Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Direct combinatorial pathway optimization Pieter Coussement, David Bauwens, Jo Maertens, and Marjan de Mey ACS Synth. Biol., Just Accepted Manuscript • Publication Date (Web): 27 Sep 2016 Downloaded from http://pubs.acs.org on September 27, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Synthetic Biology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
Direct Combinatorial Pathway Optimization Pieter Coussement, David Bauwens, Jo Maertens, and Marjan De Mey∗ Department of Biochemical and Microbial Technology, Ghent University, Coupure Links 653, 9000 Gent, Belgium E-mail:
[email protected] Abstract Combinatorial engineering approaches are becoming increasingly popular, yet they are hindered by the lack of specialized techniques for both efficient introduction of sequence variability and assembly of numerous DNA parts, required for the construction of lengthy multi-gene pathways. In this contribution, we introduce a new combinatorial multi-gene pathway assembly scheme based on Single Strand Assembly (SSA) methods and Golden Gate Assembly, exploiting the strengths of both assembly techniques. With a minimum of intermediary steps and an accompanying set of well-characterized and ready-to-use genetic parts, the developed workflow allows effective introduction of various libraries and efficient assembly of multi-gene pathways. It was put to the test, by optimizing the lycopene pathway as proof-of-principle. The here constructed libraries yield ample variation in lycopene production. In addition, good-performing transformants with a significantly higher lycopene production were obtained as compared to previously published reference strains. The best selected producer yielded threefold improvement in lycopene titers up to 448 mg lycopene / g CDW. The proposed workflow in combination with the accompanying sets of ready-to-use expression and carrier plasmids, will allow the combinatorial assembly of increasingly lengthy product pathways with minimal effort.
1 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Abbreviations Golden Gate (GG), Golden Gate assembly (GGA), Single strand assembly (SSA)
Keywords Lycopene, combinatorial pathway optimization, E. coli
1
Introduction
Nature harbors a plethora of secondary metabolites with interesting properties that comply well with modern day needs in the food, pharmaceutical, cosmetics and bio-fuel sectors, sparking industrial interest (1 ). Traditional approaches to obtain these secondary metabolites are seriously flawed by intrinsic deficiencies. For example, organic synthesis is typically hindered by the complicated multi-step synthesis routes and the extensive use of organic solvents and other non-renewable adjuvant products (2 ). On the other hand, extraction from natural resources is hampered by poor extraction efficiencies, co-extraction from complex matrices and low product yields as these metabolites only accumulate in relatively low levels (3 ). In response to these deficiencies, more environmentally friendly and economically viable biotechnological alternatives are being developed for the synthesis of an ever-increasing number of secondary metabolites (1 , 4 , 5 ). In this context, textbook organisms such as Escherichia coli and Saccharomyces cerevisiae are being transformed into true microbial cell factories, converting cheap and renewable feedstocks into high added-value products with high yields, titers and an equally high productivity (6 ). These microbial cell factories are provided with, often hard-to-express, multi-step product pathways and their metabolism is extensively optimized in order to adequately supply the various pathway precursors (6 ). This optimization typically comprises extensive expression modulation of the pathway
2 ACS Paragon Plus Environment
Page 2 of 37
Page 3 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
genes involved. In this regard, massive or even imbalanced expression can result in metabolic burden because of the excessive withdrawal of the cell’s essential building blocks, reducing cell viability and productivity (7 ), but can also result in improper protein folding and aggregation (8 ), or even in the accumulation of (toxic) pathway intermediates possibly triggering a series of regulatory events (9 ), such as feed-back inhibition, all adversely affecting the flux through the pathway. Hence, to avoid the detrimental effects of pathway expression, extensive fine-tuning of pathway expression with respect to the host’s metabolism is of capital importance (10 , 11 ). In this respect, metabolic engineers dispose of, e.g., promoter and ribosome binding site libraries, with varying transcription and translation initiation rates, to control the respective processes (12 –16 ). The development of microbial cell factories for these secondary metabolites and the de novo design of the complex product pathways has largely been relying on recent advances in DNA synthesis and synthetic biology. Techniques like Gibson assembly (17 ) and Golden Gate assembly (GGA) (18 ) have enabled the successful assembly of increasingly large DNA fragments, pathways and even of entire genomes (19 ). Moreover, extensions of these assembly techniques such as MoClo (20 ), GoldenBraid (21 ), Single-step linker-base combinatorial assembly (22 ), Single Strand Assembly (SSA) (23 ) and MAGE (24 ) have allowed expression modulation in a combinatorial fashion. In this respect, the ability to re-use existing well-characterized genetic parts, i.e., modularity, and the ability to quickly and efficiently introduce sequence variability, allowing combinatorial assembly are two features of major importance. Whereas MoClo and GoldenBraid excel in modularity as they allow to build constructs using a limited set of defined parts, they fall short with regard to the rapid and efficient introduction of genetic variability, as a result of the multiple intermediate steps, significantly reducing variability. Nevertheless, this is a key requirement when constructing combinatorial libraries. On the other hand, SSA, MAGE and single-step linker-based combinatorial assem-
3 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
bly score well on introducing sequence variability, however, these techniques are less suited for the construction of multiple DNA parts, as assembly efficiencies drastically decrease with the increasing number of DNA parts to be assembled. In conclusion, the state-of-the-art assembly methods either score well on assembling multi-piece DNA fragments or on introducing sequence variability, whereas ideally both criteria should be fulfilled. Hence, in this contribution, we propose a new assembly scheme, by combining SSA methods with Golden Gate Assembly (GGA), exploiting the strengths of both assembly techniques for the directed combinatorial optimization of multi-gene pathways. With a minimum of intermediary steps and a set of well characterized ready-to-use genetic parts, the developed workflow allows effective introduction of various libraries and efficient assembly of multiple DNA parts. Its power is demonstrated by applying the workflow to the combinatorial optimization of the multi-gene pathway for lycopene biosynthesis in E. coli . This isoprenoid belongs to a versatile class of secondary metabolites with applications in various fields, such as the pharmaceutical, flavour, fragrance, etc., markets. (25 , 26 ).
