We designed an artificial 1,5-pentanediol (1,5-PDO) biosynthetic pathway, and identified its theoretical feasibility through a metabolic network model simulation. After screening the functional enzymes in the pathway, we assembly the whole pathway in a lysine-producing strain Escherichia coli NT1003, and successfully fulfill the pathway for 1,5-PDO production from glucose. In addition, in conjunction with dry experiment of metabolic network model simulation, several engineering strategies were employed to enhance the 1,5-PDO production, including optimizing the gene expression plasmid, inhibiting acetic acid production pathway and 1,5-PDO degradation pathway, mutating the limiting enzymes and also assembling the limiting enzymes. The related parts have been registered and tested, and most of them worked well to perform their functions as we expected.

Screening the enzymes for constructing the functional modules to synthesize 1,5-PDO

Because none of natural pathways for 1,5-PDO biosynthesis is present, we first designed an artificial 1,5-PDO biosynthetic pathway in our work. In the designed pathway, lysine is used as a director precursor, which is then converted into 5-hydroxyvalerate (5-HV) and 1,5-PDO via two synthetic modules (5-HV synthesis module and 1,5-PDO synthesis module). The 5-HV synthesis module converts lysine into 5-HV via the oxidase, hydrolase, transaminase, and aldehyde reductase, while 1,5-PDO synthesis module converts 5-HV into 1,5-PDO via carboxylate reductase and aldehyde reductase. E. coli lacks some related enzymes in our designed 1,5-PDO biosynthetic pathway. We added several heterologous enzymes into the iML1515 metabolic network model of E. coli to establish the metabolic model for 1,5-PDO synthesis. Through a flux balance analysis (FBA), the maximum theoretical yield for 1,5-PDO production based on our designed pathway was determined to be 0.847 mol/mol_glucose, indicating the feasibility of our designed pathway for 1,5-PDO biosynthesis.

Figure 1. A proposed biosynthesis route for de novo production of 1,5-PDO from glucose.

Subsequently, to establish the synthetic pathway for 1,5-PDO production, the screening of desired enzymes to construct the functional modules was designed. To establish a functional 1,5-PDO synthesis module, two enzymes of carboxylate reductase and aldehyde reductase from different organisms were screened. The carboxylate reductases originated from Mycobacteroides abscessus (MAB3367) Mycolicibacterium smegmatis (MSMEG5739), Mycobacterium abscessus (MAB4714), Mycobacterium avium (MpCAR), and Mycobacterium marinum (MmCAR) were selected as candidate genes. For aldehyde reductases, the enzymes originated from Oenococcus oeni (Adh1), Oenococcus alcoholitolerans (Hsd36), Escherichia coli (AdhP) and E. coli (YahK) were selected. To establish a functional 5-HV synthesis module, the key enzyme of transaminase originated from E. coli (GabT and PatA), Paracoccus denitrificans (QLH), Streptomyces avermitilis (SAV2585), Ruegeria pomeroyi (Spo3471) and Pseudomonas aeruginosa (SpuC) were selected as candidates for screening.

Figure 2. The screening of enzymes for the establishment of functional modules for 1,5-PDO bioproduction.

