Our project design was to create a new approach to bioremediation platform using flotation to help remove or acquire particles of interest from liquid media that could be applied across industries that produce waste-water. Past iGEM teams have attempted to implement cellular flotation as a remediation platform but have demonstrated inconsistent and inconclusive cell flotation results. Thus, the goal of our wet lab team was to develop an optimized protocol that would be used to induce flotation in Escherichia coli via the expression of gas vesicles, which could be implemented into a bioreactor model designed by our dry lab team. We were able to acquire a genetically engineered gas vesicle operon arg1 from the Shapiro Lab. This construct consists of primary gas vesicle proteins GvPA and GvPC from Aphanizomenon flos-aquae, and secondary GvP R/N/F/G/L/S/K/J/T/U from Bacillus megaterium, which when combined into one construct, improved gas vesicle size and echogenic property resulting in better ultrasound imaging and confer flotation to E. coli cells (1). Since the primary goal of the Shapiro Lab was not to use the gas vesicles for flotation, their assays were not optimized for this function. Therefore, we adapted their protocol for flotation and removed illegal cut sites from the operon. We also used this adapted protocol on a new construct that we created, consisting of GvPs A and C without the secondary proteins to observe differences in flotation.
The arg1 construct (BBa_K2699000) in a pET28a expression vector with a T7 promoter inducible by Isopropyl β-D-1-thiogalactopyranoside (IPTG) was transformed into E. coli BL21 (DE3) strain. After growing the cells from a diluted overnight culture to an OD of ~0.5, they were induced with 1.6mM IPTG. After 22 hours of induction, cells were expected to exhibit buoyancy while non-induced cells were not. Buoyancy was tested by placing induced and uninduced cell cultures in falcon tubes, which were left on the benchtop and pictures were taken over the course of 5+ hours to see shifts in cell density.
(BBa_K2699000) arg1 construct:
After 5 hours on the benchtop post induction (1.6mM IPTG), E. Coli BL21 (DE3) cells transformed with arg1 showed a higher density of cells along the height of the tube vs non-induced and non-transformed cells. they also show a smaller amount of cells precipitated at the bottom of the tube, meaning fewer cells sunk down (Fig 1). This indicates that cells expressing arg1 are more buoyant than non-expressing cells, however, they are not buoyant enough to migrate to the top the culture. To test the replicability of these results the same culture was vortexed a week later and left on the bench top (Fig 2) and the same pattern in culture density change was observed indicating that the gas vesicles produced by arg1 are relatively stable and cells can retain the gained phenotype for up to a week (longer time periods not tested). Even after more than 24 hours on the bench expressing cells didn’t all precipitate at the bottom (Fig 2D). It was not tested if cells lose some of their buoyancy over time or what the percentage loss is.
Culture density starts uniformly across all constructs and we find that the arg1 retains cellular density in the suspended state whereas negative control has cellular density pelleted to the bottom of the tube (Fig 1). Any difference in culture buoyant forces will cause variation in how culture density changes over time. This method for testing buoyancy lacks a quantitative aspect, however, it is the easiest to replicate with the least interference. We found that our pET28a construct containing arg1 performed floatation in LB media over a 25 hour period after induction by IPTG, unlike the negative control.
Fig 1. Cultures of non-transformed cells (BL21 +IPTG,0.8mM), induced cells (BL21 arg1_amilCP + IPTG 1.6mM) and non-induced cells BL21arg1 after 5 hours on the bench post induction.
Fig 2. Culture form figure 1 vortexed a week later and left on the bench. a) after 2 hours on the bench.
Fig 2b) after 3 hours on the bench
Fig 2 c) after 6 hours on the bench.
Fig 2. d) after 24 hours on the bench.
To track changes in culture density overtime, pictures of the cultures were taken every hour for 6 hours (Fig 3). Two different concentrations of IPTG (1.6mM, 1.8mM) were used to test if more IPTG would enhance buoyancy. There was no observable difference between induced cells and negative controls until after 2 hours on the benchtop (fig 3c). The culture density along the height of the tube looks the same, however there are fewer cells at the bottom of the induced tubes compared to the control tubes. After 5 hours (fig 3d), more cells settled to the bottom with the induced tubes still having fewer cells than the controls however a difference in culture density along the height of the tube is observed with the induced tube having a denser culture than the control. After 6 hours (fig 3e), as more cells settle to the bottom, the cultures became clearer and the tube induced with 1.6mM seemed to have the densest culture at the top of the tube.
