Results

Tasks

1. Constructing HQTs knockout Nicotiana benthamiana.
2. Validating the chassis function through the heterologous synthesis of phaselic acid, resveratrol, and crocin.
3. Optimizing key enzymes involved in the synthesis of target products through semi-rational design and directed evolution methods.

1.Construction of HQT knockout plants

This part of the experiment was designed to reduce chlorogenic acid levels in tobacco, thereby enabling greater metabolic flow to other metabolites, thus creating a high-quality tobacco synthetic biology chassis. We identified target genes by analyzing genes related to chlorogenic acid synthesis and the metabolic flow of the pathway, and subsequently knocked them out. Metabolic flow analysis of the target products was then conducted in the knockout plants to confirm that the constructed knockout plants served as efficient expression chassis for the target products.

Identification of Genes for Knockout

During the metabolic flux analysis, we conducted knockout simulations on the three synthetic pathways of chlorogenic acid, comparing a total of six knockout methods involving the knockout of one gene and the knockout of two genes. The simulation results indicated that knocking out the HQT pathway is the most effective way to significantly reduce the production of the byproduct chlorogenic acid and increase the flux of the target product synthesis.
See details

Design of sgRNA for Knockout

We accessed the genome database of Nicotiana benthamiana (https://www.nbenth.com/) and downloaded the sequences of four HQT genes contained in Nicotiana benthamiana: NbL07g01330.1, NbL10g00550.1, NbL13g01170.1, and NbL16g00540.1. Subsequently, we utilized the online software CRISPR-P 2.0 (http://crispr.hzau.edu.cn/CRISPR2/) to predict the gene editing targets. For sgRNA design, we selected 20 bp exon sequences adjacent to the PAM site (-NGG) located near the 5'-end of the genes as the basis.

Table 1-1: Designed sgRNA Sequences

Construction of Knockout plasmids

Using homologous recombination, vectors were constructed by designing target site sequences and BsmBI recognition sites into primers. The target site sequences were incorporated between the U6 promoter and gRNA sequences via PCR. Subsequently, BsmBI endonuclease was used to insert the target site sequences into the BGFastCas9 SS Plant vector, resulting in the construction of CRISPR/Cas9 single knockout vectors for NbHQT1, NbHQT2, NbHQT3, and NbHQT4. To enhance knockout efficiency, in addition to constructing single mutants for each of the four genes, vectors were also constructed for the simultaneous knockout of NbHQT1 and NbHQT3, as well as NbHQT2 and NbHQT4, generating double knockout CRISPR/Cas9 vectors for NbHQT1+3 and NbHQT2+4.

Table 1-2: Design of Single Knockout CRISPR Primers

The plasmid was transformed into E. coli DH5α competent cells, spread onto plates, and incubated overnight. Single clones were picked for further culture. A pair of primers was designed based on the upstream and downstream sequences corresponding to the sgRNA. Using the DNA exposed after bacterial lysis as a template, PCR was performed with the aforementioned primers. The amplified bands represented the region near the two target sequences on the plasmid, with a length of approximately 300bp, which was consistent with the theoretical length. The sequencing results of the amplified products were correct, confirming the successful transformation of the knockout plasmid into E. coli DH5α.

Construction of Knockout plant

Agrobacterium tumefaciens Construction

We successfully transformed the target plasmid into Agrobacterium tumefaciens, obtaining the desired strains: single-gene knockout strains NbHQT1-EHA105, NbHQT2-EHA105, NbHQT3-EHA105, and NbHQT4-EHA105 (Fig.1-4, Fig.1-5); and double-gene knockout strains NbHQT1+3-EHA105 and NbHQT2+4-EHA105 (Fig. 1-5).

Stable Transfection, Plant Tissue Culture, Budding, Rooting, Acclimatization, Transplanting, and Cultivation to Obtain the T0 Generation

The single-gene knockout stable transformation callus of NbHQT2, NbHQT3, NbHQT4 and the double-gene knockout stable transformation callus of NbHQT2+4 could be induced and grow normally (Fig.1-6). In contrast, the callus of NbHQT1 and NbHQT1+3 gradually became enlarged and vitrified and could not differentiate into resistant buds usually (Fig.1-7).

