Modulation of the lifetime of water bound to lanthanide metal ions in complexes with ligands derived from 1,4,7,10-tetraazacyclododecane tetraacetate (DOTA)

Article abstract of DOI:10.1002/hlca.200590086

A series of di- and tetraamide derivatives of DOTA were synthesized, and their lanthanide(III) complexes were examined by multinuclear ¹H-, ¹³C-, and ¹⁷O-NMR spectroscopy, and compared with literature data of similar, know

Full text of DOI:10.1002/hlca.200590086

Helvetica Chimica Acta ± Vol. 88 (2005)
923
Modulation of the Lifetime of Water Bound to Lanthanide Metal Ions in
Complexes with Ligands Derived from 1,4,7,10-Tetraazacyclododecane
Tetraacetate (DOTA)
by Shanrong Zhanga), Xiuyan Jiangb), and A. Dean Sherry*b)c)
a) Department of Radiology, University of Washington Medical Center, P. O. Box357115,
1959 N. E. Pacific Street, Seattle, WA 98195-7115, U.S.A.
b) Department of Chemistry, University of Texas at Dallas, P. O. Box 830688, Richardson, TX 75083-0688, U.S.A.
c) Rogers Magnetic Resonance Center, Department of Radiology, University of Texas Southwestern Medical
Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9085, U.S.A.
(phone: ‡ 1-972-883-2907; fax: ‡ 1-972-883-2925; e-mail: sherry@utdallas.edu)
Â
This paper is written in celebration of the 65th birthday of Professor Andre E. Merbach
A series of di- and tetraamide derivatives of DOTA were synthesized, and their lanthanide(III) complexes
were examined by multinuclear 1H-, 13C-, and 17O-NMR spectroscopy, and compared with literature data of
similar, known complexes (Table). All ligands formed structures similar to the parent [LnIII(DOTA)]À
complexes, with four N-atoms and four O-atoms from DOTA and one O-atom from the inner-sphere water
molecules. Interestingly, the lifetimes tM of the inner-sphere, metal-bound water molecules vary widely, ranging
from nano- to milliseconds, depending on the identity of the pendent amide side chains. In general, positively
charged [LnIII(DOTA-tetraamide)]3‡ complexes display the longest residence times (high tM values), while
complexes with additional charged functional groups on the extended amides display much smaller tM values,
even when the side groups are not directly coordinated to the central Ln3‡ ions. The design of novel
[LnIII(DOTA-tetraamide)]3‡ complexes with a wide, tunable range of tM values is of prime importance for the
application of fast-responding, paramagnetic chemical-exchange-saturation-transfer (PARACEST) imaging
agents used for the study of physiological and metabolic processes.
1. Introduction. ± Magnetic-resonance imaging (MRI) has rapidly become one of
the dominant imaging modalities used in clinical medicine. In addition to providing
exquisite high-resolution anatomical images, MRI can also provide insights into
abnormal tissue structures (MT imaging) and flow (angiography). Injection of a drug to
further enhance image contrast is currently practiced in about one third of all clinical
MRI studies. Low-molecular-weight gadolinium (Gd3‡) complexes that feature a single
inner-sphere-bound water molecule in rapid exchange with bulk tissue water are most
commonly used as contrast agents (CA) [1]. The relaxation efficiency or relaxivity
(symbolized by r1) of typical Gd3‡-based CA currently in use is relatively low (r1 ꢀ
4 mmÀ1 sÀ1), much lower than theoretically possible (ca. 120 mmÀ1 sÀ1) [2]. Such low
relaxivity values do not limit the use of CA as bulk extracellular relaxation agents, but
will likely become a limitation when applying MRI to molecular imaging of selected
biological targets or specific cellular events [3].
During the early years of CA development, it was assumed that the rate of exchange
of water molecules bound to lanthanide (Ln) ions in complexes was fast enough not to
impinge upon relaxivity. However, more-recent studies, catalyzed by the pioneering
ꢀ 2005 Verlag Helvetica Chimica Acta AG, Zürich

924
Helvetica Chimica Acta ± Vol. 88 (2005)
work of Andre Merbach and his group, have revealed that this is not, in fact, the case.
For example, the water residence lifetimes (tM) of two common clinical agents,
[GdIII(DTPA)]2À and [Gd(DOTA)]À, are 303 and 240 ns, respectively [4][5], some
two orders of magnitude higher than that of the aqua Gd3‡ ion [6 ± 11]. Merbach and co-
workers [12] also discovered that tM of bound water in the neutral diamide complex
[GdIII(DTPA)(BMA)]0, was even higher, ca. 2.2 ms. It is now clear that CA that
undergo such a slow water exchange become a limitation in our attempts to create high-
relaxivity contrast agents. Subsequent physical studies, inspired by the work of
Merbach, have led to new insights into the dynamics of Ln complexes, and some of the
factors that govern water exchange in such complexes.
It has been estimated that relaxivity could be optimized by slowing the rotation of a
Gd3‡ complexif tM of that complexis on the order of 20 ± 30 ns [2]. There have been
numerous experimental attempts to arrive at this optimal water exchange rate, for
instance by 1) introduction of a pyridine into a macrocyclic ring [13 ± 16], 2)
substitution of acetate pendent arms by more bulky methylene phosphonate arms
[17], 3) insertion of additional CH2 units into either the backbone or pendent arms of
DTPA-like ligands [18 ± 20], 4) new ligand design based upon tripodal complexes [21],
and 5) ꢁlockingꢂ the chelate into a single coordination geometry having fast water
exchange [22]. Each of these approaches has resulted in the production of complexes
with more-favorable water exchange rates, yet nobody has successfully used one of
these ꢁoptimizedꢂ exchange systems to achieve the maximum relaxivity predicted by
theory by simply attaching one of these systems to a slowly tumbling macromolecule.
Clearly, these are dynamic systems, and other factors such as internal motions, bond
rotations, or water exchange limitations (when a low-molecular-weight agent is
attached to a larger protein), ultimately limit the observed relaxivity.
In the following discussion, we will compare a series of known and new complexes
of Eu3‡, Gd3‡, and Dy3‡ and the ligands 1 ± 17 with respect to lifetime of water exchange
and analytical method. The results are summarized in the Table.
The scientific curiosity about factors that control water exchange in Ln complexes
has catalyzed new interest in more-slowly exchanging systems. Aime et al. [23] were the
first to observe a separate 1H-NMR signal for the bound water molecule in
[EuIII(10)]3‡, a system with a tM value of ca. 120 ms for the M-isomer1). This was
quickly followed by similar reports for other tetraamide derivatives, including the
corresponding N-methylated compound 11 [24] (or the dimethyl variant thereof [24])
or the sterically more-crowded N-(2-phenylpropyl) analogue 13 [25]. The tM values of
the M-isomers for this series of complexes seem to vary with increasing steric bulk of
the amide pendent arms, increasing from 120 to 156 to 278 ms (in MeCN at 298 K) for
[EuIII(10)]3‡ [24][26], [EuIII(11)]3‡ [24], and [EuIII(13)]3‡ [25], respectively (Table).
However, factors other than steric bulk may also influence water exchange, as
evidenced by the observation that a Eu3‡-bound water resonance has been detected in
the high-resolution 1H-NMR spectrum of [EuIII(15)]3‡ in pure water at ambient
temperature [27]. This observation led us to investigate the use of such systems to alter
1
)
The so-called ꢁM-isomerꢂ (not to be confused with the helicity descriptor M) refers to a square-antiprism
geometry. Twisted-square-antiprism isomers, in turn, are commonly termed ꢁm-isomersꢂ (not to be
confused with the locant m for meta).

