9
[81–83] or enzymes for large-scale, industrial biosynthesis [84, 85].
It is worth noting that the consensus sequence of a protein does not
yield a true protein sequence, but an averaged one which neglects
that individual mutations are subject to epistatic effects [86]. This
means, depending on their context, combination of mutations can
have synergistic, neutral, or detrimental effects on a specific structural, biophysical, or function property. Considering this correlation is lost in the consensus sequence, the resulting proteins are not
necessarily functional and, thus, frequently have to be correlated
with additional sequence, biochemical, biophysical, or structural
information to yield proteins with the desired properties.
In contrast to the consensus sequence approach, ancestral gene
resurrection (AGR) aims to identify the true sequence of a primordial protein [87, 88]. This approach is unique in that it allows to
resurrect and experimentally study extinct proteins. Notably, from
a protein engineer’s perspective, ancestrally resurrected proteins
display a number of superior properties over their contemporary
counterparts. This includes superior folding, improved thermodynamic stability [89–91], and greater levels of substrate promiscuity
[89, 92], which, in the context of engineering synthetic proteins,
has already been exploited to reengineer the ligand specificity of
allosteric binding receptors [92]. Similar to the consensus sequence
approach, the evolutionary tree of a protein family is retraced based
on multiple sequence alignments and different statistical methods.
These include maximum likelihood, maximum parsimony, or
Bayesian reconstruction to calculate the posterior probability of a
protein sequence at every evolutionary branch point. While the
specific evolutionary ancestral resurrection algorithm is frequently
of debate—especially, if the true ancestral sequence of a protein is
to be determined in the context of evolutionary studies-this is a
lesser concern in protein engineering as long as the resurrected
protein sequences yield improved functional or biophysical
properties. For instance, AGR has been employed to improve the
thermodynamic and folding efficiency of l-arginine-specific periplasmic-binding proteins (PBPs). This turned out critical for their
efficient recombination with FPs to engineer l-arginine- specific
FRET sensors [90] and also facilitated their subsequent reengineering into l-glutamine-specific FRET sensors [92].
Finally, statistical coupling analysis (SCA) has been successfully
applied to identify co-evolving networks of residues that are distant
in primary, but continuous in tertiary structure highlighting 3D
hotspots that are functionally coupled in a protein [93, 94].
Notably, recombining AsLov2 and PDZ receptor domains with
dihydrofolate reductase (DHFR) via computationally predicted
allosteric hotspots yielded a regulated enzyme that transduces
light- and ligand-induced conformational transitions from the
receptor to the actuator [95].
Engineering Synthetic Protein Switches
[81–83] or enzymes for large-scale, industrial biosynthesis [84, 85].
It is worth noting that the consensus sequence of a protein does not
yield a true protein sequence, but an averaged one which neglects
that individual mutations are subject to epistatic effects [86]. This
means, depending on their context, combination of mutations can
have synergistic, neutral, or detrimental effects on a specific structural, biophysical, or function property. Considering this correlation is lost in the consensus sequence, the resulting proteins are not
necessarily functional and, thus, frequently have to be correlated
with additional sequence, biochemical, biophysical, or structural
information to yield proteins with the desired properties.
In contrast to the consensus sequence approach, ancestral gene
resurrection (AGR) aims to identify the true sequence of a primordial protein [87, 88]. This approach is unique in that it allows to
resurrect and experimentally study extinct proteins. Notably, from
a protein engineer’s perspective, ancestrally resurrected proteins
display a number of superior properties over their contemporary
counterparts. This includes superior folding, improved thermodynamic stability [89–91], and greater levels of substrate promiscuity
[89, 92], which, in the context of engineering synthetic proteins,
has already been exploited to reengineer the ligand specificity of
allosteric binding receptors [92]. Similar to the consensus sequence
approach, the evolutionary tree of a protein family is retraced based
on multiple sequence alignments and different statistical methods.
These include maximum likelihood, maximum parsimony, or
Bayesian reconstruction to calculate the posterior probability of a
protein sequence at every evolutionary branch point. While the
specific evolutionary ancestral resurrection algorithm is frequently
of debate—especially, if the true ancestral sequence of a protein is
to be determined in the context of evolutionary studies-this is a
lesser concern in protein engineering as long as the resurrected
protein sequences yield improved functional or biophysical
properties. For instance, AGR has been employed to improve the
thermodynamic and folding efficiency of l-arginine-specific periplasmic-binding proteins (PBPs). This turned out critical for their
efficient recombination with FPs to engineer l-arginine- specific
FRET sensors [90] and also facilitated their subsequent reengineering into l-glutamine-specific FRET sensors [92].
Finally, statistical coupling analysis (SCA) has been successfully
applied to identify co-evolving networks of residues that are distant
in primary, but continuous in tertiary structure highlighting 3D
hotspots that are functionally coupled in a protein [93, 94].
Notably, recombining AsLov2 and PDZ receptor domains with
dihydrofolate reductase (DHFR) via computationally predicted
allosteric hotspots yielded a regulated enzyme that transduces
light- and ligand-induced conformational transitions from the
receptor to the actuator [95].
Engineering Synthetic Protein Switches