Heterologous production of a sesquiterpene in Escherichia coli - Ly Thi Bich Thuy

TAP CHI SINH HOC 2015, 37(1se): 91-98 DOI: 10.15625/0866-7160/v37n1se. HETEROLOGOUS PRODUCTION OF A SESQUITERPENE IN Escherichia coli Ly Thi Bich Thuy1,2*, Nguyen Duc Bach2,3, Frank Hannemann2, Rita Bernhardt2 1Institute of Biotechnology, VAST, *ltb.thuy@ibt.ac.vn 2University of Saarland, Germany 3Vietnam National University of Agriculture ABSTRACT: Sesquiterpenes are a class of terpenes that consist of three isoprene units, with the molecular formula C15H24. They are common constituents of essential oils in plants and have been exploited as traditional medicines, spices, perfumes and flavorings. In nature, these compounds often exit in small quantities. Extracting them from plant requires a consumption of large amounts of natural resources, suffers low yields and impurities. Moreover, synthesis of these structural complex compounds by chemical processes usually requires difficult conditions, high cost and environmental hazards. To avoid these limitations, a new approach is using microorganisms, which are easy for manipulation to produce desired natural products by introducing a heterologous biosynthetic pathway into the host. In our previous work, gene coding a sesquiterpene cyclase from the myxobacteria Sorangium cellulosum So ce56 has been cloned and expressed in E. coli. However, the enzyme product seems to be a novel compound, which is not included in the database of National Institute of Standards and Technology (NIST) yet. In order to obtain sufficient amount of the sesquiterpene product for further analysis as well as application, this work investigates the production of this compound in E. coli. Genes coding for the mevalonate pathway from Streptomyces was co-expressed with gene coding for the sesquiterpene cyclase from the myxobacteria. The initial results showed that the highest yield of the desired sesquiterpene product reached 0.5 mg/50 ml culture medium. Keywords: E. coli, heterologous production, MVA pathway, MEP pathway, sesquiterpene, terpene cyclase. INTRODUCTION Sesquiterpenes belong to terpenes, a large and diverse group of natural products with the building block of isoprene units. They are produced by many plants and some insects. In plants, sesquiterpenes are involved in primary metabolisms as growth hormones, in communication and defense as attractants for pollinators, competitive phytoxins, antibiotics and herbivore repellents [1]. Due to their biological activities and physico-chemical properties, many terpenes are used as fragrances and flavors, pharmaceutical agents and insecticides [2]. In nature sesquiterpenes are often produced in small quantities. Extracting them from plants requires a consumption of large amounts of natural resources, suffers low yields and impurities. Furthermore, the synthesis of these compounds by chemical processes is difficult and expensive, as well as hazardous impact [3, 4, 5]. Therefore, in order to avoid these disadvantages, a new approach is introducing a biosynthetic pathway into microbial host organisms, which are easy for manipulation (commonly E. coli and S. cerevisiae) to produce natural compounds [6, 7, 8]. The biosynthesis of sesquiterpene can be divided into three major stages. The first stage is the biosynthesis of the active isoprene units, isopentenyl pyrophosphate (IPP) and its isomer dimethylallyl diphosphate (DMAPP). Sequentially, this precursor can be converted by various prenyltransferases to higher order terpenoid building blocks, farnesyl pyrophosphate (FPP). Finally, FPPs are converted to sesquiterpenes by enzymes called terpene cyclases (fig. 1). There are two pathways for the biosynthesis of isoprene units: the mevalonate-dependent pathway (MVA pathway) and the mevalonate-independent pathway (MEP pathway or DXP pathway) [2]. Except