Development of a Synthetic Malonyl-CoA Sensor in Saccharomyces

Subscriber access provided by UNIV OF MISSISSIPPI

Article

Development of a Synthetic Malonyl-CoA Sensor in Saccharomyces cerevisiae for Intracellular Metabolite Monitoring and Genetic Screening Sijin Li, Tong Si, Meng Wang, and Huimin Zhao ACS Synth. Biol., Just Accepted Manuscript • DOI: 10.1021/acssynbio.5b00069 • Publication Date (Web): 06 Jul 2015 Downloaded from http://pubs.acs.org on July 11, 2015

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 32

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

Submitted to ACS Synthetic Biology, revised 6/21/2015

Development of a Synthetic Malonyl-CoA Sensor in Saccharomyces cerevisiae for Intracellular Metabolite Monitoring and Genetic Screening

Sijin Li,1, 2 Tong Si,2 Meng Wang,1,3 and Huimin Zhao1, 2, 3,4*

1

Energy Biosciences Institute, University of Illinois at Urbana-Champaign, Urbana, IL 61801

2

Institute for Genomic Biology, University of Illinois at Urbana-Champaign, Urbana, IL 61801

3

Department of Chemical and Biomolecular Engineering, University of Illinois at Urbana-

Champaign, Urbana, IL 61801 4

Departments of Chemistry, Biochemistry, and Bioengineering, University of Illinois at Urbana-

Champaign, Urbana, IL 61801

* To whom correspondence should be addressed. Phone: (217) 333-2631. Fax: (217) 333-5052. E-mail: [email protected].

1 ACS Paragon Plus Environment


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

ABSTRACT Genetic sensors capable of converting key metabolite levels to fluorescence signals play an important role in monitoring intracellular compound concentrations in living cells, and serve as an efficient tool in high-throughput genetic screening. However, the development of genetic sensors in yeasts lags far behind their development in bacteria. Here we report the design of a malonyl-CoA sensor in Saccharomyces cerevisiae using an adapted bacterial transcription factor FapR and its corresponding operator fapO, which is capable of reporting intracellular malonylCoA levels. By combining this sensor with a genome-wide overexpression library, we identified two novel gene targets that improved intracellular malonyl-CoA concentration. We further utilized the resulting recombinant yeast strain to produce a valuable compound, 3hydroxypropionic acid, from malonyl-CoA and enhanced its titer by 120%. Such a genetic sensor provides a powerful approach for genome-wide screening and could further improve the synthesis of a large range of chemicals derived from malonyl-CoA in yeast.

KEYWORDS: malonyl-CoA, genetic sensor, genome-wide library, high-throughput screening, 3-hydroxypropionic acid


Page 2 of 32

Page 3 of 32

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


Introduction Combinatorial approaches have been used widely for engineering biological systems. Thanks to advances in the design, construction and optimization of large-scale libraries,1 significant progress has been achieved in library generation based on the combination and assembly of various genetic elements to create diversities either on a single gene level or on a pathway level. Representative approaches include designing pathways with enzymes derived from various microorganisms2 and construction of promoter3, ribosome binding site (RBS)4 or intergenic region5 libraries to optimize and balance the metabolic flux. Recently, more attention has been paid to genome-scale engineering6 thanks to the development of sequence-specific gene targeting tools such as zinc-finger nucleases (ZFNs),7 transcription activator-like effector nucleases (TALENs),8 RNAi9 and CRISPR-Cas (clustered regularly interspaced short palindromic repeatsCRISPR associated protein)

technologies.10 In addition, laboratory automation and robotic

systems have enabled rapid construction of a target library.11, 12 Aiming at engineering the model eukaryote and industrial production host Saccharomyces cerevisiae, new approaches named RNAi-assisted genome evolution (RAGE)13 and yeast oligo-mediated genome engineering (YOGE)14 have been developed to generate genomic libraries. Compared to library generation tools, the development of high-throughput screening technologies has lagged far behind the rapid development of phenotype diversity generation technologies, which makes screening a critical bottleneck in metabolic engineering and synthetic biology. Traditional chromatography-based approaches are effective in accurate quantification of target compounds, but the versatility is limited by the low throughput of each experiment and the long and strict sample pretreatment process.15 By contrast, fluorescence-activated cell sorting (FACS) provides a throughput as high as 109 variants per experiment16 and could serve as an



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

Page 4 of 32

attractive tool for high throughput screening, but the range of targets is limited to those affecting light scattering or fluorescence signal levels to enable physical separation. For this reason, not every kind of compounds can be sorted, and a one-of-a-kind tailored gene circuit is required to convert intracellular compound concentrations to signals detectable by FACS. Genetic sensors converting chemical concentrations into fluorescence signals via transcriptional regulation may serve as an important tool for screening and sorting.15,

17

A

successful reconstruction of a functional genetic sensor in a heterologous host requires wellcharacterized exogenous biological parts (transcription factors, operators, reporters, etc.), wellcharacterized endogenous biological parts (promoters, terminators, nuclear localization sequences, etc.), and a clear understanding of the host including transcriptional regulation network and protein expression system. Multiple transcriptional-regulation-based sensors have been developed to detect various target molecules and to screen for desired phenotypes in bacteria. 18-20 However, the development of genetic sensors in S. cerevisiae is not as prevalent as in E. coli, mainly because of the lack of available biological parts, the complexity of the transcriptional regulation network, and the lack of clear understanding of the transcriptional regulation network. Endogenous promoters responsive to specific target compounds are rare and even the native regulated promoters lack orthogonality in the native regulatory networks. Synthetic promoters with inserted heterologous operators are more versatile, but the construction of synthetic promoters rely on the well-characterized native promoters, which are limited in a small group such as TEF1, CYC1, ADH1, CMV and GAL promoters.21 The lack of versatile transcription factors also limits the successful construction of genetic sensor in S. cerevisiae. Instead of repurposing the native transcriptional regulation network in eukaryotes, well-


