Construction of the recombinant strains
The expression vector pCasKP-apr plasmid was transformed into K. pneumoniae by electrotransformation, and the resistant monoclonal KP-pCasKP-apr can be obtained on the plate. The plasmid pCasKP-apr was used as a positive control, the original strain of Klebsiella pneumoniae was used as a negative control, and pCaskp-ldhA-F(5’-GAAATGCGGCTGGTGCG-3’) and pCaskp-ldhA-R(5’-CGGTTTGGTTAGCGAGAAGAG-3’) were used as primers to perform a positive single clone colony PCR. Fig. S6 shows that the sample and the positive control band are consistent, indicating that the plasmid has been successfully transferred into the host cell K. pneumoniae. Fig. S7 shows the upstream and downstream sequences of ldhA, adhE and ack amplified by PCR. Then the upstream and downstream sequences of ldhA, adhE and ack were annealed by overlap PCR to synthesize the dsDNA repair templates (Fig. S8). Fig. S9 shows the constructed pSGKP-km-sgRNA plasmid of each target gene. They were identified by PCR with the corresponding Spacer-F and M13R as primers, and bright bands appeared at 200–300 bp, which were consistent with expectations.
The constructed pSGKP-km-sgRNA plasmid and the dsDNA repair template were co-electroporated into competent cells carrying the plasmid pCasKP-apr, and the electro-transformed host was cultured on the screening double-resistance LB agar plate to obtain resistant individual colonies (Fig. 1b-d). After curing the pCasKP-apr and pSGKP-km plasmids, The 1 kb fragment was amplified from the chromosome DNA of the ldhA(−), while 2 kb DNA fragment was obtained from K. pneumoniae KP (Fig. S10), indicating the ldhA gene has been successfully deleted from K. pneumoniae KP. Similarly, the chromosomal DNA from ldhA(−)-adhE(−) and ldhA(−)-ack(−) were also amplified with a 1 kb fragment, which are in contrast with about 2 kb DNA fragment obtained from Klebsiella pneumoniae KP (Fig. S11). This indicates that the adhE and ack genes have been successfully knocked out from ldhA(−), and means the double gene deletion strain of Klebsiella pneumoniae was constructed, which were named ldhA(−)-adhE(−) and ldhA(−)-ack(−), respectively. Through this CRISPR-Cas9 double-plasmid system gene editing method, Klebsiella pneumoniae bacteria with all the three genes of ldhA, adhE and ack deleted were also successfully constructed (Fig. 1e, Fig. S11), and it was named ldhA(−)-adhE(−)-ack(−).
Effects of ldhA, adhE and ack inactivation on metabolism of K. pneumoniae
To investigate the effects of ldhA, adhE and ack inactivation on metabolism of K. pneumoniae, batch cultures of the wild type strain and recombinant strains were performed in shake flasks with the initial glycerol concentration at 20 g L− 1. The cell growth, glycerol consumption and production of main and by-products of metabolism were monitored.
As shown in Fig. 2 and Table 3, the biomass of the recombinant strain ldhA(−) was significantly higher than that of the original strain of Klebsiella pneumoniae KP, reaching the maximum at 12h, with an OD600 of 6.96. The original bacteria also reached the maximum at 12h, with an OD600 of 5.06. The biomass of ldhA(−) increased by 37.5% compared with the original Klebsiella pneumoniae KP. In addition, the delay period of ldhA(−) was very short, indicating that it can quickly adapt to the new growth environment. The metabolism of the recombinant strain ldhA(−) has also undergone significant changes. The genetically engineered strain no longer produces lactic acid, and the production of 1,3-propanediol has also been increased. Compared with the original strain, the 1,3-PD concentration and conversion rate increased by 56.7% and 56.5%, respectively. At the same time, the concentration of 2,3-butanediol and ethanol also increased slightly. This may be due to reduced competition for intracellular NADHs by by-products and improvement in all NADH-dependent metabolic pathways.
