13C-isotope labelling for the facilitated NMR analysis of a complex dynamic chemical system

Article abstract of DOI:10.1039/c1cc15295e

¹³C-isotope labelling is presented as a novel tool for the study of complex chemical systems. ¹³C-isotope labelling permits the quantification of all 26 members of a dynamic library from a single ¹³C NMR spectrum without the need for advanced instrumentation or sophisticated experimental protocols.

Full text of DOI:10.1039/c1cc15295e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 12476–12478
www.rsc.org/chemcomm
COMMUNICATION
13C-isotope labelling for the facilitated NMR analysis of a complex
dynamic chemical systemw
Marta Dal Molin, Giulio Gasparini, Paolo Scrimin, Federico Rastrelli* and Leonard J. Prins*
Received 25th August 2011, Accepted 28th September 2011
DOI: 10.1039/c1cc15295e
13C-isotope labelling is presented as a novel tool for the study of
complex chemical systems. 13C-isotope labelling permits the
quantification of all 26 members of a dynamic library from a
single 13C NMR spectrum without the need for advanced
instrumentation or sophisticated experimental protocols.
labels and sensitivity issues arising from low compound
concentrations are typically absent. However, as chemistry is
heading towards the study of increasingly complex chemical
systems, it becomes evident that sophisticated tools, such as
targeted isotope labelling, for studying complex structures
will turn out to be essential for NMR screening. In particular
systems relying on an imine-type exchange reaction, which
is one of the most frequently applied dynamic covalent
bonds,8,9 are highly suitable for 13C isotope labelling of the
iminic carbon for the following reasons. Firstly, the iminic
position is intrinsically advantageous because the chemical
shift of the iminic carbon is very sensitive to small structural
variations. Secondly, the widespread chemically-shifted Larmor
frequencies of 13C (up to 200 ppm and more) combined
with (proton) broadband decoupling will effectively minimise
the occurrence of signal overlap, even at moderate magnetic
fields. Thirdly, being an endogenous label, the 13C-isotope
does not affect the thermodynamic stability or the structure
of the (supra)molecule under investigation, something that
cannot be excluded when heteroatoms (such as 19F or 31P)
are inserted as exogenous NMR labels. Finally, the iminic
carbon atom can be labelled by exploiting a reasonably simple
chemistry.
Here, we show that 13C-labelling permits the straightforward
identification and quantification of each member of a chemical
system containing up to 26 components. This study was aimed
at determining how neighbouring spectator groups affect the
thermodynamic stability of a series of hydrazones and imines
(P1–4X) in which an intramolecular interaction exists between
a phosphonate and an ammonium group (Fig. 1).10 In prior
studies we have shown that this information is of relevance for
the development of a model that mechanistically mimics a serine
protease.11,12 The key issue is the self-selection of functional
groups by the phosphonate target, which acts as a transition
state analogue for the hydrolysis of a carboxylic ester. In
enzymes remote spectator groups may appear inert on first
sight. Yet, by subtly altering the conformation of the enzyme
they may actually play an important role in determining the
molecular recognition event occurring in the active site.13 We
were challenged by the idea of quantifying the contribution
of different spectator groups to the thermodynamic stability of
a single compound from a direct analysis of the dynamic
system and with a minimal amount of experiments. So far,
the application of dynamic combinatorial chemistry for the
Inspired by the astonishing achievements of Nature, chemists
are now involved in the construction of functional synthetic
systems able to match up to the complexity of natural systems.1
Not only this entails the self-assembly of supramolecular
systems,2 but it also involves the study of properties emerging
from artificial networks.3 Evidently, the achievable complexity
of supramolecular constructs and networks is strongly dependent
on the availability of analytical tools able to inform on the
structure of the supramolecule or the concentration of each
network component.