To obtain the desired 1,5-PDO synthesis module, several carboxylate reductases and aldehyde reductases were evaluated. Through literature research, five carboxylate reductases (MAB3367 from M. abscessus BBa_K4800020, MpCAR from M. avium BBa_K4800021, MSMEG5739 from M. smegmatis BBa_K4800022, MmCAR from M. marinum BBa_K4800017 and MAB4714 from M. abscessus BBa_K4800023), and four aldehyde reductases (YahK from E. coli BBa_K4800027, Hsd36 from O. alcoholitolerans BBa_K4800025, Adh1 from O. oeni BBa_K4800024 and Adhp from E. coli BBa_K4800026) were codon-optimized and synthesized. Among them, MmCAR and sfp is the component of an existing part of BBa_K1655000. The gene fragments of carboxylate reductase were subcloned into the linear pRSFDuet-1 together with its activating factor phosphopantetheinyl transferase (sfp) from Bacillus subtilis, yielding the plasmids of pRSFDuet-MAB3367 BBa_K4800033, pRSFDuet-MpCAR BBa_K4800034, pRSFDuet-MSMEG5739 BBa_K4800035, pRSFDuet-MmCAR BBa_K4800060, and pRSFDuet-MAB4714 BBa_K4800036. The gene fragments of aldehyde reductase were subcloned into the linear pTrc99a by the enzyme digestion and ligation, yielding the plasmids of to pTrc99a-YahK BBa_K4800040, pTrc99a-Hsd36 BBa_K4800038, pTrc99a-Adh1 BBa_K4800037 and pTrc99a-Adhp BBa_K4800039. The plasmids were transformed into E. coli BL21 respectively, and corrected colonies were verified by PCR, sequencing, and SDS-PAGE analysis. As shown in Figure 3, all five carboxylate reductases and four aldehyde reductases have soluble expression in E. coli BL21.

Figure 3. (a) Constructing expression vectors of different carboxylate reductases. (b) SDS-PAGE analysis of of different carboxylate reductases in E. coli. (c) Constructing expression vectors of different aldehyde reductases. (d) SDS-PAGE analysis of of different aldehyde reductases in E. coli.

To obtain functional 1,5-PDO synthesis module, plasmids of pRSFDuet-MAB3367, pRSFDuet-MpCAR, pRSFDuet-MSMEG5739, pRSFDuet-MmCAR, and pRSFDuet-MAB4714 were transferred into E. coli BL21 together with pTrc99a-YahK to compare the activity of different carboxylate reductases. The plasmids of pTrc99a-Yahk, pTrc99a-Hsd36, pTrc99a-Adh1 and pTrc99a-Adhp were transferred into E. coli BL21 together with pRSFDuet-MmCAR to compare the activity of different aldehyde reductases.

Figure 4. (a) Plasmid transformation for carboxylate reductase screening assay. (b) Plasmid transformation for aldehyde reductase screening assay.

In 5-HV systhesis module, DavB and DavA have been identified capable of efficiently converting lysine to 5-aminovalerate (5-AVA) in the reported reference. To obtain the efficient 5-HV synthesis module, we turned out our attention to screen different transaminases. Six transaminases, including GabT from E. coli BBa_K4800015, PatA from E. coli BBa_K4800028, Spo3471 from R. pomeroyi BBa_K4800029, SAV2585 from S. avermitilis BBa_K4800030, SpuC from P. aeruginosa BBa_K4800032, and QLH from P. denitrificans BBa_K4800031 were codon-optimized and synthesized. The gene fragments of transaminases were subcloned into the linear pTrc99a, yielding the plasmids of pTrc99a-GabT BBa_K4800046, pTrc99a-PatA BBa_K4800041, pTrc99a-Spo3471 BBa_K4800042, pTrc99a-SAV2585 BBa_K4800043, pTrc99a-SpuC BBa_K4800045, and pTrc99a-QLH BBa_K4800044.
The plasmids were transformed into E. coli BL21 respectively, and corrected colonies were verified by PCR, sequencing, and SDS-PAGE analysis. As shown in Figure 5, all six transaminases show soluble expression in E. coli.

Figure 5. (a) Constructing expression vectors of different transaminases. (b) SDS-PAGE analysis of different transaminases in E. coli.

To compare the activity of different transaminases, the DNA fragments of Ptrc-RBS-YahK were amplified and ligated into the plasmids of pTrc99a-GabT, pTrc99a-PatA, pTrc99a-Spo3471, pTrc99a-SAV2585, pTrc99a-SpuC, and pTrc99a-QLH, yielding the desired plasmids of pTrc99a-YahK-GabT BBa_K4800047, pTrc99a-YahK-PatA BBa_K4800048, pTrc99a-YahK-Spo3471 BBa_K4800049, pTrc99a-YahK-SAV2585 BBa_K4800050, pTrc99a-YahK-SpuC BBa_K4800051, and pTrc99a-YahK-QLH BBa_K4800052. The plasmids were transferred into E. coli BL21 to compare the activity of different transaminases by detecting 5-HV production from 5-AVA.

