Team:Heidelberg/Co-Culture


Co-Culture





Preliminary co-culture experiments

Background

During our project we performed most of our assays as monocultures of our four bacterial strains Acinetobacter baylyi ADP1 (A. baylyi), Bacillus subtilis 168 (B. subtilis), Escherichia coli DH5ɑ (E. coli) and E. coli BL21. These strains are produced to be used in the laboratory environment and have a safety level 1. They should not be able to escape but their original ancestors can be found in the human gut environment as mutualists or with a more pathogenic effect [1]. In general, the gut microbiome consists of multiple bacteria species and even other organisms like protists or fungi [2], [3]. Cultivating only one species, even if the human body temperature of 37 °C is used, is not representative for the natural environment. However for our laboratory model the three laboratory strains cultured in LB medium would provide additional insights into the bacterial co-culture dynamics. We used LB medium because it allowed all species to grow, even though it is not the perfect medium for all of them (see Fig. 1).

Figure 1: Growth curves of different bacteria species in the lab. All bacteria were cultivated in 96-well plates at 37 °C from overnight cultures either undiluted (10^0) or diluted in triplicates. The optical density (OD) was measured at 600 nm every 30 minutes. The OD was corrected by subtracting the OD600 of pure medium. All three species grow in LB medium, but A. baylyi does not reach the OD of the other two species. (A) E. coli DH5ɑ in cultivated in LB medium. (B) B. subtilis 168 cultivated in LB medium. (C) A. baylyi ADP1 cultivated in LB medium. For this bacterium LB media does not provide optimal growth conditions. (D) A. baylyi ADP1 in R2A medium. The R2A medium is better suited for A. baylyi, but in order to culture it with the other two species, LB medium was used in the following experiments.

Our aim with the following co-culture experiments was to establish a highly simplified model that mimics bacteria co-cultures in the human gut. Co-cultivating two or more species is a better model environment for testing natural transformation than monocultures and co-cultures can be used to detect potential advantages or disadvantages for transformed bacteria.
For comparing different bacteria species or strains against each other, it was required to use transformed bacteria with fluorescence markers because otherwise they can’t be distinguished in liquid culture. We used for A. baylyi the mCherry protein, B. subtilis with integrated GFP and E. coli DH5α with mRFP1 or EGFP protein and E. coli BL21 with EGFP protein (see Fig. 2).

Figure 2: Examples of the used fluorescence proteins expressed by E.coli. Both red fluorescent proteins (mRFP1 and mCherry) can be easily seen by the human eye both on agar plates and in liquid culture. EGFP in small colonies and in liquid cultures is only detectable with the right wavelength, but on this plate a slight green shimmer is visible. (A) EGFP expressed by E. coli BL21. (B) E. coli BL21 expressing mCherry. (C) E. coli DH5α expressing mRFP1.

Experimental results

In the beginning we needed to make sure that we can distinguish between bacteria with EGFP and bacteria with a red fluorescent marker. For this reason we used two strains of E. coli DH5α transformed to express two distinct fluorescent proteins, EGFP with an emission at 535 nm and mRFP1 that can be detected at 635 nm. To test if the ratio can be distinguished by just measuring the emission at these two wavelengths, we pipetted both E. coli strains in different ratios. The results from our first co-culture was very promising (see Fig. 3). We were able to distinguish both the fluorescence proteins and the used ratios easily. Only the gain for GFP that we had to set on the TECAN menu before starting the measurement needed to be adjusted.

Figure 3: Distinguishing two fluorescent markers in E. coli DH5α. All monoculture bacteria strains and ratios between them have the same color in the left as well as the right plot for easier comparison. EGFP was abbreviated as GFP and mRFP1 as RFP. It can be seen that without mRFP1 (or EGFP respectively) no emission is detected. Furthermore, the bigger the portion of a fluorescent marker containing strain in a co-culture is, the higher the detected emission from this fluorescent protein is as well. (A) EGFP emission was measured at 535 nm with gain set to 50. (B) Emission of mRFP1 was measured at 635 nm with a gain of 60.

