Overview

CAU-China aimed to utilize nodule's own anaerobic environment to create an anaerobic biofactory, providing new ideas for solving the problem faced by anaerobic or hypoxic production. The purpose of this page is to present the results of four modules to provide experimental evidence for the validation of the ideas.

PHB Deletion

In order to provide sufficient raw material for the synthesis of the target products, we attempted to delete phaC2, a gene encoding an important enzyme in the PHB synthesis pathway, thereby increasing the amount of raw material for the synthesis of the target products.

Plasmid construction and validation

Firstly, the upstream and downstream about 500 bp respectively of phaC2 gene were used as UP arm(BBa_K5300029) and Down arm(BBa_K5300030) to construct ‘Up-Arm phaC2 + Down-Arm phaC2’(BBa_K5300040) fragment, and Gibson was used to introduce the composite fragment into vector pJQ200SK linearized by SmaI monoenzyme cleavage, and the recombinant plasmid was introduced into DH5α using the method of transformation, and coated onto gentamicin-resistant LB solid medium, incubated at 37°C overnight and then subjected to colony PCR using the universal primer M13F/R (Figure 1-1-1).

Colony PCR proved that the recombinant plasmid was successfully introduced, and the correct colonies were inoculated into gentamicin-resistant LB liquid medium under shaking condition overnight, and the plasmids were extracted and sent for sequencing on the second day, whose results proved that the plasmids were constructed correctly and successfully.

Triparental mating and validation

The verified correct colonies were inoculated into gentamicin-resistant LB liquid medium under shaking conditions and cultured, triparental mating was performed using Helper, Sinorhizobium fredii CCBAU45436, and two screenings were performed. The single colonies obtained from the second screening were subjected to colony PCR (Figure 1-2-1).

Colony PCR demonstrated that the UA and DA portions of the plasmid underwent homologous recombination with Sinorhizobium fredii CCBAU45436, and the phaC2 gene was successfully deleted. Notably, the ∆phaC2 mutants showed different phenotype from the wild type. The colonies were seen to be distinctly yellow when the bacteria were re-picked and streaked on TY solid medium (Figure 1-2-2).

Plant system validation

The deletion of phaC2 resulted in obvious phenotype changes in Sinorhizobium fredii CCBAU45436, and we subsequently inoculated the mutant on Jidou 17 to test whether its original nitrogen-fixing function had been altered. After 15 days of incubation, photos were taken to record the above-ground phenotype, the below-ground phenotype (Figure 1-3-1), and the rhizomatous section (Figure 1-3-2).

There were no significant differences between wild-type and mutant-inoculated plants on phenotype; while mutant strains showed darker color for rhizomatous sections. Their chlorophyll content and dry weight of above-ground parts were examined (Figure 1-3-3).

Analyzing the data of chlorophyll content and dry weight of above-ground parts after 15-day growth, none of the three groups, blank control, wild type and mutant, showed significant differences, indicating that it was not possible to prove whether the differences occurred or not under 15 days of growth. Therefore, a second round of plant experiment was carried out and verified after 21 days of culture after inoculation (Figure 1-3-4).

Chlorophyll content and above-ground part dry weight data from the second round of plant experiments demonstrated that the ∆phaC2 mutant does not affect the growth of the above-ground part of Jidou 17.

Electron microscope scan

To further observe the morphology of the mutant strain, we chose the bacterial solution of the platform stage for transmission electron microscopy observation and counted the number of its PHB particles, whose results are shown in Figure 1-4-1 and Figure 1-4-2.

The number of PHB particles in the wild-type strain and the mutant strain showed a significant difference, further validating the success of the phaC2 deletion module.

Total Lipid TLC Analysis

In order to analyze the changes in fatty acid content in ∆phaC2 strains, we used thin-layer chromatography to analyze the amount of triacylglycerol (TAG) in ∆phaC2 strains as a proxy for fatty acid content (Figure 1-5-1).

∆phaC2 strains were able to show a significant increase in the content of TAG compared to the wild type. We believed that this is due to the altered fatty acid content. This is in line with our design. The deletion of phaC2 gene leads to an increase in the substrates for fatty acid synthesis, which in turn leads to an increase in fatty acid content.

