An aryl-phosphonate appended macrocyclic platform for lanthanide based bimodal imaging agents

Article abstract of DOI:10.1039/c1cc14437e

Four ligand systems have been prepared whose characteristics are well suited to the design of bimodal MRI and luminescence probes. The lanthanide complexes display high relaxivities and luminescence quantum yields. These properties are retained at higher

Full text of DOI:10.1039/c1cc14437e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 11534–11536
www.rsc.org/chemcomm
COMMUNICATION
An aryl-phosphonate appended macrocyclic platform for lanthanide based
bimodal imaging agentsw
a
b
c
ad
¨
Matteo P. Placidi, Jorn Engelmann, Louise S. Natrajan, Nikos K. Logothetis and
Goran Angelovski*a
Received 21st July 2011, Accepted 19th September 2011
DOI: 10.1039/c1cc14437e
Four ligand systems have been prepared whose characteristics
are well suited to the design of bimodal MRI and luminescence
probes. The lanthanide complexes display high relaxivities and
luminescence quantum yields. These properties are retained at
higher magnetic fields and in a range of competitive environ-
ments including model extracellular medium and cultured cells.
of their conflicting requirements for the inner sphere hydration
number (q). MRI agents ideally need a high q to ensure an
effective magnetic relaxation, whilst maintaining a high stability.
Conversely luminescent complexes require a low value of q to
minimise the amount of non-radiative deactivation.
Lanthanide chelates appended with a phosphonate moiety
may offer an elegant solution to this problem. Previous Gd3+
complexes containing this group have shown a higher than
expected relaxivity, despite their low q values. This is attributed
to the presence of several water molecules that occupy the second
coordination sphere.10–13 Therefore we sought to design a
ligand system whose complexes would exhibit favourable char-
acteristics for both luminescence and MRI. This system consists
of a kinetically inert macrocyclic binding site covalently linked
to an aromatic phosphonate moiety.
Lanthanide complexes have received much attention recently
due to their unique magnetic and optical properties. Gd3+
chelates are used to enhance the contrast observed between
tissues in magnetic resonance imaging (MRI), providing
images at unlimited depths with a high spatial and temporal
resolution. Recent efforts have focused on improving the
efficacy or more specifically the longitudinal relaxivity (r1) of
these agents, to increase the sensitivity of this technique with
the ultimate aim of cellular imaging. Technological advances
in MRI have also increased the demand for agents that
function effectively at higher field strengths.1 In parallel, the
luminescence from lanthanide probes can be used to provide
detailed information through bioassay measurements2,3 and
high resolution images at the sub-cellular level.4 By employing
time gated techniques, interfering background autofluorescence
can be suppressed.
This system was chosen for several reasons. Due to the close
proximity of the phosphonate group, it can directly coordinate
to the lanthanide ion maintaining a low q and simultaneously
act as an antenna unit sensitising the metal centred luminescence.
Despite the fact that the complex is octa-coordinate at the
lanthanide ion, the interactions with the water molecules in the
second sphere will ensure an effective relaxivity. Finally struc-
tural variation of the ligand platforms, including length of the
alkyl chain and nature of the phenolic para substituent will
enable fine tuning of the resulting physicochemical properties.
A straightforward synthetic route was developed to obtain
L1–4 (Fig. 1). Following the monobenzylation of hydro-
quinone,14 the phenol 1 was converted to the phosphate ester
2 by reaction with diethyl phosphite in the presence of triethyl-
amine. A phospho-Fries rearrangement was then induced by
the addition of LDA resulting in 3. According to the desired
distance between the macrocycle and the aromatic phosphonate,
3 was reacted with either dibromopropane or dibromobutane
to give 4 and 5 respectively.
