Construction of the ethanol-producing Synechocystis. To biosynthesize ethanol, the ethanol-producing pathway with acetyl-CoA or pyruvate from the Calvin cycle as the initial material was usually introduced into photosynthetic cyanobacterium Synechocystis 6803 [16,17]. Acetyl-CoA was directly reduced into ethanol via acetaldehyde using bifunctional aldehyde/alcohol dehydrogenase, CoA-dependent acetaldehyde dehydrogenase, and alcohol dehydrogenase, or indirectly converted to ethanol via acetate and aldehyde:ferredoxin oxidoreductase pathway . The production of ethanol through acetyl-CoA was considered to be a less efficient route in cyanobacteria due to the low level of the reductant NADH (electron carrier), which limits the catalytic activities of NADH-dependent alcohol dehydrogenase [19, 20]. The ethanol-producing pathway was further optimized by expressing NADPH-dependent alcohol dehydrogenase, such as E. coliyqhD, which can exploit the reducing equivalent NADPH pool in Synechocysitis 6803 . Compared with NADH-dependent alcohol dehydrogenase, NADPH-dependent aldehyde reductase (as yqhD) has been confirmed to have high catalytic efficiency in the reduction of aldehydes to alcohols [21, 22, 23]. Therefore, in our study, codon-optimized gene coding NADPH-dependent aldehyde reductase from E. coli together with gene coding pyruvate decarboxylase from Z. mobilis was selected for the construction of the ethanol-producing pathway, which pyruvate was used as the initial precursor for this metabolic pathway in engineered cyanobacteria. As shown in Figure. 1, pyruvate decarboxylase catalyzes the conversion of pyruvate to acetaldehyde, which is then reduced to ethanol by NADPH-dependent aldehyde reductase (yqhD) and consumed NADPH. Compared with other NADPH-dependent alcohol dehydrogenases which have been used in previous studies like slr1192 , slr0942 , etc. E. coli's NADPH-dependent aldehyde reductase (yqhD) has been confirmed to have high catalytic efficiency in the reduction of aldehydes to alcohols [21, 22, 23].
The first ethanol producer SynBE01 was constructed by introducing Z. mobilispdc gene and E. coliyqhD gene into Synechocystis 6803, which were co-expressed with the drive of the Copper ion inducible promoter PpetE (Figure.2A) . To obtain cyanobacteria transformants with high copy transformation genes, we used a solid BG11 medium with corresponding antibiotics for repeated screening . The high-copy transformed strains were cultivated on a large scale in plastic bottles filled with BG11 liquid medium without copper ions (Spectinomycin, 10μg/ml). To verify the stable integration of ethanol biosynthetic genes into the neutral slr0168 site of the Synechocystis 6803 genome, the genomic DNA of the transformed strain was extracted and subjected to PCR analysis using primers binding to the coding regions of each transgene in the integrative locus (Figure.2B). PCR amplification products of the expected size (about 3.6kb, 2.9kb, and 2kb) for ethanol biosynthetic genes were present in the transformants but absent in the wild-type, indicating that ethanol biosynthetic genes were successfully integrated into the Synechocystis 6803 genome. These PCR analysis results provided evidence that all alien gene copies were completely segregated in the transformant and all original wild-type DNA copies were eliminated.
To determine whether the transformants expressed the integrated ethanol biosynthetic genes, we analyzed the transcription of each transgene by reverse transcriptase (RT)-PCR. The concentration of 500nM copper ion was added into the transformant Synechocystis SynBE01 cultures to drive the expression of exogenous gene pdc and yqhD. Total RNA was extracted from the transformant or wild-type cultures and digested with RNA free-DNase. The expected size fragments for ethanol biosynthetic genes were amplified from cDNA isolated from the transformant by using the specific primers. However, these fragments were not amplified in the wild-type cDNA (Figure.2C). The 16S gene was used as a positive control and was present in the transformant and wild-type cDNA which means the correctness of RNA extraction. These results indicated that the introduced ethanol biosynthetic genes were all expressed successfully in the mutant SynBE01.
To verify the impact of the introduced ethanol biosynthesis genes on the growth of Synechocystis. During the cultivation, the growth state of SynBE01 was observed and the growth curve within 14 days was drawn (Figure.2D). It was found that SynBE01 grow normally, and its growth was not inhibited by other conditions.
Strong expression of the ethanol-producing genes. To enhance the expression of ethanol-producing genes, SynBE02 was constructed by transformation of Synechocystis 6803 with plasmid pBE02 (Figure.3A), which substituted the promoter PpetE with the super promoter Pcpc560. The promoter Pcpc560 has been reported to make the expression level of functional proteins reach 15% of the total soluble protein in Synechocystis 6803, which is comparable to the protein production in E. coli .
