Modified artificial'two-step 'terpenoid production way and high-yield rimane type terpenoid by using escherichia coli

Abstract

The invention discloses a modified artificial'two-step 'terpene production way and high-yield rimane type terpenoids by using escherichia coli, hydrolase and drimenol synthase are subjected to fusion expression for the first time, and the yield is increased by 7.3 times. The phosphorylation efficiency is further improved by 3.6 times by combining a two-step method path PhoN with rational design, finally, the yield of drimenol is improved by 27 times by optimizing fermentation conditions and is improved to 398.4 mg / L from 15 mg / L, a high-yield system is successfully migrated to production of albicanol, and the yield of 3.5 g / L is obtained.

Description

Technical Field

[0001] The invention relates to the improvement of a biosynthesis method, and in particular to the transformation of an artificial "two-step method" terpene production pathway and the use of Escherichia coli to produce high-yield drimane-type terpene compounds. Background Art

[0002] Terpenoids are mixed natural products partially derived from the terpenoid pathway, mainly found in fungi, marine organisms and higher plants, with few reports in bacteria. Among them, drimane-type merosesquiterpenoids (DMTs) have attracted much attention due to their structural diversity and remarkable biological properties. To date, more than 500 DMTs with different structures and potent biological activities have been identified from natural resources. The diversity of DMTs structures endows them with a wide range of biological activities. For example, a series of metabolites with a drimane skeleton, pyripyropenes AR, have cholesterol acyltransferase (ACAT) inhibitory activity. At the same time, drimane-type natural products such as polygonialdehyde have good antifungal and antiviral activities, and are currently widely used as antibacterial enhancers.

[0003] Traditional chiral raw materials commonly used in the synthesis of DMTs (such as sclareolide and sclareol) face great challenges, such as complex synthetic routes, low product yields, and atomic diseconomy caused by carbon reduction. In order to meet these challenges, researchers have continuously explored and discovered other chiral building blocks, such as drimane-type compounds drimenol and albicanol, which have higher atom economy as direct drimane cores in DMT synthesis.

[0004] Therefore, increasing the yield of drimenol and albicanol and obtaining a large amount of compounds through cheap production will play an important role in promoting the subsequent development of their functional products. The current methods for producing drimenol and albicanol include chemical synthesis, but the steps are cumbersome and environmentally unfriendly. Previous studies have obtained drimenol and albicanol through heterologous expression, such as expressing plant-derived drimenol synthase PhDS in tobacco leaves to obtain a fresh weight yield of 392μg / g, and selecting Saccharomyces cerevisiae as the chassis for heterologous expression of albicanol synthase AncC, and only 8.3mg of product was obtained in 4L fermentation medium for 96h. There are two biosynthetic pathways for terpenoids: the classical MVA pathway and the MEP pathway. In the past decade, with the development of biosynthetic technology, the use of artificial "two-step method" to synthesize terpenoids has gradually attracted people's attention. The "two-step method" pathway reaction system adds commercial five-carbon alcohol compounds exogenously, and after two-step phosphorylation reactions, the terpenoid biosynthetic precursor compounds DMAPP and IPP can be generated. Compared with the classical pathway, the advantages are: fewer enzymes involved, the reaction process is easy to control, and it has more industrial production and application prospects. At the same time, for the first step of the two-step pathway selected in this study, the acid phosphatase PhoN, as a phosphatase that catalyzes a reversible reaction, there have been related studies that have measured the kinetics of its catalytic reaction, and found that the Km value of its phosphorylation reaction is much greater than that of the hydrolysis reaction.

[0005] The yield of drimenol produced by fermentation based on the existing "two-step method" is low, only 15 mg / L, far from the level of industrial production application.

[0006] In recent years, most of the research on optimizing the "two-step" fermentation method for producing terpenoids has focused on trying to replace phosphatases from different sources, comparing phosphatases with the best yields, and optimizing fermentation conditions. There is currently no research on modifying phosphatases in the two-step pathway to increase yields. Summary of the invention

[0007] Purpose of the invention: The purpose of the present invention is to first solve the problem of poor catalytic effect of endogenous hydrolases in the "two-step method" for producing drimenol by introducing the hydrolase SsNDH from Streptomyces and constructing a fusion expression system. By comparing the effects of linkers of different lengths on the yield, the fusion system with the optimal linker length is selected to improve the in vivo fermentation yield of Escherichia coli.

[0008] At the same time, in order to improve the phosphorylation efficiency of the first step phosphatase PhoN in the "two-step method", an experimental idea based on charge mutation and conservative evolution was designed to screen high-yield mutants and improve the efficiency of phosphorylation reactions, essentially providing a transformation idea for the "two-step method" to produce high-yield terpene compounds.

[0009] After the transformation, the high-yield system was transferred to the production of albicanol by replacing the drimenol synthase SsDMS with the albicanol synthase AncC. The fermentation conditions were optimized and the fermentation tank was used to produce a large amount of albicanol.

[0010] Technical solution: The modified acid phosphatase PhoN of the present invention is derived from the acid phosphatase PhoN of Shigella flexneri, in which the 122nd amino acid is converted from glutamic acid to arginine, the 157threonine is converted to lysine, and the 160th arginine is converted to lysine, and the amino acid sequence is shown in SEQ ID NO: 1.

[0011] The nucleotide sequence of the modified acid phosphatase PhoN is shown in SEQ ID NO:2.

[0012] The application of the modified acid phosphatase PhoN in the biosynthesis of terpenoid compounds. The application of the modified acid phosphatase PhoN in the high yield of terpenoid compounds. The application of the modified acid phosphatase PhoN in the high yield of sesquiterpenoid compounds.

