The effect of a phosphodiester linking group on albumin binding, blood half-life, and relaxivity of intravascular diethylenetriaminepentaacetato aquo gadolinium(III) MRI contrast agents

Article abstract of DOI:10.1021/jm0102351

Amphiphilic gadolinium complexes were investigated as potential magnetic resonance imaging (MRI) contrast agents. A series of complexes was synthesized in order to study the effect of hydrophilic phosphodiester groups on albumin binding, relaxivity, and b

Full text of DOI:10.1021/jm0102351

J . Med. Chem. 2002, 45, 3465-3474
3465
Th e Effect of a P h osp h od iester Lin k in g Gr ou p on Albu m in Bin d in g, Blood
Ha lf-Life, a n d Rela xivity of In tr a va scu la r Dieth ylen etr ia m in ep en ta a ceta to
Aqu o Ga d olin iu m (III) MRI Con tr a st Agen ts
Thomas J . McMurry,* David J . Parmelee, Hironao Sajiki, Daniel M. Scott, Hillori S. Ouellet,
Richard C. Walovitch, Zolta´n Tyekla´r, Ste´phane Dumas, Paul Bernard, Samuel Nadler, Katarina Midelfort,
Matthew Greenfield, J effrey Troughton, and Randall B. Lauffer
EPIX Medical, Inc., 71 Rogers Street, Cambridge, Massachusetts 02142-1118
Received May 30, 2001
Amphiphilic gadolinium complexes were investigated as potential magnetic resonance imaging
(MRI) contrast agents. A series of complexes was synthesized in order to study the effect of
hydrophilic phosphodiester groups on albumin binding, relaxivity, and blood half-life in rats.
Thus, compound 11a , a diethylenetriaminepentaacetato aquo gadolinium(III) (Gd-DTPA)
derivative with an octyl substituent, was synthesized and compared to 5b, the analogous octyl
derivative containing a phosphodiester linkage between the gadolinium chelate and the alkyl
chain. Likewise, 11b, a naphthyl Gd-DTPA derivative, was compared to the naphthyl
phosphodiester derivative 5c. A direct comparison is not available for 5a , a 4,4-diphenylcy-
clohexyl phosphodiester Gd-DTPA derivative; however, its pharmacokinetic properties mirror
those of the other phosphodiester derivatives. Although the introduction of the phosphodiester
moiety decreased log P by approximately 1.7 units, albumin binding data obtained in 4.5%
human serum albumin (HSA) indicated that derivatives containing the phosphodiester group
exhibited somewhat higher albumin affinity than their alkyl analogues (54 ( 5 and 44 ( 4%
for 5b and 11a , respectively; 40 ( 4 and 30 ( 3% for 5c and 11b, respectively). Both classes
of agents were characterized by enhanced relaxivity in the presence of 4.5% HSA (r1 ) 16-42
mM-1 s-1 at 20 MHz and 37 °C) as compared with the relaxivity values measured in phosphate-
buffered saline (PBS) alone (r1 ) 4.6-6.6 mM-1 s-1 at 20 MHz and 37 °C). Pharmacokinetic
data indicated that compound 5b had a half-life of 14.3 ( 1.8 min in the rat as compared with
a half-life of 6.20 ( 0.04 min for the non-phosphodiester analogue 11a . Similarly, the half-life
obtained for the phosphodiester 5c was 14.3 ( 1.7 min as compared with a half-life of 6.80 (
0.03 min for 11b. The percent biliary excretion was significantly lower for the phosphodiester
compounds than for non-phosphodiester analogues (17.7 ( 4.0 and 66.9 ( 3.4% for 5b and
11a , respectively; 17.0 ( 1.6 and 64.3 ( 9.0% for 5c and 11b, respectively). The percent biliary
excretion (15.8 ( 4.4%) and plasma half-life in the rat (23.1 ( 2.9 min) for 5a are consistent
with the extended plasma half-life of the other phosphodiester derivatives. Taken together,
the enhanced relaxivity and extended blood half-life of the phosphodiester derivatives support
the concept of using endogenous albumin binding to achieve blood pool-like properties for small-
molecule magnetic resonance imaging (MRI) contrast agents.
In tr od u ction
are limited by relatively short half-lives and are also
freely extravasated into background muscle, which
decreases the contrast-to-noise ratio during steady state
imaging. For routine clinical use in MRA applications,
it is desirable that the contrast agent be selectively
retained in the vascular space and have an extended
blood half-life in order to provide a sufficient plasma
concentration over a 1 h imaging window. The extended
blood half-life facilitates imaging of multiple body
regions and acquiring long, pulse-gated scans of coro-
nary arteries.
Magnetic resonance angiography (MRA) is an imaging
technique in development to permit assessment of
cardiovascular disease, including the evaluation of
stenoses and occlusions in peripheral and coronary
arteries.1-3 Selective intravascular water proton relax-
ation rate enhancement using paramagnetic contrast
agents is thought to be a critical factor in enabling
routine T1-weighted MRA examinations, partly because
of imaging complications resulting from poor flow in
blocked arteries. Gadolinium-based contrast agents are
extremely effective at causing proton relaxation rate
enhancement, and the resultant short blood T1 values
following intravenous injection permit rapid pulsing and
enable very high resolution images. Current extracel-
lular magnetic resonance imaging (MRI) contrast agents
The design of MRI “blood pool” contrast agents that
provide sufficient proton relaxation rate enhancement
selectively within the vascular space, but that are also
efficiently excreted, has been a difficult challenge. For
example, excellent blood tracer properties may be
achieved using high-molecular weight polymeric conju-
gates of gadolinium chelates (such as Gd-DTPA-BSA4
or Gd-DTPA-polylysine5) or particulate iron oxides such
* Corresponding author. Tel: 617-250-6102. Fax: 617-250-6128.
E-mail: tmcmurry@epixmed.com.
10.1021/jm0102351 CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/27/2002

3466 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16
McMurry et al.
as NC100150;6,7 however, these materials generally
have poor excretion properties. One exception to this
trend is the development of a dendrimer-based macro-
molecular agent, Gadomer 17, which is a highly mono-
disperse 17 kDa compound with a relatively short
terminal half-life (10 min in the rabbit).8
Ch a r t 1. Chemical Structures of Gd-DTPA, 5a , 5b, 5c,
11a , and 11b
An alternative approach to creating intravascular
contrast agents exploits noncovalent binding of small-
molecule gadolinium chelates to endogenous serum
proteins, such as human serum albumin (HSA). In
principle, albumin binding can protect the contrast
agent from extravasation into surrounding tissue, thus
reducing background enhancement and improving con-
trast. An additional benefit results from the decreased
rate of molecular tumbling of the protein-bound agent
which enhances the efficacy of the electron-nuclear
interaction between the gadolinium(III) ion and water
protons, leading to very short T1 relaxation times and
bright blood vessels on an MRI image.9,10 In theory, this
approach combines the practical advantages of a small-
molecule pharmaceutical with the blood tracer proper-
ties of a high-molecular weight species.
A particular challenge in the design of albumin-
targeted blood pool agents is to select substituents that
permit both strong albumin binding and long blood half-
life. In the development of small-molecule blood pool
MRI contrast agents, these have been conflicting goals:
gadolinium chelates with hydrophobic substituents tend
to be rapidly excreted through the liver. For example,
the addition of the p-ethoxybenzyl or p-butylbenzyl
moieties to gadolinium diethylenetriaminepentaacetate
(Gd-DTPA) was shown to promote HSA binding but also
to decrease plasma elimination half-life in rats due to
an increased rate of hepatocellular uptake.11,12 Recently,
it has been shown that trisodium-{(2-(R)-[(4,4-diphen-
ylcyclohexyl)phosphonooxymethyl]diethylenetriamine-
pentaacetato)(aquo)gadolinium(III)} (5a ), a phosphodi-
ester derivative of Gd-DTPA substituted with a hydro-
phobic protein binding group, has an extended blood
half-life in humans, rabbits, and monkeys compared to
Gd-DTPA.13-16 While the signal intensity of blood drops
rapidly after injection of Gd-DTPA, little change was
noted over 1 h for 5a in Phase I clinical trials. Com-
pound 5a is currently in clinical trials to assess block-
ages in arteries.17
activity relationships (SAR) in rats which provided the
groundwork for the development of small-molecule blood
pool agents, as exemplified by 5a . In particular, the
effect of phosphodiester insertion between Gd-DTPA
and hydrophobic protein-binding groups on plasma
pharmacokinetics and excretion was examined. Thus,
11a , a Gd-DTPA derivative with an octyl substituent,
was compared to 5b. Likewise, 11b, a naphthyl deriva-
tive, was compared to 5c. Corresponding data for 5a are
also reported (Chart 1).
