15N magnetic resonance hyperpolarization via the reaction of parahydrogen with 15N-Propargylcholine

Article abstract of DOI:10.1021/ja209884h

¹⁵N-Propargylcholine has been synthesized and hydrogenated with para-H₂. Through the application of a field cycling procedure, parahydrogen spin order is transferred to the ¹⁵N resonance. Among the different isomers formed upon hydrogenation of ¹⁵N- propargylcholine, only the nontransposed derivative contributes to the observed N-15 enhanced emission signal. The parahydrogen-induced polarization factor is about 3000. The precise identification of the isomer responsible for the observed ¹⁵N enhancement has been attained through a retro-INEPT (¹⁵N-¹H) experiment. T₁ of the hyperpolarized ¹⁵N resonance has been estimated to be ca. 150 s, i.e., similar to that reported for the parent propargylcholine (144 s). Experimental results are accompanied by theoretical calculations that stress the role of scalar coupling constants (J_HN and J_HH) and of the field dependence in the formation of the observed ¹⁵N polarized signal. Insights into the good cellular uptake of the compound have been gained.

Full text of DOI:10.1021/ja209884h

Article
pubs.acs.org/JACS
15N Magnetic Resonance Hyperpolarization via the Reaction of
15
Parahydrogen with N-Propargylcholine
Francesca Reineri,† Alessandra Viale,† Silvano Ellena,† Diego Alberti,† Tommaso Boi,†
Giovanni Battista Giovenzana,‡ Roberto Gobetto,† Samuel S. D. Premkumar,† and Silvio Aime*,†
†Dipartimento di Chimica IFM, University of Torino, V. P. Giuria 7, 10125 Torino, Italy
‡DiSCAFF & DFB Center, Universita
̀
degli Studi del Piemonte Orientale “A. Avogadro”, Via Bovio 6, I-28100 Novara, Italy
S
* Supporting Information
ABSTRACT: 15N-Propargylcholine has been synthesized and hydro-
genated with para-H2. Through the application of a field cycling procedure,
parahydrogen spin order is transferred to the 15N resonance. Among the
different isomers formed upon hydrogenation of 15N-propargylcholine,
only the nontransposed derivative contributes to the observed N-15
enhanced emission signal. The parahydrogen-induced polarization factor is
about 3000. The precise identification of the isomer responsible for the
observed 15N enhancement has been attained through a retro-INEPT
(15N−1H) experiment. T1 of the hyperpolarized 15N resonance has been
estimated to be ca. 150 s, i.e., similar to that reported for the parent
propargylcholine (144 s). Experimental results are accompanied by
theoretical calculations that stress the role of scalar coupling constants (JHN and JHH) and of the field dependence in the
formation of the observed 15N polarized signal. Insights into the good cellular uptake of the compound have been gained.
only, by means of the SABRE method,15 while no examples of
INTRODUCTION
■
15N hyperpolarization by parahydrogen induced polarization
(PHIP)16 on biocompatible substrates have been reported.
The possibility of generating 15N hyperpolarized resonances
that can be used for in vivo studies by means of PHIP
procedure would be outstandingly interesting. The major
limitation appears related to the availability of α-unsaturated
substrates that, upon hydrogenation, would bring the added
parahydrogen protons two and three bonds away from the 15N
center. Unfortunately, the hydrogenation of such systems
invariantly leads to the cleavage of the N−C bond and the
release of the hydrogenated moiety. Thus, one has to turn to
molecules containing an sp3 carbon between the unsaturated
group and the 15N atom. To test the feasibility of 15N
hyperpolarization for molecules containing the latter structural
arrangement, the propargylcholine derivative 1 has been
synthesized and its reaction with parahydrogen has been
tested. Compound 1 mimics the parent choline molecule and it
might have potential in vivo applications.