Helvetica Chimica Acta ± Vol. 88 (2005)
925
MRI contrast not by altering relaxation rates, but rather by changing the intensity of the
bulk-water signal detected in the imaging experiment by means of a chemical-
exchange-saturation-transfer (CEST) mechanism [28]. We and others have since
shown that such complexes offer a new avenue in the design of ꢁsmartꢂ contrast agents,
because chemical exchange in these systems is exquisitely sensitive to changes in pH
[29 ± 31], temperature [32], and metabolism [33][34]. Systems with a slow water
exchange may offer another advantage in that one can then modulate prototropic
exchange (rather than molecular exchange) to create a T1 relaxation agent that
responds to pH. This feature makes the water relaxivity of [GdIII(9)]5À highly pH-
sensitive over the important physiological range [35][36]. Indeed, [GdIII(9)]5À has
recently been used in vivo to map the distribution of tissue pH in functioning rat kidney
[37] and in brain tumors (unpublished data).
Clearly, the tunability of the lifetime of Ln3‡-bound water has become a focal point
for many investigators interested in the development of novel, ꢁsmartꢂ MRI contrast

926
Helvetica Chimica Acta ± Vol. 88 (2005)
Table. Characterization of Lanthanide Complexes of Type [LnIII(L)] (L ˆ ligand). The ligands 1 ± 17 (see
chemical formulae) and the Ln3‡ metal ions, the number of solution species1) and their ratio, the solvents used,
and the measured lifetimes tM of metal-bound water are given. All data refer to a temperature of 298 K; Wˆ
water.
1
Ligand
Ln3‡
Species
M/m )
Method
Solvent
tM [ms]
Ref.
none
Gd
Eu
Gd
Gd
Gd
Gd
Gd
±
17O-NMR
1H-NMR
17O-NMR
17O-NMR
17O-NMR
17O-NMR
17O-NMR
W
W
W
W
0.002
[8]
1
1
2
3
4
5
2
2
2
1
1
2
80 : 20
-
90 : 10
±
±
[51]
[5][52]
[53]
[22]
[22]
[46]
0.240a)
0.200a)
W (pH 7)
W (pH 7)
W (pH 7)
0.120 (M)
0.015 (m)
±
60 : 40
1.35 Æ 0.07 (M)
0.014 Æ 0.001 (m)
2.35 Æ 0.12 (M)
0.003 Æ 0.001 (m)
10.0 Æ 0.01 (M)
1.27
25.5 Æ 0.5 (M)
21.2 Æ 1.2 (M)
120 (M)
6
Gd
2
83 : 17
17O-NMR
W (pH 7)
[46]
7
8
9
Eu
Gd
Gd
Dy
Eu
Eu
1
1
1
1
2
2
±
±
±
±
1H-NMRb)
T1 (1H)
17O-NMR
17O-NMR
1H-NMRb)
CEST
CD3CN
this work
this work
this work
[35]
[23]
[42]
W (pH 7)
W (pH 7.2)
W (pH 6 ± 9)
CD3CN
10
70 : 30
±
W (pH 7)
2 (m)
89.5 Æ 16.3 (M)
11
12
13
14
15
Eu
Eu
Eu
Eu
Gd
Gd
Eu
Eu
Eu
Gd
Eu
2
2
2
1
1
70 : 30
60 : 40
±
±
±
1H-NMRb)
1H-NMRc)
1H-NMRb)
1H-NMRb)
T1 (1H)
CD3CN
CD3CN
CD3CN
156 (M)
[26]
this work
[25]
146 Æ 5 (M)
278 Æ 1 (M)
262 Æ 1 (M)
190 Æ 10
W (pH 7.4)
W (pH 7)
CD3CN
W (pH 7)
CD3CN
CD3CN
W (pH 7)
CD3CN
[47]
this work
this work
[27]
[27]
this work
[36]
17O-NMR
1H-NMRb)
1H NMRb)
1H-NMRb)
T1 (1H)
250 Æ 10
369 Æ 5 (M)
789 Æ 101 (M)
1254 Æ 98 (M)
805 Æ 8
16
17
1
1
±
±
1H-NMRb)
1346 Æ 6 (M)
[36]
a) Average. b) Direct method. c) Saturation transfer.
agents that alter image contrast by either relaxation or a CEST mechanism. It is now
known that tM in Ln complexes is exquisitely sensitive to both the size of the central
metal cation [38] and the ligand structure [39]. Merbach has played a key role in the
development of this field and, in celebration of his contributions, we wish to report
some new data, and to review some older ones, for a series of selected di- and
tetraamide ligands based on DOTA, as well as their complexes formed with Ln3‡ ions.
2. Results and Discussion. ± 2.1. Synthesis of Modified DOTA Ligands. The general
procedures for the preparation of the ligands are summarized in the Scheme. Starting
from cyclen (ˆ1,4,7,12-tetraazacyclododecane; 18), the corresponding tetraamide
derivatives with four identical pendent arms were readily prepared by reaction with
various N-alkyl or other 2-bromoacetamides in MeCN in the presence of a slight excess
of K2CO3. However, a sequence of protection/deprotection steps was required to
prepare mixed tetraamides with two types of pendent arms [40]. Typically, the 1,7-N-
atoms of cyclen (18) can be selectively protected by reacting 18 with benzyl