Figure 6. Constructing expression vectors of different transaminases to compare their activity.

By using a whole-cell catalysis process, the enzyme activity was compared. The activity of carboxylate reductase was determined by detecting the 5-HV consumption and 1,5-PDO production with HPLC. By comparing the 5-HV consumption, MmCAR and MpCAR displayed the higher substrate consumption activity. The further detection of 1,5-PDO production showed that MmCAR resulted in the highest activity for 1,5-PDO production (11.5mM). Therefore, MmCAR was selected as the carboxylate reductase in the following strain construction.

Figure 7. Comparing the activity of carboxylate reductases from different organisms.

After identifying the efficient carboxylate reductase for 1,5-PDO synthesis module, we performed the synthesis of 1,5-PDO by whole-cell catalysis to compare the activities of the four aldehyde reductases by detecting 1,5-PDO production with HPLC. The results confirmed that YahK from E. coli exhibited the highest activity for 1,5-PDO accumulation. Thus, YahK was selected to assembly the 1,5-PDO synthesis module in the following work.

Figure 8. Comparing the activity of aldehyde reductases from different organisms.

The activity of transaminase was determined by detecting the 5-HV production from 5-AVA with HPLC. The results showed the 6 transaminases all exhibit ammonia-transamination activity. Among them, GabT from E. coli exhibited the highest ammonia-transamination efficiency. Thus, GabT was selected to assembly the 5-HV synthesis module in the following work.

Figure 9. Comparing the activity of transaminases from different organisms.

Production of chemicals without natural synthetic pathways is a big challenge for synthetic biology. We attempted to engineer E. coli to produce 1,5-PDO via an artificial synthetic pathway. To assembly the functional pathways, we screened different candidate enzymes in designed synthesis modules, and efficient carboxylate reductase, aldehyde reductase and transaminase were obtained.
In this cycle, we learned that different sources of enzymes exhibit the different activity, and would affect the titer of the targeted products. In addition, the metabolic network model also shows that the enzymes in the 5-HV synthesis module and 1,5-PDO synthesis module are important for increasing the final 1,5-PDO yield. Before the construction of the producing strain, the identification of efficient parts with higher activity is an effective strategy, which provides a good basis for us to construct the 1,5-PDO-producing strain in the following work. This was the first time that the team members designed and completed the experiment independently by themselves, and everyone in the team gained a lot in the whole process.

Constructing the 1,5-PDO producing strain

Through preliminary experiments, we have identified the suitable carboxylate reductase, aldehyde reductase and transaminase in the metabolic pathway, and obtain the functional modules. Subsequently, we expressed the genes of the two modules in two plasmids and introduced them into a lysine-producing E. coli NT1003 to construct a cell factory capable of synthesizing 1,5-pentanediol from glucose. Subsequently, we used a direct fermentation to validate the feasibility to produce 1,5-PDO of our engineered strain.

Figure 10. Constructing the 1,5-PDO producing strain.

In order to produce 1,5-PDO by using E. coli NT1003, the genes of 5-HV synthesis module including davA, davB, and GabT were subcloned into the plasmid pTrc99a together under the control of the trc promoter to yield plasmid pTtrc99a-davB-davA-GabT BBa_K4800054. In detail, using the laboratory conserved plasmid pET22b-T7-davB-T7-davA as templates, the fragments of davB and davA were amplified by PCR with a consensus RBS and P_trc respectively. Subsequently, the P_trc-RBS-davB-P_trc-RBS-davA fragment was obtained by overlap PCR, which was then ligated into the linearized vector pTrc99a-GabT by In-fusion cloning method to obtain plasmid pTtrc99a-davB-davA-GabT. The plasmid was transformed into E. coli Trans-T1, and corrected colonies were verified by PCR, and sequencing.

Figure 11. (a) The schematic of plasmid pTrc99a-davB-davA-GabT. (b) The colony PCR validation diagram.