Page 5 of 32

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


characterized transcription factors derived from prokaryotes are used, which require extra efforts in heterologous expression and nuclear import. In terms of complexity, at least 4000 regulatory interactions exist in S. cerevisiae system while 1468 exist in E. coli system.22 Surprisingly, though S. cerevisiae is a model eukaryote, it remains challenging to understand its regulatory network. In the past decade, only about a dozen regulatory network engineering projects related to S. cerevisiae were published,21 which is not only much fewer than that in E. coli but also than that in mammalian cells. Due to these reasons, only a few successful examples in S. cerevisiae have been reported. For example, Umeyama and coworkers23 utilized the metO operator and MetJ repressor of E. coli in S. cerevisiae to report intracellular S-adenosylmethionine (SAM) concentrations, and succeeded in screening for SAM production enhancers from a plasmid library carrying random genomic DNA fragments. Teo and coworkers developed sensors responsive to xylose24 or fatty acid/fatty acyl-CoA25 in S. cerevisiae using bacterial repressors. In addition, adaptation of genetic sensors across different eukaryotes remains challenging. Though the fundamental functions of transcriptional regulation are similar in yeast and mammalian cells, basic building blocks of the regulatory network are not usually evolutionary conserved and metabolic reactions differ a lot.22 Some properties of S. cerevisiae hinder the simple transfer of genetic sensors between yeast and mammalian cells. For example, endogenous homologous recombination makes operators with repetitive sequences unstable. Intron density and synonymous codon usage bias require de novo design of the transcription factors. Different nuclear localization sequence (NLS) varies the binding rate of transcription factor to the operator. It has been reported that the simple transfer of a TetR-based gene circuit between yeast and mammalian cells could not work until tremendous modifications had been applied.26



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

Page 6 of 32

Therefore, we focused our effort on the development of a genetic sensor responsive to malonyl-CoA in S. cerevisiae and the enhancement of malonyl-CoA production by genome editing using a genome-wide library. Malonyl-CoA, which is a basic building block for the biosynthesis of valuable compounds such as fatty acids, polyketides, and flavonoids.27,

28

Traditional analytical approaches quantifying malonyl-CoA levels in eukaryotes such as LCMS29 or immunoassays30 require large amounts of cells, complicated sample preparation and long analysis time, which hinders rapid screening or fermentation process monitoring. Ellis and Wolfgang31 successfully constructed a malonyl-CoA sensor in mammalian cells based on a bacterial transcription factor FapR and its corresponding operator fapO from B. subtilis,32 where FapR undergoes a conformational change when it specifically binds to malonyl-CoA and dissociates from its 34 bp operator fapO, allowing access of the RNA polymerase.33 In vitro transcription assay showed FapR repression is specifically relieved by malonyl-CoA rather than other short chain acyl-CoAs (acetyl-CoA, propionyl-CoA, succinyl-CoA and butyryl-CoA).33 Such features make FapR-fapO a desirable system for the construction of malonyl-CoA specific sensors. Sensors using the FapR-fapO system have also been reported in E. coli.34, 35 However, there are no reports about the construction of a FapR-based malonyl-CoA sensor in S. cerevisiae to date. Here we report the development of a malonyl-CoA sensor in S. cerevisiae. We designed the prototype with a codon-optimized FapR and a synthetic promoter, and constructed a functional sensor in S. cerevisiae after a series of optimization on nuclear transport, repressor/operator ratio and operator localization. To demonstrate its utility, we combined this malonyl-CoA sensor with a genomic cDNA library to improve the production of 3hydroxypropionic acid (3-HP) in S. cerevisiae. By screening the genome-wide library, we


Page 7 of 32

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


successfully identified two gene targets, TPI1 and PMP1, which enhanced intracellular malonylCoA levels. The resultant strains containing a 3-HP biosynthetic pathway showed improved 3HP production, and the maximal titer was enhanced by 120%.

Results and Discussion Design and Characterization of the Malonyl-CoA Sensor in S. cerevisiae The FapR-fapO system was used to design a malonyl-CoA sensor in S. cerevisiae (Figure 1a). The FapR gene was codon-optimized to assure successful expression of this bacterial transcription factor in S. cerevisiae, and fused with a strong NLS SV40 at the C-terminus of the FapR protein to enable nuclear import. The codon-optimized FapR gene was assembled in a plasmid with a TEF1 promoter and an ADH1 terminator. To provide a quantitative readout of FapR-fapO behavior, a single plasmid harboring the fluorescent protein tdTomato as a reporter was constructed. The FapR-regulated fluorescent protein expression cassette was constructed by inserting the fapO operator into an engineered promoter GPM1 in front of tdTomato. The GPM1 promoter was modified with only one upstream activation sequence conserved to eliminate native transcriptional regulation (Figure 1b). The fapO sequence was inserted immediately in front of the TATA box, which was proven to have better induction ratio and more sensitive doseresponse curve compared to other locations.36 Functional nuclear import is a big challenge hindering successful sensor construction in S. cerevisiae. There are two NLS tags commonly used in S. cerevisiae nuclear import: SV40 and cMyc, and we chose the strong SV40 for effective nuclear import. To confirm the functional nuclear import of FapR, a FapR cassette without an NLS was also constructed. Another factor taken into concern is the ratio of FapR and fapO which may affect the sensitivity and dynamic 7 ACS Paragon Plus Environment


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

range of the sensor significantly. Here we used a strong constitutive promoter TEF1 and a codonoptimized FapR to ensure the sufficient expressionof the repressor. To further investigate the effect of different concentrations of FapR on the sensor’s behavior, we used both a single copy plasmid pRS414 and a multi-copy plasmid pRS424 to express relatively low level and high level of FapR and compared their effects. To further investigate the effect of different levels of fapO on the sensor’s behavior, we inserted a single fapO operator or a (fapO)2 operator immediately before the TATA box in the engineered GPM1 promoter and compared their effects. Strains harboring plasmids expressing FapR-NLS, FapR without NLS or blank cassette together with the GPM1-fapO promoter-tdTomato plasmid were compared by fluorescence intensities. The malonyl-CoA sensor strain harboring both pRS424-FapR-NLS and pRS415fapO-tdTomato exhibited only 30.3% of the fluorescence intensity of the control strain, while the strain expressing FapR without NLS showed almost unaffected fluorescence intensity. This significant difference indicates the SV40 NLS induces functional nuclear import and plays an essential role in the sensor (Figure 2a). To validate this sensor, we overexpressed the endogenous acetyl-CoA carboxylase gene (ACC1) in the sensor strain to increase the intracellular malonylCoA concentration.37 The fluorescence intensity was enhanced by 116.2% than the control strain harboring an empty plasmid, indicating the sensor’s positive response to malonyl-CoA increase (Figure 2b). As FapR is known to specifically bind to malonyl-CoA rather than other short chain acyl-CoAs, we investigated the sensor’s response to long chain acyl-CoA derivatives, including myristoyl-CoA and palmitoyl-CoA. We compared the fluorescence intensity of the sensor strain cultured in synthetic dropout medium supplemented with 1 g/L myristic acid or palmitic acid, which can be converted to corresponding acyl-CoAs intracellularly.25, 38 Increased long chain acyl-CoA concentrations did not induce any obvious response. The sensor strain and control 8 ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32

