New hyperpolarized contrast agents for 13C-MRI from para-hydrogenation of oligooxyethylenic alkynes

Article abstract of DOI:10.1021/ja8059733

Two alkyne derivatives, which contain one and two oligooxyethylenic chains respectively, showed to be good substrates for para-hydrogenation reactions, yielding the corresponding hyperpolarized alkenes in good yields. A suitable theory has been developed to account for the observed results, fully explaining the different para-H₂ induced effects observed upon the para-hydrogenation of symmetrically and asymmetrically substituted alkynes in ALTADENA and PASADENA modes. The oligooxyethylenic substituent provides good water solubility to the para-hydrogenated symmetrical derivative. ¹³C-MR in vitro images of the latter derivative were obtained both in acetone and in water solutions (130 mM), using the ALTADENA procedure and after application of the field cycling procedure which allows acquisition of an in-phase ¹³C carbonyl resonance. The finding that the hydrogenated product is water-soluble in contrast to the parent alkyne which is not allows for the pursuit of a fast phase-transfer separation from the organic solvent, the unreacted substrate, and the catalyst to obtain a ready-to-use water solution suitable for further in vivo MRI applications.

Full text of DOI:10.1021/ja8059733

Published on Web 10/16/2008
New Hyperpolarized Contrast Agents for 13C-MRI from
Para-Hydrogenation of Oligooxyethylenic Alkynes
Francesca Reineri,† Alessandra Viale,† Giovanbattista Giovenzana,‡
Daniela Santelia,† Walter Dastru`,† Roberto Gobetto,† and Silvio Aime*,†
Dipartimento di Chimica IFM, UniVersita` di Torino, Via P. Giuria 7, 10125 Torino, Italy,
DiSCAFF and DFB Center, UniVersita` degli Studi del Piemonte Orientale “A. AVogadro”,
Via BoVio 6, I-28100 NoVara, Italy
Received April 29, 2008; E-mail: silvio.aime@unito.it
Abstract: Two alkyne derivatives, which contain one and two oligooxyethylenic chains respectively, showed
to be good substrates for para-hydrogenation reactions, yielding the corresponding hyperpolarized alkenes
in good yields. A suitable theory has been developed to account for the observed results, fully explaining
the different para-H2 induced effects observed upon the para-hydrogenation of symmetrically and
asymmetrically substituted alkynes in ALTADENA and PASADENA modes. The oligooxyethylenic substituent
provides good water solubility to the para-hydrogenated symmetrical derivative. 13C-MR in vitro images of
the latter derivative were obtained both in acetone and in water solutions (130 mM), using the ALTADENA
procedure and after application of the field cycling procedure which allows acquisition of an in-phase 13C
carbonyl resonance. The finding that the hydrogenated product is water-soluble in contrast to the parent
alkyne which is not allows for the pursuit of a fast phase-transfer separation from the organic solvent, the
unreacted substrate, and the catalyst to obtain a “ready-to-use” water solution suitable for further in vivo
MRI applications.
(PHIP)7,15-21,29-39 methods are currently under intense scrutiny
for the preparation of 13C enriched hyperpolarized substances,
Introduction
for which interesting applications such as 13C-MRI contrast
agents have already been anticipated. With this kind of contrast
agent, images are acquired by the direct detection of the 13C
nucleus, whose endogenous signal is practically zero. The
absence of a background signal allows us to obtain images with
a high signal-to-noise ratio (SNR), where the contrast is given
by the difference in signal intensity between regions reached
by the hyperpolarized 13C molecule and uninvolved zones.8,9
Many efforts have been devoted in recent years to the
development of MR-hyperpolarization procedures, mainly driven
by the potential use of hyperpolarized molecules in Magnetic
Resonance Imaging (MRI) and Magnetic Resonance Spectros-
copy (MRS).1-3In this context, Dynamic Nuclear Polarization
(DNP)9-14,22,23 and Para-Hydrogen Induced Polarization
† Universita` di Torino.
‡ Universita` degli Studi del Piemonte Orientale “A. Avogadro”.
(1) Kauczor, H. U.; Surkau, R.; Roberts, T Eur. Radiol. 1998, 8, 820.
(2) Moller, H. E.; Chen, X. J.; Saam, B.; Hagspiel, K. D.; Johnson, G. A.;
Altes, T. A.; de Lange, E. E.; Kauczor, H. U. Magn. Res. Med. 2002,
47, 1029.
(14) Bhattacharya, P.; Chekmenev, E. Y.; Perman, W. H.; Harris, K. C.;
Lin, A. P.; Norton, V. A.; Tan, C. T.; Ross, B. D.; Weitekamp, D. P.
J. Magn. Reson. 2007, 786, 150.
(15) Ishii, M.; Emami, K.; Kadlecek, S.; Petersson, J. S.; Golman, K.;
Vahdat, V.; Yu, J.; Cadman, R. V.; MacDuffie-Woodburn, J.; Stephen,
M.; Lipson, D. A.; Rizi, R. R. Magn. Res. Med. 2007, 57, 459.
(16) Magnusson, P.; Johansson, E.; Mansson, S.; Petersson, J. S.; Chai,
C. M.; Hansson, G.; Axelsson, O.; Golman, K. Magn. Res. Med. 2007,
57, 1140.
(3) Altes, T. A.; Salerno, M. J. Thoracic Imag. 2004, 19, 250.
(4) Oros, A. M.; Shah, N. J. Phys. Med. Biol. 2004, 49, R105.
(5) Albert, M. S.; Cates, G. D.; Driehuys, B.; Happer, W.; Saam, B.;
Springer, C. S.; Wishnia, A. Nature 1994, 370, 199.