Fig 3. a) Flotation assay performed Aug 3rd. Time point 0 hours, right after cultures were placed in tubes.
Fig 3. b)Flotation assay performed Aug 3rd. Time point 1, 1 hour since cultures were placed in tubes.
Fig 3. c) Flotation assay performed Aug 3rd. Time point 2, 2 hours since cultures were placed in tubes.
Fig 3. d) Flotation assay performed Aug 3rd. Time point 3, 5 hours after cultures were placed in tubes.
Fig 3. e) Flotation assay performed Aug 3rd. Time point 4, 6 hours after cultures were placed in tubes.
We wanted to determine whether having the primary structural proteins alone was sufficient to make E. Coli buoyant so we cloned GvpA and GvpC in the pKDL071 vector, however the T7 promoter was mutated after cloning. Since the previous concentration of 2mM did not induce flotation, we reverted to using 1.6mM. With 1.6mM, all cultures looked the same after 1 hour consistent with results shown in Fig. 3. The surface of the negative controls is clear and a clear line at the interface where the culture density changes after leaving the tubes for 3 hours on the bench (fig 4c) this clearance line is seen at the top of the induced tube. At 4 hours (fig 4d) this line goes lower in all tubes but is still higher in the induced tube showing that induced tubes are again more buoyant than controls.
Fig 4. a) Flotation assay performed Sep 21st. Time point 0.
Fig 4. b) Flotation assay performed Sep 21st. Time point 1, 1 hour after cultures were placed in tubes.
Fig 4. c) Flotation assay performed Sep 21st. Time point 2, 3 hours after cultures were placed in tubes.
Fig 4. d) Flotation assay performed Sep 21st. Time point 3, 4 hours after cultures were placed in tubes.
Fig 4. e) Flotation assay performed Sep 21st. Time point 4, 5 hours after cultures were placed in tubes.
Fig 4. f) Flotation performed Sep 21st. Time point 5, 6 hours after cultures were placed in tubes.
A plasmid expressing RFP (pBbA7c_RPF) under the same induction control as arg1 was co-transformed in BL21 (DE3) to visualize the cells more easily and track their changes in height over time. We repeated the assay (Fig 5) using 2mM IPTG for induction, as this concentration showed better RFP expression. All cultures for both induced and uninduced cells showed no observable difference in cell density. This could be due to a smaller volume (15 ml) used for this test compared to previous tests (25 ml).
Fig 5. a) Flotation assay performed Oct 13th, timepoint 0.
Fig 5. b) Flotation assay performed Oct 13th. Timepoint 1, 1 hour after cultures were placed in tubes.
Fig 5. c) Flotation assay performed Oct 13th. Timepoint 2, 2 hours after cultures were placed in tubes.
Fig 5. d) Flotation assay performed Oct 13th. Timepoint 3, 3 hours after cultures were placed in tubes.
Fig 5. e) Flotation assay performed Oct 13th. Timepoint 4, 4 hours after cultures were placed in tubes.
Fig 5. f) Flotation assay performed Oct 13th. Timepoint 5, 5 hours after cultures were placed in tubes.
Fig 5. G) Flotation assay performed Oct 13th. Timepoint 6, 6 hours after cultures were placed in tubes.
Fig 6. SDS gel from flotation assay cultures on Sept 21st (fig 4). As highlighted by the yellow boxes, the only difference that is easily observed between non-induced is pBbA7c_RFP.