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


strain showed similar response: palmitic acid did not induce any response and the fluorescence ratio was not altered; myristic acid induced same level of fluorescence ratio decrease in the sensor and control strains (Figure 2c). The decreased ratio can be due to the limited oxygen exchange during cultivation because excessive myristic acid remained insoluble in the medium. To validate that the sensors are capable of reporting quantitative change of intracellular malonyl-CoA levels and to compare the effect of different FapR/fapO ratios, we sought to determine the dose-dependent response curves. It is highly likely that the engineered sensor was activated by nuclear malonyl-CoA, whose concentration correlated with that of cytosolic malonyl-CoA,24,

25

thus it is necessary to alter the intracellular malonyl-CoA concentration

gradiently. However, such goal is difficult to achieve by exogenous malonyl-CoA or genetic engineering approaches. Therefore we used cerulenin, a chemical inhibitor known to block fatty acid elongation,39 to alter the intracellular malonyl-CoA concentration, and to establish the dose dependent response. Meanwhile, malonyl-CoA is a key intermediate with low cytosolic concentration, which makes the direct extraction and quantification process in S. cerevisiae quite challenging. So far there have not been any effective approaches to extract and quantify cytosolic malonyl-CoA in S. cerevisiae CEN.PK2 strain yet. However, it has been demonstrated that the level of malonyl-CoA was positively correlated with the level of cerulenin added in the culture medium.34 Thus the response of a sensor to malonyl-CoA was reflected by the ratio of fluorescence intensity of the culture with or without cerulenin. The sensor strain with relatively high FapR level and FapR/fapO ratio exhibited only 29.8% of the fluorescence intensity of the control strain, and the strain with relatively low FapR level exhibited 50.4%. The different fluorescence intensities of the two versions of sensors indicate that the strong TEF1 promoter enables adequate FapR expression and consequent 9 ACS Paragon Plus Environment


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

Page 10 of 32

adequate repression on fluorescent protein expression, while higher level of FapR is correlated with less “leaky” expression (Figure 3a). In the high-level FapR version, the ratio of fluorescence intensity increased from 1 to 4.17, which was positively correlated with the level of cerulenin added in the culture medium ranging from 0 mg/L to 12 mg/L, indicating the accumulation of malonyl-CoA in the cell (Figure 3b). The dose-dependent response was tightly correlated to the cerulenin concentration ranging from 0 mg/L to 8 mg/L, and saturated after the cerulenin concentration was above 8mg/L. The low-level FapR version also showed a similar response from 0 mg/L to 8 mg/L, but the ratio was lower than that of the high-level version (Figure 3b). It is reasonable because the low concentration of FapR resulted in high background “leaky” expression (Figure 3a). We also investigated the effect of different fapO copy numbers. Replacing the single fapO sequence with a (fapO)2 sequence in the engineered Gpm1 promoter resulted in a lower fluorescence intensity, which was only 48.5% of the original 424FapR-fapO construct. Combination of pRS414-FapR-NLS and (fapO)2 yielded the lowest FapR/fapO ratio, while the fluorescence intensity of the 414FapR-(fapO)2 construct was higher than that of the 424FapR(fapO)2 construct due to “leaky” expression.

Meanwhile, the fluorescence intensity of the

control-(fapO)2 construct without FapR expression was also lower than that of the control-fapO construct, which was 65.2% (Figure 3c). It is likely that the (fapO)2 sequence impaired the promoter stability and lowered its transcriptional activity, resulting in a lower fluorescence intensity. The 424FapR-(fapO)2 construct showed a relatively lower dose-dependent response to malonyl-CoA concentration than the 424FapR-fapO construct. The ratio of fluorescence intensity increased from 1 to 3.69, and plateaued after the cerulenin concentration was above 8 mg/L. We speculate that the extra copy of fapO resulted in stronger transcriptional repression, 10 ACS Paragon Plus Environment

Page 11 of 32

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


and the intracellular malonyl-CoA concentration was not sufficient enough to alleviate the repression when multiple operator sites were occupied by FapR. The 414FapR-(fapO)2 construct showed a low and rapid response to the increased malonyl-CoA concentration, and quickly plateaued (Figure 3d). We speculate that is because the low FapR/fapO ratio led to leaky expression and fast FapR saturation. At the same time, the control construct expressing tdTomato and pRS424-blank showed slightly decreased fluorescence in the presence of cerulenin, which is likely due to the cellular stress response (Figure 3b and 3d). As a result, we concluded that all the different designs based on FapR/fapO ratio could reflect the change of intracellular malonyl-CoA concentrations, and the 424FapR-fapO design could serve as a better sensor because of the relatively low background noise and broad dynamic range. Therefore, this sensor construct was chosen for further screening.

Screening and Isolation of Malonyl-CoA Overproducing Strains from a Genome-wide Library To demonstrate its utility in metabolic engineering, we used our malonyl-CoA sensor in combination with a genome-wide cDNA overexpression library and to screen for yeast strains overproducing malonyl-CoA. Specifically, we constructed a library of yeast strains by transforming plasmids containing a library of gene overexpression cassettes together with the sensor plasmids and performed the screening by FACS. In each round, cells with enhanced fluorescence intensity were isolated and the corresponding plasmids conferring enhanced fluorescence intensity were extracted and retransformed. After 3 rounds of FACS, cells with enhanced fluorescence intensity were spread on plate. 72 colonies were randomly picked and cultured in liquid SC dropout medium, and 20 constructs with enhanced fluorescence intensity 11 ACS Paragon Plus Environment


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

Page 12 of 32

were picked for plasmid isolation and gene sequencing. These 20 constructs contained 5 different genes and showed over 2-fold higher fluorescence intensity (Figure 4a). The corresponding plasmids were retransformed and the fluorescence intensities of the resulting constructs were analyzed to eliminate false positives (Figure 4b). Six (6) of them that contained 3 different gene fragments (ERV14, LSC1 and SUF2) turned out to be false positives. Out of the 14 positives, 11 constructs contained an overexpression cassette for the same fragment of the PMP1 gene, and 3 constructs contained an overexpression cassette for the same fragment of the TPI1 gene. PMP1 encodes a plasma membrane proteolipid that regulates the plasma membrane proton-ATPase, and TPI1 encodes triose-phosphate isomerase that catalyzes the inter-conversion of dihydroxyacetone phosphate (DHAP) and glyceraldehyde-3-phosphate (GAP) during glycolysis. Overexpression of PMP1 or TPI1 both increased the fluorescence intensity. It should be noted that additional gene targets can be discovered by continuing the screening process and by analyzing more constructs.