Table 3
Comparison of final metabolites concentration of recombinant ldhA(−) with those of wild type K. pneumoniae KP in batch fermentations
Strains | Products concentration (g L− 1) | 1,3-PD conversion (mol mol− 1) |
1,3-PD | Lac | 2,3-BD |
K.pneumoniae KP | 3.88 | 0.08 | 4.32 | 0.23 |
ldhA(−) | 6.08 | 0 | 3.52 | 0.36 |
Figure 3 and Table 4 show the comparison of double and triple gene deletions recombinant strains relative to the original strains of Klebsiella pneumoniae KP by shaker fermentation. The biomass and growth trend of the recombinant strain ldhA(−)-ack(−) is higher than that of Klebsiella pneumoniae KP, and it’s biomass increased by 7.07% compared with that of the original bacteria KP. At the same time, the original bacteria showed a downward trend after 12 hours, but the recombinant bacteria continued to grow, indicating that after both the lactic acid and acetic acid metabolic pathways were blocked, cell growth was not harmed. However, the highest biomass of ldhA(−)-adhE(−) was 13.52% lower than that of the original bacteria, indicating that the simultaneous deletion of ldhA and adhE had a certain negative effect on cell growth. Moreover, the increase of the byproduct acetic acid yield may be one of the factors affecting the growth of the bacteria. Compared with the original strain, the biomass of the recombinant strain ldhA(−)-adhE(−)-ack(−) had no significant change. The results showed that after deleting the three genes using the CRISPR-Cas9 double plasmid system, the cells were still not lethal or significantly inhibited.
Table 4
Comparison of final metabolites concentration of recombinant ldhA(−)-adhE(−), ldhA(−)-ack(−) and ldhA(−)-adhE(−)-ack(−) with those of wild type K. pneumoniae KP in batch fermentations
Strains | Products concentration (g L− 1) | 1,3-PD conversion (mol mol− 1) |
1,3-PD | Ace | Eth | 2,3-BD |
K.pneumoniae KP | 6.04 | 0.22 | 0.41 | 1.68 | 0.37 |
ldhA(−)-adhE(−) | 7.75 | 0.37 | 0.34 | 2.81 | 0.47 |
ldhA(−)-ack (−) | 6.91 | 0.19 | 0.38 | 2.08 | 0.42 |
ldhA(−)-adhE(−)-ack (−) | 8.29 | 0.27 | 0.37 | 2.63 | 0.50 |
In addition, compared with the original strain, the concentration of 1,3-PD, the conversion rate, and the concentration of 2,3-BD of the recombinant strain ldhA(−)-ack(−) increased by 14.4%, 13.51% and 23.81%. In addition, it no longer produced acetic acid in the first 8h of the fermentation process on the basis of producing no lactic acid. And the concentration of ethanol was also 35.67% lower than that of the original strain K. pneumoniae KP. At the same time, compared with the original strain K. pneumoniae KP, the byproduct ethanol production of ldhA(−)-adhE(−), decreased significantly, by 62.23%. Furthermore, inactivation of the key enzyme for ethanol formation lead to the increased glycerol flux to 1,3-PD[17]. The concentration of 1,3-PD (7.75 g/L) and its conversion (0.47 mol/mol) were increased by 28.39% and 27.03%, respectively. However, the output of the by-product acetic acid has increased significantly, which inhibited the growth of the bacteria. It may need to be solved by improving the acid resistance of the strain. For the recombinant strain ldhA(−)-adhE(−)-ack(−), on the basis of not producing lactic acid, compared with the original strain K. pneumoniae KP, the output of ethanol was greatly reduced by 78.5%. But the acetic acid content was 22.73% higher than that of the original strain. According to the observation in Fig. 3, it is speculated that the cells may have converted to consume 2,3-BD as a substrate and produce acetic acid through other metabolic pathways. The 1,3-PD concentration and conversion rate of recombinant ldhA(−)-adhE(−)-ack(−) increased by 37.31% and 37.53%, respectively, compared with KP. And the concentration of 2,3-BD, compared with K. pneumoniae KP, increased by 56.97%. In summary, the above results show that blocking the two or three by-product synthesis pathways at the same time facilitates the flow of reduced NADH to the main product synthesis pathway, effectively promotes the generation and accumulation of the main product, and also simplifies the downstream separation of the main product.