NMR spectroscopy is an attractive analytical tool for studying
complex chemical systems. It is a non-invasive technique thus
permitting the direct analysis of the complete mixture under
equilibration conditions. In addition, the signals are very
sensitive to small structural changes.4 However, in the case
of proton NMR spectroscopy this is counterbalanced by the
fact that the chemically-shifted Larmor frequencies of 1H vary
within a limited range of values (typically 10 ppm), which most
often results in serious signal overlap at average magnetic
fields (i.e. 400 MHz). Previously, we have shown that the use
of 1H–13C HSQC spectroscopy can alleviate this problem, but
high concentrations and advanced experimental protocols were
required to study a chemical system of moderate complexity
(8 components).5 The limited use of NMR spectroscopy as an
analytical tool within the context of systems chemistry is quite
remarkable, considering that NMR is widely accepted as
the tool for studying complex biomolecular structures.6
A
common route to sensitivity enhancement in protein NMR is
the insertion of isotopic labels (notably 15N and 13C) at key
positions in the molecular structure.7 Isotope labelling in small
organic molecules is far less popular, because their much lower
complexity simply does not require the insertion of isotope
Department of Chemical Sciences, University of Padova,
Via Marzolo 1, 35131 Padova, Italy.
E-mail: federico.rastrelli@unipd.it, leonard.prins@unipd.it
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c1cc15295e
c
12476 Chem. Commun., 2011, 47, 12476–12478
This journal is The Royal Society of Chemistry 2011

View Article Online
Fig. 1 Relative thermodynamic stabilities of the 13C-isotopically
labelled imines P1–4(A/A+) and hydrazones P1–4(H/H+) as deter-
mined from 13C NMR competition experiments (the asterisk indicates
a 13C nuclide). The symbols correspond to the indicated signals in Fig.
2 and 3. The relative thermodynamic stabilities of compounds P1–4X
are arbitrarily normalized on the energy level of the imines P1–4A,
which is the thermodynamically least stable species in each series. The
differences in free energy (DG; kJ molꢀ1) are calculated from experi-
ments performed in CD3OD at 298 K.
Fig. 2 (a) 13C NMR spectrum (150 MHz, 1H decoupled DEPT-90)
and (b) 1H NMR spectrum (600 MHz, 13C decoupled) of the mixture
of scaffolds P1–P4, both hydrazides H and H+ and both amines A and
A+ at thermodynamic equilibrium (CD3OD, 298 K). The symbols
refer to the structures in Fig. 1.
determination of quantitative structure–activity relationships
(QSAR) has received very little attention.14–16
The library we have chosen is particularly challenging for
two reasons. First, it includes two sub-libraries (imines and
hydrazones), which are expected to have very different intrinsic
stabilities.17 Second, the hydrazones are present as a mixture
of geometric isomers, which cannot be resolved by chromato-
graphic means. The direct analysis of such a complex system
under equilibration conditions would be impossible with any
technique other than NMR. Aromatic aldehydes P1–P4 were
prepared with various degrees of substitution of the aromatic
ring. Within this set, compound P1, lacking the phosphonate
group, serves as an internal reference to assess the intrinsic
stabilities of the formed imines and hydrazones. All compounds
P1–P4 contain a 13C labelled aldehyde group, which was easily
introduced in the ortho-position to a MOM-protected phenol via
quenching of the BuLi-activated aromatic ring with a slight
excess of 13C-DMF (see ESIw). Hydrazides H and H+ and
amines A and A+ were chosen as complementary components
for hydrazone and imine formation, respectively, because they
allowed an assessment of the importance of the dynamic covalent
bond in terms of rigidity and spacer length. 13C-labelling of the
iminic carbon provides a single characteristic resonance for each
compound in the 13C NMR spectrum which can now be
Fig. 3 Signal assignment based on the sequential addition of scaf-
folds P1–P4 (2.5 mM each) to a solution of hydrazide H (25 mM) in
CD3OD at 298 K (600 MHz). Scaffolds were added in the following
order: (a) P1 ( ); (b) P2 ( ); (c) P4 ( ); (d) P3 ( ).
In the 13C spectrum 18 species out of 26 were quantified in just
20 minutes, whereas a thorough quantification of all signals was
possible in 12 hours. Species present at concentrations down to
0.1 mM could be detected. Signals were assigned to each library
member in a straightforward manner by gradually increasing the
complexity of the mixture. Thus, scaffolds P1–P4 were added
sequentially to a solution containing an excess of either one of the
complementary components A, A+, H, or H+. This is illustrated
in Fig. 3a–d for the addition of scaffolds P1–P4 (2.5 mM each) to
hydrazide H (25 mM). Competition experiments were then
started by mixing these sub-libraries, followed by 13C NMR
spectroscopy. In order to gain the highest possible sensitivity,
we exploited polarization transfer by the use of the DEPT pulse
sequence18 (see ESIw for details on the advantages of the DEPT
sequence in our case).