Resu lts
The alkyl-substituted DTPA gadolinium complexes
were prepared in good yields as racemic compounds by
modifications of literature methods.20,21 The synthesis
of the novel phosphodiester-linked chelates was achieved
starting from 2-(R)-hydroxymethyl-DTPA-penta-tert-
butyl ester (1).35 A modified, scaleable synthesis of this
particular intermediate has also been reported.36 Phos-
phoramidite chemistry was then used to couple 1 to
alcohols 4,4-diphenylcyclohexanol, 1-octanol, or 2-(1-
naphthyl)ethanol. Oxidation using tert-butylhydroper-
oxide gave the cyanoethyl-protected phosphorus(V) tri-
ester which was subsequently deprotected with ammonia,
methanol, and triethylamine. Complete deprotection of
the tert-butyl esters using either TFA or HCl gave the
penultimate polyaminocarboxylate ligand which was
subsequently reacted with gadolinium oxide and base
to prepare the corresponding chelate.
One key molecular feature that is responsible for the
longer blood half-life of 5a is the presence of the
hydrophilic phosphodiester group. This group appears
to decrease the rate of hepatobiliary uptake, thus
increasing blood half-life in this class of amphiphilic Gd-
DTPA derivatives. This paper focuses on key structure-
The insertion of a hydrophilic phosphodiester group
between the hydrophobic side chains and the Gd-DTPA

Effect of a Phosphodiester Linking Group
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16 3467
Ta ble 1. In Vitro Partition Coefficient (log P), Protein Binding, and Relaxivity Data of Contrast Agents
relaxivitya
observed relaxivitya
(mM-1 s-1
calculated relaxivity boundd
(mM-1 s-1
partition coefficient
percent bound
4.5% HSAb
(mM-1 s-1
PBSc
)
)
)
sample
log Pbutanol/PBS
4.5% HSA
4.5% HSA
Gd-DTPA
5a
5b
5c
11a
11b
-4.3 ( 0.3
-2.11 ( 0.06
-2.34 ( 0.06
-3.25 ( 0.15
-0.63 ( 0.04
-1.53 ( 0.03
0
4.6 ( 0.3
6.6 ( 0.4f
6.1 ( 0.3
6.0 ( 0.3
5.9 ( 0.3
5.5 ( 0.3
4.6 ( 0.3
not applicable
47 ( 4
88 ( 2e
54 ( 5
40 ( 4
44 ( 4
30 ( 3
42.0 ( 2.3f
not measured
19.5 ( 2.0
20.7 ( 2.1
16.3 ( 1.6
-
40 ( 8
40 ( 9
42 ( 10
a
b
d
r1, 0.47 T (20 MHz), 37 °C. In 0.1 mM drug, 4.5% HSA, fraction V. c Phosphate-buffered saline. Calculated using eq 1. e From
reference 22. f From reference 13.
Ta ble 2. Comparison of Rat Pharmacology Data for Phosphodiester and Non-Phosphodiester-Based Contrast Agents (dose ) 0.1
mmol/kg)
b
c
d
e
percent
Vdss
T1/2â
AUCconc
AUC1/T1
sample
biliary excretiona
(L/kg)
(min)
(mM/min)
(s-1 min)
Gd-DTPA
5a f
5b
5c
11a
6.3 ( 3.7g
15.8 ( 4.4
17.7 ( 4.0
17.0 ( 1.6
66.9 ( 3.4
64.3 ( 9.0
0.28 ( 0.04
0.24 ( 0.03
0.31 ( 0.02
0.32 ( 0.03
0.62 ( 0.05
0.44 ( 0.07
15.4 ( 1.3
23.1 ( 2.9
14.3 ( 1.8
14.3 ( 1.7
6.20 ( 0.04
6.80 ( 0.03
6.21 ( 0.09
9.35 ( 1.30
5.96 ( 0.81
5.57 ( 0.38
3.06 ( 0.13
2.58 ( 0.49
46.4 ( 4.2
285 ( 27
150 ( 17
94.2 ( 5.0
69.7 ( 3.1
40.1 ( 0.9
11b
a
b
d
In 24 h. Volume of distribution, steady state. c Elimination half-life. Area under concentration versus time curve, 0-30 min. e Area
under plasma 1/T1 versus time curve, 0-30 min. f From reference 13. Values are mean ( 1 standard deviation.
g
chelate decreased the overall hydrophobicity of the
molecules by approximately 1.7 log units as shown by
butanol buffer partition coefficients (Table 1). Despite
the lower hydrophobicity, the phosphodiester deriva-
tives bound somewhat more strongly to HSA than their
non-phosphodiester counterparts (54 ( 5 and 44 ( 4%
for 5b and 11a , respectively; 40 ( 4 and 30 ( 3% for 5c
and 11b , respectively). Under similar conditions (0.1
mM drug, 4.5% HSA, ICP detection), 5a was found to
be 88 ( 2% protein bound.22
In rats, the largest differences between the phos-
phodiester compounds and their counterparts were in
the percent of the injected dose excreted by the biliary
route. The phosphodiester derivatives had less than 20%
biliary excretion compared to greater than 60% bilary
excretion for the non-phosphodiester compounds. The
reduced hepatocellular uptake for the phosphodiester
agents resulted in longer elimination half-lives (14.3 (
1.8 and 6.2 ( 0.04 min for 5b and 11a , respectively;
14.3 ( 1.7 and 6.8 ( 0.03 min for 5c and 11b,
respectively) and greater area under the curve concen-
trations (AUCconc) from 0 to 30 min (5.96 ( 0.81 and
3.06 ( 0.13 min for 5b and 11a , respectively; 5.57 (
0.38 and 2.58 ( 4.9 min for 5c and 11b, respectively;
Table 2 and Figure 1). The low percent biliary excretion
observed for 5a (15.8 ( 4.4%) in the rat combined with
a higher albumin binding resulted in the longest blood
half-life of the group of compounds studied (23.1 ( 2.9
min).
F igu r e 1. Comparison of the 5a , 5b, or 11a plasma concen-
tration following an intravenous injection in the rat (0.1 mmol/
kg). Plots show plasma concentration as a function of time
(min). The plasma concentration of 5b is generally greater
than that of 11a , demonstrating the plasma half-life extension
ability of the phosphodiester linker. The increased half-life of
5a is reflected in the calculated area under the curve data
(AUCconc in mM/min) shown in Table 2.
Relaxivity data in PBS, 4.5% HSA, and calculated
relaxivity-bound values are also shown in Table 1. The
relaxivity of the protein-bound agent, r1b, is calculated
according to eq 1 where r1obs is the observed longitudinal
relaxivity, fb is the fraction of bound drug, r1u is the
relaxivity of the drug in the absence of protein, and fu
is the fraction of unbound drug (measured by ultrafil-
tration).
of the gadolinium complexes to albumin was accompa-
nied by an increase in observed relaxivity, the ability
of the paramagnetic complexes to alter proton relaxation
rate enhancement. Interestingly, when normalized for
albumin binding by calculating r1b, the relaxivity of the
bound complexes was similar.