(6) Middleton, H.; Balck, R. D.; Saam, B.; Cates, G. D.; Cofer, G. P.;
Guenther, R.; Happer, W.; Hedlund, L. W.; Johnson, G. A.; Juvan,
K.; Swartz, J. Magn. Reson. Med. 1995, 33, 271.
(17) Goldman, M.; Johannesson, H.; Axelsson, O.; Karlsson, M. Magn.
Res. Imag. 2005, 23, 153.
(7) Golman, K.; Axelsson, O.; Johannesson, H.; Mansson, S.; Olofsson,
C.; Petersson, J. S. Magn. Res. Med. 2001, 46, 1.
(18) Johannesson, H.; Axelsson, O.; Karlsson, M. C. R. Phys. 2004, 5,
315.
(8) Mansson, S.; Johansson, E.; Magnusson, P.; Chai, C. M.; Hansson,
G.; Petersson, J. S.; Stahlberg, F.; Golman, K. Europ. Radiol. 2006,
16, 57, and references therein.
(19) Goldman, M.; Johannesson, H. C. R. Phys. 2005, 6, 575.
(20) Johansson, E.; Olsson, L. E.; Mansson, S.; Petersson, J. S.; Golman,
K.; Stahlberg, F.; Wirestam, R. Magn. Reson. Med. 2004, 52, 1043.
(21) Golman, K.; Olsson, L. E.; Axelsson, O.; Mansson, S.; Karlsson, M.;
Petersson, J. S. Br. J. Radiol 2003, 76, S118.
(9) Golman, K.; Petersson, J. S. Acad. Radiol. 2006, 13, 932, and
references therein.
(10) Golman, K.; Zandt, R.; Thaning, M. Proc. Natl. Acad. Sci. U.S.A.
2006, 103, 11270.
(22) Svensson, J.; Mansson, S.; Johansson, E.; Petersson, J. S.; Olsson,
L. E. Magn. Reson. Med. 2003, 50, 256.
(11) Golman, K.; Zandt, R.; Lerche, M.; Perhson, R.; Ardenkjaer-Larsen,
H. Cancer Res. 2006, 66, 10855.
(23) Kolman, K.; Ardenkjaer-Larsen, J. H.; Petersson, J. S.; Mansson, S.;
Leunbach, I. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10435.
(24) Abragam, A.; Goldman, M. Rep. Prog. Phys. 1978, 41, 395.
(25) Ardenkjaer-Larsen, J. H.; Fridlund, B.; Gram, A.; Hansson, G.;
Hansson, L.; Lerche, M. H.; Servin, R.; Thaning, M.; Golman, K.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10158.
(12) Golman, K.; Petersson, J. S. Acad. Radiol. 2006, 13, 932.
(13) 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.
9
10.1021/ja8059733 CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 15047–15053 15047

A R T I C L E S
Reineri et al.
Chart 1. Chemical Structures of 1 and 2
Scheme 1. Synthesis of 1
Scheme 2. Synthesis of 2
diethylene glycol monomethyl ether (Scheme 1; see Experi-
mental Section for details). Bis[2-(2-methoxyethoxy)ethyl]acety-
lenedicarboxylate (2) was synthesized by the transesterification
of dimethylacetylenedicarboxylate with diethyleneglycol mono-
methyl ether, in the presence of H2SO4 as catalyst (Scheme 2;
see Experimental Section for details).
Both 1 and 2 are hydrogenated quite readily by the Rh catalyst
[bis(diphenylphosphino)butane](1,5-cyclooctadiene)rhod-
ium(I) tetrafluoroborate. Typically, when the reactions were
carried out in a 5 mm NMR tube containing the substrate (50
÷ 150 mM) and catalyst (17 mM) and in the presence of 4.0
atm of H2 (solvent acetone-d6), yields of ∼85% were obtained.
In Figure 1a the 13C NMR spectrum obtained upon para-
hydrogenation of 2 (2a, Scheme 3), acquired in the PASADENA
procedure, is reported. It shows two absorption/emission signals
centered at 165.53 ppm (carbonyl carbon atoms) and at 132.21
ppm (olefinic carbon atoms), respectively.
The PASADENA experiment carried out under the same
experimental conditions on the asymmetrical substrate 1 did not
afford any 13C enhanced signal (spectrum not shown). Con-
versely, a PHIP effect in the 13C spectrum of 1a was detected
(see Scheme 3) when the para-hydrogenation was carried out
in an ALTADENA experiment, as depicted in Figure 2a.
On substrate 2 the two H atoms from the para-H2 molecule
are transferred to chemically equivalent positions, while on 1
the two H atoms are added to chemically different sites.
Chemical symmetry plays a role when the parahydrogenation
The DNP hyperpolarization procedure can be in principle
applied to any molecule, provided that efficient methods for
the rapid dissolution of the hyperpolarized substrate and the
separation of the paramagnetic agent are available. Conversely,
the use of the PHIP method requires hydrogenable substrates
and implies the use of hydrogenation catalysts which must be
removed before the in ViVo administration. Its main advantage
relies on the fact that it does not require the very low
temperatures used in the DNP procedure, thus resulting in a
simpler and cheaper methodology.
To produce a 13C hyperpolarized contrast agent using this
approach, an unsaturated substrate is necessary (usually a triple
bond containing molecule, that is efficiently para-hydrogenated
in the presence of a suitable catalyst), with an adjacent carbonyl
group, whose 13C resonance is characterized by a long T1 value
(which limits the polarization loss due to relaxation). Polarization
is transferred to the carbonyl carbon nucleus through scalar
coupling with protons from para-hydrogen.7,34-38
In the first papers dealing with 13C-PHIP,36,37 the effect was
reported only for symmetric molecules, in which the para-H2
symmetry was broken by the presence of the heteronucleus.
