic acids (Figure 1) involves the terminal (ω) oxidation of fatty acids, catalyzed by a hydroxylase complex composed of cytochrome P450 monooxygenase (CYP450) and an NADPH:cytochrome P450 oxidoreductase (NCP) to form ω-hydroxy fatty acids. Subsequent oxidation of hydroxy fatty acids into oxo fatty acids or fatty acid aldehydes is catalyzed by alcohol dehydrogenase. Finally, the fatty aldehyde is oxidized by aldehyde dehydrogenase to the corresponding dicarboxylic acid (DCA). Hydroxylation of fatty acids by CYP450 and NCP is the rate-limiting step and is therefore a critical stage in the production of α,ω-DCA, as the CYP enzymes show broad substrate specificity as well as high regioselectivity.7−9 CYP450s differ from each other in the position at which CYP450s hydroxylate substrates. Hydroxylation may occur near the carboxy-terminal (α- or β-hydroxylases: e.g., CYP152), terminal methyl group (ω-hydroxylases: e.g., CYP52A, CYP153A), or subterminal methyl group (ω1,2,3-ydroxylases: e.g., © XXXX American Chemical Society

Received: August 5, 2015 Revised: September 1, 2015 Accepted: September 3, 2015

A

DOI: 10.1021/acs.jafc.5b03833 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

J. Agric. Food Chem. Downloaded from pubs.acs.org by CENTRAL MICHIGAN UNIV on 09/14/15. For personal use only.

Journal of Agricultural and Food Chemistry

for genetic recombination studies because of its rapid growth and potential for high-density cultivation on inexpensive substrates. Additionally, its well-characterized genetics and the availability of an increasingly large number of cloning vectors and mutant host strains make E. coli an attractive model for genetic studies.20 E. coli has been demonstrated as a suitable host for heterologous monooxygenase related reactions, including CYP153A, and exhibited higher catalytic activities and product yields when compared to other hosts, such as Pseudomonas.16,21,22 Moreover, using E. coli as a host for heterologous expression of CYP450s is advantageous, as E. coli lacks endogenous CYP450s that could interfere with heterologous P450 expression. The present study exploits the advantages of the E. coli model system to demonstrate a novel biotechnology system that efficiently produces long-chain α,ωdicarboxylic acids (C12 and C14) from fatty acids by introduction of the ω-oxidation pathway into E. coli.



MATERIALS AND METHODS

Microbial Strains and Plasmids. The wild-type E. coli MG1655 strain (laboratory stock) was used as the parental strain for all genetic modifications, including gene knockouts and introduction of expression constructs. The E. coli DH10B strain (laboratory stock) was used for cloning. Marinobacter aquaeolei (DSMZ, DSM 11845), B. megaterium (Korean Culture Center of Microorganisms, KCCM 11745), Saccharomyces cerevisiae (Korean Collection for Type Cultures, KCTC 7296), and C. tropicalis (American Type Culture Collection, ATCC 20336) were procured from the indicated cell culture collection centers. The Biobrick plasmids, pBbA6C and pBbE6K (A = p15A replicon; E = colE1 replicon; 6 = PLlacO promoter; C = chloramphenicol resistance; and K = kanamycin resistance), were used for expression constructs.23 Chemicals, Enzymes, and Culture Conditions. All chemicals, including fatty acids and DCAs, were purchased from Sigma-Aldrich (St. Louis, MO). Restriction enzymes (New England Biolabs, Ipswich, MA), DNA ligase (New England Biolabs, Ipswich, MA), and Phusion high-Fidelity DNA polymerase (Thermo Fischer Scientific, Waltham, MA) were used for cloning and plasmid construction. Oligonucleotides were synthesized by Macrogen (Korea). Microbial cells were grown at 37 °C (30 °C for strains carrying temperature-sensitive plasmids) in Luria−Bertani (LB) broth supplemented with suitable antibiotics (kanamycin at 50 μg/mL, ampicillin at 100 μg/mL, or chloramphenicol at 30 μg/mL).