The challenge remains to design a suitable ligand system whose
attributes suit both modalities.5–8 Through this approach,
a cocktail of magnetically and optically active complexes
with identical biodistributions could be administered, to
simultaneously exploit the advantages of both techniques
and enable a more accurate interpretation of in vivo molecular
imaging experiments.9 However, the design of a ligand system
suitable for both of these modalities is far from trivial, because
a Department for Physiology of Cognitive Processes, Max Planck
Institute for Biological Cybernetics, Spemannstr. 38, 72076
Tubingen, Germany. E-mail: goran.angelovski@tuebingen.mpg.de;
¨
Fax: +49 7071 601 919; Tel: +49 7071 601 916
b High-Field Magnetic Resonance Center, Max Planck Institute for
The ligand precursors 6 and 7 were prepared by N-alkylation
of the tris-tert butyl ester derivative of cyclen (tris-t-Bu-
DO3A15). At this stage two different synthetic routes were
envisaged for each precursor. The first involved the removal of
the phosphonate ethyl and the tert-butyl ester groups using
Me3SiBr and trifluoroacetic acid (TFA) in CH2Cl2, respec-
tively. These conditions allowed the retention of the benzyl
ether to obtain L1 and L3 as the main product, with traces of
L2 and L4 due to the partial cleavage of the ether. However, in
Biological Cybernetics, Tubingen, Germany
¨
c School of Chemistry, The University of Manchester, Oxford Road,
Manchester, UK
d Imaging Science and Biomedical Engineering, University of
Manchester, Manchester, UK
w Electronic supplementary information (ESI) available: Full synthetic
procedures, NMR and photophysical characterisation. See DOI:
10.1039/c1cc14437e
c
11534 Chem. Commun., 2011, 47, 11534–11536
This journal is The Royal Society of Chemistry 2011

View Article Online
Fig. 1 Synthesis of ligands L1–4: (i) HP(O)(OEt)2, Et3N, CCl4; (ii)
LDA, THF, À78 1C; (iii) dibromopropane or dibromobutane, K2CO3,
DMF, 60 1C; (iv) tris-t-Bu-DO3A, Cs2CO3, CH3CN, 70 1C; (v) H2,
Pd/C, EtOH, 60 1C; (vi) Me3SiBr, CH2Cl2; (vii) TFA, CH2Cl2.
Fig. 2 Absorbance (left solid line), normalised excitation (lem 545 nm,
dashed line) and emission (lex 300 nm, right solid line) spectra of TbL1.
The overlap of the absorbance and excitation spectra confirms that
sensitised emission is occurring.
both cases the two ligand systems were readily separated using
RP-HPLC. The second route involved the removal of all three
protecting groups through catalytic hydrogenation of 6 and 7,
followed by cleavage of the ethyl and tert-butyl esters, using
the same conditions described above to yield L2 and L4
respectively. The successful isolation of L1–4 was confirmed
the inner sphere hydration state (Table 1).21 TbL1–3 all have a
similar q value of 0.7. When considering the available amino
carboxylate donor set, this would suggest that the phosphonate
group is coordinating to the metal centre. TbL4 has the lowest
value of q at 0.3 and accordingly its Gd3+ analogue has the
lowest r1 value. The quantum yields (F) were also determined
in aerated water for TbL1–4 (Table 1), ranging from 0.18 to
0.27. This demonstrates the effectiveness of the sensitised
emission and the potential for these systems to be applied in
optical imaging.
1
through H, 13C, 31P NMR along with high resolution mass
spectrometry. Their Eu3+, Gd3+ and Tb3+ complexes were
prepared using standard procedures (synthesis section ESIw).
For each complex a peak corresponding to the molecular ion
with the correct isotopic pattern could be observed by electro-
spray ionisation mass spectrometry.
Sensitised emission could not be observed for the analogous
EuL1–4 complexes. The excitation spectra indicated only one
significant peak at 395 nm. When excited directly at this
wavelength, low intensity Eu3+ emission bands could be
observed at 580, 588, 617, 654 and 703 nm corresponding to
the transitions of the 5D0 excited state to the 7F0, 7F1, 7F2, 7F3
and 7F4 ground levels respectively. This behaviour most likely
occurs due to the quenching of the emissive state through a
ligand to the metal charge transfer state.22 Nevertheless,
information about the coordination environment surrounding
the Eu3+ can be extracted in the form of emission spectra. For
each complex a distinct peak at 620 nm and a splitting of
the DJ = 1 manifold into three components can be observed
(Fig. S6–S8, ESIw). This is indicative of the phosphonate
binding to the europium centre.19,23
Inversion recovery experiments in MOPS buffered aqueous
solutions were used to determine the longitudinal relaxivities
r1 of the four complexes GdL1–4 (Table 1). The r1 values are
significantly higher than those of commercially available
agents GdDTPA and GdDOTA, (4.0 and 4.2 mMÀ1 sÀ1
,
respectively, at 298 K, pH 7.4, 200 MHz).16 They are also an
improvement on other recent monohydrated phosphonate
appended DO3A systems.17–20 It is important to note that
most of the r1 of the aforementioned systems have been
determined at lower fields (0.47 T, 20 MHz); therefore it is
likely that the difference in r1 when compared to those of
GdL1–4 will be even greater at higher magnetic fields.
