Communications
The progress of the catalyzed reaction was monitored in
a chromogenic substrate to enable this comparison, there are
many naturally occurring substrates that are inaccessible to
spectrophotometry, but that can still be directly observed by
the DNP NMR measurements presented here.
1
3
real time by observing a sequence of C NMR spectra over
the course of three seconds. For each spectrum, a fraction of
the polarization that had previously been generated by the
DNP mechanism, and that had been retained by the substrate
and product molecules, was converted into NMR-observable
The increase in the intensity of the resonance stemming
from the reaction product was also linear (not shown). The
[
12]
À1
spin coherence by a variable flip angle pulse. From the
spectra shown in Figure 2, the reduction in the intensity of the
apparent rate constant was 9.0 s , which is lower than the rate
of catalysis. The reason for this difference most likely is a
shortening of the spin-lattice relaxation time when substrate
is bound to the enzyme. This effect could potentially lead to
additional information on the dynamics of the binding of
different substrates.
By using the present injection system, the first data point
could be acquired with an initial delay of 300 ms. Further
improvements of the apparatus will include the design of an
NMR flow cell, to increase the time resolution close to the
[
14]
theoretical minimum of about 10 ms.
The gain in signal provided by DNP enabled the
1
3
measurement in a single scan of C spectra without the
1
3
need for isotopic enrichment ( C concentration of 36 mm at
.1% natural abundance). A conventionally acquired NMR
spectrum of a standard solution of 25 mm BAEE under Ernst-
1
[2]
angle conditions yielded a signal/noise (S/N) ratio of 31:1 in
0.7 h. It is extrapolated that a spectrum equivalent to the first
1
spectrum in the hyperpolarized data set (S/N of 62:1 in first
scan), but using conventional signal averaging would require
over 100 days. The observation of reaction kinetics would thus
be impossible. This comparison illustrates the value of using
DNP-enhanced NMR spectroscopy for enabling the mea-
surement of rapid processes that would otherwise be inacces-
sible to NMR spectroscopy.
Figure 2. Kinetics of BAEE hydrolysis (3.3 mm) by trypsin (54 mm) in
5
0 mm potassium phosphate buffer, pH 7.6, at 278C. Time-resolved
C spectra (a) showloss of BAEE (resonance b), and growth of the BA
1
3
With the present sensitivity, a reaction can be detected at
enzyme concentrations on the order of 10 mm, which are
resonance (c). In (b) and (c), intensities from the spectra of the
reaction (*) are shown together with intensities from a reference
[15]
typical for biological tissue. Hyperpolarization selectively
spectrum (&; scaled as used for (d)). d) Linear regression of the ratio
enhances the signal of the polarized substrate by several
orders of magnitude over any background from nonpolarized
molecules. This property may prove particularly useful for
measurements in cells or in crude cell extracts, which contain
a large number of substances that would otherwise mask the
substrate to be observed, but where reaction kinetics may be
dramatically different from those under purified in vitro
À1
of intensities from (b), which yields kcat =12.1 s
.
substrate resonance (BAEE; Figure 2a,b), as well as the
appearance of product (BA; Figure 2a,c), can be seen. For a
quantitative analysis, the signal intensities were normalized
with reference values obtained from a measurement in the
absence of trypsin, which was scaled so that the resulting
intensity ratio extrapolated to t = 0 is equal to 1 (intercept in
Figure 2d). This procedure removes the effect of signal loss
arising from spin-lattice relaxation during the reaction time.
The normalized intensities are a linear function of time
[
16]
conditions. Additionally, the ability to observe individual
molecular sites by NMR spectroscopy, further enhanced by
1
3
the large chemical shift range of C nuclei, can give
simultaneous kinetic information on reactions that occur in
parallel. In more complicated situations, direct observation of
[
17]
(
Figure 2d), as is expected for a reaction where substrate
saturation transfer from an individually addressed nuclear
spin in the substrate to the product may also allow the
determination of reaction mechanisms by this method.
In conclusion, we have demonstrated that the increased
sensitivity provided by hyperpolarized NMR spectroscopy
enables time-resolved observation of enzymatic reactions
under near-physiological conditions. This approach is, how-
ever, not limited to enzyme kinetics; other applications
include the study of unidirectional chemical and biochemical
processes as diverse as polymerization reactions or protein
folding. DNP-enhanced time-resolved NMR spectroscopy is
valuable where conventional NMR spectroscopic observation
would require signal averaging, a situation that in practice is
concentration (3.3 mm) is much higher than the K value of
m
[
11] [13]
the enzyme (35.5 mm ). The rate of catalysis (kcat = 12.1 Æ
À1
1
s ) can be obtained directly from the slope of this line by
multiplication with the ratio of the substrate/enzyme concen-
tration. The error was estimated from the spread of values
obtained from different data sets, as well as from different
methods for integration of the signals. To validate the DNP
NMR measurements, we have compared the trypsin activity
with a measurement by UV/Vis spectrophotometry, which
À1
yielded a rate constant of 12.5 s for the batch of trypsin used
(
258C, pH 7.6), and is also in good agreement with a
[11]
published value under similar conditions. While we chose
5
236
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5235 –5237