The first batch of resistant shoots from callus tissue was transplanted onto rooting medium. Compared to the wild-type, the edited Nicotiana benthamiana plants exhibited slower root development, with only a small fraction of plants forming roots after approximately four weeks of culture. Additionally, the time required for rooting varied depending on the specific genes that were edited. The double-gene knockout plants demonstrated slower growth overall. In general, the single-gene knockout plants (NbHQT2, NbHQT3, and NbHQT4) rooted more quickly than the double-gene knockout plants (NbHQT2+4). (Fig.1-8)

The DNA from the CRISPR-edited tissue culture seedlings was extracted, and the Cas9 editing region was amplified by PCR (Fig.1-9) and subsequently sent for sequencing.

After the sequencing results are returned, they are compared with the correct gene sequence. Comparison of the sequencing results with the gene sequences revealed that the corresponding target NbHQT gene had been knocked out in some plants of each CRISPR strain, and single deletion mutants of NbL10g00550.1 (NbHQT2), NbL13g01170.1 (NbHQT3), and NbL16g00540.1 (NbHQT4) were successfully obtained (Fig.1-10, Fig.1-11, and Fig.1-12).

The sequencing results were compared with the gene sequences of NbHQT2 and NbHQT4, respectively. It was found that NbHQT2 and NbHQT4 had been knocked out in plant No. 5 of the NbHQT2+4-CRISPR strain, successfully resulting in the creation of the NbHQT2+4 double-gene knockout mutant (Fig.1-13).

The CRISPR-positive Nicotiana benthamiana plants with established root systems underwent hardening and transplanting (Fig.1-14).

Sowing T1 generation and then sequencing plants

Wild-type, mutant 2-4, and mutant 2-8 tobacco seedlings were transplanted into new pots, with two plants per pot, and placed in an incubator after transplantation (Fig.1-15). Based on the band size, the primers successfully amplified the gene. However, it was unclear whether the gene had been successfully knocked out, so the remaining PCR products were sent for sequencing (Fig.1-16).

Conclusion

NbHQT is a key gene involved in chlorogenic acid synthesis in tobacco. Based on our modeling results, we successfully knocked out the HQT gene in the Nicotiana benthamiana genome in various combinations, constructing a chassis with reduced chlorogenic acid levels. This breakthrough addresses the issue of high chlorogenic acid concentrations diverting significant metabolic flow away from the production of secondary metabolites, which previously limited the synthetic capacity of the chassis. By lowering chlorogenic acid levels, we have created a high-quality chassis for plant synthetic biology, enhancing the production potential of Nicotiana benthamiana and paving the way for further exploration of its ability to synthesize a wide range of high-value-added compounds.

2.Synthesis of Phaselic Acid

Background and Introduction

Phaselic acid is a common acylated acid found in plants, formed by the combination of Caffeoyl-CoA and malic acid. In order to produce this compound in our gene-edited Nicotiana benthamiana, hydroxycinnamoyl transferase (HCT) is required. An enzyme called HCT-M, derived from red clover, has a high affinity for malic acid. Introducing this gene into Nicotiana benthamiana provides the necessary activity for the synthesis of phaselic acid. Additionally, codon optimization of the expression system in Nicotiana benthamiana was carried out to verify the plant's ability to synthesize the target compound downstream of the shikimate pathway.

Plasmid Construction and Codon Optimization

We first performed PCR amplification of the codon-optimized HCT-M gene for E. coli, obtaining the target gene, followed by gel extraction for further use.

We selected the pDONR207 and pEAQ-HT-DEST1 to facilitate the efficient introduction of the gene into Agrobacterium.

Fig.2-2 Plasmid Maps of pDONR207-HCT-M and pEAQ-HT-HCT-M

Using pDONR207 as the vector, the HCT-M gene was inserted through a BP reaction. After the BP reaction, the product was transformed into Top10 competent cells and plated on LB plates with gentamycin resistance to screen for positive clones. After inoculating the selected clones into LB broth, the plasmids were extracted to obtain the entry clone plasmid. Agarose gel electrophoresis was performed on the extracted plasmids, and the plasmids were sent for sequencing. The results confirmed that the HCT-M gene was successfully inserted into the vector.