Magnetic resonance hyperpolarization techniques are under
intense scrutiny as the possibility of overcoming the sensitivity
limitations may open a number of innovative applications, in
chemistry and biology. The driving force is represented by the
development of highly innovative magnetic resonance imaging
(MRI) procedures based on the detection of key metabolites
that directly report on the specific steps of cellular machinery
(metabolic imaging). Polarized states lifetime is determined by
relaxation rate of the involved resonance and only function-
alities characterized by long relaxation time are useful for these
applications.1−4 13C resonances, usually belonging to carbonyl
groups in small-sized molecules, have been considered because
representatives of this class of molecules are easily formed in
key metabolic pathways.5−7 To now 13C-pyruvate is the most
used substrate, being a key chemical in the TCA metabolic
pathway, and the detection of its transformation allows one to
obtain information about the deviation from the healthy state,
at the cellular level.8
Among the other low γ heteronuclei particularly important in
biological systems, 15N is the most appealing one, since it is
often characterized by a very long relaxation time thanks to the
highly symmetrical environment it often owns in the molecules
of interest.
RESULTS AND DISCUSSION
■
Synthesis and Characterization of N-Propargylcholine
(1). N-Propargylcholine tosylate was prepared by the synthetic
procedure depicted in Scheme 1. First 2-hydroxyethanolamine
was methylated by reaction with formaldehyde and formic acid,
and then N,N-dimethylethanolamine was reacted with prop-
Up to now, few 15N-containing substrates have been
hyperpolarized, namely, 15N-urea,9 some amino acids,10 and,
most importantly, due to its long T1 value (120 s), 15N-
choline.11−14 All these molecules have been polarized by the
DNP method. To the best of our knowledge, 15N hyper-
polarization with parahydrogen has been achieved on pyridine
Received: October 20, 2011
Published: June 4, 2012
© 2012 American Chemical Society
11146
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
Chart 1. N-Propargylcholine Tosylate
Scheme 2. Hydrogenation of N-Propagylcholine Affords Five
Products
a
Scheme 1. Synthetic Procedure for the Preparation of
Deuterated 15N-Propargylcholine
a
Compound a is formed at the higher yields (45%).
argyltosylate to yield N-propargylcholine. For the preparation
of the 15N-enriched compound, 15N-ethanolamine was used.
Deuteration of the methylenic group adjacent to the nitrogen
atom was achieved by using deuterated propargyltosylate,
which was synthesized by reaction of lithium monoacetylide
with deuterated paraformaldehyde to yield deuterated prop-
argyl alcohol. The successive reaction of the alcohol with tosyl
chloride yields a product well-suitable for the hydrogenation
reaction to be carried out either in water or in organic solvent.
15N T1 measurements showed that, as for choline, the
quaternary nitrogen atom is characterized by a very long T1
(104 s in water at 14 T). Deuteration of the adjacent
methylenic group caused a further T1 increase to 144 s. On the
basis of the close similarity between the nitrogen surrounding
in N-propagylcholine and its hydrogenated product, it is
surmised that the latter would show similar relaxation times.
Such long relaxation time is a really favorable parameter for
development of MRI protocols based on the use of the
hyperpolarized choline derivative, since it guarantees that both
parahydrogenation and further handlings can be carried out
without significant polarization loss.
Figure 1. 1H NMR spectrum of the products deriving from the
hydrogenation of N-propargylcholine [acetone-d6/CD3OD 4:1, room
temperature (RT), 9.4 T]. tos = tosylate; coe = cyclooctene; s =
solvent.
More insight into the transposition path was gained by
2
acquiring H NMR spectrum of the products of the reaction
between N-propargylcholine and D2. Deuterium incorporation
in b/b′ has been observed only at the 2- and 3-positions to
indicate that the added deuterons are not involved in the
transposition process. Thus, the process deals with the transfer
of one of the two H(1) to the 3-position. This may take place
immediately after the hydrogenation reaction, while the just
formed alkene is still coordinated to the metal complex, via a π-
allyl intermediate as depicted in Scheme 3.17 A carbene-
mediated mechanism can be ruled out since it would yield
incorporation of both the deuterons on the 3-position.
