residue at the ligation junction can be a limitation if an appropriate
ligation site with a native Cys residue cannot be identified. In this
case, a Cys mutation has to be introduced at the desired ligation
site, which may not be possible due to a consequently building of
sulfide bridges/heteromerization and eventually loss of protein
function. However, a number of approaches involving ligation
auxiliaries, Cys surrogates, and desulfurization methods, have
been developed to allow more flexibility in the selection of ligation
sites besides Cys [11, 19–23]. With these improvements, EPL has
become applicable for the generation of a wider spectrum of
proteins.
Several types of chemical modifications have been incorporated
into PDZ domains using both strategies of EPL to investigate sidechain interactions [24], backbone interactions [25], and phosphorylations [26]. Studies of backbone H-bond interactions are especially crucial for PDZ binding because the binding of C-terminal
peptide ligands to the PDZ protein domain is facilitated by a
conserved carboxylate-binding site via backbone hydrogen amide
bonds [25]. This was deciphered by the introduction of amide-toester mutations to these and other regions of PDZ domains
through EPL. Peptide fragments containing backbone amide-toester mutations (depsipeptides) were ligated to a recombinantly
expressed protein fragment, enabling the investigation of specific
backbone hydrogen bonds in the PDZ domain protein [25].
EPL also enabled the generation of semisynthetic phosphorylated PDZ domains, specifically of PDZ domains of the postsynaptic density protein 95 (PSD-95), which allowed its
phosphoregulated PPIs to be studied [26, 27]. A riveting feature
of this work is the direct comparison of the effects of introducing
phosphomimetics versus a semisynthetic protein with a ‘true’ phosphorylation. Although phosphomimetics, Glu and Asp, are used
often to study this important PTM, structural and chemical differences of the mimetics could render them as inadequate substitutions, especially in cases involving phosphotyrosine. In contrast,
in vitro phosphorylation uses kinases to install phosphate-groups
onto proteins, but this would not be feasible in certain cases where
the enzyme is unknown, or site-specificity is desired, but the
enzyme modifies multiple sites in the same protein. Recently, an
alternative method leverages a genetic code expansion technique to
incorporate more than 200 nonproteinogenic amino acids, including phosphorylated amino acids [28]. However, an enigmatic challenge with this technology is that not all protein sites are amenable
to modifications, and the reason for this is still in contention. In
these situations, EPL is an attractive option to introduce phosphorylation site-specifically to proteins in a reliable manner.
Herein, we provide protocols that have successfully generated
PDZ domains with modifications in either the N- or C-terminal
regions. Specifically, synthesis of peptide thioesters by Boc- and
Synthetic PDZ Domains
197
ligation site with a native Cys residue cannot be identified. In this
case, a Cys mutation has to be introduced at the desired ligation
site, which may not be possible due to a consequently building of
sulfide bridges/heteromerization and eventually loss of protein
function. However, a number of approaches involving ligation
auxiliaries, Cys surrogates, and desulfurization methods, have
been developed to allow more flexibility in the selection of ligation
sites besides Cys [11, 19–23]. With these improvements, EPL has
become applicable for the generation of a wider spectrum of
proteins.
Several types of chemical modifications have been incorporated
into PDZ domains using both strategies of EPL to investigate sidechain interactions [24], backbone interactions [25], and phosphorylations [26]. Studies of backbone H-bond interactions are especially crucial for PDZ binding because the binding of C-terminal
peptide ligands to the PDZ protein domain is facilitated by a
conserved carboxylate-binding site via backbone hydrogen amide
bonds [25]. This was deciphered by the introduction of amide-toester mutations to these and other regions of PDZ domains
through EPL. Peptide fragments containing backbone amide-toester mutations (depsipeptides) were ligated to a recombinantly
expressed protein fragment, enabling the investigation of specific
backbone hydrogen bonds in the PDZ domain protein [25].
EPL also enabled the generation of semisynthetic phosphorylated PDZ domains, specifically of PDZ domains of the postsynaptic density protein 95 (PSD-95), which allowed its
phosphoregulated PPIs to be studied [26, 27]. A riveting feature
of this work is the direct comparison of the effects of introducing
phosphomimetics versus a semisynthetic protein with a ‘true’ phosphorylation. Although phosphomimetics, Glu and Asp, are used
often to study this important PTM, structural and chemical differences of the mimetics could render them as inadequate substitutions, especially in cases involving phosphotyrosine. In contrast,
in vitro phosphorylation uses kinases to install phosphate-groups
onto proteins, but this would not be feasible in certain cases where
the enzyme is unknown, or site-specificity is desired, but the
enzyme modifies multiple sites in the same protein. Recently, an
alternative method leverages a genetic code expansion technique to
incorporate more than 200 nonproteinogenic amino acids, including phosphorylated amino acids [28]. However, an enigmatic challenge with this technology is that not all protein sites are amenable
to modifications, and the reason for this is still in contention. In
these situations, EPL is an attractive option to introduce phosphorylation site-specifically to proteins in a reliable manner.
Herein, we provide protocols that have successfully generated
PDZ domains with modifications in either the N- or C-terminal
regions. Specifically, synthesis of peptide thioesters by Boc- and
Synthetic PDZ Domains
197