Successively, it was observed that polarization transfer to
heteronuclei can also be obtained on asymmetric molecules,
providing that hydrogenation takes place at a low magnetic field
(ALTADENA experiment).35
Herein we report the synthesis and parahydrogenation experi-
ments of two novel substrates, the asymmetrical 2-(2-methoxy-
ethoxy)ethyl phenylpropiolate (1) and the symmetrical bis[2-
(2-methoxyethoxy)ethyl]acetylenedicarboxylate (2) (Chart 1),
with the aim of obtaining an in-depth understanding of the
potential of these species as 13C hyperpolarized contrast agents.
A theoretical explanation of the difference between 13C hyper-
polarized spectra of symmetrical and asymmetrical substrates
in ALTADENA and PASADENA experiments is also provided.
The (oligo)oxyethylenic chains were introduced into the
alkynic substrates to obtain water-soluble para-hydrogenated
products, as aqueous contrast agent solutions are needed for in
ViVo MRI applications. Up to now only one water-soluble
substrate has been reported as a 13C hyperpolarized contrast
agent, namely 2-hydroxyethylacrylate, which upon hydrogena-
tion yields 2-hydroxyethylpropionate.17-21
Figure 1. Single scan 13C NMR spectra (in the 100-220 ppm region;
14.1 T, acetone-d6, 298 K) of 2a obtained upon para-hydrogenation of 2 in
the PASADENA experiment, recorded (a) immediately after para-
hydrogenation and (b) after relaxation (5 min). S indicates the solvent signal.
Scheme 3. Hydrogenation of 1 and 2
One aim of the present work was to improve the methodology
by making the solvent elimination step faster. It was found that
2 affords an hydrogenated product which is soluble both in
organic solvents and in water, thus allowing the hydrogenation
process to be carried out in an organic solvent solution (where
the homogeneous cationic Rh catalyst is more efficient and H2
is more soluble) and then transfer of the hyperpolarized product
to the aqueous phase.
Results and Discussion
2-(2-Methoxyethoxy)ethyl phenylpropiolate (1) was synthe-
sized by the direct esterification of phenylpropiolic acid with
9
15048 J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008

New Hyperpolarized Contrast Agents for 13C-MRI
A R T I C L E S
Figure 2. Single scan 13C NMR spectra (in the 100-220 ppm region;
14.1 T, acetone-d6, 298 K) of 1a obtained upon para-hydrogenation of 1 in
the ALTADENA experiment, recorded (a) immediately after para-
hydrogenation and (b) after relaxation (5 min). S denotes the solvent signal;
o and m indicate the ortho and meta positions of the aromatic ring,
respectively.
takes place in PASADENA conditions; otherwise, when the
ALTADENA experiment is carried out, both 13C signals
J-coupled with the parahydrogen protons in 1a (the carbonyl at
166.31 ppm and the C-1 aromatic peak at 135.60 ppm) result
in being polarized, even though to different extents. This is the
result of the scalar coupling pattern, as it will be discussed in
detail in the theoretical portion.
1
Figure 3. Single scan H NMR spectra (9.4 T, acetone-d6, 298 K) of 1a
obtained upon para-hydrogenation of 1 in the ALTADENA experiment,
recorded (a) immediately after para-hydrogenation and (b) after relaxation
(5 min). Trace b has been multiplied × 8; i denotes an impurity; * denotes
catalyst’s signals.
Table 1. Carbonyl 13C Relaxation Times of 2a
Interestingly, the olefinic (143.17 and 120.07 ppm) and the
C-1 aromatic carbon atoms of 1a are only slightly enhanced,
while strong polarization is observed for the ortho aromatic
carbon at 128.62 ppm (doublet absorption/emission), and a
weaker effect is detected for the meta carbon (130.75 ppm).
Polarization on these carbon atoms is not the result of scalar
coupling with para-hydrogen protons, but it arises from an
indirect pathway that involves the aromatic protons. Polariza-
tion is indeed detected on both ortho and meta aromatic protons
in the 1H NMR spectrum of 1a (Figure 3), as they are involved
in a dipolar interaction with the olefinic ones.
Both 1 and 2 are very soluble in organic solvents but not in
water, in spite of the presence of ethoxylic groups. However,
upon hydrogenation, 2a shows good water solubility (∼600
mM). This feature allows the para-hydrogenation reaction to
be carried out in an acetone solution and then transfer of the
alkene molecule into the aqueous phase for in ViVo administra-
solvent
O2 present
field (T)
T1 (s)
acetone-d6
D2O
D2O
D2O
D2O
no
14.1
14.1
9.4
7.0
0.5
20.0
6.3
11.1
16.0
33a
yes
yes
yes
yes
a Value estimated from the CSA and dipolar contributions.
tion. This property allows bypassing the time-consuming
procedure for solvent/catalyst elimination and dissolution in
water, thus making this compound a particularly interesting 13
hyperpolarized contrast agent for MRI applications.