1
recorded with sensitivity close to that of a H NMR spectrum.
The characteristic resonances for the iminic carbon fall within the
140–170 ppm range compared to the 8–9.5 ppm range for the
iminic proton. This much larger interval permits the study of
mixtures of far higher complexity. This is evidenced by the
observation that all 26 components (including isomers) of the
library obtained by mixing all scaffolds P1–P4, both hydrazides H
and H+ and both amines A and A+ gave well-resolved signals
in the 13C NMR spectrum (Fig. 2a). The overlapping signals in
the 1H NMR spectrum indicate that the complexity of this system
is well above the resolution limit for 1H NMR spectroscopy.
The full system analysis permitted a quantification of all 26
species present, but the obtained data were difficult to interpret
in terms of relative thermodynamic stability. These problems
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 12476–12478 12477

View Article Online
do not originate from the analytical technique itself, but rather
from the fact that near stoichiometric amounts of hydrazides
were required to allow the simultaneous coexistence of the
thermodynamically much less stable imines.17 The absence of a
buffering amount of free hydrazide introduces a cross-talk
between the various equilibria in the sense that the preferred
selection of H+ over H by one scaffold will also affect the
hydrazone ratio for the other scaffolds. Such a cross-talk is
detrimental for a proper quantification of the relative thermo-
dynamic stability of each species. For that reason, the analysis
was repeated on the separate hydrazone and imine libraries,
containing 18 and 8 components, respectively. In addition,
previous NMR study on a similar non-labelled system,
13C-labelling permits an increase in complexity from 8 to 26
signals without the need for bidimensional 1H–13C HSQC
spectroscopy, specialized NMR cryo-probes or complex
data-processing protocols.5 Most importantly, much lower
compound concentrations were required (0.1 mM compared
to 10 mM) and a total of 18 out of 26 signals could be
quantified in just 20 minutes, implying that the kinetics of
these systems can be studied with enhanced time resolution,
too. It is anticipated that the combination of advanced
NMR protocols with 13C-labelling will enable the analysis of
mixtures of even higher complexity than reported here. As we
have shown, the ability to quantify the concentrations of all
components of the system provides a unique possibility to
determine quantitative structure–activity relationships in a
straightforward manner.
individual competition experiments between P1–4A and P1–4
H
were used to correlate the energy scales of the imines and
hydrazones (see ESIw). From these studies it emerges in a
straightforward manner how substituents on the scaffolds
P1–P4 affect the thermodynamic stability of the corresponding
hydrazones and imines (Fig. 1). The interpretation of these
results requires some considerations. First, in contrast to our
previous studies using P1 and P2,5,11,19 the competition experi-
ments involving different scaffolds now occur simultaneously
in the same NMR tube in a parallel set of equilibria. On one
hand, this guarantees identical screening conditions but, on
the other hand, it may complicate the analysis in the case of
crosstalk between the platforms (see above). This appears not
to be the case, since the measured differences in thermodynamic
stabilities between hydrazones P1H–P1H+ (0.0 kJ molꢀ1) and
P2H–P2H+ (0.8 kJ molꢀ1) are nearly identical to those
obtained previously from separate experiments (0.2 kJ molꢀ1
and 0.7 kJ molꢀ1).5,11 Second, it should be noted that these
experiments do not reveal information on the thermodynamic
stabilities between the different scaffolds.z In fact, all energy
levels have been arbitrarily normalized on imines P1–4A, which
are taken as a reference within each scaffold series. Third, in
this preliminary analysis we have not taken hydrazone isomerism
into account and have treated each component P1–4X as a single
species.y
Financial support from the European Research Council under
the European Community’s Seventh Framework Programme
(FP7/2007–2013)/ERC Starting Grant agreement no. 239898
and COST (CM0703) is acknowledged.
Notes and references
z In the presence of an excess of amines and hydrazides there is
no competition between scaffolds for imine- or hydrazone bond
formation.
y A full account including a treatment of hydrazone isomerism and
all energetic contributions involved will be reported in due course.