Hydrogenation of N-Propargylcholine (1). Hydro-
genation of N-propargylcholine (1), carried out either in a
mixture of acetone and methanol or in water, catalyzed by
[Rh(diphenylphosphinobutane)(cyclooctadiene)][BF4] and
[Rh(1,4-bis[(phenyl-3-propanesulfonate)phosphine]butane)-
(norbornadiene)][BF4] catalysts, respectively affords five
different products (Scheme 2, Figure 1). Their characterization
PHIP Effects in the 1H NMR Spectrum. PHIP effects have
been observed in the 1H spectrum upon reaction with
parahydrogen of N-propargylcholine. The absorption/emission
pattern is similar for the three considered isotopomers (i.e., N-
1
has been pursued by NMR spectroscopy by means of H-
1
COSY and H−15N HMBC experiments.
Compound a is the main product (about 45%) of both
experimental conditions. b and b′ derivatives are formed by
transposition of the double bond from a, resulting in the cis
(b′) and trans (b) isomers of the alkene in which the olefin is
adjacent to N. Due to the intrinsic instability of such
configuration, b and b′ partially undergo propargyl elimination,
yielding dimethylethanolamine and partially deuterated
CH3CH2CH(OCH3)2 (in acetone/CD3OD) or CH3CH2CHO
(in D2O). The fifth reaction product (c) corresponds to the
complete hydrogenation of the substrate (about 10%).
Scheme 3. Double Bond Transposition Leading from a to b/
b′
11147
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
propargylcholine, 15N-labeled N-propargylcholine, and deuter-
ated, 15N-labeled N-propargylcholine) and does not change
upon passing from organic solvent to water.
reported earlier by Bargon et al.18 Scalar coupling values
between parahydrogen protons and nitrogen in the observed
reaction products are all quite small. The protons in the 3-
position (Scheme 2) are four bonds away from the nitrogen
nucleus and their J coupling value with 15N can be
approximated to zero. The protons in the 2-position experience
The 1H-spectrum of the products obtained from the reaction
of N-propargylcholine with para-H2 in acetone-d6/CD3OD is
reported in Figure 2. It clearly shows enhanced signals for a, b,
3
a JHN coupling that is about 2 Hz for b′-like systems, having a
trans configuration, 3 Hz for a-like systems, and 5 Hz for b-like
systems, having cis configuration.12,19 It has also been reported
14
2
1
that J coupling between 15N and H are smaller than 1 Hz.
On this basis it can easily be shown that the expected
enhancement on 15N would be much smaller than the one
usually observed on 13C. As reported in Figure 3b the
maximum theoretical enhancement attainable on 15N is
expected to be 1 order of magnitude lower than the one
observed on 13C. Moreover, the splitting between the
absorption and emission peaks is smaller, thus contributing to
the mutual cancellation of the two polarized peaks. As a matter
of fact, no 15N-polarization was observed following a para-
hydrogenation ALTADENA experiment.16,20
It is also well-established21 that heteronuclear hyperpolariza-
tion can be manipulated by low magnetic field cycle. This
procedure is based on the fact that at 0.5 G (earth field) the
Larmor frequencies of magnetically active nuclei are still
different, so the weak coupling condition between heteroatoms
applies. On the contrary, when the magnetic field is further
lowered to 1 mG, isotropic mixing between all the magnetically
active nuclei is obtained.
The variations of the spin states populations during field
cycle can be treated using the product operator formalism and a
three spins system (HH′X). This system is quite a simplification
of the real spin systems, as it does not take into account all the
protons of the product molecules. However, it allows one to
demonstrate how polarization intensity of the heteronucleus
can be changed upon field cycle application.
1
Figure 2. H NMR spectrum of the products of the reaction of N-
propargylcholine with para-H2 (one scan, RT, acetone-d6/CD3OD 4:1,
9.4 T).
1
b′, and c products. Among them, the most intense H-PHIP
effects are those ones arising from b/b′, in spite of their lower
concentration. From the transposition mechanism it derives
that on product b/b′ parahydrogen protons are added in
positions 3 and 2, and then polarization is transferred to the
methyl protons. Therefore, the highest proton polarization is
observed on the methyl signal.