Overexpression of TPI1 and PMP1 Increased the Production of 3-Hydroxypropionic Acid We utilized the newly discovered gene targets to enhance the production of 3-HP, an attractive value-added chemical that serves as the precursor to a series of chemicals such as acrylates.40 One approach to synthesize 3-HP in bacteria or yeast utilizes malonyl-CoA as the basic substrate to produce the target compound. A bi-functional enzyme caMCR from Chloroflexus aurantiacus acts as both an NADPH-dependent malonyl-CoA reductase and a 3-hydroxypropionate dehydrogenase, which converts malonyl-CoA to malonic-semialdehyde and then to 3-HP.41-43 However, a vast amount of optimization work would be required to enhance the 3-HP production, mainly due to the limited supply of malonyl-CoA. In the first step of our engineering


Page 13 of 32

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


approach, we investigated the effect of the 2 single genetic modifications by overexpressing PMP1 and TPI1 individually in a pRS423 plasmid, and found highly enhanced production of 3HP. Overexpression of PMP1 enhanced the titer by 100% at 72h, with maximal enhancement of 116% at 24h. Overexpression of TPI1 enhanced the titer by 120% at 72h (Figure 5b). Corresponding to the enhanced 3-HP production, the lower cell densities of the PMP1 and TPI1 overexpression strains reflected the repressed cell growth capability due to the competition between malonyl-CoA synthesis and cell growth (Figure 5a). We further combined the two overexpression targets to improve 3-HP production, but without any success (Figure 5d). In fact, the 3-HP titer was only enhanced by 47% at 72h, indicating an antagonistic effect due to the cooverexpression of PMP1 and TPI1. PMP1 is a small single-membrane span proteolipid that is a key regulatory subunit of the proton-ATPase PMA1.44

45

PMA1 is an important proton pump which regulates the pH of the

cytoplasm and generates the electrochemical proton gradient driving secondary active transport of nutrients,46 for example, biotin. Biotin, or vitamin H, is a cofactor essential for carboxylation reactions in S. cerevisiae. However, the biosynthesis pathway of biotin in S. cerevisiae is defective and its biotin supply relies entirely on the uptake of exogenous biotin from the medium. The biotin transporter in S. cerevisiae (VHT1) has been identified as a proton-biotin symporter enabling intracellular biotin accumulation.47, 48 During malonyl-CoA synthesis catalyzed by an acetyl-CoA carboxylase, biotin serves as a vector transferring the carboxyl-group from the donor bicarbonate to the acceptor acetyl-CoA to form malonyl-CoA.49,

50

We conclude that the

constitutive overexpression of PMP1 facilitates the forming of a proton gradient, thus increasing the rate of malonyl-CoA synthesis by accelerating biotin transportation.



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

Page 14 of 32

TPI1 catalyzes the inter-conversion between DHAP and GAP in glycolysis.51 Although this step of glycolysis does not involve any cofactors, the following glycolysis reactions converting GAP to pyruvate produces 2 molecules of ATP. Conversely, the glycerol synthesis pathway converting DHAP to glycerol does not produce ATP. It has been reported that the absence of TPI1 activity induces a strong accumulation of DHAP that leads to glycerol overproduction.52, 53 As a result, it is very likely that the constitutive overexpression of TPI1 enhances the production of GAP, resulting in more metabolic flux redirected and more ATP produced. Malonyl-CoA synthesis requires 1 molecule of ATP per molecule of malonyl-CoA, and the extra supply of ATP in the cytoplasm via glycolysis is beneficial to malonyl-CoA overproduction. Taken together, we could also explain the antagonistic effect of PMP1 and TPI1 cooverexpression. TPI1 requires neutral cytoplasmic pH environment for optimal activity. However, the overexpression of PMP1 helps to build up a proton gradient by pumping out protons. Although the VHT1 transporter co-transports biotin and proton into cytoplasm and complements the loss of protons, the transport efficiency is limited by the expression level of VHT1. As a result, the effect of overexpressed PMP1 on cytoplasmic pH may exceed the effect of VHT1, and the cytoplasmic pH may be altered from neutral to basic. Although TPI1 is overexpressed, the overall activity of TPI1 in the co-expression strain may not be as efficient as in the TPI1 overexpression strain because of the basic pH condition. Meanwhile, the heavy metabolic burden of simultaneously overexpressing TPI1 and PMP1 may lower the production of malonyl-CoA. In general, overexpression of PMP1 or TPI1 may enhance malonyl-CoA production via two different mechanisms, but the mechanisms are incompatible with each other.


Page 15 of 32

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


Thus the co-overexpression of PMP1 and TPI1 showed an antagonistic effect instead of further enhancement on malonyl-CoA production.

Conclusions We successfully designed a malonyl-CoA sensor in S. cerevisiae using FapR and fapO from bacterium. This sensor allows us to convert intracellular malonyl-CoA levels to fluorescence signals, enabling simpler and faster monitoring and screening in S. cerevisiae. This novel malonyl-CoA sensor in S. cerevisiae serves as a useful tool for providing new mechanistic insight into the production of this metabolic intermediate. As proof of concept, we combined this sensor with a genome-wide cDNA overexpression library to improve the intracellular malonylCoA levels. Altering key metabolite levels in yeast remains a challenging task because of the highly-regulated metabolic network, vast putative gene targets, and the lack of efficient screening/selective

tools.

Existing

strategies

to

increase

intracellular

malonyl-CoA

concentrations focus on the activity of ACC1 and related regulatory proteins54 or on increasing and coupling the supply of acetyl-CoA and NADPH,42 which are both directly related to acetylCoA carboxylation. By virtue of the integrated sensor-genetic screening, we succeeded in discovering two novel gene targets, PMP1 and TPI1, and achieved over 100% improvement of 3HP production. Thus this sensor provides a novel way to study the complex malonyl-CoA metabolism and to enhance the production of fatty acids, lipids and other valuable compounds.

