Protein Structure and Dynamics by NMR in Solution
95
ensemble average and that measurable spectral parameters represent all conformations interchanging in the ns to ms time range. Therefore, the NMR spectroscopist only gets a blurred picture of the molecule: when several spectral parameters seem mutually inconsistent, this may simply originate from averaging processes during the NMR experiment.
3.1
Chemical Shift
The origin of the chemical shift is that the moving charges of the electron cloud
around the nucleus under observation induce a local magnetic field which adds
to the applied field, this is called the shielding of the nucleus. As a result, nuclei
which do not exhibit exactly the same (chemical and structural) environment,
have different resonance frequencies or chemical shifts. Although the chemical
shift is a small effect (= 10- 6 ), partially resolved resonances can generally be
observed for individual nuclei in a biomolecule. The chemical shift of a resonance (6) is expressed with respect to a reference line in ppm units: 6 = (vV ref)/v 0' where v 0 is the Larmor frequency. The Larmor frequency depends on the
magnetic field strength and typically takes values between 500 and 800 MHz for
the proton in high resolution NMR spectroscopy. Whereas the separation in Hz
between the resonances of two different spins increases with the magnetic field,
the chemical shift difference in ppm remains constant. In considering that the
NMR line-width is field-independent, it becomes clear that one of the major benefits of higher magnetic fields is the larger separation between resonances (called
spectral resolution). This accounts for the endless race to higher static magnetic
fields which is only limited by production costs and available technology.
A reference is needed for the measurement of the chemical shift; it can be
either a small molecule added to the sample (such as tetramethylsilane,TMS or 3trimethylsilylpropionate, TSP) or a residual solvent line. In the first case, it is
assumed that its chemical shift is insensitive to the environmental conditions
(pH, ionic strength or temperature) and in the latter case, empirical correction is
frequently needed. Several authors have proposed relationships between I3C, 15N
and lH chemical shift scales based on precise knowledge of their gyro magnetic
ratio (c.f. Wishart et al. 1995).
The shielding of a nucleus is anisotropic (i.e. it varies with the orientation of
the molecule in the magnetic field) and can therefore only be described by a rank
2 tensor (011,022,033)' However, in solution, the molecules generally undergo fast
rotation (due to the brownian motion) and do not exhibit any preferential orientation. Because of this averaging, the tensor reduces to its trace: 0 = 1/3 (0 II + 022
+ 033)' In this case, complex relaxation measurements (see below) are required
for the detection of chemical shielding anisotropy (CSA). On the other hand, in
the case of partially aligned molecules, CSA leads to the variation of the observed
chemical shift (Ottiger et al. 1997). In this case the chemical shift is not simply
given by the trace of the tensor but a weighted average of the individual components.
95
ensemble average and that measurable spectral parameters represent all conformations interchanging in the ns to ms time range. Therefore, the NMR spectroscopist only gets a blurred picture of the molecule: when several spectral parameters seem mutually inconsistent, this may simply originate from averaging processes during the NMR experiment.
3.1
Chemical Shift
The origin of the chemical shift is that the moving charges of the electron cloud
around the nucleus under observation induce a local magnetic field which adds
to the applied field, this is called the shielding of the nucleus. As a result, nuclei
which do not exhibit exactly the same (chemical and structural) environment,
have different resonance frequencies or chemical shifts. Although the chemical
shift is a small effect (= 10- 6 ), partially resolved resonances can generally be
observed for individual nuclei in a biomolecule. The chemical shift of a resonance (6) is expressed with respect to a reference line in ppm units: 6 = (vV ref)/v 0' where v 0 is the Larmor frequency. The Larmor frequency depends on the
magnetic field strength and typically takes values between 500 and 800 MHz for
the proton in high resolution NMR spectroscopy. Whereas the separation in Hz
between the resonances of two different spins increases with the magnetic field,
the chemical shift difference in ppm remains constant. In considering that the
NMR line-width is field-independent, it becomes clear that one of the major benefits of higher magnetic fields is the larger separation between resonances (called
spectral resolution). This accounts for the endless race to higher static magnetic
fields which is only limited by production costs and available technology.
A reference is needed for the measurement of the chemical shift; it can be
either a small molecule added to the sample (such as tetramethylsilane,TMS or 3trimethylsilylpropionate, TSP) or a residual solvent line. In the first case, it is
assumed that its chemical shift is insensitive to the environmental conditions
(pH, ionic strength or temperature) and in the latter case, empirical correction is
frequently needed. Several authors have proposed relationships between I3C, 15N
and lH chemical shift scales based on precise knowledge of their gyro magnetic
ratio (c.f. Wishart et al. 1995).
The shielding of a nucleus is anisotropic (i.e. it varies with the orientation of
the molecule in the magnetic field) and can therefore only be described by a rank
2 tensor (011,022,033)' However, in solution, the molecules generally undergo fast
rotation (due to the brownian motion) and do not exhibit any preferential orientation. Because of this averaging, the tensor reduces to its trace: 0 = 1/3 (0 II + 022
+ 033)' In this case, complex relaxation measurements (see below) are required
for the detection of chemical shielding anisotropy (CSA). On the other hand, in
the case of partially aligned molecules, CSA leads to the variation of the observed
chemical shift (Ottiger et al. 1997). In this case the chemical shift is not simply
given by the trace of the tensor but a weighted average of the individual components.