Team:HKUST-Hong Kong/mic.html

I. Introduction
In order to quantitatively demonstrate the effect of indole charity as well as our construct’s ability to negate it, we have decided to perform a series of minimum inhibitory concentration (MIC) tests. In these tests, we subjected different strains and mixtures of E. coli to an antibiotic gradient and cultured them overnight (18 hours). The OD₆₀₀ readings of each test were recorded, and they will be shown in later sections for comparison. It is important to note that for each test, we did incubations in both 15ml Falcon tubes (2ml culture) and 1.5ml microcentrifuge tubes (1ml culture) to observe whether oxygen supply would affect the population distribution.[Top]

II. Wild Type (RR1) MIC Test

Phase 1 - Kanamycin MIC test

Experimental Design and Aim:
RR1 is a derivative of the common Escherichia coli strain K12 and is not known to have any antibiotic resistance other than for streptomycin. Hence it was arbitrarily chosen as the non-resistant ‘wild type’ for our tests. A simple MIC test was conducted for RR1 to serve as a benchmark for comparison with later experiments; and kanamycin, an aminoglycoside, was opted as the antibiotic of choice. This was primarily for two reasons:

First, the kanamycin resistance gene incorporated into our selection plasmids functions by producing mutated ribosomes that are insensitive to kanamycin. Unlike some other forms of resistance where antibiotic molecules are directly inactivated, this method not only ensures that the antibiotic levels remain relatively constant throughout the experiment, but also prevents the appearance of satellite colonies during plating.

Another reason is because kanamycin can be both bacteriostatic and bactericidal, depending on its concentration and the microbe’s resistance. As our experiments involve plating out cultures for colony counting, it is useful to have a clear differentiation between cells severely affected by kanamycin (bactericidal effect kicks in and removes vulnerable cells) and those that are sustained by indole (cells either kept in stasis or are unaffected, and thus will have colonies). This allows us to better observe the potency of indole charity when we apply kanamycin at below working concentrations.

Results:
The MIC of RR1 was found to lie between 6~9µg/ml.

Click to enlarge

Phase 2 - Kanamycin MIC test with indole supplement

Experimental Design and Aim:
Indole has been proposed as a key signalling molecule produced by unstressed (high resistant) E. coli as a form of ‘charity’ that grants stressed (low resistance) cells passive immunity against antibiotics. This enables such stressed individuals to continue to survive and proliferate. Indole functions by inducing the expression and activity of multidrug efflux pumps to expel antibiotics and toxins, as well as activating oxidative-stress protective mechanisms to minimize DNA damage.[1] In an attempt to ascertain and quantify this effect, we repeated the kanamycin MIC test, this time supplementing the LB medium with different concentrations of indole (300µM to 1mM).

Results:
The effect of indole on the MIC for RR1 varied under different concentrations. At 300µM, which was the documented natural concentration of indole maintained by unstressed E. coli [1], we saw a clear increase in MIC as shown by a shift of the curve to the right of the non-indole MIC curve. The slower rate of decline of OD₆₀₀ (an estimation of cell concentration), also indicated that at 300µM, indole is helping RR1 survive better in kanamycin. The same trend is observed when indole is increased to 500µM, in which RR1 performs even better, showing significant growth even past the MIC of RR1 cultured in 300µM indole.

On the other hand, further increasing the concentration of indole to 1mM did not seem to yield higher MICs. Rather, the results indicated that RR1 performed similarly in 1mM indole and in normal LB, at times even worse. We have several possible explanations for this. First, indole is inherently toxic. It is possible that at 1mM, the toxicity of indole overcame the benefits it provided, and instead began to kill rather than protect cells. Another possibility is that over-promoted expression of passive immunity mechanisms due to higher than natural concentrations of indole over-exhausted cell resources, leading to cell senescence or even death. [Top]

Click to enlarge

III. Mixed Culture MIC Tests

Phase 1 - Wild type (RR1) with RFP-labelled kanamycin resistance strain (RFP) (99:1)

Experimental Design and Aim:
As mentioned previously, when E. coli cultures are subjected to antibiotic selection pressure, a small number of naturally resistant individuals, at some cost to themselves, provide protection to other more vulnerable cells by producing indole, resulting in an overall enhancement of the survival capacity of the population in stressful environments. To mimic this naturally occurred phenomenon, a kanamycin resistant strain, which represents the mutants, was introduced into the RR-1 at 1:99 ratio. This kanamycin resistant strain was labeled with RFP for easy recognition. The ratio of kanamycin resistant strain (KanR/RFP) to RR-1 was recorded for comparison with that of later mixed culture assays.

Results:
Here we can clearly see the effect of indole charity from the result. Even under 25µg/ml kanamycin, which is half of the recommended working concentration and almost 3 times the MIC of RR1, we are still able to observe significant growth from RR1. In all the concentrations we tested, RR1 remains to compose the majority of the population after overnight culturing. It is particularly interesting to note that even though we were using increasing concentrations of kanamycin, the ratio of RFP to RR1 colonies on our plates remains relatively constant, with RFP occupying around 30-40% of the total population. There did not seem to be a correlation between kanamycin concentration and population ratio when less than 25µg/ml kanamycin was used.

