(143c) High-Throughput Screening Of Combinatorial Regulatory Protein Libraries for Isolation of Novel Endogenous Molecular Biosensors

HIGH-THROUGHPUT
SCREENING OF COMBINATORIAL REGULATORY PROTEIN LIBRARIES FOR ISOLATION OF NOVEL
ENDOGENOUS MOLECULAR BIOSENSORS

Christopher Frei, Joseph Gredell,Shuai Qian, Patrick C. Cirino

University of Houston, S294
Engineering Bldg 1, Houston, TX, 77204-4004, USA

T: 832-842-6306, csfrei@uh.edu

Endogenous molecular biosensors based on effector-dependent
transcriptional regulatory proteins are emerging and powerful biomolecular
tools for specific, sensitive, and high-throughput detection/monitoring of
intracellular molecules [1].  The naturally
occurring proteins couple molecular recognition with changes in gene expression. 
To direct the evolution of molecular recognition toward compounds of interest,
we construct large libraries of regulatory protein variants and screen for
ligand responsiveness using a GFP reporter system[2].  These customized molecular biosensors
enable high-throughput screening of combinatorial biosynthesis libraries to
improve enzymatic or microbial production of metabolites [3, 4]. 

Our group has shown success engineering AraC, a well-studied
native E. coli dual-regulatory protein.  By generating multiple AraC
saturation mutagenesis libraries and using positive and negative FACS-based
screening, we have isolated variants with altered specificity toward D-arabinose
(D-ara), mevalonate (mev), and triacetic acid lactone (TAL)[2-4].  We then explored the limits of engineering
AraC for altered molecule recognition.  We sought variants that respond to many
compounds including p-coumaric acid (pCA), theophylline, propionic acid,
vanillin, malonic acid, levulinic acid, and several others. In the majority of
cases AraC variants remaining after sorting showed minimal ligand
responsiveness, although a notable exception was pCA-responsive variants.  Improvements
in biosensor properties of variants isolated from the initial saturation
mutagenesis library screening (e.g. reduced leaky expression and enhanced
sensitivity) were also sought by further directed evolution.  Properties of
select variants and mutation patterns that correlate with these properties will
be described.

The isolation of responsive clones is both tedious and not
guaranteed, and to streamline the process we aimed to optimize multiple
parameters in the high-throughput screen.  First, schemes with different
combinations of positive and negative sorting were tested for a subset of
target ligands, and the responses of the populations and individual variants
present at the sorting endpoints were compared.  The influence of reporter gene
copy number and copy number variability on sorting results was also studied.  As
expected, stable chromosomal integration of a single reporter gene reduced
expression variability.  Also, replacing GFP with a variant having attenuated
stability reduces the signal from leaky AraC variants, amplifies the on/off
expression ratio, and could improve the accuracy of the positive sorting step.  

Finally, high-throughput sequencing of individual gene populations
surviving different stages of sorting was also performed to develop a more
comprehensive understanding of the sequence evolution of specific clones.  This
sequencing study and the resulting insights into future screening and AraC
protein library design will be discussed.

1.            Gredell, J.A., C.S. Frei, and P.C. Cirino, Protein
and RNA engineering to customize microbial molecular reporting.
Biotechnology Journal, 2012. 7(4): p. 477-499.

2.            Tang, S.Y., H. Fazelinia, and P.C. Cirino, AraC
regulatory protein mutants with altered effector specificity. Journal of
the American Chemical Society, 2008. 130(15): p. 5267-5271.

3.            Tang, S.-Y., O. Akinterinwa, and P.C. Cirino, A
novel endogenous reporter of triacetic acid lactone (TAL) enables screening for
improved TAL production by E. coli. submitted.

4.            Tang, S.-Y. and P.C. Cirino, Design and
Application of a Mevalonate-Responsive Regulatory Protein. Angewandte
Chemie International Edition, 2011. 50(5): p. 1084-1086.