Similar to SynBE01, DNA and RNA of SynBE02 were extracted to verify the construction of mutant strain and the successful expression of exogenous gene pdc and yqhD. The extracted DNA of SynBE02 was also subjected to PCR analysis with primers binding to the coding regions of each transgene in the integrative locus (Figure.3B). PCR products (about 3.9kb, 2.9kb, and 2kb) were amplified successfully with the extracted DNA of SynBE02 as the template. At the same time, when using the extracted DNA of wild-type 6803 as a template to amplify the target product, there are no bands in the gel electrophoresis pattern (Figure.3B). This evidence indicates that SynBE02 successfully integrated the target fragment into the corresponding site. The total RNA of SynBE02 was also extracted to determine the expression level of ethanol-producing genes in SynBE02. The expected fragments for ethanol biosynthetic genes were also amplified successfully with the cDNA which transformed from RNA in SynBE02 as the templet (Figure.3C). Total RNA extracted from SynBE01 and SynBE02 were also used for a quantitative (RT)-PCR analysis. The Table.2 results showed that pdc and yqhD gene expression was increased by 1.27-fold and 1.70-fold in the SynBE02 strain when compared with the SynBE01 strain. These results indicate that the exogenous ethanol-producing genes in SynBE02 were successfully expressed, and compared with SynBE01, the super-strong promoter Pcpc560 effectively increased the expression of ethanol-producing genes.
Under the same culture conditions as SynBE01, SynBE02 was also cultured in a BG11 liquid medium containing 10μg/ml spectinomycin antibiotic. Without optimizing any culture conditions, the growth curve of SynBE02 is shown in Figure.3D. The ethanol producer SynBE01 and SynBE02 were constructed on the expression platform that knocks out the gene slr0168. It has been confirmed that the knockout of slr0168 has no effect on the growth of Synechocystis 6803, and our observations are consistent with this conclusion.
HPLC detection of metabolites. During the observation of SynBE01 and SynBE02, we collected the expressed cyanobacterial samples accurately and prepared for the detection of ethanol production. To measure ethanol production, the SynBE01 medium was collected regularly after adding copper ions to drive expression. SynBE02 samples were regularly collected during the observation of growth status. The ethanol content in the sample was determined by high-performance liquid chromatography (HPLC) (Agilent Technologies 1200 Series) system equipped with a refractive index detector (RID) using Biorad Aminex HPX-87H column (300×7.8 mm). The column was eluted with 0.5mM of H2SO4 at a flow rate of 0.6 mL/min at 50°C . The ethanol production graph shows that on the 14th day when the OD730 reached 1.12, the maximum ethanol production of SynBE01 was 389 mg/L. On the 9th day, when OD730 reached 0.75, the ethanol production of SynBE02 reached 591.7 mg/L. At the highest yield, the dry cell weights of SynBE01 and SynBE02 were 0.3g and 0.315g, respectively. By weighing and corresponding calculations, the maximum ethanol output of SynBE02 is 1.97g/dry cell weight (1g) and the maximum output of SynBE01 is 0.9g/dry cell weight (1g). So, the ethanol production of SynBE02 is much higher than SynBE01. The only difference between the SynBE02 and SynBE01 is the promoter. It has been found that the super-strong promoter can effectively enhance the expression of ethanol-producing genes. Therefore, enhancing the expression of ethanol-producing genes in Synechocystis 6803 can improve ethanol production.
Inhibition of competing phosphoenolpyruvate synthase activity. It has been suggested that the deletion of the related competitive pathways may contribute to increasing the production yield of ethanol in the engineered cyanobacteria. A previous experiment carried out by Wu et al. indicated that disrupting the ADP-glucose pyrophosphorylase gene leads to the accumulation of the poly-β-hydroxybutyrate (PHB) in Synechocystis 6803 through blocking the glycogen biosynthetic pathway. To divert the PHB carbon flux to the ethanol biosynthetic pathway, Gao et al. constructed a Synechocystis 6803 mutant in which the position of genes phaAB coding the enzymes polyhydroxyalkanoate-specific bketothiolase and polyhydroxyalkanoate-specific acetoacetyl-CoA reductase was substituted by the pdc gene from Z. mobilis and endogenous slr1192 gene [10, 26, 27]. No significant increase in the production of ethanol was observed in the Synechocystis 6803 mutant cultures. This result may be the suggestion that the physiological level of the precursor metabolite is regulated by many different metabolic pathways. However, the genes phaAB is used by acetyl-CoA as the precursor metabolite to produce PHB, and thus the disruption of the PHB metabolic pathway may not directly influence the concentration of pyruvate. To investigate if the ethanol productivity is influenced by the related competitive pathways, regulating the concentration of the precursor pyruvate, the Synechocystis SynBE03 strain was constructed by knocking out the gene slr0301.