Figure 1. ω-Oxidation pathway. Biotransformation of fatty acids by the sequential oxidation of the terminal methyl group yields the dicarboxylic acid product. Protein ID (Genbank): CYP153A (ABM17701.01); NADPH:cytochrome P450 oxidoreductase (ACZ37122.1); alcohol dehydrogenase (AAA344101.1); and aldehyde dehydrogenase (U.S. Patent no. WO 2013006730 A2).

improve the fatty acid uptake, resulting in high productivity.18,19 In addition to this, to obtain the maximum yield, the toxicity of accumulated H2O2 has to be reduced.6,16 To the best of our knowledge, to date there are no reports of the production of α,ω-DCAs directly from fatty acid substrates, through utilization of a CYP450-dependent ω-oxidation pathway in Escherichia coli, a highly versatile organism that has been widely used for heterologous protein production and Table 1. Primers Used for PCR and Cloninga primer name FadR_del_F FadR_del_R FadE_del_F FadE_del_R Maq_F Maq_R NCP_F NCP_R AlkL_F AlkL_R ADH1_F ADH1_R ALD5_F ALD5_R ALD_F a

primer sequence TCTGGTATGATGAGTCCAACTTTGTTTTGCTGTGTTATGGAAATCTCACTGTGTAGGCTGGAGCTGCTTC AACAACAAAAAACCCCTCGTTTGAGGGGTTTGCTCTTTAAACGGAAGGGAATTCCGGGGATCCGTCGACC CCATATCATCACAAGTGGTCAGACCTCCTACAAGTAAGGGGCTTTTCGTTGTGTAGGCTGGAGCTGCTTC TTACGCGGCTTCAACTTTCCGCACTTTCTCCGGCAACTTTACCGGCTTCGATTCCGGGGATCCGTCGACC ATAGATCTTTTAAGAAGGAGATATACATATGCCAACACTGCCCAGAAC CTGTTCAGTGCTAGGTGAAGGAATGCTGCCGCCGCTGCCGCCGCTGCCGCC ACTGTTCGGTGTCAGTTTGACCATCAACCTGGAA GCGGCAGCATTCCTTCACCTAGCACTGAACAGTCTGCTAAAAAAGTACGCA AAAAGGCAGAAAACGCTATAATACGCCGCTGC ATTCTCGAGCGATATCGATCGTTATTACCCAGCCCACACGTCTTTTG GGTCGATCGTTTAAGAAGGAGATATACAT ATGAGTTTTTCTAATTATAAAGTAATCGCGATGCCG ATCTCGAGTTATTAGAAAACATATGACGCACCAA TTTAAGAAGGAGATATACACATGTCTATCCCAGAAACTCA CTTACTCGAGTTATTTAGAAGTGTCAACAACGT ATACATATGTCTTTGCCAGTCGTCACCAA GTGTATATCTCCTTCTTAAAAGATCCTTAAGTGAGCTTAATTCTAACAG AGGCCTCGAGAATTGTGAGCGGATAACAATTGAC

Underline formatting indicates restriction sites. B

DOI: 10.1021/acs.jafc.5b03833 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