Methods Strains, Media, and Cultivation Conditions



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

Page 16 of 32

Saccharomyces cerevisiae CEN.PK2-1C (MATa; ura3-52; trp1-289; leu2-3,112; his3∆1; MAL28C; SUC2) was used for gene cloning and sensor validation. 3-HP producing strain ST68755 was a gift from Dr. Irina Borodina (Technical University of Denmark, Denmark) and was engineered with the LEU2 and HIS3 marker rescued. Escherichia coli DH5α was used for recombinant DNA manipulation. Recombinant yeast strains were cultivated in synthetic complete (SC) dropout media consisting of 0.17% yeast nitrogen base without amino acids and ammonium sulfate, 0.5% ammonium sulfate, 0.05% amino acid dropout mix (MP Biomedicals, Solon, OH) supplemented with 2% D-glucose. E. coli strains were grown in Luria-Bertani broth (Fisher Scientific, Pittsburgh, PA) supplemented with 100 µg/mL ampicillin. S. cerevisiae strains were grown in baffled shake-flasks at 30 °C and 250 rpm for aerobic growth. E. coli strains were grown at 37 °C and 250 rpm. All other chemicals not specified were purchased from Sigma-Aldrich (St. Louis, MO).

DNA Manipulation The FapR gene was codon-optimized and synthesized by GenScript (Piscataway, NJ). The FapR gene with or without an SV40 sequence at the 3’ end was assembled with the TEF1 promoter and the HXT7 terminator into the pRS414 or pRS424 plasmid to yield pRS414-FapR-NLS, pRS424-FapR-NLS or pRS424-FapR. The tdTomato gene was cloned from the ptdTomato plasmid purchased from Clontech (Mountain View, CA) and assembled with the GPM1p-fapO or GPM1p-(fapO)2 promoter and the ADH1 terminator into pRS425 to yield pRS415fapOtdTomato or pRS415-(fapO)2-tdTomato. The yeast homologous recombination based DNA assembler method was used to construct the sensor plasmids and candidate genes.56 Briefly, polymerase chain reaction (PCR) was used to generate DNA fragments with homology arms at


Page 17 of 32

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


both ends, which were purified with a QIAquick Gel Extraction Kit (Qiagen, Valencia, CA) and co-transformed along with the linearized backbone into S. cerevisiae. All genetic elements including promoters, coding sequences, and terminators were PCR-amplified from their corresponding genomic DNAs. The Wizard Genomic DNA Purification Kit (Promega, Madison, WI) was used to extract the genomic DNA from various yeast species, according to the manufacturer’s protocol. The resultant yeast plasmids were isolated using Zymoprep Yeast Plasmid Miniprep II Kit (Zymo Research, Irvine, CA) and re-transformed into E. coli DH5α competent cells, according to the manufacturer’s protocols. Single colonies of the E. coli transformants were then inoculated into LB liquid media. Plasmids were isolated from E. coli using the QIAprep Spin Miniprep Kit (QIAGEN, Valencia, CA) and checked by diagnostic PCR or restriction digestion. All restriction enzymes were obtained from New England Biolabs (Ipswich, MA). These plasmids were transformed into the CEN.PK2-1C strain using the standard lithium acetate method.57 The resultant transformation mixtures were selected on SC dropout plates supplemented with 2% glucose.

Sensor Activity Assays Fluorescence intensity was used to evaluate the signal strengths. Host cells transformed with different sensor plasmids were grown overnight in 2 mL SC dropout media supplemented with 2% glucose in 14 mL culture tubes at 30 °C and 250 rpm agitation. 200 µL of the overnight culture was inoculated into 2 mL fresh SC media supplemented with 2% glucose and different amounts of cerulenin if necessary in 14 mL culture tubes at 30 °C, 250 rpm for approximately 24 hour. For long chain acyl-CoA response assays, 200 µL of the overnight culture was inoculated into 2 mL fresh SC media supplemented with 2% glucose and 1 g/L myristic acid or 1g/L palmitic acid



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

Page 18 of 32

in 14 mL culture tubes at 30 °C, 250 rpm for approximately 24 hour. Subsequently, 20 µL cell culture was transferred to 180 µL water in a Corning 96-well, clear bottom fluorescence plate (ThermoFisher Scientific, Rockford, IL). Cell density and the tdTomato fluorescent signal were simultaneously detected using a Tecan Infinite M1000 PRO microplate reader (Morrisville, NC). Cell density was read at 600 nm and the excitation and emission wavelengths for tdTomato were set at 559 ± 20 nm and 581 ± 20 nm, respectively. Fluorescence intensity (a.u.) was normalized to cell density. All experiments were performed in triplicates.

Library Creation and High-throughput Screening A normalized cDNA library was prepared as follows. The total RNAs from the CEN.PK2-1c strain were isolated using the RNeasy mini kit (QIAGEN, Valencia, CA). A cDNA library was synthesized using the In-Fusion SMARTer Directional cDNA Library Construction Kit (Clontech Laboratories, Mountain View, CA), while the double-stranded cDNA library was treated using the Trimmer-2 cDNA normalization kit (Evrogen, Moscow, Russia) according to the manufacturer’s instructions. A TEF1 promoter-PmeI-PGK1 terminator expression cassette was assembled into a pRS416 plasmid to make the helper plasmid. The cDNA library was cloned into the linearized helper plasmid using the Gibson assembly cloning kit (New England Biolabs), and the mix was electroporated into E. coli DH5α strain and amplified on LB plate. More than 106 independent E. coli colonies were obtained.

The library plasmids were extracted and

transformed into the sensor strain using the standard lithium acetate method.57 After transformation, the yeast cells were recovered at 30 °C with shaking for 1 h in YPA medium (1% yeast extract, 2% peptone and 0.01% adenine hemisulfate) with 2% glucose. The cells were then centrifuged and resuspended in SC medium with 2% glucose for growth at 30 °C with shaking


Page 19 of 32

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


overnight. On the next day, cells were analyzed on a BD FACS Aria III cell sorting system (San Jose, CA) and 4000-10000 tdTomato-positive cells with high fluorescence intensity were collected. After plating on solid SC medium with 2% glucose at 30 °C for 2 days, colonies were separated and plasmids were extracted using the Zymoprep Yeast Plasmid Miniprep II Kit (Zymo Research, Orange, CA). The plasmids were then electroporated into DH5α competent cells for amplification. The extracted plasmids were then subjected to the next round of screening and overexpression. Confirmed plasmids were sequenced to identify the origin of the gene fragments. Identified genes were assembled into a pRS423 plasmid with a TEF1 promoter and a PGK1 terminator.