The genes of 1,5-PDO synthesis module including MmCAR and YahK were subcloned into the plasmid pACYCDuet together under the control of trc promoter to yield plasmid pACYCDuet-sfp-MmCAR-YahK BBa_K4800053. In detail, based on the sequence of an existing part of BBa_K1655000, it was codon-optimized and synthesized to obtain the fragment of sfp-MmCAR, which was then ligated into the pACYCDuet plasmid under the control of trc promoter. The Ptrc-YahK fragment was amplified with plasmid pTrc99a-YahK as the template, and In-Fusion cloning was then used to ligate into the above vectors to yield plasmid pACYCDuet-sfp-MmCAR-YahK. The plasmid were transformed into E. coli Trans-T1, and corrected colonies were verified by PCR, and sequencing.

Figure 12. (a) The schematic of plasmid pACYCDuet-sfp-MmCAR-YahK. (b) The colony PCR validation diagram.

The plasmid of pTrc99a-davB-davA-GabT and pACYCDuet-sfp-MmCAR-YahK were co-transformed into a lysine producing strain E. coli NT1003 to construct an engineered 1,5-PDO producing strain NT1003-P1.

The 1,5-PDO production of the engineered strain E. coli NT1003-P1 was determined by the HPLC analysis. After the fermentation of 72 hours, the fermentation broth contained 2.82 mM (0.293g/L) of 1,5-pentanediol, which indicated that the synthetic pathway we designed was experimentally feasible, and the cell factory that could produce 1,5-PDO from glucose has been successfully developed. However, the 1,5-PDO titer was still lower than we expected, and the intermediate of 5-HV was largely accumulated.

Figure 13. The 1,5-pentanediol production by the engineered strain of E. coli NT1003-P1.

In this cycle, we successfully constructed an engineered strain of E. coli NT1003-P1 to produce 1,5-PDO from glucose. We learned that the assembly of multiple genes in pathway into different plasmids with different replicon could make the pathway functional work in the strain. However, the 1,5-PDO titer is far below than we expected. Based on the dry expriment group, several potential gene targets are predicted for the strain optimization. We designed the experiment protocol and discussed with other professors, and they gave us some advices.
Prof. Liang Liya from Dalian University of Technology advised us to change a plasmid with high copy number to enhance the expression of the 1,5-PDO synthesis module, as the plasmid pACYCDuet-sfp-MmCAR-YahK contained a replicon with low copy number.
Prof. Cao Yingxiu, from Tianjin University, exhibited keen interest in our modular designing idea, and she advised us to adjust the metabolic pathway to improve the product titer.
In view of dry experiment simulation results, as well as the professor’s advice, team members designed the following two cycles to modify the strain of E. coli NT1003-P1, hoping to further improve 1, 5-PDO production. One is the optimization of expression plasmid, and another one is the deletion of branched genes.

The optimization of 1,5-PDO production by changing plasmid with high copy number

By combining FSEOF, and OptKnock algorithms in the metabolic network model, several gene overexpression targets that potentially involved in the improvement of 1,5-PDO production were predicted. As a large amount of 5-HV was accumulated in the fermentation broth in our wet experiment, we focused on the genes in the 1,5-PDO synthesis module among these overexpression targets. After consulting with experts, we recognized that using pACYCDuet plasmid to express the 1,5-PDO synthesis module might not be the optimal choice, After in-depth discussions within the team, we decided to adopt a high-copy number plasmid of pRSFDuet to express the 1,5-PDO synthesis module. The resulting plasmid was co-transformed with 5-HV synthesis module into the E. coli NT1003 to evaluate the effect on 1,5-PDO production.

