Metabolite biosensors are powerful tools for basic biological research, medical diagnostics, and biotechnological applications. However, a generalizable strategy for developing new metabolite biosensors when an existing sensor cannot be found in nature, is a persistent challenge. Furthermore, while transcription factor biosensors have the broadest range of applications, the pool of naturally occurring transcription factor biosensors is small. There is however, a wealth of metabolite binding proteins that can be found in nature, that bind many metabolites, but are unable to regulate transcription. Therefore, the primary objective of my thesis was to develop a methodology by which a metabolite binding protein could be converted into a metabolite responsive transcription factor. Toward this goal, two hypothesis and literature driven approaches were investigated in order to determine the feasibly of converting a model metabolite binding protein into a metabolite responsive transcription factor. The split protein (SP) strategy ultimately resulted in a functional metabolite responsive transcription factor by fusing the BCR-ABL1 zinc finger DNA-binding domain (ZFP) internally to maltose binding protein (MBP) at amino acid 316R. This initial demonstration validated the feasibility of the idea that a metabolite binding protein can be converted into a metabolite responsive transcription factor. Next, in order to investigate the generalizability of the SP conversion strategy, the biosensor engineering by random domain insertion (BERDI) method was developed to construct and test all possible insertions of the ZFP into MBP. Because the original biosensor developed by the SP strategy used a previously published split of MBP (316R) that resulted in an enzymatic biosensor, the BERDI method was developed so that reliance on this type of information would not be required for future biosensor development. In addition, the BERDI method is not specific to MBP as the metabolite binder, therefore the method can be used on any metabolite binding protein enabling this method to be as generalizable as possible. Using the BERDI method, three new splits of MBP were found to generate maltose responsive biosensor illustrating not only that the method can generate novel biosensors with no reliance of previously published information, but also that one metabolite binding protein can result in several functional biosensors. Finally, to apply the BERDI method to metabolite binding enzyme for a biotechnologically relevant metabolite, farnesyl pyrophosphate (FPP), a screening strain for the inducible overproduction of FPP was developed. This strain enables the same cell to inducible over produce FPP such that cells containing potential FPP responsive biosensors experience a change in internal FPP levels to enable screening for functional FPP responsive biosensors. This strain enables the BERDI method to be applied to FPP binding enzymes to continue the investigation into the generalizability of the SP biosensor conversion strategy. Overall, this work demonstrates the feasibility of a potentially transformative technology and lays the ground work for future investigations and applications of the BERDI method for biosensor engineering and development.

Last modified

• 04/18/2018

Creator

• Andrew Younger

DOI

• https://doi.org/10.21985/N24103

Subject

• Interdepartmental Biological Sciences (IBiS) Graduate Program

Keyword

• Bioengineering
• Transcription factor
• Synthetic biology
• Promoter engineering
• Protein engineering
• Metabolic engineering

Date created

• 2017-01-01

Resource type

• Dissertation

Rights statement

• In Copyright