3-HP Production and Quantification Cells with the identified PMP1 gene or TPI1 gene assembled in pRS423 plasmids were inoculated into 2 mL fresh SC media with 2% glucose in 14 mL culture tubes at 30 °C, 250 rpm for 48 hour. Subsequently, seed cultures were washed and transferred into 25 mL fresh SC media with 2% glucose in a 125 mL baffled shaker flask at an initial OD600 of about 1, and cultured under aerobic conditions for 72 h. Cells harboring a pRS425 plasmid together with a pRS423 plasmid with the PMP1 cassette and the TPI1 cassette respectively were also cultured to study the effect of the combination of these two overexpression targets. Cell density was read at 600 nm in a Tecan Infinite M1000 PRO microplate reader (Morrisville, NC) using a Corning 96-well, clear bottom fluorescence plate (ThermoFisher Scientific, Rockford, IL). 3-HP concentration was determined using Shimadzu HPLC equipped with a Bio-Rad HPX-87H column (Bio-Rad Laboratories, Hercules, CA), a Shimadzu RID-10A refractive index detector (Shimadzu Scientific Instruments, Columbia, MD) and a Shimadzu SPD-10A UV/Vis detector (Shimadzu



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

Page 20 of 32

Scientific Instruments, Columbia, MD) following the manufacturer’s protocol. The HPX-87H column was kept at 65°C using a Shimadzu CTO-20AC column oven. 0.5 mM sulfuric acid solution was used as a mobile phase at a constant flow rate of 0.6 mL/min. 50 µL of filtered sample was injected into the HPLC system with a Shimadzu SIL-20AC HT auto sampler, and each run was stopped at 25 minutes after the injection. The concentrations were determined using a standard curve generated using a series of external standards of 3-HP purchased from Sigma-Aldrich (St. Louis, MO). All experiments were performed in triplicates.

Author Information SL and HZ designed the study, analyzed data, and wrote the manuscript. SL and MW performed the experiments. All authors have read and approved the final manuscript.

Acknowledgements This work was supported by the Energy Biosciences Institute. Dr. Han Xiao and Jiazhang Lian are acknowledged for their help in troubleshooting and discussions. Especially, we appreciate Dr. Irina Borodina for providing the ST687 strain for 3-HP production. We also thank Dr. Barbara Pilas and Dr. Lucas Li at the Roy J. Carver Biotechnology Center, and Dr. Jianping Wang at the Energy Biosciences Institute for their assistance in FACS and LC-MS analysis.

Conflict of Interest Disclosure The author’s claim no conflict of interest.


Page 21 of 32

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


References [1] Keasling, J. D. (2010) Manufacturing molecules through metabolic engineering, Science 330, 1355-1358. [2] Kim, B., Du, J., Eriksen, D. T., and Zhao, H. (2013) Combinatorial design of a highly efficient xylose-utilizing pathway in Saccharomyces cerevisiae for the production of cellulosic biofuels, Appl. Environ. Microbiol. 79, 931-941. [3] 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 Biotechnol. 7, 34. [4] Salis, H. M., Mirsky, E. A., and Voigt, C. A. (2009) Automated design of synthetic ribosome binding sites to control protein expression, Nat. Biotechnol. 27, 946-950. [5] Pfleger, B. F., Pitera, D. J., Smolke, C. D., and Keasling, J. D. (2006) Combinatorial engineering of intergenic regions in operons tunes expression of multiple genes, Nat. Biotechnol. 24, 1027-1032. [6] Si, T., Xiao, H., and Zhao, H. (2014) Rapid prototyping of microbial cell factories via genome-scale engineering, Biotech. Adv., dx.doi.org/10.1016/j.biotechadv.2014.1011.1007. [7] Urnov, F. D., Rebar, E. J., Holmes, M. C., Zhang, H. S., and Gregory, P. D. (2010) Genome editing with engineered zinc finger nucleases, Nat. Rev. Genet. 11, 636-646. [8] Sun, N., and Zhao, H. (2013) Transcription activator-like effector nucleases (TALENs): A highly efficient and versatile tool for genome editing, Biotechnol. Bioeng. 110, 18111821. [9] Na, D., Yoo, S. M., Chung, H., Park, H., Park, J. H., and Lee, S. Y. (2013) Metabolic engineering of Escherichia coli using synthetic small regulatory RNAs, Nat. Biotechnol. 31, 170-174. [10] Cong, L., Ran, F. A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P. D., Wu, X., Jiang, W., and Marraffini, L. A. (2013) Multiplex genome engineering using CRISPR/Cas systems, Science 339, 819-823. [11] Linshiz, G., Stawski, N., Poust, S., Bi, C., Keasling, J. D., and Hillson, N. J. (2012) PaRPaR laboratory automation platform, ACS Synth. Biol. 2, 216-222. [12] Chao, R., Yuan, Y., and Zhao, H. (2014) Recent advances in DNA assembly technologies, FEMS Yeast Res. 15, 1-9. [13] Si, T., Luo, Y., Bao, Z., and Zhao, H. (2014) RNAi-assisted genome evolution in Saccharomyces cerevisiae for complex phenotype engineering, ACS Synth. Biol. 4, 283291. [14] DiCarlo, J. E., Conley, A. J., Penttilä, M., Jäntti, J., Wang, H. H., and Church, G. M. (2013) Yeast oligo-mediated genome engineering (YOGE), ACS Synth. Biol. 2, 741-749. [15] Dietrich, J. A., McKee, A. E., and Keasling, J. D. (2010) High-throughput metabolic engineering: advances in small-molecule screening and selection, Annu. Rev. Biochem 79, 563-590. [16] Becker, S., Schmoldt, H.-U., Adams, T. M., Wilhelm, S., and Kolmar, H. (2004) Ultra-highthroughput screening based on cell-surface display and fluorescence-activated cell sorting for the identification of novel biocatalysts, Curr. Opin. Biotechnol 15, 323-329. [17] Zhang, F., and Keasling, J. (2011) Biosensors and their applications in microbial metabolic engineering, Trends Microbiol. 19, 323-329. 21 ACS Paragon Plus Environment


