Our ternary microbial symbiosis system is consist of three chassis: Escherichia coli, Synechococcus elongatus , and Azotobacter caulinodans. S. elongatus fix CO₂ through photosynthesis and secretes the product, sucrose as the carbon source for heterotrophic chassis. A. caulinodans convert the N₂ in Mars atmosphere into the universal nitrogen source, NH₄⁺. E. coli plays the role of production chassis.

To achieve high stability and production efficiency, two strategies are applied to our symbiosis system. First, we hope to build up an effective nutrient circulation among the above three microorganisms. Second, we construct intercellular genetic circuit feedback to regulate the nutrient flux in our system. In the genetic circuit, starvation promoters in E. coli are sensors for nutrient signals. The lux quorum sensing system undertakes the function of transmitting intercellular signals and acts on S. elongatus and A. caulinodans to regulate nutrient fixation efficiency.

For implementation, suitable microbial fermentation equipment and pattern for our system are particularly important. We proposed a novel fermentation model named Separate-Immobilized Fermentation. To see more details of the designs of our project, please go to our Description page.

We will present the experimental results on nutrient circulation and intercellular genetic circuit on the three chassis in order, and present immobilized culturing-related data in the fourth part.

The carbon source produced by S. elongatus is provided to E. coli in the form of sucrose. However, some previous papers have reported that wild-type E. coli K12 strains can't utilize the sucrose to grow and produce very well^[1]. To improve the sucrose utilization efficiency, we have expressed a fructofuranosidase (SacC) from Mannheimia succiniciproducens. SacC will be secreted to intercellular space and hydrolyze sucrose into glucose and fructose. Then these two monosaccharides can be uptaken through the PEP-dependent phosphotransferase system efficiently (Figure 1).

A series of constructs expressing SacC with different transcriptional activity was constructed and then transformed into E. coli DH5α (Figure2). Based on the activities of Anderson promoters have been well quantified, the order of expression level of these four constructs is expected as follows: J23101-SacC > J23107-SacC > J23109-SacC > Empty.

Then we cultured all these four strains in an M9 minimal medium which contains sucrose as the sole carbon source. Their growth is quantified by measuring optical density at 600 nm (OD₆₀₀). All of the groups expressing SacC grow faster than Empty in the first 14 hours. Furthermore, compared with J23109-SacC, the SacC high expression groups J23101-SacC and J23107-SacC showed higher growth rates, along with the dose-dependent phenomenon (Figure 3A). After culturing for 10 hours, we take out the culture medium to test the remaining sucrose. The residual sucrose concentrations of culture medium in the SacC-positive groups were significantly lower than that of Empty (Figure 3B). The higher the SacC expression level, the lower the remaining sucrose concentration. These demonstrate that the appropriate expression level of SacC can help E. coli use sucrose as the sole carbon source to grow.

Unexpectedly, Empty reaches a higher OD₆₀₀ at the stationary phase. We hypothesize that this difference may be caused by a metabolism pathway difference between SacC-positive groups and Empty. After 14 hours, SacC-positive groups and Empty may accumulate different by-products. For example, fast-growing E. coli will produce more acetic acid, a common phenomenon called overflow metabolism^[18]. To test this hypothesis, we measure the pH value of the medium after culturing for 48 hours. pH value decrease was found in SacC high expression groups, which indicates that they produce more organic acids (Figure 4). Previous research has proved that organic acids like acetate can inhibit the growth of E. coli, which may cause the lower stationary phase of SacC-positive groups^[19]. Moreover, acetate can not be utilized as efficiently as glucose and fructose, so acetate production is a waste of carbon source. This might be another reason for the lower stationary phase.

Considering E. coli is the production chassis in our system, we set E. coli as the core and initiator of the genetic circuit. Our engineered E. coli is given the ability of nutrient starvation response, and then 'tell' the nutrition situation to the nitrogen fixation bacteria and cyanobacteria through a quorum sensing system.

The stationary phase sigma factor RpoS and the signal molecule cAMP are the major elements in the regulation of transcription under carbon source starvation conditions.^{[2, 3]} These two regulators are involved in two pathways of starvation response, respectively (Figure 5). RpoS is an alternative sigma factor. As with other sigma factors, RpoS binds to the core RNA polymerase and controls the expression of a specific but large set of genes related to stress resistance. Under the conditions of nutrient deprivation or other stresses, the intracellular abundance of RpoS will increase, which will accelerate the transcription initiation of the RpoS-dependent genes (Figure 5A). The other starvation response pathway uses a cAMP receptor protein (CRP) as a regulator and the sigma factor RpoD for transcription initiation. Under carbon source starvation conditions, the intracellular cAMP concentration will increase and further activate CRP. Then CRP binds to a CRP-binding sequence located upstream of the transcription start site and activates this CRP-dependent promoter (Figure 5B).