Fig.2-3 Colony Formation and Colony PCR Detection Results After BP Reaction

After the BP reaction, we extracted plasmids and sequenced them. Once the BP reaction was successful, we performed the LR reaction using the entry clone plasmid and destination vector to obtain the expression plasmid. The product was plated on kanamycin-resistant LB plates to screen for positive clones. After inoculating the selected clones into LB broth, the plasmids were extracted to obtain expression plasmids. Agarose gel electrophoresis was performed on the extracted plasmids, and the plasmids were sent for sequencing. The results confirmed that we successfully constructed the expression plasmid.

Fig.2-5 Colony Formation and Colony PCR Detection Results After LR Reaction

We transformed the expression plasmid into Agrobacterium. The transformed Agrobacterium was cultured in LB medium containing antibiotics, where single colonies successfully grew. We performed colony PCR and agarose gel electrophoresis. The results confirmed that the target gene was successfully introduced into Agrobacterium.

Transient Transformation

The Agrobacterium culture was prepared into a suspension, with the OD600 adjusted to approximately 0.9, for transient transformation in tobacco plants. Using a syringe, the suspension was injected into the underside of tobacco leaves. The injected plants were covered with black plastic bags and placed in an incubator for cultivation. The injection sites were marked.

Fig.2-9 Tobacco Injection Results

LC-MS/MS Results

The injection areas of the tobacco leaves were cut, freeze-dried for 24 hours, and metabolites were extracted using 80% methanol for LC-MS/MS analysis. The results showed a significant increase in phaselic acid concentration after Agrobacterium injection. This indicates that the introduced HCT-M gene was successfully expressed in tobacco and catalyzed the synthesis of phaselic acid from endogenous malic acid and caffeoyl-CoA. Additionally, the mutant tobacco plants produced more phaselic acid than the wild-type plants, suggesting that Nicotiana benthamiana with HQT knockout has a higher capacity for phaselic acid synthesis.

Conclusion

By introducing the exogenous HCT-M gene, Nicotiana benthamiana was able to synthesize phaselic acid. Furthermore, the mutant tobacco demonstrated an enhanced ability to synthesize phaselic acid compared to the wild-type tobacco. This indicates that after knocking out the HQT gene, Nicotiana benthamiana gained an increased capacity to synthesize target products downstream of the shikimate pathway. However, since the mutant tobacco missed the optimal time for transfection, chlorogenic acid had already begun to accumulate, reducing the plant's ability to synthesize phaselic acid. In future experiments, earlier transfection times should be considered to achieve better synthesis performance.

3.Synthesis of Resveratrol

This part aims to validate whether the increased p-Coumaryl-Coenzyme A, an essential chemical in the MEP pathway, could be used to synthesize medicinal secondary metabolite resveratrol in bulk. Considering that Nicotiana benthamiana lacks the key enzyme -- Stilbene synthase in catalyzing p-Coumaryl-Coenzyme A to produce resveratrol, we focus on the widely verified stilbene synthase gene VvSTS originating from grapes. We transferred it into tobacco and analyzed the production of resveratrol. Moreover, in the upstream of p-Coumaryl-Coenzyme A, L-Phe represses the DHAP synthase through a negative feedback loop, resulting in a reduction in p-Coumaryl-Coenzyme A. DHAP synthase gene YIARO3 from Saccharomyces is the mutant of ARO3 and can inhibit the negative feedback. Thus, we hope to improve the resveratrol production by over-expressing YIARO3 in our chassis. Finally, we conducted LC-MS/MS to analyze resveratrol production.

Construction of Expression Plasmid

VvSTS and YIARO3 fragments were synthesized and optimized by a professional company. We added attB sequences at both ends of the genes. We used Gateway clone and exerted BP reaction first to attach VvSTS and YIARO3 to vector pDONR207 to form entry clones.

The sequencing results showed that the critical regions were utterly consistent with the standard sequences, and no mutations occurred, indicating success in constructing entry clones.

Next, we conducted an LR reaction to attach VvSTS and YIARO3 to expression vector pEAQ-HT-DEST1 and to regulate expression by CaMV35S strong promoter. The sequencing results showed no mutation in critical regions, indicating that pEAQ-DEST1-YIARO3 and pEAQ-DEST1-YIARO3 expression plasmids were constructed successfully.

