Determining the quantitative link between protein function and cellular fitness can be challenging as even the modestly sized genome of Escherichia coli is comprised of thousands of genes. Using an appropriate survey of fitness across a range of selective conditions, we can reduce the complexity of this system by tightly linking cellular fitness to the function of one protein essential for growth within that selective environment. Several model proteins have been studied in this fashion whereby in vitro protein parameters are used to predict cellular fitness as a function of selection strength. Underlying this approach, however, is the idea that the reverse relationship is also true: analysis of cellular fitness can be used to predict protein physicochemical properties.
In this study I present a physiological model that uses cellular fitness as a proxy to predict the biochemical properties of the tetracycline efflux pump, TetB, and a family of strategically chosen single amino acid variants. TetB is a member of the Major Facilitator Superfamily (MFS) of transporters which have a conserved protein fold and for which we have a general understanding of how protein structure relates to function. We first performed growth rate analysis on our host strain without tet(B) at a wide range of drug concentrations to obtain global parameters that describe the baseline response of our cellular system. Growth analysis was also performed on strains expressing a chromosomal copy of tet(B) or variant allowing for a quantitative measurement of the fitness effects produced by TetB. Using both sets of fitness data and in vivo protein concentration, our model was able to predict physicochemical pump parameters relating to substrate binding affinity and pumping efficiency for TetB and variants which match the current knowledge of how MFS transporter structure influences function. Taken together, this study shows that cellular fitness in strong selective conditions can be used to characterize efflux pumps, a class of proteins which are classically challenging to characterize using classical in vitro biochemistry techniques. Additionally, this analysis opens up the possibility of characterizing protein libraries from high-throughput growth rate assays.