Improvement of isoprene production in Escherichia coli by rational optimization of RBSs and key enzymes screening
- 362 Downloads
As an essential platform chemical mostly used for rubber synthesis, isoprene is produced in industry through chemical methods, derived from petroleum. As an alternative, bio-production of isoprene has attracted much attention in recent years. Previous researches were mostly focused on key enzymes to improve isoprene production. In this research, besides screening of key enzymes, we also paid attention to expression intensity of non-key enzymes.
Firstly, screening of key enzymes, IDI, MK and IspS, from other organisms and then RBS optimization of the key enzymes were carried out. The strain utilized IDIsa was firstly detected to produce more isoprene than other IDIs. IDIsa expression was improved after RBS modification, leading to 1610-fold increase of isoprene production. Secondly, RBS sequence optimization was performed to reduce translation initiation rate value of non-key enzymes, ERG19 and MvaE. Decreased ERG19 and MvaE expression and increased isoprene production were detected. The final strain showed 2.6-fold increase in isoprene production relative to the original strain. Furthermore, for the first time, increased key enzyme expression and decreased non-key enzyme expression after RBS sequence optimization were obviously detected through SDS-PAGE analysis.
This study prove that desired enzyme expression and increased isoprene production were obtained after RBS sequence optimization. RBS optimization of genes could be a powerful strategy for metabolic engineering of strain. Moreover, to increase the production of engineered strain, attention should not only be focused on the key enzymes, but also on the non-key enzymes.
KeywordsIsoprene RBS sequence optimization T.I.R. Metabolic engineering Enzyme screening
Isoprene is an important platform chemical used for the commercial production of synthetic rubber and various other compounds, such as pesticides, medicines, oil additives, fragrances, and biofuels [1, 2]. Currently, 800,000 tons of isoprene monomer are produced annually from cracking petroleum, and over 95% of isoprene is used for rubber manufacture . However, the common problems of petroleum, such as irrecoverability, fluctuating price, high energy consumption and high environmental pollution, limit sustainable supply of isoprene in future . As an alternative, microbial biosynthesis of isoprene has attracted increasing attention and has been explored in the last decade .
The MVA pathway is commonly adopted for isoprene production (Fig. 1b). Three key enzymes, mevalonate kinase (MK/ERG12), isopentenyl-diphosphate isomerase (IDI) and IspS, impede the metabolic flux seriously. To improve isoprene production, most researches focused on these key enzymes. MK was identified as a bottleneck by targeted proteomics analysis. Higher level production of amorpha-4, 11-diene was obtained when the MK expression was up-regulated by selecting a strong promoter . IDI, catalyzing the transformation between IPP and DMAPP, was proved to be another key enzyme in isoprenoids production. 1.4-fold increase of β-carotene production would be achieved by introducing a strong promoter for IDI expression . In addition, summarized from the previous report, all known IspSs showed low kcat and high Km for DMAPP and restricted isoprene production seriously . A Gal4p (a promoter) controlled expression system lead increased IspS expression and a fourfold increase of isoprene production . Similar isoprene increase was achieved by enhancing expression of IspS through codon-optimization and adjustment of ribosome binding site (RBS) sequence . Further isoprene production increase was obtained by inserting the MVA pathway into a high copy plasmid with a strong promoter . Directed evolution has been performed for key enzymes. F310L and A570T mutations were identified after directed evolution of ISPS and a 27-fold increase of isoprene was obtained . Except for the modification of the natural MVA pathway, novel pathway which circumvented the rate-limiting steps has been explored. Isoprene was synthesized from mevalonate by two steps catalyzed by OleTJE from Jeotgalicoccus species and OhyAEM from Elizabethkingia meningoseptica, shortening the MVA pathway by three steps and avoiding the three rate-limiting steps . In summary, overexpression of key enzymes by selection of stronger promoter and RBS sequence and circumvention of key enzymes by a novel pathway has been usually applied for isoprene production.
On the other hand, screening the enzymes from different species with better characteristics for isoprene production was another powerful method. For isoprene production, enzymes from specific organisms were usually selected and screening enzymes from various species deserve our attention. However, little research was focused on key enzyme screening, especially the combinatorial analysis of them. Through the metabolic pathway, intracellular balance was obtained by the expression of key enzymes and non-key enzymes. However, all the engineering methods mentioned above aimed at improving expression of key enzymes and regulation of expression level of non-key enzymes was rarely reported. Moreover, when engineering methods, such as RBS sequence optimization, were applied for strains, the productivity data was usually detected; however, the change of enzyme expression was rarely analyzed.