The aromatic moiety also acts as an effective antenna for
each of the TbL1–4 complexes. Upon excitation of the chromo-
phore, characteristic Tb3+ emission bands could be observed
at 489, 547, 588 and 622 nm corresponding to transitions from
the 5D4 to 7F6, 7F5, 7F4 and 7F3 states respectively (Fig. 2;
Fig. S1–S4, ESIw). The luminescence lifetimes were measured
by time-resolved measurements in H2O and D2O to determine
Additional evidence for the interaction of the phosphonate
with the Eu3+ centre could be provided by 31P NMR spectro-
scopy (Fig. S9–S12, ESIw). For each of the complexes, a broad
shifted peak could be observed at a range of frequencies from
À75 to À116 ppm, the direction and magnitude of which have
been observed previously in complexes where the phosphonate
group is in close proximity to the paramagnetic center.24
To gain further insight into how these systems will function
in vivo, the inversion recovery and lifetime measurements were
repeated in Dulbeco’s modified Eagle medium (DMEM),
supplemented with fetal calf serum. This solution was designed
to mimic the extracellular medium that surrounds cells in vivo
and is used to grow cell cultures. The r1 of the complexes
GdL1–4 remained high, with only modest decreases ranging
from 8 to 28% (Table S3, ESIw). Luminescence lifetime
measurements in the same medium also resulted in a small
increase in the lifetimes, between 1–14% (Table S4, ESIw).
Table 1 Longitudinal relaxivities (7 T, 25 1C), luminescence lifetimes,
hydration numbers and quantum yields, of GdL1–4 and TbL1–4 respectively
r1/mMÀ1 sÀ1
GdL
,
tD2Oa/
ms
tH2Oa/
ms
q,
TbL
FH2Ob,
TbL
Complex
LnL1
LnL2
LnL3
LnL4
7.20
7.33
6.65
5.23
3.40
3.23
3.10
3.04
2.01
1.97
1.90
2.20
0.71
0.69
0.72
0.33
0.18
0.18
0.27
0.27
a
b
Lifetimes quoted are subject to Æ10% error. Quantum yields
determined relative to Na3Á[Tb(dpa)3] in 0.1 M Tris buffer (pH 7.4)
at 278 nm excitation, F = 0.265; estimated error Æ20%.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 11534–11536 11535

View Article Online
for use in time-gated imaging techniques. These agents are still
effective when tested in more biologically relevant environ-
ments such as model extracellular medium and in the presence
of cells. Moreover, two of the agents possess a phenol group
which opens the possibility for further coupling reactions to
other functional molecules (e.g. a targeting vector or an
internalisation unit) which can further broaden their potential
applications.
We are thankful to Dr Sven Gottschalk and Ms Hildegard
Schulz, for help performing the biological experiments.
Financial support of the Max-Planck Society and the
German Research Foundation (DFG, grant AN 716/2-1) is
gratefully acknowledged. The European COST Action D38
‘‘Metal-Based Systems for Molecular Imaging Applications’’
is also acknowledged.
Fig. 3 Supernatant (grey) and apparent cell pellet (black) relaxivities
of GdL1–4 and GdDOTA recorded on a 3T MRI scanner at 21 1C.
Notes and references
This shows that the presence of biologically important ions
and amino acids does not significantly inhibit the signals
arising from both types of agents.
1 P. Caravan, Chem. Soc. Rev., 2006, 35, 512–523.
2 J. Yuan and G. Wang, TrAC, Trends Anal. Chem., 2006, 25,
490–500.
To further examine their biocompatability, additional
experiments were performed in a MRI scanner operating at
3 T (128 MHz) by mixing varying concentrations of GdL1–4
(20, 40, 60 and 80 mM) with 3T3 mouse fibroblasts. The
T1 values were obtained from the recorded phantom images
of both the supernatant and the cell pellet. The obtained
results demonstrate that the agents retain their high relaxivities
in the presence of cells within the supernatant (Fig. 3). Since
the exact concentration of the contrast agent in cell pellets
(consisting of extracellular and intracellular volumes) is hard
to assess, an apparent relaxivity r1,cell was determined by
taking the slope of the measured relaxation rates plotted
against the applied concentration. For exclusively extracellular
complexes r1,cell should be lower than the relaxivity in solution
and proportional to the reduced extracellular volume fraction
since the actual concentration within the cell pellet is lower
(compared to the applied concentration). As expected, there
is a loss in r1,cell ranging from 19 to 29% for GdL1–3 within
the cell pellet. A reduction in r1,cell is also observed for the
extracellular agent GdDOTA to a similar extent, under the
same conditions. Interestingly for GdL4, such an effect was
not observed. Rather, the values are almost identical, indi-
cating that an interaction with the cells is occurring (Fig. 3).
To investigate this observation, the complexes GdL1–4 were
incubated with 3T3 cells, the supernatant was removed, the
cells were washed and the T1 values were re-measured. There
is a clear enhancement of the cellular relaxation rate R1,cell
only for GdL4 within the cell pellet by 22% (at 80 mM) relative
to the control. Further studies would be necessary to explain
whether GdL4 simply binds to the cell membrane or inter-
nalises into the cells.