[0013] A fusion hydrolase is a fusion of SsNDH protein from Streptomyces showdoensis and drimenol synthase SsDMS, and the amino acid sequence is shown in SEQ ID NOs: 3, 5, 7, and 9.

[0014] The fusion hydrolase has a nucleotide sequence as shown in SEQ ID NOs: 4, 6, 8, and 10.

[0015] A method for producing a drimane-type terpenoid compound comprises the following steps: (1) raw materials DMAA and ISO are subjected to the action of acid phosphatase PhoN and phosphokinase IPK to generate DMAPP and IPP; (2) isopentenyl pyrophosphate isomerase (IDI) is further added to promote the mutual conversion between DMAPP and IPP; (3) farnesyl diphosphate (FPP) is formed under the action of FPP synthase IspA; (4) drimenol is generated under the action of drimenol synthase SsDMS and hydrolase; or albicanol is generated under the action of albicanol synthase AncC.

[0016]

[0017] In the production method, the acid phosphatase PhoN in step (1) is the phosphatase PhoN derived from Shigella flexneri or the modified acid phosphatase PhoN.

[0018] In the production method, the hydrolase in step (4) is an endogenous hydrolase of Escherichia coli or an exogenous hydrolase SsNDH.

[0019] In the production method, the drimenol synthase SsDMS or hydrolase in step (4) may be replaced by the fusion hydrolase.

[0020] The invention discloses an Escherichia coli for efficiently producing drimane-type terpene compounds, comprising an Escherichia coli cell into which expression vectors for Ipk (nucleotide sequence as shown in SEQ ID NO: 11), IspA (nucleotide sequence as shown in SEQ ID NO: 12), IDI (nucleotide sequence as shown in SEQ ID NO: 13), the modified acid phosphatase PhoN, and AncC (nucleotide sequence as shown in SEQ ID NO: 14) are introduced; or an Escherichia coli cell into which expression vectors for Ipk (nucleotide sequence as shown in SEQ ID NO: 11), IspA (nucleotide sequence as shown in SEQ ID NO: 12), IDI (nucleotide sequence as shown in SEQ ID NO: 13), the modified acid phosphatase PhoN, and the fusion hydrolase are introduced.

[0021] 1. The present invention is the first to fuse the hydrolase from Streptomyces with the drimenol synthase for expression, thereby increasing the yield from 15 mg / L to 110 mg / L, an increase of 7.3 times.

[0022] 2. The present invention is the first to use a "two-step" approach to PhoN through rational design and charge and conservative modification, increasing the yield from 110 mg / L to 398.4 mg / L, the phosphorylation efficiency by 3.6 times, and the total yield by 27 times.

[0023] 3. The high-yield system constructed by the present invention was successfully transferred to the production of albicanol, and through the optimization of fermentation conditions and fermentation tank production, a yield of 3.5 g / L was obtained.

[0024] Beneficial effects: The present invention is the first to express the fusion of hydrolase and drimenol synthase, which increases the yield by 7.3 times. In combination with the rational design and transformation of the two-step pathway PhoN, the phosphorylation efficiency is further increased by 3.6 times, and the drimenol yield is increased by 27 times, from 15 mg / L to 398.4 mg / L. The high-yield system is successfully transferred to the production of albicanol, and the yield of 3.5 g / L is obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Figure 1 is a diagram of the transformation strategy of cyclase fusion hydrolase SsNDH; Figure A is a schematic diagram of using endogenous hydrolases of E. coli, introducing exogenous hydrolase SsNDH, directly fusing SsNDH with SsDMS, and introducing GSG linker in the middle of the fusion expression protein, and explores the effect of GSG linker length of 1-4 on the fermentation yield of drimenol; Figure B is a bar graph of the yield of different construction methods in Figure A, Endo-hy represents the use of endogenous hydrolases of E. coli, Exo-hy represents the introduction of exogenous hydrolases without fusion expression, fus-n represents the fusion expression of hydrolase and drimenol synthase, and the addition of GSG linker, n represents the number of times the linker is repeated, the blue bar graph has a yield unit of mg / L, and the orange bar graph has a yield unit of mg / g; Figure C shows the hydrolase fused to the cyclase predicted by AlphaFold2. This figure shows the three-dimensional structure of the hydrolase and cyclase fusion protein predicted by AlphaFold2. The fusion direction refers to the order in which the hydrolase SsNDH is located upstream of the cyclase SsDMS in the reference gene cluster, the hydrolase is located at the N-terminus (red), and the cyclase is located at the C-terminus (blue); the figure shows that there is a disordered loop between the two enzymes, and the blue represents the introduced GSG linker;

