In addition to verifying the structure and stability of each
compound, it is also necessary to confirm that the compound is
soluble at the expected concentration. This is readily done by
comparison of the compound resonance integrals with the integral
of an internal standard (such as 50 μM DSS (see Note 3)). Compounds that are not soluble at the required concentration should
not be incorporated into the screening library.
A final step in characterization of the compounds to be
included in the fragment library is to analyze the water-LOGSY
spectrum acquired on each compound sample. The water-LOGSY
spectrum is sensitive to association with slowly tumbling macromolecules and as such can be used as an indicator of compound selfassociation to form microaggregates. Although many factors influence the intensity and magnitude of the water-LOGSY spectrum,
self-association will dominate the observed spectrum in a sample
containing only compound and buffer. If the compound is selfassociating, a positive water-LOGSY spectrum is observed and the
compound should be excluded from the library; the formation of
microaggregates can give rise to many artifactual signals and complicates analysis of the water-LOGSY spectra acquired during the
fragment screening procedure.
After data analysis and subsequent selection of compounds
suitable for incorporation in the fragment library, it is useful to
store the reference spectra for each compound in a spectral database. The procedure for doing this will obviously depend on the
analysis software used, but most modern NMR analysis packages
include a spectral database functionality.
3.2 Characterization
of the Target Protein
Characterization of the protein is a key part of establishing the
NMR fragment screening conditions. The protein should be as
pure as possible (>95% by SDS-PAGE), and ideally, the identity
of the protein should be confirmed (for example, by intact mass
spectrometry or peptide-mass fingerprinting [40, 41]). Samples of
the protein in a suitable buffer should be prepared (see Note 4), and
1 H 1D NMR spectra of the protein should be acquired in order to
confirm that the protein is folded correctly.
1 H 1D or
15 N–
1
H
HSQC spectra should be used to identify interactions with buffer
components such as metal ions or detergents [42].
Acquisition of NMR spectra should be repeated at intervals to
confirm that the protein is stable under these conditions for the
expected duration of the fragment screening. It is also advisable to
examine the protein by SDS-PAGE and/or intact mass spectrometry after the final experiment to confirm that no degradation has
occurred. If multiple batches of the protein are to be used, 1D
1 H
NMR spectra can be used to confirm the batch-to-batch consistency of both protein and buffer.
Fragment Screening by NMR
257
compound, it is also necessary to confirm that the compound is
soluble at the expected concentration. This is readily done by
comparison of the compound resonance integrals with the integral
of an internal standard (such as 50 μM DSS (see Note 3)). Compounds that are not soluble at the required concentration should
not be incorporated into the screening library.
A final step in characterization of the compounds to be
included in the fragment library is to analyze the water-LOGSY
spectrum acquired on each compound sample. The water-LOGSY
spectrum is sensitive to association with slowly tumbling macromolecules and as such can be used as an indicator of compound selfassociation to form microaggregates. Although many factors influence the intensity and magnitude of the water-LOGSY spectrum,
self-association will dominate the observed spectrum in a sample
containing only compound and buffer. If the compound is selfassociating, a positive water-LOGSY spectrum is observed and the
compound should be excluded from the library; the formation of
microaggregates can give rise to many artifactual signals and complicates analysis of the water-LOGSY spectra acquired during the
fragment screening procedure.
After data analysis and subsequent selection of compounds
suitable for incorporation in the fragment library, it is useful to
store the reference spectra for each compound in a spectral database. The procedure for doing this will obviously depend on the
analysis software used, but most modern NMR analysis packages
include a spectral database functionality.
3.2 Characterization
of the Target Protein
Characterization of the protein is a key part of establishing the
NMR fragment screening conditions. The protein should be as
pure as possible (>95% by SDS-PAGE), and ideally, the identity
of the protein should be confirmed (for example, by intact mass
spectrometry or peptide-mass fingerprinting [40, 41]). Samples of
the protein in a suitable buffer should be prepared (see Note 4), and
1 H 1D NMR spectra of the protein should be acquired in order to
confirm that the protein is folded correctly.
1 H 1D or
15 N–
1
H
HSQC spectra should be used to identify interactions with buffer
components such as metal ions or detergents [42].
Acquisition of NMR spectra should be repeated at intervals to
confirm that the protein is stable under these conditions for the
expected duration of the fragment screening. It is also advisable to
examine the protein by SDS-PAGE and/or intact mass spectrometry after the final experiment to confirm that no degradation has
occurred. If multiple batches of the protein are to be used, 1D
1 H
NMR spectra can be used to confirm the batch-to-batch consistency of both protein and buffer.
Fragment Screening by NMR
257