73
protein [13]. Notably, computational evolutionary simulations [14]
and the observation that ancestral proteins share similarities with
consensus proteins in terms of both sequence and thermostability
[8, 15, 16] suggest that the high thermostability of ancient proteins
might also originate from bias in the maximum- likelihood method
commonly used for APR, specifically, that more probable (often stabilizing) residues are chosen at every site, while less probable (often
destabilizing) residues are neglected [14]. The implication of this
potential bias for protein engineering is that APR may produce thermostable proteins regardless of their hypothetical age.
One problem that may be encountered in the construction of
SBP-based FRET sensors is that the conformational change of the
SBP may not translate into a change in distance between the N- and
C-termini, which is often the case when the termini are located on
the same domain of the bilobal SBP. A small change in the distance
between the termini of the SBP results in a small change in FRET
efficiency between the bound and unbound states of the sensor,
which limits its dynamic range. In this circumstance, the dynamic
range of the sensor can be improved by circular permutation, a
method that allows the termini of a protein to be relocated [2, 6].
Circular permutation is achieved by rearranging the sequence of the
SBP such that the original termini are connected by a flexible linker
and new termini are created by disconnecting the sequence at a different location, which is chosen to maximize the distance change
between the termini upon ligand binding. Since circular permutation is often destabilizing [17], the use of APR to obtain a thermostable SBP and the use of circular permutation to improve the
dynamic range of the resulting FRET sensor are complementary.
Here, we show how APR can be used to create thermostable
SBPs for the construction of robust FRET sensors. In this example, the maximum-likelihood statistical framework is used for
phylogenetic analysis and reconstruction of ancestral sequences,
although a variety of alternative methods are also available
[18–20]. We focus on the reconstruction of ancestral SBPs, but
similar approaches could be used to improve the thermostability of
any protein, provided that a suitable sequence dataset can be
obtained. We also show how the dynamic range of the sensor derived
from an ancestral SBP can be improved by circular permutation
prior to the insertion of the SBP into the FRET sensor construct.
2 Materials
1. SeaView version 4.6 (http://doua.prabi.fr/software/seaview).
2. ProtTest version 3.4 (https://github.com/ddarriba/prottest3).
3. FigTree version 1.4.2 (http://tree.bio.ed.ac.uk/software/
figtree).
2.1 Software
Improving FRET Sensors by Ancestral Gene Resurrection
protein [13]. Notably, computational evolutionary simulations [14]
and the observation that ancestral proteins share similarities with
consensus proteins in terms of both sequence and thermostability
[8, 15, 16] suggest that the high thermostability of ancient proteins
might also originate from bias in the maximum- likelihood method
commonly used for APR, specifically, that more probable (often stabilizing) residues are chosen at every site, while less probable (often
destabilizing) residues are neglected [14]. The implication of this
potential bias for protein engineering is that APR may produce thermostable proteins regardless of their hypothetical age.
One problem that may be encountered in the construction of
SBP-based FRET sensors is that the conformational change of the
SBP may not translate into a change in distance between the N- and
C-termini, which is often the case when the termini are located on
the same domain of the bilobal SBP. A small change in the distance
between the termini of the SBP results in a small change in FRET
efficiency between the bound and unbound states of the sensor,
which limits its dynamic range. In this circumstance, the dynamic
range of the sensor can be improved by circular permutation, a
method that allows the termini of a protein to be relocated [2, 6].
Circular permutation is achieved by rearranging the sequence of the
SBP such that the original termini are connected by a flexible linker
and new termini are created by disconnecting the sequence at a different location, which is chosen to maximize the distance change
between the termini upon ligand binding. Since circular permutation is often destabilizing [17], the use of APR to obtain a thermostable SBP and the use of circular permutation to improve the
dynamic range of the resulting FRET sensor are complementary.
Here, we show how APR can be used to create thermostable
SBPs for the construction of robust FRET sensors. In this example, the maximum-likelihood statistical framework is used for
phylogenetic analysis and reconstruction of ancestral sequences,
although a variety of alternative methods are also available
[18–20]. We focus on the reconstruction of ancestral SBPs, but
similar approaches could be used to improve the thermostability of
any protein, provided that a suitable sequence dataset can be
obtained. We also show how the dynamic range of the sensor derived
from an ancestral SBP can be improved by circular permutation
prior to the insertion of the SBP into the FRET sensor construct.
2 Materials
1. SeaView version 4.6 (http://doua.prabi.fr/software/seaview).
2. ProtTest version 3.4 (https://github.com/ddarriba/prottest3).
3. FigTree version 1.4.2 (http://tree.bio.ed.ac.uk/software/
figtree).
2.1 Software
Improving FRET Sensors by Ancestral Gene Resurrection