Regulation module

In order to ensure that the synthesis of the target product is accomplished without affecting the nitrogen-fixing function of the rhizobium itself, we designed the regulation module to enable the induction of the synthesis of the target product under dual conditions so as to utilize the rhizobial system with the highest efficiency.

Promoter validation

First, we validated the glnK promoter (BBa_K5300017) alone and explored its expression under different nitrogen concentrations. We constructed a “promoter + fluorescent protein” system in pBBR1MCS-2 (Figure 2-1-1) to reflect the ability of the promoter based on the expression of the fluorescent protein. Since glnK is regulated by nitrogen in both rhizobium and E. coli and the growth cycle of rhizobium is longer, we chose to validate the system in E. coli. The successfully constructed plasmid was introduced into E. coli (DH5α) and colony PCR was performed for verification (Figure 2-1-2).

Subsequently, we explored the induction conditions by using M9 medium without nitrogen source as culture conditions and adding different concentrations of ammonium chloride as an additional nitrogen source. The OD600 value of the bacterial solution was determined after overnight shock incubation, and the fluorescence intensity of the bacterial solution was measured using a fluorescence spectrophotometer (Figure 2-1-3).

Obviously, there was a significant difference between the fluorescence expression intensity under nitrogen-free conditions and that in the presence of a rich nitrogen source, successfully verifying that the glnK promoter was repressed by high nitrogen conditions.

In order to verify the function of nifH (BBa_K5300016) promoter, we constructed a composite circuit of nifH promoter and gfp (BBa_K5300015), which was seamlessly cloned into pBBR1MCS-2 (BBa_K5300023) along with “glnK promoter+sgRNA-GFP (BBa_K5300025) for subsequent experiments. (Figure 2-1-4).

Since Cas12k (BBa_K5300021) was not expressed and the glnK promoter was repressed by high-nitrogen conditions, we cultured rhizobia with this plasmid in high-nitrogen (1 g/L NH4Cl) M9 medium. Under these conditions it was thought that the glnK promoter would be unable to express subsequent genes and that gfp expression would not be repressed. The fluorescence intensity was detected after 24 h of incubation under both low and high oxygen conditions, respectively (Figure 2-1-5).

Apparently, the expression intensity of nifH promoter was up-regulated under low-oxygen conditions; while under high-oxygen conditions, the expression intensity of nifH promoter was down-regulated. It was successfully verified that the nifH promoter was repressed by hyperoxic conditions.

Plasmid construction and validation

We first amplified the glnK promoter and nifH promoter from the genome of Sinorhizobium fredii CCBAU45436, amplified gfp and designed sgRNAs from pRJPaph-bjGFP plasmid, and ligated them using overlap PCR. To verify the effect of the regulation module alone, we connected it with the nuoA (BBa_K5300013) promoter-Cas12k (BBa_K5300021) and the linearized pBBR1MCS-2 (Figure 2-2-1). The recombinant plasmid was transformed into E. coli (DH5α).

Triparental mating and validation

The verified correct colonies were inoculated into kanamycin-resistant LB liquid medium under shaking condition. Then we conducted triparental mating using the correct colony we obtained above, Helper and Sinorhizobium fredii CCBAU45436. Then colony PCR was performed on colonies grown on TY/NA/Kan solid medium incubated at 28°C for 24 h using the universal primer M13F/R.

Fluorescence intensity verification

We utilized its fluorescence expression intensity to verify whether the regulation module functioned properly. We set up four sets of experiment with high nitrogen and high oxygen, low nitrogen and high oxygen, high nitrogen and low oxygen, as well as low nitrogen and low oxygen for verification. Subsequently, we used M9 medium without nitrogen source,then added different concentrations of ammonium chloride to form nitrogen concentration gradient. We added petroleum jelly after boiling the medium to isolate medium against atmosphere, creating anaerobic conditions. The OD600 value of the bacterial solution was determined after a period of shaking and incubation, and the fluorescence intensity of the bacterial solution was determined using a fluorescence spectrophotometer (Figure 2-4-1 and Figure 2-4-2).