15N-PHIP. It is well-established that hyperpolarization is
transferred from parahydrogen to heteronuclei through scalar
coupling.18 Polarized spin order IzHINz can be directly obtained in
parahydrogenation experiments, and enhancement is a function
of the ratio R = (JNH − JNH′)/JHH′, where H and H′ are the two
parahydrogen protons, H being the most strongly coupled with
15N. As can be observed in Figure 3a, the heteronuclear
polarization (P) is null when the two protons experience the
same coupling with the heteroatom, it tends to zero when J
coupling between the two protons vanishes, and it is a
maximum for (JNH − JNH′)/JHH′ ≈ √8, in accordance with that
It has already been reported that, after parahydrogen addition
to a molecule, the parahydrogen density operator is perturbed
and then a new steady state is reached.22 The present study
starts from the steady-state condition which is obtained on an
HH′X spin system when the reaction with parahydrogen is
carried out at the earth magnetic field (EF)22
EF
ZZHH′
EF
IzHIzH′ + c
(ZQ HH′
σEF = E/4 + c
) (ZQ
)
x
HH′
x
EF
(ZQ HH′
EF
ZZHX
+ c
) (ZQ )y + c
(IzH − IzH′)IzX
HH′
(1)
y
where ZQHH′ is zero quantum coherence between the two
protons.
When the magnetic field is nonadiabatically lowered (nearly
1 mG), the equilibrium state is suddenly perturbed by the fully
coupled Hamiltonian (ZF denoting zero field)
Figure 3. (a) Polarization transfer to proton-heteronucleus spin order
HX
IzHIzX as a function of coefficient R = |JHX−JH′X|/JHH′. (b) c
zz
HZF = −νH(IzH + IzH′) − νNIzX + J [IzHIzH′ + ZQ Hx H′
]
̂
comparison between maximum theoretical enhancement that can be
HH′
HX
achieved on a nucleus X for which R = 1.7 (corresponding to c
=
zz
+ JHX[IzHIzX + ZQ HX] + JH′X[IzH′IzX + ZQ H′X
]
0.8; blue line) and another with R = 0.2 (corresponding to cHzzX = 0.08;
red line). The first is related to J coupling values typical of a carbonyl
group adjacent to double bond, the second one is obtained using J
coupling values of propargylcholine hydrogenation product a.
x
x
(2)
after which a new steady state is reached
11148
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
HH′
HH′
σZF = E/4 + czHzH′IzHIzH′ + c (ZQ HH′)x + c
ZQ x
ZQ y
HX
HX
(ZQ HH′)yczHzX H X
Iz Iz + c (ZQ HX)x + c (ZQ HX)
y
ZQ x
ZQ y
H′X H′ X
H′X
H′X
+ c Iz Iz + c (ZQ H′X) + c (ZQ H′X
)
y
zz
ZQ x
ZQ y
+ czH(IzH − IzH′) + czXIzX
(3)
All the c coefficients are a function of J coupling and of the
1
frequency difference between H and the heteronucleus, which
is related to the magnetic field strength. In our case, the amount
of longitudinal magnetization on 15N (coefficient czN) has been
calculated as a function of magnetic field (Figure 4).
Figure 5. cNz as a function of transport time from 1 to 500 mG.
Molecule a (JHH = 10 Hz, JHN = 3 Hz, JH′N = 0) was used for these
simulations. Diagrams were obtained starting from 0 mG: (a) slow
transport (5 s), ⟨cNz ⟩B=0.5G = −0.15; (b) fast transport (1 s) ⟨czN⟩B=0.5G
=
−0.09. Diagrams were calculated starting from 2.5 mG: (c) slow
transport (5 s) ⟨cNz ⟩B=0.5G = −0.45; (d) fast transport (1 s) ⟨czN⟩B=0.5G
−0.45.