[0026] Figure 2 Figure 1 is a rational transformation strategy diagram of the phosphatase PhoN; Figure A shows the docking results of PhoN and DMAP molecules, the orange structure is the homologous protein 1D2T structure, and the pink structure is the simulated protein PhoN structure; PhoN docking with ligand molecules shows the surrounding Residues within; Figure B is the fermentation result of alanine scanning of catalytic pocket residues; To verify the accuracy of docking, alanine scanning mutation test was performed on PhoN catalytic residues, and it was found that the activity was significantly lost, indicating that the simulated PhoN protein structure has a high accuracy; Figure C is the simulation result of mutants based on charge modification; charge mutation was performed on active pocket residues, and its effect on the charge around the pocket was simulated, where the redder the color, the more negative charge is accumulated, and the bluer the color, the more positive charge is accumulated; the position selected by the circle is the location of the active pocket; the figure shows that the positive charge near the active pocket of the E122R / K mutant increased significantly; Figure D is the fermentation result of the mutant strain with increased positive charge in the active pocket of PhoN; This figure is the fermentation yield diagram of mutation designed based on charge, and the result is similar to the conclusion drawn from the simulation, The yield of E122R and E122K mutants is higher than that of wild type, indicating that increasing the positive charge of the active pocket is beneficial to increasing the yield; Figure E is an evolutionary analysis of the conservative residues of PhoN family proteins; the figure shows the conservation of PhoN family proteins in evolution, and the degree of conservation is divided into 9 levels, red represents conservative residues, green represents non-conservative residues, and non-conservative residues closer to the active pocket are marked with arrows in the figure; Figure F is a WebLogo diagram of non-conservative residues in evolution, and the arrows point to non-conservative residues in evolution; Figure G is the fermentation results of mutation transformation of non-conservative residues of PhoN; the fermentation results of mutant strains designed according to the conservative analysis in Figure F showed that the drimenol yield of the three mutant strains S90G, T157K and R160K was slightly higher than that of the wild type, indicating that improving conservation is beneficial to increasing yield;

[0027] Figure 3 Figure 1 is the effect of combined mutations on PhoN yield and the purification and crystallization of high-yield triple mutant proteins; Figure A is the fermentation result of combined mutation of dominant mutants; combined mutant fermentation, the dominant mutants screened by the two strategies were combined and fermented separately. It can be seen from the figure that the yield of the triple mutant combined with E122R was significantly improved, among which the E122R / T157K / R160K triple mutant had the highest yield; Figure B is the high-yield triple mutant PhoN E122R / T157K / R160K Figure 3 shows protein expression, purification and crystallization of PhoN. E122R / T157K / R160K Crystalline α-helix structure diagram (PDB ID: 8YC1); Figure D is a comparison of PhoN E122R / T157K / R160K Compared with the wild-type PhoN, the three mutant residues have slight differences in position; Figure E shows PhoN E122R / T157K / R160K The wild-type PhoN and the ligand DMAP were subjected to molecular dynamics simulation for 100 ns. The orange and blue lines represent PhoN, respectively. E122R / T157K / R160K Compared with the wild-type PhoN simulation results, the results showed that the mutant prolonged the time of stable interaction between the enzyme and the ligand;

[0028] Figure 4 Optimization of shake flask fermentation conditions for albicanol; Figure A shows the effect of the amount of five-carbon alcohol added on the yield of albicanol; the figure shows the addition of five-carbon alcohols with different concentration gradients, and the preference of five-carbon alcohols DMAA and ISO is explored. The result shows that the fermentation yield of 30mM ISO is the best; Figure B shows the effect of IPTG addition on the yield of albicanol. This figure explores the effect of different IPTG concentrations on the yield. When the inducer addition concentration is 0.05mM, the fermentation yield is the highest; Figure C shows the effect of glycerol addition on the yield of albicanol. This figure explores the effect of different glycerol additions on the yield. When 2% glycerol is added, the fermentation yield is the highest; Figure D shows the effect of fermentation time on the yield of albicanol. This figure explores the effect of different fermentation times on the yield. When the fermentation is on the 4th day, the yield is the highest;

[0029] Figure 5 The fermenter produces albicanol. The graph shows the growth of bacteria in the fermenter over time. 600 , the changes of glucose and product albicnaol, the inducer IPTG was added when OD reached 30, and the first DMAA was added, and the second DMAA was added at the 34th hour of fermentation. DETAILED DESCRIPTION

[0030] The artificial "two-step method" pathway was selected to produce terpenoid production precursors DMAPP and IPP, and the raw materials DMAA and ISO were input to generate DMAPP and IPP under the action of acid phosphohydrolase PhoN and phosphokinase IPK. In addition, isopentenyl pyrophosphate isomerase (IDI) was added to promote the mutual conversion between DMAPP and IPP. Next, farnesyl diphosphate (FPP) was formed under the action of FPP synthase IspA, and drimenol and albicanol were generated under the action of bacterial drimenol and albicanol synthase SsDMS and AncC.

[0031]

[0032] The present invention is based on laboratory preliminary research. First, a three-plasmid system for producing drimenol is constructed. Acid phosphatase PhoN from Shigella flexneri is selected as the first step phosphokinase of the two-step pathway. Phosphokinase IPK from Thermoplasmaacidophilum is selected as the enzyme of the second step. Five-carbon alcohol substrates 3-methyl-2-butene-1-ol (DMAA) and 3-methyl-3-butene-1-ol (ISO) are added during the fermentation process. After two consecutive phosphorylation reactions, terpenoid biosynthetic precursors DMAPP and IPP are obtained. Further, farnesyl pyrophosphate FPP is generated under the action of farnesyl pyrophosphate synthase IspA from E. coli. At the same time, isopentenyl pyrophosphate isomerase IDI is introduced. Then, isopentenyl pyrophosphate is generated under the action of drimenol synthase SsDMS (amino acid sequence NCBI accession number: A0A2P2GK84). After that, drimenol can be generated under the action of endogenous hydrolases in E. coli. The plasmid construction method is shown in Example 1.