In the present work, to increase isoprene production, we focused on the three key enzymes (MK, IDI and IspS) and non-key enzymes, acetyl-CoA acetyltransferase/HMG-CoA reductase (MvaE) and diphosphomevalonate decarboxylase (ERG19) through the whole pathway. Based on the previously constructed isoprene-producing strain , screening of key enzymes (MK, IDI and IspS) from different organisms and RBS sequence optimization were conducted firstly. Then enzymes with higher isoprene production were obtained and combinatorial analysis of the screened enzymes was carried out. In addition, RBS sequence optimization of non-key enzymes, MvaE and ERG19 and the effects on enzyme expression and isoprene production were examined. Furthermore, expression of key enzymes and non-key enzymes were analyzed through SDS-PAGE analysis to confirm the effect of RBS sequence optimization.
Strains and plasmids
Escherichia coli DH5α was used for gene cloning while E. coli Bl21(DE3) was used for expression of heterogenous genes and isoprene production. In our previous study, an engineered strain, which was defined as LMJ0 in this study, was constructed with the pYJM14, carrying the lower MVA pathway (genes ERG12, ERG8, ERG19 and IDI from Saccharomyces cerevisiae), and pYJM20, carrying upper MVA pathway (mvaE and mvaS genes from Enterococcus faecalis) and IspSpa from Populus alba (Additional file 1: Table S1).
Media and culture conditions
LB medium with appropriate antibiotics (100 μg/mL ampicillin or 34 μg/mL of chloramphenicol) was used for gene cloning. Modified M9 medium, adding appropriate antibiotics, was prepared as described for isoprene production under shake-flask fermentation .
Constrcution of plasmids and strains
All plasmids and primers (synthesized by GENEWIZ, Suzhou) used in this study are listed in Additional file 1: Tables S1 and S2, respectively. For IDI substitution, gene IDIbl (Genebank No. KND06900), IDIbs (Genebank No. AIY99819), IDImj (Genebank No. WP_010870377) and IDIsa (Genebank No. KII20428) were codon-optimized by JAVA Codon Adaptation Tool  and synthesized by GENEWIZ company. IDIbl fragment was amplified by 2 × PCR Bestaq™ MasterMix (abm, Canada) using IDIbl-F/IDIbl-R, digested by Sca I and Pst I (Thermo Scientific, USA) and ligated by T4 DNA Ligase (Thermo Scientific,USA) to the linearized pYJM14 which was digested by the same enzyme, Sac I and Pst I. pT-EEE-IDIbl was constructed. The other plasmids were constructed by the similar strategy, using primers listed in Table S2, correspondingly. IspSib (Genebank No. JP105673) and IspSmp (Genebank No. HW399219) were analyzed by ChloroP 1.1 Server  to eliminate the localization sequence of chloroplastid. IspS pa MT (L494P) and ERG12MT (N66 K/I152 M) were obtained by site-directed mutagenesis (TIANGEN, Beijing), using IspS pa MT -F/IspS pa MT -R and ERG12MT-1-F/ERG12MT-1-R, ERG12MT-2-F/ERG12MT-2-R as primers (Table S2).
Shake-flask cultures and GC analysis of isoprene
Strains were constructed by co-transformation of two plasmids into Bl21(DE3). Single colony was picked into seed culture (LB medium) and cultured at 37 °C overnight. Seed culture was transformed into 100 mL modified M9 medium in a sealed flask and cultured at 37 °C to OD600 of 0.6–0.8, when induction was conducted with 0.5 mM IPTG, then cultivation was continued at 30 °C for 48 h. At 3 h and 6 h after induction, 1 mL of the fermented liquid was collected by centrifugation at 5000×g for 15 min. The cell pellet was preserved at −20 °C for SDS-PAGE analysis. After 48 h cultivation, OD600 was detected. 1 mL of the gas samples from headspace of the sealed vials were analyzed by GC (Agilent 7890A, America) equipped with a flame ionization detector (FID). A HP-AL/S column (25 m × 320 μm × 8 μm) was used with nitrogen as carrier gas. The temperatures of oven, detector and injector were 50 °C, 150 °C and 50 °C, respectively.
