Team:CAU China/Design

Design

Our project is intended to employ the synthetic biology methods to engineer E. coli cells which can utilize cellulose as well as improve the added value of the products from cellulose degradation. The modular approach allows us to divide the entire problem into two subsections that can be processed in parallel. To deal with the lengthy pathway and other considerations such as metabolic burden, we come up with a simple solution in order to optimize our project design and complete the experimental tasks as much as possible within the scheduled time.

Expression of cellulases in Escherichia coli

The cellulose can be degraded by the synergy of β-1, 4-endoglucanase (BBa_K118023), β-1,4-exoglucanase (BBa_K118022) from Cellulomonas fimi , and we employ β-glucosidase (BBa_K3279007) from Streptomyces coelicolor to release glucose. We transfer the genes encoding these enzymes into E. coli to allow the cell to obtain the ability to break cellulose into glucose. The enzymes are induced by IPTG and activities are determined by 3,5-dinitrosalicylic acid (DNSA) method separately. We set several inducing conditions to obtain the maximum enzyme activity. We also plan to join three genes together in tandem on the vector and transfer the plasmid into E. coli cells, so that cells can degrade cellulose completely.

Bacteria Surface Display Using N terminus of Ice-nucleation Protein

Ice-nucleation protein (INP) is a secreted outer membrane protein. It is widely distributed in Pseudomonas syringae, Pseudomonas fluorescens and other Gram-negative bacteria. Compared with other surface carrier proteins, ice-nucleation protein has the advantage of stably expressing heterogenous proteins and displaying proteins with larger molecular weight⁶. Compared with the intact INP, which is made up by N domain, C domain and internal sequences in between, or INP-NC used by team Edinburgh in 2011, which is a composite of N domain and C domain, we adopt N terminus of INP, which only contains N domain of INP and hence is much shorter and easier to fuse with the target proteins. Nevertheless, a new C terminus is required. So we add a linker sequence to accomplish the scheme.

Figure 1 Different strategies of surface display by INP(adapted from team Edinburgh)A. Target proteins fused with intact INP, B. Target proteins fused with INP-NC(used by Edinburgh), C. Target proteins fused with INP-N (Our scheme)

We intend to fuse the cellulases with the INP N-terminal sequence to accomplish the purpose of degrading cellulose on the surface of bacteria as well as downsizing the INP carrier. The fusion effect on enzyme activity can be detected by determining the cellulose degradation ability of fusion proteins. We also intend to detect the presence of the target protein by immunofluorescence staining.

Figure 2 The circuit of cellulase expression system

Astaxanthin

Figure 3 Astaxanthin synthesis pathway

In the astaxanthin biosynthesis section, we need to construct six astaxanthin synthesis genes (see description) into one plasmid and transform the constructed plasmid into E. coli cells.

Escherichia coli BL21 (DE3), which is often used for prokaryotic expression of proteins, was selected as the chassis organism for this experiment, while Escherichia coli DH5$\alpha$ was used for plasmid preservation.

In the selection of plasmids, considering that plasmids with strong promoters and of high copies often cause metabolic pressure of strains. Besides, the accumulation of secondary metabolites may cause toxicity to cells, and plasmids of high copies often get lost unexpectedly in bacteria¹. On the contrary, plasmids of low copies can optimize the level of products by alleviating metabolic pressure². Therefore, we choose low copy number plasmids as the carriers of astaxanthin synthesis genes. The vectors selected in this experiment were pACYC184-M and ptrc99A-M, which were donated by Zhang Xueli, a researcher at Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences.

In the selection of gene source, the strain provides the target gene should be easy to culture, and its genome sequence should be known. Moreover, considering that the chassis organism is E. coli, so the microorganism which provides the target gene should be prokaryotic. Based on the above reasons, we finally decided to obtain the first two genes CrtE and CrtB from Rhodobacter sphaeroides, the third gene CrtI from Rhodospirillum rubrum, and the fourth gene CrtY from Pantoea agglomerans. The CrtZ gene was derived from Pantoea ananatis, which was difficult to obtain, so we chose to synthesize it by the company. The BKT gene was derived from an Eukaryote, Chlamydomonas reinhardtii, which has different codon preferences than E. coli, so the BKT gene was synthesized by the company and the codon was optimized.

In the process of constructing astaxanthin producing strain, considering the long gene pathway, we divided the construction of engineered strain into three steps: lycopene production, $\beta$-carotene production and astaxanthin production.

The selection of restriction enzyme sites should consider the cutting efficiency and the restriction enzyme sites should not exist on the target gene sequence. Based on these two requirements, we selected three efficient restriction enzyme sites in pACYC184-M and ptrc99A-M plasmids, named KpnI, BamHI and HindIII respectively. Six genes need to be transferred into our pathway, and the number of restriction enzyme sites is limited. Therefore, we considered to connect genes in each step with overlap PCR firstly, and then splicing them into the vector by restriction enzyme sites. Considering there is no RBS on pACYC184-M and ptrc99A-M plasmids, and the genes in E. coli are polycistronic, we added an RBS (BBa_B0034⁴) before each gene to ensure the successful expression of each protein.

In the actual process of construction of lycopene producing strain, we successfully use overlap PCR to connect the first two genes (CrtE, CrtB), but the CrtI gene has been unable to overlap with the CrtE-CrtB junction product. Therefore, we use the seamless cloning kit to connect CrtE - CrtB junction product and CrtI, thus successfully build the lycopene production strains. Figure 1 shows the atlas of successfully constructed pACYC184-M-EBI plasmid.

Figure 4 Recombinant plasmid atlas for lycopene production

On the basis of pACYC184-M-EBI plasmid, we only need to insert the fourth gene CrtY between BamHI and HindIII, and then we obtained the constructed plasmid pACYC184-M-EBI-Y for $\beta$ -carotene production (see Figure 2 for plasmid atlas ).

Figure 5 Recombinant plasmid atlas for $\beta$-carotene production

Due to the lack of available cleavage sites on the plasmid, only HindIII remained for the construction of CrtZ and BKT. Considering that the cloning with single restriction enzyme sites will produce more false positives clonies, we consider to connect the last two genes CrtZ and BKT with overlap PCR, then use seamless cloning kit to clone them into the HindIII restriction enzyme site of plasmid pACYC184-M-EBI-Y. If the construction is successful, the strain used to produce astaxanthin can be obtained, and the atlas information is shown in figure 3.

Figure 6 Recombinant plasmid atlas for astaxanthin production

However, the efficiency of seamless cloning was low, so we also consider to insert the last two genes CrtZ and BKT into ptrc99A-M (see figure 4), and co-transforming E. coli BL21 with the recombinant plasmid pACYC184-M-EBI-Y.

Figure 7 Atlas for coconversion of astaxanthin production plasmid

In order to determine the activity of individual enzymes in the astaxanthin synthesis pathway, we induced individual enzymes separately. In this part of job, there is no need to consider the metabolic pressure of the strain, and the strong promoter contributes to the high expression of the target protein. Thus, we selected the plasmid pET-30a (+) as our vectors and the T7 promoter to activate the target gene expression. The restriction enzyme sites were BamHI and HindIII. In the experiment, we used gradient IPTG concentration and gradient temperature to induce each protein, exploring the optimal IPTG concentration and temperature of single protein expression and detected the enzyme activity of single enzyme. So that we can provide a more detailed description of our new parts. The recombinant plasmid profiles of each enzyme are shown in the table below.