Fig 7. OD comparison of induced and
The following flotation assay [fig. 8 (a-i)] was captured over the span of 25 hours, and is summarized in figure 7. A flotation phenotype was observed in BL21 cells co-transformed with arg1_pET28 and pBbA7C_ RFP. This culture retained the buoyancy phenotype 25 hours after being placed on the benchtop (1.6mM IPTG) (fig 8 h,i), which shows the robustness of our phenotype. The induced BL21 arg1 + RFP culture showed a higher density along the height of the tube, as well as a smaller mass of cells sinking to the bottom compared to the uninduced BL21 arg1 + RFP (fig.8 a-i). To quantify this observation we decided to create an OD profile of the induced and uninduced BL21 pET28a_arg1 + pBbA7c_RFP cultures by removing and reading sequential aliquots off the top layer of the tube. The last mL of each sample contained the semi-solid settled cell mass at the bottom of each tube. To compare the number of cells that had completely sunk to the bottom of the tube we pelleted the last mL, discarded the supernatant, and measured the cell mass. The induced cell pellet mass was half of the uninduced cell pellet mass (fig 7). We used a two-sample t-test to check whether the mean of the induced and none induced cultures significantly differed. The null hypothesis was that there is no difference between the two samples, while the alternative hypothesis was that there is a difference. With a p-value of p = 0.000027, n =28 and alpha = 0.01, we were able to reject the null hypothesis in favour of the alternative hypothesis This show that the induced cells were significantly more buoyant than the uninduced cells as they remained at a higher levels in the media for a longer period of time.
Fig 8. a) Flotation performed Oct 15th. Time Point 0.
Fig 8. b) Flotation performed Oct 15th. Timepoint 1, 1 hour after cultures were placed in tubes.
Fig 8. c) Flotation performed Oct 15th. Timepoint 3, 3.5 hours after cultures were placed in tubes.
Fig 8. d) Flotation performed Oct 15th. Timepoint 4, 4 hours after cultures were placed in tubes.
Fig 8. e) Flotation performed Oct 15th. Timepoint 4, 5 hours after cultures were placed in tubes.
Fig 8. f) Flotation performed Oct 15th. Timepoint 4, 5 hours after cultures were placed in tubes.
Fig 8. g) Flotation performed Oct 15th.
Timepoint 5, 8.5 hours after cultures were placed in tubes.
Fig 8. h) Flotation performed Oct 15th. Timepoint 6, 25 hours after cultures were placed in tubes.
Fig 8. i) Flotation performed Oct 15th. Timepoint 7, 25 hours after cultures were placed in tubes.
Overall we obtained inconsistent results. A few flotation assays show a clear difference between induced and non-induced BL21 arg1. We have reason to believe that this is due to inefficient arg1 expression. Firstly, the reporter gene amilCP that we introduced to arg1 was not expressed. Secondly, the SDS-PAGE gel showed no indication of a difference between induced and uninduced arg1. To confirm that this was due to the inability of arg1 to be expressed, we also induced the RFP gene in the pBbA7c expression vector under the same induction control as the arg1 operon. In the SDS-PAGE gel (fig 6), there was an obvious difference between RFP protein levels in induced and uninduced cells, indicating effective expression of the protein after induction.
Inefficient expression of arg1 could be due to E. Coli strain differences between our strain BL21 (DE3) and the strain used in the Shapiro Lab paper (BL21-AI) (1). Although, this shouldn’t have an effect on arg1 expression since the main difference between the strains is that the T7 polymerase is inducible by L-arabinose in BL21-AI and IPTG in BL21 (DE3). The expression of RFP in the BL21(DE3) cell line suggests that using this strain should not have been a problem since there was enough T7 polymerase to express the protein.
After excluding the secondary proteins, we did not observe a difference in flotation. Our pKDL071_A+C construct acquired mutations in the promoter region, which could explain why flotation was not induced. As a result of this, we cannot be sure of whether this new construct would be better, worse, or similar in efficacy of forming the gas vesicles essential for flotation.
In the cases where we did observe a difference in flotation, the cells shift from a homogenous mixture to settling more slowly than the negative controls. This is contrary to our expectation of having a distinct cell layer on the surface, which suggests the gas vesicles do not confer enough buoyancy to the cell with an upward force strong enough to allow them to significantly increase their height in the culture, but rather a force only strong enough to prevent them from sinking as quickly as cells without the induced operon.
Further experiments will be carried out to test the effects of secondary protein permutation on flotation. There will also be an investigation into the reason behind the inefficient induction of the operon.
Bourdeau, R. et al. Acoustic reporter genes for noninvasive imaging of microorganisms in mammalian hosts. Nature 553, 86-90 (2018).