[0033] Example 1

[0034] Construction of three plasmids for producing drimenol

[0035] pCDFDuet-Ipk-ispA-IDI plasmid was constructed, and primers were designed (Forward: TCATCACCACAGCCAGGATATGGACCCTTTTACCATGATGATT, Reverse: GCATTATGCGGCCGCAAGCTTTTATTTAAGCTGGGTAAATGCAGATAA) to clone ipk, ispA, and idi genes (nucleotide sequences are shown in SEQ ID NOs: 11-13) from the original plasmid pLD10012 in the laboratory by PCR. The pCDFDuet-1 vector was double-digested with BamHI and HindIII (Takara Company) and the synthase gene was homologously recombined into the vector by ClonExpress II One Step Cloning Kit and transformed into E. coli DH5α. Positive clones were selected by PCR identification and identified by sequencing (sequencing primers: ACYCDuetUP1: GGATCTCGACGCTCTCCCT, T7 terminator: GCTAGTTATTGCTCAGCGG).

[0036] Construction of pRSFDuet-PhoN plasmid and design of primers (Forward:

[0037] The phoN gene was cloned from the original plasmid pLD10012 in the laboratory by PCR, and the pRSFDuet-1 vector was double-digested with BamHI and HindIII (Takara Company), and the synthase gene was homologously recombined into the vector by ClonExpressIIOne Step Cloning Kit and transformed into E.coli DH5α. Positive clones were selected by PCR and identified by sequencing (sequencing primer: DuetDOWN 1: GATTATGCGGCCGTGTACAA).

[0038] pETDuet-SsDMS plasmid construction and primer design (Forward:

[0039] The ssdms gene (nucleotide sequence is shown in SEQ ID NO: 15) was cloned from Streptomyces pyrotechnica by PCR, the pETDuet-1 vector was digested with KpnI (Takara) and the synthase gene was homologously recombined into the vector by ClonExpressIIOne Step Cloning Kit and transformed into E.coli DH5α, the positive clones were selected by PCR and identified by sequencing (sequencing primer: DuetDOWN1: GATTATGCGGCCGTGTACAA).

[0040] Example 2

[0041] Production of drimenol and albicanol by in vivo fermentation of Escherichia coli and post-processing method

[0042] Before fermentation, the strain to be fermented was inoculated into a test tube of 5 ml LB liquid medium containing 0.1% resistance, and cultured in a 37°C shaker at 230 rpm for 12 hours as seed liquid. After that, the seed liquid was inoculated into a 250 ml conical flask containing 50 ml TB medium at 0.5%, and 0.1% of the corresponding antibiotics were added. 2% glycerol was added before sterilization of the TB medium. After inoculation, the culture was placed in a 37°C shaker at 230 rpm for about 2.5 to 3 hours until the OD 600After reaching 1.3-1.4, the mixture was quickly cooled on ice, and then 0.25 mM isopropyl-β-D-thiogalactoside (IPTG) inducer was added in a clean bench, and the mixture was cultured in a shaker at 18°C ​​and 200 rpm for 72 h.

[0043] After the fermentation is finished, take two ep tubes, one of which is weighed with an analytical balance, and the fermentation liquid is vortexed and mixed, and 1 ml is taken and added to the ep tube, and the supernatant is poured out after centrifugation at 12000rpm in an eppendorf high-speed centrifuge for 5min, and the weighed ep tube is placed in a 60°C oven to dry to constant weight, and weighed again with an analytical balance, and the dry weight (DCW) of 1 ml fermentation liquid is deducted from the weight of the empty tube. After pouring out the supernatant from the other ep tube, 500ml of acetone is added, and the bacterial precipitation and acetone are vortexed and mixed, and then ultrasonicated for 4h, with an ultrasonic frequency of 40kHz and a power of 200W. After centrifugation at 12000rpm in a high-speed centrifuge for 20min, the supernatant acetone layer is taken and passed through an organic filter membrane and analyzed by HPLC, with an injection volume of 5μ L each time, and the fermentation yield is calculated based on the peak area, and the yield of drimenol produced by the three plasmids is calculated to be 14.7mg / L.

[0044] Example 3

[0045] Construction of drimenol synthase fusion hydrolase expression plasmid

[0046] The current production system relies on the endogenous hydrolase of E. coli. In order to increase the fermentation yield of drimenol, it is considered to introduce exogenous hydrolases, such as Figure 1 As shown in A, the SsDMS cycle of drimenol synthase in Streptomyces cushowdoensis was investigated.

[0047] The surrounding genes were analyzed and there was a SsNDH protein with a functional annotation of hydrolase upstream (the amino acid sequence NCBI accession number is A0A2P2GK86, and the nucleotide sequence is shown in SEQ ID NO: 16). It was constructed and expressed on the same plasmid as SsDMS. Each gene was preceded by a T7 promoter. The fermentation yield of drimenol was 63.0 mg / L, which was 4.3 times higher than that without adding exogenous hydrolase, confirming the speculation that the catalytic efficiency of endogenous hydrolase was insufficient. Afterwards, in order to further increase the yield, the hydrolase and cyclase were tried to be fused and expressed. The fusion protein was predicted by AlphaFold2, as shown in the attached figure. Figure 1C, the C-terminus of the hydrolase and the N-terminus of the cyclase both have a disordered loop domain. The presence of this disordered loop makes the active pocket of the hydrolase close to the top of the βγ domain of the active pocket of the cyclase, allowing the product of the cyclization reaction, pyrophosphate, to reach the catalytic center of the hydrolysis reaction more quickly. Therefore, a fusion expression plasmid was constructed, and the yield was 99.7 mg / L when the two proteins were directly fused, which was 1.6 times higher than that of non-fusion expression. Afterwards, in order to extend the disordered loop region in the middle of the fusion protein, referring to the method of fusion expression of terpenoid synthase and p450 oxidase, a flexible Gly-Ser-Gly (GSG) linker was added, and the optimal length of the repeat sequence was explored. The results showed that when (GSG) n When linker n=3, the yield was the highest, 110.8 mg / L, slightly higher than the yield without linker and 7.5 times higher than the fermentation yield without hydrolase, indicating that the fusion of hydrolase has a significant effect on the yield increase. The fermentation data are shown in Figure 2. Figure 1 B, therefore, the subsequent study chose to add (GSG) 3linker fusion expression protein for subsequent optimization.

