Design
Sum
mary
This year in 2023, the OUC-China iGEM team utilized a newly discovered fungus called M. aphidis XM01 from the mangroves in Hainan to regulate its metabolic pathways. We engineered the fungus to synthesize medium-chain fatty acids and express the corresponding enzymes to produce medium-chain α-olefins and 10-hydroxydecanoic acid. This achievement holds great industrial value.
Upon realizing the difficulty of removing benzo[a]pyrene often found in waste cooking oil, our team designed a light-controlled circuit for degradation. This circuit utilizes light as a trigger to activate the pathway responsible for breaking down benzo[a]pyrene, allowing for its removal from the waste oil.
To ensure the safety and security of the project, we also designed a NeuAc riboswitch to prevent gene leakage, thereby enhancing the safety of the project.
Part1:
MEL synthesis system
Our chassis organism, derived from the mangroves in Hainan, has the ability to utilize oils and fats as substrates to produce Mannosylerythritol lipids (MELs). MELs are composed of a hydrophilic core, 4-O-β-d-pyran mannosylerythritol, and a hydrophobic tail. They are microbial surface-active agents obtained through microbial fermentation and are completely based on renewable substrates. The hydrolysis products of MELs contain a significant amount of medium-chain fatty acids with carbon chain lengths ranging from 8 to 12, as shown in the diagram below.
Figure 1 MEL hydrolysis process and products
The MEL biosynthetic cluster was initially described in Ustilago maydis but was later found in two strains of M. antarcticus as well. It consists of five genes encoding enzymes involved in the biosynthesis of MELs. These genes are:
EMT1 (Erythritol-Mannose Transferase): This gene encodes a glycosyltransferase responsible for transferring a mannose moiety to erythritol, forming the core structure of MELs.
MAC1 and MAC2 (Mannose Acyltransferases): These genes encode acyltransferases that transfer fatty acids to the core structure of MELs, resulting in the hydrophobic tail formation.
MAT1 (Mannose Acetyltransferase): This gene encodes an acetyltransferase that adds an acetyl group to the MELs, contributing to their overall structure.
MMF1 (Mannitol-Mannose Facilitator): This gene encodes a cell-export protein that facilitates the export of MELs from the cell, enabling their secretion.
In the selected strain XM01 of our engineered bacteria, the gene cluster responsible for MEL biosynthesis is depicted in the diagram below.
Figure 2 MEL biosynthetic gene cluster
However, XM01 faces a significant challenge when it comes to the direct synthesis of MELs. During MEL production, it tends to accumulate a large amount of intracellular lipids[4], with cellular lipid content exceeding 60%. This is due to the coexistence of different fatty acid pathways in the XM01 strain, leading to the simultaneous production of a high quantity of MELs and the accumulation of intracellular lipids.
After brainstorming, we decided to enhance MEL synthesis by knocking out key genes involved in intracellular lipid synthesis. Previous research has indicated a close association between fungal intracellular lipid synthesis and genes such as DGA1, LRO1, and ARE1. Complete knockout of the LRO1 gene, in particular, results in the inability to generate intracellular lipids. Therefore, we chose to construct a double knockout strain of the ARE1 and DGA1 genes.
By removing these genes, we expect to mitigate the issue of intracellular lipid accumulation, thereby improving the efficiency of MEL synthesis. This genetic modification strategy has the potential to optimize our engineered XM01 strain for enhanced MEL production.
Figure 3 Intracellular lipid metabolism pathway
Figure 4 Double gene knockout schematic diagram
Due to various issues with natural substance extraction and chemical synthesis processes [14-15], we ultimately chose the biocatalysis approach. However, biocatalysis currently faces the challenge of low yields. First, through literature research, we identified several host organisms capable of producing medium-chain fatty acids, including Saccharomyces cerevisiae (brewer's yeast), Mycobacterium vaccae, and Escherichia coli. The corresponding yields of medium-chain fatty acids by these microorganisms are shown in the following table. Due to the inherent toxicity of medium-chain fatty acids to these microorganisms, their production is limited, making it difficult to significantly increase the yield of medium-chain fatty acids.
