6
has been made in the computational design of protein structures,
protein assemblies, protein-protein interactions, ligand and substrate specificities, as well as catalytic mechanisms [17–21]-yet,
progress in the computational design of synthetic protein switches
with tailored response functions has been limited. Notably, synthetic protein switches are dynamic entities and undergo conformational transitions that are critically important for their function,
yet challenging to analyze and even more challenging to predict,
control, and engineer in a systematic fashion. The majority of synthetic protein switches have thus been designed based on an intuitive molecular understanding of protein structure and function
while computational strategies increasingly assist in the rational
optimization of key functional or biophysical properties.
The protein database (PDB) features over 120,000 solved protein
structures that can be exploited for the structure-guided engineering of protein switches by (semi-)rationally recombining binding
receptors with enzymes, fluorescent, or bioluminescent proteins.
Protein structures are readily accessible through structural visualization programs such as PyMol (DeLano WL, 2002 The PyMOL
Molecular Graphics System) that provide an indispensable design
aid. For instance, in domain insertion strategies, an allosteric
receptor is typically inserted into surface exposed loop regions
such that ligand-induced conformational changes are efficiently
transmitted to the actuator modulating its function. In this way,
synthetic protein switches and sensors have been engineered based
on GFP [22–24], β-lactamase [25–27], tyrosine protein kinases
[28–30], xylanase [31], and PQQ-dependent glucose dehydrogenase (GDH) [32]. Similarly, alternative frame folding relies on a
thorough structural analysis to identify, duplicate, and modify
structural elements that are important for the binding or catalytic
function of a synthetic protein switch [33–35]. Structurally related
to synthetic protein switches engineered by domain insertion are
split protein complementation sensors that reassemble into a functional protein upon induced localization of the two protein halves.
Here, structural intuition frequently guides the choice of the split
sites that separate a protein into two structurally well-defined
subdomains. In this way, it has been possible to reengineer a number of split luciferases to report on intracellular signaling events
[36–39], split tobacco etch virus (TEV) proteases to sense and
actuate intracellular signaling function [40–43], split tyrosine protein kinases to actuate cellular signaling functions and screen for
drugs [44, 45], and split PQQ-dependent GDH as a universal
biosensor platform [46]. In addition, the construction of modularly organized protein switches based on structurally distinct allosteric receptors and actuators benefits from high-resolution
structural information as it provides clues about the position and
relative orientation of the N- and C-termini that assist in the
2.1 Design
by Molecular Intuition
Viktor Stein
has been made in the computational design of protein structures,
protein assemblies, protein-protein interactions, ligand and substrate specificities, as well as catalytic mechanisms [17–21]-yet,
progress in the computational design of synthetic protein switches
with tailored response functions has been limited. Notably, synthetic protein switches are dynamic entities and undergo conformational transitions that are critically important for their function,
yet challenging to analyze and even more challenging to predict,
control, and engineer in a systematic fashion. The majority of synthetic protein switches have thus been designed based on an intuitive molecular understanding of protein structure and function
while computational strategies increasingly assist in the rational
optimization of key functional or biophysical properties.
The protein database (PDB) features over 120,000 solved protein
structures that can be exploited for the structure-guided engineering of protein switches by (semi-)rationally recombining binding
receptors with enzymes, fluorescent, or bioluminescent proteins.
Protein structures are readily accessible through structural visualization programs such as PyMol (DeLano WL, 2002 The PyMOL
Molecular Graphics System) that provide an indispensable design
aid. For instance, in domain insertion strategies, an allosteric
receptor is typically inserted into surface exposed loop regions
such that ligand-induced conformational changes are efficiently
transmitted to the actuator modulating its function. In this way,
synthetic protein switches and sensors have been engineered based
on GFP [22–24], β-lactamase [25–27], tyrosine protein kinases
[28–30], xylanase [31], and PQQ-dependent glucose dehydrogenase (GDH) [32]. Similarly, alternative frame folding relies on a
thorough structural analysis to identify, duplicate, and modify
structural elements that are important for the binding or catalytic
function of a synthetic protein switch [33–35]. Structurally related
to synthetic protein switches engineered by domain insertion are
split protein complementation sensors that reassemble into a functional protein upon induced localization of the two protein halves.
Here, structural intuition frequently guides the choice of the split
sites that separate a protein into two structurally well-defined
subdomains. In this way, it has been possible to reengineer a number of split luciferases to report on intracellular signaling events
[36–39], split tobacco etch virus (TEV) proteases to sense and
actuate intracellular signaling function [40–43], split tyrosine protein kinases to actuate cellular signaling functions and screen for
drugs [44, 45], and split PQQ-dependent GDH as a universal
biosensor platform [46]. In addition, the construction of modularly organized protein switches based on structurally distinct allosteric receptors and actuators benefits from high-resolution
structural information as it provides clues about the position and
relative orientation of the N- and C-termini that assist in the
2.1 Design
by Molecular Intuition
Viktor Stein