=
(500 mG, earth field) is strongly dependent on the speed of
transport. In fact, it can be observed from Figure 5 that, if
transport takes 5 s instead of 1 s, the absolute value of cNz is
almost doubled. On the contrary, if the starting B0 coincides
with that at which heteronuclear magnetization is a maximum
(2.5 mG for molecule a), the speed of transport does not
influence the amount of longitudinal 15N magnetization, the
value of which remains close to that achieved at low field.
It is evident that, in order to optimize the amount of
longitudinal polarization on the heteronucleus, the magnetic
field strength must be exactly measured; that is not a trivial task
in the μ-metal cylinder that has been used for our measure-
ments. However, is important to notice that longitudinal
Figure 4. The amount of longitudinal magnetization (cNz ) at very low
magnetic field (from 1 to 10mG) as a function of the magnetic field
strength. These values have been calculated for product a (blue circles)
and product b (red) having (a) JHH = 10 Hz, JHN = 3 Hz, JH′N = 0 Hz
and (b) JHH = 4 Hz, JHN = 2 Hz, JH′N = 0 Hz, respectively.
After that the steady-state condition at zero field has been
reached (a few seconds), the magnetic field shield is slowly
(adiabatically) removed. During this adiabatic passage from
zero to earth field the density operator is subjected to a time-
dependent Hamiltonian
polarization (INz ) is, in all these cases, higher than the calculated
HN
value of polarized spin order (c ).
zz
Following the application of field cycle to parahydrogenated
15N-propargylcholine, polarized 15N is indeed obtained (Figure
6).
H
H(t) = −νH(t)(Iz + IzH′) − νX(t)IzX + J IHIH′
̂
HH
The chemical shift of the enhanced peak (56.34 ppm) does
not allow one to discriminate whether it pertains to a and/or b/
b′ and/or c. Furthermore, as the magnetic field strength into
the μ-metal cylinder is not exactly known, it is not possible to
know, in principle, which is the most polarized molecule.
Surely, the 15N signal for d (36.36 ppm) is not polarized.
Because of the very small difference between the 15N chemical
shifts of a, b/b′, and c, it is not possible to distinguish which of
the different species leads to the observed heteronuclear
polarization. To tackle this task the selective inverse INEPT
sequence has been applied to the parahydrogenated products.
This experiment allows one to identify those protons that are
scalarly coupled to the polarized 15N nucleus.12 In the resulting
spectrum (Figure 7), all the signals related to a are present,
while no signal can be observed for b/b′ and c. This finding
clearly indicates that the polarization transfer yielding the 15N
enhancement has occurred mainly in a.
+ JHXIHIX + JH′XIH′IX
(4)
where the Larmor frequencies are time-dependent while J
coupling is constant.
Density matrix during the adiabatic passage has been
numerically calculated by applying a time-dependent Liou-
ville−Von Neumann equation
̂
̂
σ(t+Δt) = exp(−2iπH(t+Δt)Δt)σ(t) exp(2iπH(t+Δt)Δt)
(5)
From the calculated time-dependent density operator, longi-
tudinal magnetization of the heteronucleus, represented by the
coefficient cNz , has been extracted and reported as a function of
time (Figure 5).
Simulations of the adiabatic transport have been carried out
for molecule a. If the starting magnetic field is zero, it can be
derived from Figure 4 that the absolute value of longitudinal
magnetization on 15N is minimum. In this case, the amount of
longitudinal magnetization obtained at the end of transport
The longitudinal relaxation time of molecule a has been
estimated by following the 15N polarization decay upon the
application of small flip angles to acquire the enhanced signals
11149
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
the hydrogenation product mixture (15, 0.95, and 0.36 mM)
for time lags increasing from 10 to 480 min. Figure 8 reports a
Figure 6. 15N spectrum of parahydrogenated 15N-propargylcholine.
(a) single-scan spectrum recorded immediately after parahydrogena-
tion and field cycling and (b) 80-scan spectrum recorded after
relaxation (acetone-d6/CD3OD 4:1, RT, 9.4 T). The signal enhance-
ment is about 3000 times. As the magnetic field profile is not exactly
known inside the cylinder, polarization may be further increased by
using an electromagnetic field shield.