2 2.1
Materials and Methods Chemicals, oligonucleotides and molecular biology
All reagents were purchased from Sigma-Aldrich (Diegem, Belgium), unless otherwise stated. Agarose and ethidium bromide were purchased from Thermo Fisher Scientific (Erembodegem, Belgium). Qiagen kits (Hilden, Germany) were used for all DNA preparations. Oligonucleotides were purchased from Integrated DNA Technologies (Leuven, Belgium) and are listed in Supplementary Table S.1. Sequencing services were conducted by Macrogen (Amsterdam, The Netherlands). Molecular procedures were performed according to the standard procedure (27 ).
4 ACS Paragon Plus Environment
Page 4 of 37
Page 5 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
2.2
Strains
E. coli One Shot TOP10 ElectrocompTM cells (Life Technologies, Gent, Belgium) were solely used for cloning purposes. For lycopene production experiments, two different background expression hosts were selected to express the combinatorially assembled pathway, i.e., E. coli K-12 MG1655 DE3 ∆recA ∆endA ∆araA::pTrc-MEP (Trc-MEP) and E. coli K-12 MG1655 DE3 ∆recA ∆endA ∆araA::pT7 MEP (T7-MEP), where MEP stands for the dxs-idi-ispDF operon. The latter two strains were kindly provided by Dr. Ajikumar Parayil (25 ). A full list of the plasmids and strains used, are listed in Supplementary Tables S.2 and S.3, respectively.
2.3
Media and culture conditions
The culture medium Luria Broth (LB) consisted of 1% tryptone-peptone (Difco, BD, Erembodegem, Belgium), 0.5% yeast extract (Difco) and 1% sodium chloride (VWR, Leuven, Belgium). Luria Broth Agar (LBA) is similarly composed to LB, be it for the addition of 10 g/L agar (Difco). If required, the culture medium was supplemented with appropriate antibiotics. Stock concentrations for antibiotics were 100 mg/mL for ampicilin, 25 mg/mL for chloramphenicol, and 50 mg/mL for kanamycin. Antibiotic stocks were diluted 1000x for cell culture experiments. After transformation, cultures were plated out on LBA containing the appropriate antibiotic and grown overnight at 37 ◦ C. To allow sacBR counter selection, 5% sucrose was added to the medium, whilst NaCl was removed. Individual colonies were picked with an automated colony-picker (QPix2, Genetix/Molecular Devices, California, USA) and inoculated into sterile 96-well flat-bottomed microtiter plates enclosed by a sterile Breathe-Easy sticker containing 150 µL LB medium per well, supplemented with the appropriate antibiotics. The microtiter plates (Greiner Bio-One, 655096) were incubated for 24 h at 37 ◦ C and 250 rpm (50 mm shaking amplitude) on a LS-X AppliTek shaker(AppliTek, Nazareth, Belgium). 5 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2.4
Plasmids and plasmid construction
In this study, two different sets of plasmids were constructed to allow directed combinatorial pathway assembly, i.e., carrier plasmid s (pCPs) (see Supplementary Figure S.1) and expression plasmid s (pEXs) (see Supplementary Figure S.2). pCP originates from plasmid pUC57, which contains a pMB1 ori and an ampR selection marker. Silent point mutations were introduced in ampR to render it non sensitive to BsaI. This set of pCP plasmids also carries an operon devoid of any coding sequences, which comprises the constitutive p22 promoter and ribosome binding site from De Mey, et al.(12 ) and a different terminator from the BIOFAB library (14 ) was used for every pCP. This part is flanked by two unique BsaI cut sites, designed for ordered GGA. In addition, the BsaI sites are properly oriented to allow the assembly of various pCP-derived genes into an expression plasmid (pEX) (see Figure 1). pEX plasmids, are a set of E. coli compatible vectors carrying a sacBR operon, enabling counter selection, flanked by two BsaI cut sites. Successful Golden Gate assembly (Golden Gate assembly) results in the removal of the sacBR operon and the BsaI recognition sites due to the orientation of the BsaI cut sites on the pEX plasmids. Moreover, these pEX vectors contain a kanR selection marker and an ori. In this study, two different ori ’s were evaluated, i.e., the low-copy pSC101 (∼ 5 copies, pEX-SC101) and medium-copy pMB1 (∼ 15-20 copies, pEX-BR322). Lycopene reference plasmids were constructed by picking up the carotenoid pathway gene cluster of Erwinia herbicola (crtEIB ) from pAC-Lyc (28 ), and by cloning it in both pEX’s. The DNA parts required for plasmid construction were amplified using PrimeSTAR-HS (TaKaRa Bio, Saint-Germain-en-Laye, France) with the appropriate primers (Supplementary Table S.1). Subsequently, these parts were purified with the appropriate Qiagen kit and finally assembled using Circular Polymerase Extension Cloning (CPEC) (29 ). The CPEC assembly reaction mixture contained 0.5 µM dNTP’s, 0.75 µL DMSO, 0.5 µL Q5 polymerase, and 4 µL Q5 reaction buffer (New England BioLabs, Ipswich, USA). Temperature profile for 6 ACS Paragon Plus Environment
Page 6 of 37
Page 7 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
CPEC started with 30 sec at 98 ◦ C. Next were 15 cycles of 10 sec at 98 ◦ C, followed with 30 sec at 55 ◦ C and the necessary elongation time at 72 ◦ C. The final step was 5 min at 72 ◦ C. All plasmids used are listed in Supplementary Table S.2.