As a measure of the potential MRI blood signal, the
1/T1 relaxation rate of plasma samples was also mea-
sured (Figure 2). The AUC-1/T1 values for the phos-
phodiester derivatives were roughly twice that of the
closely related non-phosphodiester analogues.
r1obs ) r1bfb + r1ufu
(1)
The data in Table 1 indicate that noncovalent binding

3468 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16
McMurry et al.
buffer) partition coefficients ranged from -0.63 to -3.25;
butanol was selected as a substitute for the typical
octanol because of the experimental difficulty in ac-
curately determining low concentrations of chelates in
the more traditional octanol/water system. Consistent
with expectations, the addition of a phosphodiester
moiety to the chelate backbone further increased the
hydophilicity of the gadolinium chelates and reduced log
P by about 1.7 units. In the two cases examined (Table
1, 11a vs 5b and 11b vs 5c), the increased hydrophilicity
of the phosphodiester was accompanied by a slight
increase in equilibrium albumin binding.
The maintenance of good, even moderately improved,
HSA binding upon the addition of the very hydrophilic
phosphodiester group to an already extremely hydro-
philic Gd-DTPA dianion (2-) is striking. These results
are contrary to the common notion that albumin binding
is simply lipophilicity driven and suggest that, in this
case, the interaction with the protein more than com-
pensates for the strong hydration of the anionic func-
tional group. The proposed structure of HSA binding
sites afforded by X-ray analysis suggests sock-shaped
hydrophobic sites lined by positively charged residues
near the “ankle”25 which could provide a site for
electrostatic interaction with the phosphodiester. Ad-
ditional SAR studies to further elucidate the nature of
the binding interaction are in progress.
F igu r e 2. Comparison of relaxation rate enhancement (∆1/
T1) measured in rat plasma following an intravenous injection
of 5a , 5b, or 11a (0.1 mmol/kg). Plots show plasma ∆1/T1 as a
function of time (min). The plasma relaxation rate enhance-
ment is generally greater for 5b than 11a primarily due to
the higher concentration of the phosphodiester agent at any
given time point. The increased efficacy of 5a relative to 5b
or 11a or Gd-DTPA is reflected in the larger area under the
Blood Ha lf-Life. The addition of lipophilic albumin
binding moieties to Gd-DTPA chelates generally results
in a blood half-life shorter than that of the parent,
unsubstituted DTPA chelate because of hepatobiliary
extraction. This characteristic was exploited in the
design of the liver imaging agent gadolinium ethoxy
benzyl (EOB) DTPA.26 Therefore, a significant challenge
in the design of effective blood pool imaging agents was
the combination of significant albumin binding with
extended blood half-life.
curve for the plasma relaxation rate enhancement (AUC1/T
,
1
s-1 min).
Discu ssion
Syn th esis. Alkyl- and aryl-substituted chelates de-
rived from amino acid starting materials have found
wide application in both MRI and nuclear imaging
applications.10 The synthetic route chosen for the syn-
thesis of the racemic polyaminocarboxylate ligands 11a
and 11b (Scheme 1) is analogous to those previ-
ously reported for the synthesis of p-NCS-Bz-DTPA and
EOB-DTPA.20,23 The synthesis of chiral key intermedi-
ate 2-(R)-hydroxymethyl-DTPA-penta-tert-butyl ester
(1), which contains the basic functionality necessary for
metal chelation, has been reported previously in the
literature.35,36 The penta-tert-butyl esters were purified
by flash chromatography and the polyaminocarboxylates
by C18 Sep-Pak reversed phase chromatography. Clean
transformation to the gadolinium complexes was ac-
complished using 0.5 equiv of Gd2O3 and stoichiometric
quantities of sodium hydroxide or N-methyl-D-glucam-
ine (NMG) base.
Phosphoramidite chemistry, the method of choice for
automated polynucleotide synthesis,24 was used to
prepare the novel unsymmetrical phosphodiesters 5a ,
5b, and 5c. In this work, 2-(R)-hydroxymethyl-DTPA-
penta-tert-butyl ester (1, Scheme 1) was converted to a
common phosphoramidite intermediate which could
then be coupled to a variety of albumin binding moieties
prior to the formation of the desired gadolinium com-
plexes. The phosphodiester linking group was found to
be remarkably robust under synthetic conditions used
for chelation (95 °C, pH 2-4).
While some rational design criteria might be useful
in designing HSA binders, the success of the phosphodi-
ester in reducing biliary excretion was unexpected.
Aside from the requirement of some hydrophobic resi-
dues, structure-activity relationships for hepatocellular
uptake are less understood.27 Perhaps the added hy-
drophilicity was sufficient to tip the balance away from
rapid hepatocellular uptake while, at the same time, the
phosphodiester placement maintained good HSA bind-
ing. Regardless of the mechanism, reduction of hepa-
tobiliary extraction was critical to the improvement in
blood half-life for the phosphodiester derivatives. With
hepatobiliary elimination minimized, the relatively slow
glomerular filtration of the unbound drug dominated the
excretion pathway, thus extending the blood half-life.
Other functional groups, such as carboxylate or sul-
fonate, may function in the same manner.28-30
Rela xivity. The currency of MRI contrast agent
efficacy is proton relaxation enhancement, as it is this
effect that translates to a brighter image during the
MRI examination. The data in Table 1 indicate that the
observed reversible protein binding was clearly coupled
to an increase in relaxivity by the paramagnetic com-
plexes. Relaxivity is the paramagnetic relaxation rate
enhancement normalized to 1 mM and is dependent on
both field strength and temperature.9,10 A larger relax-
ivity indicates a greater enhancement effect and, pro-
Albu m in Bin d in g. The gadolinium DTPA chelates
described herein are negatively charged (2- or 3-),
extremely hydrophilic molecules. Butanol/water (PBS

Effect of a Phosphodiester Linking Group
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16 3469
Sch em e 1. Synthetic Scheme for the Phosphodiester Gadolinium Complexes 5a , 5b, and 5c
Sch em e 2. Synthetic Scheme for the Non-Phosphodiester DTPA Gadolinium Complexes 11a and 11b
vided other pharmacokinetic factors are equivalent, a
more efficacious MRI contrast agent. Relaxivity is a
complex biophysical phenomenon dependent on a num-
ber of key molecular properties, including the rotational
diffusion rate of the paramagnetic chelate unit, the
exchange rate of the coordinated water, and the elec-
tronic properties of the gadolinium ion. For the type of
Gd-DTPA chelates discussed here, theory predicts that
relaxivity should increase due to slowing of the overall
rotational diffusion rate of the paramagnetic chelate.
The observed increase in relaxivity at 20 MHz (a
clinically relevant field strength) upon binding to HSA
was approximately 3 times higher for the moderately
bound compounds 5b , 11a , 5c, and 11b and about 7
times higher for 5a , the most efficacious agent of the
group.
The relaxivity of the bound contrast agent was
calculated from the measured percent bound at 0.1 mM
drug in 4.5% HSA, the total observed relaxivity in 4.5%
HSA, and the relaxivity of the free (non-protein-bound)
agent in PBS alone. For these calculations, a single
binding site was assumed. The bound relaxivity values
reported here (Table 1) were significantly larger than
values obtained for covalently labeled Gd-DTPA poly-

3470 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16
McMurry et al.
mers (r1 ) 6-18 mM-1 s-1 at 20 MHz)5,31-33 and
superior to that reported for Gd-DTPA conjugated
dendrimers (r1 ) 18-36 mM-1 s-1 at 25 MHz).34
Interestingly, the calculated bound relaxivity of the
various gadolinium complexes was quite similar (40-
47 mM-1 s-1), indicating that the complexes experience
similar local rotational properties and may occupy
similar sites on HSA. Studies of the effect of site
occupancy on relaxivity for this class of compounds are
in progress.
P h osp h or a m id ite In ter m ed ia te (2). To a stirred solution
of the penta-tert-butyl ester 1 (107.45 g, 0.15 mmol) and
diisopropylethylamine (53 mL, 304 mmol) in distilled CH2Cl2
(600 mL) was added 2-cyanoethyl-N,N-diisopropylchlorophos-
phoramidite (40 mL, 179 mmol) at RT under N2. The mixture
was stirred at RT for 2 h. The reaction mixture was washed
with ice cold 10% NaHCO3 solution (700 mL), H2O (500 mL),
and brine (500 mL), and then dried over MgSO4. The organic
layer was evaporated to afford the crude product as a pale
yellow oil (150 g). This crude oil was used for coupling reactions
without further purification. Rf (ether/hexanes ) 2:1) ) 0.41.