Figure 8. Cellular uptake of hydrogenation products derived from
parahydrogenation of propargylcholine (1 mM total concentration of
product molecules). Bottom spectrum: MCF-7 cells extract in normal
grow medium. Top spectrum: cells extract after the addition of the
mixture of choline derivatives added to the medium. Middle spectra:
cell extracts at increasing incubation time. Methyl signals of
phosphocholine and its parent compounds are indicated with P-
Chol and P-(a,b,b′,c).
1
selection of H NMR spectra of cellular extracts to summarize
the results of the cellular uptake experiments. Upon increasing
the incubation time, the progressive growth of a peak at 3.03
1
ppm is observed. This peak is not present in the control H
NMR spectrum corresponding to cell lysate. In the latter
solution (top trace in Figure 8), the methyl signal of a, b, b′,
and c is detected at 3.01 ppm when the hydrogenation products
mixture is added. A peak at 3.01 ppm is also observed at long
incubation times (see fifth and sixth traces from bottom in
Figure 8). On this basis, the peak at 3.03 ppm has been
assigned to the phosphocholine derivatives of the a, b, b′, c
choline mimics mixture (the chemical shift separation is the
same order of magnitude as that observed between choline and
phosphocholine, at 3.08 and 3.10 ppm, respectively; see the
Supporting Information). The observed behavior indicates that,
upon entering the cells, the a, b, b′, c choline mimics are readily
phosphorilated and only at later times the signal for the
unphopshorilated derivatives is detected.23
Although the methyl signal is the same for all the molecules
derived from hydrogenation of propargylcholine, the percent-
age of compound a with respect to other hydrogenation
products can be derived from the intensity of the olefinic signal
at 5.47 ppm with respect to the methyl one (see Supporting
Information). From this comparison one may draw the
conclusion that the relative percentage of this compound is
maintained in the cells extract, thus suggesting that the uptake
of all the hydrogenation products takes place at about the same
rate.
Figure 7. Inverse INEPT spectra of parahydrogenated 15N-
propargicholine-d2. (a) Inverse INEPT spectrum of the 15N hyper-
polarized molecule: signals from H(2) and H(3), methylic and
methylenic protons of product a; (b) inverse INEPT spectrum
recorded after 15N polarization has returned to equilibrium; (c)
1
standard H NMR single-scan spectrum at equilibrium (acetone-d6/
CD3OD 4:1, RT, 14 T).
at various time intervals. This procedure afforded a T1 value of
ca. 140 s, which is similar to that measured for the starting
alkyne molecule, as expected.
In Vitro Essays on Cell Cultures. In order to get some
more insights into the potential of 15N-propargylcholine for
further in vivo imaging studies, we went to carry out cellular
uptake experiments using the hydrogenation product mixture
containing a (45%), b and b′ (24%), c (9%), and d (20%).
Breast cancer cells (MCF7) have been chosen for this
experiment, as it has been reported23 that either choline
transport and choline kinase activity is up-regulated in this
tumor cell line. Moreover, it has also been demonstrated that
the rate of choline phosphorylation, in these cells, is more than
2 orders of magnitude faster than the rate of transport.
Uptake experiments were carried out on cells seeded in 60
mm Pedri dishes and incubated with different concentrations of
The amount of compound a taken up by cells is reported in
Figure 9. From these data, using the values obtained for [a] =
11150
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
between earth and zero field. Among the different isomeric
structures that form upon hydrogenation of propargylcholine, it
has been shown that only one molecule (a, Scheme 1) is
effectively 15N polarized under the employed experimental
setup. Its identification has been possible through a retro-
INEPT experiment that has allowed polarization transfer from
1
15N to the H resonances.
In the investigated system the 15N resonance maintains a
very long T1 value in analogy with that previously reported for
the parent choline.
Cellular uptake experiments have demonstrated that internal-
ization of the choline derivatives takes place by the same
membrane transporters as choline. Furthermore, the concen-
tration of the polarized molecule a has been evaluated to be
enough to be observed in tissues and in particular in breast
cancer cells, for which the altered choline metabolism and
uptake have been previously demonstrated.
