Biomedical Engineering Reference
In-Depth Information
under consideration. For example, Krueger-Koplin et al.
109
screened 25
detergents against three different proteins; single-helix R. sphaeroides LH1,
two-helix E. coli subunit c and four-helix S. aureus Smr. Some detergents were
found to give high-quality spectra for one of the test proteins, but poor quality
spectra for others. Overall, the lyso-phosphatidylglycerol class proved the most
suitable. However, this prediction has not been borne out for other systems.
105
Overall a number of general guidelines seem to emerge relevant for NMR
studies. A low CMC and aggregation number are important considerations.
High values for the CMC lead to very viscous solutions, reducing the rate of
molecular tumbling, whilst a high aggregation number leads to large protein-
detergent complexes again increasing the tumbling time. For example, in our
experience, whilst dioctanoylphosphatidylcholine (c8-diPC) has been suggested
in some studies as a suitable phosphocholine detergent,
96
the large aggregates
formed
111
resulted in extremely poor-quality NMR spectra. Avoiding highly
denaturing detergents is crucial. Sodium dodecylsulphate, used in some studies,
is generally found to destroy native activity, rendering it unsuitable for structural
studies of more complex proteins.
108
Similar problems are experienced with
extremely short-chain detergents which, with a large headgroup size relative to
chain length and thus a more conical profile, result in high micelle curvature and
are found to be more denaturing relative to the longer chain-length analogues
with a more cylindrical profile.
108
Effects may also result from poor mimicking
of the membrane lateral pressure profile.
112,113
There is some evidence that
mimicking the native headgroups found in the membrane may also be
beneficial,
109
and may prove important for protein functionality.
114
Charged
headgroups are thought to cause repulsion between micelles, limiting protein
aggregation and thus prolonging sample lifetime.
109
In our experience with pSRII, we found protein translational diffusion
measurements based on the BPP-LED sequence
115,116
to be of particular use in
characterising the size of protein-detergent aggregates formed, and thus to give
an initial estimate of the expected quality of spectra (Figure 12.3). These
measurements can be carried out at low protein concentrations, making this
an effective screening mechanism, before larger sample concentrations are
prepared for 2D studies. In our case, initial studies were carried out in n-
dodecyl-b-
D
-maltoside (DDM). Whilst the protein was found to be
functional and stable, high-quality spectra could only be gained in this
system with highly deuterated protein and at high temperatures (y55 uC).
Analysis by size-exclusion chromatography indicated the suspected presence
of a heterogeneous distribution of solubilised monomers and dimers, whilst
translational diffusion measurements indicated a relatively slow diffusion
rate [Figure 12.3(b) and (g)] and large aggregate size of ca. 120-150 kDa. A
variety of detergents were screened and diheptanoylphosphatidylcholine (c7-
DHPC) was found to be the most suitable, giving high-quality [
1
H,
15
N]
TROSY spectra at 50 uC
105
and a faster translational diffusion rate
[Figure 12.3(a) and (g)]. Size-exclusion chromatography indicated a single
species
with
a
molecular
weight
of
ca.
50-70
kDa,
believed
to
be
a