31P NMR (CD3CN): δP 138.1, 138.2. A small portion of 2 was
purified by flash chromatography on silica gel (EtOAc/hexanes/
Con clu sion s
1
Et3N 90:10:1 ratio) for analysis. H NMR (CDCl3): δ 1.16 (d,
The human serum albumin binding, blood half-life,
and relaxivity of amphiphilic gadolinium complexes
containing a phosphodiester moiety were compared with
those of complexes which are lacking the negatively
charged, hydrophilic linker. A modest increase in albu-
min binding was observed for the compounds containing
the phosphodiester linker. Moreover, a dramatic shift
from predominantly biliary excretion to mostly renal
excretion was observed following the addition of the
phosphodiester moiety. Concomitantly, an extended
blood half-life was observed for the phosphodiester
compounds, presumably because glomerular filtration
is a relatively slow process compared to biliary excretion.
Because of the extended blood half-life and pronounced
relaxivity enhancement observed in albumin solution,
the gadolinium chelate phosphodiester compound 5a is
now being investigated in blood vessel enhancement
studies.
J ) 16.4 Hz, 2H), 1.42 and 1.43 (2s, 45H), 2.6-2.73 (m, 3H),
2.75-2.87 (m, 5H), 2.93-3.03 (m, 1H), 3.33-3.48 (m, 6H),
3.48-3.7 (m, 7H), 3.7-3.92 (m, 3H). 13C NMR (CDCl3): δ 20.2,
20.3, 24.5, 24.6, 28.01, 28.03, 28.1, 42.9 (δ, J ) 10.1 Hz), 43.0,
52.5, 52.7, 52.8, 53.7, 53.8, 53.9, 54.2, 55.8, 56, 56.1, 58.2 (d, J
) 8.6 Hz), 58.5 (d, J ) 8.1 Hz), 60.9 (2d ≈ tapp, J ) 8.1 Hz),
63.1 (d, J ) 15.1 Hz), 63.6 (d, J ) 15.6 Hz), 80.2, 80.3, 80.6,
117.6, 117.7, 170.5, 171, 171.1, 171.2, 171.23. Exact mass calcd
for C44H83N5O12P: 904.577587. Found: 904.574520. Anal.
(C44H82N5O12P) C, H, N, P.
Cya n oeth yl-(4,4-d ip h en ylcycloh exyl)-(R)-p h osp h on o-
oxym eth yl-p en ta -ter t-bu tyl-DTP A (3a ). To a crude phos-
phoramidite intermediate 2 (prepared from 107.47 g (152.7
mmol) of 1-(R)-hydroxymethyl-DTPA-penta-tert-butyl ester)
and 4,4-diphenylcyclohexan-1-ol (39.51 g, 0.157 mol) in dis-
tilled CH3CN (900 mL) was added 1H-tetrazole (10.96 g, 0.156
mol). The reaction mixture was stirred overnight at RT, when
31P NMR (CH3CN) spectroscopy showed complete conversion
of 2 (δp ) 147.3 and 147.4) into the expected phosphite (δp )
134.2 and 134.0). To the latter mixture was added tert-
butylhydroperoxide (Aldrich, 90% solution, 17.62 mL, ca. 0.158
mol), and the mixture was stirred for 1.5 h. The solution was
tested with peroxide paper (EM separation) in order to indicate
the absence of residual peroxide, and then concentrated in
vacuo. The gummy residue was partitioned between diethyl
ether (1000 mL) and H2O (750 mL). The organic layer was
washed with H2O (500 mL), ice cold 10% NaHCO3 solution
(500 mL), and brine (2 times, 500 mL each), and then dried
over MgSO4 and evaporated. The residue was used for the next
reaction without further purification. Rf (CHCl3/MeOH ) 10:
1) ) 0.47. 31P NMR (CH3CN): δP -8.1.
Exp er im en ta l Section
Unless otherwise noted, all materials were obtained from
commercial suppliers and used without further purification.
THF was distilled from potassium benzophenone ketyl im-
mediately prior to use. Methylene chloride was distilled over
calcium hydride. All column chromatography was carried out
by flash methods with silica gel (230-400 mesh, EM separa-
tion). 2-(R)-Hydroxymethyl-DTPA-penta-tert-butyl ester (1)
was obtained from 2-(R)-hydroxymethyldiethylenetriamine
trihydrochloride as described in ref 35. Reactions were moni-
tored by thin-layer chromatography (TLC) performed on
aluminum-backed silica gel 60 F254, 0.2 mm plates (EM
separation), and compounds were visualized under UV light
(254 nm), Ninhydrin-Plus reagent, or Dragendorff’s reagent
(both Alltech) with subsequent heating. Routine proton NMR
spectra were recorded at 300 MHz in CDCl3 with TMS as the
internal standard except for the spectra recorded in D2O (CD3-
Cya n oeth yl-(n -octyl)-(R)-p h osp h on ooxym eth yl-p en ta -
ter t-bu tyl-DTP A (3b). Compound 3b was prepared from
n-octanol (16.9 mmol) and crude phosphoramidite intermediate
2 (obtained from 1-hydroxymethyl-DTPA-penta-tert-butyl ester
1, 4.40 g, 6.40 mmol) by the same procedure described for 3a .
The product was purified by silica gel column chromatography
(CHCl3/MeOH) to give 2.71 g (44.7% total yield) from 2 as a
1:1 mixture of diastereoisomers. Rf (CHCl3/MeOH ) 10:1) )
CN ref). Coupling constants (J ) are reported in hertz (Hz). 31
P
1
0.33. 31P NMR (CD3CN): δP -7.79. H NMR (CDCl3): δ 0.8-
NMR spectra were obtained at 121.4 MHz and referenced to
H3PO4. Fast atom bombardment-mass spectrometry (FAB-MS)
samples were run at the University of California, Berkeley,
CA, and electrospray samples were obtained on a Hewlett-
Packard 1100 MSD instrument operating in the negative ion
mode, unless otherwise noted. Elemental analyses were per-
formed by Galbraith Laboratories, Inc., Knoxville, TN. Solu-
tions of the N-methyl-D-glucamine salt of Gd-DTPA (molecular
weight ) 938) were either prepared from Gd2O3, H5DTPA, and
N-methyl-D-glucamine (NMG) or obtained commercially (Mag-
nevist, Berlex, Wayne, NJ ). Analytical HPLC was performed
using either a Thermohypersil C4 column (150 × 4.6 mm, 5
µm HyPURITY) or a YMC-Pack TMS column (250 × 4.6 mm,
3 µm) at a flow rate of 1.5 mL/min and a gradient of 5-20% B
over 5 min, and then 20-100% B over 18 min. HPLC solvent
A comprised 970 mL of phosphate buffer (0.146 M phosphate,
0.175 mM EDTA, pH 6.85) and 30 mL of ACN, while HPLC
solvent B comprised 400 mL of phosphate buffer (0.146 M
phosphate, 0.175 mM EDTA, pH 6.85) and 600 mL of ACN.
Detection was performed at 220 nm (compounds 5c, 11b) or
197 nm (compound 5b).
0.9 (m, 3H), 1.18-1.38 (m, 10H), 1.41 and 1.42 (2s, 45H), 1.6-
1.64 (m, 2H), 2.53-2.68 (m, 1H), 2.72-2.82 (m, 6H), 2.82-
2.93 (m, 1H), 3.02-3.13 (m, 1H), 3.27-3.56 (m, 10H), 4.0-
4.11 (m, 2H), 4.14-4.33 (m, 4H). 13C NMR (CDCl3): δ 14.3,
19.7, 19.8, 22.8, 25.5, 28.27, 28.32, 29.27, 29.30, 30.4 (d, J )
7.1 Hz), 31.9, 52.5, 52.9, 53, 53.5, 53.6, 53.7, 53.9, 56, 56.3,
60.5 (2d ≈ tapp, J ) 8.6 Hz), 61.9, 68.2 (2d ≈ tapp, J ) 7.6 Hz),
68.5 (2d ≈ tapp, J ) 4 Hz), 80.40, 80.44, 80.5, 80.6, 116.2, 116.3,
169.8, 170.10, 170.14, 170.34, 170.4. Anal. (C46H85N4O14P) C,
H, N, P.