We used pACYCDuet-sfp-MmCAR-YahK as a template and amplified the PT7-YciA-Ptrc-sfp-MmCAR-Yahk fragment by PCR, and homologous recombination of PT7-YciA-Ptrc-sfp-MmCAR-Yahk with the linearized pRSFDuet vector was carried out by the In-fusion cloning method to yield plasmid of pRSFDuet-sfp-MmCAR-YahK BBa_K4800055. The plasmids were transformed into E. coli Trans-T1, and corrected colonies were verified by PCR, and sequencing. Subsequently, the plasmids of pTrc99a-davB-davA-GabT and pRSFDuet-sfp-MmCAR-YahK were co-transformed into a lysine producing strain E. coli NT1003 to construct an engineered 1,5-PDO producing strain NT1003-P2.

Figure 14. (a) The schematic of plasmid pRSFDuet-sfp-MmCAR-YahK. (b) The colony PCR validation diagram.

The yield of 1,5-PDO produced by the engineered strain NT1003-P2 was determined by the HPLC analysis. After a total fermentation period of 84 hours, 8.09 mM (0.843g/L) 1,5-PDO was produced by the engineered strain E. coli NT1003-P2, which is 3.5-fold higher than that in E. coli NT1003-P1.

Figure 15. Fermentation of the engineered strain E. coli NT1003-P2 to produce 1,5-PDO from glucose.

In this cycle, we have learned how to optimize gene expression for the increased product titer. Due to the difference in the replicon, different plasmid has different copy number in the strain, which would affect the gene expression level. Therefore, the selection of appropriate plasmid is also an effective strategy for improving strain producing ability.

Knockout of genes in the branched pathway to improve 1,5-PDO production

By using OptKnock algorithms in the metabolic network model, several gene deletion targets that potentially involved in the improvement of 1,5-PDO production were predicted. Among them, the ackA-pta gene encodes enzymes that convert pyruvate to acetic acid, the main by-product during the 1,5-PDO fermentation. After discussions with the experts, we also attempt to evaluate the effect of deleting these two genes on the 1,5-PDO production. In addition, in the previous study of the laboratory, it was found that the dehydrogenase of YcjQ was involved in the 1,5-PDO degradation. Therefore, in order to increase the yield, we also decided to knock out the YcjQ gene. Eventually, we decided to construct two strains NT1003-ΔackA-pta and NT1003-ΔYcjQ by the CRISPR/Cas9.

Figure 16. Metabolic pathway for 1,5-PDO production in NT1003. The genes marked red would be knocked out.

Firstly, we extracted the pCas9 and pTarget-sgRNA plasmids from laboratory preserved strains. In order to knock out these three genes, the upstream and downstream sequences (about 1000 bp) flanking ackA-pta and YcjQ were overlap PCR-amplified (BBa_K4800010 and BBa_K4800011). Next, we designed the N20 sequence of ackA-pta and YcjQ by using the CHOPCHOP website (http://chopchop.cbu.uib.no), synthesized them on primers and ligated the N20 into the pTarget-sgRNA plasmid by PCR. With these tools, we used CRISPR/Cas9 strategy to knock out target genes of ackA-pta and YcjQ by homologous recombination. Finally, the corrected colonies were verified by PCR to determine the deletion of ackA-pta and YcjQ.
The plasmid of pTrc99a-davB-davA-GabT, and pRSFDuet-sfp-MmCAR-YahK were co-transformed into E. coli NT1003-ΔackA-pta and E. coli NT1003-ΔYcjQ to construct the engineered 1,5-PDO producing strain NT1003-P2-ΔackA-pta, NT1003-P2-ΔYcjQ respectively.

Figure 17. Schematic diagram of CRISPER/Cas9.

The yield of 1,5-PDO produced by these strains was analyzed by HPLC. In comparison to engineered strain NT1003-P2, the knockout of ackA-pta increased the 1,5-PDO titer by 16%, and the titer in strain NT1003-P2-Δacka-pta reached 9.3 mM (0.97 g/L). Meanwhile, the deletion of YcjQ gene markedly improved the final 1,5-PDO titer, 1.75-fold higher than that in the strain NT1003-P2. These results demonstrated that NT1003-P2-ΔYcjQ performed better, and we had decided to adopt this strain in the following experiments.