C
As mentioned above, the relaxation rate of the hyperpolarized
CA is crucial for the acquisition of in ViVo 13C images. Therefore
the T1 of the 2a 13C carbonyl carbon atom was measured under
different conditions (solvent, magnetic field strength, in the
presence and absence of oxygen). The obtained data are reported
in Table 1. The relatively large molecular size of 2a shortens
T1 with respect to the parent dimethylmaleate. In fact, the
carbonyl carbon atom T1 (measured in acetone-d6 solutions at
14.1 T) is 20 s for 2a and 40 s for dimethylmaleate, respectively.
By measuring T1 at different field strengths it was possible to
obtain an estimate of dipolar and CSA contributions, the latter
being the dominant term. From these data it is possible to
anticipate a T1 of 33 s at 0.5 T (corresponding to a field value
still used in clinical MRI). The observed T1 values appear to
be long enough to allow the detection of 13C images, even after
its transfer into the aqueous phase and in the presence of oxygen.
For the acquisition of 13C images of hyperpolarized 2a, 13C
labeled 2 (13C enriched on one of the carbonyl groups) was
synthesized (see Experimental Section).
(26) 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.
(27) Gerfen, G. J.; Becerra, L. R.; Hall, D. A.; Griffin, R. G.; Temkin,
R. J.; Singel, D. J. J. Chem. Phys. 1995, 102, 9494.
(28) Un, S.; Prisner, T.; Weber, R. T.; Seaman, M. J.; Fishbein, K. W.;
McDermott, A. E.; Singel, D. J.; Griffin, R. G. Chem. Phys. Lett. 1992,
189, 54.
(29) Eisenberg, R. Acc. Chem. Res. 1991, 24, 110, and references therein.
(30) Natterer, J.; Bargon, J. Prog. Nucl. Magn. Reson. Spectrosc. 1997,
31, 293, and references therein.
(31) Duckett, S. B.; Sleigh, C. J. Prog. Nucl. Magn. Reson. Spectrosc. 1999,
34, 71, and references therein.
(32) Bowers, C. R.; Weitekamp, D. P. Phys. ReV. Lett. 1986, 57, 2645.
(33) Pravica, M. G.; Weitekamp, D. P. Chem. Phys. Lett. 1988, 145, 255.
(34) Natterer, J.; Schedletzky, O.; Barkemeyer, J.; Bargon, J. J. Magn.
Reson. 1998, 133, 92.
(35) Stephan, M.; Kohlman, O.; Niessen, H. G.; Eichhorn, A.; Bargon, J.
Magn. Reson. Chem. 2002, 40, 157.
Hyperpolarization effects observed on the 13C resonances of
parahydrogenated molecules, in both ALTADENA and PASA-
DENA experiments, stem from longitudinal two spin order terms
(IHz ICz ) and result in adsorption/emission signals that, as such,
can not be exploited in the 13C image acquisition, as their
resultant is practically zero. Thus it is necessary to convert the
longitudinal spin order into longitudinal magnetization (ICz )
(36) Barkemeyer, J.; Haake, M.; Bargon, J. J. Am. Chem. Soc. 1995, 117,
2927.
(37) Haake, M.; Barkemeyer, U.; Bargon, J. J. Phys. Chem. 1995, 99,
17539.
(38) Bargon, J.; Bommerich, U.; Kadlecek, S.; Ishii, M.; Fischer, M. C.;
Rizi, R. R. Proc. Intl. Soc. Magn. Reson. Med. 2005, 13, 2572.
(39) Barkemeyer, J.; Haake, M.; Bargon, J. J. Am. Chem. Soc. 1995, 117,
2727.
9
J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008 15049

A R T I C L E S
Reineri et al.
Figure 4. Single scan 13C NMR spectrum (14.1 T, acetone-d6, 298 K) of
13C labeled 2a obtained upon para-hydrogenation of 13C-2 (ALTADENA)
recorded (a) immediately after para-hydrogenation and application of the
field cycling procedure and (b) after relaxation (5 min). The signal at 172.48
ppm is attributed to the fully hydrogenated alkane carbonyl group; i denotes
an impurity; S indicates the solvent signal. Signal enhancement in spectrum
(a) is about 250.
Figure 6. Single scan 13C spectrum (100-220 ppm region; 14.1 T, 298
K) of 13C labeled 2a extracted in D2O after para-hydrogenation of 13C-2
(ALTADENA, CDCl3/acetone-d6 6:1) and application of the field cycling
procedure, recorded (a) immediately after phase separation; (b) after
relaxation (5 min). The two different signals at 165.99 and 164.82 ppm are
attributed to 2a dissolved in D2O (13 mM) and in CDCl3 (still present in
trace, 117 mM), respectively.
Nevertheless, the signal enhancement observed in D2O is still
sufficient to allow acquisition of the 13C image of a tube
containing a D2O solution of the para-hydrogenated compound
(130 mM), which is reported in Figure 5b.
The change in the solubility properties between the parent
alkyne 2 and its hydrogenated derivative 2a has been exploited
to set up a novel method for separating the hyperpolarized
product from the reaction mixture. Upon carrying out the para-
hydrogenation reaction in a solvent not miscible with water,
such as chloroform or dichloromethane, one may obtain an
aqueous solution of the hyperpolarized product simply by phase
transfer to a small amount of water added upon completion of
the para-hydrogenation reaction. Thus the nonhydrogenated
substrate, the catalyst, and the organic solvent will not pollute
the aqueous phase. The possibility of attaining the hyperpolar-
ized product solution in a single step will be highly advanta-
geous, as the time-consuming solvent and catalyst elimination,
during which polarization decays, is avoided. This goal was
accomplished in the present work by performing the para-
hydrogenation of 2 in a CDCl3/acetone-d6 6:1 mixture (a low
fraction of acetone is necessary for maintaining the catalyst
activity) in the NMR tube and then quickly adding some D2O,
shaking the tube, and let it stand for a few seconds to allow
phase separation. The D2O phase was then transferred into a
second tube for NMR detection. The 13C NMR spectrum of an
aqueous solution of 2a obtained by this procedure is shown in
Figure 6a. Of course the signal enhancement (∼100) is lower
than that obtained in the acetone-d6 solution (Figure 4), due to
polarization decay occurring during the phase-transfer step.