To achieve the nutrient starvation response, we first selected three endogenous carbon starvation promoters as the sensors: PyciG (BBa_K4115018), PcstA (BBa_K4115003 or BBa_K118011), and PcsiE (BBa_K4115016). These promoters use different transcriptional factors and sigma factors (Table 1).

To check if the selected promoters can be activated by nutrient limitation (starvation), we constructed report genes as Figure 6A. E. coli transformed with such report genes are cultured in an M9 medium containing glucose with different concentrations as the sole carbon source. After 6 hours, the cultures are taken for measuring fluorescence intensity (FI) and OD₆₀₀. The fluorescence intensities normalized by OD₆₀₀ indicate the relative promoter activities.
After communicating with our partnership team BUCT_China, we decide to add an LVA degradation tag on the C-terminal of sfGFP. (To see more collaborations with BUCT_China, you can go to our partnership page.) sfGFP has a very long half-life in the cytoplasm, which makes it not suitable for indicating some immediate change in gene expression. LVA tag can reduce the half-life. So in principle, using sfGFP with LVA tag as the reporter can more realistically reflect changes in promoter activity (Figure 6B).

The following data in Figure 7 demonstrates that all three selected promoters have a higher promoter activity under the low glucose concentration. Also, sfGFP with LVA is a better reporter for PcstA and PcsiE (Figure 7B). After adding the LVA tag, the relative FI/OD₆₀₀ decreased since the shorter half-life. Promoter activity fold changes at different glucose concentrations were also increased with the addition of the LVA tag. So we choose sfGFP with LVA as the reporter in experiments of Figure 8 and Figure 9. Unexpectedly, LVA tag increases the relative FI/OD₆₀₀ of PyciG abnormally (Figure 7A). This may be due to LVA tag competitively inhibiting the degradation of sigmaS, for RpoS degrades in the same pathway with LVA-tagged proteins.^[2] Another interesting finding is on J23101. It is usually known as a constitutive promoter, but its activity changes with the global metabolism level. At high glucose concentrations, the activity of J23101 was significantly higher than that at low glucose concentration.

We mainly evaluate the quality of promoters from two aspects:
First, promoter activity is an important evaluation index. Empirically, we believe that promoter activity not lower than the J23101 (a constitutive promoter with moderately strong activity) is necessary for production and genetic circuits. Second, fold-change is another important perspective for promoters, especially for those being used in complicated genetic circuits. Fold-change can be defined as the ratio of the activity in the activated state to the non-activated state. For our starvation promoters, the practical definition is the ratio of promoter activity under low glucose concentration to that under high glucose concentration. We want the fold change to be as large as possible.
According to the above data in Figure 7, we find that PcstA has the largest fold-change and proper activity among the three starvation promoters. To further increase its fold change and activity, we make some mutants on its CRP-binding sequence and Fis-binding sequence (Figure 8). To see the detailed design strategy of these mutants, please go to our Improve page.

We first cloned all of the report gene constructs into pUC high-copy number backbone to see if our mutants have better properties. E. coli transformed with these constructs are cultured in M9 minimal medium with different glucose concentrations. A significant promoter activity increase was observed on PcstA_Mutant1 (Figure 9A). However, some unwanted phenomena also appeared. First, the leakage expression of PcstA_Mutant1 under high glucose concentration increases, which results in the decrease of fold-change. Second, the high copy of PcstA_Mutant1 in cells has some side effects on growth (Figure 9B). We hypothesize that the high copy of PcstA_Mutant1 causes these side effects. CRP regulates the gene expression related to central metabolisms like glycolysis and pentose phosphate pathway. For example, CRP activates the expression of glk(glucokinase), which catalyzes the first step of glucose metabolism. So we suppose lack of CRP will decrease the glucose utilizing of E. coli. The mutant on the CRP-binding sequence of PcstA_Mutant1 increases its affinity to CRP. If PcstA_Mutant1 keeps a high copy in cells, it may compete the CRP with the endogenous CRP-dependent promoter. Under the high glucose concentration, even though the abundance of CRP in cells is low, the high copy and high affinity of PcstA_Mutant1 cause the small amount of CRP to be still used to activate PcstA_Mutant1. This might be a possible explanation of the higher leakage activity of PcstA_Mutan1. Less CRP proteins can be used to regulate endogenous gene expression and cause some side effects on global metabolism^[20].