Helvetica Chimica Acta ± Vol. 88 (2005)
927
Scheme. Synthesis of Modified DOTA Ligands
chloroformate and NaOH at ca. pH 2 ± 3. This is then followed by alkylation of the
remaining two N-atoms, followed by removal of the (benzyloxy)carbonyl protective
groups by catalytic hydrogenation (H2, Pd/C), and, finally, by addition of the second
type of amide substituents.
2.2. Determination of the Lifetime (tM) of Bound Water. There are numerous NMR
methods available for measuring tM for Ln-bound water, including variable-temper-
ature (VT) 17O-NMR [1], T1 nuclear-magnetic-relaxation-dispersion (NMRD) profiles

928
Helvetica Chimica Acta ± Vol. 88 (2005)
[41], VT-1H-NMR band-shape simulations [27], VT water-relaxivity measurements
[27][36], and CEST spectral fitting [42]. Each method has its advantages and
limitations. Several examples will be discussed below.
2.2.1. VT-17O-NMR. This method, widely applied by Merbach and colleagues [1],
has proven effective in measuring tM over a wide range. It is especially effective for Ln
complexes whose relaxation enhancement is relatively large (Gd, Dy, Tb, etc.), but
ineffective for Ln complexes whose relaxation enhancement of the 17O-NMR signal is
small (Sm, Eu, Yb, etc.). The method is also less sensitive when tM for Ln3‡-bound
water is either too small (fast exchange) or too large (slow exchange). Examples of
such extremes may be seen in the VT-17O-NMR curves for [GdIII(2)]À, [GdIII(4)]À, and
[GdIII(9)]5À, as shown in Fig. 1. Obviously, the tM value for [GdIII(2)]À can be more
reliably determined because of the features presented by these data, i.e., a character-
istic point of inflection [22]. In contrast, the lifetimes of [GdIII(4)]À and [GdIII(9)]5À are
less accurate due to their featureless 17O-NMR curves (no point of inflection).
Fig. 1. Variable-temperature 17O-NMR linewidth curves for [GdIII(2)]À, [GdIII(4)]À, and [GdIII(9)]5À. The
curves were fitted to the SBM theory according to published procedures [12].
1
2.2.2. Variable-Temperature H-NMR Relaxation Time. If the exchange of bound
water is extremely slow, the inner-sphere contribution will be negligible at low
temperatures, and will gradually increase at higher temperatures. Examples of this
behavior from the literature include [GdIII(15)]3‡ and [GdIII(17)]3‡ [27][36]. Here, we
show that [Gd(8)]5À presents a similar behavior (Fig. 2). By fitting the data to the
standard Solomon ± Bloembergen ± Morgan (SBM) theory, a tM value of 1.27 ms was
obtained. The advantage of this method is its relative simplity, but it is suitable only for
systems with tM values on the order of microseconds.

Helvetica Chimica Acta ± Vol. 88 (2005)
929
Fig. 2. Temperature dependence of the 20-MHz relaxivity of [GdIII(8)]5À in aqueous solution at pH 7.0. The solid
line reflects the best fit of the data to the SBM theory, the dotted lines show the individual outer-sphere (OS)
and inner-sphere (IS) contributions [27][36].
2.2.3. Saturation Transfer. Magnetization-transfer techniques have historically
found many applications in chemical and biological systems [43]. Recently, it has
been found that the bound water molecule in a series of Eu complexes with DOTA
tetraamides exchanges so slowly that a separate, highly shifted bound water signal
could be directly observed by high-resolution 1H-NMR [23 ± 27]. This offers the
opportunity to use an entirely different mechanism for producing image contrast based
upon CEST. Paramagnetic versions of slow-chemical-exchange molecules are now
referred to as PARACESTagents [44]. If chemical exchange in these systems meets the
slow-to-intermediate exchange regime condition (Dw  tM »1), then the degree of
saturation of bulk water (Mt/M0) will depend upon the chemical exchange rate (kobs
and the relaxation properties of the system according to Eqn. 1.
)
ꢀ
ꢁ
ꢀ
ꢁ
Mt
M0
1
kobsT1sat
ꢁ1 ‡ kobsT1sat
ꢁ1 ‡ kobsT1sat†
ˆ
‡
exp À
 t
ꢁ1†
ꢁ1 ‡ kobsT1sat
†
†
T1sat
Here, Mt and M0 are signal intensities of bulk water with and without saturation at
the exchanging site, respectively, for a period of time t; T1sat is the spin ± lattice
relaxation time of bulk water H-atoms during steady-state saturation at the bound-
water site; kobs is the pseudo-first-order exchange rate between bulk water and the H-
atoms of bound water, equivalent to the concentration ratio of the bound-water site
relative to bulk water, divided by the lifetime tM of the bound-water site. Thus, Eqn. 1 is
an alternative way to measure tM of Ln complexes. An example of such data is given in