2.5
Library construction
Three different oligonucleotide-based methods to introduce sequence variability (SSA methods) were evaluated, i.e., a Gibson-based assembly method (SSA-Gibson (23 )), and two CPECbased methods using a single primer SSA-CPEC-1P and two primers SSA-CPEC-2P, respectively. SSA-Gibson was performed as described in Coussement, et al.(23 ). In short, 100 ng of purified backbone was used for SSA using an at 98 ◦ C melted primer mix with a final concentration of 400 nM per oligonucleotide. For the CPEC-based methods, the above described CPEC protocol was used, with similar concentrations of backbone and oligonucleotides as compared to SSA-Gibson. The plasmid backbone was amplified using PrimeSTAR-HS (TakaraBio). A subsequent DpnI (New England Biolabs) treatment was performed to remove template DNA. Gel-extraction allowed to simultaneously check the fragment length and purify the backbone. Concentrations were measured using the NanoDrop (Thermo Fisher Scientific). All oligonucleotides used and plasmids obtained, are listed in Supplementary Table S.1 and Table S.2, respectively. The completed assembly reactions were transformed in E. coli One Shot TOP10 ElectrocompTM .
2.6
Pathway assembly
The directed combinatorial assembly step relies on combining the appropriate pCP-derived libraries with the corresponding pEX for Golden Gate assembly(18 ). After transformation of the pCP libraries, 1 mL transformed culture was used as inoculum for 15 mL LB with the appropriate antibiotic, to biologically amplify the individual libraries. After 24 h incubation at 30 ◦ C, the libraries were extracted using the appropriate Qiagen kit. Subsequently, Golden Gate assembly was performed to assemble the different pCP-derived libraries into 7 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
the final pEX expression vector. In short, the Golden Gate assembly reaction mix (18 ) was supplemented with 100 ng pEX and equimolar amounts of the pCP-derived libraries. MilliQ water is added up to 20 µL. Electrotransformation was used to introduce the final pathway libraries in the various expression hosts. Cells were made competent using the protocol described by Sambrook, et al.(27 ). Cultures were plated on LBA with appropriate antibiotics, lacking NaCl, but with 5% sucrose to allow for counter selection.
2.7
Fluorescence measurements/Evaluation of SSA methods
The success of the SSA methods was evaluated based on two criteria, i.e., the number of colony forming units (CFUs) and the generated variation in fluorescence, which correspond to the size of the library and the introduction of genetic/phenotypic variability, respectively. CFUs were determined 24 h after transformation. Variation in fluorescence was determined by analyzing 288 robotically picked colonies grown in microtiter plates (Greiner Bio-One, 655201) at 37 ◦ C for 24 hin LB medium. Fluorescence output (F) was quantified using a Tecan M200 infinite PRO (Tecan, Mechelen, Belgium). Excitation and emission wavelengths used are 588 nm and 633 nm, respectively. Optical density at 600 nm (OD600 ) was measured for biomass correction. The criterion used for fluorescence is F/OD600 .
2.8
Lycopene assay
From every library 288 individual colonies were randomly picked by an automated colonypicker (QPix2, Genetix) and cultivated in polypropylene microtiter plates containing 150 µL LB medium per well, supplemented with the appropriate antibiotic and 0.1 mM IPTG to induce the MEP pathway. The microtiter plates were incubated for 48 h at 37 ◦ C and 250 rpm (50 mm shaking amplitude). After incubation, OD600 was measured for biomass correction using a Tecan Infinite M200 Pro. Optical density was converted to g CDW using the OD600 - g CDW correlation from 8 ACS Paragon Plus Environment
Page 8 of 37
Page 9 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
Ren, et al.(30 ). The plate was cooled (4 ◦ C) and centrifuged at 3000 rpm (2224 rcf) for 10 min in a Rotixa swing bucket centrifuge (Hettich, Tuttlingen, Germany). The supernatant was discarded by inverting the plate and by gently tapping the bottom. The pellet was subsequently resuspended in 150 µL of extraction solution, i.e., a mixture of 50% tetrahydrofuran (THF), 20% acetone, 30% glycerol (35%) and 0,1% α-tocoferol, which also acts as a lysing solution. After 10 min incubation at 50 ◦ C, this solution was centrifuged as described above. As such lycopene, which dissolves well in this extraction solution, can be separated from cell debris, which would interfere with the lycopene measurement. 100 µL of the red-coloured supernatant was transfered to propylene 96-well plates for direct absorbance measurements. Lycopene production was quantified based on its three characteristic absorbance wavelengths of lycopene, namely 443, 471 and 502 nm, using a standard curve obtained with commercial lycopene.
2.9
Statistical calculations
Statistical calculations and analyses were performed in R(31 ). Error bars represent the calculated standard deviation, unless otherwise stated. Boxplots were created in R using the default settings.