Still a small amount of chloroform solution is present, thus
yielding an additional signal for 2a dissolved in the latter phase
at 164.82 ppm. When comparing it to the peak from the aqueous
phase (165.99 ppm), one must take into account that its intensity
reflects the higher solubility of 2a in chloroform (117 mM) than
in water (13 mM). Thus, from the intensity ratio between the
two different signals it is possible to estimate that the residual
chloroform amount in the aqueous phase is ∼3%.
Figure 5. 13C single shot RARE images of 5 mm NMR tubes containing
hyperpolarized 13C labeled 2a (axial projections). (a) Tube containing 130
mM HP-13C- 2a in degassed acetone-d6; (b) tube containing a D2O 130
mM solution of HP-13C-2a (residual acetone-d6 in the solution is ∼4% after
the fast distillation procedure).
before proceeding to the MRI acquisition. This can be achieved
by a magnetic field cycling procedure, consisting of two
asymmetric transformations of the magnetic field: in the first
passage the sample is quickly (non-adiabatically) thrown from
a 50 µT (earth’s field) to nearly zero (0.1 µT) magnetic field,
and then it is slowly brought back (adiabatically) to the earth’s
magnetic field. The whole process takes only a few seconds.
The 13C spectrum in Figure 4, obtained after application of
the magnetic field cycle on 13C labeled 2a, shows an in-phase
enhanced 13C signal (165.53 ppm) related to the carbonyl 13C
atom, while the polarization on olefinic carbons is lost, probably
due to the faster relaxation of these nuclei caused by dipolar
interaction with directly bound protons. The corresponding 13
C
image of the NMR tube containing HP 13C-2a was then
acquired, and it is reported in Figure 5a. The estimated
concentration of the hydrogenated product in the solution was
130 mM.
A good polarization level was maintained even after acetone
elimination (achieved by fast vacuum distillation, as described
in detail in the Experimental Section) and dissolution of the
residue in D2O. The 13C enhancements in the high resolution
spectra of acetone-d6 (figure 4) and D2O (spectrum not shown)
solutions were 250 and 50, respectively. The lower value in
D2O is due to polarization loss caused by relaxation during
nebulization, acetone distillation, and transfer of the sample from
the distillation apparatus to the NMR tube and due to the
relaxation enhancement in the aqueous solution, as demonstrated
by the comparison between the measured T1 values in acetone-
d6 and D2O (see Table 1). The increased relaxation rate in water
is mainly due to the presence of oxygen but also to a lower
molecular mobility (longer τc due to increased viscosity).
Although the detected signal enhancement appears sufficient
for image acquisition, further improvements are expected to
make this methodology, based on the differential solubility of
the para-hydrogenated molecule and the parent substrate, the
technique of choice for the acquisition of 13C-MR images of
hyperpolarized molecules.
Theory: Polarization Transfer to 13C in Symmetrical and
Asymmetrical Molecules. Polarization transfer from para-
1
hydrogen protons to 13C is mainly driven by H-13C scalar
9
15050 J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008

New Hyperpolarized Contrast Agents for 13C-MRI
A R T I C L E S
coupling in the para-hydrogenation products.39 As experimen-
tally observed, para-H2 addition to chemically different sites as
in molecule 1 allows 13C polarization to be obtained only when
it takes place out of the spectrometer (ALTADENA experiment),
while para-hydrogenation of the symmetric molecule 2 leads
to heteronuclear polarization in both ALTADENA and PASA-
DENA experiments. In the latter case the loss of the singlet
state, necessary to exploit para-hydrogen hyperpolarization, is
due to asymmetrical coupling of the two protons to the 13C
nucleus. The difference between the two cases is of general
applicability, and it can be demonstrated by describing the
evolution of the para-hydrogen density operator σpara on the
product molecule using an approach, introduced by Canet et
al.,40,41 based on the assumption that a steady-state condition
is reached by the para-H2 density operator after addition to the
substrate. This method, more convenient than others that require
the solution of the evolution equation, is summarized below.
The para-H2 density operator is described by eq 1,42 and the
hydrogenation reaction can be considered as a sudden change
of the Hamiltonian governing the two spins system.
where only the J-Hamiltonian is considered, as the Zeeman part
does not effect para-H2. This equation means that a certain term
Gj is polarized if it does not commute with HJ.
13C Hyperpolarization in ALTADENA and PASADENA
Experiments. Let us consider the parahydrogen addition to a
symmetrical substrate such as 2. Only one of the two carbonyl
groups contains a 13C atom; therefore the symmetry of the two
hydrogen atoms is broken by asymmetric coupling with it and
an AA′X spin system is formed. In this case the strong coupling
condition between protons is maintained inside the spectrometer.
On the contrary, with an asymmetrical substrate such as 1 the
two parahydrogen protons are added in chemically different
environments and strong coupling between protons is achieved
only outside the magnet, when an ALTADENA experiment is
carried out.
As a consequence the J-Hamiltonian that transforms para-
hydrogen after its addition to a symmetric substrate results in
eq 7 in both ALTADENA and PASADENA experiments, while
with asymmetric molecules it results in eq 7 only in the latter
case.