Considering E. coli is the production chassis in our ternary microbial symbiosis system, side effects on growth and global metabolism are serious problems. To eliminate the undesirable properties of PcstA_Mutant1, we shifted all constructions to the low-copy number pET backbone and repeat the experiment (Figure 10). Another PcstA variant, PcstA_Mutant2 is tested in this-time experiment. The bacteria growth is back to normal in all of the groups, which may partially support our hypothesis on the growth. Mutants of PcstA show significantly higher activities as well as larger fold-changes with low-copy number backbone. Compared with PcstA_Mutant1, PcstA_Mutant2 shows higher activity and more obvious does-dependency. However, the leakage expression of PcstA_Mutant2 is higher than that of PcstA_Mutant1, which results in a decrease in fold-change. The data of relative promoter activities and fold changes of PcstA variants are summarized in Table 2.

In conclusion, after two cycles of optimization, by mutating the regulation-related sequence and replacing the plasmid backbone, we finally get Pcst_Mutant1 (on pET backbone) as the best construct for the nutrient response. We are never satisfied with the fold-change. For the future plan, we intend to use an 'amplifier' strategy to further optimize the properties of this nutrient response part.

We choose a widely used quorum sensing system, lux for this purpose. The whole lux system can be separated into two parts: the signal sender and the signal receiver. 3O C6 HSL is the signal molecule of quorum sensing. The signal sender expresses an AHL synthetase, LuxI (BBa_C0161). The signal receiver needs to express a composite part device BBa_K4115039, which is mainly based on LuxR (BBa_C0062).
By design, the signal receivers should be cyanobacteria and nitrogen fixation bacteria. However, we met some problems with expressing the receiver device in those two chassis. To test the feasibility of our constructed quorum sensing system, we first expressed the receiver device in E. coli.
To examine whether the signal receiver can function properly, we induced the reporter (sfGFP) expression in the receiver with 3OC6 HSL at concentrations ranging from 10^-4 to 10^-14 M. After 3 hours induction, the cultures are used for measuring FI and OD₆₀₀. The data was fitted with a four-parameter logistical curve model as shown below.

a is the maximal sfGFP output, b is the basal sfGFP output, [3OC6 HSL] is the supplied inducer concentration, sfGFP is the fluorescence response of the device for that inducer concentration, EC50 is the inducer concentration that results in half maximal activation of the device, and h is proportional to the value of the steepest slope along the curve (Hill coefficient) that indicates the responsiveness of the device to the input.^[7] Our data fitted the logistical curve successfully (Figure 11). Concentration-dependent changes in promoter activity were observed between 10^-10 and 10^-7 M. 10^-7 M 3OC6 HSL is sufficient to induce the maximum activity of lux pR (BBa_R0062).

After confirming that our receiver is valid, we used the supernatant of the signal sender culture to activate the signal receiver (Figure 12). 10^-7 M 3O C6 HSL is supplied for positive control. The supernatant successfully induced the receiver to produce efficient sfGFP expression compared to the negative control (Blank). Referring to Figure 11, the concentration of 3O C6 HSL in the supernatant is about 10^-9 M.

In conclusion, these experiments demonstrate our quorum-sensing constructs can achieve intercellular communication between bacteria.

Since cyanobacteria are a brand new chassis for our team, we first explored how S. elongatus grow under our culture conditions. We measure the growth curves of two strains, UTEX2973 (Figure 13) and HL7942 (Figure 14) over 700 hours to test their growth situation.

Since the strain used in the following experiments was S. elongatus HL7942, we focused on analyzing the growth curve of S. elongatus HL7942. The curve reflects that its growth trend is closer to a straight line rather than a logistic curve. This is mainly caused by cell-cell shading.^[8] In other words, because we put the flask with bacterial fluid into a horizontal shaking table for culture, there's poor mixing of bacterial fluid at the centre and outer of the flask. The light received by the bacteria in the centre of the flask is absorbed by the outer bacteria. Thus, with the centre of the flask as the centre of the circle, the growth situation of S. elongatus is different at different radii. When we mixed the whole bacterial fluid and measured the growth curve, its image was in a straight line. To see more details, please go to our Model page.

As an autotrophic microorganism, S. elongatus can produce biomass using CO₂ as the substrate. However, in natural conditions, S. elongatus prefer to store biomass for use but not provide for other organisms. To overcome this problem, we integrated a sucrose permease, cscB expression box to S. elongatus genome.