Figure 5 Synthesis of medium-chain fatty acids by microbial method
To address the aforementioned issues, we decided to use XM01 as a precursor for the synthesis of medium-chain fatty acids (MCFAs), knowing that it can produce MEL (which can be hydrolyzed to yield a large amount of MCFAs) [5]. By doing so, we can utilize the metabolic products of microorganisms instead of relying on petrochemical raw materials, effectively resolving the problem of material scarcity. Additionally, this approach significantly simplifies the complex chemical synthesis process.
Moreover, by not directly synthesizing MCFAs, we can avoid cell toxicity issues that inhibit cell growth and limit productivity. This strategy helps overcome the challenge of low yields associated with direct biocatalysis.
The first step in the metabolism of used cooking oil as a substrate involves the hydrolysis of triglycerides into free fatty acids and glycerol. This process is catalyzed by extracellular lipases or esterases. The fatty acids then enter the fatty acid metabolic pathway, while the free fatty acids are converted to acyl-CoA, which serves as the fatty acyl donor for various steps in the synthesis of intracellular triacylglycerols (TAG), ultimately leading to the production of intracellular oil.
Fungal intracellular lipids mainly consist of neutral lipids such as triglycerides and sterol esters. Two key steps in the synthesis of intracellular oil are the one-step synthesis of diacylglycerol to triacylglycerol and the one-step synthesis of sterol esters from sterols. These steps are catalyzed by three members of the acyl-CoA:diacylglycerol acyltransferase (DGAT) gene family, namely, acyl-CoA:cholesterol acyltransferase (ACAT), diacylglycerol acyltransferase (DGAT), and phospholipid:diacylglycerol acyltransferase (PDAT). Through analysis of the XM01 strain's genome, it was found that the three members of the o-acyltransferase gene family correspond to the genes ARE1, DGA1, and LRO1 in XM01.
Based on information from references [3], it was discovered that the LRO1 gene corresponds to the diacylglycerol acyltransferase (PDAT), which is an acyl-CoA-independent enzyme involved in the synthesis of triacylglycerol (TAG) by cleaving an acyl group from the sn-2 position of phosphatidylcholine or phosphatidylethanolamine, using diacylglycerol as the acyl acceptor molecule. Both phosphatidylcholine and phosphatidylethanolamine are phospholipids present in cell membranes. During the logarithmic growth phase, cells must selectively add, remove, or remodel membrane phospholipids in different organelles in response to cellular cycles and stress signals. This implies that the intracellular oil synthesis mediated by the LRO1 gene provides a large energy supply for cell growth and division while also regulating the differentiation and growth of cell membrane organization. The metabolic pathway mediated by the LRO1 gene links membrane remodeling in organelles with lipid storage. Consequently, knocking out the LRO1 gene has a significant impact on cell growth. Therefore, we ultimately did not choose it as the target for knockout.
Video 1 Part1: MEL synthesis system
Part2:
Medium chain fatty acid derivatives synthesis system
α-olefins are monoolefins with a double bond at the end of the carbon chain. In the industry, α-olefins generally refer to high-carbon olefins starting from C4 and above. Currently, the most widely used α-olefins are of medium to long chain lengths (C6 to C18), with straight-chain alkenes being more commonly utilized. Since the early stages of industrial development, α-olefins have been extensively used as raw materials for detergents, plasticizers, and lubricating oils. Even today, α-olefins have become the main raw materials for synthetic industrial lubricants and are also used in military industries.
We utilize the OleTJE P450 enzyme to catalyze the single-step decarboxylation of fatty acids to produce α-olefins, which is a relatively simple approach in the biosynthesis of α-olefins. It directly uses free fatty acids as substrates to synthesize α-olefins in one step. This process is simple and easy to control, making it highly regarded for its potential in designing biological systems for α-olefin production. The degradation mechanism is shown in the diagram below:
Figure 6 P450 catalytic process and synthesis of α-olefin
10-Hydroxydecanoic acid is a pharmaceutical intermediate that can be used in the production of antitussive and expectorant drugs. It is also utilized in the synthesis of 10-undecenoic acid (royal jelly acid), a compound found in bee products, as well as in the development of new drugs for the treatment of Alzheimer's disease, like donepezil, and other pharmaceutical intermediates. In the cosmetics industry, 10-Hydroxydecanoic acid serves several purposes:
Figure 7 Synthesis of 10-hydroxydecanoic acid pathway
In the ω-hydroxylation enzyme system, we have chosen the alkBGT[9] system, which consists of three enzymes: AlkB, AlkG, and AlkT, derived from the plasmid of Pseudomonas putida GPo1. AlkB is a membrane-bound monooxygenase and a key component of the alkane hydroxylation enzyme system. AlkG and AlkT are electron transfer proteins involved in the oxidation of NADH and electron transfer within the system. The operating mechanism of the alkane hydroxylation system in vivo is as follows: AlkT consumes NADH and generates electrons, AlkG transfers the electrons to AlkB, and finally, AlkB introduces a single oxygen atom from molecular oxygen at the terminal position to achieve hydroxylation reaction.