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

Page 22 of 32

[18] Dietrich, J. A., Shis, D. L., Alikhani, A., and Keasling, J. D. (2012) Transcription factorbased screens and synthetic selections for microbial small-molecule biosynthesis, ACS Synth. Biol. 2, 47-58. [19] Tang, S.-Y., and Cirino, P. C. (2011) Design and application of a mevalonate-responsive regulatory protein, Angew. Chem. Int. Ed. 50, 1084-1086. [20] Siedler, S., Stahlhut, S. G., Malla, S., Maury, J., and Neves, A. R. (2014) Novel biosensors based on flavonoid-responsive transcriptional regulators introduced into Escherichia coli, Metab.Eng. 21, 2-8. [21] Blount, B. A., Weenink, T., and Ellis, T. (2012) Construction of synthetic regulatory networks in yeast, FEBS Lett. 586, 2112-2121. [22] Herrgård, M. J., Covert, M. W., and Palsson, B. Ø. (2004) Reconstruction of microbial transcriptional regulatory networks, Curr. Opin. Biotechnol 15, 70-77. [23] Umeyama, T., Okada, S., and Ito, T. (2013) Synthetic gene circuit-mediated monitoring of endogenous metabolites: identification of GAL11 as a novel multicopy enhancer of Sadenosylmethionine level in yeast, ACS Synth. Biol. 2, 425-430. [24] Teo, W. S., and Chang, M. W. (2015) Bacterial XylRs and synthetic promoters function as genetically encoded xylose biosensors in Saccharomyces cerevisiae, Biotechnol. J. 10, 315-322. [25] Teo, W. S., Hee, K. S., and Chang, M. W. (2013) Bacterial FadR and synthetic promoters function as modular fatty acid sensor- regulators in Saccharomyces cerevisiae, Eng. Life Sci. 13, 456-463. [26] Nevozhay, D., Zal, T., and Balázsi, G. (2013) Transferring a synthetic gene circuit from yeast to mammalian cells, Nat. Commun. 4, 1451. [27] Xu, P., Ranganathan, S., Fowler, Z. L., Maranas, C. D., and Koffas, M. A. (2011) Genomescale metabolic network modeling results in minimal interventions that cooperatively force carbon flux towards malonyl-CoA, Metab.Eng. 13, 578-587. [28] Liu, T., Vora, H., and Khosla, C. (2010) Quantitative analysis and engineering of fatty acid biosynthesis in E. coli, Metab. Eng. 12, 378-386. [29] Gao, L., Chiou, W., Tang, H., Cheng, X., Camp, H. S., and Burns, D. J. (2007) Simultaneous quantification of malonyl-CoA and several other short-chain acyl-CoAs in animal tissues by ion-pairing reversed-phase HPLC/MS, J. Chromatogr. B 853, 303-313. [30] López, M., Lelliott, C. J., Tovar, S., Kimber, W., Gallego, R., Virtue, S., Blount, M., Vázquez, M. J., Finer, N., and Powles, T. J. (2006) Tamoxifen-induced anorexia is associated with fatty acid synthase inhibition in the ventromedial nucleus of the hypothalamus and accumulation of malonyl-CoA, Diabetes 55, 1327-1336. [31] Ellis, J. M., and Wolfgang, M. J. (2012) A genetically encoded metabolite sensor for malonyl-CoA, Chem. Biol. 19, 1333-1339. [32] Schujman, G. E., Paoletti, L., Grossman, A. D., and de Mendoza, D. (2003) FapR, a bacterial transcription factor involved in global regulation of membrane lipid biosynthesis, Dev. Cell 4, 663-672. [33] Schujman, G. E., Guerin, M., Buschiazzo, A., Schaeffer, F., Llarrull, L. I., Reh, G., Vila, A. J., Alzari, P. M., and De Mendoza, D. (2006) Structural basis of lipid biosynthesis regulation in Gram positive bacteria, EMBO J. 25, 4074-4083. [34] Xu, P., Wang, W., Li, L., Bhan, N., Zhang, F., and Koffas, M. A. (2013) Design and kinetic analysis of a hybrid promoter–regulator system for malonyl-CoA sensing in Escherichia coli, ACS Chem. Biol. 9, 451-458. 22 ACS Paragon Plus Environment

