Abstract
The utilization of microbial communities instead of single strain populations in biotechnological processes has been under investigation over the last decades. While some natural communities have already for example been employed for the production of fermented food the establishment of synthetic microbial communities for the production of valuable compounds or bioremediation is still under investigation. One of the most important aspects that has to be considered to maintain microbial communities for biotechnological applications stably is to be able to tune growth and growth rate of individual populations to prevent outgrowth and unbalancing of the community. To address this challenge, we aimed to develop a multicellular network in two engineered Escherichia coli populations with which the growth of one population is controlled in a density-dependent manner by the respective other population and vice versa. In Synthetic Biology standardization and modularization represent a key aspect in order to be able to predict the performance of circuits and networks. For this to be possible all the utilized parts needed to be characterized. To get an understanding of gene expression strength as well as the dynamics over time, we characterized the promoters we implemented in our synthetic network. We expressed representatively a fluorescent protein from the promoters and thus monitored the expression of GFP under different conditions. To regulate the population growth in our synthetic microbial community we took three different approaches and investigated their characteristics and suitability to be implemented: 1. Antibiotic resistances 2. RNA polymerase and 3. Contact-dependent inhibition. For the antibiotic resistance, we utilized the chloramphenicol acetyltransferase as a mechanism to regulate population growth. The expression of the corresponding cat gene was placed under the control of a quorum sensing promoter (PluxI or PlasI). While the cognate transcription factor (LuxR or LasR) was expressed in the same cell the cognate communication molecule (acyl-homoserine lactone: AHL) (AHL-C6 or AHL-C12) was produced in cells of the other population. Thus the resistance to chloramphenicol (CAM) should only be conferred by high cell density of the respective other population when enough cognate AHL is produced to induce the gene expression of the resistance cassette. However, our data showed that the cells grew in the presence of CAM, even when no AHL or transcription factor was available. In consecutive rounds of design improvements plasmid backbones, promoters, and also the antibiotic resistance cassette were exchanged for alternative variants or versions but did not result in a system exhibiting the desired performance. This led us to conclude that the implementation of antibiotic resistance cassettes could not achieve the regulation of population growth in bacteria. The RNA polymerase approach was performed by controlling the expression of the rpoBC operon which encodes the β and β’ subunit of the RNA polymerase. Additionally, we replaced the Las quorum sensing system with the Rpa system (RpaR, RpaI, AHL-pC, and Prpa*). First experiments showed that the cells did not exhibit any differences in growth or growth rate in comparison to the wild-type. Therefore, we improved the design by tagging each component with an LVA tag to enhance protein degradation and thus shortening the protein half-life; all genes were also chromosomally integrated to reduce gene copy number. The results showed that the growth and growth rate were reduced; nevertheless, we did not obtain a quorum sensing-dependent growth switch. To conclude, even though the overall aim could not be achieved so far we got valuable insights into quorum sensing-regulated RNA polymerase expression. Implementation of further improvements is promising, and we identified measures which could be taken next in order to design and build a functional RNA polymerase-based system. Additionally, we explored the potential of contact-dependent growth inhibition (CDI) to be utilized in a similar system as previously described. The CDI system is a touch-dependent and receptor-mediated toxin delivery system which can inhibit neighboring cells. We characterized two CDI systems from two E. coli strains, EC93 and EC869, under different culture and induction conditions in coculture with either CDI negative cells or two CDI strains. We observed that CDI is very effective in inhibiting growth however the cells carrying the pore-forming protein toxin exhibited a stronger toxin effect than DNase toxin-carrying cells. We can report that induction type and levels, inoculation ratios, and selected toxins influence the dynamics and changes in community composition over time. We also coupled the expression of CDI to quorum sensing resulting in quite controlled and stable cocultures over the course of 26 hours. Valuable characteristics were determined which lay the foundation for developing downstream applications harnessing these contact-dependent inhibition properties in microbial communities. To conclude, CDI is a very interesting and promising tool for growth regulation that has to be further explored and adapted to be employed in a growth regulating application. To summarize, for all three mechanisms, the antibiotic resistance cassettes, the RNA polymerase, and contact-dependent inhibition we explored their suitability to be employed to regulate population growth in a density-dependent manner. We obtained valuable data exhibiting the difficulties yet potentials of each system and how they can be further improved to obtain the desired systems performance.