This system exhibits highly specific terminal hydroxylation activity towards fatty acids and has a broad substrate specificity. It can oxidize various substrates including C5-C16 alkanes, cycloalkanes, alkenes, thioether fatty acids, and fatty acids. It shows the highest catalytic activity towards medium-chain fatty acids. Since many hydroxylation enzymes have a preference for longer carbon chain fatty acid substrates, the alkane hydroxylation enzyme system holds great promise in the production of medium-chain hydroxy fatty acids.
Figure 8 Hydroxylase reaction principle
In addition, it has been suggested that the intracellular NADH in E. coli is not sufficient for the continuous biosynthesis of 10-hydroxydecanoic acid in cell factories. Therefore, it is necessary to design an efficient coenzyme regeneration cycle system to improve the biosynthesis efficiency of 10-hydroxydecanoic acid. In this study, glucose dehydrogenase (GDH) from Bacillus subtilis was selected as the driving enzyme for NADH cofactor recycling.
In order to further provide the catalytic capacity of the system, we designed and constructed an intracellular self-assembled multi-enzyme complex using the SpyCareer-SpyTag system to assemble the cofactors GDH and AIKT to achieve more electron transfer and an ordered coenzyme transfer channel between the two enzymes, thus achieving efficient biosynthesis [11].
Figure 9 Construction of dual enzyme complex system using SpyCatcher-SpyTag
Through the addition of cofactor and the construction of two-enzyme complex, we hope to gradually improve the production of 10-hydroxydecanoic acid.
Figure 10 Synthesis of 10-hydroxydecanoic acid
Different sources and types of hydroxylases have a direct impact on the final yield and type of products. Currently known fatty acid hydroxylases include P450 monooxygenase, lipoxygenase, hydratase and alkane hydroxylase. Among them, alkane hydroxylase is highly specific for the terminal light ylation of fatty acids and has a wide substrate spectrum, which can oxidize alkanes, cycloalkanes, alkenes, thioether fatty acids and fatty acids of C5-C16. As most hydroxylases prefer long-chain fatty acids, this alkane hydroxylase system has a good prospect in the production of medium-chain hydroxy fatty acids. In addition, in many studies of E. coli hydroxylation, the hydroxylation of non-terminal methyl groups is still a major obstacle in biocatalysis. A key advantage of alkane hydroxylase AIkBGT is that it only hydroxylates terminal methyl groups. This will make them have great development potential in the production and application of medium-chain ω-hydroxy fatty acids.
After adding GDH as a cofactor to the 10-hydroxydecanoic acid catalytic system, we thought about constructing a two-enzyme complex system to improve the catalytic efficiency. We first selected a series of linker to try, as shown in the following table. Homology modeling was performed by I-TASSER and alphafold2 software, and it was found that the addition of linker would have a certain impact on the active centers of AlkT and GDH. For more results, refer to the model page.
Figure 11 selected linker
Through homology modeling, it can be found that when two larger proteins are linked through a linker, the conformation of their original active center is often easily destroyed. Based on the data query, we decided to use SpyCather-SpyTag. These two short peptides can be self-assembled in bacteria after being linked to their respective proteins, which can greatly reduce the breakage of their own structure, and at the same time, due to their small molecular weight, the expression efficiency of proteins can be greatly improved.