SDS-PAGE analysis of protein expression
The preserved cell pellet was resuspended in 100 μL lysis buffer (Beyotime, Shanghai) and then placed on ice for 1 h. Centrifugation at 5000×g for 10 min was conducted to separate the soluble protein and other cell fragments. Concentration of the soluble protein was detected by a BCA protein assay kit (Beyotime, Shanghai). SDS-PAGE analysis was performed to detect the protein expression.
Translation initiation rate (T.I.R.) of different genes were analyzed by RBS Calculator [15, 16]. All results were expressed as mean ± standard deviation (SD). The results were analysed by OriginPro 9.0 and column charts were made. The showed figures were made by Adobe Photoshop CS5.
Accession numbers for the various genes
The codon-optimized gene sequences were submitted into the GeneBank and the accession numbers were provided. The accession numbers of codon-optimized IDIbl, IDIbs, IDImj and IDIsa were MH084474, MH084475, MH084476, MH084477. The accession numbers of codon-optimized IspSib and IspSmp were MH084470 and MH084471. The accession numbers of codon-optimized MKcv and MKmm were MH084472 and MH084473.
In our previous study, an isoprene-producing strain was constructed through overexpressing the hybrid MVA pathway and IspSpa from P. alba in E. coli. 287 mg/L isoprene production was achieved under shake flask condition . MK, IDI and IspS were identified as bottlenecks through the isoprene producing pathways (Fig. 1b). Release bottlenecks through utilizing enzymes with better performance and RBS sequence optimization were conducted. In addition, RBS sequence optimization of non-key enzymes, MvaE and ERG19 were performed.
Improvement of isoprene production through enzyme screening and enhancing RBS strength of key enzymes
As a key enzyme, MK (ERG12) was demonstrated to be an essential regulatory point in MVA pathway and it is feedback inhibited by downstream intermediates, such as DMAPP, IPP, FPP and GPP. Several MKs from different organisms were proved to show preferable characteristics and were utilized in this research. A N66K/I152M mutation of ERG12 (ERG12MT) which showed 148% increase of specific activity and 33% decrease of feedback resistance, was reported . However, the strain engineered with ERG12MT in this research have the same isoprene production as the wide type (Fig. 2b and Additional file 1: Fig. S2b). MK from Corynebacterium variabile (MKcv) and MK from Methanosarcina mazei (MKmm) were also selected for MK optimization. Decreased isoprene production resulted after substitution of MKcv (Fig. 2b and Fig. S2b). After MKmm substitution and RBS strength enhancement, 1.4-fold increase (402 mg/L) of isoprene production was obtained (Fig. 2b, gray column) and decreased OD600 values was detected (Additional file 1: Fig. S1b). In conclusion, the strain with MKmm and RBS strength enhancement was proved to be the better producer.
IspS is an enzyme which catalyzes the production of isoprene from the isoprenoids intermediate, DMAPP, with pyrophosphate elimination. IspSpa from P. alba was utilized in our original strain. A positive mutant, IspS pa MT , was constructed and utilized for isoprene production. However, no improved isoprene production was detected (Fig. 2c and Additional file 1: Fig. S2c). Then IspS from Mucuna pruriens (IspSmp) and IspS from Ipomoea batatas (IspSib) were utilized for IspS optimization. Improved isoprene production, 327 mg/L and 504 mg/L (1.9-fold increase) were achieved, separately (Fig. 2c). The OD600 values of different strains were showed in Additional file 1: Fig. S1c. Furthermore, RBS sequence optimization of IspSib was performed and no increased isoprene production was achieved (data not show). It is speculated that the original RBS sequence of IspSib, with a 35.845 kau T.I.R. value (Fig. 2c, white dot), is strong enough for IspSib expression. Further RBS sequence optimization may result in excessive IspS expression, which may lead to reduced expression of other proteins. The unbalanced protein expression may lead to decreased isoprene production. In conclusion, 1.9-fold increase of isoprene production resulted after utilizing IspSib.