J. Agric. Food Chem. Downloaded from pubs.acs.org by CENTRAL MICHIGAN UNIV on 09/14/15. For personal use only.

Journal of Agricultural and Food Chemistry Strain Construction. Knockout of fadR and fadE in MG 1655 was performed using the lambda Red and FLP-mediated site-specific recombination system as described previously.24 Primers used for gene deletion and cloning are shown in Table 1. The monooxygenase (CYP153A), NADPH:cytochrome P450 oxidoreductase (NCP), alcohol dehydrogenase, and aldehyde dehydrogenase were amplified using Phusion high-Fidelity DNA polymerase from the genomic DNA of M. aquaeolei (DSM 11845), B. megaterium (KCCM 11745), S. cerevisiae (KCTC 7296), and C. tropicalis (ATCC 20336). First, the fusion construct components of CYP153A from M. aquaeolei and NCP from B. megaterium were amplified using the primers Maq_F and Maq_R and NCP_F and NCP_R, respectively, and these two genes were linked by overlap extension PCR with a 3xGGS tandem region in the linker between CYP and NCP.16 The pBbA6C-FP plasmid was constructed by cloning of the fusion construct (CYP-NCP) into the BglII and XhoI site of the pBbA6C plasmid. The Pseudomonas putida-alkane transporter gene (alkL) was synthesized and cloned into pGEM-B1 by Bioneer (Korea). AlkL was then amplified using primers AlkL_F and AlkL_R from the pGEM-B1 plasmid. Subsequently, PvuI- and XhoI-digested pBbA6C-FP and alkL were ligated to generate the pBbA6C-FP/AlkL plasmid. Alcohol dehydrogenase (ADH1) was amplified from S. cerevisiae (KCTC 7296) using primers ADH1_F and ADH1_R, and aldehyde dehydrogenase (ALD5) was amplified from C. tropicalis (ATCC 20336) using primers ALD5_F and ALD5_R. These two genes were linked by overlap extension PCR and cloned into the NdeI/XhoI site of the pBbE6K plasmid to study the activities of ADH and ALD on the first intermediate of ω-oxidation pathway, ω-hydroxy fatty acid. Subsequently, primers ALD_F and ADH1_R were used to amplify ADH1 and ALD5 from the pBbE6K plasmid; these products were then digested with XhoI. Finally, the XhoI digested pBbA6C-FP/AlkL plasmid and the ALD-ADH PCR product were ligated to yield the final construct, the pBbA6C-ω plasmid. For biotransformation studies, pBbA6C-ω was transformed into MG1655:ΔfadE or MG1655:ΔfadR:ΔfadE, which were designated as Eω and REω, respectively. Protein Expression and SDS-PAGE Analysis. Protein expression in recombinant cells, grown until OD600 reached 0.5, was induced with 0.1 mM of isopropyl β-D-1-thiogalactopyranoside (IPTG). After incubation at 25 °C for 16 h, the cells were collected, sample loading buffer was added, and the mixture was heated at 95 °C for 20 min. The whole cell lysate samples were run on 10% sodium dodecyl sulfatepolyacrylamide gels. Protein bands were visualized by staining with Coomassie Brilliant Blue R-250. Whole Cell Biotransformation. A single colony of recombinant E. coli cells (Eω and REω strains) was selected and grown overnight in 5 mL of Luria−Bertani (LB) medium at 37 °C with rotation at 200 rpm. Five hundred microliters of the culture was inoculated into 50 mL of LB medium and grown overnight at 37 °C with rotation at 200 rpm. Subsequently, this culture was inoculated (1:100) into 1 L of Luria−Bertani (LB) medium, and expression of target genes was induced with 0.1 mM IPTG in addition to 0.25 mM of 5aminolevulinic acid (ALA) at an OD600 of 0.5−0.6. Following the addition of IPTG and ALA, temperature was reduced to 25 °C to facilitate soluble expression of the target genes, with rotation at 200 rpm for 20 h. The recombinant cells were harvested by centrifugation at 12 000g for 15 min at 4 °C, washed twice with potassium phosphate buffer (0.1 M, pH 7.4), and used as biocatalysts in the biotransformation studies. The biotransformation was conducted in 250 mL shake flasks by adding a fatty acid (C12 or C14) substrate stock solution containing 20% of FA in DMSO to a final concentration of 1g/L into 0.1 M potassium phosphate buffer (pH 7.4) containing the following: 50 gcww/L of the induced resting cells, 0.5% Tween 80, 1X filter sterilized trace element solution25 (g/L: 2.4 g of FeCl3·6H2O, 0.3 g of CoCl2·H2O, 0.15 g of CuCl2·2H2O, 0.3 g of ZnCl2, 0.3 g of Na2MO4·2H2O, 0.075 g of H3BO3, and 0.495 g of MnCl2·4H2O), and 30 μg/mL of chloramphenicol. The biotransformation was carried out in the presence and absence of cosubstrates (1% (w/v) glycerol and 0.4% (w/v) D-glucose). The biotransformations were conducted in a shaking incubator (30 °C and 200 rpm), and samples were collected at