plants using both the MVA and MEP pathways for isoprenoid synthesis, eukaryotes use exclusively the MVA pathway to convert acetyl-CoA to IPP and DMAPP precursors [10, 11]. In contrast, in prokaryotes, isoprenoids are synthesized primarily via the MEP pathway [10]. It is known that E. coli does not synthesize monoterpenes, sesquiterpenes and diterpenes. Therefore, in order to turn E. coli into a suitable host for the production of terpenoids, a series of biosynthesis genes encoding enzymes catalyzing several steps from (FPP) to the desired isoprenoids have been introduced into E. coli [9]. In E. coli, IPP and DMAPP are synthesized via the MEP pathway for the prenylation of tRNAs and the synthesis of FPP, which is used for quinone and cell wall biosynthesis [9, 12]. However, the level of FPP is rather low and by no means sufficient for biotechnological applications. In addition, MVA pathway is known to be a superior biosynthetic route for delivering high-level precursors to terpene synthases. Therefore, this pathway is often used for heterologous production of isoprenoids in E. coli instead of the native pathway MEP [9, 12, 13, 14]. Recently, we have characterized a terpene cyclase called GeoA from the myxobacteria So ce56. The enzyme was able to convert FPP into a new sesquiterpene with 89% similar to valencene, an aroma component of citrus fruit and citrus-derived odorants [15]. To produce sufficient amounts of this product for further analysis, in vitro conversion is not feasible because (i) the substrate for this conversation (FPP) is very expensive and (ii) the enzyme GeoA is unstable. This research presents an alternative method for the production of this compound by introducing MVA pathway and co-expression with the terpene cyclase in E. coli. Figure 1. Biosynthesis pathways of Sesquiterpene GAP: glyceraldehyde-3-phosphate; IPP: isopentenylpyrophosphate; DMAPP: dimethylallylpyrophos-phate; GPP: geranylpyrophosphate; FPP: farnesylpyrophosphate. Figure 2. Mevalonate vector pACMev [9] MATERIALS AND METHODS Materials Farnesyl pyrophosphate (FPP) was purchased from Sigma-Aldrich (St. Louis, MO). Purified terpene cyclase was obtained from previous work [15]. Luria-Bertani (LB) and Terrific broth (TB) were purchased from Beckton-Dickinson (Germany). E. coli BL21(DE3) and C43(DE3) were purchased from Novagen (Wisconsin, USA). pETG vector containing gene coding the terpene cyclase GeoA was construted from pET17b in previous work [15]. The mevalonate vectors pACMev (fig. 2) and pACMv were kindly provided by Prof Norihiko Misawa, Research Institute for Bioresources and Biotechnology, Ishikawa Prefectural University, Japan. The plasmid pACMev contains a cluster composed of five genes of the mevalonate pathway (encoding 3-hydroxy-3-methylglutaryl CoA (HMGCoA) synthase, HMG-CoA reductase, MVA kinase, phospho-mevalonate (PMVA) kinase, and diphosphomevalonate (DPMVA) decarboxylase) as well as the IPP isomerase gene from Streptomyces sp. strain CL190 [9]. The plasmid pACMv is derived from pACMev by removing the genes encoding HMG-CoA reductase and HMG-CoA synthase and adding of a DNA fragment containing a multiple cloning site between HindIII and EcoRV [16]. In vitro assay An in vitro assay was used to produce the product of GeoA for the comparison with whole cell conversion. An in vitro assay was carried out in a volume of 500 µl of 10 mM potassium phosphate buffer (pH 7.4) containing GeoA (30 µg), FPP (50 µM) and MgCl2 (10 mM). Reactions were incubated for 1h at 30°C in a thermomixer under constant shaking. Reactions were stopped by adding 500 µl of n-hexane and thoroughly mixing of the samples. Extraction was done twice with 2 volumes of n-hexane. The n-hexane phases were pooled and concentrated to 100 µl in a vacuum-drier before they were analyzed by GC-MS. Production of sesquiterpene