Figure 18. The production of 1,5-PDO. (a) Comparison of production between NT1003-P2 and NT1003-P2-ΔackA-pta. (b) Comparison of production between NT1003-P2 and NT1003-P2-ΔYcjQ. (c) Comparison of production between NT1003-P2, NT1003-P2-ΔackA-pta and NT1003-P2-ΔYcjQ.

In this cycle, we learned that metabolic network model simulation is very helpful for guiding strain engineering, which can reduce the workload of the experiment. This also has reference significance for the work of other teams. In addition, we realized that knockout of genes in the branched pathway was a very effective strategy in microbial metabolic engineering for the improvement of strain performance.
After the optimization of expression plasmid and also deletion of branched pathways, the production 1,5-PDO was increased from 8 mM (0.83 g/L) to 14 mM (1.46 g/L). However, the accumulation of 5-HV was still detected. We thought that the activity of 1,5-PDO synthesis module was required to be further improved. Therefore, we further discussed our idea with some professors. Prof. Ren Xinkun from Nanjing university proposed that the engineering of limiting enzyme of MmCAR can be employed to improve the final production. Prof. Xu Jiaxing from Huaiyin Normal University advised us an strategy of enzyme assembly based on the protein scaffold, which might also be benefit for increasing enzyme activity. Based on their suggestion, we further designed the following two cycles of experiment.

Engineering carboxylate reductase for improved 1,5-PDO production

As the suggestion by Prof. Ren Xinkun, we attempted to improve the catalytic activity of MmCAR by a rational engineering strategy. After reviewing relevant literature and consulting with Prof. Ren, we decided to carry out targeted mutagenesis of MmCAR by rational design and verified the effect of mutantion on its catalytic activity. The A-domain crystal structure of carboxylate reductase (MmCAR) from Mycobacterium marinum was not available in the PDB database. Therefore, we submitted amino acid sequences of MmCAR including A-domain and PCP arm to the phyre2 website (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index) to build a protein model. After obtaining the PDB file of the protein structure, we used Autodock software to dock the protein structure with the 5-hydroxyvaleric acid (at least 50 docking) to find the 10th result with the lowest binding energy. For each docking result, the orientation of small molecules binding to the pocket was analyzed to see if it corresponded to the actual orientation. From structure of MmCAR binding to AMP and 5-hydroxyvaleric acid, it was found that the amino acids Q302 and T390 were on the side of 5-hydroxyvaleric acid and AMP away from the active center site, respectively. As both substrates are small polar molecules, we decided to mutant Q302 and T390 of MmCAR to acidic amino acid of glutamate (E) to change the polarity of the residues, and evaluated the effect of these mutants on MmCAR catalytic activity.

Figure 19. The targeted mutant amino acid sites of Q302 and T390 in MmCAR.

Based on the plasmid of pRSFDuet-sfp-MmCAR-YahK, we designed point mutation primers and cloned the whole plasmid using PCR. The linearized plasmid was isolated and purified by agarose gel electrophoresis, and then transferred into E. coli Trans-T1 to obtain the targeted plasmids of pRSFDuet-sfp-MmCAR^Q302E-YahK BBa_K4800059 and pRSFDuet-sfp-MmCAR^T390E-YahK BBa_K4800057. The corrected colonies were verified by sequencing. The plasmids of pRSFDuet-sfp-MmCAR^Q302E-YahK and pRSFDuet-sfp-MmCAR^T390E-YahK were transferred into E. coli BL21 respectively to evaluate the effect of mutation on the activity of 1,5-PDO synthesis module. The mutant plasmid with higher activity was further transformed into E. coli NT1003-ΔYcjQ together with pTrc99a-davB-davA-GabT to obtain the engineered strain NT1003-P3-ΔYcjQ for 1,5-PDO production from glucose.

Figure 20. (a) The pRSFDuet-sfp-MmCAR^Q302E-YahK plasmid. (b) The pRSFDuet-sfp-MmCAR^T390E-YahK plasmid. (c) Sequencing results indicate the successful mutation of pRSFDuet-sfp-MmCAR^Q302E-YahK plasmid. (d) Sequencing results indicate the successful mutation of pRSFDuet-sfp-MmCAR^T390E-YahK plasmid.