On the left side of the figure the expression host, the pEX and
the construction technic are indicated, whilst the boxplot on the left show the absorbance variation over OD for 288 colonies from the library. 471 nm is a characteristic absorbance peak for lyco
3
Results and Discussion
The development of novel microbial cell factories remains cumbersome and time consuming despite the substantial advances in the fields of synthetic biology and systems biology. Since rational methods for pathway optimization are seriously flawed for example by the frag-
9 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
mentary knowledge of the various regulatory mechanisms (32 ) affecting the flux through the pathway (33 ), (semi-)combinatorial approaches in combination with high-throughput screens are popularly applied to address such multi-variate optimization questions. In response to the compelling need for ever more efficient and fast construction methods for combinatorial pathway optimization, we introduce a novel scheme for combinatorial pathway assembly and a set of ready-to-use plasmids designed to this end. This workflow comprises two major steps, i.e., a first step to introduce sequence variability using degenerate oligonucleotides, enabling expression variation for each individual step in the pathway, and a second step to assemble the multi-gene pathway using GGA. Next, the resulting pathway library is transformed in the expression host and the obtained transformant library is screened for the desired phenotype (see Figure 1). The developed ready-to-use parts and the workflow are designed to allow high-throughput applications and the optimal reuse of genetic parts, enabling the direct combinatorial optimization of novel pathways with minimal effort. Its usefulness was evaluated by combinatorially optimizing the multi-gene lycopene pathway. [Figure 1. About here]
3.1
Step 1: Introducing sequence variability
In a first step of the workflow, sequence variability is introduced to enable expression modulation of each individual reaction of the pathway. To this end, two novel variation-introducing assembly methods based on CPEC and making use of degenerate oligonucleotides are presented and benchmarked against our earlier described SSA method (SSA-Gibson) (23 ), which has proven to be a powerful tool for sequence variability introduction. The two CPEC-based assembly methods, i.e., the single primer (SSA-CPEC-1P) variant and the two primer (SSACPEC-2P) variant, and SSA-Gibson mainly differ by the enzymes used and the operating temperature. Whereas SSA-CPEC only depends on the activity of a polymerase, SSAGibson needs the activity of a ligase, an endonuclease and a polymerase. The use of a single 10 ACS Paragon Plus Environment
Page 10 of 37
Page 11 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
enzyme, renders the reconciliation of differences in temperature optima obsolete. In this respect, SSA-Gibson is an isothermal reaction at 50 ◦ C, whilst both SSA-CPEC variants are PCR reactions, with a likewise temperature profile. SSA-Gibson nor SSA-CPEC-1P are able to amplify the final construct, in contrary to the SSA-CPEC-2P. To compare the different techniques, a spacer-randomized promoter library was introduced, conserving both the -35 and -10 boxes upstream of the reporter mKate2 (34 ). To this end, two features were evaluated, i.e., the methods’ efficiency, reflected by the counted CFUs, and the ability to introduce genetic variability, reflected by the variation in biomasscorrected fluorescence of the created libraries. [Figure 2. About here] Figure 2 compares the CFUs obtained for the three methods tested. Whereas SSA-Gibson and SSA-CPEC-1P yielded a similar number of CFUs, SSA-CPEC-2P yielded a significantly higher number of CFUs. Since for all three methods equal amounts of DNA were used both for the backbones and the oligonucleotides, the observed increase in CFUs should be due to differences in assembly mechanism. In contrast to SSA-Gibson and SSA-CPEC-1P, which merely allow the introduction of the degenerate oligonucleotides into the pCP plasmid backbone, the mechanism underlying SSA-CPEC-2P additionally allows amplification of the assembled plasmids. Thus, SSA-CPEC-2P resembles a whole plasmid amplification reaction (35 ), resulting in a significantly higher number of assembled plasmids, and thus transformants. Figure 2 shows the fluorescence of the promoter libraries constructed using the aforementioned methods. Whilst the SSA-Gibson-based promoter library is very similar to earlier described promoter libraries (12 , 23 ), both SSA-CPEC-based libraries show an aberrant distribution. These differences are probably due to extensive oligonucleotide melting for the SSACPEC-based methods. The iterative melting at 95 ◦ C likely allows oligonucleotides with strong secondary structures to be melted and subsequently assembled into the dsDNA con-
11 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
struct. In contrast, the moderate operating temperature (50 ◦ C) of the SSA-Gibson method, releases less secondary structures which might prohibit introduction of these sequences in the final construct. Introduction of such secondary structures in the promoter sequence might negatively affect the affinity of RNAP for the promoter sequence, reducing transcription initiation and resulting in an over-representation of weaker promoters in the resulting libraries as compared to (12 , 23 ). This could explain the distribution differences between these libraries. Despite these differences, a similar range of promoter strengths was obtained for SSA-Gibson and SSA-CPEC-1P. For a screened library of 288 colonies, the strongest promoter reached just more than half the expression strength of both other methods (SSAGibson and SSA-CPEC-1P). This is probably due to the small sample size as this method resulted in a double amount of CFU’s and sequencing showed proper sequence randomization. Hence, all three methods can be used to introduce sequence variability using simple degenerate oligonucleotides.
3.2
Step 2: Directed multi-gene pathway assembly
In the second step of the workflow, the constructed libraries for the individual pathway genes (step 1) are assembled into a final expression plasmid using an ordered Golden Gate assembly. To this end, two sets of standardized and well-defined building blocks, i.e., a set of ten pCP plasmids (see Supplementary Figure S.1) and a set of pEXs plasmids (see Supplementary Figure S.2) required to assemble the complete pathway (see Figure 1), were designed. Essential parts of these building blocks are the BsaI cut sites, which allow ordered and direct pathway assembly. Purposely, a R/Bioconductor script (36 , 37 ) was developed (Supplementary listings S.1) to generate in silico multiple BsaI sites (see Supplementary Table S.2) complying with the guidelines formulated in Engler, et al.(38 ), i.e., multiple complementary and palindromic sequences should be avoided, and should differ by at least two nucleotides. An additional restriction that was introduced for these designed sticky ends is 12 ACS Paragon Plus Environment
Page 12 of 37
Page 13 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
the GC content of 50%. This set of optimally in silico designed BsaI cut sites ensure the ordered construction of the pathway since each BsaI cut site generates a unique sticky end which is complementary with the sticky end of either the consecutive pCPs-derived library or of the pEXs-derived backbone part. The set of pCP plasmids are used to generate the various gene library parts flanked by two sequentially numbered and BsaI-generated sticky ends, e.g. pCP(1-2), to allow ordered pathway assembly (see Figure 1). A total of 10 pCP plasmids (see Supplementary Table S.2), each with a different couple of unique Golden Gate assembly cut sites, were constructed. Pathway genes can be cloned into these sequential pCP plasmids. Using GGA, the final combinatorial pathway is assembled from the BsaI created parts starting from the pCP plasmids and the appropriate pEX plasmid (see Figure 1). These pEX vectors comply with three specific requirements, i.e., they contain a different selection marker compared to the pCP plasmids, in order to avoid cross-contamination from surviving pCP plasmids, a sacBR counter selection marker, to avoid false positives, and the proper BsaI cut sites. Since the plasmid copy number plays an important role for plasmid-based expression, two different pEX plasmid sets were created, one set with a low-copy ori (pSC101, ∼ 5 copies) and one set with a medium-copy ori (pMB1, ∼ 15-20 copies).
