spectrometer and the type of sample changer used, and the manufacturer’s instructions should be followed for these details. A 1D
1 H NMR spectrum with solvent suppression should be acquired on
each sample, along with the required LO-NMR binding
experiments.
After acquisition of the first set of LO-NMR experiments, a
small volume of competitor ligand should be added to the required
concentration as discussed previously. The set of LO-NMR experiments should then be acquired again, allowing comparison of the
observed binding before and after the displacement step. If no
potent competitor is available, this step can be excluded, although
subsequent validation steps will be required to be stringent (see
Subheading 3.8).
3.5 Data Analysis
The signal in each LO-NMR experiment reflects different aspects of
the sample and of the binding of any ligands present. A brief
overview of these aspects will be discussed below, followed by a
discussion of the analysis of these data.
3.5.1 1D
1
H NMR
As mentioned, the
1 H 1D NMR experiment contains information
on almost every component present in the sample, the exceptions
being molecules that contain no observable resonances (such as
phosphate buffer, metal ions, or deuterated solvents) and those
macromolecules that are tumbling so slowly as to give rise to signals
that are too broad to be observed (such as protein or compound
aggregates). Of particular importance are the resonances from the
ligands, the protein, and the buffer components.
Resonances from the ligands should be compared with the
reference spectra acquired for the isolated compounds. Minor perturbations in the spectra of the compounds are typically the result
of small differences in the pH or ionic strength of reference and
mixture samples. However, significant differences between the
observed and reference spectra are often indications of compound
degradation (either of the DMSO stocks themselves or subsequent
to their being diluted in aqueous solution) or of interactions
between compounds present in the mixture. Such differences can
significantly complicate analysis or give rise to artifactual results.
Importantly, ligand resonances should not change significantly
between the “before” and “after” addition of competitor; loss of
ligand signal between these two conditions is typically associated
with compound precipitation and should not be interpreted as
displacement (particularly in the STD experiment, see below).
Inspection of resonances from the protein, particularly within
the methyl envelope (between 0.7 and 1.1 ppm) and any resolved
shifted aliphatic resonances (below 0.7 ppm) can readily identify
differences for samples where the protein has precipitated or denatured. This typically occurs as a function of the one or more
compounds present in a mixture and is one of the most common
causes of false positives and negatives in screening experiments.
Fragment Screening by NMR
261
1 H NMR spectrum with solvent suppression should be acquired on
each sample, along with the required LO-NMR binding
experiments.
After acquisition of the first set of LO-NMR experiments, a
small volume of competitor ligand should be added to the required
concentration as discussed previously. The set of LO-NMR experiments should then be acquired again, allowing comparison of the
observed binding before and after the displacement step. If no
potent competitor is available, this step can be excluded, although
subsequent validation steps will be required to be stringent (see
Subheading 3.8).
3.5 Data Analysis
The signal in each LO-NMR experiment reflects different aspects of
the sample and of the binding of any ligands present. A brief
overview of these aspects will be discussed below, followed by a
discussion of the analysis of these data.
3.5.1 1D
1
H NMR
As mentioned, the
1 H 1D NMR experiment contains information
on almost every component present in the sample, the exceptions
being molecules that contain no observable resonances (such as
phosphate buffer, metal ions, or deuterated solvents) and those
macromolecules that are tumbling so slowly as to give rise to signals
that are too broad to be observed (such as protein or compound
aggregates). Of particular importance are the resonances from the
ligands, the protein, and the buffer components.
Resonances from the ligands should be compared with the
reference spectra acquired for the isolated compounds. Minor perturbations in the spectra of the compounds are typically the result
of small differences in the pH or ionic strength of reference and
mixture samples. However, significant differences between the
observed and reference spectra are often indications of compound
degradation (either of the DMSO stocks themselves or subsequent
to their being diluted in aqueous solution) or of interactions
between compounds present in the mixture. Such differences can
significantly complicate analysis or give rise to artifactual results.
Importantly, ligand resonances should not change significantly
between the “before” and “after” addition of competitor; loss of
ligand signal between these two conditions is typically associated
with compound precipitation and should not be interpreted as
displacement (particularly in the STD experiment, see below).
Inspection of resonances from the protein, particularly within
the methyl envelope (between 0.7 and 1.1 ppm) and any resolved
shifted aliphatic resonances (below 0.7 ppm) can readily identify
differences for samples where the protein has precipitated or denatured. This typically occurs as a function of the one or more
compounds present in a mixture and is one of the most common
causes of false positives and negatives in screening experiments.
Fragment Screening by NMR
261