We were able to clearly see that high oxygen conditions always inhibit the expression of fluorescent proteins, whether under high or low nitrogen conditions. Low nitrogen conditions also inhibit the expression of fluorescent proteins, whether under high or low oxygen conditions. This is consistent with our assumptions. Under high oxygen conditions, the nifH promoter is repressed, which reduces the expression of downstream genes, thereby the fluorescence intensity is reduced. Under low nitrogen conditions, the glnK promoter is promoted, and the expression of sgRNA is up-regulated, directing the Cas12k protein to repress the expression of fluorescent proteins. However, both low-nitrogen and high-oxygen conditions induced high expression, which is hypothesized to be the result of uninhibited constitutive expression of the promoter, as well as the possibility of leaky expression.

Synthesis module

In order to verify the validity of the “nodule factory”, we chose to utilize it for the production of dipeptidyl aldehyde and PUFA (namely polyunsaturated fatty acids).

Dipeptide aldehyde

Plant system validation

In order to verify that our idea works properly, we first conducted the synthesis of dipeptide aldehydes. We constructed a composite circuit of nifH promoter and dipeptide aldehyde synthesis gene cluster (Figure 3-1-1). To verify whether the plasmid would stress the growth of nodules, we constructed a composite circuit of the nifH promoter and half of the sequence of the dipeptidyl aldehyde synthesis gene cluster (Fig. 3-1-2), which encodes a product that cannot synthesize complete dipeptidyl aldehyde.

The plasmid for dipeptide aldehyde synthesis in E. coli was transferred into Sinorhizobium fredii CCBAU45436 using triparental mating. The verified correct strain was inoculated into Jidou 17 and harvested after 21 days of growth, and photographs were taken to record the above-ground phenotype, below-ground phenotype (Fig. 3-1-3), and nodules section (Fig. 3-1-4).

For phenotypes and nodule sections, there were no significant differences between plants inoculated with wild type and strains containing dipeptide aldehyde synthesis plasmid. Their chlorophyll content in leaves and dry weight of above-ground parts were examined (Figure 3-1-5).

For chlorophyll content of plants after 21 days of growth, there was a significant difference between plants inoculated with rhizobium and the blank control. There was no significant difference between the different strains. For dry weight of above-ground portions, inoculation with recombinant-plasmid-containing bacteria led to a decrease in dry weight, whereas whether or not dipeptidyl aldehydes were synthesized did not affect dry weight of above-ground portions.

Real-time fluorescence quantitative PCR

We used RT-qPCR to quantify gene transcription levels. We extracted RNA from +bgc33-sfp strain and wild-type strain grown for 21 days respectively and conducted reverse transcription. qPCR was performed using cDNA as a template. 16S rRNA was used as an internal reference to compare the expression of the target gene in the +bgc33-sfp (BBa_K5300032) strain with the endogenous nifH gene. We designed three pairs of primers (bgc33-a, bgc33-b, and bgc33-c) for bgc33 (BBa_K5300031) gene to characterize the transcript levels of different regions of the gene in terms of the transcript levels of three fragments, and performed three biological replicates (Figure 3-2-1).

Obviously, we found that the expression of bgc33 gene decreased following the increase of its distance from the promoter. The transcript levels of the first two segments of bgc33 gene were up-regulated compared to the expression of the endogenous nifH gene, while the expression of the third segment was down-regulated. We hypothesized that the decreased transcript levels of the second half of the gene may be due to the fact that the gene is long and the mRNA is susceptible to degradation or forms secondary structure during transcription, which leads to post-transcriptional instability in complete translation of the gene.

Mass Spectrometry Results

To verify whether the gene cluster was phenotypically successful in synthesizing dipeptide aldehydes, we used direct mass spectrometry to analyze the mass-to-charge ratio of dipeptide aldehydes (Figure 3-3-1).

We performed direct mass spectrometry on the wild-type strain and the +bgc33-sfp strain and analyzed the +bgc33-sfp strain according to a gradient of 100-fold dilution, 10-fold dilution of the solution and the stock solution. We did not find new peaks in the diluted liquid, but surprisingly, when we used the stock solution, we detected new peaks not found in the wild type. Although the amount was extremely low, we finally found the new product. Based on the results of real-time fluorescence quantitative PCR, we believed that the amount of products is greatly reduced under the influence of low transcription levels, and therefore we could only detect peaks with a very small area.

