By linking inaK and mlrA to construct an extracellular enzyme display system on E. coli, we engineer a microcystin (MC) degrader bacteria that leads to 2100-fold decrease in toxicity, which has no influence on the normal metabolism of cyanobacteria microcystis aeruginosa.

The following are our Proof-of-Concept experiments to verify our design and proposed implementations.

Construction & Verification of Plasmid

We connected gene inaK and mlrA consecutively onto the shuttle vector pET-23b using restrictive enzyme digestion and PCR. The recombinant plasmid was then transformed into the competent state of E. coli BL21 (DE3). We selected positive transformants from the recombinant strain to expanded culture. The recombinant plasmid was then extracted to be identified with restriction endonuclease digestion.

We were mainly interested in genes inaK and mlrA. Because their sequences were designed to be consecutive, it was also necessary for us to verify whether the combination of inaK and mlrA (indicated as “total”) was present in the recombinant plasmid. The size of the gene fragments and their respective restrictive endonuclease digestion sites are listed as follows:

We separated the targeted genes from the plasmid and performed PCR on them. This way, we obtained larger quantities of the targeted genes, and the gel electrophoresis graph would be of better clarity. The sequences of the primers we used are listed as follows:

We used a 200 bp ladder marker in gel electrophoresis. As shown in FIG.1, there were 3 significant bands respectively around 550bp, 1000bp and 1550bp, each corresponding to inaK, mlrA and combination of the two. This indicated that the plasmid construction had been successful.

Protein Expression

We designed an extracellular enzyme display system with InaK as the anchoring motif and MlrA being displayed. Both proteins were theorized to be located on the outer membrane of our recombinant E. coli. Through cell fractionation and SDS-PAGE, we verified their locations. We used 3 types of cells: 1) E. coli transformed with pET23b-inaK+mlrA plasmid; 2) E. coli transformed with pET23b-mlrA plasmid; 3) E. coli transformed with empty pET23b plasmid as blank group.

We first separated different components of the cell by cell fractionation with ultracentrifuge (Li et al., 2004; Shi et al., 2001). The samples of outer membrane, inner membrane and cytoplasm were obtained and stored at -4°C overnight. SDS-PAGE gel electrophoresis was performed the next day. We then used Coomassie Bright Blue to stain the gel and observe the proteins in each sample. (Liu et al., 2019)

In summary, our subcellular fractionation followed by SDS-PAGE and Coomassie staining successfully identified the locations of the expressed proteins, and verified the construction of our extracellular enzyme display system.

Degradation Assay with HPLC

We first decided to measure the concentration of cyclic microcystins using high performance liquid chromatography. Using samples of known cyclic microcystin-LR concentrations (0.1 mg/L, 0.25 mg/L, 0.5 mg/L, 1 mg/L, 2.5 mg/L, 5 mg/L, 10 mg/L), we derive a standard curve that correlates HPLC area to mass concentration. The R Squared value =0.997, indicating a reliable fit.

We cultured recombinant E. coli in LB Broth at 37 °C for 12 hours prior to the experiment. For the degradation assay, the concentration of Microcystin-LR was at 1 mg/L, and no other components were added except for the degrader bacteria. Microcystin-LR was purchased from Taiwan Algal Science Inc. at >95% purity. The total volume of each reaction system was 20 mL. Both the experiment group and the control group were triplicated.

We cultured the reaction systems at 37 °C, and retrieved samples of 1 milliliter after each hour. Samples were centrifuged at 10000 rpm for 1 minute to separate the bacteria from the clear solution. This also stopped the reaction from proceeding. The clear solution from centrifuge were then placed into HPLC vials for analysis. The intensity of both the cyclic Microcystin-LR and the linearized Microcystin-LR were measured. (Qin et al., 2019)

The results are as shown below. Our recombinant demonstrated capability to degrade microcystins, and could reach complete degradation at lower concentrations. This is favorable for practical applications, as the microcystin concentrations in natural environments would not be as high as those in an laboratory setting.

From mass spectrometry, and comparison with standards derived in previous studies, we were also able to verify the degradation product: linearized microcystin-LR (m/z = 1013).

Influence on Cyanobacteria

When learning stakeholders’ attitudes on using GMOs for bioremediation, we noted a particular concern for the recombinant bacteria’s influence on the environment. After field research at our local Qinhuai River and discussing relevant experiment experience with other teams, we decided to further look into the interaction between our bacteria and the natural algae. We will focus on these 3 aspects:

1.The surviving population of algae. This would be measured by directly counting the algae cells in a small sample from the culture.

2.The metabolic activity of algae. This would be measured by performing qPCR on gene sequences 16S, psaB, psbD, rbcL, fbp.

3.The microcystin production of algae. The concentration of microcystin produced by algae at experimental scope is rather low, and is unsuitable for HPLC detection. The production would also be measured by performing quantitative PCR on genes mcyA, mcyB, mcyD, mcyG that are responsible for microcystin synthesis. (Nishizawa et al., 1999)

FIG.7 Schematic of E.coli-Algae Co-Culture

Created with Biorender

Experimental Setup

The experiment group had bacteria with pET23b-inaK+mlrA plasmid, and the blank group had no bacteria. Each group was triplicated. 120 milliliters of algae – E.coli mixture were placed into each Erlenmeyer flask. The initial E. coli density was 1.52±0.24×109 CFU/mL.

The flasks were incubated at 25 degrees Celsius for 7 consecutive days, and samples were retrieved every 24 hours. Retrieval of RNA and reverse transcription was performed prior to qPCR.

Population of Algae

In the 7-day timespan, there was no statistically significant difference observed between the population density of the control and treatment group. The data for both the control and treatment groups demonstrated a classic logistic growth model. We therefore conclude that our recombinant bacteria has no influence on the growth of microcystis aeruginosa population.

Metabolic Activity of Algae

In order to reflect the metabolism intensity of cyanobacteria, we selected several genes to perform qPCR on. Gene 16S was used as the reference gene in both treatment and control groups. The genes and their primers are listed as the following:

We observed no statistically significant difference between the gene expression of the treatment and control groups. Data obtained on Day 7 are as shown below.

The values of both treatment and control groups were normalized using the reference 16S gene results. For clarity, the expression level of all control groups were taken as 1, and treatment group levels were calculated accordingly.

Microcystin-LR Synthesis of Algae

At an experimental scope, the quantity of algae was not sufficient to generate concentrations of microcystin-LR high enough for HPLC detection. In order to reflect the microcystin production levels of microcystis aeruginosa, we selected several genes (Nishizawa et al., 1999) to perform qPCR on. Gene 16S was used as the reference gene in both treatment and control groups. The genes and their primers are listed as the following:

We observed no statistically significant difference between the gene expression of the treatment and control groups. Data obtained on Day 7 are as shown below.

The values of both treatment and control groups were normalized using the reference 16S gene results. For clarity, the expression level of all control groups were taken as 1, and treatment group levels were calculated accordingly. Although the engineered E. coli could not inhibit cyanobacteria’s further production of microcystis aeruginosa, we believe its degradation capability is already sufficient.

Overall, our bacteria demonstrated no influence on the life activities of cyanobacteria, and thus has excellent environmental compatibility.