Construction of plasmid

We expected that our S. elongatus can improve the ability to secrete sucrose by expressing CscB, sucrose permease. In order to integrate CscB gene and the screening marker gene into the genome of S. elongatus, we constructed a plasmid used to perform genetic manipulation on S. elongatus containing cscB gene and kanamycin resistance gene located between the homologous arms upstream and downstream of the neutral site III. We finally constructed a pUC57-NS3-2-Plac_UV5-cscB-lacI-KanR-NS3-1 (Parts: BBa_K4115045) plasmid (Figure 15).

To complete this work, we conducted several molecular cloning experiments. We linearized the vector and fragment to be used and recombined them by Gibson assembly.
The final constructed plasmid was sent for sequencing and the results proved its correct sequence (Figure 16).

Transformation and recombination of S. elongatus

We used S. elongatus HL7942 for gene manipulation and coated two kanamycin-selective BG11 plates. After 5 days of culturing, several small colonies occurred. After 8 days of culturing, we picked two colonies of each plate and inoculated them into the liquid medium (Figure 17).

After 18 days of culturing, we obtained S. elongatus that reached the logarithmic growth period. To test whether the target gene, cscb and lacI, was successfully integrated into the S. elongatus genome, we did colony PCR at different annealing temperatures to exclude non-specific binding. The result of the agarose gel electrophoresis is shown in Figure 18. We recovered the target bands of engineering bacteria from tracks whose annealing temperature is 59℃ and 60℃ and sent them for sequencing. The results showed that we successfully inserted the target gene with the correct sequence (Figure 19).

Test on the capacity of engineering S. elongatus to produce sucrose

In order to test the ability of the engineering bacteria, we set several control groups and experiment groups under different conditions. These conditions included whether 1mM IPTG or 100mM NaCl existed in the medium. Set three repetitions for each group. We detected the sucrose concentration of supernatant and OD685 value every 24 hours and lasted for 3 days. We compared the accumulation of sucrose secreted by a unit number of S. elongatus in a certain period of time by comparing the ratio of sucrose concentration in the supernatant to OD₆₈₅ value of S. elongatus (Figure 20).

The results show that under any conditions, engineering bacteria have a better capacity for sucrose production than wild-type ones. In addition, engineering bacteria have the best capacity when 1mM IPTG and 100mM NaCl exist at the same time among all of the different conditions. In order to understand the changing trend of the capacity for sucrose production of engineering bacteria more intuitively over time, we made a graph based on the ratio of the capacity of engineering bacteria and wild-type ones (Figure 21).

The figure shows that the ratios under each condition increase over time except for the one with both IPTG and NaCl, yet the latter begins to increase after 48 hours of culturing. The reason for this phenomenon may be caused by cscb. As a sucrose permease, cscB can not only transport sucrose from intracellular to extracellular but also reverse transport. As a co-transporter protein, CscB co-transports protons with sucrose along the proton concentration gradient. During the first 24 hours, engineering S. elongatus secreted so much sucrose and H⁺ that may greatly change the liquid surrounding. As a result, during the second 24 hours, engineered S. elongatus may import some sucrose and grow much faster. Thus, the ratio went down. However, this problem will not occur in our designed system for our culture system is a large mobile phase. The surrounding of the liquid won't be changed that greatly.

To study the response of S. elongatus to signal molecules, we decided to integrate Plux and LuxR into S. elongatus genome. LuxR gene encodes R-protein that can receive the signal of hunger, AHL, and Plux is a promoter that is activated by LuxR in concert with HSL. In addition, we chose sfGFP as an reporter protein used to indicate the intensity of Plux. In addition. In order to integrate LuxR gene, the screening marker gene, and sfGFP gene with Plux into the genome of S. elongatus, we constructed a plasmid used to perform genetic manipulation on S. elongatus containing genes or sequences above located between the homologous arms upstream and downstream of the neutral site I. We finally constructed a pUC57-NS3-2-KanR-J23101-B0034-LuxR-B0015-Lux pR-B0034-sfGFP-B0015-NS3-1 (Parts: BBa_K4115046) plasmid (Figure 22).

We linearized vector pUC57-NS3-2-NS3-1 and fragment KanR-LuxR-sfGFP for Gibson assembly. The agarose gel electrophoresis results of linearization are shown in Figure 23. Then the constructed plasmid is sent for sequencing. The results showed that we successfully constructed this plasmid (Figure 24).

We transformed S. elongatus-cscB using this plasmid. As of the writing of this text, no colony has been produced in the selected culture medium. Due to the slow growth of S. elongatus and time constraints, we were not able to repeat the experiment again.