To enable expression in Nicotiana benthamiana, we transferred the pEAQ-DEST1-YIARO3 and pEAQ-DEST1-YIARO3 plasmids into Agrobacterium GV3101. The sequencing results showed no mutation in critical regions, indicating successful transfers.

LC-MS/MS Analysis to Resveratrol Production

We transiently transfected the tobacco with Agrobacterium, carrying targeted genes, and ran an LC-MS/MS non-target metabolism analysis. We drew the standard curve with the data and calculated the hqt - knockout plants and wild-type resveratrol production.

Above all, we drew the standard curve according to the correspondence between the area and concentration of resveratrol standards. Compared to wild-type tobacco, T0-generation hqt2 - knockout plants increased resveratrol production somewhat, regardless of VvSTS single gene transfer and VvSTS and YIARO3 two gene transfer. These results matched our anticipation.

However, we found that resveratrol production of mutant and wild-type tobacco transferred by VvSTS and YIARO3 was slightly lower than that of single VvSTS. We speculated that the sample volume was too small and that the plant difference significantly influenced the results. Consequently, we enlarged the sample volume when transferring T1 - generation hqt2 - knockout tobacco. We mixed the samples from the same gene combination group to narrow the difference between different plants and leaf layers. We observed that hqt2 - knockout plants produced more resveratrol than the wild-type when injected with a single VvSTS (p < 0.05), which was what we expected.

Nevertheless, hqt2 - knockout plants transferred with VvSTS and YIARO3 were lower than that of wild-type. Similarly to T0 - generation, resveratrol production transferred with two genes was significantly decreased compared to VvSTS transfer, and it was more evident in hqt2 - knockout plants. We suspected that DHAP synthase encoded by YIARO3 was competitively combined with DHAP, resulting in possible influence in other metabolic pathways and finally reducing resveratrol production. The speculation needed to be validated.

In conclusion, we successfully used Gateway to construct VvSTS and YIARO3 expression plasmids and transferred them into hqt2- knockout plants through Agrobacterium to verify resveratrol production increase in hqt2 - knockout plants. This part proved that our chassis has advantages in synthesizing secondary metabolites. However, it remains to test whether YIARO3 can improve resveratrol production.

4.Synthesis of Crocin

We chose crocin as the synthetic product for analysis to evaluate the metabolic changes in other metabolic pathways such as mevalonate pathway (MEP pathway) besides the shikimate pathway. By reading the relevant literature, we selected the multiple gene superposition synthesis method, which expresses four genes related to crocin synthesis (GjCCD4a, GjALDH2C3, GjUGT74F8, and GjUGT94E13) in Nicotiana Benthamiana.

Plasmid construction

Integrating multiple genes into the plasmid will make the plasmid too large resulting in the reduction of plasmid transformation efficiency and stability and increasing experimental operation complexity and difficulty. We planned to construct four plasmids for each gene. Thus, the efficiency of plasmid construction and the expression of exogenous genes in the host would be improved. We cloned GjCCD4a, GjALDH2C3, GjUGT74F8 and GjUGT94E13 into the pEAQ-DEST1 plasmid vector and used the 35S promoter to constitutively express the exogenous gene.

Construct GjCCD4a、GjUGT74F8 and GjUGT94E13 expression vector by gateway clone.

Gateway consists of 2 main reactions: BP reaction and LR reaction. We commissioned Sangon Biotech (Shanghai) Co. Ltd to synthesize CDS sequences for four genes, which sequenced was codon-optimized. And these sequences contain attB sites at both ends. By BP enzyme, we recombined the target fragment into the gateway cloning vector pDONR207 (Fig.4-1). Then the constructed plasmid was transformed into E. coli TOP10 (Fig.4-2). Finally, the colony PCR products were sent for sequencing (Fig.4-3). The sequencing results showed that three entry plasmids successfully constructed.

After confirmation of the successful construction of the entry plasmids, in the LR reaction, the entry plasmid pDONR207-gene was recombined with the expression vector pEAQ-DEST1 under the action of LR enzyme to form the final expression clone. The constructed expression plasmid was transfected into E. coli TOP10 (Fig.4-4). Then the constructed plasmids were transformed into E. coli TOP10 (Fig.4-5). Subsequently, the PCR products were sent for sequencing (Fig.4-6). The sequencing results showed that three expression plasmids successfully constructed.

