Result
This year, we carried out experiments to optimize the isobutanol production in Zymomonas mobilis. Our results were divided into two parts: the first part was the isobutanol production pathway engineering, including the optimization of gene expression and the interference of ethanol production through CRISPRi system; the second part was associated with the carbon concentrating and fixation, including a synthetic carboxysome assembly and the Calvin cycle pathway transformation. The details of our results were displayed as follows.
Part Ⅰ Isobutanol Biosynthesis
Overview
As shown in the Part Improvement, the kdcA gene inserted in the genome of Z. mobilis was better than in the plasmid, and the over-expression of the optimized synthetic operon Bsals2-ilvC-ilvD (A5 plasmid) was necessary to divert carbon flux to isobutanol production. In order to further increase the isobutanol production in Z. mobilis, we decided to increase the gene copy number of kdcA in the genome, as well as to further optimize the expression of the operon Bsals2-ilvC-ilvD. In addition, considering that the ethanol metabolism pathway can divert carbon flux, another strategy to delete or suppress the ethanol pathway is needed to redirect carbon flux from ethanol to isobutanol biosynthesis. Therefore, the dCpf1 assisted CRISPRi system was employed to inhibit the pdc expression, which is the key gene involved in the ethanol production. All the plasmids and recombinant strains constructed in this part were shown in Table 1-1 and Table 1-2.
Table 1-1 The plasmids constructed for the isobutanol biosynthesis
| Plasmids |
Description |
| A5 |
pEZ15A-Ptet-Bsals2-Peno-ilvC-ilvD |
| A6 |
pEZ15A-Ppdc-Bsals2-Peno-ilvC-ilvD |
| A7 |
pEZ15A-PBAD-Bsals2-Peno-ilvC-ilvD |
| D1 |
pEZ39P-cpf1-gRNA-pdc |
Table 1-2 The strains constructed for the isobutanol biosynthesis
| Strains |
Description |
| ZMQ |
0038::Ptet-dcpf1 |
| ZMQ9 |
0038:Ptet-dcpf1,1650:Pgap-kdcA |
| ZMQ11 |
0038:Ptet-dcpf1.1650:Pgap-kdcA,1547:Pgap-kdcA |
| ZMQ9-A5 |
ZMQ9 containing plasmid A5 |
| ZMQ11-A5 |
ZMQ11 containing plasmid A5 |
| ZMQ11-A6 |
ZMQ11 containing plasmid A6 |
ZMQ11-A7 |
ZMQ11 containing plasmid A7 |
| ZMQ11-A6-D1 ZMQ11 |
ZMQ11 containing plasmid A6 and D1 |
Results 1—Two copies of kdcA integrated in the genome
The dcpf1 gene driven by tetracycline-inducible promoter Ptet was firstly integrated into the genome at the chromosomal locus of ZMO0038, and the recombinant strain ZMQ was used as the parent strain for subsequent isobutanol pathway engineering. With the codon-optimized kdcA gene driven by Pgap integrated into ZMO1650 and ZMO1547, two new recombinant strains ZMQ9 and ZMQ11 were obtained. Then the optimized A5 plasmid was transferred saperately in the two recombinants, resulting in the strain ZMQ9-A5 and ZMQ11-A5.
Fig.1-1 Cell growth and isobutanol production of ZMQ9-A5 and ZMQ11-A5 with different tetracycline concentrations.
Fig.1-1 displayed the cell growth and isobutanol production of ZMQ9-A5 and ZMQ11-A5 with different concentrations of tetracycline. Under lower tetracycline induction, the isobutanol production of ZMQ9-A5 reached a maximum of 0.58 g/L, and the ZMQ11-A5 reached a maximum of 0.93 g/L. Under higher tetracycline induction, the isobutanol production of ZMQ9-A5 reached a maximum of 3.45 g/L. The results indicated that the double-copy kdcA in Z. mobilis was superior to single-copy kdcA. However, the A5 plasmid contained a tetracycline-inducible promoter Ptet, and the dcpf1 was driven by Ptet as well, which may mutually influence each other due to the tetracycline concentrations. Higher concentration of tetracycline was required for the higher isobutanol production.