3.3
Workflow and parts validation: test case lycopene
The proposed workflow and the constructed sets of plasmids were put to the test using the lycopene pathway as proof-of-principle. Lycopene is a bright red carotene that can be found in, i.a., tomatoes. It has applications as food colourant and antioxidant. Its bright red colour renders it an ideal molecule to validate the developed combinatorial pathway assembly tools. The lycopene biosynthesis pathway, starting from farnesyl diphosphate (FPP), consists of three reactions, catalyzed by CrtE, CrtI and CrtB (see Figure 3). To allow for sufficient precursor supply, dxs-idi-ispDF should be overexpressed. [Figure 3. about here] 13 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
3.3.1
Creation of individual gene libraries
The coding sequences of crtE, crtI and crtB were introduced into the plasmids pCP(12), pCP(2-3) and pCP(3-4), respectively, using CPEC assembly. For these three coding sequences, all three SSA methods were used for the introduction of the promoter libraries. The successful introduction was confirmed by sequencing. In conclusion, for each of the three pathway genes (crtI, crtE and crtB ) three promoter libraries were obtained, introduced using either SSA-Gibson, SSA-CPEC-1P or SSA-CPEC2P.
3.3.2
Combinatorial construction of the lycopene pathway and assaying thereof
In the subsequent step the various individual gene libraries, constructed with one of the three SSA methods, were assembled into the final pEX plasmids, i.e., pEX-SC101 and pEX-BR322, to obtain the six complete lycopene pathway libraries. These combinatorial assembled pathway libraries were transformed in two expression hosts, i.e., E. coli K-12 MG1655 DE3 ∆recA ∆endA ∆araA::pT7-MEP (T7-MEP) and E. coli K-12 MG1655 DE3 ∆recA ∆endA ∆araA::pTrc-MEP (Trc-MEP), to evaluate the effect of the genetic background on lycopene production. From each of the resulting 12 combinatorial libraries, 288 individual colonies were randomly picked, cultivated in MTPs and for each colony lycopene production was assessed using a lycopene assay. As reference strains, T7-MEP pEX-SC101 ref, Trc-MEP pEX-SC101 ref, T7-MEP pEX-BR322 ref and Trc-MEP pEX-BR322 ref were incorporated in the experimental design. Figure 4 depicts the outcome of the 12 combinatorial libraries and four reference strains using boxplots. In both background strains the methylerythritol phosphate (MEP) pathway (dxs-idiispDF operon) is genomically expressed from an inducible promoter, either pT7 or pTrc. Upon induction, the MEP operon is expressed, allowing overproduction of isopentenyl pyrophosphate (IPP), a lycopene building block. At the evaluated inducer (IPTG) concentra14 ACS Paragon Plus Environment
Page 14 of 37
Page 15 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
tion, expression from a pT7 is higher compared to that from a pTrc, which also results in an increased lycopene production as can be observed when comparing the appropriate reference strains with the same expression vector (T7-MEP pEX-SC101 ref vs Trc-MEP pEX-SC101 ref and T7-MEP pEX-BR322 ref vs Ttrc-MEP pEX-BR322 ref)(See Figure 4).
15 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
[Figure 4. about here] Analogously, the impact of the used expression vector on lycopene production can be evaluated. Whereas the pMB1 ori (pEX-BR322) yields a medium copy number plasmid, the pEX-SC101 yields a low copy one. Figure 4 shows that this higher copy number results in a higher lycopene production for both reference strains (e.g., T7-MEP pEX-SC101 ref vs T7-MEP pEX-BR322 ref). The here constructed libraries yield ample variation in lycopene production. In addition, good-performing transformants with a significantly higher lycopene production were obtained as compared to the reference strains. The differences between the various constructed libraries is however less clear. No significant trends can be observed neither with respect to the copy number (pSC101 vs. pBR322) nor to the used SSA method (SSAGibson vs. SSA-CPEC-1P vs. SSA-CPEC-2P). In this regard, it is impossible to specifically designate certain differences between the various libraries to one of these parameters, as lycopene production is the complex outcome of the expression vector and host used, and the 3 promoter-gene combinations. An increased sample size, for example by using a more high-throughput screening set-up, as compared to the library size might possibly allow to elucidate the impact of the different parameters on the production variation. Nevertheless, Figure 4 clearly demonstrates that the assembly scheme and the set of pEX and pCP plasmids, are a powerful combinatorial pathway optimization tool, as many producers are obtained with a significantly increased lycopene production compared to their respective reference strains. Figure 5 depicts the five best ranked strains for every expression host - plasmid combination. The best lycopene producer (448 mg lycopene / g CDW) produces twice as much lycopene in comparison with Rad, et al. (198 mg lycopene / g CDW) (39 ). [Figure 5. about here]
16 ACS Paragon Plus Environment
Page 16 of 37
Page 17 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
4
Conclusion
Due to the limitations of rational approaches, combinatorial pathway optimization is becoming increasingly popular. However, besides a high-throughput screen, the latter requires efficient techniques to combinatorially assemble increasingly lengthy metabolic product pathways. Since, none of the existing assembly techniques score well on both introducing sequence variability and the ability to assemble multiple DNA fragments, we introduce, in this contribution, a novel workflow for combinatorial pathway optimization that satisfies both criteria. This scheme exploits the strengths of Single Strand Assembly (SSA) methods, i.e., the ability to efficiently and rapidly introduce sequence variability, and of Golden Gate Assembly, i.e., the ability to successfully assemble multiple DNA fragments into one reaction, by combining both methods into one integrated workflow. To this end, a complete set of ready-to-use carrier pCP and expression pEX plasmids were developed, which allow ordered and direct pathway assembly. These vectors were specifically designed to avoid