Cya n oeth yl-(2-(1-n a p h th yl)eth yl)-(R)-p h osp h on ooxy-
m eth yl-p en ta -ter t-bu tyl-DTP A (3c). Compound 3c was
prepared from 2-(1-naphthyl)ethanol (1.6 g, 9.5 mmol), crude
phosphoramidite intermediate 2 (3.5 g, 3.8 mmol), and tetra-
zole (0.53 g, 7.6 mmol) in CH3CN (40 mL) as described above
for compound 3a . After the mixture was oxidized with tert-
butylhydroperoxide (1.2 mL) for 1 h, the solution was concen-
trated and partitioned between EtOAc (30 mL) and water (10
mL). The organic layer was extracted with brine, dried
(MgSO4), and evaporated to give 4.7 g of crude product 3b as

Effect of a Phosphodiester Linking Group
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16 3471
a 1:1 mixture of diastereoisomers which was immediately
converted to compound 4c. A portion was purified by silica
gel chromatography (CH2Cl2/MeOH) for characterization pur-
poses. Rf (CHCl3/MeOH ) 10:1) ) 0.22. 31P NMR (CD3CN):
δP -4.12. 1H NMR (CDCl3): δ 1.41, 1.416, and 1.418 (3s, 45H),
2.52-2.65 (m, 3H), 2.76 (br s, 4H), 2.82-2.93 (m, 1H), 2.82-
2.93 (m, 1H), 3.02-3.12 (m, 1H), 3.27-3.54 (m, 12H), 4.0-
4.24 (m, 4H), 4.36-4.47 (m, 2H), 7.41 (d, J ) 5.3 Hz, 2H),
procedure described for 4b in 2 M NH3-MeOH (35 mL) at RT
for 18 h. The phosphodiester intermediate was purified by
flash chromatography on silica gel as the triethylammonium
salt (1.82 g, 1.75 mmol, 54.7%). The tert-butyl esters were
deprotected (1.71 g, 1.64 mmol) in 12 N HCl (7 mL) and
dioxane (15 mL) for 22 h at RT to give the product 4c (0.71 g,
1.1 mmol, 65.5%) after desalting on a C18 reverse phase silica
gel column (from H2O to 1:4 CH3CN/H2O). 31P NMR (MeOD):
δ 2.3. MS: m/e 656.5 [(M - H)-]. 1H NMR (D2O): δ 2.86-3.56
(m, 19H), 3.95 (AB system, J ) 6.7 Hz, 1H), 4.0 (AB system,
J ) 6.7 Hz, 1H), 4.36 (s, 2H), 7.14-7.35 (m, 4H), 7.48-7.55
(m, 1H), 7.6-7.66 (m, 1H), 7.92 (d, J ) 8.2 Hz). 13C NMR
(MeOD): δ 35.0 (d, J ) 7.1 Hz), 50.5, 53.1, 53.8, 55.1, 55.5,
55.7, 58.6 (d, J ) 7.1 Hz), 63.6 (d, J ) 3.8 Hz), 67.1 (d, J ) 5
Hz), 124.9, 126.69, 126.74, 127.3, 128.3, 128.6, 129.8, 133.4,
135.3, 135.7, 169.3, 174.3, 175.4. Anal. (C27H36N3O14P‚1.1NaCl)
C, H, N, P. FAB-MS exact mass calcd for C27H37N3O14P+:
658.2013. Found: 658.2015.
7.46-7.58 (m, 2H), 7.72-7.79 (m, 1H), 7.83-7.88 (m, 1H). 13
NMR (CDCl3): δ 19.1, 19.2, 27.9, 28, 33.4, 33.5, 52.2, 52.57,
C
52.60, 53.2, 53.3, 53.5, 55.7, 56, 60.1, and 60.2 (2d ≈ tapp, J 1
)
7.6 Hz, J 2 ) 7.1 Hz), 61.5 and 61.6 (2d ≈ tapp, J ) 5 Hz), 67.5
and 67.6 (2d ≈ tapp, J ) 5.5 Hz), 67.9 and 68 (2d ≈ tapp, J )
7.1 Hz), 80.47, 80.51, 80.53, 80.6, 116.4, 116.5, 123.2, 125.3,
126.1, 127.2, 127.4, 128.7, 131.7, 132.66, 132.69, 170.4, 170.67,
170.71, 170.89, 170.94. Anal. (C50H79N4O14P‚H2O) C, H, N, P.
FAB-MS exact mass calcd for
Found: 991.5402.
C
50H79N4O14P+: 991.5409.
Tr isod iu m -{(2-(R)-[(4,4-d ip h en ylcycloh exyl)p h osp h o-
n ooxym et h yl]d iet h ylen et r ia m in ep en t a a cet a t o)(a q u o)-
ga d olin iu m (III)} (5a ). A test tube was charged with gado-
linium oxide (Gd2O3) (0.95 g, 2.6 mmol), compound 4a
(purity: 98 wt %, 3.76 g, 5 mmol), and sodium hydroxide (10
mL of a 1.0 N solution, 10 mmol). The mixture was stirred at
95 °C for 6 h. The reaction mixture was cooled to room
temperature, and the pH was adjusted to ∼7.5 by adding 4.5
mL of 1.0 N (4.5 mmol) NaOH solution in small portions. Half
of the solvent was removed by evaporation at 95 °C. The
remaining mixture was cooled to room temperature and loaded
on a 10 g Sep-Pak column. Elution with water followed by
freeze drying yielded 4.70 g (4.8 mmol, 96%) as a white powder.
Anal. (C33H40GdN3Na3O15P) C, Gd, H, N. FAB-MS, 981 [(M -
H2O + Na)+], 959 [(M - H2O + H)+], 937 [(M - H2O - Na +
2H)+], 915 [(M - H2O - 2Na + 3H)+]. N-Methyl-D-glucamine
(NMG) used as base instead of sodium hydroxide produces the
N-methyl-D-glucammonium (NMG) salt, 5a -NMG. Anal.
(C54H95N6O30PGd) C, H, N, Gd. +FAB-MS: 1478 [(M - H2O
+ H)+], 1283 [(M - H2O - NMG + 2H)+], 1088 [(M - H2O -
2NMG + 3H)+].
Tr is-[N-m et h yl-D-glu ca m in e]-{(2-(R)-[oct ylp h osp h o-
n ooxym et h yl]d iet h ylen et r ia m in ep en t a a cet a t o)(a q u o)-
ga d olin iu m (III)} (5b). A test tube was charged with gado-
linium oxide (0.18 g, 0.5 mmol), compound 4b (0.70 g, 1.05
mmol), N-methylglucamine (NMG) (0.38 g, 2 mmol), and
deionized water (3.5 mL). The mixture was stirred at 95 °C
for 6 h; the solution was cooled to room temperature, and the
pH was adjusted to 7. The total volume was adjusted to 5.0
mL with deionized water, and the solution was heated again
to 95 °C for 10 min. The cooled solution was filtered through
a 0.2 µm syringe filter to give an aqueous solution of the titled
compound (5.6 µg/mL Gd). MS: 767.2 [(M - H2O + H)-]. FAB-
MS exact mass calcd for C23H40GdN3O14P+: 771.1489. Found:
771.1514. HPLC purity: 94% (YMC-Pack TMS).
4,4-Dip h en ylcycloh exyl-p h osp h on ooxym et h yl-DTP A
(4a ). The solution of 3a in 2 M NH3-MeOH (1000 mL) was
stirred at RT overnight. The solvent was evaporated. The
residue was treated with Et3N (21.3 mL, 153 mmol) and
purified with silica gel column chromatography (divided into
two portions of ca. 80 g each) (diameter 5.5 cm × height 40
cm, from CH2Cl2 only to 95:5 CH2Cl2/Et3N, silica gel was
treated with 5% Et3N in CH2Cl2 prior to loading) to afford
purified intermediate (138.8 g, 89%, 3 steps total from 5). Rf
(CHCl3/MeOH ) 10:1) ) 0.12. 31P NMR (CDCl3): δp -7.37.