The effect of MmCAR mutation on the activity of 1,5-PDO synthesis module was determined by detecting the level to convert 5-HV to 1,5-PDO with HPLC analysis. The 1,5-PDO production of the engineered strain NT1003-P3-ΔYcjQ was determined by the HPLC analysis. From the catalysis results, we found that the 1,5-PDO synthesizing activity of the 1,5-PDO synthesis module could be increased by 50% when the amino acid site at position 302 was mutant to glutamic acid (E) from glutamine (Q) in MmCAR. However, when the amino acid site at position 390 was changed to glutamic acid (E) from threonine (T) in MmCAR, the activity of 1,5-PDO sysnthesis module was moderately affected. Therefore, we selected mutant MmCAR^Q302E for the following strain construction, and obtained the engineered strain of E. coli NT1003-P3-ΔYcjQ. Through a fermentation experiment, the mutation of MmCAR^Q302E could moderately increase the final 1,5-PDO production, and 14 mM (1.45 g/L) of 1,5-PDO in the engineered E. coli NT1003-P3-ΔYcjQ was finally obtained from glucose.

Figure 21. (a) Whole-cell catalytic results of mutant 1,5-PDO sysnthesis module to convert 5-HV to 1,5-PDO. (b) The 1,5-pentanediol production by the engineered strain of E. coli NT1003-P3-ΔYcjQ.

In this cycle, we learned that the types of the amino acid sites that make up enzymes have a critical influence on enzyme activity. Rational engineering of enzymes is an effective strategy to improve enzyme activity, which is also a powerful tool to improve the limiting enzyme activity in a metabolic pathway. In addition, we learned that the cultivation condition could also affect the enzyme activity. Although the mutation of enzyme significantly improve the activity of 1,5-PDO synthesis module, it did not significantly improve the 1,5-PDO production in the final fermenting strain. As a next plan, the optimization of cultivation conditions, as well as the further engineering MmCAR can be considered.

The immobilization of carboxylate reductase by a self-assembling protein scaffold system

In our communication with Prof. Xu Jiaxing, we learned that certain protein scaffolds can be used to enhance the catalytic efficiency of enzymes. After reviewing the literature and discussing within the group, we focused on the EutM from Salmonella enterica. EutM is known to self-assemble as filament-like structure in vivo, and has been developed as a powerful protein scaffolding system to rapidly co-immobilize enzymes for improving the biocatalysis efficiency. Therefore, we chose the EutM protein scaffold to immobilize the mutated MmCAR in vivo. At the same time, we intended to use the SpyTag-SpyCatcher system to help MmCAR and EutM bind better. The whole-cell conversion experiment and fermentation experiment were subsequently designed to verify whether the immobilization of the mutated MmCAR by the protein scaffold can further improve the 1,5-PDO production.

The genes of certain protein scaffolds including SpyTag, EutM and SpyCatcher were subcloned into the plasmid pRSFDuet-sfp-MmCAR^Q302E-YahK together to yield plasmid pRSFDuet-sfp-MmCAR^Q302E-SpyTag-EutM-SpyCatcher-YahK BBa_K4800062. In detail, using the laboratory conserved plasmid pETDuet-T7-EutM-SpyCatcher as templates, the fragments of EutM and SpyCatcher were amplified by PCR with a consensus RBS and trc promoter, which was then ligated into the linearized vector pRSFDuet-sfp-MmCAR^Q302E-YahK by In-fusion cloning method. Subsequently, the MmCAR^Q302E-Spytag fragment was obtained by modifying the 5' end of the primer, which was then ligated into the linearized vector pRSFDuet-sfp-EutM-SpyCatcher-YahK by In-fusion cloning method. The corrected colonies were verified by PCR, and sequencing. The plasmids of pRSFDuet-sfp-MmCAR^Q302E-YahK and pRSFDuet-sfp-MmCAR^Q302E-SpyTag-EutM-SpyCatcher-YahK were transferred into E. coli BL21 respectively to evaluate the effect of mutation on the activity of 1,5-PDO synthesis module. The plasmids of pRSFDuet-sfp-MmCAR^Q302E-SpyTag-EutM-SpyCatcher-YahK and pTrc99a-davB-davA-GabT were co-transformed into E. coli NT1003-ΔYcjQ to construct the engineered 1,5-PDO producing strain NT1003-P4-ΔYcjQ.