cross-contamination and false positives. The BsaI cut sites, required for GGA, were designed in silico, with a view to promoting a successful assembly, e.g., palindromic sequences were avoided and their GC content was optimized. In a first step, various pathway gene libraries are constructed using SSA methods. Two novel SSA methods, i.e., SSA-CPEC-1P and SSA-CPEC-2P were introduced to generate sequence diversity using degenerate oligonucleotides and compared to our previous described Gibson-based SSA method. To this end, spacer-randomized promoter libraries were constructed, which were benchmarked against earlier described promoter libraries (12 , 23 ). Though differences in terms of both the library size and distribution could be observed, all of them are suited to effectively and efficiently introduce sequence variability. In a second step the complete pathway is combinatorially assembled using Golden Gate assembly, which allows assembly of multiple DNA fragments in one reaction. To put the fully integrated workflow and the various accompanying sets of pEX and pCP plasmids 17 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
to the test, they were applied for the combinatorial optimization of the 3-gene lycopene pathway, consisting of crtE, crtI, and crtB. The pathway libraries were constructed using the aforementioned SSA methods, and expressed in two different background expression hosts. Consistent and significant differences in overall lycopene production could be observed in function of the expression host and vector for the reference pathway. Such clear differences could not be observed for the combinatorial pathway libraries. This is probably due to multiple factors (copy number, promoter) counterbalancing each other and the relatively small sample size. Nonetheless, huge differences in lycopene production between individual transformants were observed, varying from 0 to a maximum of 448 mg lycopene / g CDW, which is about double the lycopene production reported by Rad, et al. (198 mg lycopene / g CDW) (39 ). In conclusion, the proposed workflow in combination with the accompanying ready-to-use sets of expression and carrier plasmids will allow combinatorial assembly of ever increasingly lengthy product pathways with minimal effort.
5
Author contributions
Pieter Coussement (PC), David Bauwens (DB), Jo Maetens (JM) and Marjan De Mey (MDM) were involved in the conception and design. PC, DB and JM drafted the manuscript. Acquisition of data: PC and DB, data analysis and interpretation: PC. All authors revised the manuscript critically.
Acknowledgement Pieter Coussement was supported by a fellowship of the Institute for the Promotion of Innovation through Science and Technolgy in Flanders (IWT-Vlaanderen). This research was also supported by the ERA-IB project ”ChitoBioEngineering“, the project ”Nano3Bio“ from the EU 7th Framework Programme (FP7/2007-2013) under grant agreement number 18 ACS Paragon Plus Environment
Page 18 of 37
Page 19 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
613931 and by the FWO project CondEx (FWO - G.0321.13N).
19 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 1: Overview of the complete workflow for direct combinatorial pathway engineering. In the first step 1.a. sequence variation is introduced in each step (n) of the pathway. In step 1.b. these individual libraries are combined into a single expression plasmid. Subsequent transformation (2.) and screening (3.) retrieves the strain with the desired phenotype, i.e., carrying the combinatorial optimized pathway.
20 ACS Paragon Plus Environment
Page 20 of 37
1.0
● ● ●
0.6 0.4 0.2 0.0
60
CFUs x1000 [−]
0.8
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
0
50
100
150
200
40
20
0 SSA_Gibson
250
name
Colony number [−]
1.0
● ● ● ● ● ●
0.8
60
●
CFUs x1000 [−]
Norm. fluorescence / OD [−]
(a) Figure
0.6 ● ●
0.4
● ●
0.2 0.0
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
0
50
100
150
200
40
20
0 SSA_CPEC_1P
250
name
Colony number [−] (b) Figure 1.0
●
60
0.8
CFUs x1000 [−]
Norm. fluorescence / OD [−]
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
Norm. fluorescence / OD [−]
Page 21 of 37
0.6 0.4 ●
0.2 0.0
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
0
50
100
150
200
250
40
20
0 300
Colony number [−]
SSA_CPEC_2P
name
(c) Figure
Figure 2: Promoter libraries generated with the SSA method (a.) with corresponding CFU’s of the library, SSA-CPEC-1P (b.) and SSA-CPEC-2P (c.). To facilitate comparison, the library data were normalized. Absolute maxima are 36016, 37745, and 21368 for SSA-Gibson, SSA-CPEC-1P, and SSA-CPEC-2P, respectively. For CFU count, n=3, error bars represent standard deviation.
21 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 3: Lycopene pathway. crtE, crtI, and crtB need to be heterologous expressed as these are not present in the E. coli gene set. In order to achieve significant production the dxs-idi-ispDF operon needs to be overexpressed to ensure sufficient amounts of lycopene building blocks.
22 ACS Paragon Plus Environment
Page 22 of 37
Page 23 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
OD corrected lycopene production (ABS471 / OD600) Plasmid
Construct
Trc-MEP
pEX-SC101
ref
ref
Strain
Trc-MEP
pEX-BR322
Trc-MEP
pEX-SC101
SSA-Gibson
Trc-MEP
pEX-BR322
SSA-Gibson
Trc-MEP
pEX-SC101
SSA-CPEC-1P
Trc-MEP
pEX-BR322
SSA-CPEC-1P
Trc-MEP
pEX-SC101
SSA-CPEC-2P
Trc-MEP
pEX-BR322
SSA-CPEC-2P
T7-MEP
pEX-SC101
ref ref
T7-MEP
pEX-BR322
T7-MEP
pEX-SC101
SSA-Gibson
T7-MEP
pEX-BR322
SSA-Gibson
T7-MEP
pEX-SC101
SSA-CPEC-1P
T7-MEP
pEX-BR322
SSA-CPEC-1P
T7-MEP
pEX-SC101
SSA-CPEC-2P
T7-MEP
pEX-BR322
SSA-CPEC-2P
Figure 4: Boxplots of the reference strains and the generated pathway libraries, based on 288 randomly picked colonies per library, using direct combinatorial pathway optimization. Lycopene output was measured using absorbance at 471 nm and was corrected using OD600 . The table lists the different host strains used for expression, namely Trc-MEP: ∆araA::pTrcMEP; T7-MEP: ∆araA::pT7 -MEP. The column plasmid indicates which pEX plasmid was used either pEX-SC101 or pEX-BR322 with copy numbers of 5 and 20, respectively. The construct column shows what is being expressed, either the reference construct or the pathway libraries, generated using different methods.