Fed-batch fermentation and metabolic flux redistribution
As results described above, the recombinant ldhA(−) and ldhA(−)-adhE(−)-ack(−) both exhibited great ability of cell growth and 1,3-PD production. The further investigation of metabolic flux redistribution in fed-batch fermentation was performed in 1 L fermentor using wild type K. pneumoniae KP as the control. Concentration of major metabolites (1,3-PD, 2,3-butanediol, lactate, ethanol and acetate) were determined. The fermentation results of ldhA(−) were shown in Fig. 4 and Table 5. And the fermentation results of ldhA(−)-adhE(−)-ack(−) were shown in Fig. 5 and Table 6.
Table 5
Comparison of final metabolites concentration of recombinant ldhA(−) with those of wild type K. pneumoniae KP in fed-batch fermentations
Strains | Products concentration (g L− 1) | Glycerin consumption (g) | 1,3-PD conversion (mol mol− 1) |
1,3-PD | Ace | Lac | 2,3-BD |
K.pneumoniae KP | 36.62 | 1.72 | 0.41 | 4.10 | 103.3 | 0.34 |
ldhA(−) | 40.52 | 1.76 | 0 | 7.45 | 102.57 | 0.38 |
Table 6
Comparison of final metabolites concentration of recombinant ldhA(−)-adhE(−)-ack(−) with those of wild type K. pneumoniae KP in fed-batch fermentations
Strains | | Products concentration (g L− 1) | Glycerin consumption (g) | 1,3-PD conversion (mol mol− 1) |
1,3-PD | Ace | Lac | Eth | 2,3-BD |
K.pneumoniae KP | 36.39 | 0.82 | 0.40 | 0.19 | 8.61 | 95.55 | 0.36 |
ldhA(−)-adhE(−)-ack (−) | 30.71 | 1.20 | 0 | 0 | 8.48 | 66.17 | 0.43 |
It can be seen from Fig. 4 and Table 5 that the highest biomass OD600 (10.74) of the recombinant strain ldhA(−) was 7.82% lower than the OD600 (11.58) of the original strain. However, the growth rate of recombinant bacteria was increased, indicating that it had good growth performance, and the deletion of the ldhA gene did not cause significant inhibition of the cells. With 24 h as the end point of fermentation, the recombinant bacteria consumed 102.57 g glycerol and produced 40.52 g/L 1,3-PD and 7.45 g/L 2,3-BD. Compared with the original strain K. pneumoniae, the concentration and conversion rate of 1,3-PD and the concentration of 2,3-BD were increased by 10.65%, 11.77% and 81.71%, respectively. In addition, the recombinant bacteria ldhA(−) no longer produced by-product lactic acid, and the production level of acetic acid was basically the same as that of the original bacteria. The above results indicate that blocking the lactic acid synthesis pathway effectively reduces the competition for reducing power, so that more NADH flows to the synthesis pathways of 1,3-PD and 2,3-BD.
It can be seen from Fig. 5 and Table 6 that the biomass OD600 (7.79) of the recombinant strain ldhA(−)-adhE(−)-ack(−) was lower than the OD600 (11.56) of the original strain K. pneumoniae KP. The maximum biomass of recombinant bacteria was even nearly 4 OD less than that of wild-type bacteria. This indicated that the growth of the bacteria was significantly inhibited after the deletion of the three genes. The above results may be caused by the following reasons: First, the increase of acetic acid production inhibited cell growth; Secondly, the deletion of multiple genes may increase the lethality of the strain. The third is that the massive reduction of by-products leads to the disturbance of the reducing power of NADH and NAD + which maintains the redox balance in the cell. Compared with the original strain K. pneumoniae, the recombinant strain ldhA(−)-adhE(−)-ack(−) no longer produceed by-products of lactic acid and ethanol, and glycerol fluxed to the main products 1,3-PD and 2,3-BD increased accordingly. Since the growth of ldhA(−)-adhE(−)-ack(−) was inhibited by the increase of acetic acid concentration, the decrease in biomass reduced the production of 1,3-PD by 18.5%. However, due to the redistribution of metabolic flux, the conversion rate of 1,3-PD (0.43 mol/mol) was somewhat higher than that of the original strain (0.36 mol/mol), which increased by 19.44%. The results show that ldhA(−)-adhE(−)-ack(−) with few by-products and high 1,3-propanediol conversion rate has a certain potential for large-scale production of major products.