PT  - JOURNAL ARTICLE
AU  - R. R. Cheng
AU  - E. Haglund
AU  - N. Tiee
AU  - F. Morcos
AU  - H. Levine
AU  - J. A. Adams
AU  - P. A. Jennings
AU  - J. N. Onuchic
TI  - Guiding the design of bacterial signaling interactions using a coevolutionary landscape
AID  - 10.1101/116947
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 116947
4099  - http://biorxiv.org/content/early/2017/06/02/116947.short
4100  - http://biorxiv.org/content/early/2017/06/02/116947.full
AB  - The selection of amino acid identities that encode new interactions between two-component signaling (TCS) proteins remains a significant challenge. Recent work constructed a co-evolutionary landscape that can be used to select mutations to maintain signal transfer interactions between partner TCS proteins without introducing signal transfer between non-partners (crosstalk). A bigger challenge is to introduce mutations between non-natural partner TCS proteins using the landscape to enhance, suppress, or have a neutral effect on their basal signal transfer rates. This study focuses on the selection of mutations to a response regulator (RR) from Bacilus subtilis and its effect on phosphotransfer with a histidine kinase (HK) from Escherichia Coli. Twelve single-point mutations of the RR protein are selected from the landscape and experimentally expressed to directly test the theoretical predictions on the effect of signal transfer. Differential Scanning Calorimetry is used to monitor any protein stability effects caused by the mutations, which could be detrimental to proper protein function. Of these proteins, seven mutants successfully perturb phosphoryl transfer activity in the computationally predicted manner between the TCS proteins. Furthermore, brute-force exhaustive mutagenesis approaches indicate that only 1% of mutations result in enhanced activity. In comparison, of the six mutations predicted to enhance phosphotransfer, two mutations exhibit a significant enhancement while two mutations are comparable to the wild-type. Thus co-evolutionary landscape theory offers significant improvement over traditional large-scale mutational studies in the efficiency of selecting mutations for protein engineering and design.