930
Helvetica Chimica Acta ± Vol. 88 (2005)
Fig. 3 for [EuIII(12)]3‡. By adding different amounts of water (5 ± 100 ml) to
[EuIII(12)(H2O)]3‡(CF3SO3À )3 (30 mg) in CD3CN (0.4 ml), the relative concentrations
of the bound and bulk pools change, and this is reflected by large differences in the
equilibrium magnetization, Mt/M0, of bulk water at steady-state. Fitting of those data to
Eqn. 1 gives an average tM value of 146 Æ 5 ms, which is similar to values reported for
other DOTA tetraamide complexes, as determined by other methods (Table). One
should keep in mind that Eqn. 1 has an approximation of approaching steady-state, and
assumes complete saturation of the bound pool. It practice, however, full saturation of
the bound pool is experimentally difficult to achieve. Recently, we reported a numeric
method for solving the modified Bloch equations that does not require such
assumptions [42].
Fig. 3. 1H-NMR Saturation-transfer experiments with [Eu(12)(H2O)](CF 3SO3À )3. To a soln of the complex
(30 mg) in anhydrous CD3CN (0.4 ml) were added 5, 20, 50, or 100 ml of H2O. The saturation frequency was set
to the bound H2O resonance (55 ppm rel. to bulk H2O at 0 ppm), with a B1 value of 1200 Hz and different
saturation duration times (0 ± 10 s). The experiments were performed on a Varian Inova-500 spectrometer with
a PFG/ID probe. The solid lines reflect the fits of the data to Eqn. 1 (see text) for a tM value of 146 Æ 5 ms.
2.3. Factors Affecting tM. 2.3.1. Lanthanide Coordination Geometry. Early NMR
studies of the [LnIII(DOTA)]À complexes showed that two slowly interconverting
species are present in aqueous solution, and that the populations of these two species
are sensitive to the size of the central Ln3‡ cation [45]. These two coordination isomers
have since been assigned to the M- and m-isomers1) of square-antiprism and twisted-
square-antiprism geometries, respectively, related to each other by rotation of the side-
chain CH2C(O) groups attached to the four central N-atoms of DOTA. Interestingly, it
was subsequently discovered that these two species give rise to quite different tM values,
the m-isomer showing much faster exchange compared to the M-isomer. For example,
in the case of [EuIII(10)]3‡, tM is 120 ms for the M-, but only 2 ms for the m-isomer

Helvetica Chimica Acta ± Vol. 88 (2005)
931
[23][24][26]. Analogously, values of 1.35 vs. 14 ns, and 2.35 vs. 3 ns were found for the
‡
‡
M- and m-isomers of [GdIII(5)] and [GdIII(6)] , respectively [46]. Such dramatic
differences were recently confirmed by the successful separation and characterization
of the M- and m-isomers of the same molecule differing only in chirality (tM ˆ 120 ns
for [GdIII(3)]À vs. 15 ns for [Gd(4)]À) [22] .
2.3.2. Complex Rigidity. Ligands 2 and 3 differ from DOTA (1) by an additional 4-
nitrobenzyl group on the macrocyclic backbone (for 2), or by a combination of
backbone substituents including Me groups in the a-positions of the pendant arms (for
3). This provides ligands with increasing rigidities in the series [GdIII(1)]À <
[GdIII(2)]À < [GdIII(3)]À. Interestingly, the measured tM values for this series decrease
with increasing rigidity [tM values of 240 ms with ligand 1 (average of two coordination
isomers), 200 ns with 2 (average of two coordination isomers), and 120 ns with 3 (single
M-isomer)]. This indicates that water exchange becomes faster when the ligand
architecture is rigidified by substitution. One could argue that the tM values reported
for the Gd complexes with 1 and 2 have an exchange component that partially reflects
the presence of the faster-exchanging m-isomer; so the differences described above
may be even larger if exchange could be measured for the pure M-isomer in each case.
A second illustration of this is provided by the complexes [GdIII(8)]5À and [GdIII(9)]5À.
Here, water exchange in the less-rigid complex with ligand 9 is ca. 20-times slower than
in the more-rigid complexwith 8 (tM ˆ 26 vs. 1.3 ms, resp.). This limited series may
provide a better indicator of the effects of rigidity, because, in this case, both complexes
exist nearly exclusively as the M-isomers.
2.3.3. Side Chains of Ligands. The data summarized in the Table reveal that other
chemical characteristics of the extended amide side chains, especially polarity and
charge, also affect tM, even though they are not typically involved in coordination to the
central Ln ion. In general, for [EuIII(DOTA-tetraamide)] complexes, tM correlates with
the polarity of the ligand side chains, i.e., phosphonates ꢂ carboxylates»alkyl groups ꢂ
simple amides. For example, the measured tM values for one such series falls in the
order [EuIII(17)]3‡ (1.3 ms) [36] ꢂ [EuIII(16)]3‡ (1.25 ms) > [EuIII(15)]3‡ (780 ms)
[27]»[EuIII(13)]3‡ (278 ms) [25] > [EuIII(12)]3‡ (146 ms) and [EuIII(11)]3‡ (156 ms)
[24] > [EuIII(10)]3‡ (120 ms) [23], as determined in each case for the corresponding
M-isomers in hydrous MeCN solution. A correlation between tM and a calculated
water-accessibility parameter has been reported for an even more-extensive series of
Eu3‡ complexes [39].
Complexes with negatively charged side chains generally display lower tM values
‡
than their neutral counterparts. For example, the tM values of [EuIII(5)] and
‡
[EuIII(6)] (1.3 ± 2.3 ms) are ca. five times smaller than that of [EuIII(7)]3‡ (10 ms); and
the tM value measured for [GdIII(9)]5À (25 ms, at pH 7.2) is notably smaller than that of
the corresponding ester, [GdIII(17)]3‡ (805 ms, at pH 7.4). Similarly, the tM value found
for [EuIII(14)]À (262 ms, pH 7.4) is somewhat lower than the one for the corresponding
ester, [EuIII(15)]3‡ (382 ms, pH 7.4) [27][47].
2.4. Molecular vs. Prototropic Exchange. Aime et al. [48] recently published a review
article on the relaxivity contributions from the prototropic exchange vs. exchange of
water in Gd3‡ complexes. It has been observed that prototropic exchange is accelerated
in complexes of the type discussed here in the presence of acids and/or bases, and that
an increase in prototropic exchange results in an increase in relaxivity below and above

