design of new compounds to mimic the four C-terminal residues of
the protein partner. This ligand was able to displace the reference
peptide probe in a concentration-dependent manner. However, the
affinity of the compound for MAGI3-PDZ was relatively weak, as a
competition was observed with concentrations in the 100 μM
range. In another study, the same compound was able to disrupt
the interaction between the Frizzed-7 Wnt receptor and the PDZ
domain of Dishevelled [25], leading to the downregulation of the
canonical Wnt signaling and suppression of tumor cell growth.
According to this study, the compound was among the first nonpeptide inhibitors to show therapeutic efficacy through the disruption of a PDZ PPI. Finally, this indole core was also used to target
the PDZ domain of NHERF1 [26]. Similar to the study from Fujii
et al., the DOCK tool was employed to suggest putative inhibitors
designed around the indole moiety [23]. In contrast to previous
work, an additional carboxylate group was incorporated at the end
of the flexible aliphatic sidechain of the compound to mimic the
aspartic residue from the reference partner.
In 2013, Vogrig et al. published small compounds able to
disrupt the PSD-95-PDZ1/5-HT2A receptor interaction leading
to an antihyperalgesic activity [27]. These inhibitors were discovered using a structure-based approach that combined molecular
modeling and NMR. A series of indole analogues were synthesized
on the basis of docking studies using AutoDock Vina software
[28]. Their ability to bind to the first PDZ domain (PDZ1) of
the PSD-95 protein was then assessed using NMR experiments.
The best compound exhibited a moderate IC 50 value of 190 μM.
PDZ inhibitors, which were able to disrupt the PSD-95PDZ2/GluN2B PPI, were discovered using a rational “click chemistry” strategy [29]. The aim was to mimic the TAV/SAV tripeptide
PSD-95 ligand using triazole-containing compounds that were
easily synthesized from reactive azide and alkyne moieties. The
triazole heterocycle was chosen because it was previously reported
as a potential amide bioisostere while being more rigid and not
recognized by protein peptidases. One triazole-containing compound inhibited the PSD-95-PDZ2/GluN2B interaction with an
affinity similar to that measured for the SAV tripeptide. ITC experiments concluded that the compound had a low affinity for the
PSD-95-PDZ2 domain (K d value in the 600 μM range). Finally,
molecular docking simulations using the Glide tool [30] suggested
that the triazole-containing compound interacts with the PDZ2
domain in a similar way as the TAV tripeptide.
In 2010, Thorsen et al. reported an organic compound
(FSC231) able to bind to the PDZ domain from PICK1 [9]. The
inhibitor was identified using a fluorescent polarization assay by
screening approximately 44,000 compounds. This inhibitor exhibited an affinity similar to that measured for peptide ligands
(C-terminal end from endogenous protein partners) in the 10 μM
Rational Design of PDZ Domain Inhibitors
281
the protein partner. This ligand was able to displace the reference
peptide probe in a concentration-dependent manner. However, the
affinity of the compound for MAGI3-PDZ was relatively weak, as a
competition was observed with concentrations in the 100 μM
range. In another study, the same compound was able to disrupt
the interaction between the Frizzed-7 Wnt receptor and the PDZ
domain of Dishevelled [25], leading to the downregulation of the
canonical Wnt signaling and suppression of tumor cell growth.
According to this study, the compound was among the first nonpeptide inhibitors to show therapeutic efficacy through the disruption of a PDZ PPI. Finally, this indole core was also used to target
the PDZ domain of NHERF1 [26]. Similar to the study from Fujii
et al., the DOCK tool was employed to suggest putative inhibitors
designed around the indole moiety [23]. In contrast to previous
work, an additional carboxylate group was incorporated at the end
of the flexible aliphatic sidechain of the compound to mimic the
aspartic residue from the reference partner.
In 2013, Vogrig et al. published small compounds able to
disrupt the PSD-95-PDZ1/5-HT2A receptor interaction leading
to an antihyperalgesic activity [27]. These inhibitors were discovered using a structure-based approach that combined molecular
modeling and NMR. A series of indole analogues were synthesized
on the basis of docking studies using AutoDock Vina software
[28]. Their ability to bind to the first PDZ domain (PDZ1) of
the PSD-95 protein was then assessed using NMR experiments.
The best compound exhibited a moderate IC 50 value of 190 μM.
PDZ inhibitors, which were able to disrupt the PSD-95PDZ2/GluN2B PPI, were discovered using a rational “click chemistry” strategy [29]. The aim was to mimic the TAV/SAV tripeptide
PSD-95 ligand using triazole-containing compounds that were
easily synthesized from reactive azide and alkyne moieties. The
triazole heterocycle was chosen because it was previously reported
as a potential amide bioisostere while being more rigid and not
recognized by protein peptidases. One triazole-containing compound inhibited the PSD-95-PDZ2/GluN2B interaction with an
affinity similar to that measured for the SAV tripeptide. ITC experiments concluded that the compound had a low affinity for the
PSD-95-PDZ2 domain (K d value in the 600 μM range). Finally,
molecular docking simulations using the Glide tool [30] suggested
that the triazole-containing compound interacts with the PDZ2
domain in a similar way as the TAV tripeptide.
In 2010, Thorsen et al. reported an organic compound
(FSC231) able to bind to the PDZ domain from PICK1 [9]. The
inhibitor was identified using a fluorescent polarization assay by
screening approximately 44,000 compounds. This inhibitor exhibited an affinity similar to that measured for peptide ligands
(C-terminal end from endogenous protein partners) in the 10 μM
Rational Design of PDZ Domain Inhibitors
281