Design of Feces Analysis

E. coli (DH5alpha) expressing amilGFP, were cultivated using both agar plates as well as liquid cultures. The first test was done by weighing 5 grams of horse manure in glass petri dishes which were then gently mixed with 5 mL tap water. Various mixtures of either solid or liquid fluorescent chromoprotein expressing colonies were gently mixed with the feces in varying concentrations and quantities. These samples were analyzed using an ultraviolet gel reader in combination with an overhead UV lamp, so that both sides of the petri dish were illuminated.

Figure 1: Solid cultures of amilGFP (left) and RFP (right) gently mixed into horse manure. The purpose of this test was to determine if there was any background florescence and if colonies would be seen on horse manure at all.

Quantification

To focus on quantification, a similar experiment was undertaken using a plate reader. The expansion of the experiment via different sample preparation methods was to see if there was a cheap method (filtering), that would show the presence of the reporter system more easily and quantitatively. From the initial experiment that was previously described, it was decided that amilGFP would be the best candidate for this follow up experiment. With this in mind, a liquid culture of OD600 = 2.13 expressing amilGFP was used in all sample preparations below.

The first sample preparation consisted of horse feces (a mixed batch was prepared with 10 g of feces from 5 different horses) mixed with a liquid culture of amilGFP at different concentrations (1, 3, 8, 15, 25, 50 and 75 %), with the total weight of mixed samples totalling at 1 g. In the second and the third sample preparation, 1 g of horse feces was mixed with LB-media and liquid culture of amilGFP at different concentrations (5, 10, 15, 25, 50, 75 and 100%). In the second and third sample preparations the amount of horse feces was kept at a constant 1 g whilst LB-media was used to dilute the liquid culture so that different concentrations could be attained and the total weight of mixed samples was 5 g.

Figure 2: Sample preparation of plate reader samples. The right shows the samples being put onto a shaker in order to speed up the flow-through time.

The first sample preparation (consisting of a mixtures of horse feces and amilGFP liquid culture) was put directly into a 24-well plate with triplicates of each concentration. This method was done for all samples for all preparation methods. The second preparation of samples was put into a filter (Munktell’s A3-90-700, 70 mm, poorsize 10 µm) and the third preparation into another type of filter (ICA coffee filter “unbleached 102”, cut into circles at 70 mm, est. poorsize 10-15 µm). When 1 mL of liquid could be obtained it was put into a 24-well plate. In order to speed up the process of filtering the sample a shaker was used.

Results

Regarding the second experiment aimed to establish what concentrations would be required in order for fluorescence to be measured with a plate reader. The results can be seen in figure 3 and figure 4 below. Figure 3 shows the fluorescence of samples where feces were mixed with amilGFP and figure 4 shows the fluorescence of the filtrate. The results where feces had been mixed with liquid culture of amilGFP show inconclusive results, with no relationship between the concentration of amilGFP and the fluorescence intensity. However, the results from the samples where liquid cultures of amilGFP, LB-media and feces were mixed and filtered through a coffee filter and a Munktell filter shows some linearity in the fluorescence intensity.

Figure 3: Unfiltered samples at varying concentrations, x-axis and intensity of light being emitted from the sample on the y-axis. As one can see, there is unfortunately no correlation between the different concentrations. I.e. the lowest concentration of amilGFP shows the highest intensity followed by the sample containing the highest concentration.

Figure 4:This graph shows the data from the filtrate, and shows some linearity as could be hypothesized. There is a negative slope from higher concentrations on the right side of the graph to the lower concentrations on the left, showing that higher concentrations of GFP can be seen via the plate reader in filtered horse manure.

Conclusion

It would be interesting to redo the experiment and to see if there would be any change over time. Other settings for the plate reader also might yield different results. Furthermore it would be interesting to redo the experiment with lower concentrations and a smaller interval to come to a conclusion on what the lowest amount bacteria that is required for detection of fluorescent bacteria in feces is. There was also some inconsistency in the fluorescent intensity measured between triplicate samples, and more duplicates would likely make the results more accurate.

Chromoproteins may possibly be a viable strategy as a reporter system in horse manure if a proper amount of cells can be grown. We do not know how cultures would be excreted out of a horse, whether in somewhat solid colonies, completely liquid, or a combination of both.

Building upon this, we decided to go forward and test a new reporter system that responds specifically to bilirubin, since bilirubin is present in mammalian intestines and would potentially prevent extraneous readings. This followup experiment can be seen at our UnaG Page.