23 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 5: For every expression host - plasmid combination, the lycopene production of the five best ranked colonies is plotted and compared with the respective reference strains. Error bars represent standard errors using the 3 characteristic absorbance signals of lycopene (443 nm, 471 nm and 502 nm).
24 ACS Paragon Plus Environment
Page 24 of 37
Page 25 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
References 1. Jullesson, D., David, F., Pfleger, B., and Nielsen, J. (2015) Impact of synthetic biology and metabolic engineering on industrial production of fine chemicals. Biotechnology advances 2. Koch, K., Podlech, J., Pfeiffer, E., and Metzler, M. (2005) Total synthesis of alternariol. The Journal of organic chemistry 70, 3275–3276. 3. Trantas, E. A., Koffas, M., Xu, P., and Ververidis, F. (2015) When plants produce not enough or at all: metabolic engineering of flavonoids in microbial hosts. Frontiers in Plant Science 6 . 4. Fisher, A. K., Freedman, B. G., Bevan, D. R., and Senger, R. S. (2014) A review of metabolic and enzymatic engineering strategies for designing and optimizing performance of microbial cell factories. Computational and Structural Biotechnology Journal 11, 91–99. 5. Choi, S., Song, C. W., Shin, J. H., and Lee, S. Y. (2015) Biorefineries for the production of top building block chemicals and their derivatives. Metabolic Engineering 28, 223–239. 6. Keasling, J. D. (2010) Manufacturing molecules through metabolic engineering. Science (New York, N.Y.) 330, 1355–8. 7. Jones, K. L., Kim, S.-W., and Keasling, J. (2000) Low-copy plasmids can perform as well as or better than high-copy plasmids for metabolic engineering of bacteria. Metabolic engineering 2, 328–338. 8. Kopito, R. R. (2000) Aggresomes, inclusion bodies and protein aggregation. Trends in cell biology 10, 524–530. 9. Seo, S. W., Yang, J.-S., Cho, H.-S., Yang, J., Kim, S. C., Park, J. M., Kim, S., and
25 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Jung, G. Y. (2014) Predictive combinatorial design of mRNA translation initiation regions for systematic optimization of gene expression levels. Scientific reports 4 . 10. Neubauer, P., Lin, H. Y., and Mathiszik, B. (2003) Metabolic load of recombinant protein production: inhibition of cellular capacities for glucose uptake and respiration after induction of a heterologous gene in Escherichia coli. Biotechnology and bioengineering 83, 53–64. 11. Pitera, D. J., Paddon, C. J., Newman, J. D., and Keasling, J. D. (2007) Balancing a heterologous mevalonate pathway for improved isoprenoid production in Escherichia coli. Metabolic engineering 9, 193–207. 12. De Mey, M., Maertens, J., Lequeux, G. J., Soetaert, W. K., and Vandamme, E. J. (2007) Construction and model-based analysis of a promoter library for E. coli: an indispensable tool for metabolic engineering. BMC biotechnology 7, 34. 13. De Mey, M., Maertens, J., Boogmans, S., Soetaert, W. K., Vandamme, E. J., Cunin, R., and Foulqui´e-Moreno, M. R. (2010) Promoter knock-in: a novel rational method for the fine tuning of genes. BMC biotechnology 10, 26. 14. Mutalik, V. K., Guimaraes, J. C., Cambray, G., Lam, C., Christoffersen, M. J., Mai, Q.A., Tran, A. B., Paull, M., Keasling, J. D., Arkin, A. P., and Endy, D. (2013) Precise and reliable gene expression via standard transcription and translation initiation element. Nature methods 10, 1–42. 15. Salis, H. M., Mirsky, E. A., and Voigt, C. a. (2009) Automated design of synthetic ribosome binding sites to control protein expression. Nature biotechnology 27, 946–50. 16. Leonard, E., Ajikumar, P. K., Thayer, K., Xiao, W.-H., Mo, J. D., Tidor, B., Stephanopoulos, G., and Prather, K. L. J. (2010) Combining metabolic and protein engineering of a terpenoid biosynthetic pathway for overproduction and selectivity con-
26 ACS Paragon Plus Environment
Page 26 of 37
Page 27 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
trol. Proceedings of the National Academy of Sciences of the United States of America 107, 13654–9. 17. Gibson, D. G., Young, L., Chuang, R.-y. Y., Venter, J. C., Iii, C. A. H., Smith, H. O., and America, N. (2009) Enzymatic assembly of DNA molecules up to several hundred kilobases. Nature Methods 6, 12–16. 18. Engler, C., Kandzia, R., and Marillonnet, S. (2008) A one pot, one step, precision cloning method with high throughput capability. PLoS ONE 3, e3647. 19. Smanski, M. J., Bhatia, S., Zhao, D., Park, Y., Woodruff, L. B. a., Giannoukos, G., Ciulla, D., Busby, M., Calderon, J., Nicol, R., Gordon, D. B., Densmore, D., and Voigt, C. a. (2014) Functional optimization of gene clusters by combinatorial design and assembly. Nature Biotechnology advance on. 20. Weber, E., Engler, C., Gruetzner, R., Werner, S., and Marillonnet, S. (2011) A modular cloning system for standardized assembly of multigene constructs. PloS one 6, e16765. 