Page 23 of 32

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


[35] Liu, D., Xiao, Y., Evans, B., and Zhang, F. (2015) Negative feedback regulation of fatty acid production based on a malonyl-CoA sensor-actuator, ACS Synth. Biol. 4, 132-140. [36] Wang, M., Li, S., and Zhao, H. (2015) Design and engineering of intracellular-metabolitesensing/regulation gene circuits in Saccharomyces cerevisiae, Biotechnol. Bioeng., 10.1002/bit.25676. [37] Wattanachaisaereekul, S., Lantz, A. E., Nielsen, M. L., and Nielsen, J. (2008) Production of the polyketide 6-MSA in yeast engineered for increased malonyl-CoA supply, Metab.Eng. 10, 246-254. [38] Zhang, F., Carothers, J. M., and Keasling, J. D. (2012) Design of a dynamic sensorregulator system for production of chemicals and fuels derived from fatty acids, Nat. Biotechnol. 30, 354-359. [39] Omura, S. (1976) The antibiotic cerulenin, a novel tool for biochemistry as an inhibitor of fatty acid synthesis, Bacteriol.Rev. 40, 681. [40] Suthers, P. F., and Cameron, D. C. (2005) Production of 3-hydroxypropionic acid in recombinant organisms, US Patent 6852517. [41] Rathnasingh, C., Raj, S. M., Lee, Y., Catherine, C., Ashok, S., and Park, S. (2012) Production of 3-hydroxypropionic acid via malonyl-CoA pathway using recombinant Escherichia coli strains, J. Biotechnol. 157, 633-640. [42] Chen, Y., Bao, J., Kim, I.-K., Siewers, V., and Nielsen, J. (2014) Coupled incremental precursor and co-factor supply improves 3-hydroxypropionic acid production in Saccharomyces cerevisiae, Metab. Eng. 22, 104-109. [43] Hügler, M., Menendez, C., Schägger, H., and Fuchs, G. (2002) Malonyl-Coenzyme A reductase from Chloroflexus aurantiacus, a key enzyme of the 3-hydroxypropionate cycle for autotrophic CO2 fixation, J. Bacteriol. 184, 2404-2410. [44] Navarre, C., Ghislain, M., Leterme, S., Ferroud, C., Dufour, J. P., and Goffeau, A. (1992) Purification and complete sequence of a small proteolipid associated with the plasma membrane H+-ATPase of Saccharomyces cerevisiae, J. Biol. Chem. 267, 6425-6428. [45] Mousson, F., Coïc, Y.-M., Baleux, F., Beswick, V., Sanson, A., and Neumann, J.-M. (2002) Deciphering the role of individual acyl chains in the interaction network between phosphatidylserines and a single-spanning membrane protein, Biochemistry 41, 1361113616. [46] Serrano, R., Kielland-Brandt, M. C., and Fink, G. R. (1986) Yeast plasma membrane ATPase is essential for growth and has homology with (Na+ + K+), K+- and Ca2+-ATPases, Nature 319, 689-693. [47] Rogers, T. O., and Lichstein, H. C. (1969) Characterization of the biotin transport system in Saccharomyces cerevisiae, J. Bacteriol. 100, 557-564. [48] Stolz, J., Hoja, U., Meier, S., Sauer, N., and Schweizer, E. (1999) Identification of the plasma membrane H+-biotin symporter of Saccharomyces cerevisiae by rescue of a fatty acid-auxotrophic mutant, J. Biol. Chem. 274, 18741-18746. [49] Roggenkamp, R., Numa, S., and Schweizer, E. (1980) Fatty acid-requiring mutant of Saccharomyces cerevisiae defective in acetyl-CoA carboxylase, Proc. Natl. Acad. Sci. USA 77, 1814-1817. [50] Levert, K. L., Waldrop, G. L., and Stephens, J. M. (2002) A biotin analog inhibits acetylCoA carboxylase activity and adipogenesis, J. Biol. Chem. 277, 16347-16350. [51] Alber, T., and Kawasaki, G. (1981) Nucleotide sequence of the triose phosphate isomerase gene of Saccharomyces cerevisiae, J. Mol. Appl. Genet. 1, 419-434. 23 ACS Paragon Plus Environment


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

[52] Compagno, C., Boschi, F., and Ranzi, B. M. (1996) Glycerol production in a triose phosphate isomerase deficient mutant of Saccharomyces cerevisiae, Biotechnol. Progr. 12, 591-595. [53] Overkamp, K. M., Bakker, B. M., Kötter, P., Luttik, M. A. H., van Dijken, J. P., and Pronk, J. T. (2002) Metabolic engineering of glycerol production in Saccharomyces cerevisiae, Appl. Environ. Microbiol. 68, 2814-2821. [54] Shi, S., Chen, Y., Siewers, V., and Nielsen, J. (2014) Improving production of malonyl Coenzyme A-derived metabolites by abolishing Snf1-dependent regulation of Acc1, mBio 5, e01130-01114. [55] Jensen et al. (2013) Microbial production of 3-hydroxypropionic acid. Patent in application [56] Shao, Z., Zhao, H., and Zhao, H. (2009) DNA assembler, an in vivo genetic method for rapid construction of biochemical pathways, Nucleic Acids Res. 37, e16. [57] Gietz, R. D., Schiestl, R. H., Willems, A. R., and Woods, R. A. (1995) Studies on the transformation of intact yeast-cells by the LiAc/s-DNA/PEG procedure, Yeast 11, 355360.


Page 24 of 32

Page 25 of 32

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


Table 1 List of plasmids. Plasmids pRS424-blank pRS424-FapR pRS424-FapR-NLS pRS414-FapR-NLS pRS415-fapO-tdTomato pRS415-(fapO)2-tdTomato pRS423 pRS425 pRS423-PMP1 pRS423-TPI1 pRS425-PMP1 pRS416-ACC1 Library plasmids

Description pRS424-TEF1p-HXT7t pRS424-TEF1p-FapR-HXT7t pRS424-TEF1p-FapR-SV40-HXT7t pRS414-TEF1p-FapR-SV40-HXT7t pRS415-GPM1p-fapO-tdTomato-ADH1t pRS415-GPM1p-(fapO)2-tdTomato-ADH1t pRS423, multi-copy plasmid with HIS3 marker pRS425, multi-copy plasmid with LEU2 marker pRS423-TEF1p-PMP1-HXT7t pRS423-TEF1p-TPI1-HXT7t pRS425-TEF1p-PMP1-HXT7t pRS416-TEF1p-ACC1-PGK1t pRS416-TEF1p-gene fragments-PGK1t



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

Page 26 of 32

Figure Legends Figure 1. Design of a malonyl-CoA sensor. (a) scheme (b) sequence Figure 2. Sensor validation: (a) Fluorescence intensity of strains expressing pRS415-fapOtdTomato together with pRS424-FapR-NLS, pRS424-FapR and pRS424-blank. (b) Fluorescence intensity enhanced by ACC1 overexpression. (c) Response to long chain acyl-CoAs. Error bars indicate standard deviations from triplicates. Figure 3. Dynamic response of sensors designed with different FapR/fapO ratios. Sensor activities were compared by fluorescence intensity of cells cultured in synthetic dropout medium supplemented with glucose (a and c). Dose-response curve was reflected by fluorescence intensity altered by different levels of cerulenin (b and d). Ratio is determined by the fluorescence intensity of strains cultured with cerulenin over that of strains without cerulenin. Error bars indicate standard deviation from duplicates. Figure 4. Fluorescence intensity of the screened constructs before re-transformation (a) and after re-transformation (b). Error bars indicate standard deviations from triplicates. Figure 5. Fermentation performance of engineered 3-HP producing strains with overexpressed PMP1, TPI1 (a and b) or co-overexpressed PMP1 and TPI1 (c and d) respectively. Error bars indicate standard deviations from triplicates.


Page 27 of 32

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



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 2


Page 28 of 32

Page 29 of 32

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



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


Page 30 of 32

Page 31 of 32

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



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

Graphic for TOC 266x104mm (293 x 293 DPI)

ACS Paragon Plus Environment

Page 32 of 32

Development of a Synthetic Malonyl-CoA Sensor in Saccharomyces

Recommend Documents