Exact mass calcd for C53H85N3O14P: 1018.576919. Found:
1018.577300. A portion of the tert-butyl ester intermediate
(135.2 g, 132 mmol) was dissolved in a mixture of concentrated
HCl (Fisher Scientific trace metal grade, 150 mL) and ether
(300 mL) and stirred at RT overnight. Ether was evaporated
off, and the aqueous solution was treated with 8 M NaOH
solution to adjust the pH to ∼1.5. The resulting white
precipitate was filtered and washed with H2O (5 times, 200
mL each) and ether (2 times, 150 mL each). The white solid
product was dried under a vacuum desiccator (P2O5) at RT for
24 h to give the purified product 4a (80.49 g, 77-82%). 31P
NMR (D2O + NaOD, pD 13.5): δp -0.3. 1H NMR (D2O): δ 1.49
(m, 2H), 1.73 (m, 2H), 1.98-2.28 (m, 4H), 2.42-2.57 (m, 6H),
2.69 (d, 1H, J ) 16.4 Hz), 2.78 (m, 1H), 2.92-3.07 (m, 8H),
3.12 (d, J ) 16.4 Hz, 1H), 3.68 (m, 2H), 4.08 (m, 1H), 6.97-
7.35 (m, 10H). 13C NMR (MeOD): δ 31, 34.1 (br s), 46.6, 50.5,
53.2, 54, 55.2, 55.5, 55.8, 58.8 (br s), 63.8 (br s), 75 (br s), 126.6,
127.9, 128.3, 129.2, 129.4, 147.9, 149.4, 169.3, 174.4, 175.5.
Anal. (C33H44N3O14P‚H2O) C, H, N, P.
n -Octyl-p h osp h on ooxym eth yl-DTP A (4b). A solution of
n-octyloxy phosphate cyanoethyl ester 3b (2.17 g, 2.3 mmol)
in 2 M NH3-MeOH (25 mL) was stirred at RT for 18 h. The
solvent was evaporated. The residue was purified by flash
chromatography on silica gel (from a 99:1:1 to a 96:4:1 CH2-
Cl2/MeOH/Et3N ratio) to afford the penta-tert-butyl ester-
protected n-octylphosphodiester product as a triethylammo-
nium salt (1.95 g, 2.0 mmol, 87%). The foaming solid (1.27 g,
1.3 mmol) was then dissolved in dioxane (5 mL), and 10 mL
of 12 N HCl was added. The solution was stirred at RT for 24
h. The dioxane was evaporated, and the pH was adjusted to
≈1.6 with 8 N NaOH. The solid was filtered and washed with
a HCl solution (pH ≈ 1.6) and then with ether. The white solid
was desalted on a C18 reverse phase silica gel column (Sep-
Pak pre-packed cartridge, Waters) (from H2O to 1:4 CH3CN/
H2O) to give the pure product 4b (0.44 g, 56.7%). 31P NMR
Tr is-[N-m eth yl-D-glu ca m in e]-{(2-(R)-[(2-(1-n a p h th yl)-
eth yl) p h osp h on ooxym eth yl]d ieth ylen etr ia m in ep en ta -
a cet a t o)(a qu o)ga d olin iu m (III)} (5c). Gadolinium oxide
(0.15 g, 0.4 mmol), compound 4c (0.73 g, 0.84 mmol), and
N-methylglucamine (0.31 g, 1.6 mmol) were reacted in deion-
ized water (3 mL) and isolated as described above for com-
pound 5b to give an aqueous solution of 5c (5.6 µg/mL Gd).
MS: 809.2 [(M - H2O + H)-]. FAB-MS exact mass calcd for
C
27H34GdN3O14P+: 813.1020. Found: 813.1027. HPLC pu-
rity: 99% (Thermohypersil C4).
1
(D2O): δP 2.3. MS: m/e 614.5 [(M - H)-]. H NMR (D2O): δ
2-Am in od eca n oic Acid Eth ylen ed ia m in e Am id e (7a ).
Racemic methyl 2-aminodecanoate hydrochloride 6a (6.94 g,
29.2 mmol) was prepared by the method of O’Donnell.18 6a .
1H NMR (CDCl3): δ 0.8-0.9 (m, 3H), 1.2-1.4 (m, 12H), 1.42-
1.6 (m, 3H), 1.62-1.76 (m, 1H), 3.42 (dd, J 1 ) 5.5 Hz, J 2 ) 7.4
Hz, 1H), 3.7 (s, 3H). 13C NMR (CDCl3): δ 13.8, 22.4, 25.3, 28.9,
29.11, 19.13, 31.6, 34.7, 51.5, 54.1, 176.4. Anal. (C11H23NO2)
C, H, N, P. Compound 6a was converted to the free base by
treatment with triethylamine (6.1 mL, 43.8 mmol) in methanol
(10 mL). Diethyl ether was added to precipitate triethylam-
monium chloride, which was then removed by filtration. The
0.8-0.9 (m, 3H), 1.0-1.22 (m, 10H), 1.36-1.56 (m, 2H), 3.06-
3.18 (m, 2H), 3.18-3.48 (m, 4H), 3.48-3.96 (m, 15H). 13C NMR
(CDCl3): δ 23.7, 26.9, 30.4, 31.8 (d, J ) 7.6 Hz), 33, 50.6, 53.2,
54, 55.2, 55.5, 55.7, 58.9 (d, J ) 7.1 Hz), 64 (d, J ) 5.1 Hz),
66.9, 67.2 (d, J ) 6 Hz), 169.3, 174.3, 175.5. Anal. (C23H42N3O14-
P‚0.55NaCl) C, H, N, P. FAB-MS exact mass calcd for
C23H43N3O14P+: 616.2483. Found: 616.2484.
2-(1-Na p h th yl)eth yl-p h osp h on ooxym eth yl-DTP A (4c).
A solution of 2-(1-naphthyl)ethanol phosphate cyanoethyl ester
3c (3.21 g, 3.2 mmol) was deprotected following the same

3472 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16
McMurry et al.
filtrate was concentrated. Ethylenediamine (75 mL) was
added, and the mixture was stirred overnight at RT under
nitrogen. Ethylenediamine was removed by evaporation at
reduced pressure, and the resulting solid was dried in vacuo
to give 5.94 g (25.9 mmol, 89% yield). 1H NMR (CDCl3): δ 0.82
(t, J ) 6.6 Hz, 3H), 1.21 (br s, 12H), 1.36-1.53 (m, 1H), 1.71-
1.85 (m, 1H), 2.20 (br s, 4H), 2.82 (t, J ) 5.8 Hz, 2H), 3.25-
3.38 (m, 3H), 7.64 (t, J ) 5.4 Hz, 1H). 13C NMR (CDCl3): δ
13.7, 22.2, 25.5, 28.9, 29.1, 31.4, 34.9, 40.80, 40.83, 55, 175.5.
FAB-MS exact mass calcd for C12H28N3O+: 230.2232. Found:
230.2230.
by flash chromatography (20% EtOAc/hexanes) to give 9a as
a light orange oil (0.54 g, 13.4 mmol, 72% yield). 1H NMR
(CDCl3): δ 0.82 (m, 3H), 1.22 (m, 14H), 1.44 (s, 45H), 2.24 (m,
1H), 2.7 (m, 6H), 3.3-3.55 (m, 10H). 13C NMR (CDCl3): δ 14,
22.6, 26.8, 28, 28.05, 28.09, 29.3, 29.5, 29.9, 31.1, 31.8, 52.4,
53, 53.1, 56, 56.5, 60.7, 80.2, 80.4, 80.6, 170.6, 171.1, 171.5.
Anal. (C42H79N3O10‚0.12CH2Cl2) C, H, N. FAB-MS exact mass
calcd for C42H80N3O10+: 786.5844. Found: 786.5823.
2-((2-Na p h th yl)m eth yl)d ieth ylen etr ia m in e P en ta -ter t-
bu tyl Aceta te (9b). Compound 9b was prepared from 2-((2-
naphthyl)methyl)diethylenetriamine trihydrochloride 8b (4.22
g, 12 mmol) by the same procedure described for 9a . Product
9b was purified by flash chromatography (25% EtOAc/hex-
Na p h th yla la n in e Eth ylen ed ia m in e Am id e (7b). Race-
mic 2-naphthylalanine methyl ester hydrochloride 6b (3.53 g,
13.4 mmol) was prepared by the method of O’Donnell.18 6b.