in E. coli E. coli C43(DE3) carrying plasmids pETG and pAC-Mev (or pAC-Mv) was precultured in TB medium containing ampicillin (100 µg/ml) and chloramphenicol (30 µg/ml) at 37°C overnight under constant shaking. The overnight culture was used to inoculate (1% v/v) 30 ml TB medium containing ampicillin and chloramphenicol in a 50 ml flask. The culture was grown at 37°C under constant shaking. When the OD600 reached 0.6-1.0, 1mM IPTG and 0.5 mg/ml of D-mevalonolactone (Sigma, Germany) were added and the cells were further cultivated at 27°C under constant shaking. After 3 hours, the culture was overlaid with 10 ml decane (Sigma, Germany) and incubated for another 48-72 hours under constant shaking. Aliquots of 100 µl of the overlaid-decane phase were then taken for GCMS analysis. GC/MS analysis of the produced sesquiterpene The sesquiterpene product from n-hexan phase were analyzed via GC-MS, on an Agilent 6890N (GC) and Agilent 5973N (MS) instrument under the following conditions: EI 70 eV with a scan width of m/z 50-600, source temperature 250°C, column DB-5ms or HP-5ms (both Agilent Technologies); injection temperature 250°C, interface temperature 250°C (200°C); carrier gas He, flow rate 1.2 ml/ min., constant flow mode; splitless injection, column temperature program: 50°C hold for 0.5 min, then raised to 320°C at a rate of 10°C/min and then hold 5 min (which is in total 32.5 min). All products were identified by comparison of their EI-MS spectra with those of the NIST MS search 2.0. RESULTS AND DISCUSSION E. coli strain and culture medium Since poor expression of GeoA in E. coli would limit the terpene production, optimizing the expression conditions was a primary task. At the beginning of this work, BL21(DE3) was tested for the expression of terpene cyclase GeoA. This strain commonly has been used as a host for the synthesis of a foreign protein using pET vectors. However, GeoA can only be expressed in this strain when co-expressed with the molecular chaperones GroEL and GroES (unpublished data). Therefore, C43(DE3), a mutant strain of E. coli BL21(DE3) was applied. This strain is often used to overcome the toxicity when recombinant proteins are over-expressed using the bacteriophage T7 RNA polymerase expression system [17]. We found that GeoA could be expressed in C43(DE3) without co-expression with chaperones. In addition, LB and TB culture media were also tested because the carbon source is known as a factor, which could influence on the protein expression level [18]. The protein expression from whole cell lysate was analyzed by SDS-PAGE (fig. 3). As shown in figure 3, bands corresponding to GeoA were observed 20 hours after induction in TB medium. In contrast, no expression was observed when LB medium was used. For this reason, the C43(DE3) strain and TB medium were used for the production of terpene compounds. kDa 1 2 3 4 5 116® 14,4® 18,4® 25® 35® 45® 66,2® ←GeoA Figure 3. SDS-PAGE of the expression of GeoA in E. coli C43(DE3) Lane 1 - cell lysate at 0 h of expression; Lane 2 - cell lysate after 20 h of expression in LB medium; Lane 3 - protein marker (pEQLab); Lane 4 - cell lysate after 44 h of expression in TB medium; Lane 5 - cell lysate after 20 h of expression in TB medium. Production of sesquiterpene in E. coli Among well-understood microorganisms, E. coli is an attractive host for the production of terpenoids because of their fast-growing, low cost and ease to manipulate [9, 12]. Many studies on pathway-engineering have been performed for increasing intracellular pools of IPP and DMAPP [19, 20]. For example, FPP amounts increased significantly when the intrinsic 1-deoxy-D-xylulose-5-phosphate synthase or 1-deoxy-D-xylulose 5-phosphate reductoisomerase gene in the MEP pathway was over-expressed in E. coli, resulting in an increase of the desired terpenoid yield [21]. Similarly, the FPP content was enhanced several times when the IPP isomerase gene derived from green alga Haematococcus pluvialis or yeast S. cerevisiae was expressed in E. coli [22]. Although engineering the native MEP pathway has received a lot of attention, introducing the heterologous mevalonate pathway into E. coli seems to be more attractive, since its higher capacity in delivering high-level precursors to terpene synthases for large-scale production of isoprenoids [9, 12, 14]. Besides, introducing just a foreign partial (bottom) mevalonate pathway into E. coli and supplementing the culture medium with mevalonate significantly increased lycopene production in E. coli [13]. Based on these observations, we introduced mevalonate pathway gene into E. coli and co-expressed them with GeoA for the sesquiterpene production. Two mevalonate vectors (pAC-Mev and pAC-Mv) have been reported in context with efficient synthesis of α-humulene [9] and ß-eudesmol [23] in E. coli. Figure 4. Mass spectra of the detected sesquiterpene from in vivo conversion in the overlaid decane phase Figure 5. Mass pectra of sesquiterpene product from in vitro conversion Figure 6. Sesquiterpene X production by E. coli C43(DE3) haboring different plasmids “+M” or “–M” indicates “with addition of” or “without addition of” D-mevalonolactone, respectively; nd - not detected In this work, E. coli C43(DE3) carrying plasmids haboring pAC-Mev or pAC-Mv together with pETG were cultured in TB medium supplemented with mevalonate. Since sesquiterpenes are volatile, the culture was overlaid with decane, an organic solvent with high hydrophobicity and low volatility to capture volatile compounds. This solvent has to be proven not to affect cell growth [24]. After 48 hours of incubation at 27οC, 100 µl of the overlay-decane phase was sampled for GCMS analysis. The result showed that there was a sesquiterpene present in the decane phase. The mass spectrum of the detected compound was compared to that of the sesquitepene from in vitro conversion. The relative ion abundances for the sesquiterpene produced in vivo (fig. 4) matched perfectly with the values of the product from the in vitro conversion of FPP by the terpene cyclase GeoA (fig. 5) for all ion fragments: 41, 55, 77, 91, 107, 120, 133, 147, 161, 175, 189, 204 m/z ions (Figure 4 and 5). In contrast, a control extract prepared from E. coli C43(DE3) transformed with only pAC-Mev (or pAC-Mv) did not give any products, confirming the fact that E. coli does not produce any sesquiterpene compounds. The results obtained here indicated that the sesquiterpene product of GeoA was produced in the engineered E. coli strain. Interestingly, when mevalonate (0.5 mg/ml) was supplied to the medium, sesquiterpene production was improved about 2.5 folds (Figure 6). Since E. coli is not able to utilize mevalonate without the introduction of the foreign mevalonate pathway gene [13], this result indicates that the introduced mevanolate pathway was active in this strain. Using valencene as external standard for GCMS, the highest sesquiterpene yield was estimated to be 0.5 mg/50ml culture medium. Furthermore, the product yield from the strain carrying pAC-Mev was higher than that from the strain carrying pAC-Mv (about 1.2 times) (fig. 6). In the control reaction (i.e. E. coli carrying only pAC-Mev or pAC-Mv), the sesquiterpene product was not detected, confirming that E. coli itself did not produce this compound from the isoprenoid precursor. However, the yield of sesquiterpene (≤ 10 mg/l) was much lower than the yield of α-humulene (≈ 1 g/l culture) achieved by Harada et al. [9] using the same vector pAC-Mev to boost isoprenoid synthesis. According to our observation, cell growth was considerably inhibited by decane, which might cause the comparatively low yield. Although the conversion time was