Results 2—Promoter Replacement
To avoid the mutual influence of two Ptet promoters in ZMQ11-A5, the Ptet promoter in the A5 plasmid was replaced with constitutive promoter Ppdc or PBAD, constructing the plasmid A6 and A7. After transferred into ZMQ11, the recombinant strain ZMQ11-A6 and ZMQ11-A7 were obtained. The isobutanol production of ZMQ11-A6 and ZMQ11-A7 were evaluated with different concentrations of tetracycline and arabinose (Figs.1-2).
Fig.1-2 Cell growth and isobutanol production of ZMQ11-A6 and ZMQ11-A7.
As expected, the isobutanol production in ZMQ11-A6 and ZMQ11-A7 were both signifcantly higher than in the ZMQ11-A5 with low concerntation of inducers induction. The maximum isobutanol titer in ZMQ11-A6 reached 4.12 g/L without tetracycline induction, and it in ZMQ11-A7 reached 2.58 g/L with 5 mg/mL arabinose induction. These results indicated that the isobutanol production increased without tetracycline competition. The isobutanol production in ZMQ11-A6 was higher than in ZMQ11-A7, suggesting that Ppdc promoter was more efficient than the PBAD promoter in isobutanol production. The ZMQ11-A6 with the highest isobutanol production was selected for pdc interference through CRISPRi strategy.
Results 3—dCpf1 assisted CRISPRi system to inhibit the pdc expression
The plasmid D1 containing gRNA targeting pdc was transferred into the recombinant strain ZMQ11-A6, and a new strain ZMQ11-A6-D1 was obtained. The isobutanol production in ZMQ11-A6-D1 was investigated with different tetracycline concentrations, the results were shown in Fig.1-3.
Fig.1-3 Isobutanol production and growth curve of ZMQ11-A6-D1 at different tetracycline concentrations
Two repetitive experiments were performed. In the first experiment, the highest isobutanol production reached 5.7 g/L, and it was 5.58 g/L in the second experiment. The isobutanol production in ZMQ11-A6-D1 was higher than in ZMQ11-A6, which indicated that the pdc was silenced, and more carbon flux was red redirect iverted from ethanol production to isobutanol biosynthesis.
Part Ⅱ CO2 Concentrating and Fixation
Overview
Considering the issue of carbon dioxide emissions in the isobutanol production pathway, our project also designed to introduce the heterologous Calvin cycle pathway coupled with a carboxysome for CO2 concentrating and CO2 fixation, thus, the CO2 released during the isobutanol fermentation can be absorbed and further transformed into isobutanol again, achieved an efficient zero-carbon isobutanol generation. All the plasmids and recombinant strains constructed in this part were shown in Table 2-1 and Table 2-2.
Table 2-1 The plasmids constructed for the CO2 concentrating and fixation
| Plasmids |
Description |
| pEZ-PrkA |
prkA from Synechococcus elongatus PCC 7942 driven by Ptet |
| pEZ-Rbc-PrkA |
prkA from Synechococcus elongatus PCC 7942 driven by Ptet and rbc ( encoding Rubisco) from Thiobacillus denitrificans driven by Peno |
| pEZ-GFP |
gene encoding mNeonGreen driven by Ptet |
| pEZ-Shell |
genes encoding CcmK1, CcmK2, CcmO, CcmL from Synechococcus elongatus PCC 7942, and mNeonGreen driven by Ptet |
| pEZ-Shell-PrkA |
genes encoding CcmK1, CcmK2, CcmO, CcmL from Synechococcus elongatus PCC 7942, and mNeonGreen, CA, PrkA driven by Ptet |
| pEZ39p-Core |
genes encoding CcmM, CcmN from Synechococcus elongatus PCC 7942, and Rubisco driven by Ptet |
Table 2-2 The strains constructed for the CO2 concentrating and fixation
| Strains |
Description |
| 8b-PrkA |
8b derivative containing plasmid pEZ-PrkA |
| 8b-Rbc-PrkA |
8b derivative containing plasmid pEZ-Rbc-PrkA |
| 8b-GFP |
8b derivative containing plasmid pEZ-GFP |
| 8b-Shell |
8b derivative containing plasmid pEZ-Shell |
| 8b-Shell-PrkA |
8b derivative containing plasmid pEZ-Shell-PrkA |
| 8b-Complete |
8b derivative containing plasmid pEZ-Shell-PrkA and pEZ39p-Core |
Results 1—Carboxyl body characterization
Carboxysome shell associated genes ccmK1, ccmK2, ccmO, and ccmL, as well as a fluorescent protein gene mNeonGreen fused to ccmK1 were cloned into the shuttle vector pEZ15Asp generating the plasmid pEZ-Shell. The shell protein CcmK1 fused to mGFP (mNeonGreen enconding protein) was used to evaluate the assembly of carboxysomal proteins in Z. mobilis. After transferred into Z. mobilis strain 8b, the xylose-engineered mutant strain, a new recombinant strain 8b-Shell was constructed. The fluorescence of the cell was detected by flow cytometry with different concentrations of tetracycline induction (0, 0.2, 0.8 µg/mL).