Figure 22. (a) pRSFDuet-sfp-MmCARQ302E-SpyTag-EutM- SpyCatcher-YahK plasmid (b) MmCAR by the protein scaffold. (c) The colony PCR validation diagram.

The effect of assembly on the activity of 1,5-PDO synthesis module was determined by detecting the level to convert 5-HV to 1,5-PDO. The 1,5-PDO production of the engineered strain NT1003-P4-ΔYcjQ was determined by the HPLC analysis. From the catalysis results, we found that the assembly of MmCAR into the scaffold of EutM could significantly increase the ability of the 1,5-PDO producing module, 1.8 fold-higher than that without the assembly. Through a fermentation experiment, the assembly of MmCAR could increase 1,5-PDO production to 23 mM (2.4 g/L) in the engineered E. coli NT1003-P4-ΔYcjQ.

Figure 23. (a) Whole-cell catalytic results of scaffold 1,5-PDO sysnthesis module to convert 5-HV to 1,5-PDO. (b) The 1,5-pentanediol production by the engineered strain of E. coli NT1003-P4-ΔYcjQ.

In this cycle, we learned that the assembly of enzymes based on the protein scaffold is an efficient strategy for improving the titer of the targeted product. In comparison to the rational enzyme engineering strategy, the enzyme assembly was more convenient, especially aiming the enzymes that lack understanding of catalytic mechanisms. Through the literature study and communication with experts, we also learned that the assembly of multiple enzymes in a metabolic pathway is also an efficient strategy for the microbial metabolic engineering. In the future work, we will further optimize the assembly strategy to improve 1,5-PDO production.

The overall goal of the project was to develop a cell factory to produce 1,5-PDO from glucose through the synthetic biology. Firstly, an artificial 1,5-PDO biosynthetic pathway was designed. The screening of the functional enzymes successfully fulfill the pathway for 1,5-PDO production from glucose in a lysine-producing strain Escherichia coli NT1003. To further improve 1,5-PDO production, we optimized the gene expression in a high copy number plasmid. Subsequently, we knocked out the branched pathways involving in acetic acid accumulation and 1,5-PDO degradation. What is more, focusing on the limiting enzymes of MmCAR, we engineered the enzyme by a rational engineering strategy, and obtained the mutant enzymes with higher activity. Meanwhile, we employed an assembly strategy to further improve the activity of the limiting enzymes. Based on these engineering strategies, an 1,5-PDO producing strain was successfully developed and optimized. Figure 24 shows the results obtained after each attempt.

Figure 24. Results involved 1,5-PDO content obtained at each step.

[1].Cen, X. et al. Metabolic Engineering of Escherichia coli for De Novo Production of 1,5-Pentanediol from Glucose. ACS. Synth. Biol. 10, 192-203. (2021).

[2].Chen, C. et al. Metabolic engineering of Escherichia coli for the efficient production of 5-hydroxyvaleric acid. Process. Biochem. 130, 625-633.(2023)

[3].Xin, W. et al. Efficient production of 5-aminovalerate from l-lysine by engineered Escherichia coli whole-cell biocatalysts. Journal of Molecular Catalysis B: Enzymatic. 134,115-121.(2016).

[4].Zhang, G. et al. Self-Assembling Protein Scaffold System for Easy in Vitro Coimmobilization of Biocatalytic Cascade Enzymes. ACS. Catal. 8(6), 5611-5620 (2018).

[5].Zhang, G. et al. Developing a Protein Scaffolding System for Rapid Enzyme Immobilization and Optimization of Enzyme Functions for Biocatalysis. ACS. Synth. Biol. 8(8), 1867-1876. (2019).