[0048] Specific operation method:

[0049] Construction of pETDuet-Hydro-SsDMS plasmid (non-fusion expression): Design primers (Forward: TCATCACCACAGC CAGGATGTGACGGCGCACGAAACG, Reverse: GCATTATGCGGCCGCAAGCTTTCACGTCGGACGGGTGGT) to PCR out hydrolase A0A2P2GK86 from the Streptomyces howdoensis genome, double-digest the pETDuet-SsDMS plasmid with BamHI and HindIII, and homologously recombinant the hydrolase gene into the vector, which was then transformed into E. coli DH5α, and positive clones were selected by PCR identification and identified by sequencing (sequencing primer: DuetDOWN1: GATTATGCGGCCGTGTACAA).

[0050] Construction of pETDuet-Hydro-SsDMS(R) plasmid (fusion expression): Primer pairs (Forward: TCATCACCACAGCCAGGATGTGACGGCGCACGAAACG, Reverse: TGATGCGTTCATCGTCGGACGGGTGGTGGG) and (Forward: GTCCGACGATGAACGCATCACCGACGC, Reverse: GCATTATGCGGCCGCAAGCTTTCATCGGGAGCAGCCTTCG) were designed to PCR amplify hydrolase SsNDH and drimenol synthase SsDMS from pETDuet-Hydro-SsDMS plasmid, respectively. The pETDuet-1 plasmid was double-digested with BamHI and HindIII, and the hydrolase gene was homologously recombined into the vector and transformed into E. coli DH5α. Positive clones were selected by PCR identification and identified by sequencing (sequencing primers: DuetDOWN1: GATTATGCGGCCGTGTACAA, pET Upstream Primer: ATGCGTCCGGCGTAGA).

[0051] Table 1 Primer correspondence table of linkers of different lengths

[0052]

[0053] GSG linkers of different lengths were added to the fusion protein. The construction method was the same as that of the fusion plasmid pETDuet-Hydro-SsDMS(R). The difference was that the linkers of different lengths replaced the corresponding Forward primers of SsNDH and Reverse primers of SsDMS. The primer sequence table is shown in Table 1.

[0054] Example 4

[0055] AlphaFold2 predicts hydrolase and cyclase fusion proteins

[0056] Use the AlphaFold2 online website ( https: / / colab.research.google.com / github / sokrypton / ColabFoldblob / main / AlphaFold2.ipynb#scrollTo=6xbvRNrwnJqj ) predicts the protein structure of hydrolase fused to drimenol synthase, and the protein structure with (GSG)3linker at the connection is predicted. The pdb file of the fusion protein is predicted and displayed using PyMol as follows Figure 1 C.

[0057] Example 5

[0058] SWISS-MODEL homology modeling identifies the active pocket

[0059] To determine the active pocket for the interaction between PhoN and its ligand, homology modeling of PhoN was performed using SWISS-MODEL ( https: / / swissmodel.expasy, or g / ) website, input the fasta format sequence file of protein PhoN, search for templates, and find the homologous protein 1D2T with a sequence similarity of up to 84.4%. By comparing the active pocket residues and conserved motif residues of 1D2T with PhoN, the conserved residues and active pocket of protein PhoN were determined.

[0060] Determine the active pocket. The PhoN structure was simulated on the AlphaFold2 online website, and the PhoN homologous protein 1D2T was found using SWISS-MODEL homology modeling, with a sequence similarity of 79.9%. The two proteins were compared using the visualization software PyMol and found to be highly similar to the simulated PhoN protein 1D2T, with an RMSD value of According to literature reports, the conserved motifs of the phosphohydrolase of 1D2T protein, KX6RP-PSGH-SRX5HX3D, are 115K, 122R, 123P, 147P, 148S, 149G, 150H, 182S, 183R, 189H, and 193D, respectively, and the conserved motifs corresponding to the PhoN protein are 133K, 140R, 141P, 165P, 166S, 167G, 168H, 200S, 201R, 207H, and 211D, respectively. At the same time, the active sites of 1D2T are 115K, 122R, 148S, 149G, 150H, 183R, 189H and 193D, and the active sites of PhoN are 133K, 140R, 166S, 167G, 168H, 201R, 207H and 211D. In order to further determine the accuracy of the conservative site selection, these 8 sites were mutated to A respectively. The mutant construction method is shown in Example 6. By comparing the yield changes through in vivo fermentation, it was found that the activity of the mutant strain was almost lost, which verified that these sites were active pockets. The fermentation results are shown in Figure 2 B.

[0061] Example 6

[0062] Mutant construction method

[0063] Based on the construction of pRSFDuet-PhoN plasmid, mutant intermediate primers were designed. The upstream forward primer was unified as TCATCACCACAGCCAGGATATGAA GCGCCAGCTGTTTAC. The upstream reverse primer and downstream forward primer were different according to different mutant strains, as shown in Table 2. The downstream reverse primer was unified as GCATTATGCGGCCGCAAGCTTTTACTTTTTCTGATTATTGGCAAATTC. PCR was first performed with the upstream and downstream primer pairs respectively. The PCR product fragment was used as a template and then the upstream forward primer and downstream reverse primer were selected for one PCR to obtain the full-length mutant PhoN fragment. The pRSFDuet-1 vector was double-digested with BamHI and HindIII (Takara Company), and the synthase gene was homologously recombined into the vector by ClonExpress IIOne Step Cloning Kit and transformed into E. coli DH5α. The positive clones were selected by PCR identification and identified by sequencing (sequencing primer: DuetDOWN1: GATTATGCGGCCGTGTACAA).

