Bacteria are exposed to a highly complex and variable environment. To optimize their fitness they need to properly allocate their limited resources to several tasks. For example, E. coli needs to decide which carbon catabolic genes it expresses at what level when multiple sugars are present. It is well known that E. coli preferentially consumes glucose, which represses the expression of most other sugar specific enzymes. Here, we combine theory and experiment to investigate how E. coli decides between multiple non-glucose substrates. We find a hierarchical order where the sugar supporting the highest growth rate reduces the expression of enzymes catabolizing other sugars. The hierarchy is determined by the sensitivity of the sugar specific genes to the master regulator CRP. In a simple mathematical model we describe carbon catabolic gene expression under integral feedback control via CRP. This control not only defines the sugar hierarchy, but also ensures that catabolic enzyme expression level is precisely tuned to what is required for optimal growth. Gene expression control remains optimal even when competitive substrates or mutations impair the activity of catabolic enzymes. We show this experimentally for the lac operon. Our findings may help to design bacterial strains for optimal use in biotechnology and understand how biological systems can perform optimally in a diverse set of conditions.