PUFA

Plasmid construction and validation

We first seamlessly cloned the regulation module and each two fragments of PUFA1-5 with the PCR linearized pUC19 vector in order to amplify three large fragments (i.e., regulation-PUFA1, PUFA2-3, and PUFA4-5) using the recombinant plasmid as a template. The three large fragments were seamlessly cloned with the PCR linearized pBBR1MCS-2 cloning, and the recombinant plasmids were transformed into E. coli (DH5α) (Figure 3-4-1).

The sequencing results of this plasmid were disappointing, as it showed several point mutations and even nonsense mutations, making our experiment impossible. We were unable to repeat the experiment again due to time constraints. And in order to validate our synthesis module, we therefore designed a synthesis circuit that utilizes a constitutive promoter for expression. We repaired the point mutation in the pfa biosynthetic gene cluster and seamlessly cloned and ligated it to the nuoA promoter to pBBR1MCS-2. (Figure 3-4-2)

Triparental mating and validation

The verified correct colonies were inoculated into gentamicin-resistant LB liquid medium under shaking conditions and cultured, triparental mating was performed using Helper, Sinorhizobium fredii CCBAU45436, and two screenings were performed. The single colonies obtained from the second screening were subjected to colony PCR utilizing the universal primer M13F/R.

RT-qPCR results

We used RT-qPCR to quantify gene transcription levels. We extracted RNA from the bacterial solution of the +pfa strain and the wild-type strain, respectively and conducted reverse transcription. qPCR was performed using cDNA as a template. 16S rRNA was used as an internal reference to compare the expression of the target genes in the +pfa strain relative to the endogenous nuoA gene. We designed three pairs of primers (pfa1, pfa2, pfa3) of the three core enzyme genes of the pfa gene to characterize the amount of transcription in different segments of the gene in terms of the transcript levels of the three fragments and performed three biological replicates (Figure 3-6-1).

Obviously, we found that the expression of the pfa gene decreased following the increase of its distance from the promoter. The transcript levels of all three segments of pfa gene were up-regulated compared to the expression of the endogenous nuoA gene, but the overall trend was a gradual decrease. We hypothesized that the decrease in transcript levels in the second half of the gene may be due to the fact that the gene is long and the mRNA is prone to degradation or forms secondary structure during transcription, which leads to post-transcriptional instability in complete translation.

Total lipid TLC analysis

In order to pre-analyze the changes in fatty acid content in the +pfa strain, we analyzed the amount of triacylglycerol (TAG) in the +pfa strain using thin-layer chromatography to represent the fatty acid content (Figure 3-7-1).

Gas Chromatography Results

In order to directly verify that the gene cluster was actually successfully expressed and synthesized the target product, we used gas chromatography to isolate, purify and characterize the fatty acid content above ten carbons. We sent the wild-type strain and the +pfa strain for gas chromatographic identification (Figure 3-8-1).

It’s a pity that we did not identify the presence of DHA and EPA. But with the bad news came good news. We identified a fatty acid in the +pfa strain that was not present in the wild-type strain - a long-chain saturated fatty acid with fourteen carbons. And the content of this fatty acid varied dramatically, from nonexistent to 5% of the total. We read the references and found that the expression of gene cluster is very complex. pfa biosynthetic gene clusters are capable of expressing several enzymes, each plays a different role (Figure 3-8-2).

Since new fatty acids that never appeared in the wild type appeared in our +pfa strain, it indicates that the expression of the gene cluster was successful to some extent. Based on the results of real-time fluorescence quantitative PCR, we analyzed that the failure to produce the target product was due to the greatly reduced transcription of the second half of the cluster. In Figure 3-8-2, we found that the oxidase was encoded only by pfa3 (BBa_K5300009), which is capable of forming unsaturated bonds. Also, the gas chromatography we used for time reasons was food grade and less accurate. Therefore there is a high probability that pfa3 failed to express the oxidase in large quantities thus DHA and EPA were not detected with the low precision of the assay.

To verify our speculation, we organized the data and reanalyzed it. pfa1 (BBa_K5300004) expresses the product of ER, the reductase, which is able to change unsaturated bonds to saturated bonds. We therefore counted the proportion of unsaturated fatty acids among the eighteen-carbon fatty acids and the changes in the content of different types of fatty acids (Fig. 3-8-3).

