Chemistry Reference
In-Depth Information
approach, the global folds that can be accessed without the requirement for NOE
assignment have the exciting potential to increase the number of membrane
proteins structures that can be characterized by solution NMR.
5.3.3 Paramagnetic Relaxation Enhancements
As shown in the UCP2 study, RDCs on their own are usually not sufficient to
determine a unique protein conformation, and long-range distance restraints are
often required to resolve structural ambiguities. PREs have frequently been used to
provide this type of distance information for membrane proteins both in the absence
and presence of RDC data, particularly when long-range NOEs are not available
(listed in Table
2
, Method VII). PREs arise from dipolar interactions involving the
large dipole of an unpaired electron that leads to efficient loss of polarization and
coherence in surrounding nuclear spins. There is a relatively straightforward rela-
tionship between the paramagnetically enhanced transverse relaxation rate of a
proton and the inverse sixth power of its distance from the unpaired electron (
r
6
),
allowing this measurement to be converted into distance restraints for structure
determination [
359
,
360
]. This is usually achieved by measurement of paramagnetic
spin-induced changes in backbone amide
1
H peak intensities [
360
], although greater
accuracy can be achieved by either full [
361
] or two-point [
362
]R
2
relaxation
measurements [
363
]. The range of distances that are accessed by the PRE measure-
ment depends on the type of paramagnetic center introduced into the protein, but
typically provides distance information on spins as far away as 25
˚
or more.
There are several different types of spin labels that can be introduced into
the target protein, usually via covalent modification of a unique cysteine residue
and/or inclusion of a metal-chelating peptide in the protein sequence (reviewed in
[
364
,
365
]). Many of the metal-chelating tags used to induce alignment for measure-
ment of RDCs (Sect.
5.3.1
) can also be used to bind to paramagnetic metals with slow
electronic relaxation properties (e.g., Mn
2+
and Cu
2+
) that give rise to a significant
PRE. However, for membrane proteins, the vast majority of PRE-based restraints
have used the sulfhydryl-reactive nitroxide spin label, 1-oxyl-2,2,5,5-tetramethyl-3-
pyrroline-3-methyl-methanethiosulfonate (MTSL) [
38
,
39
,
41
,
67
,
88
,
136
,
354
,
366
].
To obtain sufficient numbers of PRE-based restraints, it is usually necessary to
introduce spin-labels at a diverse range of sites. For example, nine different MTSL-
labeled DsbB samples (approximately 1 MTSL-label for every 20 residues), were
required to generate an average of 6.5 PREs per residue [
39
]. Choosing where to
place these spin-labels requires identification of sites that can provide informative
PRE-based restraints without disrupting the structure. This makes it necessary to
have some knowledge of the structure being studied, with secondary structure
determination from backbone assignments being the first step in this process.
Once labeled, the NMR spectrum is used to confirm that the probe does not alter
structure, since only local shift changes should be observed in this case.
For any type of paramagnetic center used, accurate translation of PREs into
distance restraints requires that the entire population be spin-labeled, since the extent