Add a His-tag to the vector by Gibson assembly

Since no His-tag was added at the end of the gene sequence when we designed. We added His-tag to each gene end by Gibson assembly (Fig.4-7). Then the target fragment is obtained by PCR, and the vector was linearized.

The target bands were cut, and Gibson assembly was performed using a ClonExpress II One Step Cloning Kit, incubated at 50°C for 5 min, and the constructed plasmids were inserted into E. coli TOP10 (Fig.4-8). Finally, the PCR products were sent for sequencing (Fig.4-9).

So far, the expression vector of GjCCD4a, GjUGT74F8 and GjUGT94E1 had been built successfully.

Construct GjALDH2C3 expression vector by Gibson assembly.

After several failed BP reactions to GjALDH2C3, we realized that not all genes could be used to construct expression vectors by Gateway cloning, so we used Gibson assembly for expression vector construction.

Due to the large size of the vector, to increase the efficiency of linearized PCR of the vector, we removed unnecessary sequences when designing the primers (Fig.4-10).

We obtained the target linearized gene by PCR and performed the Gibson assembly after gel extraction. Then we transformed the product into E. coli TOP10. Then 28 single colonies for colony PCR were selected randomly (Fig.4-11).

Electrophoresis results showed that the target bands appeared in 12 lanes, and we confirmed that the pEAQ-DEST1-ALDH was successfully constructed after sequencing.

After the expression plasmids of all genes were constructed, we inserted them into Agrobacterium GV3101 by chemical transformation, and then selected single colonies for colony PCR (Fig.4-13) and sent the PCR products for sequencing. Sequencing results showed that the plasmid was successfully transferred into Agrobacterium(Fig.4-14).

Transient transfection and metabolome extraction

We mixed different strains of Agrobacterium tumefaciens (GjCCD4a-GV3101, GjALDH2C3-GV3101, GjUGT74F8 -GV3101, and GjUGT94E13-GV3101) and added them to the suspension solution for injection into the leaves (Fig.4-15). After injection, the leaves were subjected to a 24-hour dark treatment.

After 48 hours of light cultivation, the leaves were cut for metabolome extraction (Fig.4-16). A round of concentrated extraction was performed. A total of five blank control tubes and nine experimental tubes of samples were prepared.

Crocin LC-MS/MS Result

We generated a standard curve for crocin I according to the LC-MS/MS results of the crocin I standard (Fig.4-17). Our samples were divided into four groups: Wild-type control, CR-nbhqt2 control, wild-type experimental, and CR-nbhqt2 experimental. Since the amount of each sample bottle in the WT experimental group could not reach the detection standard, we performed a sample mixing, which means there is only one result for the WT experimental group. Moreover, the value obtained after analysis can be used as the theoretical average of a one-sample t-test.

After analyzing the sample results (Fig.4-18), we found that crocin I was not detectable in the WT and CR-nbhqt2 control groups. However, crocin I was detected in both the WT and CR-nbhqt2 experimental groups. Compared to the WT experimental group, the CR-nbhqt2 experimental group showed a significant increase in crocin content. These results indicate that we successfully transfected the crocin metabolic pathway into Nicotiana Benthamiana and achieved its expression.

Summary

In conclusion, by transiently expressing four essential genes for crocin synthesis in Nicotiana Benthamiana, we found that the improved Nicotiana Benthamiana could synthesize more crocin I than wild-type Nicotiana Benthamiana, which means when the chlorogenic acid synthesis is turned down, the MEP synthesis will increase. In our experiment, we successfully synthesized crocin I. It indicates that the modified Nicotiana Benthamiana has a high application value. The crocin and other rare compounds with terpenoids as precursors are expected to be produced in large quantities in our Nicotiana Benthamiana.