21. Sarrion-Perdigones, A., Falconi, E. E., Zandalinas, S. I., Ju´arez, P., Fern´andez-del Carmen, A., Granell, A., and Orzaez, D. (2011) GoldenBraid: an iterative cloning system for standardized assembly of reusable genetic modules. PloS one 6, e21622. 22. Ramon, A., and Smith, H. O. (2011) Single-step linker-based combinatorial assembly of promoter and gene cassettes for pathway engineering. Biotechnology letters 33, 549–55. 23. Coussement, P., Maertens, J., Beauprez, J., Van Bellegem, W., and De Mey, M. (2014) One step DNA assembly for combinatorial metabolic engineering. Metabolic engineering 23, 70–77. 24. Wang, H. H., Isaacs, F. J., Carr, P. A., Sun, Z. Z., Xu, G., Forest, C. R., and Church, G. M. (2009) Programming cells by multiplex genome engineering and accelerated evolution. Nature 460, 894–8. 27 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
25. Ajikumar, P. K., Xiao, W.-H., Tyo, K. E. J., Wang, Y., Simeon, F., Leonard, E., Mucha, O., Phon, T. H., Pfeifer, B., and Stephanopoulos, G. (2010) Isoprenoid pathway optimization for Taxol precursor overproduction in Escherichia coli. Science (New York, N.Y.) 330, 70–4. 26. Klein-Marcuschamer, D., Ajikumar, P. K., and Stephanopoulos, G. (2007) Engineering microbial cell factories for biosynthesis of isoprenoid molecules: beyond lycopene. Trends in biotechnology 25, 417–24. 27. Sambrook, J., Fritsch, E. F., and Maniatis, T. New York ; 1989; pp 931–957. 28. Cunningham, F. X., Sun, Z., Chamovitz, D., Hirschberg, J., and Gantt, E. (1994) Molecular structure and enzymatic function of lycopene cyclase from the cyanobacterium Synechococcus sp strain PCC7942. The Plant Cell 6, 1107–1121. 29. Quan, J., and Tian, J. (2009) Circular polymerase extension cloning of complex gene libraries and pathways. PloS one 4, e6441. 30. Ren, Q., Henes, B., Fairhead, M., and Th¨ony-Meyer, L. (2013) High level production of tyrosinase in recombinant Escherichia coli. BMC Biotechnology 13, 18. 31. R Core Team, R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria, 2015. 32. Maertens, J., and Vanrolleghem, P. A. (2010) Modeling with a view to target identification in metabolic engineering: a critical evaluation of the available tools. Biotechnology progress 26, 313–331. 33. Julsing, M. K., Koulman, A., Woerdenbag, H. J., Quax, W. J., and Kayser, O. (2006) Combinatorial biosynthesis of medicinal plant secondary metabolites. Biomolecular engineering 23, 265–279.
28 ACS Paragon Plus Environment
Page 28 of 37
Page 29 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Synthetic Biology
34. Lukyanov, S., and Chudakov, D. Modified Fluorescent Proteins and Methods for Using Same. 2013; http://www.google.com/patents/US20130344591, US Patent App. 13/912,040. 35. Parikh, A., and Guengerich, F. P. (1998) Random mutagenesis by whole-plasmid PCR amplification. Biotechniques 24, 428–431. 36. Team, R. C. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria, 2015. 37. Gentleman, R. C., Carey, V. J., Bates, D. M., and others, (2004) Bioconductor: Open software development for computational biology and bioinformatics. Genome Biology 5, R80. 38. Engler, C., and Marillonnet, S. (2011) Generation of families of construct variants using golden gate shuffling. Methods in molecular biology (Clifton, N.J.) 729, 167–181. 39. Rad, S. A., Zahiri, H. S., Noghabi, K. A., Rajaei, S., Heidari, R., and Mojallali, L. (2012) Type 2 IDI performs better than type 1 for improving lycopene production in metabolically engineered E. coli strains. World Journal of Microbiology and Biotechnology 28, 313–321.
29 ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Paragon Plus Environment
Page 30 of 37
1 2 3 4 5 6 7 8 9
0.8 0.6 0.4 0.2 0.0
ACS Synthetic Biology
●
● ●
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
ACS Paragon Plus Environment
0
50
100
150
200
Colony number [−]
250
60
CFUs x1000 [−]
Norm. fluorescence / OD [−]
1.0
Page 31 of 37
40
20
0 SSA_Gibson
name
ACS Synthetic Biology
●
● ● ● ● ● ●
0.8 0.6
● ●
0.4
●
0.2 0.0
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
ACS Paragon Plus Environment
0
50
100
150
200
Colony number [−]
250
Page 32 of 37 60
CFUs x1000 [−]
Norm. fluorescence / OD [−] 1 2 3 4 5 6 7 8 9
1.0
40
20
0 SSA_CPEC_1P
name
1 2 3 4 5 6 7 8 9
ACS Synthetic Biology
●
60
0.8
CFUs x1000 [−]
Norm. fluorescence / OD [−]
1.0
Page 33 of 37
0.6 0.4 ●
0.2 0.0
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
ACS Paragon Plus Environment
0
50
100
150
200
Colony number [−]
250
300
40
20
0 SSA_CPEC_2P
name
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
ACS Paragon Plus Environment
Page 34 of 37
Page 35 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
ACS Synthetic Biology
ACS Paragon Plus Environment
ACS Synthetic Biology
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
ACS Paragon Plus Environment
Page 36 of 37
Page 37 of ACS 37 Synthetic Biology
1 2 3 4 5 6
ACS Paragon Plus Environment