After the functionality of the plate reader for this application was tested, the interspecies co-cultures were started. B. subtilis with GFP, E. coli DH5α with EGFP and mRFP1 and A. baylyi with mCherry. Each species was paired with one other species, for that reason we needed A. baylyi and B. subtilis to express different fluorescent proteins and two E. coli strains with two different markers respectively. The experimental setup was the same as described above, only the gains were changed. The results can be seen in figure 4.
In most combinations the ratio determined how strong the emission value was. Of course the monoculture of the species had the strongest emission at the emission value of its marker. Furthermore a mixture with 80 % of this species had a stronger signal then one with only 50 or 20 %. These observations are consistent with what we would expect to see when both species grow side by side without interfering with each other, other than competing for nutrients.
But there are two exceptions from this observation. One of them is the unexpected increase of the A. baylyi monoculture in the assay with E. coli EGFP after 15 hours (see Fig. 4 A). The A. baylyi strain used only expressed mCherry and shouldn’t have an emission at 535 nm. We suspect that this increase is the result of cross contamination between the mixed cultures and A. baylyi alone and therefore the increase happened because E. coli with EGFP grew in that well.
The other important result is the co-culture of A. baylyi together with B. subtilis (see Fig. 4 E+F). In this combination A. baylyi doesn’t stand a chance against B. subtilis and even with 80 % A. baylyi, this bacteria species only barely manage to grow. For A. baylyi the 37 °C used in these co-culture assays isn’t optimal, but in monoculture and competing with E. coli (see Fig. 4 B) A. baylyi still grew. It seems like at the chosen conditions (LB media and 37 °C) B. subtilis outcompetes A. baylyi or maybe it produces substances that further decreases A. baylyi’s viability.

Figure 4: Co-cultures of E. coli, A. baylyi and B. subtilis. EGFP was abbreviated as GFP, mRFP1 as RFP and mCherry is written as mcherry. (A), (C) and (E) show emission at 535 nm, the wavelength EGFP emits at, measured with a gain of 70. (B), (D) and (F) show emission at 635 nm, the wavelength mRFP1 and mCherry emit at, measured with gain set to 60. (A+B) Co-culture of E. coli BL21 EGFP and A. baylyi ADP1 mCherry. C+D) Co-culture of B. subtilis with GFP and E. coli DH5α mRFP1. (E+F) Co-culture of A. baylyi ADP1 mCherry and B. subtilis with GFP.

Discussion

Overall these assays gave us confidence to use co-cultures in further experiments. Only the weak performance of A. baylyi is a problem that needs to be tackled. Another side observation was that E. coli EGFP contaminated A. baylyi monoculture which started to emit with green fluorescence. It can not took up the E. coli EGFP plasmid because the origin of replication was not suitable for A. baylyi.
We compared only two species at a time, because we did not have access to a fluorescence protein with an emission value differing from 535 nm and 635 nm. With additional fluorescence proteins and plate readers with more color filters, it would be possible to increase the number of comparable species and strains. Another approach for three species and only two markers that we couldn’t execute because of too short access to the TECAN would have been to use two species with markers and the third without one and comparing the results to the two marked species alone. The comparison would have looked like in table 1.

Table 1: Planned co-culture to analyse multi-cultures. The co-cultres on the left represent the measured co-culture. To analyse three strains wihtout a third fluorescence marker our results from the co-culture measurement can be compared to a multi-culture measurment, with one strain not containing a marker. The required multi-cultures are shown in the right column.

Measured co-culture Planned co-culture
E. coli BL21 EGFP + A. baylyi ADP1 mCherry E. coli BL21 EGFP + A. baylyi ADP1 mCherry + B. subtilis WT
B. subtilis with GFP + E. coli DH5α mRFP1 B. subtilis with GFP + E. coli DH5α mRFP1 + A. baylyi WT
A. baylyi ADP1 mCherry + B. subtilis with GFP A. baylyi ADP1 mCherry + B. subtilis with GFP + E. coli WT
Even without further comparisons between three species at a time, we used the information we learned here to design our further experiments:
With a selective advantage we tried to manipulate a strain’s ability to outcompete other strains in a co-culture.
Furthermore, we adjusted the setup to combine all three species to test natural transformation in co-cultures. Here we decrease the temperature to 30 °C to support A. baylyi and hinder B. subtilis.

References

[1] Tchaptchet, S., & Hansen, J. (2011). The Yin and Yang of host-commensal mutualism. Gut microbes, 2(6), 347–352. https://doi.org/10.4161/gmic.19089

[2] Laforest-Lapointe, I., & Arrieta, M. C. (2018). Microbial Eukaryotes: a Missing Link in Gut Microbiome Studies. mSystems, 3(2), e00201-17. https://doi.org/10.1128/mSystems.00201-17

[3] Loftus, M., Hassouneh, S. A., & Yooseph, S. (2021). Bacterial associations in the healthy human gut microbiome across populations. Scientific reports, 11(1), 2828. https://doi.org/10.1038/s41598-021-82449-0