1H NMR (CDCl3): δ 3.0 (dd, J 1 ) 7.9 Hz, J 2 ) 13.4 Hz, 1H),
3.25 (dd, J 1 ) 5.0 Hz, J 2 ) 13.4 Hz, 1H), 3.7 (s, 3H), 3.82 (dd,
J 1 ) 5.3 Hz, J 2 ) 7.9 Hz, 1H), 7.28-7.32 (m, 1H), 7.38-7.48
(m, 2H), 7.63 (br s, 1H), 7.72-7.83 (m, 3H). 13C NMR (CDCl3):
δ 41.1, 51.8, 55.6, 125.4, 125.9, 127.1, 127.35, 127.43, 127.8,
128.0, 132.2, 133.2, 134.6, 175.2. Anal. (C14H15NO2) C, H, N,
P. Compound 6b was converted to compound 7b by the same
procedure described for 7a . The product was isolated as an
1
anes) to give an orange oil (8.38 g, 10.3 mmol, 86%). H NMR
(CDCl3): δ 1.32 (s, 9H), 1.40 (s, 36H), 2.53 (dd, J 1 ) 6.5 Hz, J 2
) 13.2 Hz, 1H), 2.77-2.93 (m, 2H), 2.99 (dd, J 1 ) 6 Hz, J 2
)
13.7 Hz, 1H), 3.14-3.26 (m, 1H), 3.34 (s, 2H), 3.36 (s, 4H),
3.47 (s, 4H), 7.32-7.43 (m, 3H), 7.63 (br s, 1H), 7.66-7.78 (m,
3H). 13C NMR (CDCl3): δ 27.8, 27.9, 37.2, 52.1, 52.6, 53.3, 55.6,
55.7, 55.9, 62.7, 80.2, 80.4, 124.8, 125.4, 127.2, 127.4, 127.7,
131.8, 133.3, 138.0, 170.4, 170.9, 171.1. Anal. (C45H71N3O10
C, H, N.
)
1
off-white solid (3.39 g, 13.2 mmol, 98%). H NMR (CDCl3): δ
2-Octyld ieth ylen etr ia m in e P en ta a cetic Acid (10a ). The
2-octyldiethylenetriamine penta-tert-butyl acetate 9a (10.54
g, 13.4 mmol) was dissolved in dioxane and concentrated HCl
(1:1 ratio, trace metal grade, 150 mL) and was stirred at room
temperature for 16 h. The reaction mixture was concentrated
under reduced pressure to provide 8.07 g of a tan foam. A
portion of this material (4.57 g) was chromatographed on a
1.05-1.35 (br m, 4H), 2.77 (t, J ) 6.1 Hz, 2H), 2.90 (dd, J 1
)
13.7 Hz, J 2 ) 8.9 Hz, 1H), 3.29 (m, 2H), 3.39 (dd, J 1 ) 13.7
Hz, J 2 ) 4.2 Hz, 1H), 3.69 (dd, J 1 ) 8.9 Hz, J 2 ) 4.2 Hz, 1H),
7.36 (dd, J 1 ) 8.4 Hz, J 2 ) 1.8 Hz, 1H), 7.4-7.54 (m, 3H), 7.64
(br s, 1H), 7.74-7.84 (m, 3H). 13C NMR (CDCl3): δ 41.1, 41.3,
41.7, 56.2, 125.5, 126.0, 127.1, 127.2, 127.5, 127.8, 128.2, 132.1,
133.2, 135.2, 174.4. Anal. (C15H19N3O‚0.3H2O) C, H, N. FAB-
C
18 reversed phase silica gel column (from H2O to 20-60% CH3-
MS exact mass calcd for
258.1602.
C
15H20N3O+: 258.1606. Found:
CN/H2O) applying 1.5 g per run. Acetonitrile was evaporated
under a stream of argon, and the resulting aqueous solution
was lyophilized, producing a flocculent white solid (2.97 g, 5.9
mmol). 1H NMR (D2O, CD3CN ref, pH <1): δ 0.74 (t, 3H),
1.10-1.30 (m, 12H), 1.4 (m, 1H), 1.70 (m, 1H), 2.86 (dd, 1H),
3.20-3.04 (m, 3H), 3.70-3.40 (m, 5H), 4.45 (m, 4H), 4.26 (m,
4H). 13C NMR (MeOD): δ 14.5, 23.7, 27.7, 28.4, 30.4, 30.5, 30.8,
33, 52, 52.3, 53.7, 55.2, 56.5, 57.1, 60.9, 171.5, 172.9, 174. Anal.
(C22H39N3O10) C, H, N. MS: m/e 504.5 [(M - H)-].
2-Octyldieth ylen etr iam in e Tr ih ydr och lor ide (8a).2-Ami-
nodecanoic acid ethylenediamine amide 7a (5.94 g, 25.9 mmol)
was suspended in 50 mL of THF. Borane‚THF (130 mL, 1.0
M) was added slowly, and the mixture was then refluxed under
argon for 16 h. The excess borane was quenched by careful
addition of methanol (100 mL: Caution, Exothermic!). The
reaction mixture was concentrated under reduced pressure.
Dry ethanol (150 mL) was added, and the solution was
saturated with HCl gas at 0 °C. The mixture was brought to
reflux for 24 h, after which the solution was cooled in an ice
bath. The resulting solid was filtered, washed with diethyl
ether, and vacuum-dried to give 6.08 g (18.7 mmol, 72%) of a
white powder. 1H NMR (MeOD): δ 0.74-0.8 (m, 3H), 1.1-
1.42 (m, 12H), 1.52-1.80 (m, 2H), 3.22-3.44 (m, 6H), 3.46-
3.63 (m, 1H).13C NMR (MeOD): δ 14.4, 23.7, 26, 30.29, 30.35,
32.1, 32.9, 37, 46.5, 50.4, 50.6. Anal. (C12H32Cl3N3) C, H, N.
FAB-MS exact mass calcd for C12H30N3+: 216.2440. Found:
216.2444.
2-((2-Na p h th yl)m eth yl)d ieth ylen etr ia m in e Tr ih yd r o-
ch lor id e (8b). Compound 8b was prepared from racemic
naphthylalanine ethylenediamine amide 7b (3.39 g, 13.2
mmol) by the same procedure described for 7a . Product 8b was
obtained as an off-white solid (4.22 g, 12 mmol, 91%). 1H NMR
(D2O): δ 3.02 (dd, J 1 ) 8.4 Hz, J 2 ) 14.2 Hz, 1H), 3.12-3.5
(m, 5H), 3.84-3.97 (m, 1H), 7.29 (d, J ) 8.7 Hz, 1H), 7.36-
7.46 (m, 2H), 7.68 (br s, 1H), 7.72-7.84 (m, 3H). 13C NMR
(MeOD): δ 37, 38, 46.5, 50.4, 51.5, 127.4, 127.6, 128, 128.7,
129.7, 130.2, 132.9, 134.2, 135. Anal. (C15H24Cl3N3) C, H, N.
FAB-MS exact mass calcd for C15H22N3+: 244.1814. Found:
244.1814.
2-((2-Na p h th yl)m eth yl)d ieth ylen etr ia m in e P en ta a ce-
tic Acid (10b). Compound 10b was prepared from 2-((2-
naphthyl)methyl)diethylenetriamine penta-tert-butyl acetate
9b (4.41 g, 5.42 mmol) by the same procedure described for
10a . Product 10b was purified by chromatography on a C18
reversed phase silica gel column (from H2O to 20% CH3CN/
H2O) and isolated as a fluffy white powder (2.3 g, 4.31 mmol,
80%). 1H NMR (D2O): δ 2.83 (dd, 1H), 3.0-3.19 (m, 3H), 3.28-
3.4 (m, 2H), 3.91-3.49 (m, 9H), 4.19 (s, 4H), 7.34-7.49 (m,
3H), 7.71 (s, 1H), 7.75-7.85 (m, 3H). 13C NMR (MeOD): δ 35.2,
51, 52.8, 53.7, 55.9, 56.7, 61.7, 126.9, 127.4, 128.3, 128.6, 128.7,
129, 129.7, 133.8, 135, 136.3, 170.7, 173.8, 175.1. Anal.
