In order to look for phages for specific organism’s TolC proteins, we had to first make sure the proteins were being produced. In all, we isolated the TolC genes for Salmonella, Vibrio cholera, Yersenia pestis, Proteus, Pseudomonas, Klebsiella, and E. coli K12. These were put under the control of the pBAD promoter and a strong E. coli RBS, and the E. coli tolC signaling sequence was put upstream the different organisms’ tolC genes. Using the Western Blots, we were able to show that the proteins were being transcribed, and transcription levels were raised in the presence of the pBAD promoter’s inducer- arabinose (at 0.1% working solution). Note: basal levels of transcription in Luria Broth are sufficient for function (see Kirby-Bauer assays which were done without the inducer) and phage binding (see the example phage plate in Modeling > Experimental Verification, which was also done without any inducer).

Pictured are Example Kirby-Bauer assays. Note that the zones of inhibition are small for the WT E. coli (top left)
and much larger for ΔtolC E. coli (top right).
Some transgenic tolC proteins can restore efflux function to ΔtolC E. coli,like that
of Proteus (bottom left), but not all work- for an example see Pseudomonas's tolC (bottom right)

The Kirby-Bauer assays were the second part of our expression tests. While the western blots tested for the production of the proteins, the Kirby-Bauer assay is testing whether or not the expressed TolC is having an impact on sensitivity to two efflux-resisted antibiotics: Novobiocin and Erythromycin. The Kirby-Bauer assay is a simple test that involves plating what would become a lawn of bacteria (using Top Agar); on top of this filter disks containing an antibiotic are placed. The antibiotics diffuse from the disks and inhibit the growth of bacteria within a certain distance based on the concentration of antibiotic at that distance and the minimum inhibitory concentration (MIC) of the antibiotic against the specific bacteria plated.

Kirby-Bauer assays were analyzed by taking a picture of the plates from the same height (using a stand) and keeping a ruler in the image. The pictures were analyzed using the ImageJ software package. A conversion was set between pixels and centimeters using the “Set Scale” function. A circle was made to reflect the zone of clearing and its properties were exported as a .csv. Excel was used to convert from the circumference to the diameter of the zone of clearing, and each diameter was expressed as a percentage of the ΔtolC E. coli’s zone of clearing’s diameter.

Results of Kirby-Bauer assays done on ΔtolC E. coli with our transgenic molecules.
Green signifies results that indicate a restoration of efflux function in ΔtolC E. coli.

Our project aims to act as a chassis for collecting and isolating new bacteriophages specific to the TolC proteins of pathogenic strains of bacteria, using a safe lab-strain of E. coli. To be used in such a way, we modified a preexisting protocol for isolating bacteriophages from the environment. In doing so, we added three rounds of what we referred to as “phage solution enhancement.” The idea behind this is that the environmental phage source is incubated with a larger volume of overnight-grown ΔtolC E. coli for ~20 minutes; this should be enough time for any bacteriophages capable of infecting these knockout-tolC cells to have an opportunity to bind a host bacterium, but not enough time to replicate and release new phage particles into the environment. These incubations are then chilled and centrifuged. The pellet created at the bottom contains the ΔtolC E. coli and any non-TolC specific bacteriophages which bound/infected the ΔtolC E. coli. The supernatant, which contains any bacteriophages which did not bind, is moved into a new tube, treated with chloroform to kill any remaining bacteria, and moved into another new tube. This becomes the new phage solution and the process is repeated twice more. The protocols are included in the table below.

In order to justify these steps, we attempt to quantify the effect of our process of phage solution enhancement. In order to test the protocol’s effectiveness, our experiment made use of two phages. One TolC-binding phage, TLS phage, and one non-TolC-binding phage we isolated ourselves called Phage #4 (we found that mipA, fhuA, fecA, and flgH knocokout E. colis, were insensitive to this phage). To begin the experiment we infected bacterial cultures (separately) in order to proliferate our phages; 1 mL of overnight grown WT E. coli (wild type) was mixed with 10 mL of LB and 250 µL of isolated phage solution. After incubating overnight at 37°C, these tubes were centrifuged. The supernatants were moved to new tubes; these were treated with chloroform, centrifuged, and the supernatants were again moved to new tubes (these are our starting phage solutions). From there four new tubes were labeled (TLS+, TLS-, 4+, and 4-). The “+” tubes became the “enhanced” tubes, and the “-“ tubes were left “unenhanced”; both “TLS” tubes were started with 5 mL of the TLS starting phage solution, and both “4” tubes were started with 5 mL of the Phage #4 starting phage solution. The “+” tubes were enhanced (see steps 2-9 under “TolC Specific Bacteriophage Screen” above) adding 30 mL of volume to the original 5mL of phage solution. To compensate for the gained volume, the “-“ tubes had 30 mL of sterile LB added.

Expecting high phage concentrations in some of our new phage solutions, we performed 1:10 serial dilutions (of which 100µL of the total 1000µL of each dilution would be used) on all of our new phage solutions down to a concentration of 10^-12. All these dilutions were mixed with 250 mL of overnight grown WT E. coli culture and 4 mL of molten LB top agar; the mixtures were immediately plated. After ~24 hours of incubation at 37°C, the plates were analyzed. Plates with a countable number of plaques (between 30 and 300) were chosen for each type of phage solution. These were used to calculate the number of plaque forming units per milliliter (PFU/mL). The “+” solutions’ PFU/mL’s were divided by those from the matching “-“ solution, to obtain a fractional change in phage concentration from the enhancement process. We expected a fractional change of approximately 1 from TLS+/TLS- and a number significantly less than 1 for Phage #4.