However, we suspect that if we test the remaining range of 25-50µg/ml, there will be a critical value where RFP out-competes RR1 and subsequently dominates the population, which would indicate the limit of the effect of indole charity.

Click to enlarge

Phase 2 - Wild type (RR1) with kanamycin resistance T4MO (GFP)

Experimental Design and Aim:
In order to interfere with indole charity work and thus achieve more efficient selection with antibiotics, we introduced a plasmid expressing the gene for a mutated form of Toluene-4-Monooxygenase (T4MO) into our construct. This variation of T4MO allows it to catalyse the oxidation of indole and hence 'deactivates' indole. Furthermore, the main product of this reaction is 7-hydroxyindole, a derivative known to inhibit biofilm formation in E. coli, which we hope will have synergistic effects in increasing the effectiveness of antibiotics.

In this test, we mixed RR1 and T4MO in a 1 to 1 ratio in volume and incubated the culture overnight. However, as we were unable to complete the strain that relies on antibiotics-free transformation (E.CRAFT), we used a T4MO-GFP/KanR plasmid for regular antibiotic-selection transformation and applied the transformed E. coli instead. As such we have made an assumption prior to the experiment that the indole degradation rate of T4MO will be able to counteract the indole produced by the currently-resistant T4MO strain, in effect treating it as if it were both the resistant strain and the indole system hijacker.

Results:
While we did not achieve close-to-complete elimination of RR1, the ratio between T4MO and RR1 colonies was clearly different from that of RFP and RR1. Rather than maintaining a fixed ratio, T4MO was seen to gradually out-compete RR1, with a sudden increase seen at the normal MIC limit of RR1 (~10µg/ml). This suggests the weakening of indole charity by T4MO, though not to the extent that it could completely eliminate RR1 at half the recommended working concentration of kanamycin.

We have two possible explanations for this. First, the plasmid containing T4MO is primarily maintained by kanamycin selection, and thus there would have been inevitable plasmid loss as we were using below-working concentrations of kanamycin. This would have impacted the efficiency of T4MO as well as the colony ratio of RR1 to T4MO. Another reason is that it is possible that indole is not the sole extracellular molecule providing passive immunity to antibiotics. While we might have extinguished charity from indole, other signalling molecules might still be protecting RR1.

Click to enlarge

Here we have also compiled graphs to compare the ratios of RFP/RR1 cultures to that of T4MO/RR1 ones after overnight incubation. It is clear from the data that the selection efficiency for resistant individuals increased markedly, which might indicate that indole charity work is indeed disrupted, favouring the survival of resistant individuals. [Top]

Click to enlarge

IV. Conclusion

Our results show preliminary evidence that indole at the right concentration enhances wild type E. coli’s resistance to antibiotics, and that interfering with the indole signalling pathway is indeed a potential method of enhancing the effectiveness of antibiotic selection. [Top]

V. Future Plans

Phase 3 - Wild type (RR1), RFP-labeled kanR, and GFP-labeled T4MO/Bcr 3-way mixed culture

As we have not managed to engineer the non-antibiotic selection strain in time, we could not create a GFP-labeled T4MO strain that does not require antibiotics for plasmid maintenance. However, in the event that we can, we will attempt to re-perform the three-way mixed culture MIC tests using this strain to evaluate the true effect of T4MO in indole quorum-sensing disruption.

In this strain, we will include the Bcr gene (coding for a multidrug efflux pump) regulated by pLac to serve as a way to artificially increase the resistance of this mutant strain when exposed to more potent concentrations of kanamycin.

It is hoped that by introducing this strain into the population, indole charity work will be even more greatly restricted due to higher plasmid copy maintenance, so that the efficiency of selection for resistant individuals increases. In addition to this, we also have preliminary plans to include a suicide system (tentative candidate: Toxin-Antitoxin systems) inside the plasmid as a failsafe in the event that horizontal gene transfer occurs and causes unintended ecological impacts. [Top]

VI. BioBrick Construction - Bcr Multidrug Efflux Pump

Bcr is a type of multidrug efflux pump, which are integral membrane proteins that utilize cellular energy to actively extrude antibiotics or biocides actively out of the cell. It belongs to the major facilitator superfamily (MFS), and is known to contribute to multidrug resistance in E. coli.

Under normal growth conditions, a large number of drug efflux pumps are thought to be weakly expressed. In particular, literature documents Bcr to confer varying degrees of resistance to several kinds of antibiotics when overexpressed; including bicyclomycin (selection-capable), tetracycline (8-fold MIC increase*), and kanamycin (4-fold MIC increase*).

In our iGEM project, we planned to construct a BioBrick with a pLac promoter (BBa_R0010) driving the expression of Bcr. The reason behind this is to take advantage of the additive effect of IPTG on pLac activation. We hope that by varying the concentration of IPTG, we can control the level of expression of Bcr and thus manipulate the mutant E. coli’s MIC to certain antibiotics. However, due to limited time, we did not manage to finish this construct. Yet other iGEM teams may still obtain our coding sequence for Bcr BBa_K524100) and attempt their own tests. [Top]

*: compared with wild type

[1] http://www.nature.com/nature/journal/v467/n7311/abs/nature09354.html
[2] http://www.scielo.br/pdf/gmb/v26n2/a17v26n2.pdf