932
Helvetica Chimica Acta ± Vol. 88 (2005)
ca. pH 2 and 9 ± 10 , respectively [49]. This early observation prompted us to examine
more-closely the water relaxivity of [GdIII(9)]5À as a function of pH. In this interesting,
1
pH-sensitive complex, the H relaxivity of water steadily increases between pH 8.5
(r1 ˆ 4.5 mmÀ1 sÀ1) and pH 6 (r1 ˆ 9.8 mmÀ1 sÀ1) [35]. It has been shown that this effect is
not due to an increase in the exchange of water molecules over this pH range [35], but
is rather due to an increase in prototropic-exchange rate catalyzed by the appended
phosphonate groups. The phosphonate groups in this complexundergo four distinct
protonations between pH 8.5 and 6. This implies that the monoprotonated species, with
a single PO3HÀ function, are most efficient at catalyzing proton exchange between the
Gd3‡-bound water H-atoms and bulk solvent. Such pH-dependent relaxivities are not a
feature of current clinical Gd3‡-based contrast agents, where water exchange is
relatively fast (tM < 240 ns), so the retarded exchange feature of [GdIII(9)]5À makes this
complexan attractive agent for mapping extracellular pH in tissues by MRI [37].
2.5. Size of the Metal Cation. Merbach and co-workers [12] found for a series of
[Ln(DTPA)(BMA)] complexes that the metal-bound water lifetimes tM decrease from
1.88 ms (Nd3‡) to 0.16 ms (Ho3‡) and suggested that this reflects a change in water
exchange mechanism from interchange (Ia) to dissociative (D) type upon going from
larger towards smaller cations. More recently, Parker and co-workers [25] reported that
the tM value of [EuIII(13)]3‡ might be as much as 500 times higher than that of the
corresponding Yb3‡ complex. We have also observed that the tM values vary
dramatically with Ln ionic radii for the [LnIII(15)]3‡ series [38]. It is especially
interesting that the tM values in [EuIII(15)]3‡ [27] and [Eu(17)]3‡ [36] are about twice as
high as those of the corresponding Gd3‡ complexes, even though Eu3‡ and Gd3‡
complexes are usually considered isostructural because of their similar ionic radii (0.95
and 0.94 Š, respectively).
2.6. Solvent Effects. The data summarized in the Table also indicate that the life-
times of Ln3‡-bound water for identical complexes measured in hydrous MeCN vs. bulk
water may differ substantially. For example, tM values of 789 vs. 369 ms, and 120 vs.
89 ms, have been determined for [EuIII(15)]3‡ and [EuIII(11)]3‡, respectively, in MeCN
vs. water [27]. This may be related to the solvent accessibility, as suggested previously
[39], or reflect a larger contribution from prototropic exchange in bulk water. Given
that the mechanism of water exchange in these tetraamide complexes occurs via a D-
mechanism, greater solvent accessibility to the Eu3‡-bound water molecule may aid
dissociation of the bound water by H-bonding with bulk water.
Conclusions. ± Water exchange has arguably become the most-interesting physical
parameter in Ln complexes that may be manipulated by clever chemistry and used to
advantage in MRI applications. We have summarized in this article the NMR methods
used to measure water exchange, and have highlighted the influence of complex charge,
ionic radius of the central metal, solvent accessibility, coordination geometry, and
molecular rigidity on the lifetimes tM of inner-sphere bound water. It is now possible to
design Ln complexes that display tM values ranging over sixorders of magnitude, i.e.,
from nano- to milliseconds.
The following general observations can be made: 1) [Ln(DOTA-tetraamide)]
complexes have much higher tM values compared to [Ln(DOTA-diamide)] complexes;
2) additional charged functional groups on the pendent amide arms reduce the tM