Figure 12 Glucose 1-dehydrogenase-----SpyCatcher Rubredoxin-NAD(+) reductase-----SpyTag
Video 2 Part2: Medium chain fatty acid derivatives synthesis system
Part3:
Benzopyrene degradation systemm
Due to the small amount of benzo (a) pyrene in kitchen waste oil, due to its strong carcinogenic ability, it has adverse effects on the environment and human health. Therefore, we designed a related circuit to degrade benzo (a) pyrene. At the same time, in order to achieve controllability of degradation, we designed a red light switch to control the expression of degradation enzymes [1,2,6,7]. The gene circuit is shown below:
Figure 13 Benzo[a]pyrene degradation pathway
Among them, phyA is a functional phytochrome receptor composed of an apolipoprotein and a phytochrome mobile protein (chromophore). Since chromophore cannot be synthesized in yeast, a similar compound, phycocyanobilin (PCB), purified from cyanobacteria, was added to the medium. PCB is readily taken up by yeast cells and is bound by the phytochrome apoprotein to form phytochrome photoreceptors.
Under red light (λmax = 660 nm) or far-red light (λmax = 730 nm), PhyA reversibly changes its conformation to be able to bind to FHY1, which binds PhyA to BD (DNA binding domain) and PHY1 to AD (transcription activation domain) via linker. The principle of yeast double hybrid [12-13] was used to realize the expression of the target gene under the induction of red light.
Figure 14 Red light system schematic
Transcription activation proteins can bind to specific sequences on DNA to initiate the transcription reaction of corresponding genes. The Binding Domain (BD) and Activation Domain (AD) are two independent domains on the transcription activation protein, and both of them are required for gene transcription activation. Based on this principle, yeast two-hybrid experiments have been designed to verify the interaction between the two proteins. Currently, yeast two-hybrid experiments have used two systems, LexA system and Gal4 system.
Following the design of the 2012TU_Munich team, we chose LexA as our DNA-binding domain. This is because, in contrast to GAL4-based systems, no knockdown of the yeast endogenous GAL4/GAL80 gene is required. It therefore also does not interfere with endogenous yeast metabolism and signaling systems because it recognizes only one specific prokaryotic DNA sequence, the so-called LexA binding site. Unlike the GAL4-based system, we do not need a special strain carrying a GAL4/80 deletion, so theoretically every yeast strain could work.
Meanwhile, based on the study of Oxana Sorokina et al. (2009), we found that the combination of PhyA and FHY1 had a faster and better response to red light. Therefore, PhyA and FHY1 were selected as the two main proteins in the red light system among the many phytochrome and its interaction factors.
Figure 15 Different photosensitive pigments and their interacting factors
Video 3 Part3: Benzopyrene degradation systemm
Part4:
Kill system
To avoid the leakage of foreign genes into the environment, we designed NeuAc riboswitch. "Sequence of the NeuAc riboswitch, which consists of an RNA aptamer and a hammerhead ribozyme." Its mechanism of action is that NeuAc binds to the RNA aptamer and causes a conformational change of the whole switch, leading to self-cleavage of the hammerhead ribozyme, resulting in a gap. Then, under the action of ReJ ribozyme in Escherichia coli cells, the genes connected to it are cleaved, thus reducing the mRNA abundance of downstream expressed genes [16-17]. To ensure that the switch could function effectively, spacer regions were added at both ends of the switch, 5-AAAAATAAAAAGAAAAA-3 at the 5' end and 5-AAACAAACAAA-3 at the 3 'end. We selected ΦX174 downstream as the cleavage protein to achieve E. coli suicide.
Figure 16 kill switch
The suicide switch is designed so that it can only survive in the culture medium or fermentation broth, without which it can no longer survive. This can often be achieved by inducible promoters, auxotrophic strains, riboswitches. Inducible promoters require inducible substances that are either toxic, will be absorbed and utilized, or involve antibiotics. Auxotrophic strains often have difficulty controlling the amount of nutrients in the fermentation broth and do not know when they are exhausted for repleniation. The riboswitch involved in thiaminepyrophosphate, TPP), S-adenosyl-L-methionine (SAM), purines and their derivatives, amino acids, phosphorylated sugars, and metal ions are also toxic or can be absorbed and utilized. So we finally chose NeuAc riboswitch, its aptamer N-acetylneuraminic acid cannot be synthesized in E. coli, and it is non-toxic and not metabolized by E. coli. (Some E. coli strains have knocked out the key gene for N-acetylneuraminic acid degradation.).
