Overview

To allow our modified E. coli to be applied in the situations we envisioned and in practice, we began by:

Improve the expression of adsorption proteins

All our plasmids contain promoters that are regulated by lacI. To turn on downstream expression, a certain amount of IPTG is required, then we created an engineering cycle for IPTG input.

Cycle 1

Design

We added 0.5mM IPTG after having a meeting with our technician, who gave our suggestion based on his experience.

Build

Then we constructed CsgA-MBP3 plasmid that containing promoter controlled by IPTG. After harvested E. coli containing CsgA-MBP3 plasmid, 0.5mM IPTG was added for 8 hours culture.

Test

We measured the effect of induction of E.coli by 0 and 0.5mM IPTG through RT-qPCR.

Learn

After obtaining the results, we wondered if adding more IPTG could increase the expression of the target gene, so we shared the data with the modelling group to find an answer with the help of the computational simulation.

Cycle 2

Design

According to the requirement of experimental group, modelling group needed to design a model based on the mechanism that IPTG binds with LacI to adjust the promoter activity of tac promoter. Before we built the model, we made some assumptions based upon biological pathway(details seen in the modeling page). The purpose of the model was to find the relationship between concentration of IPTG input and tac promoter activity and to respond optimal concentration of IPTG for the maximal expression of MBP.

Build

On the basis of assumptions and biological pathway, we obtained the following equation:

Test

In matlab, we input 0~2 $\mu M$ IPTG to simulate the model and obtained the relationship between the concentration of IPTG and tac promoter activity:

Learn

We successfully demonstrated that with the concentration of IPTG input increases, the promoter activity of tac promoter grows up but reach a peak around 1 mM and remain at this promoter activity after that in theory.

Cycle 3

Design

According to the computational simulation outcome from modeling group, experimental group chose 0.5 mM and 1 mM IPTG as addition to one type of modified E.coli (CsgA-MBP3).

Build

To verify the conclusion from modeling group, experimental group transformed plasmid containing CsgA (CsgA-MBP3) into E.coli and designed primers based on CsgA sequence.

Test

Experimental group measured the effect of induction of E.coli by 0, 0.5, 1mM IPTG through RT-qPCR.

The result showed the expression of CsgA-MBP3 is highest in the group added 1mM IPTG.

Learn

We have successfully demonstrated, both theoretically and experimentally, that 1 mM of IPTG best facilitates the activity of the tac promoter in the plasmid.

Improve tolerance of E.coli to Metal Ions

We prioritized the design to improve the E.coli's survival rate in order for the modified E.coli to survive in effluents with high metal ion content in the future.

Cycle 1

Design

We had planned for MBP3 and metallothioneins (MTs) to be displayed on the surface of E. coli in order to adsorb metal ions, and we thought this design would also improve the tolerance of E. coli for metal ions. To enable the modified E. coli to be employed in wastewater with greater metal ion concentrations, it was our first thought to further improve the fraction of adsorbed proteins (MBP3 and MTs) on the surface of E. coli. We therefore decided to knock out the CsgA gene from the E. coli by Crispr/Cas9 before transforming plasmids carrying CsgA and adsorption proteins (CsgA-MBP3, CsgA-SmtA, SUMO-CsgA-SmtA, CsgA-ShMT3, SUMO-CsgA-ShMT3), resulting in transformed E. coli with MBP3 or MTs in each curli fimbria.

Build

Based on the sequence of CsgA, we designed the gRNA and constructed it into the pCrispr plasmid. After the transformation of pCas9 and pCrispr plasmids, we harvested knock-out CsgA E.coli (ΔCsgA-E.coli). Then we constructed SUMO-CsgA-ShMT3 plasmid and transformed it into ΔCsgA-E.coli.

Test

To verify the knock out of CsgA, we measured the expression of CsgA by RT-qPCR.

The result of RT-qPCR showed that CsgA gene was knocked out from E.coli.

Then we tested the metal tolerance of modified E. coli, which using ΔCsgA-E.coli as the chassis bacteria.

We measured the Absorbance values of each group after eight hours of addition of different concentration CuSO$_4$ and CdCl$_2$ (the Absorbance values of each group were the same at begin).

Learn

Contrary to what we expected, the use of ΔCsgA-E.coli as a chassis bacterium did not result in a significant increase in metal ion tolerance.

We speculated that this is because wild-type E. coli expresses more curli fimbriae forming biofilms, which were reported to have the function of adsorbing heavy metal ions$^1$. Therefore, we changed our initial design and used MG1655 E. coli instead of ΔCsgA-E. coli as the chassis bacterium.

Cycle 2

Design

Based on the outcome from last cycle, we decided to use MG1655 E. coli as the chassis bacterium.

Build

We constructed SUMO-CsgA-ShMT3 plasmid and transformed it into ΔCsgA-E.coli and MG1655 E. coli.

Test

Then we tested the metal tolerance of two modified E. coli, one using ΔCsgA-E.coli as the chassis bacteria and the other using MG1655 E.coli.

Learn

Improve detection method of metal ions adsorbed by modified E.coli

Cycle 1

Design

To measured the concentrations of metal ions adsorbed by E.coli, we had tried to obtain the supernatant by centrifugation and used ICP-MS to determine the concentrations of metal ions in the supernatant that were not bound by E. coli and the concentrations of the added metal solutions. This is because ICP-MS instruments do not allow large crystals or organic matter to be present in the measured material and all samples need to be filtered prior to testing. Then, the actual concentration of metal ions adsorbed by E.coli was calculated by the difference between two measured values.