Figure 9. MCF-7 uptake experiments. Although the experiments have
been carried out using the a, b, b′, c hydrogenation products mixture,
the reported quantities (in molecules/cell) refers to compound a (45%
in the hydrogenation mixture, see text). Dashed-dotted line, triangles
[a] = 6.57 mM; dashed line and squares, [a] = 0.5 mM; straight line
and dots, [a] = 0.2 mM.
ASSOCIATED CONTENT
■
0.5 mM, it can be derived that, after 1 min, the concentration of
internalized a is 1 × 108 molecules/cell. In solid tissues, the
number of cells/mL is on the order of 3 × 108; thus, compound
a reaches the concentration of 0.5 mM after 1 min. Considering
that the attainable polarization is 3000 times and that the
relaxation rate is 140 s, after 1 min enhancement is about 1900
times. As a solution containing 1 M 15N-choline gives a S/N of
10, we can say that the amount of internalized a is enough to be
observed using the hyperpolarized product.
S
* Supporting Information
Experimental procedures, synthesis of substrate molecules, and
calculations of the time-dependent density operator during field
cycling using an on-purpose function written with Matlab 7.0.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
silvio.aime@unito.it
The cellular uptake of compound a appears to be effected, to
some extent, by the concentration of the choline in the
incubation medium. Figure 10 shows that, when [a] = 0.2 mM
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We gratefully acknowledge Regione Piemonte (bando POR-
FESR 2007, Asse 1, Misura I.1.1) for financial support.
■
REFERENCES
■
(1) Mansson, S.; Johansson, E.; Magnusson, P.; Chai, C. M.;
Hansson, G.; Petersson, J. S.; Stahlberg, F.; Golman, K. Eur. Radiol.
2006, 16, 57−67 and references therein.
(2) Viale, A.; Reineri, F.; Santelia, D.; Ellena, S.; Gobetto, R.; Aime, S.
Q. J. Nucl. Med. Mol. Imag. 2009, 53, 604−617 and references therein.
(3) Viale, A.; Aime, S. Curr. Op. Chem. Biol. 2010, 14, 90−96 and
references therein..
(4) Gallagher, F. A.; Kettunen, M. I.; Brindle, K. M. Prog. Nucl. Magn.
Reson. Spectrosc. 2009, 55, 285−295 and references therein.
(5) Kohler, S. J.; Yen, Y.; Wolber, J.; Chen, A. P.; Albers, M. J.; Bok,
R.; Zhang, V.; Trop, J.; Nelson, S.; Vigneron, D. B.; Kurhanewicz, J.;
Hurd, R. E. Magn. Res. Med. 2007, 58, 65−69.
(6) Nelson, S. J.; Vigneron, D.; Kurhanewicz, J.; Chen, A.; Bok, R.;
Hurd, R. Appl. Magn. Reson. 2008, 34, 533−544.
Figure 10. Straight line, dots: cellular uptake rate of a ([a] = 0.2 mM)
incubated with choline (0.3 mM) in the grow medium. Dashed line,
squares: cellular uptake with [a] = 0.2 mM, without choline.
(7) (a) Chekmenev, E. Y.; Hovener, J.; Norton, V. A.; Harris, K.;
Batchelder, L. S.; Bhattacharya, P.; Ross, B. D.; Weitekamp, D. P. J.
Am. Chem. Soc. 2008, 130, 4212−4213. (b) Reineri, F.; Viale, A.;
Ellena, S.; Boi, T.; Daniele, V.; Gobetto, R.; Aime, S. Angew. Chem., Int.
Ed. 2011, 32, 7350−7353.
(8) Albers, M. J.; Bok, R.; Chen, A. P.; Cunningham, C. H.; Zierhut,
M. L.; Yi Zhang, V.; Kohler, S. J.; Tropp, J.; Hurd, R. E.; Yen, Y.;
Nelson, S. J.; Vigneron, D. B.; Kurhanewicz, J. Cancer Res. 2008, 68,
8607−8615.
