Team:Paris Bettencourt/Atb exp


Team IGEM Paris 2011

Antibiotics cross-resistance experiments

A non-hereditary exchange of antibiotic resistance between two Bacillus subtilis strains: one example of a situation that could involve nanotubes!

In one of the main experiments of the founding paper [1], one strain is resistant to chloramphenicol(CmR) and the other resistant to lincomycin (LinR). When cultivated separately in the presence of the antibiotic corresponding to their respective resistances, each strain can survive. The strains cease to grow in the presence of the other antibiotic (CmR strains on lincomycin or LinR strains on Chloramphenicol). However, when the two strains were mixed and cultivated on plates containing both antibiotics, both strains continued to grow, suggesting that bacteria of the two strains had shared antibiotic resistance proteins. The transfer of resistance was only temporary.

Indeed, when each strain was again plated in isolation, it no longer displayed double resistance. In addition, the same experiment was done with chloramphenicol resistant strain and kanamycin resistant strain. In line with previous results, only cells in the mixed population were able to grow on the antibiotic plate containing both Cm and Kan. Genotypic analysis of the surviving cells revealed that they were exclusively KanR, implying that the kanamycin resistant cells received in a non-heredetery manner the chloramphenicol acetyl transferase protein (or its coding RNA). This was expanded to thousands of colonies and revealed that indeed CmR cells rarely survive. This assay enables delineation between "donor" (CmR) and "recipient"(KanR) strains, providing an approach to follow the directionality of molecular exchange.

Thereby, questions around the nature of these interactions can be raised. Which molecules diffuse? Why do we observe a unidirectional transfer of resistance? How can such a small quantity of cells and molecules allow the other cells to survive?

In order to elucidate some of these questions, we focused on the second experiment involving CmR and KanR strains and proposed an alternative explanation for this phenomenon.

We first re-did the main experiment: mixing two strains of different antibiotic resistances together on plates containing both antibiotics (Cm and Kan). In parallel, we checked the kinetics of bacterial growth on separate plates, confirming that the two strains have the same division rate. Analysis of the mixed solid supported culture (overnight incubation at 37°C) indicated that CmR strains did not survive and only the KanR strain were able to grow. Even if there were rare survivors, they could not justify the significant growth of the KanR strain. This leaves us to conjecture that the CmR strain could locally modify the medium.

The Guardian Angel theory

They were here, they are perhaps still here and they watch over the other cells so that they can grow, but we can’t see them! What kind of creatures could they be?

In order to study the crime scene, we washed with LB liquid all bacteria that had grown in the mixed spot by up and downpipetting . The wash was then collected and re-spotted on a separate spot, but on the same plates containing both antibiotics. Simultaneously, we spotted on the control spot (spots on which the CmR strains were now absent) new KanR cells. Much to our surprise, new KanR cells grew in these conditions. These results suggest that CmR cells can affect their environment and enable strains that are not resistant to chloramphenicol to survive in originally hostile medium. The death of the CmR bacteria could bequest their resistance enzymes - in this case chloramphenicol acetyl transferase (cat).

The Schrödinger cat

The best way to see if our bacteria are alive or not is to open the box of "Schrödinger chloramphenicol acetyl transferase (cat)". For this experiment, the two strains were spotted together but separated with a filter on Cm plate. KanR strains grow significantly by being in indirect contact with Cm. Furthermore, it can also grow on the LB-agar medium were used to live CmR cells on a filter (cf figure). These results strongly indicate that the resistance exchange is not due to nanotubes, and suggest a detoxification of the medium by CmR cells.

To investigate the nature of the resistance diffusion, we did again the first experiment with E.coli strains harboring chromosomal resistance to kanamycin and chloramphenicol (cf Notebook). We also obtained the same results as with the B.subtillis strains: mixed population were able to grow on plate containing both antibiotics. So we can rule out the possibility of non-conjugative plasmid exchange. Others hypotheses can explain the survival of these cells.

  • Living bacteria could “pump” enough antibiotic molecules to detoxify the environment. In this case, CmR strain’s time of survival to kanamycin supposed to be enough to afford this phenomenon.
  • The other hypothesis is that by dying, CmR cells spill out active resistance proteins in the medium. The action mode conditions of chloramphenicol acetyl transferase (cat) supports this possibility. Optimal pH of acetylation is 7.5, which is in the range of LB Agar pH we work with. Thus, all cells in the surroundings of kanamycin killed chloramphenicol-resistant cells profit from the ex vivo activity of cat protein .


Our results suggest that CmR cells are first killed by kanamycin, which is a bactericide antibiotic. When dying, they offer resistance to chloramphenicol to KanR cells, who are not totally destroyed by chloramphenicol, which is a bacteriostatic antibiotic.

Results of experiment with CmR and LinR strains may be also explain by the same succession of events. Both antibiotics are bacteriostatic, and are not enough efficient to kill all bacteria, and both resistance protein are poured out in the same time in the medium.