Ayoub et al. conducted a virtual screen of six million compounds against the
first bromodomain of BRDT; the top 0.1% were filtered to 200 compounds by
precision filtering and removal of potential pan-assay interference compounds.
Twenty of these compounds were purchased and tested in a fluorescence polarization (FP) assay, of which nine had activity against BRDT. The lead compound, a
dihydropyridine, was further verified as a BET inhibitor using differential scanning
fluorimetry, a protein-observed NMR method, and finally the binding pose was
elucidated by a co-crystal structure with the BET bromodomain BRD4 BD1. Due
to the higher binding affinity for BRD4 BD1, the potency of the molecule was
improved leading to in-cell activity in MM.1.S cells by downregulating Myc
expression [80].
4.2 Direct-Binding Experiments with Bromodomains
4.2.1 NMR Methods Have Been Instrumental in Bromodomain
Inhibitor Discovery
As previously noted, early research by Zhou and co-workers using NMR
methods was instrumental in confirming the tertiary structure of bromodomains
and understanding binding-site interactions between acetylated histones and
bromodomains [17]. Since then, NMR methods have been widely used in screening
assays and to characterize bromodomain inhibitors. In a NMR binding assay, the
difference in the chemical shift, intensity, or phase of the resonance is analyzed
between two experiments: (1) ligand- or protein-only reference and (2) ligand +
protein. The two broad classes of NMR binding assays are ligand- and proteinobserved, in which the resonances of ligands or the protein are monitored, respectively. Both ligand- and protein-observed NMR methods are sensitive techniques
that can detect a wide range of binding affinities (nM to mM) making them
applicable assays for any stage of inhibitor development. NMR methods used in
drug discovery of bromodomains include, but are not limited to, HSQC, PrOF, STD,
and CPMG NMR. These methods are described below.
Both ligand- and protein-observed NMR methods offer distinct advantages and
disadvantages. For example, ligand-observed experiments require relatively low
concentrations of protein (5–10 μM per experiment), and the behavior of ligands
in solution can be monitored by resonance height, line width, or chemical shift.
Additionally, mixtures of ligands can be screened in a single NMR tube with no
deconvolution needed, because the identity of each ligand resonance is known.
However, ligand-observed methods give no information on protein stability or
binding-site location. In contrast, protein-observed methods require larger amounts
of isotopically labeled protein (typically 50–200 μM of
19 F-,
13 C-, or
15 N-labeled
protein per experiment). An advantage of protein-observed methods is the ability to
glean protein structural information from the spectrum. Because these methods
observe protein resonances, a disadvantage of protein-observed experiments using
308
W. C. K. Pomerantz et al.
first bromodomain of BRDT; the top 0.1% were filtered to 200 compounds by
precision filtering and removal of potential pan-assay interference compounds.
Twenty of these compounds were purchased and tested in a fluorescence polarization (FP) assay, of which nine had activity against BRDT. The lead compound, a
dihydropyridine, was further verified as a BET inhibitor using differential scanning
fluorimetry, a protein-observed NMR method, and finally the binding pose was
elucidated by a co-crystal structure with the BET bromodomain BRD4 BD1. Due
to the higher binding affinity for BRD4 BD1, the potency of the molecule was
improved leading to in-cell activity in MM.1.S cells by downregulating Myc
expression [80].
4.2 Direct-Binding Experiments with Bromodomains
4.2.1 NMR Methods Have Been Instrumental in Bromodomain
Inhibitor Discovery
As previously noted, early research by Zhou and co-workers using NMR
methods was instrumental in confirming the tertiary structure of bromodomains
and understanding binding-site interactions between acetylated histones and
bromodomains [17]. Since then, NMR methods have been widely used in screening
assays and to characterize bromodomain inhibitors. In a NMR binding assay, the
difference in the chemical shift, intensity, or phase of the resonance is analyzed
between two experiments: (1) ligand- or protein-only reference and (2) ligand +
protein. The two broad classes of NMR binding assays are ligand- and proteinobserved, in which the resonances of ligands or the protein are monitored, respectively. Both ligand- and protein-observed NMR methods are sensitive techniques
that can detect a wide range of binding affinities (nM to mM) making them
applicable assays for any stage of inhibitor development. NMR methods used in
drug discovery of bromodomains include, but are not limited to, HSQC, PrOF, STD,
and CPMG NMR. These methods are described below.
Both ligand- and protein-observed NMR methods offer distinct advantages and
disadvantages. For example, ligand-observed experiments require relatively low
concentrations of protein (5–10 μM per experiment), and the behavior of ligands
in solution can be monitored by resonance height, line width, or chemical shift.
Additionally, mixtures of ligands can be screened in a single NMR tube with no
deconvolution needed, because the identity of each ligand resonance is known.
However, ligand-observed methods give no information on protein stability or
binding-site location. In contrast, protein-observed methods require larger amounts
of isotopically labeled protein (typically 50–200 μM of
19 F-,
13 C-, or
15 N-labeled
protein per experiment). An advantage of protein-observed methods is the ability to
glean protein structural information from the spectrum. Because these methods
observe protein resonances, a disadvantage of protein-observed experiments using
308
W. C. K. Pomerantz et al.