In the system, A. caulinodans is responsible for nitrogen fixation by nitrogenase. A. caulinodans is different from many other rhizobia. It can perform nitrogen fixation in the free state while other rhizobia only do this when are symbiotic with plants^{[9, 10]}. Besides, A. caulinodans has been relatively well studied and genetically modified in previous literature, which provides us with sufficient experience and guidance to carry out the operation^{[11, 12]}. Therefore, we chose A. caulinodans ORS571 strain as our final nitrogen provider in the project. We got the A. caulinodans ORS571 (AmpR+) strain from Professor Huawei Liu of College of Life Sciences, Northwest A & F University as a gift.

All subsequent experimental validation about A. caulinodans ORS571 should be based on successful genetic operations of it. The genetic operation of microorganisms always involves the transformation of heterologous DNA fragments into the cell. In most previous literature, the transformation operation for A. caulinodans has been triparental mating, which is considered a complex and difficult method to operate. By contrast, electrotransformation is a much simpler and less time-consuming method. We performed eletrotransformation to transform the pBBR1MCS-2 (KanR) into A. caulinodans ORS571 cells as proof of successful genetic operation (Figure 25). By reviewing the literature and summarizing the methods, we got a reliable protocol for the electrotransformation of A. caulinodans and did experimental validation^[13]. To see more details about the electrotransformation, you can go to the Experiments Page.

In the nutrition regulation genetic circuit, A. caulinodans acts as the nitrogen output, which receives intercellular signals to control its output rate. This requires two aspects: one is the response of intercellular signals, and the other is the controllable regulation of nitrogen fixation rate.

We constructed the LuxR-lux pR-sfGFP sequence into pBBR1 backbone. The LuxR-lux pR-sfGFP sequence can respond to the 3O C6 HSL signal, which has been demonstrated in E. coli. The broad-spectrum shuttle plasmid of Gram-negative bacteria, pBBR1, can replicate in A. caulinodans. By our construction, we hoped to make "pBBR1-luxR-lux pR-sfGFP" transformed A. caulinodans be able to respond to intercellular signal molecule 3O C6 HSL.

After PCR validation, pre-experiments of fluorescence measurement were performed. The results showed that the luxR system seemed to be not workable in A. caulinodans.

We transformed the same plasmid into E. coli and performed trial induction, of which the result was good, so there was no problem with the sequence itself.

We assumed that the RNA polymerase in A. caulinodans is unable to interact with luxR protein, which makes the luxR system fail to respond to signaling molecules. Thankfully, previous articles have already had successful results on the intercellular signal response in A. caulinodans through other signal systems^{[11, 12]}. So our concepts in this part can also be proved by reference. In the future, we plan to optimize our experimental conditions and change our signal system for our independent proof of concept.

In A. caulinodans, the nitrogen fixation is tightly controlled by the concentration of intracellular ammonium and this counts for the nitrogenase master regulator nifA^[14]. To achieve controllable regulation of nitrogen fixation rate, endogenous regulators must be removed. In previous research, knocking out of genomic nifA gene and then complementing by nifA controlled by inducible promoters was proved to be a viable strategy^{[11, 12]}. So we first transformed the suicide plasmid into A. caulinodans to construct the ΔnifA strain, and then we transformed the shuttle plasmid into the ΔnifA strain to complement the inducible nifA gene.
We learned from research that the knockout of a target gene can be achieved by inserting a DNA sequence into a specific site on the genome using homologous recombination. Therefore, we got the genome sequence of A. caulinodans ORS571 from Genbank and designed the suicide plasmid (BBa_K4115041).

The pBR322 origin of the suicide plasmid cannot drive a replication in A. caulinodans. The only way the BleoR gene can be passed on as bacteria proliferates is to be integrated into the genome. Therefore, after eletrotransforming the suicide plasmid into A. caulinodans and performing the bleomycin selection, the colonies growing on the plate are likely to be the ΔnifA strain. To verify positive colonies, colony PCR was performed.

After colony PCR, one ΔnifA colony was found. The gel map of the PCR result is shown in Figure 32. After preliminary analysis, the BleoR sequence has been inserted into the genome, and the nifA gene has been knocked out. But there are still some phenomena that cannot be explained well just by PCR and electrophoresis. The Band-Ⅲ coming from the negative control template wild-type genome seems to have a similar length as Band-Ⅰ and Band-Ⅳ but much shallower; Both the wild-type and ΔnifA genome lanes have smaller non-specific Band-Ⅳ and Band-Ⅴ; The Band-Ⅶ coming form the template ΔnifA genome, seems to have the similar length as Band-Ⅷ, but much shallower. So further sequencing validation to figure out what all of these bands are is necessary.