Isoprene production was further improved through three screened enzymes utilization, IDIsa, MKmm and IspSib. To further improve isoprene production, combination of the three enzymes was conducted. Combinatorial optimization of IDIsa and IspSib was conducted and improved isoprene production, 599 mg/L, 2.1-fold increase comparing to the original strain, was achieved, higher than the strains when IDIsa and IspSib were utilized singly (Fig. 2d). However, no further increase of isoprene production resulted when additional MKmm substitution was added to the strains (Fig. 2d and Additional file 1: Fig. S2d). In the strain of modified with both MKmm and IspSib, 512 mg/L of isoprene production was achieved, same as the strain modified with only IspSib (Fig. 2d), indicating that MK node need more modifications. The strain modified with MKmm and IDIsa show the same isoprene production as the strain modified with only MKmm substitution and reduced production compared to the strain modified with only IDIsa substitution. The OD600 values of different strains were showed in Additional file 1: Fig. S1d. These result indicated that when IDIsa or IspSib or both of them were utilized, MKmm substitution lead to no improvement of isoprene production, even decreased production, which indicated that the MK node need more research, such as RBS sequence optimization of MKmm and finding new MKs from other organisms. In conclusion, the strain modified with IDIsa and IspSib showed the highest isoprene production, 2.1-fold increase comparing to the original strain.
Enhancing isoprene production through weakening RBS strength of non-key enzymes
Desired enzyme expression after RBS sequences optimization
In this research, enzyme screening and RBS sequence optimization of key enzyme, IDI, were conducted. IDIbs substitution lead to decreased isoprene production. However, in another study, introduction of IDIbs into carotene-producing strain lead to improved productivity . In view of the complexity of strain inner environment, the different performance of IDIbs in different engineered strain was reasonable to understand. Type 2 IDI showed better activities than type 1 IDI in lycopene-producing strain . IDIsa, a Type 2 IDI, showed better performance in this study. IDIsa has been studied as a classic enzyme for catalytic mechanism analysis of type 2 IDI and the lower MVA pathway from S. aureus have been widely utilized for isoprenoids production [21, 22]. However, for engineering of cell factories for isoprene production, IDIsa was utilized at the first time. Increased IDIsa expression and isoprene production were detected simultaneously after the predicated T.I.R. value was enhanced through RBS sequence optimization. It is reasonable to speculate that more isoprene production will be obtained with further optimization of RBS sequence, considering that only two RBS sequences of IDIsa were explored.
Modification of another key enzyme, MK (ERG12), was performed. A ERG12 mutation (ERG12MT) which show better characteristics was constructed and the unchanged isoprene production indicated that the higher enzyme activity detected in vitro did not mean better performance in vivo . In another study, MKcv has been attempted for isoprene production and 11.5-fold increase was achieved, even through the lower kcat/K m DMAPP (0.05), comparing to ERG12 (kcat/K m DMAPP ) . Different from other MKs, activity of MKmm was proved to be not inhibited by downstream intermediates . Dramatically increase of isoprene production, about 11–12 fold, was obtained in strain utilized MKmm and MKcv . However, in this study, relatively small increase was detected after modification of MKmm and no increase was detected for MKcv. The different performance of MK in different system indicated that the production of the target product was substantially determined by the whole system, not just one or two genes, even though optimization of one gene can lead to a great difference sometimes. It is speculated that MK in this system may not be a key enzyme and optimization of MK has little effect on the metabolic flux in the engineered strain.-
IspS from gray poplar (Populus alba × Populus tremula) was isolated firstly in 2001 and IspSs from polar and kudzu were widely utilized for isoprene bio-production . We summarized the previous studies for IspSs from other species and utilized the IspSs with better performance in our expression system. A L494P mutation of IspSpa showed higher kcat (2.1) and lower K m DMAPP (3.6), comparing to the wide type, kcat (1.5) and K m DMAPP (7) . However, in our research, isoprene production remained unchanged after utilizing L494P mutation, which indicated that high kcat and low K m DMAPP not always means high productivity and complicated intracellular environment cannot be changed by only one enzyme. Engineered strain with IspSmp substitution was reported to have significantly increased isoprene production . However, when the same IspSmp was introduced into our isoprene-producing system, only subtle increase was obtained. In consideration of T.I.R. value of IspSmp is the highest among all the IspSs used in this research (Fig. 2c), it is speculated that no further increase in production would obtained when T.I.R. value of IspSmp is increased after RBS sequence optimization. A novel IspS from I. batatas (IspSib) was identified through genome mining and performed better than other IspSs . Similar isoprene production increase was detected in this study. IspSib show better characteristics and deserve further research.