5.Directed Evolution

For the production of resveratrol, wild VvSTS enzyme derived from grapes still has significant room for improvement in its catalytic activity. Therefore, we have further optimized the enzymatic activity and thermal stability of VvSTS through semi-rational design. For the production of phaselic acid, the catalytic activity of HCT-M derived from red clover is limited due to its weak affinity for caffeoyl-CoA, which tends to bind with feruloyl-CoA instead, resulting in a large amount of feruloyl malic acid by-products. To address these issues, on one hand, we have optimized the catalytic activity and thermal stability of the enzyme through semi-rational design. On the other hand, we have conducted directed evolution of the enzyme, hoping to screen for mutants with high substrate affinity for caffeoyl-CoA.

Design Principle and Process

In order to screen for mutants with high substrate affinity for caffeoyl-CoA, we conducted directed evolution of HCT-M and performed enzyme activity assays on various variants.

To construct plasmids carrying different HCT-M mutants, we performed error-prone PCR to introduce random mutations into the HCT-M gene and then used homologous recombination to ligate the mutated gene fragments into the pET-28a plasmid backbone. The constructed plasmids were transformed into E.coli BL21, and protein expression was induced with IPTG. The mutated HCT-M proteins were purified and the enzyme activities were tested to screen for mutants with higher catalytic activity.

Plasmid Construction

Control Plasmid Construction

PCR amplification was performed on the HCT-M fragment and the pET-28a plasmid to obtain the HCT-M gene fragment and the pET-28a backbone.

After agarose gel electrophoresis of the PCR fragments and gel extraction, homologous recombination was used to ligate the pET-28a-HCT-M plasmid, which was then transformed into E.coli DH5α. Colony PCR and sequencing results indicated successful ligation of the HCT-M gene to the vector. The plasmid from E.coli DH5α was extracted and transformed into E.coli BL21.

Random Mutant Plasmid Construction

Error-prone PCR was performed on the HCT-M to obtain randomly mutated HCT-M gene fragments; PCR was performed on the pET-28a plasmid backbone to obtain the pET-28a backbone.

After agarose gel electrophoresis of the PCR fragments and gel extraction, homologous recombination was used to ligate the pET-28a-HCT-M plasmid containing random mutations, which was then transformed into E.coli DH5α. Colony PCR indicated successful ligation of the HCT-M gene to the vector. The plasmid from E.coli DH5α was extracted and transformed into E.coli BL21.

Protein Expression and Activity Assay

Protein Purification

Single colonies were picked and cultured in LB medium containing kanamycin at 37 ℃, 220 rpm overnight. The bacterial culture was transferred into 50mL LB medium with kanamycin resistance at a 1% inoculation rate, cultured at 37 ℃, 220 rpm for 3 hours to achieve absorbance between 0.6-0.8. IPTG was added to a final concentration of 0.2mM, and the culture was induced at 24 ℃, 150 rpm for 16 hours to express the HCT-M protein.

After centrifugation and lysis of the bacteria, total bacterial protein was extracted. SDS-PAGE results showed that out of 13 proteins, 9 were successfully purified. Four mutant proteins still did not show clear bands, which may be due to the fact that after random mutations in the protein sequence, proteins may have been formed inclusion bodies or are unable to expose the His-tag, leading to the failure of purification of these four mutant proteins.

Enzyme Activity Assay

Protein concentration was determined using the Bradford protein concentration detection kit, with absorbance measured at 595 nm to calculate the protein concentration. In the reaction system, the concentration of tartaric acid was 300 μM, the concentration of caffeoyl-CoA was 300 μM, and the enzyme concentration was 100 ng/mL, with the reaction carried out at 30 ℃ for 30 minutes. The results showed that compared to the control group, the enzyme activity of mutant3 was increased, while the enzyme activities of mutant5, mutant6, and mutant7 were decreased, and the rest of the mutants did not show significant enzyme activity. This indicates that directed evolution can change the catalytic activity of the enzyme by altering key residues of the enzyme.

6.Semi-rational Design

Although directed evolution has achieved some success, it requires lots of experimental screening and has a low probability of obtaining favorable mutations. Therefore, we adopted a semi-rational design approach to screen for mutation sites through molecular docking, dynamic simulation, and other methods. Then, we established a mutation library and verified its effectiveness through wet experiments. The results showed that among the 7 HCT-M mutants, 2 had higher enzyme activity and thermal stability than the wild type. And among the 12 VvSTS mutants, 5 mutants had stronger enzyme activity at the optimal temperature than the wild type, and 4 mutants had stronger thermal stability than the wild type. Our work will help increasing the production of resveratrol and caffeoyl-CoA in tobacco chassis, and demonstrate that semi-rational design is a efficient protein optimization strategy with broad application prospects in plant synthetic biology.