[0064] Table 2. Mutant primers

[0065]

[0066]

[0067] Example 7

[0068] Molecular docking to identify active pocket residues

[0069] AutodockTools-1.5.7 was used for docking. The catalytic product DMAP with a pyrophosphate group was selected for docking with PhoN. The docking box size was 46*46*46, and the grid spacing was The center of the box is the default receptor active center, with coordinate values ​​of (x=5.983, y=3.649, Z=-0.559). The GeneticAlgorithm genetic algorithm was used in the ligand conformation search process, and the number of GA runs was set to 100. After the docking, the ligand conformation with the lowest binding energy was taken as the optimal conformation. The active pocket and the residues that interacted with the ligand DMAP were determined by molecular docking. After removing the conserved residues, the remaining residues near the pocket were 122E, 125G, 158R and 171R.

[0070] The energy of the docking result is -5.95, Figure 2A. The residues that have hydrogen bond interactions with DMAP are 133K, 140R, 165P, 166S, 167G, 168H and 207H. Figure 2 A, all are conservative motifs, no mutation sites are designed. In addition to the conservative motifs, the distance from the substrate There are also 122E, 125G, 158L, 171I within the range of Figure 2 A, These four residues are located around the isopentenyl hydrophobic chain of the substrate and are also the residues on the outside of the active pocket, which has an important influence on the entry of the substrate into the active cavity.

[0071] Example 8

[0072] It is designed to mutate positions 122, 125, 158 and 171 into positively charged basic residues. Since the imidazole ring of the histidine side chain has a greater impact on steric hindrance and its basicity is slightly weaker than arginine and lysine, it is considered to mutate residues 122, 125, 158 and 171 into arginine and lysine respectively to simulate the charge change of the active pocket after mutation. The method is shown in Example 8, and the simulation results are shown in Figure 2 C, where the “#” in the wild type represents the location of the active pocket, and the positions selected by the circles are the locations of the mutation sites. According to the simulation results, the E122R / K mutant not only significantly increases the positive charge near the side chain of the extra arginine, but also significantly increases the positive charge around the active pocket, which is in line with the expectations of the mutation design. After the 125G position mutated to R / K, it was found that the extra side chain not only had a poor effect on increasing the positive charge in the pocket, but also caused a certain barrier to the pocket, affecting the entry of the substrate. Since 158L and 171I are not on the surface of the protein, the mutation has little effect on the increase in surface positive charge near the pocket. The above speculation was verified by in vivo fermentation experiments. The fermentation results are shown in Figure 2. Figure 2 D. The drimenol production of the E122R mutant reached 177.6 mg / L, which was 1.6 times higher than that of the wild-type strain (110.8 mg / L), indicating that increasing the positive charge in the pocket did have an effect on improving the phosphorylation reaction. After 125G mutated to R and K, the production dropped significantly, reaching only 17.1 mg / L and 8.8 mg / L, which was consistent with the simulation results that the G125R / K mutant hindered the pocket. The production of the L158R / K and I171R / K mutants was significantly reduced. Considering that the two sites are not located on the protein surface, the phenomenon of reduced production cannot be explained only from the charge perspective. It is speculated that the mutation has an adverse effect on the interaction between the substrate and the protein.

[0073] Mutants simulate charge changes

[0074] The pdb file of the protein PhoN is displayed in PyMOL to generate the protein contact potential of vacuum electrostatics. The protein shows surface charge, red represents negative charge and blue represents positive charge. The redder (bluer) the color, the more negative (positive) charge is accumulated.

[0075] To simulate the effect of different mutants on the surface charge of the active pocket, first select the Mutagenesis-protein option under the Wizard option in the PyMOL toolbar, simulate the mutation of residues 122E, 125G, 158R and 171R into basic residues arginine and tyrosine, respectively, analyze and compare the effects of the mutant residues on the surface charge of the active pocket, and perform experimental verification.

[0076] Example 9

[0077] Conservation analysis of PhoN family proteins

[0078] Use ConSurf WEB SERVER to perform conservative analysis ( https: / / consurf.tau.ac.il / consurf_index.php ), select amino acid conservation analysis, upload the pdb file predicted by protein PhoN, select PSI-BLAST for homologous protein comparison, select 1 for the number of iterations, set the E-value to 0.0001, and select SWISS-PROT for the protein database; select 500 sequences to be analyzed, select MUSCLE for the multiple sequence alignment method, select Bayesian for the method of calculating evolutionary conservation, select Best model for the evolutionary model, and then submit the analysis task.

[0079] The analysis results are divided into 9 levels according to the degree of conservatism, such as Figure 2 E, red represents conservation, green represents non-conservation, combined with the active pocket in Example 7, which is highly conserved in evolution, near the active pocket Within the distance, there are 6 non-conserved residues at positions 90S, 135Y, 153K, 157T, 160R, and 222I, which are used as target mutation sites. The high-frequency amino acid residues at these sites are counted, and the statistical results are shown in Figure 2 F. Ten mutants, S90A, S90G, Y135K, Y135H, K153T, T157K, R160K, R160T, I222V, and I222A, were further designed. The pRSFDuet-PhoN plasmid in the three-plasmid system was replaced with PhoN with different mutation sites for co-expression fermentation. The fermentation results are shown in Figure 2G, among which the drimenol production of the three mutants S90G, T157K and R160K were 125.8, 134.8 and 135.4 mg / L, respectively, which were slightly higher than the production of 110.8 mg / L of the wild-type strain, indicating that mutating the phosphatase PhoN towards the conservative residues of the family protein evolution is beneficial to improving the catalytic efficiency of the enzyme. It is speculated that in the long process of protein family evolution, the enzyme always evolves towards a more stable, more efficient and other beneficial directions. It is feasible to use this strategy to transform the enzyme to improve the catalytic efficiency of the enzyme and then achieve a high-yield solution.