Helvetica Chimica Acta ± Vol. 88 (2005)
933
values greatly, even though they may not be directly involved in coordination to the
central Ln cation; 3) tM values are very sensitive to the ionic radius of the central Ln
ion, and can even vary for ions with very similar ionic radii such as Eu3‡ and Gd3‡; and
4) the solvent obviously influences water exchange in these complexes. For
[EuIII(DOTA-tetraamide)] complexes, the lifetime of bound water is typically ca.
twice as fast in bulk water compared to hydrous MeCN. Hopefully, these observations
may be helpful in design of new Ln-based contrast agents for MRI that have optimal
dynamic properties for either fast (Gd-based relaxivity) or slow (other Ln-based
CEST) contrast applications.
This work was supported by Grants from the NIH (RR-02584 and CA 84697), the Robert A. Welch
Foundation (AT-584), and the Texas Advanced Technology Program (ATP).
Experimental Part
The following chemicals and solvents were purchased from Sigma-Aldrich (Milwaukee, WI), and used
without further purification: benzyl chloroformate, isobutylamine, bromoacetylbromide, tert-butyl bromoace-
tate, diisopropyl(ethyl)amine, 10% Pd/C, 1,4-dioxane, anh. MeCN, anh. Et2O, and anh. EtOH. Cyclen
(ˆ1,4,7,12-tetraazacyclododecane; 18) was purchased from Strem Chemicals (Newburyport, MA). Ligand 8
(and the corresponding GdIII complex) was synthesized by Fredrick Tajada (Ph.D. Thesis, University of Texas at
Dallas, 2002). Column chromatography (CC) was performed with SilicAR-150 silica gel (60 ± 200 mesh;
Mallinckrodt). Hydrogenations were performed in a Parr hydrogenation apparatus. NMR Data: d in ppm rel. to
Me4Si as internal standard.
2-Bromo-N-(2-methylpropyl)acetamide (reagent for the synthesis of 7, 12, and 16). In a N2-purged, three-
neck 100-ml flask, isobutylamine (4.97 ml, 0.05 mol) was dissolved in benzene (60 ml), and K2CO3 (8.293 g,
0.05 mole) was added. Bromoacetyl bromide (4.36 ml, 0.05 mol) was added dropwise over 30 min, while
maintaining the temp. at 08. The mixture was then warmed to r.t. and stirred overnight. The solids were removed
by filtration, and the solvent was evaporated under reduced pressure. The remaining residue was dissolved in
Et2O (20 ml), washed with H2O (3 Â 20 ml), and the combined org. extract was dried (Na2SO4) overnight. The
remaining solid was removed by filtration, and the solvent was evaporated in vacuo to afford colorless crystals
(7.77 g) in 80% yield. 13C-NMR (75 MHz, CDCl3): 165.42 (CˆO); 47.51 (NHCH2); 29.51 (BrCH2); 28.40 (CH);
20.05 (2 Me).
N,N',N'',N'''-Tetrakis(2-methylpropyl)-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetamide (12). To a
soln. of cyclen (18; 0.86 g, 0.005 mol) in MeCN (20 ml) were added 2-bromo-N-(2-methylpropyl)acetamide
(3.88 g, 0.020 mol) and K2CO3 (3.0 g, 0.022 mol). The mixture was heated at 60 ± 708 overnight. The resulting
solid was removed by filtration, CHCl3 was added, and the mixture was heated to reflux. The soln. was
concentrated and cooled to yield 12 (2.86 g, 91%). Colorless solid. 1H-NMR (270 MHz, CDCl3): 6.97 (t, NH);
3.05 (s, CH2CO); 2.70 (s, CH2CH2); 1.75 (m, CH); 0.88 (d, Me). 13C-NMR (75 MHz, CDCl3): 170.69 (CˆO);
57.83 (NCH2); 50.20 (CH2 (ring)); 46.24 (CH2NH); 28.26 (CH); 19.70 (Me).
Bis(phenylmethyl) 1,4,7,10-tetraazacyclododecane-1,7-dicarboxylate (19). Prepared according to a litera-
ture procedure [40].
Bis(phenylmethyl) 4,10-Bis[2-[(2-methylpropyl)amino]-2-oxoethyl]-1,4,7,10-tetraazacyclododecane-1,7-di-
carboxylate (20). In a 100-ml three-neck flask, 19 (2.2 g, 0.005 mol) was dissolved in MeCN (20 ml). Then, Et(i-
Pr)2N (1.5 g) in MeCN (7 ml) was added at 08, followed by 2-bromo-N-(2-methylpropyl)acetamide (2.13 g,
0.011 mol) in MeCN (7 ml). The mixture was stirred at 08 for 2 h, then warmed to 608, and stirred for an
additional 20 h at this temp. The solvent was removed, and the residue was partitioned between Et2O (100 ml)
and H2O (20 ml). The org. layer was washed with H2O (2 Â 20 ml) and 5% aq. NaOH soln. (2 Â 20 ml). The org.
layer was dried (Na2SO4) overnight. The solvent was removed in vacuo to afford the crude product as light-
yellow oil, which was purified by CC (SiO2; Et2O) to yield the title compound as a colorless oil (2.73 g, 82%).
1H-NMR (270 MHz, CDCl3): 7.28 (m, arom. H); 5.02 (s, COOCH2); 3.42, 2.73 (CH2 (ring)); 3.38 (NCH2); 3.12
(NHCH2); 2.96 (br., NH); 1.71 (br., CH); 0.83 (d, Me). 13C-NMR (75 MHz, CDCl3): 170.95 (CH2CO); 156.85
(COO); 136.31, 128.68, 128.38, 128.19 (arom. C); 67.51 (NCH2); 58.85 (OCH2); 55.94, 48.51 (CH2 (ring)); 46.66
(NHCH2); 28.58 (CH); 20.31 (Me).