Design Principles and Processes

Semi-rational design and codon optimization

Semi-rational design

Through three rounds of semi-reational design, we successfully screened 13 HCT-M mutants and 12 VvSTS mutants for validation. Firstly, we simulated single point saturation mutations of HCT-M and VvSTS on a computer, and used a pre trained protein language model to score all mutants in terms of stability and ligand affinity. We selected the mutants with higher scores. On this basis, protein and ligand molecular docking was performed to compare the binding stability of mutant and wild-type proteins and ligands, and to exclude mutants that are not conducive to stable binding. Finally, molecular dynamics simulations were conducted on the mutants selected in the first two rounds to evaluate the energy and stability of the entire reaction system from a physical perspective, and obtain mutants for validation by wet experiments

Codon Optimization

Considering that we will heterologous express HCT-M from Trifolium repens and VvSTS from grapes in Escherichia coli, it is highly likely that using the original sequence directly will result in a decrease in protein expression due to differences in amino acid coding preferences among different species. The common strategy for codon optimization is to replace codons with those preferred by the host organism and adjust their GC content to increase mRNA stability and adapt to the host organism's level. The genes of HCT-M and VvSTS were both codon optimized and synthesized by Sangon Biotech, and we conducted experiments at IDT.The effectiveness of codon optimization has been verified.

Plasmid Construction and Primer Design

The synthesized HCT-M and VvSTS gene both carry Histag and are loaded onto the pET-28a (+) plasmid, which was synthesized by Sangon Biotech. The reason for choosing pET-28a (+) is that it is a plasmid vector with a lac operon and T7 promoter, which can induce overexpression of exogenous genes through IPTG, effectively controlling the expression of exogenous genes and reducing their harmful effects on host cells. At the same time, high concentrations of expressed proteins help us perform enzyme function testing.

In order to perform site directed mutagenesis on wild-type genes, we designed primers based on the selected mutation sites. The primers carry corresponding mutation sites, which are marked in green in the table. The naming convention for mutants is: original amino acid abbreviation - mutation site - mutated amino acid abbreviation. For example, H294L is mutated from histidine at position 294 to leucine. The naming convention for primers is: mutation site of amino acid - abbreviation of mutated amino acid - F (forward primer)/R (reverse primer).

Table6-1: HCT-M Primer Sequence Carrying Mutation Sites

Table6-2: VvSTS Primer Sequence Carrying Mutation Sites

Construction of Point Mutation Plasmids

We used mutation primers to perform PCR reaction on wild-type plasmids and obtained linearized plasmids with mutation sites. After digesting the wild-type plasmids with dpn1 enzyme, the mutation products were transferred into competent E. coli DH5α, and plasmids were extracted and sequenced from E. coli DH5α after expanded cultivation.

As shown in the sequencing results, we have successfully constructed all the required mutants.

Enzyme Expression and Activity Detection

We induced protein expression and purification of the successfully constructed mutant HCT-M gene and VvSTS gene. The activity validation of HCT-M enzyme is characterized by detecting the remaining amount of malic acid, while the activity validation of VvSTS enzyme is characterized by directly detecting the amount of resveratrol.

HCT-M

For HCT-M, due to time and budget constraints, seven mutants including H294L, H294Y, A419T, A389N, K26F, K26S, and K26V were selected for protein characterization. After inducing IPTG expression, in order to unify the concentrations of various enzymes in subsequent enzymatic reactions, we measured the purified protein concentration using the Bradford method.

By analyzing the absorbance data, the concentration of mutant protein was determined as shown in the table below.

Table6-3: Protein concentration

The mutant protein and wild-type protein were diluted to the same concentration and subjected to the following enzymatic reaction: 300μM caffeoyl-CoA, 300μM malic acid, 50ng/μL HCT-M enzyme, and phosphate buffer to adjust the reaction system to pH 6.5. The reaction time is 30 minutes, and three temperature gradients are set to verify thermal stability. The temperature gradients are 30 ℃ (the optimal temperature for HCT-M enzyme), 35 ℃, and 40 ℃, respectively. Using specific enzyme activity as a measure of activity.