different time points for product quantification and identification by (gas chromatography) GC. Effects of Temperature on DCA Production. To examine the effect of temperature on DCA production, the biotransformation was conducted at 25, 30, and 37 °C using the engineered E. coli strain, REω. Biotransformations were conducted in a shaking incubator (200 rpm), samples were collected at different time points, and products were quantified using GC. Biotransformations in Mini-Bioreactor. Because REω performed better in shake flask conditions than did Eω, REω was used for further biotransformation studies that were carried out in a homemade, mini-bioreactor (250 mL capacity) that had precise monitoring and control systems for pH, temperature, and agitation speed. The biotransformation medium and conditions were maintained as for the shake flask experiments. Agitation speed was set at 1200 rpm. The biotransformations were conducted at 30 °C, and samples were collected every 2 h for product quantification by GC. Effects of 5-Aminolevulinic Acid and Thiourea on DCA Production. To determine the effect of heme precursors, such as 5aminolevulinic acid (ALA), on the production of DCA, 0.25 mM ALA and 0.1 mM IPTG were added at the time of heterologous gene induction in REω. For analyzing the effects of thiourea, thiourea was added to the biotransformation reaction for a final concentration of 1 mM.6 The biotransformations were conducted in a shaking incubator (30 °C and 200 rpm), and samples were collected at different time points for product quantification by GC. Production of LDCAs from Coconut Oil Hydrolysate. Coconut oil was hydrolyzed according to the method described by Chua et al.26 with slight modifications. Briefly, the reaction mixture contained 10 mL of coconut oil (Samchun, Korea), 1.75 mL of distilled water, 13.25 mL of hexane, and 0.25 g of lipase lipozyme from Mucor miehei (Sigma-Aldrich, St. Louis, MO). The mixture was incubated at 40 °C with rotation at 200 rpm for 12 h. After hydrolysis, water and hexane were removed by rotary evaporation at 60 °C. The resulting residue was used for biotransformation studies with REω. The biotransformation medium and conditions were maintained as described in a previous section. The biotransformations were conducted in a shaking incubator (30 °C and 200 rpm), and samples were collected for product identification by GC. Product and Data Analysis. For the quantification and identification of α,ω-DCAs, the cell-free supernatant from the culture media of recombinant E. coli was acidified with HCl. DCAs produced through biotransformation were extracted twice with an equal volume of ethyl acetate containing 1 g/L methyl nonadecanoate (C19 FA) as an internal standard and then subjected to derivatization with (trimethylsilyl) diazomethane solution (TMS). The silylated derivatives were analyzed and quantified by a gas chromatograph (GC) (Agilent 7890A, U.S.) equipped with a flame ionization detector (FID) and an Agilent capillary column-DB WAX (30 m length, 0.320 mm internal diameter, 0.50 μm film thickness). The temperature program employed was 120 °C for 2 min, followed by a linear temperature gradient, increasing 10 °C/min to 220 °C. The concentration of biotransformation products was determined using calibration curves (R2 > 0.99). Identification was performed on GC (Agilent 7890A) equipped with a mass selective detector (MSD, Agilent 5975C). Samples were prepared as for GC/FID. Mass spectra were obtained by electron impact ionization at 70 eV. The expected reaction products were identified by their characteristic mass fragmentation pattern and compared to authentic standards. All of the experimental data were subjected to one-way analysis of variance (ANOVA) or multivariate analysis of variance (MANOVA) followed by Tukey’s test using SPSS (Version 11) software to determine levels of significance. Probability (P) values of