The well-studied regulatory system for competence development in B. subtilis provided a genetic set-up based on quorum sensing that we used as a proof of principle (Figure 1). [4] B. subtilis constantly secretes the ComX pheromone, a 9- to 10-amino acid oligopeptide, as a signaling molecule (a). By rising cell-density, the ComX-concentration in the surrounding medium increases until it reaches a threshold and activates ComP, a membrane-spanning protein kinase (b). The kinase reacts to the accumulation of ComX by phosphorylating the response regulator ComA (c) which then works as a transcription factor by binding to several promoters and enhancing their activity (d) [5,6].

The most important promoter regulated by ComA and involved in competence development is the promoter of the srfA operon. This operon contains not only genes for the production of the antibiotic surfactin (e) but also for ComS (f), another small peptide that prevents the degradation of the autoregulated transcription factor ComK (g). ComK activates expression of more than 100 genes, including comG, that is part of the transformation machinery (h) [7,8]. Consequently, B. subtilis can take up DNA from the environment either as direct nutrient source or incorporate the DNA via homologues recombination into its own genome.

Based on this genetic background, we wanted to set up a simple experiment that shows communication between two B. subtilis strains. The basic idea is comparable to a radio broadcast: you need a transmitter that sends out the radio waves and a receiver that receives the waves and converts them into an acoustical signal. Transferring this idea over to our project means that we need a sender strain (SeSt) that secretes the ComX pheromone, and a receiver strain (ReSt) that reacts to the ComX-stimulation with an easy-detectable output (Figure 2).

In our setup, we fused target promoters of the competence machinery of B. subtilis to the lux operon and measured luminescence output. Additionally, the ReSt needed to be comX-deficient to prevent autoinduction. We encapsulated the ReSt within Peptidosomes and incubated it together with the SeSt (either wild type or a ComX-overproducing strain) in the surrounding medium. Thus, we expected ComX that is produced outside the Peptidosme to diffuse through the membrane and stimulate the ReSt. As a result, we should obtain a glowing Peptidosome mediated by the lux operon, which is only present in the ReSt.

Since the ComX pheromone itself does not act as a transcription factor we examined promoters that are either under the control of the response regulator ComA (P_srfA, P_rapA & P_rapF) or regulated by the competence transcription factor ComK (P_comK and P_comG). Surprisingly P_rapA and P_rapF seem to be constitutive as well as P_comK mut that is regulated by ComK itself. The unexpected behaviour of P_comK mut could be a consequence of the single nucleotide exchange that was necessary to remove a restriction site since small changes of a promoter sequence can influence the binding of regulatory elements. After an initial decrease of activity P_comG exhibits a weak increase of activity during stationary phase. However, for our purpose the most promising promoter is P_srfA. Although it possesses a high basal activity in the wild type (WT) it reveals a tenfold increase of expression during the transition to stationary phase. In the associated comX-knockout strain (ΔcomX) the promoter activity only shows basal activity. This clearly demonstrates the ComX-dependent expression of the srfA operon.

We continued by examining the influence of different media on the activity of P_srfA. This step is important because competence development is not only dependent on quorum sensing but also on nutritional conditions. As expected MNGE, the transformation medium of B. subtilis, turned out to be the most convenient medium for our purposes (Figure 8). In rich medium (LB) we observed the lowest promoter activity for the knockout strain as well as the lowest increase of activity in wild type. However, in both minimal media (MCSE and MNGE), we could observe a higher increase of luminescence, although there is a temporary decline of activity after 3 hours of incubation in MCSE-medium. Therefore, we decided to continue using MNGE-medium for all following experiments.

After screening of promoter activity and media-dependence of P_srfA we examined the influence of ComX-overproduction via induction with xylose (Figure 9). In strains with an additional copy of comX (WT), an induction with xylose (+ xylose) did not lead to higher luminescence but to an earlier increase in comparison to the non-induced WT (+ dH2O). In knockout strains (∆comX), we could compensate the comX-deficiency via an induction of ComX-production. Surprisingly in those cultures the luminescence increased even earlier as in WT with an additional induction of the ComX-production. The increase of luminescence in the comX-deficient strain without xylose-induction (∆comX + dH2O) served as negative control. In conclusion, we could demonstrate an impact of ComX-overproduction.

Via plate reader assays, we have already proven the functionality of our genetic constructs combined in one single strain. So, the next step was to demonstrate that the system also works if the constructs are separated in two different strains. Therefore, we continued with co-cultures of SeSt and ReSt using cell culture inserts (ThinCerts^TM). Thereby, we could finally show that the SeSt stimulates the ReSt, containing a P_srfA-luxABCDE fusion, by using luminescence as output (Figure 10). In the control, without SeSt, the comX-deficient ReSt exhibited only a weak luminescence signal because of the promoter’s basal activity. Hence, our system, consisting of SeSt and ReSt, is qualified as a proof of principle for communication between two distinct strains.

In Figure 11 A, B and C illustrate the experimental set-up whereas D, E and F display the results of the well scan analysis. While areas with luminescence less or equal the maximal luminescence measured for SeSt (≤ 40 RLU) are blue-colored, areas with luminescence above SeSt level (≥ 40 RLU) are pink-colored. Therefore, we can show that the increase of luminescence is restricted to the Peptidosome (Figure 11, E). The mean luminesescence determined for the encapsulated ReSt is 90 RLU. There are few areas with values that are negligibly above the threshold (41 to 62 RLU) causing pink-colored squares located outside the expected position of the Peptidosome. This is a consequence of the chosen threshold and provides no evidence for cells leaving the Peptidosome.