H1 C1
H1C1 z
H ) JH1H2(IzH1IHz 2 + IxH1IHx 2 + IyH1IyH2) + JH2C1IzH2ICz 1 + J
I Iz
E
4
σpara
)
- IHz IHz - IHx IHx - IyHIHy
(1)
(7)
Thus the evolution of the density operator σpara upon addition
to the substrate is given by the Liouville-von Neumann
equation
The same expression can be written for the other 13C nucleus
(C2); in fact only one position at a time can be occupied by a
13C atom because of its low natural abundance.
dσ
If a PASADENA experiment is performed, the Hamiltonian
by which para-hydrogen is transformed is
para ) -i[H, σpara
]
(2)
dt
H1 C1
H1C1 z
HPasAySmAmDENA ) JH1H2IHz 1IzH2 + JH2C1IzH2ICz 1 + J
I Iz
(8)
where H is the product molecule Hamiltonian.
After reaction completion, parahydrogen reaches a new steady
As assessed above, a product operator is polarized if it does
not commute with H. We focus on the longitudinal order IHz 2ICz 1
which gives the antiphase doublets observed in 13C polarized
spectra. This term is invariant with respect to eq 8, while
applying Hamiltonian of eq 7 one obtains
state (σstat) on the product molecule, such that
dσ
stat ) -i[H, σstat] ) 0
(3)
dt
which contains, besides the para-H2 product operators (eq 2),
other terms that, deriving from parahydrogen evolution, are
polarized. Hyperpolarization is transferred according to σstat as
follows:
1
2
1
2
H1 H2 C1
H1H2 x
H, IH2IC1 ) JH1H2Iy Ix Iz
H1 H2 C1 - J
I Iy Iz
(9)
[
]
z
z
In other words the longitudinal magnetization term does not
commute with (IxH1IHx 2 + IyH1IHy 2), which is related to the strong
coupling condition achieved in ALTADENA experiments or
in PASADENA carried out on symmetric molecules. On the
contrary, in PASADENA experiments on asymmetric molecules
the strong coupling condition is not obtained.
E
4
σstat
)
-
astatGj
(4)
∑
j
j
where
h
Gj ) ΠS(ij)
The reported method allowed us to derive that the amount
of polarization transfer toward 13C is proportional to the ratio
i)1
40
with Si is Ei, Iix, Iyi, or Iiz and asjtat represent the contribution of
each Gj product operator to the steady state.
Therefore eq 3 becomes eq 5,
RJ,C1 ) (JH2C1 - JH1C1)/JH1H2
,
which is the same as that
obtained by Bargon for symmetric molecules.36 Being the
coupling constants in 1a JH2,C1 ) 10.0 Hz, JH1,C2 ) 15.8 Hz,
JH1,C1 ) -1.8 Hz, JH2,C2 ) -2.5 Hz, and JH1,H2 ) 12.6 Hz,
then RJ,C1 ) 0.94 and RJ,C2 ) 1.45: this means that polarization
is expected to be higher on the carbonyl signal with respect to
the aromatic one. This is indeed the case. Hyperpolarization
observed on the aromatic ortho 13C carbon cannot be due to
scalar coupling with para-hydrogen protons. The antiphase
doublet clearly refers to the term IzCorthoIzHortho, and polarization
on Hortho (seen in Figure 3) must be transferred from para-
hydrogen protons by means of dipolar interaction.
[H , G ]astat ) 0
(5)
∑
J
j
j
j
(40) Aime, S.; Gobetto, R.; Reineri, F.; Canet, D. J. Chem. Phys. 2003,
119, 8890.
(41) Aime, S.; Gobetto, R.; Reineri, F.; Canet, D. J. Magn. Reson. 2006,
178, 184.
(42) Canet, D.; Aroulanda, C.; Mutzenhardt, P.; Aime, S.; Gobetto, R.;
Reineri, F. Concept Magnet. Reson. A 2006, 28A, 321.
(43) A higher signal enhancement was observed in the ALTADENA
experiment without application of the field cycling with respect to
that observed when the field cycling was used (shown in Figure 4)
because in the second case some polarization was lost during the
procedure.
Considering the J coupling values of 2a (JH2,C1 ) 24.0 Hz,
JH1,C1 ) -2.5 Hz, JH1,H2) 12.0 Hz), an RJ,C1 value of 2.20 is
obtained; i.e. polarization transfer to the carbonyl 13C atom in
2a is expected to be more efficient than that in 1a. This is in
9
J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008 15051

A R T I C L E S
Reineri et al.
generator, model HG300. 13CO2 (99% enriched) was purchased
from Eurisotop, France.
accordance with the observed carbonyl signal enhancements
obtained in ALTADENA experiments without field cycling,
which are 1500 for 2a (spectrum not shown) and 800 for 1a
(Figure 1), respectively.
NMR spectra were recorded on a Bruker Avance 600 spectrom-
eter operating at 600 MHz for the proton and 150.9 MHz for 13C.
For the detection of the PHIP effect single scan acquisitions were
carried out. Single scan 13C images were obtained on a Bruker
Avance 300 spectrometer (7 T) equipped with a Microimaging
probe, using the RARE sequence, with a RARE factor of 32, matrix
32 × 32, echo time 8.465 ms.
Conclusions
The possibility of obtaining 13C-PHIP on asymmetrical
molecules only when the para-hydrogenation reaction takes place
at low magnetic field strength (ALTADENA experiment), while
with symmetrical substrates this is observed even when para-
hydrogenation occurs at high magnetic field strength (PASA-
DENA experiment) was elucidated by means of the steady-state
density operator approach. This method allowed to directly
demonstrate why the strong coupling condition between protons
is necessary to 13C polarization transfer.