Cu(II) concentration (adsorbed by E.coli) = Total added Cu(II) concentration - unbound Cu(II) concentration

Build

Test

The results showed that the actual concentrations were less than 50% of the configured solutions, which was clearly abnormal.

Learn

We speculated that this was due to excessive macromolecules blocking the pores of the filter membrane in the mixed system of bacteria and metal solutions, resulting in a large number of metal ions being unable to pass through.

Improve detection method of metal ions adsorbed by modified E.coli

Cycle 2

Design

Based on our speculation, we changed our method and decided to measure the concentrations of metal ions adsorbed by E.coli directly. Unlike the original method, the precipitate after centrifugation was used as the measurement object, but microwave digestion was used to break down large crystals or organic matter in solution. After the substance has been broken down, the metal ions we want to measure pass more easily through the pores of the filter membrane.

Build

We measured the concentration of CuSO$_4$ and CdCl$_2$ ready to be added to E. coli with ICP-MS again.

Test

After measuring the concentrations of CuSO$_4$ and CdCl$_2$ with the new method, the results showed that the actual CuSO$_4$ concentration was 91.43% of the configured concentration and the actual CdCl$_2$ concentration was 88.69% of the configured concentration. This was a significant improvement in accuracy compared to the original method.

Learn

After obtaining results close to the actual concentration, we decided to apply this method to measure the concentration of metal solutions adsorbed by modified E. coli.

Improve the concentration of metal ions adsorbed by modified E.coli

Cycle 1

Design

In terms of absorption capability, metal binding peptide and Metallothioneins are functional and novel parts of our biofilm system design, in which we aim to comprehensively build and test. An atomic level simulation is tremendously helpful and typically generates substantial insight about how the target component functions, simultaneously proving the feasibility and robustness of the system.Here we design a computational framework to provide insightful guidance before experimental measurement to optimize our systematical performance at molecular level, including Autodock, Gromacs and relevant intelligent approach for simulation and analysis.

Build

Here we build a comparison and validation workflow consists of five groups of docking simulation under the computational framework-Autodock vina, which is subsequently simulated using gromacs software package, in synergy to prove the binding performance between target heavy metal ions and the designed protein complex.

Test

The Test session of Autodock simulation shows the feasible binding affinity with multiple kinds of heavy metal ions in electroplating and mining sector, indicating enormous potential of our designed complex(group 5) in terms of bioremediation capability.

Figure 13

After completing the Gromacs simulation workflow(setup, solvation, minimization, equilibration(NPT,NVT)),detailed analysis in Modeling session(RMSD,RMSF,PCA Gibbs Free Energy Landscape) also proves rational performance of our parts design.

Learn

In this Cycle regarding Molecular level modeling,computational speculation(Autodock,Gromacs,etc) of our designed protein provides theoretical basis and practical guidance for our heavy metal absorption system. We will further proceed to rational design and more intelligent approach like using Deep Learning framework to modify simulation and scoring algorithm based on domain insights are expected, in collaboration with experimental research to construct and validate a efficient bioremediation system.

Cycle 2

Design

We decided to measure the adsorption of metal ions by modified E.coli that grew in 600μM CuSO$_4$ according to the results of E.coli growth curves in different concentrations of CuSO$_4$. The result showed that modified and control E.coli can grow in 600μM CuSO$_4$ (for more information about the result of this part, please visit our Results).

Build

600μM CuSO$_4$ were added into the culture of E.coli for 8 hours.

Test

We used ICP-MS to measure the concentration of metal ions adsorbed by modified E.coli grew in 600μM CuSO$_4$.

However, the results showed no significant difference in the adsorption of metal ions between the modified E. coli and the control group.

Learn

According to one equation about absorption in the model: $\pmb{[Cu^{2+}·absorbed] = \frac{n_3[CsgA·MT][Cu^{2+}]^{n_3}}{K_{D2} + [Cu^{2+}]^{n_3}}}$, with the concentration of unbound copper ions increases, the maximal absorption of copper ions increases.(Since small amount of CsgA·MT in the experiment led to lower absorption of copper ions compared with total input, the concentration of unbound copper ions could be considered to approximate the concentration of input). Therefore, the results might be more significant with higher concentration of $Cu^{2+}$ input.

Cycle 3

Design

Based on the suggestions from cellular and population level modeling, we did the experiment again while E.coli was growing in 1800μM CuSO$_4$ this time. Before testing the adsorbed ability of modified E.coli growing in 1800μM CuSO$_4$, we have measured the growth curve of E.coli, which verified that modified E.coli could still grow in 1800μM CuSO$_4$.

For more information about the result of this part, please visit our Results.

Build

1800μM CuSO$_4$ were added into the culture of E.coli for 8 hours.

Test

We used ICP-MS to measure the concentration of metal ions adsorbed by modified E.coli grew in 1800μM CuSO$_4$. Then, we compared the adsorption result of modified E.coli grew in 600μM and 1800μM CuSO$_4$.

Learn

The result showed that E.coli grew in higher concentration of CuSO$_4$ (1800μM) could adsorbed more Cu(II) than in lower concentration of CuSO$_4$ (600μM), which verified the suggestion offered by modeling.

We also successfully verified that our E. coli can work robustly in a wide range of different concentrations of CuSO$_4$ solution and perform even better at higher concentrations, which could allow our project to be applied to more wastewater treatment scenarios in the future.