Design of a methotrexate-controlled chemical dimerization system and its use in bio-electronic devices

Zhong Guo,Caryn Hepburn,Patricia Walden,Oleh Smutok,Tom Nebl,Janet Newman,Thomas S Peat,Richard J Suderman,Evgeny Katz,Jacobus P J Ungerer,Kirill Alexandrov,Jake Parker,Artem Melman,Wayne A Johnston

doi:10.1038/s41467-021-27184-w

Abstract

Natural evolution produced polypeptides that selectively recognize chemical entities and their polymers, ranging from ions to proteins and nucleic acids. Such selective interactions serve as entry points to biological signaling and metabolic pathways. The ability to engineer artificial versions of such entry points is a key goal of synthetic biology, bioengineering and bioelectronics. We set out to map the optimal strategy for developing artificial small molecule:protein complexes that function as chemically induced dimerization (CID) systems. Using several starting points, we evolved CID systems controlled by a therapeutic drug methotrexate. Biophysical and structural analysis of methotrexate-controlled CID system reveals the critical role played by drug-induced conformational change in ligand-controlled protein complex assembly. We demonstrate utility of the developed CID by constructing electrochemical biosensors of methotrexate that enable quantification of methotrexate in human serum. Furthermore, using the methotrexate and functionally related biosensor of rapamycin we developed a multiplexed bioelectronic system that can perform repeated measurements of multiple analytes. The presented results open the door for construction of genetically encoded signaling systems for use in bioelectronics and diagnostics, as well as metabolic and signaling network engineering.

Highlights

Natural evolution produced polypeptides that selectively recognize chemical entities and their polymers, ranging from ions to proteins and nucleic acids
In the current work we explored the design parameters for construction of small molecule-mediated protein dimerization systems and the resulting protein biosensors and bio-electrodes
We started our study on the assumption that it is possible to develop artificial protein complexes that are held together by an interface composed of the ligand and structural elements of both proteins

Summary

Introduction

Natural evolution produced polypeptides that selectively recognize chemical entities and their polymers, ranging from ions to proteins and nucleic acids. Efforts to engineer allosteric systems broadly fall into two categories: (a) protein designs where new or re-designed allosteric sites are coupled to the active sites through long range intradomain interactions[2], and (b) utilization of ligand-binding domains that pass their conformational changes onto reporter domains thereby controlling their activity[4]. Induced dimerization (CID) and protein biosensor development efforts have both approached this challenge by either creating semisynthetic systems where the signal is generated through competitive displacement of modified ligand, or through computational and selection-based engineering of genuine high affinity ternary complexes[10,11,12]. This is not expected to occur in the latter case where “anchor” domain:ligand affinity and specificity will determine the behavior of the ternary complex

Methods

Results

Conclusion