8
an allosterically regulated Ca
2+
-sensitive Kemp Eliminase was engineered by introducing a binding site and reactive groups for a
Kemp Eliminase reaction into the EF hand of calmodulin, while
preserving its natural propensity to undergo a conformational transition from compact to extended upon binding Ca
2+
[77]. Similarly,
the ligand specificity of the bacterial transcription factor LacI was
computationally reengineered to recognize fucose, gentiobiose,
lactitiol, and sucralose [78], while preserving the natural propensity of LacI to bind DNA in a ligand-dependent fashion. However,
preserving natural allosteric transitions while introducing new
ligand specificities is nontrivial, and in case of bacterial transcription factors additionally involved experimental screening and selection of a large library of mutant protein switches [78].
In contrast, predictably engineering the conformational transitions that underlie synthetic protein switches have so far met with
limited success. This particularly applies to integrated designs,
where allosteric changes are regulated through complex networks
of amino acids in the tertiary structure of a protein that are difficult
to recapitulate in a rational manner. In contrast, for modularly
organized protein switches with structurally distinct receptor, actuator, and AI-domains, the behavior of the connecting linkers can
be described with synthetic polymer models to assist balancing
steric strain in ligand-bound and unbound conformational states.
In one example, the worm-like chain (WLC) model was successfully applied to quantify the behavior of Gly-Ser-rich linkers connecting two FPs undergoing resonance energy transfer in a
Zn
2+
-specific protein sensor [79]. Yet, these models have so far
primarily been used to rationalize the behavior of a linker postexperimentally, but not engineer linkers a priori.
Beyond structure-guided protein engineering, the evolutionary
history of proteins provides a rich source of information that can
be computationally analyzed to derive useful functional and biophysical properties of proteins. Notably, next-generation sequencing technologies have generated an unprecedented wealth of
sequence data that provides a detailed snapshot on the evolution of
proteins and protein families. This data is increasingly mined and
analyzed using sophisticated computational algorithms to extract
valuable information on how the primary structure of a protein
correlates with key biophysical and functional properties.
In the simplest case, the consensus sequence of a protein can
highlight functionally and structurally important residues that are
conserved within a protein family [80]. Enriching proteins with
conserved consensus motifs has previously been shown to improve
their thermal and conformational stability that constitutes a critical
parameter in the development of recombinant proteins for many
biotechnological applications including therapeutic binding agents
2.3 Design
by Statistical
Sequence Analysis
Viktor Stein
an allosterically regulated Ca
2+
-sensitive Kemp Eliminase was engineered by introducing a binding site and reactive groups for a
Kemp Eliminase reaction into the EF hand of calmodulin, while
preserving its natural propensity to undergo a conformational transition from compact to extended upon binding Ca
2+
[77]. Similarly,
the ligand specificity of the bacterial transcription factor LacI was
computationally reengineered to recognize fucose, gentiobiose,
lactitiol, and sucralose [78], while preserving the natural propensity of LacI to bind DNA in a ligand-dependent fashion. However,
preserving natural allosteric transitions while introducing new
ligand specificities is nontrivial, and in case of bacterial transcription factors additionally involved experimental screening and selection of a large library of mutant protein switches [78].
In contrast, predictably engineering the conformational transitions that underlie synthetic protein switches have so far met with
limited success. This particularly applies to integrated designs,
where allosteric changes are regulated through complex networks
of amino acids in the tertiary structure of a protein that are difficult
to recapitulate in a rational manner. In contrast, for modularly
organized protein switches with structurally distinct receptor, actuator, and AI-domains, the behavior of the connecting linkers can
be described with synthetic polymer models to assist balancing
steric strain in ligand-bound and unbound conformational states.
In one example, the worm-like chain (WLC) model was successfully applied to quantify the behavior of Gly-Ser-rich linkers connecting two FPs undergoing resonance energy transfer in a
Zn
2+
-specific protein sensor [79]. Yet, these models have so far
primarily been used to rationalize the behavior of a linker postexperimentally, but not engineer linkers a priori.
Beyond structure-guided protein engineering, the evolutionary
history of proteins provides a rich source of information that can
be computationally analyzed to derive useful functional and biophysical properties of proteins. Notably, next-generation sequencing technologies have generated an unprecedented wealth of
sequence data that provides a detailed snapshot on the evolution of
proteins and protein families. This data is increasingly mined and
analyzed using sophisticated computational algorithms to extract
valuable information on how the primary structure of a protein
correlates with key biophysical and functional properties.
In the simplest case, the consensus sequence of a protein can
highlight functionally and structurally important residues that are
conserved within a protein family [80]. Enriching proteins with
conserved consensus motifs has previously been shown to improve
their thermal and conformational stability that constitutes a critical
parameter in the development of recombinant proteins for many
biotechnological applications including therapeutic binding agents
2.3 Design
by Statistical
Sequence Analysis
Viktor Stein