Video 4 Part4: Kill system
Refer
ences
[1]Sorokina, O., Kapus, A., Terecskei, K. et al. A switchable light-input, light-output system modelled and constructed in yeast. J Biol Eng 3, 15 (2009). https://doi.org/10.1186/1754-1611-3-15
[2]Hochrein L, Machens F, Messerschmidt K, Mueller-Roeber B. PhiReX: a programmable and red light-regulated protein expression switch for yeast. Nucleic Acids Res. 2017 Sep 6;45(15):9193-9205. doi: 10.1093/nar/gkx610. PMID: 28911120; PMCID: PMC5587811.
[3]Athenstaedt K. Yali0e32769g (dga1) and yali0e16797g (lro1) encode major triacylglycerol
synthases of the oleaginous yeast yarrowia lipolytica [J]. BBA Molecular and Cell Biology of
Lipids, 2011, 1811(10) : 587-596.
[4]Dai Kitamoto T N, Hiroshi Yanagishita1 et al. Fatty - acid metabolism of mannosylerythritol
lipids as biosurfactants produced by Candida antarctica [J]. Journal of the American Chemical
Society, 1993, 42(5) : 346-358.
[5]Yu G, Wang X, Zhang C, et al. Efficient production of mannosylerythritol lipids by a marine
yeast Moesziomyces aphidis XM01 and their application as self-assembly nanomicelles [J]. Marine Life Science & Technology, 2022, 4(3) : 373-383.
[6]Shimizu-Sato S , Huq E , Tepperman J M ,et al.A light-switchable gene promoter system[J].Nature Biotechnology, 2002, 20(10):1041-1044.DOI:10.1038/nbt734.
[7]Levskaya, A., Weiner, O., Lim, W. et al. Spatiotemporal control of cell signalling using a light-switchable protein interaction. Nature 461, 997–1001 (2009). https://doi.org/10.1038/nature08446
[8]Zhao E M , Zhang Y , Mehl J ,et al.Optogenetie regulation of engineered cellular metabolism for microbial chemical production[J].Nature, 2018, 555(7698):683-687.
[9]Qiaofei He, George N. Bennett, Ka-Yiu San, Hui Wu*. Biosynthesis of medium-chain ω-hydroxy fatty acids by AlkBGT of Pseudomonas putida GPo1 with native FadL in engineered Escherichia coli. Frontiers in Bioengineering and Biotechnology. 2019. 7:273.
[10]Zakeri B , Fierer J O , Celik E ,et al.Peptide tag forming a rapid covalent bond to a protein, through engineering a bacterial adhesin[J].Proceedings of the National Academy of Sciences of the United States of America, 2012(12).DOI:10.1073/PNAS.1115485109.
[11]Liu, L., Yin, L., Huang, F., et al. Construction of intracellular self-assembling multi-enzyme complexes for efficient biosynthesis using SpyTag/SpyCatcher. Chinese Journal of Biotechnology, (07), 75-82. DOI: 10.13523/j.cb.20180710.
[12]Li, H., Qin, X., Song, P., et al. A LexA-based yeast two-hybrid system for studying light-switchable interactions of phytochromes with their interacting partners. Biotechnology Bulletin: English Edition, 2021, 2(2), 12. DOI: 10.1007/s42994-021-00034-5.
[13]Zhu, K., Zhang, J., Liu, S. Research progress on the yeast transcription factor Gal4. Chinese Journal of Biotechnology, 2011, 31(01), 81-85. DOI: 10.13523/j.cb.20110116.
[14]Sun, L., Wu, J. Research progress and application status of medium-chain fatty acids. Chinese Journal of Food Additives, 2022, 33(12), 50-55. DOI: 10.19804/j.issn1006-2513.2022.12.007.
[15] Wang, J. Evolution of N-acetylneuraminic acid-producing recombinant Escherichia coli based on a riboswitch screening system. (Doctoral dissertation). Shandong University, 2016.
[16] Chen, M., Yong, R., Tang, Z. Construction and validation of a kanamycin artificial riboswitch. Journal of Applied and Environmental Biology, 2023, 29(02), 340-345. DOI: 10.19675/j.cnki.1006-687x.2022.01015.
[17] Staijen I E , Beilen J B V , Witholt B .Expression, stability and performance of the three-component alkane mono-oxygenase of Pseudomonas oleovorans in Escherichia coli[J].European journal of biochemistry, 2000, 267(7):1957-65.DOI:10.1046/j.1432-1327.2000.01196.x.