Skip to main content

Costs and optimality of gene expression levels in Escherichia coli

Periodic Reporting for period 1 - OPTEX (Costs and optimality of gene expression levels in Escherichia coli)

Reporting period: 2017-02-01 to 2019-01-31

Microbes have shaped the evolution of all of earths life forms. For most higher organisms microbes have important implications during health and disease. Currently human society is faced with the unprecedented spread and evolution of antibiotic-resistant pathogenic bacteria. Sustainable counter measures include better treatment strategies. To tackle this problem a detailed understanding of general bacterial growth properties is more urgent than ever. Growth and resistance are controlled by different genes and their expression level. Some gene products need to be available at exactly the right amount to perform cellular functions and maximize fitness also in the presence of antibiotics. For these reasons, it has been hypothesized that microbes evolved towards a state in which expression levels are optimally tuned to maximize fitness in a given environment. Here we set out to test this hypothesis on a whole-genome level to allow for a systematic and unbiased approach. We aimed to dissect general cellular cost-factors from gene specific effects which allow us to systematically uncover the mechanistic causes of the cost of protein overexpression. This is of importance to judge optimality. Optimality is a common objective in man-made systems and it will be intriguing to learn how natural selection has shaped microbes according to this concept. This can be exploited in antibiotic susceptibility assays.
During this project we have screened a whole-genome overexpression library in Escherichia coli to identify the cost of each gene. We have started with a sub-set of the whole collection, encompassing about 400 genes, and screened several different levels of expression using a robotic high-throughput setup for liquid bacterial cultures. Next, we aimed at extending the screen to the whole-genome level of 4000 genes. For that we used another robotic high-throughput setup for colony growth. The advantage is that a denser format is possible being able to screen the whole library at several, and not but a few, expression levels. This allows for the generation of high-resolution, so called, fitness functions for each gene which can be used to quantify the cost of overexpression. In the framework of this project we have found that natural regulation can result in an expression response which makes E. coli less susceptible to antibiotics. We have also found some genes, particularly from the membrane fraction, for which the exact expression level seems to be crucial to avoid detrimental effects. Those are probably under tight control and might have been optimized during the course of evolution. These results have so far been presented at several international conferences and during seminars in the systems biology community. A manuscript describing the expression-level dependent susceptibility is in preparation. Another manuscript describing overall effects of expression levels is in preparation. Both will be submitted for publication and in the event of getting published open access and visibility of the EU funding will be assured.
The work done in this project has been presented at a local graduate student meeting to inform students about possibilities to do scientific research. Science and in particular the importance of microbiology has been presented at the, so called, ‘Children’s University Cologne’ (KinderUniKöln) sparking interest in the natural sciences. During a conference a contact has been made with theoretical biologists to foster a collaboration aiming at modelling the cost-factor of expression. This might lead to a quantitative understanding of the limits to expression levels. The concepts developed within this project might be used to adapt antibiotic-susceptibility assays which are the next step on the path to better informed antibiotic treatment strategies.
optimalityofexpressionlevels.png