Fig.2-1A displayed that no fluorescence was detected in parental strain as a negative control. Fig.2-1B showed the fluorescence detection in 8b-GFP as a positive control. The fluorescence was high even with no tetracycline induction. Fig.2-1C showed the fluorescence in 8b-Shell expressing complete shell components. The fluorescence was significantly improved with the increase of tetracycline concentrations. The results in Fig.2-1 suggested that the recombinant carboxysome shell plasmid was successfully transformed in Z. mobilis.
Fig.2-1 Fluorescence intensity of 8b, 8b-GFP and 8b-Shell.
Results 2—Intracellular localization of carboxysome
Fig.2-2 Fluorescence image of 8b-Complete
8b-Complete was further constructed to express the complete carboxysome as well as the Calvin cycle pathway with both pEZ-Shell-PrkA and pEZ39p-Core transformed in Z. mobilis. Expression of carboxysome components in 8b-Complete was induced by tetracycline. Through the fluorescence detection by fluorescence microscope, it was confirmed that the carboxysome was successfully assembled in the cytoplasm in Z. mobilis, since the fluorescence intensity in 8b-Complete was discerned as green puncta as shown in Fig.2-2.
Results 3—Carboxysome structure characterization by TEM microscope
Fig.2-3 Thin-section TEM images of intracellular carboxysome in 8b, 8b-Shell and 8b-Complete.
The formation of carboxysome structure was further verified by thin-section transmission electron microscopy (Fig.2-3). As expected, the carboxysome-like structures (∼400 nm in diameter) with a high internal protein density were observed in both recombinant 8b-Shell and 8b-Complete cells but are invisible in the parental 8b cells without carboxysome-expressing vectors.
Fig.2-4 Thin-section TEM images of extracellular carboxysomes isolated from 8b-Shell and 8b-Complete.
The recombinant carboxysome was then purified by sucrose density gradient centrifugation, following the induction of tetracycline. The purified carboxysome structure was verified by thin-section transmission electron microscopy as well. The images in Fig.2-4 showed that both the purified carboxysomes from 8b-Shell and 8b-Complete were 50 nm in diameter. However, they did not exhibit manifestly a regular icosahedral shape and symmetry. All these results suggested that these carboxysome proteins were densely packed in the carboxysome-like structures in Z. mobilis.
Results 4—Fermentation analysis of 8b-Complete
Fig.2-5 Cell growth, sugar utilization, and ethanol production of Z. mobilis recombinant strains.
The capability of carbon fixation by recombinant strain 8b-Complete was evaluated under aerobic and anaerobic conditions, compared with 8b-PrkA and 8b-Rbc-PrkA with the Calvin Cycle pathway inserted but without carboxysome expression. As shown in Fig.2-5, the single expression of rbc and prkA (8b-Rbc-PrkA) significantly inhibited the cell growth of Z. mobilis under aerobic and anaerobic conditions (almost no bacterial growth), while the co-overexpression of carboxysome and Calvin Cycle pathway (8b-Complete) dramatically eliminated the inhibition. The result suggested that the Calvin Cycle pathway, essentially the toxic intermediate in the process of carbon fixation was encaptured in carboxysome. In addition, the ethanol production detection showed that higher ethanol titer was obtained by 8b-Complete compared with the control 8b-GFP, indicating that a certain amount of CO2 may be fixed for ethanol production in 8b-Complete.
Summarize：
So far, we have successfully increased the isobutanol production in Z. mobilis to 5.7 g/L, and we have demonstrated the successful assembly of synthetic carboxysome in the cell, which can be used for carbon dioxide fixation. We would further integrate the two parts above into a single strain, to achieve the zero-carbon isobutanol production.