Chemistry Reference
In-Depth Information
contribution to
R
1
r
. For this purpose, a strong
B
1
field strength, which makes
y
close
to 90
, is employed.
Limitations of applicable B
1
field strength
and the rf duty cycle depend on
individual probes. In general, at
15
N resonance frequency at 61 MHz, we recom-
g
N
B
1
/2
p >
mend that a spin lock field
2 kHz be applied for 60-80 ms to determine
15
N
R
1
r
. The
g
N
B
1
/2
p
was calculated assuming that the entire chemical shift range
for amide backbone
15
N signals in diamagnetic proteins is
900 Hz (
15 ppm) at
61 MHz, which corresponds to sin
0.9. The spin lock duration was estimated
based on approximate
15
N
R
2
of a folded 10-20 kDa protein at room temperature.
Although the highest measurement accuracy is obtained when data is recorded until
the magnetization decays sufficiently (typically for a time
y >
1/
R
1
r
), it may not
be possible to satisfy this condition for
15
N sites which relax slowly (as in unfolded
proteins or small peptides) without reducing
B
1
which leads to a reduction in sin
> ¼
.
Errors
in
R
1
r
resulting from off-resonance effects may be significant but can be
corrected. It is an advantage of the
R
1
r
experiment that, even when a signal is off-
resonance from the rf carrier frequency and for which sin
y
is small, an accurate
R
1
r
value can be obtained using equation (
1
). Although the correction requires an
R
1
value, this will be available when
R
1
data is recorded to characterize fast backbone
dynamics using model-free analysis. As described above, the accuracy of
R
1
r
measurements decreases for signals located far off-resonance. Compared to
CPMG
R
2
that is described in Sect.
2.1.2
,
R
1
r
values do not need to be recorded
at different carrier-frequencies without discarding any data.
Cross-correlation
interference by
1
H-
15
N dipolar interaction (DD) and
15
NCSA
hastobesuppressedtodetectaccurate
15
N transverse relaxation rates (Fig.
1
). The
cross term is suppressed by flipping the sign of the DD term by applying
1
H180
pulses at a rate greater than the decay rate of the two
15
N-
1
H J-coupled components
[
35
-
37
]. However, in a weak
B
1
field, the two J-coupled components undergo
y
Fig. 1 (a)
15
N-H dipolar and (b)
15
N CSA contribution to longitudinal and transverse relaxation
rates (
R
1
, and
R
2
) as a function of a correlation time. The rates were calculated assuming a simple
Lorentzian spectral density function,
J
(
2
).
Solid
and
dotted lines
indicate rates
calculated assuming at 900 MHz and 600 MHz instruments, respectively
2
o
)
¼ t
/(1 +
o
t