[0080] Example 10

[0081] Combination mutation

[0082] The three mutation sites of the advantageous mutant strains S90G, T157K and R160K selected based on the conservative mutations were first combined in pairs, and the construction method was the same as in Example 6. The results are as follows: Figure 3 As shown in A, the yields of the double mutants S90G / T157K, S90G / R160K, and T157K / R160K were 162.1, 150.0, and 162.7 mg / L, respectively, which were 1.5, 1.4, and 1.5 times higher than the wild type 110.8, respectively. However, the effect of the combination of the three mutants was very low, only 7.9 mg / L. Then, the dominant mutation site E122R selected based on the charge distribution of the active pocket and the mutation sites S90G, T157K, and R160K selected based on conservative transformation were combined into three mutation sites, and the results are shown in Figure 3 As shown in A, the yield of the three combinations of triple mutation sites is significantly higher than that of the single E122R mutation and the conservative combination mutation, among which E122R / T157K / R160K 三三The yield of the site mutation combination reached 398.4 mg / L, which was 2.2 times higher than the yield of the single-site E122R mutant strain, 2.4 times higher than the yield of the conservatively designed T157K combination R160K double mutant strain, and 3.6 times higher than the yield of the wild-type strain. This shows that the mutation sites designed by the two strategies have a cumulative effect on improving the yield of enzyme modification. The design of the E122R site is to improve the catalytic efficiency of phosphorylation in the "phosphorylation-hydrolysis" reversible reaction of acid phosphatase PhoN, thereby increasing the yield, and the screening of the T157K / R160K double mutation is based on conservative analysis. The mutant strain is presumed to improve the stability of the phosphatase PhoN. Therefore, Expasy is used to predict the change of the isoelectric point of the mutant strain. The method is shown in Example 11, and the prediction results are shown in Table 3. The isoelectric point of the wild-type PhoN is 6.08Mw, and the E122R mutant is 6.08Mw. The isoelectric point of the mutant strain was significantly improved to 6.88Mw, which was related to the increase in positive charge near the active pocket. The isoelectric point of the T157K / R160K double mutation was 6.42Mw, indicating that the conservative mutation changed the overall protein from acidic to neutral, increasing protein stability, while the isoelectric point of the best triple mutant strain was 7.72Mw, making the overall protein alkaline, indicating that increasing alkalinity can increase the phosphorylation activity of PhoN, which is consistent with the previous speculation that increasing positive charge can increase the interaction with the pyrophosphate group and increase phosphorylation activity.

[0083] Embodiment 11

[0084] Isoelectric point prediction

[0085] Expasy was then used to predict the change in isoelectric point of the mutants. The amino acid sequences of different mutants were input and the prediction results are shown in Table 3. The isoelectric point of the wild-type PhoN was 6.08Mw, and the isoelectric point of the E122R mutant was significantly increased to 6.88Mw, which was related to the increase in positive charge near the active pocket. The isoelectric point of the combination of T157K and R160K was 6.42Mw. w , indicating that conservative mutations change the overall protein from acidic to neutral, increasing protein stability, while the isoelectric point of the best triple mutant was 7.72Mw, which made the protein alkaline as a whole, indicating that increasing alkalinity can increase the phosphorylation activity of PhoN, which is consistent with the previous speculation that increasing positive charge can enhance the interaction with pyrophosphate groups and increase phosphorylation activity.

[0086] Table 3 Relationship between the change of isoelectric point of mutants and the yield of buxinol

[0087]

[0088]

[0089] Example 12

[0090] Protein expression and purification

[0091] PhoN E122R / T157K / R160K The plasmid was transformed into E. coli BL21 (DE3). The strain was cultured in 1L LB at 37°C and shaken at 230 rpm until OD 600 The pH reached 0.6. The culture was cooled to 4°C and induced by adding 0.25 mM IPTG. The cells were grown for 20 hours at 18°C ​​and 200 rpm under shaking conditions. After harvesting the cells by centrifugation at 4000 rpm for 15 minutes at 4°C, the cells were resuspended in lysis buffer (50 mM Tris, pH 8.0, containing 500 mM NaCl and 10% glycerol), lysed by sonication, and then centrifuged at 15000 rpm for 20 minutes at 4°C. The supernatant containing soluble PhoN was passed through a HisTrap column equipped with Purification was performed by nickel affinity chromatography using an FPLC system (GE Healthcare Biosciences). Desalting was performed using a PD-10 desalting column (GE Healthcare Biosciences) and concentration was performed using an Amicon Ultra-15 concentrator (Millipore) in 50 mM Tris, pH 7.5, containing 100 mM NaCl and 10% glycerol to obtain a protein with an N-terminal His6 tag. Protein crystallization was performed ( Figure 3 B).