934
Helvetica Chimica Acta ± Vol. 88 (2005)
N,N'-Bis(2-methylpropyl)-1,4,7,10-tetraazacyclododecane-1,7-diacetamide (21). Compound 20 (see above;
1.1 g, 1.65 mmol) was dissolved in anh. EtOH (20 ml), and 10% Pd/C (200 mg) was added. The mixture was
hydrogenated at 20 ± 25 psi for 48 h. The catalyst was filtered off, and the solvent was removed in vacuo to afford
21 as a colorless oil (0.643 g, 98%). 1H-NMR (270 MHz, CDCl3): 7.57 (t, NHCH2); 7.19 (br., NH); 3.20 (s,
CH2CO); 2.90 (dd, CH2); 2.70 (br., CH2 (ring)); 1.65 (m, CH); 0.72 (d, Me). 13C-NMR (75 MHz, CDCl3): 171.32
(CˆO); 59.91 (NCH2); 52.65 (CH2 (ring)); 46.85 (CH2 (ring)); 46.62 (NHCH2); 28.37 (CH); 20.29 (Me).
Tetraethyl [(4,10-Bis{2-[(2-methylpropyl)amino]-2-oxoethyl}-1,4,7,10-tetraazacyclododecane-1,7-diyl)-
bis[(1-oxo-2,1-ethanediyl)iminomethylene]]bis(phosphonate) (16). To a soln. of 21 (0.643 g, 1.61 mmol) in
anh. MeCN (10 ml) were added K2CO3 (0.446 g) and diethyl {[(bromoacetyl)amino]methyl}phosphonate
(0.930 g, 3.23 mmol) [50]. The mixture was stirred at 408 for 14 h, and then at 608 for 8 h. The solvent was
removed on a rotary evaporator, and the residue was partitioned between Et2O (20 ml) and H2O (10 ml). The
org. layer was washed with H2O (2 Â 10 ml), 5% aq. NaOH soln. (10 ml), and H2O (2 Â 10 ml), dried (Na2SO4)
overnight, and evaporated in vacuo to afford 16 (0.821 g, 63%). 13C-NMR (75 MHz, CDCl3): 171.44 (CˆO);
170.90 (CˆO); 62.71 (CH2 Me); 60.17 (NCH2); 58.29 NCH2); 54.32, 52.93 (CH2 (ring)); 46.47 (CH2CH); 34.63
(d, CH2P); 28.57 (CH); 20.27 (Me); 16.53 (CH2Me).
Bis(1,1-Dimethylethyl) 10-Bis[2-[(2-methylpropyl)amino]-2-oxoethyl]-1,4,7,10-tetraazacyclododecane-1,7-
diacetate (7). Compound 21 (0.485 g, 1.22 mmol) was alkylated with tert-butyl bromoacetate (0.475 g,
2.44 mmol) in the presence of K2CO3 (0.337 g), as described above, to afford 7 (0.598 g) as a pale yellow oil.
13C-NMR (75 MHz, CDCl3): 171.49 (CˆO); 170.92 (CˆO); 50 ± 52 (CH2 (ring)); 57.95 (NCH2); 56.18 (NCH2);
46.24 (NHCH2); 28.28 (Me2CH); 19.66 (Me2CH), 81.15 (CMe3); 27.52 (CMe3).
Lanthanide Complexes. The complexes were synthesized by mixing the appropriate macrocyclic ligand with
a lanthanide chloride (or triflate) in either MeCN or water according to established procedures [27].
NMR Spectroscopy and Data Analysis. Longitudinal-relaxation measurements for water H-atoms were
performed with an MRS-6 analyzer operating at 20 MHz (Institut Jozef Stefan, Ljubljana, Slovenija) and an
inversion-recovery pulse sequence (15 experiments, eight scans). The experiments were conducted under
thermostated conditions (Æ0.58) by means of an air-flow heater equipped with a copper thermocouple, and
measured with an Omega CN76000 microprocessor-based temperature/processor controller. The NMR samples
were allowed to equilibrate for 10 min at each temp. before data acquisition.
1H- and 13C-NMR spectra of all ligands were recorded on a JEOL Eclipse-270 spectrometer. Variable-
temperature (VT) 1H-, 13C-, and 17O-NMR spectra were recorded on a Varian INOVA-500 spectrometer at 500,
125.6, and 67.8 MHz, respectively. Again, the probes were allowed to equilibrate for at least 10 min before data
acquisition. An NMR band-shape analysis of the Ln3‡-bound and bulk water resonances was performed by
means of a program written in Fortran 77 for a two-site exchange system with unequal populations. The molar
ratios of bound vs. bulk water were determined directly from the low-temperature 1H-NMR integration values.
The VT-17O-NMR linewidth data were fitted by following procedures published by Merbach and co-workers
[12].
REFERENCES
Â
[1] ꢁThe Chemistry of Contrast Agents in Medical Magnetic Resonance Imagingꢂ, Eds. A. E. Merbach, E.
Â
Toth, John Wiley & Sons, Chichester, 2001, p. 471.
[2] S. Aime, M. Fasano, E. Terreno, Chem. Soc. Rev. 1998, 27, 19.
[3] E. T. Ahrens, U. Rothbächer, R. E. Jacobs, S. E. Fraser, Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8443.
[4] K. Micskei, L. Helm, E. Brücher, A. E. Merbach, Inorg. Chem. 1993, 32, 3844.
Â
Á
[5] D. H. Powell, G. Gonzalez, V. Tissieres, K. Micskei, E. Brücher, L. Helm, A. E. Merbach, J. Alloys Compd.
1994, 207 ± 208, 20.
[6] C. Cossy, L. Helm, A. E. Merbach, Inorg. Chem. 1988, 27, 1973.
[7] C. Cossy, L. Helm, A. E. Merbach, Inorg. Chem. 1989, 28, 2699.
[8] K. Micskei, D. H. Powell, L. Helm, E. Brücher, A. E. Merbach, Magn. Reson. Chem. 1993, 31, 1011.
[9] D. H. Powell, A. E. Merbach, Magn. Reson. Chem. 1994. 32, 703.
[10] P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem. Rev. 1999, 99, 2293.
[11] V. Comblin, D. Gilsoul, M. Hermann, V. Humblet, V. Jacques, M. Mesbahi, C. Sauvage, J. F. Desreux,
Coord. Chem. Rev. 1999, 185 ± 186, 451.
[12] D. Pubanz, G. Gonzalez, D. H. Powell, A. E. Merbach, Inorg. Chem. 1995, 34, 4447.