For the seven mutants, only three have activity, namely A419T, A389N, and K26F. At 30 ℃, the activity of A419T is lower than that of the wild type, while the activity of A389N and K26F is higher than that of the wild type. When the temperature is 35 ℃, the wild-type HCT-M loses its activity, but A419T, A389N, and K26F still have activity. At 40 ℃, A389N and K26F remain active. This indicates that semi-reational design can effectively construct enzymes with higher enzyme activity and higher thermal stability.

VvSTS

Due to the low solubility of VvSTS, inclusion bodies are easily formed. In order to increase the amount of normal VvSTS collected, we used TB medium as the source of bacterial nutrition and induced expression conditions of 0.1 mM IPTG, 21 ℃, 120 rpm, and 20 h. Finally, we successfully obtained sufficient protein concentration for the reaction through gravity column purification and ultrafiltration.

Afterwards, an enzyme activity assay experiment was conducted with a system of 50 μ L, consisting of the following components: VvSTS 0.4 μ g, coumarin-CoA 50 μ mol/L, malonyl-CoA 100 μ mol/L, pH 7.0, reaction time 60 min, and three temperature gradients (50 ℃, 55 ℃, 60 ℃).

After the reaction, the concentration of the product resveratrol was detected using the Resveratrol Immune Competition Kit (ELISA), and the results are as follows:

The reaction activity of wild-type VvSTS did not change significantly under three temperature gradients, indicating that VvSTS itself has good heat resistance. In terms of thermal stability, mutant W268S showed better reaction activity than WT at 55 and 60 ° C, while S219A and P269K showed better reaction activity than WT at 55 ° C, indicating that the semi-reational design constructed enzymes with better thermal stability. In terms of catalytic activity, the mutants with stronger catalytic activity than WT at the optimal temperature include S219A, W268L, S220A, and L114I, indicating that semi-reational design has constructed mutants with higher catalytic activity.

Outlook

The high chlorogenic acid accumulation in Nicotiana benthamiana limits its potential for synthesizing other products. To develop tobacco as a versatile chassis for the short-term, efficient synthesis of desired plant natural products, we first constructed a genome-scale metabolic model (GSMM) and used COBRA algorithms to formulate the best metabolic flow adjustment strategies. Subsequently, the critical gene NbHQTs for chlorogenic acid synthesis in tobacco was knocked out using CRISPR/Cas9, and a low chlorogenic acid level tobacco chassis was obtained. We then introduced heterologous synthetic pathways for phaselic acid, resveratrol, and crocin into the chassis and improved the synthesis capability by engineered critical enzymes. The chassis could quickly synthesize the desired functional products to fulfill different requirements and is expected to be developed into an efficient biological reactor for the biopharmaceutical industry, thereby promoting the green transformation and upgrading of the traditional tobacco cultivation industry.

During the iGEM competition cycle, the number of repetitions and technical iterations of our experiments is limited. Currently, the validation results for the potential of chassis synthesis are very limited and the credibility is questionable. It is necessary to select more types of compounds from other metabolic pathways, such as upstream and downstream in the MVA pathway, for exogenous synthesis to functionally validate our tobacco chassis. In addition, we only selected two key enzymes, VvSTS and HCT-M, for thermal stability and catalytic activity modification in order to increase the production of phaselic acid and resveratrol. However, other key enzymes used for synthesizing other products such as crocin were not optimized during the experimental period. For other compounds such as crocin, more efforts need to be made in enzyme modification and metabolic pathway design before applying our chassis for efficient production.

In the future, we will further optimize the chassis of Nicotiana benthamiana by knocking out more genes such as nicotine synthesis genes that occupy a large amount of metabolic flow, in order to release more metabolic flow potential for the synthesis of other downstream secondary compound products.

In terms of factory production, we will use soilless cultivation techniques to cultivate our chassis plants on a large scale in artificially controlled greenhouses, achieving the large-scale and systematic construction of plant synthetic biology "bioreactors" based on Nicotiana benthamiana as the chassis organism. We will also actively cooperate with biopharmaceutical companies to strive for the industrial implementation of the project.