3 J. C. G. Bunzli, Chem. Rev., 2010, 110, 2729–2755.
4 C. P. Montgomery, B. S. Murray, E. J. New, R. Pal and D. Parker,
Acc. Chem. Res., 2009, 42, 925–937.
5 L. Pellegatti, J. Zhang, B. Drahos, S. Villette, F. Suzenet,
G. Guillaumet, S. Petoud and E. Toth, Chem. Commun., 2008,
6591–6593.
6 A. Nonat, C. Gateau, P. H. Fries and M. Mazzanti, Chem.–Eur. J.,
2006, 12, 7133–7150.
7 I. Nasso, C. Galaup, F. Havas, P. Tisnes, C. Picard, S. Laurent,
L. V. Elst and R. N. Muller, Inorg. Chem., 2005, 44, 8293–8305.
8 J. E. Jones, A. J. Amoroso, I. M. Dorin, G. Parigi, B. D. Ward,
N. J. Buurma and S. J. A. Pope, Chem. Commun., 2011, 47,
3374–3376.
9 C. S. Bonnet and E. Toth, C. R. Chim., 2010, 13, 700–714.
´
10 Z. Kotkova, G. A. Pereira, K. Djanashvili, J. Kotek, J. Rudovsky´ ,
P. Hermann, L. Van der Elst, R. N. Muller, C. F. G. C. Geraldes,
I. Lukes and J. A. Peters, Eur. J. Inorg. Chem., 2009, 119–136.
11 S. Aime, M. Botta, E. Terreno, P. L. Anelli and F. Uggeri, Magn.
Reson. Med., 1993, 30, 583–591.
12 M. Botta, Eur. J. Inorg. Chem., 2000, 399–407.
13 J. Rudovsky, M. Botta, P. Hermann, A. Koridze and S. Aime,
Dalton Trans., 2006, 2323–2333.
14 R. Frlan, S. Gobec and D. Kikelj, Tetrahedron, 2007, 63,
10698–10708.
15 A. Dadabhoy, S. Faulkner and P. G. Sammes, J. Chem. Soc.,
Perkin Trans. 2, 2002, 348–357.
16 J. B. Livramento, C. Weidensteiner, M. I. M. Prata, P. R. Allegrini,
C. Geraldes, L. Helm, R. Kneuer, A. E. Merbach, A. C. Santos,
P. Schmidt and E. Toth, Contrast Media Mol. Imaging, 2006, 1,
30–39.
17 V. Kubicek, J. Rudovsky, J. Kotek, P. Hermann, L. V. Elst,
R. N. Muller, Z. I. Kolar, H. T. Wolterbeek, J. A. Peters and
I. Lukes, J. Am. Chem. Soc., 2005, 127, 16477–16485.
18 P. Lebduskova, P. Hermann, L. Helm, E. Toth, J. Kotek,
K. Binnemans, J. Rudovsky, I. Lukes and A. E. Merbach, Dalton
Trans., 2007, 493–501.
19 I. Mamedov, A. Mishra, G. Angelovski, H. A. Mayer, L. O. Palsson,
D. Parker and N. K. Logothetis, Dalton Trans., 2007, 5260–5267.
20 J. Rudovsky, P. Cigler, J. Kotek, P. Hermann, P. Vojtisek,
I. Lukes, J. A. Peters, L. Vander Elst and R. N. Muller,
Chem.–Eur. J., 2005, 11, 2373–2384.
21 A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker,
L. Royle, A. S. de Sousa, J. A. G. Williams and M. Woods,
J. Chem. Soc., Perkin Trans. 2, 1999, 493–503.
22 J. C. G. Bunzli, P. Froidevaux and J. M. Harrowfield, Inorg.
Chem., 1993, 32, 3306–3311.
In summary, these four ligand systems show great potential
for application in bimodal imaging, satisfying many of the
prerequisites required for both luminescence and MRI. They
are readily prepared from cheap starting materials, employing
an effective transformation to yield a multifunctional aryl
phosphonate moiety. Incorporating Gd3+ leads to high relax-
ivity agents that retain their effectiveness at high magnetic
fields, whilst inclusion of Tb3+ results in systems with long
lived luminescence lifetimes and high quantum yields, suitable
23 P. Atkinson, B. S. Murray and D. Parker, Org. Biomol. Chem.,
2006, 4, 3166–3171.
24 J. Henig, I. Mamedov, P. Fouskova, E. Toth, N. K. Logothetis,
´
G. Angelovski and H. A. Mayer, Inorg. Chem., 2011, 50, 6472–6481.
c
11536 Chem. Commun., 2011, 47, 11534–11536
This journal is The Royal Society of Chemistry 2011