The symmetrical molecule 2 represents a good step along
the way of applying para-hydrogenated substrates in MRI as
13C hyperpolarized agents. In fact, in spite of its size, the
carboxylate resonance in 2a displays a relatively long T1,
reaching a value of 33 s at the field strenght 0.5 T. Furthermore,
it is easily soluble in acetone, where it can be hydrogenated in
high yields, and its hydrogenated product 2a is also soluble in
water, thus making its use possible for in ViVo MRI applications.
Good in Vitro 13C images were obtained from solutions of the
para-hydrogenated 2a product either in acetone or in D2O at 7
T.
13C relaxation times were measured using the inversion recovery
pulse sequence.
Para-enriched hydrogen (52%) was prepared storing H2 over
Fe2O3 at 77 K for 1 h.
Para-hydrogenation reactions in ALTADENA conditions were
carried out in a 5 mm NMR tube equipped with a Young valve.
[Bis(diphenylphosphino)butane](1,5-cyclooctadiene)rhodium(I) tet-
rafluoroborate (5 mg) was dissolved in acetone-d6 (0.4 mL) and
activated by reaction with H2. Normal H2 was then replaced by the
para-H2 enriched mixture (4 atm), after addition of the substrate
(0.02 mmol for 1H NMR spectra or 0.06 mmol for 13C NMR
spectra/images acquisition). The hydrogenation was carried out by
shaking the tube for 10 s (yields ) 80% for 1a and 85% for 2a).
For the acquisition of an in phase 13C resonance, magnetic field
cycling was applied to the hydrogenated sample: this was pursued
by quickly inserting the tube into a µ-metal shield (field strength
0.1 µT) and then slowly removing the shield (the entire field cycling
procedure required 3-5 s). The sample was finally inserted into
the spectrometer for acquisition of the high resolution NMR
spectrum.
2a is characterized by water solubility and signal enhancement
similar to those observed under the same experimental condi-
tions for the already reported 2-hydroxyethylpropionate (derived
from para-hydrogenation of 2-hydroxyethylacrilate).17-21 How-
ever, it shows some advantages, namely:
Para-hydrogenation reactions in PASADENA conditions were
carried out by bubbling para-H2 directly into the NMR tube for
∼10 s (inside the spectrometer). The tube contained 0.4 mL of an
acetone-d6 solution of activated catalyst and substrate at the same
concentrations as those in the ALTADENA experiments. The
bubbling was then stopped, and the NMR spectra were immediately
acquired.
(a) The 13C relaxation time of 2a is somewhat longer than
that of 2-hydroxyethylpropionate (20 s for 2a, 15 s for
2-hydroxyethylpropionate in degassed acetone-d6 solu-
tions at 14.1 T), possibly because in 2a the dipolar
relaxation is less efficient due to the minor number of
protons adjacent to the carbonyl group, and because the
OH group in 2-hydroxyethylpropionate can give rise to
intermolecular interactions, slowing down the reorienta-
tion process and thus shortening T1.
For MR image acquisitions, the ALTADENA procedure was
applied. After completing the magnetic field cycle, the 5 mm NMR
tube was inserted into the NMR spectrometer and the 13C image
in pure acetone was rapidly acquired.
Elimination of acetone was achieved by means of a fast
distillation procedure. After the field cycling step, the tube
containing the acetone solution of the para-hydrogenated sample
was opened and the solution was transferred into a syringe
containing 0.75 mL of degassed D2O and then injected through a
capillary into a flask connected to a vacuum pump. Nebulization
occurred, and the more volatile acetone solvent was pumped off,
while the remaining water solution was collected on the bottom of
the flask. Argon was then inflated into the flask, and the solution
was forced into a plastic tube by pressure, thus being transferred
into the tube for MR acquisition. Residual acetone-d6 in the
(b) 2a shows a lower lipophilicity with respect to 2-hydroxy-
ethylpropionate, as demonstrated by the measurement of
the partition coefficient (P) in water/n-octanol of both
substances (see Experimental Section). Since lipophilicity
is commonly accepted as an indirect measure of
toxicity,44,45 one may conclude that 2a is expected to be
less toxic than 2-hydroxyethylpropionate.
Finally, the change in solubility properties between the parent
alkyne 2 and its hydrogenated derivative 2a represents an
important result in view of its potential in ViVo applications. In
fact para-hydrogenation of this kind of molecules can be carried
out in an organic solvent not miscible with water, and the
parahydrogenated compound can be extracted from the reaction
mixture simply by water addition. The occurrence of a phase
separation provides a quick route to “ready-to-use” solutions
(catalyst- and organic-solvent-free) for in ViVo administration.
2
collected solution was ∼4%, as determined by D spectroscopy.
Samples for water extraction were prepared by para-hydrogenat-
ing 2 in CDCl3/acetone-d6 (6:1) in the NMR tube equipped with
the Young valve (same procedure and conditions as for pure acetone
samples). After the field cycling application, the tube was quickly
opened and 0.4 mL of degassed D2O were added; the tube was
shaken vigorously for 3 s and then allowed to stand for 5 s, during
which phase separation occurred. By means of a syringe, the water
solution was transferred into another NMR tube. Due to the short
procedure times, some residual CDCl3 (about 3%) was present in
the water solution. The quantity of 2a transferred into the water
phase was estimated to be about 10% of the total (corresponding
to a final concentration in water of ∼13 mM).
Experimental Section
Solvents were stored over molecular sieves and purged with
nitrogen before use. Hydrogen was produced by a CLAIND
(44) Wei, L.; Yu, H.; Cao, J.; Sun, Y.; Fen, J.; Wang, L. Chemosphere
Measurement of the partition coefficient (P) in water/n-octanol
for 2a and 2-hydroxyethylpropionate was carried out by a mi-
1999, 38, 1713
.