Luckily, all the unwanted bands have been identified by DNA sequencing as other sites on the genome which is far from nifA. The Band-Ⅳ has been identified as the BleoR gene. The results show that the gene knock-out has been achieved, and the ΔnifA strain has been constructed.

Due to limited laboratory conditions, we have not been able to do a direct characterization of nitrogenase activity. Through literature review, we found an alternative plan. The nifH gene expresses a subunit of nitrogenase whose promoter PnifH is tightly controlled by transcriptional activators nifA. Therefore, the part PnifH-sfGFP can effectively reflect the expression of nitrogenase and the content of nifA. We chose the broad-spectrum shuttle plasmid of Gram-negative bacteria, pBBR1 as the backbone, so the plasmid can replicate in A. caulinodans. Therefore, the shuttle plasmid pBBR1-PnifH-sfGFP was built for characterization of the change of nitrogenase expression by knock-out of genomic nifA gene (BBa_K4115042).

After knocking out the genomic nifA gene, the regulation of nitrogen fixation in A. caulinodans should be achieved by controllable nifA expression. In order to do that, we constructed the composite part Plac-nifA-LacI to make nifA gene under IPTG induction. Moreover, the part PnifH-sfGFP was added to reflect the expression of nitrogenase and the content of nifA. Therefore, the shuttle plasmid pBBR1-PnifH-sfgfp-Plac-nifA-LacI (BBa_K4115043) was built for regulation of nitrogenase expression and characterize such expression .

After PCR validation and pre-experiment of fluorescence measurement, there was no positive result. We repeated a few times, but unfortunately, all the experiments failed.

At first, there were no positive bands for the experimental group in PCR validation. Some colonies did grow under kanamycin selection, and a negative control plate with no colony showed that the kanamycin did work. So we assumed that those plasmids were transformed into cells (since kanamycin selection worked), but systematic false negatives occurred in colony PCR. Maybe the DNA polymerase worked badly in colony PCR of A. caulinodans. Then, we directly carried out the pre-experiment of trial IPTG induction and fluorescence measurement of the colonies. The results showed that the "pBBR1-PnifH-sfGFP" transformed cells (both wild-type and ΔnifA) showed no obvious fluorescence, and the "pBBR1-PnifH-sfGFP-Plac-nifA-LacI" transformed ΔnifA cells showed no obvious fluorescence after induction of adequate IPTG.

Due to the time limitation, the experiment had to end though there was no successful result. Thankfully, previous articles have already had successful results in the same experiment, so our concepts in this part can also be proved by reference^{[11, 12]}. In the future, we plan to optimize our experimental conditions for our independent proof of concept.

The separate-immobilized fermentation pattern we proposed should be how the three microbial symbiotic system is implemented in practice. To prove this concept, the three microorganisms should survive under some of the same medium condition, immobilized cells should remain viable for a period of time, and material exchange between immobilized cells needs to be verified. So we first set up a medium condition in which three microorganisms could separately live. Then, microorganisms were immobilized and tested for viability in the gel. After that, all three immobilized microorganisms were cultured in a container, under the same culture medium, and tested for viability. Lastly, the material exchange between immobilized cells was validated.

We had known from previous research about the CoBG11 medium protocols for the coculture of Synechococcus and other heterotrophic microorganisms. Based on this, we proposed our CoBG11-Sucrose medium protocols by adding sucrose as solo carbon source to simulate the sucrose secretion of S. elongatus. Moreover, ammonium chloride in the medium can be used as a simulation of ammonium production by A. caulinodans. To see the ingredients of the culture medium, you can go to the Experiments Page.

Through the resulting growth curve, we found that there is a significant rise in the growth curve of E. coli and S. elongatus for a period of time (Figure 36.a&b). This indicated that E. coli and S. elongatus can live and grow in CoBG11-Sucrose medium . However,the growth curve of A. caulinodans quickly reached a plateau and remained at a low OD, which indicated that it cannot live and grow in CoBG11-Sucrose medium (Figure 36.c). After research review, our hypothesis was confirmed. A. caulinodans prefers to use organic acid as its carbon source rather than sugar^[15]. Therefore, when the medium contains only sugar as solo carbon source, A. caulinodans cannot live and grow.

We found that E. coli would excrete significant amounts of acetate when growing aerobically on glucose as the sole carbon source, which SacC-positive E. coli growing on sucrose will be in a similar situation^[16]. In this case, A. caulinodans can take acetate as its solo carbon source. Therefore, we proposed a new CoBG11-Acetate protocol by adding acetate as solo carbon source to simulate the acetate secretion of E. coli. In this experiment, we culture the A. caulinodans both in CoBG11-Sucrose and CoBG11-Acetate to prove that the change of carbon sources can make A. caulinodans successfully live in CoBG11 medium. To see the ingredients of the culture medium, you can go to the Experiments Page.