[0092] Embodiment 13

[0093] Protein crystallization and structure determination

[0094] PhoN E122R / T157K / R160K After initial screening and optimization, the protein was observed to be suitable for structural determination in 0.1M sodium acetate trihydrate (pH = 4.0), 0.2M magnesium formate dihydrate and 18% PEG 4000 solution. The crystal structure with a resolution of 2.7A was obtained. In order to collect X-ray data, the crystal was quickly immersed in a buffer containing 15% glycerol and immediately frozen in liquid nitrogen after being taken out. E122R / T157K / R160K The X-ray diffraction data were collected at the BL02U1 beamline of the Shanghai Synchrotron Radiation Center (Shanghai Light Source). The diffraction data were processed using iMosflm, and the phase was performed by molecular replacement in PHASER5. The molecular replacement experiment used the PhoN protein structure simulated by AlphaFold2 as the starting model. The warpNtrace routine in ARP / wARP was used to automatically build the initial model. Manual modeling was then performed in COOT. Model refinement (including TLS) was performed in PHENIX6. The final resolved protein structure is a trimer, and the monomer structure is shown in Figure 3 C. Comparison with PhoN E122R / T157K / R160KThe structural differences with wild-type PhoN are that the three mutant residues have slight differences in position, see Figure 3 D;PhoN E122R / T157K / R160K The wild-type PhoN and the ligand DMAP were subjected to molecular dynamics simulation for 100 ns. The blue and black lines represent PhoN, respectively. E122R / T157K / R160K The simulation results with wild-type PhoN are shown in Figure 3 E, The mutant prolongs the time that the enzyme and ligand interact stably.

[0095] Table 4 PhoN E122R / T157K / R160K Crystal structure collection and optimization

[0096]

[0097]

[0098] Embodiment 14

[0099] Optimization of albicanol fermentation conditions

[0100] In order to further improve the yield of albicanol in Escherichia coli fermentation, the optimal concentration and method of addition of five-carbon alcohol, the amount of IPTG added, the amount of glycerol added and the fermentation time were explored. Figure 4 ,After experimental verification, the optimal yield of 1804.9 mg / L was obtained by adding 30 mM ISO, 2% glycerol, 0.05 mM IPTG and fermenting for 4 days.

[0101] Embodiment 15

[0102] Fermentation tanks for the production of albicanol

[0103] Given that gram yields of albicanol have been achieved in shake flask production, in order to optimize the production of albicanol, we expanded the fermenter production to obtain more albicanol. In the early fermentation stage, the dissolved oxygen level was maintained at 30% by controlling agitation and aeration. The pH was maintained at 7.0 by automatic addition of NH3-H2O. When the initial D-glucose was exhausted, D-glucose was supplemented by adding feed medium to keep the D-glucose concentration between 0.1 g / L and 1 g / L. Due to the toxicity of isopentenyl alcohols, we adopted a staged feeding method to minimize the adverse effects on strain growth. In the first stage, when OD 600When it reached 30, the fermentation tank temperature was lowered to 18°C ​​and IPTG was added for induction. At the same time, 8 grams of DMAA and 0.1mM IPTG were added. In the second stage, the albicanol production reached 1.6g / L at 34 hours. At this time, 8 grams of DMAA was added to ensure sufficient supply of precursors. Finally, the highest albicanol production reached 3.5g / L at 53 hours. Figure 5 This means that the engineering of the kinase PhoN has gained generalizability.

Claims

1. A modified acid phosphatase PhoN, characterized in that The 122nd amino acid of the acid phosphatase PhoN from Shigella flexneri is converted from glutamic acid to arginine, the 157threonine is converted to lysine, and the 160th arginine is converted to lysine. The amino acid sequence is shown in SEQ ID NO:

1.

2. The modified acid phosphatase PhoN according to claim 1, characterized in that The nucleotide sequence is shown in SEQ ID NO:

2.

3. Use of the modified acid phosphatase PhoN according to claim 1 in the biosynthesis of terpenoids.

4. A fusion hydrolase, characterized in that: The SsNDH protein and drimenol synthase SsDMS derived from Streptomyces cushowdoensis are fused, and the amino acid sequence is shown in SEQ ID NO: 3, 5, 7, and 9.

5. The fusion hydrolase according to claim 4, characterized in that The nucleotide sequences are shown in SEQ ID NOs: 4, 6, 8, and 10.

6. A method for producing a drimane-type terpenoid compound, characterized in that: The following steps are involved: (1) The raw materials DMAA and ISO generate DMAPP and IPP under the action of acid phosphatase PhoN and phosphokinase IPK; (2) In addition, isopentenyl pyrophosphate isomerase (IDI) is added to promote the mutual conversion between DMAPP and IPP; (3) farnesyl diphosphate (FPP) is formed under the action of FPP synthase IspA; (4) drimenol is generated under the action of drimenol synthase SsDMS and hydrolase; Alternatively, albicanol is produced under the action of albicanol synthase AncC.

7. The production method according to claim 6, characterized in that: The acid phosphatase PhoN in step (1) is the phosphatase PhoN derived from Shigella flexneri or the modified acid phosphatase PhoN according to claim 1.

8. The production method according to claim 6, characterized in that: The hydrolase in step (4) is an endogenous hydrolase of Escherichia coli or an exogenous hydrolase SsNDH.

9. The production method according to claim 6, characterized in that: The drimenol synthase SsDMS or hydrolase in step (4) may be replaced by the fusion hydrolase described in claim 4.

10. An Escherichia coli for efficiently producing drimane-type terpene compounds, characterized in that: Escherichia coli cells containing expression vectors for Ipk, IspA, IDI, the modified acid phosphatase PhoN according to claim 1, and AncC; Or an Escherichia coli cell containing an expression vector into which Ipk, IspA, IDI, the modified acid phosphatase PhoN according to claim 1, and the fusion hydrolase according to claim 4 are introduced.

Images