prolonged to 48 hours after induction, the optical density of the culture was very low (OD600≈1.5). Therefore, we believe that the sesquiterpene production can be further improved by using other systems instead of decane to capture the volatile products. CONCLUSION Sesquiterpene production in E. coli C43(DE3) using the mevalonate gene cluster of pAC-Mev and pAC-Mv has been investigated. The highest yield of the product reached 0.5 mg/ 50 ml culture. The results from this work suggested that larger amounts of the sesquiterpene product for further analysis can be obtained by optimizing culture conditions, e.g. carbon source, or by substitution of the expression vector and host strain or the use of codon-optimized genes using the E. coli codon preferences. Higher yields can also be achieved by using a fermentor where culture parameters like pH, dissolved oxygen, and substrate concentration can be controlled. Acknowledgement: This work was supported by a PhD fellowship of the Vietnamese Ministry of Education and Training and DAAD (Deutscher Akademischer Austauschdienst) to Thuy T. B. L. We are very grateful to Prof. Dr. Norihiko Misawa for providing the plasmid pAC-Mv and thankful to Dominik Pistorious, Department of Pharmaceutical Biotechnology, Saarland University, for his kind help in GC-MS analysis. REFERENCES McGarvey D. J., Croteau R., 1995. Terpenoid Metabolism. The Plant Cell., 7: 1015-1026. Tholl D., 2006. Terpene synthases and the regulation, diversity and biological roles of terpene metabolism. Curr. Opin. in Plant Biol., 9: 1-8. Danishefsky S. J., 1996. Total synthesis of baccatin III and taxol. J. Amer. Chem. Soc., 118: 2843-2859. Nicolaou K. C., 1997. Total synthesis of eleutherobin. Angew. Chem. Int. Ed. 36: 2520-2524. Avery M. A., Chong W. K. M., Jennings-White C., 1992. Stereoselective total synthesis of (+)-artemisinin, the antimalarial constituent of Artemisia annua L. J. Amer. Chem. Soc., 114: 974-979. Takashi A., Tomoya K., Hiroshi Y., 2007. Terpenes Isolated from Bark and Wood on the Seiridium Canker Infected Chamaecyparis lawsoniana. Bull. Col. Edu. Ibaraki. Univ. (Nat Sci)., 56: 15-22. Muntendam R., Melillo E., Ryden A., Kayser O., 2009. Perspectives and limits of engineering the isoprenoid metabolism in heterologous hosts. Appl. Microbiol. Biotechnol., 84: 1003-1019. Misawa N., 2011. Pathway engineering for functional isoprenoids. Curr. Opin. Biotechnol., 22: 1-7. Harada H., Yu F., Okamoto S., Kuzuyama T., Utsumi R., Misawa N., 2009. Efficient synthesis of functional isoprenoids from acetoacetate through metabolic pathway-engineered Escherichia coli. Appl. Microbiol. Biotechnol., 81: 915-925. Boucher Y., Doolittle W. F., 2000. The role of lateral gene transfer in the evolution of isoprenoid biosynthesis pathways. Mol. Microbiol., 37: 703-716. Rohdich F., Hecht S., Gärtner K., Adam P., Krieger C., Amslinger S., Arigoni D., Bacher A., Eisenreich W., 2002. Studies on the nonmevalonate terpene biosynthetic pathway: metabolic role of IspH (LytB) protein. Proc. Natl. Acad. Sci., 99: 1158-1163. Martin V. J., Pitera D. J., Withers S. T., Newman J. D., Keasling J. D., 2003 Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nat. Biotechnol., 21: 796-802. Yoon S. H., Park H. M., Kim J. E., Lee S. H., Choi M. S., Kim J. Y., Oh D. K., Keasling J. D., Kim S. W., 2007. Increased β-Carotene Production in Recombinant Escherichia coli Harboring an Engineered Isoprenoid Precursor Pathway with Mevalonate Addition.. Biotechnol, Biotechnol. Prog., 23: 599-605. Vadali R. V., Fu Y., Bennett G. N., San K. Y., 2005. Enhanced lycopene productivity by manipulation of carbon flow to isopentenyl diphosphate in Escherichia coli. Biotechnol. Prog., 21: 1558-1561. Thuy Thi Bich Ly, Bach Duc Nguyen, Frank Hannemann, Rita Bernhardt, 2014. Cloning, expression and characterization of a terpene cyclase