Preliminary results indicate that steric hindrance in scaffold P4 induces
the formation of two non-planar isomers in P4H in addition to the
regular two isomers originating from E/Z isomerism of the amide
bond, which are common to all hydrazones.
1 G. M. Whitesides and R. F. Ismagilov, Science, 1999, 284, 89–92.
2 J. M. Lehn, Angew. Chem., Int. Ed. Engl., 1990, 29, 1304–1319.
3 R. F. Ludlow and S. Otto, Chem. Soc. Rev., 2008, 37, 101–108.
4 W. C. Pomerantz, E. B. Hadley, C. G. Fry and S. H. Gellman,
ChemBioChem, 2009, 10, 2177–2181.
5 G. Gasparini, B. Vitorge, P. Scrimin, D. Jeannerat and L. J. Prins,
Chem. Commun., 2008, 3034–3036.
6 J. Cavanagh, W. J. Fairbrother, A. G. Palmer III, M. Rance and
N. J. Skelton, Protein NMR spectroscopy: principles and practice,
2nd edn, Academic Press, 2007.
7 L. P. McIntosh and F. W. Dahlquist, Q. Rev. Biophys., 1990, 23,
1–38.
8 S. J. Rowan, S. J. Cantrill, G. R. L. Cousins, J. K. M. Sanders and
J. F. Stoddart, Angew. Chem., Int. Ed., 2002, 41, 898–952.
9 C. D. Meyer, C. S. Joiner and J. F. Stoddart, Chem. Soc. Rev.,
2007, 36, 1705–1723.
10 L. J. Prins and P. Scrimin, Angew. Chem., Int. Ed., 2009, 48,
2288–2306.
11 G. Gasparini, L. J. Prins and P. Scrimin, Angew. Chem., Int. Ed.,
2008, 47, 2475–2479.
12 G. Gasparini, M. Dal Molin and L. J. Prins, Eur. J. Org. Chem.,
2010, 2429–2440.
13 A. Fersht, Structure and Mechanism in Protein Structure,
W.H. Freeman and Company, 1999.
14 M. G. Woll and S. H. Gellman, J. Am. Chem. Soc., 2004, 126,
11172–11174.
15 M. K. Chung, K. Severin, S. J. Lee, M. L. Waters and
M. R. Gagne, Chem. Sci., 2011, 2, 744–747.
16 R. F. Ludlow, J. Liu, H. X. Li, S. L. Roberts, J. K. M. Sanders and
S. Otto, Angew. Chem., Int. Ed., 2007, 46, 5762–5764.
17 F. A. Carey and R. J. Sundberg, Advanced Organic Chemistry,
vol. A, Springer, 2008.
18 D. M. Doddrell, D. T. Pegg and M. R. Bendall, J. Magn. Reson.,
1982, 48, 323–327.
This analysis provides a quantitative insight into the correlation
between structure and thermodynamic stability. A first glance
immediately reveals some key features. First, the anticipated
stabilizing interactions between phosphonate and ammonium
groups in both P2A+ (2.6 kJ molꢀ1) and P2H+ (0.8 kJ molꢀ1),
which is further enhanced upon the insertion of a methyl group
in the ortho-position to the phosphonate group (scaffold P3).
Second, the anticipated much higher intrinsic stability of hydra-
zones compared to imines. Nonetheless, after a closer look it
emerges that the overall picture is actually rather complicated. For
example, substituents affect the intrinsic stability of imines and
hydrazones (DGP H–P A = 16.8 kJ molꢀ1 against 13.6 kJ molꢀ1
2
2
for DGP H–P A) and the presence of the methoxy substituent in
1
1
P4 has a destabilizing effect in particular on the hydrazones P4H
and P4H+.y It is evident that such an interplay between
stabilizing and destabilizing energetic contributions always plays
a role in determining the composition of a dynamic system.
However, such information can only emerge if the thermodynamic
stability of all system components can be determined.
In summary, these results show that 13C isotopic labelling of
a dynamic library is a very attractive tool to study complex
dynamic systems by NMR spectroscopy. Compared to a
19 G. Gasparini, F. Bettin, P. Scrimin and L. J. Prins, Angew. Chem.,
Int. Ed., 2009, 48, 4546–4550.
c
12478 Chem. Commun., 2011, 47, 12476–12478
This journal is The Royal Society of Chemistry 2011