(9) Golman, K.; Ardenkjaer-Larsen, J. H.; Petersson, J. S.; Mansson,
S.; Leunbach, I. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 10435−10439.
(10) Hall, D. A.; Maus, D. C.; Gerfen, G. J.; Inati, S. J.; Becerra, L. R.;
Dahlquist, F. W.; Griffin, R. G. Science 1997, 276, 930−932.
and [choline] = 0.3 mM, the observed uptake of a is
significantly decreased. Thus, one may conclude that a enters
into MCF-7 cells by means of the same transporters used by
choline.24
CONCLUSIONS
■
This work has shown that it is possible to hyperpolarize the 15
N
resonance of an unsaturated derivative of choline by reaction
with parahydrogen if the spin order derived from parahydrogen
is changed into net 15N polarization by magnetic field cycling
11151
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152

Journal of the American Chemical Society
Article
(11) Cudalbu, C.; Comment, A.; Kurdzesau, F.; van Heeswijk, R. B.;
Jannin, S.; Denisov, V.; Zirik, D.; Gruetter, R. Phys. Chem. Chem. Phys.
2010, 15, 5818−5823.
(12) Gabellieri, C.; Reynolds, S.; Lavie, A.; Payne, G. S.; Leach, M.
O.; Eykyn, T. R. J. Am. Chem. Soc. 2008, 130, 4598−4599.
(13) Sarkar, R.; Comment, A.; Vasos, P. R.; Jannin, S.; Gruetter, R.;
Bodenhausen, G.; Hall, H.; Kirik, D.; Denisov, V. P. J. Am. Chem. Soc.
2009, 131, 16014−16015.
(14) Pfeilstikcer, J. A.; Ollerenshw, J. E.; Norton, V. A.; Weitekamp,
D. P. J. Magn. Reson. 2010, 205, 125−129.
(15) Adams, R. W.; Aguilar, J. A.; Atkinson, K. D.; Cowley, M. J.;
́
Elliott, P. I.; Duckett, S. B.; Green, G. G.; Khazal, I. G.; Lopez-Serrano,
J.; Williamson, D. C. Science 2009, 323, 1708−1711.
(16) Natterer, J.; Bargon. J. Prog. Nucl. Magn. Res. Spect. 1997, 31,
293−315.
(17) Kwim, W. p-Allyl systems in catalysis. In Transition Metals in
Homogeneous Catalysis; Marcel Dekker Inc.: New York, 1971.
(18) Barkemeyer, J.; Haake, M.; Bargon, J. J. Am. Chem. Soc. 1995,
117, 2927−2928.
(19) Annual Reports on NMR Spectroscopy; Academic Press Inc.: San
Diego, CA, 1993; Vol. 25.
(20) Reineri, F.; Viale, A.; Giovenzana, G.; Santelia, D.; Dastru, W.;
̀
Gobetto, R.; Aime, S. J. Am. Chem. Soc. 2008, 130, 15047−15053.
(21) (a) Johannesson, H.; Axelsson, O.; Karlsson, M. C. R. Phys.
2004, 5, 315−324. (b) Reineri, F.; Viale, A.; Dastru, W.; Gobetto, R.;
̀
Aime, S. Contrast Med. Mol. Imag. 2010, 6, 77−84.
(22) Natterer, J.; Schedletzky, O.; Barkemeyer, J.; Bargon, J.; Glaser,
S. J. J. Magn. Reson. 1998, 133, 92−97. Aime, S.; Gobetto, R.; Reineri,
F.; Canet, D. J. Chem. Phys. 2003, 119, 8890−8896.
(23) Katz-Brull, R.; Degani, H. Anticancer Res. 1996, 16, 1375−1380.
(24) Cudalbu, C.; Comment, A.; Kurdzesau, F.; van Heeswijk, R. B.;
Jannin, S.; Denisov, V.; Zirik, D.; Gruetter, R. Phys. Chem. Chem. Phys.
2010, 15, 5818−5823.
11152
dx.doi.org/10.1021/ja209884h | J. Am. Chem. Soc. 2012, 134, 11146−11152