For strain engineering, most of the approaches can be divided into two types, rational engineering and adaptive evolution. The strategies regulating enzyme expression, including RBS sequence optimization, belong to rational engineering. For isoprene production, rational engineering approaches are mostly performed. In the last 17 years, almost thirty articles about microbial isoprene production have been published and, to our knowledge, only one research is about adaptive evolution, directed evolution of isoprene synthetase [11, 33]. RBS sequence optimization of key-enzymes and non-key enzymes was proved to be a useful strategy in this study and strain with improved isoprene production was obtained. However, RBS sequence optimization, including all the rational engineering, not always work well. Decreased isoprene production also resulted after the RBS sequence optimization of IspSib in this study. On the other hand, directed evolution of isoprene synthetase was performed and isoprene production was increased by threefold. Adaptive evolution is a powerful strategy for strain engineering, without considering the enzymes, the metabolic fluxes the products toxicity, etc. However, difficulty in high-throughput screening of isoprene limit its application . When this problem was resolved, combination of rational engineering and adaptive evolution may be a better strategy for subsequent study.
In the previous study, most researches focused on the key enzymes through the whole MVA pathway and almost no articles showed any interest in non-key enzymes. However, as we mentioned above, the production of the target product was the result of the whole internal environment, which was determined not only by the key enzymes, but also by the non-key enzymes, such as MvaE and ERG19. It is the first time that down-regulation of non-key enzymes help to increase target productivity. For metabolic engineering, modification of metabolic pathway should not only focused on up-regulation of key genes, but also down-regulation of non-key genes (Fig. 5). We should analyze the RBS strength of multiple genes when engineering an strain, except for the enzyme activity and affinity.
The steps in the metabolic engineering of E. coli for isoprene production in this work
Isoprene production (mg/L)
Improvement over original strain (fold)
Enzyme screening and RBS sequence optimization of key enzyme, IDI
Enzyme screening and RBS sequence optimization of key enzyme, MK
Enzyme screening of key enzyme, IspS
Combinatorial optimization of key enzymes, IDI, MK and IspS
RBS sequence optimization of non-key enzyme, MvaE and ERG19
ML, MX and HZ conceived and designed the research. ML carried out the majority of the experiments. HC, CL and JG helped to design and construct the plasmids used in this study. ML analyzed the data and drafted the manuscript. HZ and RN help to analyzed the data and reviewed the manuscript. All authors read and approved the final manuscript.
The authors thank Xungang Tan at the Institute of Oceanology, Chinese Academy of Sciences (IOCAS) for technical assistance.
The authors declare that they have no competing interests.
Availability of data and materials
All data generated or analysed during this study are included in this published article.
Consent for publication
Ethics approval and consent to participate
This study was supported by National Natural Science Foundation of China (NSF No. 31400084), Hainan’s Key Project of Research and Development Plan (No. ZDYF2017155), Taishan Scholars Climbing Program of Shandong (No. TSPD20150210), Youth Innovation Promotion Association CAS No. 2017252.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 13.JAVA Codon Adaptation Tool. http://www.jcat.de/. Accessed 30 Dec 2015.
- 14.ChloroP 1.1 Server. http://www.cbs.dtu.dk/services/ChloroP/. Accessed 30 Dec 2015.
- 16.The ribosome binding site calculator https://salislab.net/software/. Accessed 20 Apr 2016.
- 18.Bai R, Wyss M, Bai RY. New modified mevalonate kinase, which exhibits sensitivity to feedback inhibition, useful for manufacturing a medicament for treating disorders associated with decreased activity of mevalonate kinase. 2012. US8338155B2.Google Scholar
- 23.Mihara Y, Rachi H, Nishio Y, Katashkina YJ, Kazieva DE, Andreeva GI. Method of producing isoprene monomer. 2015. US2015/0275233.Google Scholar
- 26.Beck ZQ, Bott RR, Rife CL, Wells DH, Miller JV, Beck Z, Bott R, Rife C, Wells D, Miller J, et al. Generating variant of isoprene synthase with improved activity, expression or stability involves identifying location in the synthase for amino acid substitutions using three dimensional structure of the synthase, and modifying the synthase. 2010. US8518686-B2.Google Scholar
- 27.Hayashi Y, Harada M, Takaoka S, Fukushima Y, Yokoyama K, Nishio Y, Tajima Y, Mihara Y, Nakata K, Tayima Y, et al. New polynucleotide useful for producing isoprene monomer, composition specific base pair sequence. 2013. US8962296-B2.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.