(45) Dai, J.; Jin, L.; Yao, S.; Wang, L. Chemosphere 2001, 42, 899
.
9
15052 J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008

New Hyperpolarized Contrast Agents for 13C-MRI
A R T I C L E S
croshake-flask procedure.46 40 mg of 2a or 2-hydroxyethylpropi-
onate were dissolved in 2 mL of phosphate buffer (0.01 M, pH )
7.4) saturated n-octanol and mixed with 2 mL of n-octanol saturated
phosphate buffer at 25 °C. The mixture was vigorously shaken for
3 h, and then the two phases were separated and centrifuged
individually. The concentration in each phase was determined by
UV photometry. The partition coefficient P was calculated as P )
EtOAc 7:3; 3:7). Pure 2 was obtained as a colorless oil (264 mg;
yield: 30%; Rf [petroleum ether/EtOAc 6:4] ) 0.2). 1H NMR
(acetone-d6): δ ) 4.51 (t, 4H), 3.85 (t, 4H), 3.72 (t, 4H), 3.62 (t,
4H), 3.43 (s, 6H); 13C{1H}-NMR (acetone-d6): δ ) 151.93 (2C),
75.41 (2C), 66.58 (2C), 68.70 (2C), 70.76 (2C), 72.31 (2C), 58.52
(2C).
Synthesis of 13C-Labeled Acetylenedicarboxylate. Propiolic
acid (2.0 mL, 32.5 mmol) was placed in a flask containing THF
(200 mL, freshly distilled from sodium and benzophenone). The
solution was cooled at 0 °C, and then a solution of n-butyllithium
in hexane (1.6 M, 40.6 mL, 65 mmol) was slowly added. The
solution was then cooled at -50 °C, and air was removed. A 13CO2
(99%) cylinder was connected to the reaction flask, and the gas
was bubbled into the solution at RT. The 13CO2 uptake was
monitored by a digital manometer connected to the reaction flask.
The obtained yellow Na acetylenedicarboxylate salt was filtered,
washed three times with THF, and dried under vacuum. A
suspension of the salt (3.64 g, 28.9 mmol) in methanol (20 mL)
was then prepared and was added to a solution of 96% sulfuric
acid (6.6 mL) in methanol (10 mL) in an ice bath (0 °C). The
solution was then warmed to RT and stirred for 4 days. Water (∼50
mL) was added, and the ester was extracted with diethylether (5 ×
40 mL). The organic fraction was washed with water and a saturated
sodium bicarbonate solution and then dried over CaCl2. The dryer
was removed by filtration, and the solvent was evaporated under
vacuum. Dimethyl acetylenedicarboxylate was purified by column
chromatography on silica gel eluting with a petroleum ether/ethyl
Cn-octanol/Cbuffer, obtaining P2a ) 0.371 ( 0.03 and P2-hydroxyethylpropionate
) 2.388 ( 0.08.
Synthesis of 1. Pyridine (25 mL) was introduced into a round
bottomed flask, and phenylpropiolic acid (1.000 g, 6.84 mmol) and
diethyleneglycol monomethyl ether (1.116 g, 9.28 mmol) were
added. A solution of N,N-dicyclohexylcarbodiimide (DCC, 1.806
g, 8.75 mmol) in pyridine (5 mL) was then slowly dropped into
the flask, and the mixture was shaken at RT for 24 h. An aqueous
solution of HCl 10% was then added (pH ) 5), and the solution
was extracted with three portions of diethylether (30 mL). The
organic phase was dried over sodium sulfate, filtered, and rotary-
evaporated. The product was purified by column chromatography
with a petroleum ether/diethylether 9.9/0.1 mixture as eluent,
yielding 900 mg of 1 as a colorless oil (yield 45%). 1H NMR
(acetone-d6): δ ) 7.63 (d, 2H), 7.53 (dd, 1H), 7.48(d, 2H), 4.43 (t,
2H), 3.65 (t, 2H), 3.57 (t, 4H), 3.45 (t, 2H), 3.26 (s, 3H); 13C{1H}-
NMR (acetone-d6): δ ) 153.89 (1C), 133.40 (2C), 131.59 (1C),
129.53 (2C), 119.89 (1C), 86.16 (1C), 81.04 (1C), 72.37 (1C), 70.96
(1C), 69.03 (1C), 58.57 (1C).
Synthesis of 2. Dimethyl acetylenedicarboxylate (400 mg; 2.8
mmol) was added to a mixture formed of diethyleneglycol mono-
methyl ether (10 mL) and H2SO4 (0.5 mL). The reaction mixture
was stirred under nitrogen at 65 °C for 8 h. The solution was then
cooled to RT and diluted with diethyl ether (20 mL), and then
it was washed twice with H2O and once with brine. The organic
layer was dried over sodium sulfate and rotary-evaporated. The
residue was purified by column chromatography (petroleum ether/
1
acetate (95:5) mixture. Yield 44%. H NMR (methanol-d4): δ )
3.90; 13C{1H}-NMR (methanol-d4): δ ) 153.45 (2C), 75.66 (2C,
dd, J13C,13C ) 72.3 Hz; 17.7 Hz), 54.75 (2C).
Acknowledgment. We gratefully acknowledge discussions with
Prof. D. Canet. This work was carried out in the frame of the EU-
NEST STREP Project (No. 005032). Support from MUR (Firb
Project No. RBIP06293N) and EU-EMIL NoE is also aknowledged.
(46) OECD, Paris, 1981 Test Guideline 107, Decision of the council C(81)
30 final.
JA8059733
9
J. AM. CHEM. SOC. VOL. 130, NO. 45, 2008 15053