Artificial Pattern Formation in E.Coli

David Van Valen

Oren Schaedel

Michael Amori

Plasmid Design

Materials and Methods

Results

Conclusions

Abstract

Weiss et al created a band pass filter using genetically engineered building blocks and used them to create spatial patterns of gene expression. We attempted to reconstruct this experiment in order to investigate effects of AHL diffusion on the generation of spatial patterns as a function of distance, location, and timing. To reconstruct this experiment, we first transformed DH5a cells with plasmids received from the Weiss group. We then set out to perform two sets of experiments on our transformed cells. First, we wished to characterize the band pass filter by measuring the expression of GFP as a function of AHL concentration. Second, we wished to examine the types of spatial patterns we could generate by varying the location of AHL emitting cells. Unfortunately our experiments were impeded by a number of technical dilemmas and we were only able to verify the band-pass nature of one plasmid strand.

Plasmid Design

The design of pHD and pLD allows the concomitant transcription of CI and LacIm1 by the transcriptional regulator LuxR (which is activated by AHL). LacIm1 inhibits the Lac promoter on which GFP is present and CI represses the lambda promoter, onto which the LacI is transcribed. Thus at high AHL concentrations, both CI and LacIm1 are transcribed shutting off GFP expression. CI inhibition is stronger than LacIm1 shutting off the native LacI, which represses GFP, therefore at medium AHL concentrations, LacIm1 is less effective in pLac inhibition and GFP is turned on. At low AHL concentrations, CI is transcribed at basal levels, allowing transcription of LacI (WT) which represses GFP transcription.

When describing this as a function of distance, sender cells emit AHL which diffuses as a gradient throughout the plate. High concentrations are close to the sender cells, and low concentrations are present far form the cells. Plating cells with both pLD and pHD around sender cells forms a fluorescent ring due to the gradient of AHL. 3 different versions of pHD were constructed, high sensitivity (HD3), medium sensitivity (HD2) and low sensitivity (HD1). HD3 cells have a low copy number of LuxR, therefore higher AHL concentrations are required to successfully create the high detect filter. HD2 cells have a WT LuxR, and HD1 cells have a hypersensitive mutant of LuxR, activated by lower concentrations of AHL.

Figure 1 Design of a band-pass filter

Materials and Methods

We developed a number of protocols that helped streamline our experiments

Transformation
Transformation of the cells was done by electroporation. Frozen cells were mixed with 1 ul of each plasmid and labeled strain BD{x}, or transformed with one plasmid HD. Electroporation was carried out in cuvettes, cells were transferred to LB (no antibiotics) and after 1 hour of recovery transferred to liquid LB with appropriate antibiotics. Cells were incubated at 37^oC over night, and plated the next day onto LB-agar plates with appropriate antibiotics. Glycerol stocks were prepared.

In order to test whether transformation was successful we grew cells in a 96 well plate for AHL induction assay (discussed later).

Cell Culture

Pipette 5 mL of LB into a falcon tube. The antibiotics added depends on the strain that use used. No antibiotics are added for wildtype cultures, chloramphenicol for the single plasmid HD, and chloramphenicol plus kanamycin for the double strain LD. 15uL of chloramphenicol and 10uL od kanamycin are added per 5mL of media. Once the culture media is prepared, we inoculate the culture with cells from our frozen stocks. The cells are grown in the 37 °C room overnight

Induction

The cells for the induction experiment should be placed in cell culture the previous night. We always prepare the wildtype cells in addition to the transformed cells because measurements of their fluorescence are needed to calculate the fold change. 30 uL of the cultured cells are placed in 8mL of M9 media with the appropriate antibiotics several hours before the planned experiment because M9 produces less auto-fluorescence. The experiment also required a solution of 100mM AHL dissolved in phosphate buffer.

The M9 solution of cells is then diluted to an OD600 value of .05. For our experiments we induced the cells with concentrations of AHL ranging from 100-.01. The most effective way to arrange our plate was to use the following configuration

	1	2	3	4	5	6	7	8	9	10	11	12
A	X	100	90	80	70	60	50	40	30	20	10	X
B	X	10	9	8	7	6	5	4	3	2	1	X
C	X	1	.9	.8	.7	.6	.5	.4	.3	.2	.1	X
D	X	.1	.09	.08	.07	.06	.05	.04	.03	.02	.01	X
E	X	0									0	X
F	X	WT	WT	WT	WT	WT	WT	WT	WT	WT	WT	X
G	X	M9	M9	M9	M9	M9	M9	M9	M9	M9	M9	X
H	X	X	X	X					X	X	X	X

To achieve these concentrations, we first performed a 1:10 dilution of the 100mM AHL stock. It should be noted that if we wished to explore different orders of magnitude, all we have to do is either perform a higher dilution of the stock, or just use the stock directly. We then pipetted 220 uL of cells into row A and 180 uL into rows B-D. The following volumes of the 10 mM AHL solution were added:

Cell	Concentration (uM)	Amount to add (uL)
A2	100	2.2
A3	90	1.98
A4	80	1.76
A5	70	1.54
A6	60	1.32
A7	50	1.11
A8	40	.88
A9	30	.66
A10	20	.44
A11	10	.22

Once the AHL is added, we then performed a serial 1:10 dilution of row A. 20 uL from row A are pipetted into row B, 20 uL of row B are pipetted into row C, and 20 uL from row B are pipetted into row D. The space between the wells is filled with dH₂O and the plate is covered with the special plastic top generously provided by Dhananjay Tampe to prevent evaporation during culture. The plate is then cultured overnight in the plate reader. The proper setup file is accessible on snowdome, as are the MATLAB files we created to interpret the data.