Obviously, we found that the content of unsaturated fatty acids in the +pfa strain decreased significantly compared to the wild type strain, which is in line with our speculation. Therefore, it can be proved to a certain extent that our synthesis module is successful and the related genes are able to be expressed normally. And in order to obtain the final target product as well as to increase the yield, it is necessary to optimize the subsequent experiments.

Suicide circuit

Plasmid construction and validation

We first amplified the glnK promoter, nifH promoter, vapC (BBa_K5300019) and vapB (BBa_K5300018) from the genome, amplified mCherry (BBa_K5300014) from the pRJPaph-mChe plasmid and designed the sgRNAs. We screened the loci from the genome, and amplified UP-feo-suicide (BBa_K5300027) and DOWN-feo-suicide (BBa_K5300028) by PCR. Then we performed seamless cloning to link them to pJQ200SK (BBa_K5300034) (Figure 4-1-1).

The recombinant plasmid was transferred into E. coli, and the correct recombinant plasmid was verified by colony PCR.

Triparental mating and validation

The verified correct colonies were inoculated into gentamicin-resistant LB liquid medium under shaking conditions and cultured, triparental mating was performed using Helper, Sinorhizobium fredii CCBAU45436, and two screenings were performed. The single colonies obtained from the second screening were subjected to colony PCR utilizing the universal primer M13F/R.

Colony PCR demonstrated that the plasmid successfully underwent homologous recombination with Sinorhizobium fredii CCBAU45436, and the suicide circuit was successfully introduced into chassis. It is noteworthy that the module showed a different phenotype from the wild type when cultured for more than 3 d after introduction, i.e., the colonies became transparent teardrop-shaped.

Fluorescence intensity verification

We utilized its fluorescence expression intensity to verify whether the suicide module functioned properly. We set up four sets of experiment with high nitrogen and high oxygen, low nitrogen and high oxygen, high nitrogen and low oxygen, as well as low nitrogen and low oxygen for verification. Subsequently, we used M9 medium without nitrogen source,then added different concentrations of ammonium chloride to form nitrogen concentration gradient. We added petroleum jelly after boiling the medium to isolate medium against atmosphere, creating anaerobic conditions. The OD600 value of the bacterial solution was determined after a period of shaking and incubation, and the fluorescence intensity of the bacterial solution was determined using a fluorescence spectrophotometer (Figure 4-3-1).

In a low oxygen environment, different nitrogen concentrations significantly affect the expression intensity of fluorescent proteins. This is consistent with our design. When changed to high oxygen environment, the change of nitrogen concentration could not significantly affect the fluorescent intensity, which may be caused by the environmental instability due to the expression of toxin. At the same time, we can also find that the fluorescent intensity of the bacterial solution cultured in high oxygen environment was higher than that in low oxygen environment, which may be related to the concentration of the bacterial solution fluctuates more in high oxygen environment by the change of toxin.

Change in OD600 value of bacterial solution

In order to further verify whether the circuit leads to the death of the bacteria, we selected the strain with suicide circuit plasmid with OD600 value of 1, added different concentrations of NH4Cl, and investigated the changes of OD600 values after same interval under hyperoxic conditions (Figure 4-4-1).

Obviously, we found that after a period of time, the OD600 value of the bacterial solution with additional nitrogen source showed a decreasing trend. And the decreasing trend was more obvious with the increase of the concentration of nitrogen source. Under this experimental condition, it can be considered that our suicide circuit is regulated by nitrogen concentration and the overall pathway construction is successful.

Phenotype validation

In the process of triparental mating, we found that the correct exchanged strains did not express significant trait changes in the medium used for the first screening, whereas the phenotype of hyaline colonies appeared in the medium used for the second screening. We believe that the suicide module played a role in causing the strain to accumulate too much toxin and die. We therefore re-cultured the strain with the suicide module in nitrogen-free M9 medium shaking conditions. The cultured bacteria were spread on TY solid medium to observe their growth (Figure 4-5-1).

The toxin-antitoxin system functions under high nitrogen and high oxygen conditions, and the excess toxin content leads to colony death. We were able to observe that colonies from three-day cultures showed a milky white color, while colonies from five-day cultures showed a transparent color. This is the most direct evidence that the suicide circuit was successfully constructed.