In Synechosytis, the precursor pyruvate is produced through the EMP pathway. Phosphoenolpyruvate is catalyzed by pyruvate kinase to generate pyruvate and ATP. The gene slr0301 in Synechocystis 6803 also encodes phosphoenolpyruvate synthase, which controls the conversion of pyruvate to phosphoenolpyruvate reversal pathway . To increase the carbon flux to pyruvate, the initial precursor of ethanol production, SynBE03 was generated as the last ethanol producer by knocking out phosphoenolpyruvate competitive pathways. The ethanol-producing genes were incorporated into the site of the slr0301 gene encoding phosphoenolpyruvate synthase and expressed in Synechocystis 6803 under the drive of the promoter Pcpc560. The PCR results of genome DNA extracted indicated that the gene pdc and yqhD were integrated into Synechocystis 6803 genome, suggesting the successful construction of the engineered cyanobacterium SynBE03 lacking phosphoenolpyruvate competitive pathways (Figure.4A-B). Furthermore, the RT-PCR results indicated that the exogenous gene pdc and yqhD in SynBE03 were successfully expressed, as shown in Figure.4C. For quantitative RT-PCR analysis, the total RNA of SynBE03 was extracted and converted to cDNA. As shown in Table2, in multiple parallel experiments, the average △CT values of pdc and yqhD in SynBE03 were -0.93, -0.96. The relative abundance of different mRNA molecules was estimated using 2-△△CT. The difference of the RNA abundance of the pdc gene between strain SynBE03 and SynBE02 is 2^-(-0.93-(-0.25)). The difference of the RNA abundance of the yqhD gene between strain SynBE03 and SynBE02 is 2^-(-0.96-0.22). The data demonstrated that pdc and yqhD gene expression was increased by 1.6-fold and 2.26-fold in the SynBE03 strain when compared with the SynBE02 strain.
In terms of ethanol production, in our experiment, we found that the modified strain SynBE03 which knocked out slr0301 improved significantly when compared with SynBE02. SynBE03 reached the highest output of 878mg/L on the 9th day, its OD730 was 0.81. We weighed the dry weight of the algae cells (DCW) at the highest yield, the dry cell weight of SynBE03 was 0.43g. By weighing and corresponding calculations, the maximum ethanol output of SynBE03 is 2.79g/dry cell weight (1g). The output of SynBE02 is much lower than SynBE03. SynBE03 has the highest ethanol production which is consistent with the analysis of RT-qPCR results. The knockout of the gene slr0301 promotes the expression of exogenous ethanol-producing genes and effectively increases the yield of ethanol synthesis. Knocking out the competitive phosphoenolpyruvate synthesis pathway can effectively increase the expression level of ethanol-producing genes and ethanol production in Synechocystis 6803.
To verify the impact of slr0301 disruption on cell growth, we transferred the plasmid pBE409 into Synechocystis 6803 and observed the growth status. During the period of cultivation under constant light condition, we observed that the mutant strain lacking the gene slr0301 showed no obvious difference in the cell growth compared with wild-type Synechocystis 6803 (Figure.4D, Figure.5). Therefore, the knockout of the slr0301 gene had no obvious effect on the growth of Synechocystis 6803.
In our experiment, as shown in Figure.6, our highest output is 878mg/L/9days. In terms of output, our results are still far from Gao and Rajendran Velmurugan [10, 16]. However, the cyanobacteria were cultivated with a column photo-bioreactor and optimized for cultivation in their experiment. In their experiment, they still need a relatively long cultivation period and a high concentration of cultivation to achieve high yield. For example, as shown in Table.3, in Gao’s experiment, the ethanol production reached 5.50gL-1 with the OD730 reached 12. That means its average output just reached 0.458g/OD730. Our experiment provides a modified strain that can achieve relatively high yields through simple cultivation and has the advantages of simple production and short production cycle in industrial production. It has great application prospects in industrial production. In our experiment, we also found that ethanol production increased significantly by knocking out the competitive phosphoenolpyruvate synthesis pathway and enhancing the expression of ethanol-producing genes in Synechocystis 6803. It will serve as the basis for optimizing ethanol synthesis in the future.
Future study directions. All the cyanobacteria described in this article are cultivated in plastic air bottles. Ethanol is easy to volatilize. Under normal culture conditions, ethanol volatilization will cause experimental deviation. So, we plan to design a bioreactor to cultivate cyanobacteria. Blowing a certain amount of CO2 into the cyanobacteria will effectively promote the growth of cyanobacteria and increase the metabolic yield of ethanol. Under normal conditions, our cyanobacteria were only cultivated to OD730 =1.3. Compared with other experiments, we need to further optimize the experimental conditions to optimize the growth status of cyanobacteria.
As documented in previous studies, Synechocystis can grow in the maximum concentration of 10.6g/L ethanol without significant impacts on cells [16, 27]. Nowadays, the maximum yield of ethanol has not yet been obtained in engineered cyanobacteria due to the physiological impacts of ethanol tolerance and metabolic pathways on the cyanobacterial cells. Thus, to further improve ethanol production, the next work will be performed to increase the expression of ethanol tolerance genes, such as slr0982 , which may contribute to reducing the effect of high-yield ethanol on the cyanobacteria cell growth. Another alternative approach is to modify the ethanol-producing metabolic pathways in Synechocystis 6803 through introducing exogenous genes such as maeB and Ycf21, which may increase the carbon flux to the initial precursor pyruvate, or deleting other related competitive pathways such as the lactic acid metabolism pathway. Also, enhancing the expression of ethanol-producing genes will become the focus of the future. Transferring the plasmid pBE02 into the modified strain SynBE03 potentially double the expression of ethanol-producing genes.