from the myxobacteria So ce56. Vietnamese Journal of Biotechnology, 32(4): 69-79. Yu F., Okamoto S., Harada H., Yamasaki K., Misawa N., Utsumi R., 2011. Zingiber zerumbet CYP71BA1 catalyzes the conversion of α-humulene to 8-hydroxy-α-humulene in zerumbone biosynthesis. Cell. Mol. Life. Sci., 68: 1033-1040. Dumon-Seignovert L., Cariot G., Vuillard L., 2004. Thetoxicity of recombinant proteins in Escherichia coli: a comparison of overexpression in BL21(DE3), C41(DE3), and C43(DE3). Protein. Expr. Purif., 37: 203-206. Lee P. C., Mijts B. N., Schmidt-Dannert C., 2004. Investigation of factors influencing production of the monocyclic carotenoid torulene in metabolically engineered Escherichia coli. Appl. Microbiol. Biotechnol., 65: 538-546. Kim S. W., Keasling J. D., 2001. Metabolicengineering ofthe nonmevalonate isopentenyl diphosphate synthesis pathway in Escherichia coli enhances lycopene production. Biotechnol. Bioeng., 72: 408-415. Reiling K., Yoshikuni Y., Martin V. J. J., Newman J., Bohlmann J., Keasling J. D., 2004. Mono and Diterpene Production in Escherichia coli. Biotechnol. Bioeng., 87: 200-212. Albrecht M., Misawa N., Sandmann G., 1999. Metabolic engineering of the terpenoid biosynthetic pathway of Escherichia coli for production of the carotenoids β-carotene and zeaxanthin. Biotechnol. Lett., 21: 791-795. Yuan L. Z., Rouviere P. E., Suh W., 2006. Chromosomal promoter replacement of the isoprenoid pathway for enhancing carotenoid production in E. coli. Metab. Eng., 8: 79-90. Yu F., Harada H., Yamasaki K., Okamoto S., Hirase S., Tanaka Y., Misawa N., Utsumi R., 2008. Isolation and functional characterization ofa beta-eudesmol synthase, a new sesquiterpene synthase from Zingiber zerumbet Smith. FEBS. Lett., 582: 565-572. Wang C., Yoon S. H., Shah A. A., Chung Y. R., Kim J. Y., Choi E. S., Keasling J. D., Kim S. W., 2010. Farnesol production from Escherichia coli by harnessing the exogenous mevalonate pathway. Biotechnol. Bioeng., 107: 421-429. SẢN XUẤT SESQUITERPENE BẰNG Escherichia coli Lý Thị Bích Thủy1,2, Nguyễn Đức Bách2,3, Frank Hannemann2, Rita Bernhardt2 1Viện Công nghệ sinh học, Viện Hàn lâm KH & CN Việt Nam 2Trường Đại học Tổng hợp Saarland, CHLB Đức 3Học viện Nông nghiệp Việt Nam TÓM TẮT Sesquiterpene là những hợp chất thuộc nhóm terpene chứa 3 đơn vị isoprene với công thức phân tử là C15H24. Đây là nhóm hợp chất phổ biến trong tinh dầu thực vật, được sử dụng cho nhiều mục đích khác nhau như thuốc, gia vị, nước hoa và hương liệu. Những hợp chất này thường tồn tại với lượng rất nhỏ trong tự nhiên. Việc khai thác các hợp chất này từ thực vật làm có thể dẫn đến cạn kiệt nguồn tài nguyên thiên nhiên và sản phẩm có hàm lượng thấp hoặc lẫn nhiều tạp chất. Do cấu trúc phức tạp, việc tổng hợp các chất này bằng con đường hóa học đòi hỏi điều kiện khắc nghiệt, giá thành cao và ảnh hưởng tới môi trường. Để khắc phục những hạn chế này, xu hướng hiện nay là sử dụng các vi sinh vật dễ thao tác điều khiển để sản xuất các hợp chất mong muốn bằng cách đưa vào vật chủ các gene mã hóa cho các enzyme tham gia vào con đường sinh tổng hợp chất đó. Trong nghiên cứu trước của chúng tôi, gen mã hóa cho một enzyme sesquiterpene cyclase từ myxobacteria So ce56 đã tách dòng, biểu hiện ở E. coli. Sản phẩm chuyển hóa của enzyme này là một hợp chất sesquiterpene mới. Nhằm thu được một lượng đủ lớn sesquiterpene này cho các phân tích tiếp theo cũng như ứng dụng của hợp chất này, chúng tôi đã khảo sát quá trình sản xuất hợp chất này ở E. coli bằng cách đồng biểu gene mã hóa con đường mevalonate có nguồn gốc từ Streptomyces và gene mã hóa enzyme terpene synthase có nguồn gốc từ myxobacteria. Kết quả bước đầu cho thấy hàm lượng sesquiterpene đạt được 0.5 mg/ 50 ml môi trường. Từ khóa: E. coli, heterologous production, MVA pathway, MEP pathway, sesquiterpene, terpene cyclase. Ngày nhận bài: 22-10-2014