After replacing the carbon source, the growth of A. caulinodans was significantly improved. The result confirms that A. caulinodans can live and grow in CoBG11 medium when acetate is the solo carbon source.

To verify the feasibility of "Separate-Immobilized Fermentation", a preliminary immobilization method needs to be designed. After research, we got an achievable protocol, which immobilizes the cells by calcium alginate embedding^[17]. Then, immobilized cells were periodically sampled and lysed through phosphate buffer, and the OD value of lysate was measured to reflect the viability of immobilized cells (Figure 39.d). Besides, for bacteria expressing the green fluorescent protein, fluorescence intensity can be obtained by taking fluorescence images of the lysate to reflect viability. To see more details about the immobilization, you can go to the Experiments Page.

After a week of lysis of immobilized cells and OD or fluorescence measurement, both OD survival and fluorescence survival curves did not show a significant downward trend (Figure 40.a-c). The results showed that we had successfully produced immobilized E. coli and S. elongatus, which maintained high viability in the gel.

Besides, the sfGFP and tagRFP marked immobilized E. coli were observed under a fluorescent confocal microscope. The result would help us to have a clear idea what a single cell is like when being immobilized in calcium alginate gel.

To verify the feasibility of the "Separate-Immobilized Fermentation", three microorganisms need to be immobilized and cultured in the same medium condition. In past experiments, it has been demonstrated that E. coli can survive in CoBG11 medium containing sucrose as a solo carbon source, while A. caulinodans can survive in CoBG11 medium containing acetate as a solo carbon source. In this experiment, sucrose and acetate were added to the CoBG11 medium (called CoBG11-PLUS) to simulate the sucrose secretion of S. elongatus and the acetate secretion of E. coli. To see the ingredients of the culture medium, you can go to the Experiments Page.

The immobilized E. coli expressing red fluorescent protein appears faint red under natural light, immobilized A. caulinodans appears white, and immobilized S. elongatus appears green. By color, we can distinguish three immobilized microorganisms, and then sample and measure the OD value to draw the inventory curve.

After a week of lysis of immobilized cells and OD measurement, OD survival curves did not show a significant downward trend during a week. The results showed that all three immobilized microorganisms can maintain high viability in the gel and in the same medium condition.

The effective substance exchange between immobilized cells is important because all the symbiotic nutrition relationships and intercellular genetic circuit regulation rely on this exchange. We did experimental validation of the intercellular communication in immobilized E. coli through the LuxR system as a demo of intercellular signal communication of our system as well as an intuitive proof of concept that substances can circulate between immobilized cells.

In this experiment, we applied Quorum sensing LuxR system to the intercellular communication in immobilized E. coli. There were two kinds of immobilized E. coli: one carries the LuxI gene, which can synthesize the signal molecule 3OC6HSL; the other kind of E. coli carries the LuxR-Plux-sfgfp sequence, which can respond to the 3OC6HSL signal by expressing sfGFP and emit green fluorescence under blue light. The above two kinds of immobilized E. coli were placed in the same medium to verify the intercellular communication in immobilized microorganisms. Meanwhile, E. coli carrying no LuxI gene and E. coli carrying the LuxR-Plux-sfGFP sequence were placed together to be the negative control.

After two days of co-culture and induction, the two experimental groups showed obvious fluorescence enhancement compared with the negative control groups, which indicated the success of intercellular communication in immobilized E. coli.

[1] Lee, J.W., Choi, S., Park, J.H., Vickers, C.E., Nielsen, L.K., and Lee, S.Y. (2010). Development of sucrose-utilizing Escherichia coli K-12 strain by cloning beta-fructofuranosidases and its application for L-threonine production. Appl Microbiol Biotechnol 88, 905-913. 10.1007/s00253-010-2825-7.

[2] Battesti, A., Majdalani, N., and Gottesman, S. (2011). The RpoS-mediated general stress response in Escherichia coli. Annu Rev Microbiol 65, 189-213. 10.1146/annurev-micro-090110-102946.

[3] Franchini, A.G., Ihssen, J., and Egli, T. (2015). Effect of Global Regulators RpoS and Cyclic-AMP/CRP on the Catabolome and Transcriptome of Escherichia coli K12 during Carbon- and Energy-Limited Growth. PLoS One 10, e0133793. 10.1371/journal.pone.0133793.