The final result of the experiment is to calculate the fold change, which is given by

, where all values are taken relative the values of the M9 media.

Cell Plating

This experiment was orchestrated in order to test whether the band pass filter functioned as a function of distance from a point source. Cells emitting AHL were plated in the center of the petri dish, and BD or HD cells were plated around. A double layer of M9-agar was plated in the following fashion: the bottom layer consisted of M9-agar at 1.5% with appropriate antibiotics was poured and let to dry. A second layer of M9-agar 0.7% was plated on top premixed with the receiving cells, and chloramphenicol only (sender cells have only Chloramphenicol resistance). After second layer dried, filter paper was cut at the diameter of a 1 ml tube, loaded with sender cells and placed in the center of the petri dish. Two plates were made per strain. Controls for this setup were a WT dish, and a sender dish W/O receiver cells.

Results

Figure 2 Fluorescence response of transformed cells to AHL

Unfortunately, we only had one successful trial of the induction experiment. While we were unable to calculate fold-change of gene expression in this experiment, the fluorescence per cell clearly shows a band pass behavior for the BD1 plasmid. From our first experiment, we thought that the concentration points we used were simply to spread out to properly observe the peak of BD2, and did not include high enough concentrations to observe the peak for BD3. We aimed to repeat the experiment for the BD2 and BD3 strains using more a better range of concentrations.

Unfortunately, we ran into a problem during the next set of experiments. We expect our cells to be in the log phase of growth, but when we ran the experiment again, we observed a constant OD600, implying no growth. Plotting the OD600 as a function of AHL concentration, we saw the following for BD2:

Figure 3 OD600 as a function of AHL concentration for BD2

From the graph, we can see that the OD600 decreases as the concentration of AHL increases? Why is that? After some investigation, we were able to determine that the acidic ethyl acetate that the AHL was dissolved in was actually dissolving polystyrene. Almost everything that we used in the lab was made of polystyrene, including the plates used for the plate reader. We attempted to redo the experiment using lower concentrations of AHL, and hence less solvent but we were unsuccessful. Our solution was to order AHL as a solid compound and dissolve it in a phosphate buffer to make a 100mM solution. Working with AHL in a different buffer solves the ethyl acetate problem, but unfortunately we had issues with cell growth in the M9 media. As a result, none of our most recent experiments can provide any useful data.

In parallel with our efforts to characterize the plasmids, we also plated them to see if we would get the pattern formation described by Weiss et al. The experiment was rather lengthy, and we were only able to attempt it once. We did not see any fluorescence on the plates. Our main suspicion is that either the cells were unable to grow in the media, or that the production of AHL and its diffusion through the agar was so high and so fast that we missed the window where the concentration in the plate was within the necessary concentration range.

Conclusions

Since we don’t have many results, we can advise of steps to improve future attempts.

1. AHL should be ordered a week in advance, solidified in PBS, pH 6.5 aliquoted and frozen.
2. Calibrating cells to AHL sensitivity after transformation: on a plate arrange 4 orders of magnitude for strains BD2 and BD3 ranging from 0.01mM to 100 mM AHL, and 2 orders of magnitude for BD1 from 0.001 to 0.1 mM. Controls are M9 medium, WT (untransformed) DH5a cells and relevant strain without AHL induction.
3. Testing fluorescence as a function of distance: prepare M9 medium autoclave with 1.5% agar, and a stirrer, let sit in heat source. Prepare a separate medium of M9 with 0.7% agar and stirrer, and let sit on a heat source. Once it reaches a temperature you’re comfortable to leave your hands on for a few seconds (don’t stick a thermometer in side), add relevant antibiotics and pour 1.5% medium into Petri dish. It takes around 15 minutes to dry. In that time, spin down sender cells, and resuspend the pellet in fresh M9 in 20 fold volume decrease. Pour 0.7% agar-M9 mixed with relevant strain onto first layer and let dry. During that time, plate 100ul of cells onto filter paper. We recommend that the filter paper used should be received as small circles pre-cut and trimmed and sterile. Cutting filters on-site slices filter threading which was observed after 2 days (filter fiber was seen in 10x magnification 1 day after plating). Furthermore cutting the filter increases the risk of contamination. According to the materials and methods section in Weiss et al, a fluorescence response was observed after 15 hours. This is due to transformation efficiency and should not be taken at face value.

We had several experiments which came to mind:

1. Placing filter paper at various locations and observing pattern formation.
2. Plating strains at different times with different fluorescence colors.
3. Sharpening the bands on the agar plates: since the fluorescence is a function of the diffusion coefficient of agar (which can be altered as a function of agar %), a relatively high agar concentration should sharpen the bands of the strains. It would be interesting to see if there is a difference between the sharpness levels of the different strains as a function of agar concentration.
4. If we were able to obtain an inhibitor of AHL, we may be able to engineer symmetry breaking in our system. If you look at the diffusion equation, once you include a term that accounts for the reaction, then you break the invariance of the equation under actions of SO(2). This means that the concentration of AHL would no longer be invariant under rotations and you would expect the spatial pattern of gene expression to experience the same symmetry breaking. This experiment would be even cooler if were able to engineer the cells to be genetically identical because then our system would be able to mimic the symmetry breaking you see in development.