Helvetica Chimica Acta ± Vol. 88 (2005)
935
[13] S. Aime, M. Botta, S. Geninatti Crich, G. Giovenzana, G. Jommi, R. Pagliarin, M. Sisti, Inorg. Chem. 1997,
36, 2992.
[14] S. Aime, M. Botta, S. Geninatti Crich, G. Giovenzana, R. Pagliarin, M. Sisti, E. Terreno, Magn. Reson.
Chem. 1998, 36, S200.
[15] S. Aime, M. Botta, L. Frullano, S. Geninatti Crich, G. Giovenzana, R. Pagliarin, G. Palmisano, F. Riccardi
Sirtori, Massimo Sisti, J. Med. Chem. 2000, 43, 4017.
[16] S. Aime, E. Gianolio, D. Corpillo, C. Cavallotti, G. Palmisano, M. Sisti, G. B. Giovenzana, R. Pagliarin,
Helv. Chim. Acta 2003, 86, 615.
[17] I. Lukes, P. Hermann, PCT Int. Appl. 2004, 70.
Â
Â
[18] R. Ruloff, E. Toth, R. Scopelliti, R. Tripier, H. Handel, A. E. Merbach, Chem. Commun. 2002, 2630.
Â
Â
[19] L. Burai, E. Toth, A. E. Merbach, Chem. Commun. 2003, 2680.
Â
Â
[20] S. Laus, R. Ruloff, E. Toth, A. E. Merbach, Chem.±Eur. J. 2003, 9, 3555.
[21] M. K. Thompson, M. Botta, G. Nicolle, L. Helm, S. Aime, A. E. Merbach, K. N. Raymond, J. Am. Chem.
Soc. 2003, 125, 14274.
[22] M. Woods, Z. Kovacs, S. Zhang, A. D. Sherry, Angew. Chem., Int. Ed. 2003, 42, 5889.
[23] S. Aime, A. Barge, M. Botta, A. S. De Sousa, D. Parker, Angew. Chem., Int. Ed. 1998, 37, 2673.
[24] S. Aime, A. Barge, J. I. Bruce, M. Botta, J. A. K. Howard, J. M. Moloney, D. Parker, A. S. de Sousa, M.
Woods, J. Am. Chem. Soc. 1999, 121, 5762.
[25] A. S. Batsanov, A. Beeby, J. I. Bruce, J. A. K. Howard, A. M. Kenwright, D. Parker, Chem. Commun. 1999,
1011.
[26] F. A. Dunand, S. Aime, A. E. Merbach, J. Am. Chem. Soc. 2000, 122, 1506.
[27] S. Zhang, K. Wu, M. C. Biewer, A. D. Sherry, Inorg. Chem. 2001, 40, 4284.
[28] S. Zhang, P. Winter, K. Wu, A. D. Sherry, J. Am. Chem. Soc. 2001, 123, 1517.
[29] S. Aime, A. Barge, D. Delli Castelli, F. Fedeli, A. Mortillaro, F. U. Nielsen, E. Terreno, Magn. Reson. Med.
2002, 47, 639.
[30] S. Aime, D. Delli Castelli, E. Terreno, Angew. Chem., Int. Ed. 2002, 41, 4334.
[31] S. Zhang, L. Michaudet, S. Burgess, A. D. Sherry, Angew. Chem., Int. Ed. 2002, 41, 1919.
[32] E. Terreno, D. Delli Castelli, G. Cravotto, L. Milone, S. Aime, Invest. Radiol. 2004, 39, 235.
[33] S. Aime, D. Delli Castelli, F. Fedeli, E. Terreno, J. Am. Chem. Soc. 2002, 124, 9364.
[34] S. Zhang, R. Trokowski, A. D. Sherry, J. Am. Chem. Soc. 2003, 125, 15288.
[35] S. Zhang, K. Wu, A. D. Sherry, Angew. Chem., Int. Ed. 1999, 38, 3192.
[36] S. Zhang, K. Wu, A. D. Sherry, Invest. Radiol. 2001, 36, 82.
[37] N. Raghunand, C. Howison, A. D. Sherry, S. Zhang, R. J. Gillies, Magn. Reson. Med. 2003, 49, 249.
[38] S. Zhang, K. Wu, A. D. Sherry, J. Am. Chem. Soc. 2002, 124, 4226.
[39] S. Aime, A. Barge, A. S. Batsanov, M. Botta, D. Delli Castelli, F. Fedeli, A. Mortillaro, D. Parker, H.
Puschmann, Chem. Commun. 2002, 1120.
[40] Z. Kovacs, A. D. Sherry, Synthesis 1997, 759.
[41] X. Zhou, P. Caravan, R. B. Clarkson, P.-O. Westlund, J. Magn. Reson. 2004, 167, 147.
[42] D. E. Woessner, S. Zhang, M. E. Merritt, A. D. Sherry, Magn. Reson. Med. 2005, 53, 790.
[43] S. Forsen, A. Hoffman, J. Chem. Phys. 1964, 40, 1189.
[44] S. Zhang, M. Merritt, D. E. Woessner, R. E. Lenkinski, A. D. Sherry, Acc. Chem. Res. 2003, 36, 783.
[45] S. Aime, M. Botta, M. Fasano, M. P. M. Marques, C. F. G. C. Geraldes, D. Pubanz, A. E. Merbach, Inorg.
Chem. 1997, 36, 2059.
[46] S. Zhang, Z. Kovacs, S. Burgess, S. Aime, E. Terreno, A. D. Sherry, Chem.±Eur. J. 2001, 7, 288.
[47] S. Zhang, A. D. Sherry, ꢁParamagnetic CEST Agents: NH versus OHꢂ, in ꢁProceedings of the International
Society for Magnetic Resonance in Medicineꢂ, 10th Scientific Meeting & Exhibition, 2002, Hawaii.
[48] S. Aime, M. Botta, M. Fasano, E. Terreno, Acc. Chem. Res. 1999, 32, 941.
[49] S. Aime, A. Barge, M. Botta, D. Parker, A. S. de Sousa, J. Am. Chem. Soc. 1997, 119, 4767.
[50] A. D. Sherry, S. Zhang, K. Wu, US Pat. 6,746,662, 2004.
[51] S. Aime, M. Botta, G. Ermondi, Inorg. Chem. 1992, 31, 4291.
[52] D. H. Powell, O. M. N. Dhubhghaill, D. Pubanz, L. Helm, Y. S. Lebedev, W. Schlaepfer, A. E. Merbach, J.
Am. Chem. Soc. 1996, 118, 9333.
[53] M. Woods, Z. Kovacs, R. Kiraly, E. Brücher, S. Zhang, A. D. Sherry, Inorg. Chem. 2004, 43, 2845.
Received April 14, 2005