[4] Rozen, Y., Dyk, T.K., LaRossa, R.A., and Belkin, S. (2001). Seawater Activation of Escherichia coli Gene Promoter Elements: Dominance of rpoS Control. Microb Ecol 42, 635-643. 10.1007/s00248-001-1012-x.

[5] Gasperotti, A., Going, S., Fajardo-Ruiz, E., Forne, I., and Jung, K. (2020). Function and Regulation of the Pyruvate Transporter CstA in Escherichia coli. Int J Mol Sci 21. 10.3390/ijms21239068.

[6] Marschall, C., and Hengge-Aronis, R. (1995). Regulatory characteristics and promoter analysis of csiE, a stationary phase-inducible gene under the control of sigma S and the cAMP-CRP complex in Escherichia coli. Mol Microbiol 18, 175-184. 10.1111/j.1365-2958.1995.mmi_18010175.x.

[7] Kylilis, N., Tuza, Z.A., Stan, G.B., and Polizzi, K.M. (2018). Tools for engineering coordinated system behaviour in synthetic microbial consortia. Nat Commun 9, 2677. 10.1038/s41467-018-05046-2.

[8] Broddrick, J.T., Rubin, B.E., Welkie, D.G., Du, N., Mih, N., Diamond, S., Lee, J.J., Golden, S.S., and Palsson, B.O. (2016). Unique attributes of cyanobacterial metabolism revealed by improved genome-scale metabolic modeling and essential gene analysis. Proc Natl Acad Sci U S A 113, E8344-E8353.

[9] Jiang, N., Liu, W., Li, Y., Wu, H., Zhang, Z., Alexandre, G., Elmerich, C., and Xie, Z. (2016). A Chemotaxis Receptor Modulates Nodulation during the Azorhizobium caulinodans-Sesbania rostrata Symbiosis. Appl Environ Microbiol 82, 3174-3184.

[10] Lee, K.-B., De Backer, P., Aono, T., Liu, C.-T., Suzuki, S., Suzuki, T., Kaneko, T., Yamada, M., Tabata, S., Kupfer, D.M., et al. (2008). The genome of the versatile nitrogen fixer Azorhizobium caulinodans ORS571. BMC Genomics 9, 271.

[11] Ryu, M.H., Zhang, J., Toth, T., Khokhani, D., Geddes, B.A., Mus, F., Garcia-Costas, A., Peters, J.W., Poole, P.S., Ané, J.M., et al. (2020). Control of nitrogen fixation in bacteria that associate with cereals. Nat Microbiol 5, 314-330.

[12] Haskett, T.L., Paramasivan, P., Mendes, M.D., Green, P., Geddes, B.A., Knights, H.E., Jorrin, B., Ryu, M.-H., Brett, P., Voigt, C.A., et al. (2022). Engineered plant control of associative nitrogen fixation. Proceedings of the National Academy of Sciences 119, e2117465119.

[13] 张俊彪 (2017). Azorhizobium caulinodans ORS571 Ⅲ型分泌系统效应蛋白的预测、鉴定及相关功能分析 (西北农林科技大学).

[14] Demtröder, L., Pfänder, Y., Schäkermann, S., Bandow, J.E., and Masepohl, B. (2019). nifA is the master regulator of both nitrogenase systems in Rhodobacter capsulatus. Microbiologyopen 8, e921.

[15] 郝保海 (2012). 根瘤菌比较基因组学研究及蛋白质互作网络预测 (华中农业大学).

[16] Bernal, V., Castaño-Cerezo, S., and Cánovas, M. (2016). Acetate metabolism regulation in Escherichia coli: carbon overflow, pathogenicity, and beyond. Applied Microbiology and Biotechnology 100, 8985-9001.

[17] 赵景联 (1989). 固定化大肠杆菌细胞生产γ-氨基丁酸的研究. 生物工程学报 5.

[18] Basan, M., Hui, S., Okano, H., Zhang, Z., Shen, Y., Williamson, J.R., and Hwa, T. (2015). Overflow metabolism in Escherichia coli results from efficient proteome allocation. Nature 528, 99-104. 10.1038/nature15765.

[19] Pinhal, S., Ropers, D., Geiselmann, J., and de Jong, H. (2019). Acetate Metabolism and the Inhibition of Bacterial Growth by Acetate. J Bacteriol 201. 10.1128/JB.00147-19.

[20] Shimada, T., Fujita, N., Yamamoto, K., and Ishihama, A. (2011). Novel roles of cAMP receptor protein (CRP) in regulation of transport and metabolism of carbon sources. PLoS One 6, e20081. 10.1371/journal.pone.0020081.