(C25H31N3O10) C, H, N.
Bis-[N-m et h yl-D-glu ca m in e]-{(2-(oct yld iet h ylen et r i-
a m in e)d ieth ylen etr ia m in ep en ta a ceta to)(a qu o)ga d olin -
iu m (III)} (11a ). A test tube was charged with gadolinium
oxide (0.145 g, 0.4 mmol), compound 10a (0.457 g, 0.84 mmol),
N-methylglucamine (0.31 g, 1.6 mmol), and deionized water
(3.0 mL). The mixture was stirred at 95 °C for 6 h; the solution
was cooled to room temperature, and the pH was adjusted to
7.0 with 0.07 g of NMG. The volume was adjusted to 4.0 mL,
and the solution of the product was filtered through a 0.2 µm
filter. Gadolinium content, calcd: 200 mM. Found (ICP): 196
mM. A portion of this material was isolated by reversed phase
chromatography on a C18 silica gel column (from H2O to 20-
50% CH3CN) and lyophilized to give a white solid. FAB-MS
exact mass calcd for C22H34GdN3Na3O10+: 727.1178. Found:
727.1189. Anal. (C36H70GdN5O21‚3H2O) C, Gd, H, N.
2-Octyld ieth ylen etr ia m in e P en ta -ter t-bu tyl Aceta te
(9a ). 2-Octyldiethylenetriamine trihydrochloride 8a (6.08 g,
18.7 mmol) was suspended in dry DMF (100 mL). Diisopro-
pylethylamine (47 mL, 280 mmol) was added followed by tert-
butylbromoacetate (24.2 mL, 0.15 mmol). The mixture was
stirred at room temperature for 16 h under argon. The DMF
and excess reagents were removed by evaporation under
reduced pressure. The residue was partitioned between con-
centrated aqueous NaHCO3 and dichloromethane. The organic
solution was washed one time each with water and saturated
aqueous NaCl, and then dried over Na2SO4, filtered, and
concentrated under reduced pressure. The residue was purified
Bis-[N-m eth yl-D-glu ca m in e]-{(2-((2-n a p h th yl)m eth yl)-
d ie t h yle n e t r ia m in e p e n t a a ce t a t o)(a q u o)ga d olin iu m -
(III)} (11b). Compound 11b was prepared from compound 10b
(0.590 g, 1.05 mmol) and gadolinium oxide (0.181 g, 0.5 mmol)
by the same procedure described for 11a . A clear solution of
the gadolinium complex was obtained. MS: m/e 685.9 [(M -
H2O + H)-]. FAB-MS exact mass calcd for C25H27GdN3-

Effect of a Phosphodiester Linking Group
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16 3473
+
Na2O10
:
733.0733. Found: 733.0749. HPLC purity: 99%
biexponential equation
(YMC-Pack TMS).
Ra d iola belin g P r oced u r e. Radiochemical trace labeling
was accomplished using 153Gd chloride as described in the
literature.19 Samples of 153Gd chloride of 99% radionucleotide
purity and known specific activity (10-12 mCi/mg) were
obtained from Dupont New England Nuclear (North Billerica,
MA). Solutions of gadolinium chelates (0.1-0.25 mM) were
reacted with 153Gd chloride at pH 4-5 at 80 °C for 1-3 h.
Specific activity of the labeled compounds was typically 200-
300 µCi/mmol, with values for free gadolinium generally less
than 0.2% as determined by TLC.19 The pH of the reaction
mixture was adjusted between pH 7 and 8 with N-methyl-D-
glucamine or sodium hydroxide. HPLC experiments confirmed
that this procedure resulted in no degradation of the chelate,
as evidenced by a single 153Gd chelate peak at a retention time
consistent with that of the unlabeled material. Gamma count-
ing was performed using a Cobra II gamma counter (Packard,
Downers Grove, IL) operating with a counting window of 15-
140 keV. Samples were counted for 30 min or until deviations
of >1% sigma were achieved.
C ) A exp(-Rt) + B exp(-ât)
where C is the plasma concentration, A is the distribution
coefficient, R is the distribution rate constant, B is the
elimination coefficient, â is the elimination rate constant, and
t is the time. The elimination half-life (Te1/2), distribution half-
life (Td1/2), and apparent volume of distribution (Vd) were
calculated by means of the following formulas:
Te1/2 ) 0.693/â
Td1/2 ) 0.693/R
Vd ) dose/B
Area under the curve (AUC) was calculated for each individual
animal by J MP using the trapezoidal rule
Bu ta n ol Bu ffer P a r tition Coefficien ts. Because of the
overall hydrophilic nature of the gadolinium chelates, partition
coefficients were measured in butanol/phosphate-buffered
saline (PBS) instead of octanol/PBS. 153Gd-labeled chelate
([Gd]T ) 0.1 mM in 0.5 mL of PBS) was equilibrated at RT for
1-2 h with PBS-saturated butanol (0.5 mL). The vials were
centrifuged at 2000g for 5 min to ensure that the layers were
separated. Aliquots (100 µL) from each phase were removed
and counted in a Packard Cobra II gamma counter. The
partition coefficient (P) was calculated as follows (CPM )
counts per minute):
n
Ci + Ci+1
AUC0ft
)
(ti+1 - ti)
x
∑
(
)
[
]
2
i)1
where AUC0ft is the time period desired beginning with 0, n
is the number of time points, and Ci is the plasma concentra-
tion at time ti. For excretion studies, rats were housed in
metabolic cages for the collection of urine and feces and were
sacrificed at 24 h. Tissue samples, urine, and feces were
analyzed for 153Gd content by gamma counting. The percent
injected dose was determined as follows, where % ID is percent
of injected dose and CPM is counts per minute.
x
average CPM/µL of butanol
P )
average CPM/µL of PBS
P r otein Bin d in g. The protein binding studies were per-
formed by ultrafiltration using 153Gd-radiolabeled chelates 5b,
5c, 11a , 11b, and Gd-DTPA. A detailed analysis of the binding
properties of 5a is presented elsewhere.22
total CPM sample
injected dose (in CPM)
% ID )
× 100
In Vitr o Rela xivity. Proton T1 relaxation times were
determined at 20 MHz using a Minispec PC 120/125/VTs
(Bruker Canada, Milton, ON) nuclear magnetic resonance
(NMR) process analyzer. The temperature was regulated at
37 ( 0.2 °C with a circulating bath. The longitudinal (T1)
relaxation rate was determined by the use of an inversion
recovery pulse sequence and the fitting of the intensity data
from eight experimental time points to an exponential. Two
scans were averaged for each data point with a recycle delay
that was 5 times longer than the T1 observed for the sample.
The T1 was determined in triplicate, and the average T1
measurement was calculated.
Solutions of HSA (4.5 wt %, fraction V, 96-99% albumin,
Sigma Chemical Co., St. Louis, MO) were prepared by adding
HSA to phosphate-buffered saline (PBS, 10 mM sodium
phosphate, 150 mM sodium chloride, pH 7.4). A relaxivity
titration consisted of five sequential 1-10 µL syringe additions
of a solution of gadolinium chelate (1-10 mM) to a 10 mm
NMR tube containing 500-1000 µL of phosphate-buffered
saline or 4.5% HSA solution. After the addition of each chelate,
the sample was thoroughly mixed by manual agitation and
returned to the Bruker Minispec for the T1 determination.
Concentrations of the 1-10 mM chelate solutions were deter-
mined by either UV absorbance at 220 nm or ICP analysis.
Protein binding studies were initiated in duplicate by adding
10 mM 153Gd-labeled chelate solution to a vial containing 500
µL of 4.5% HSA to give a total gadolinium concentration of
0.1 mM. The contents of the vials were mixed by vortex, and
a 25-50 µL aliquot from each sample was pipetted to a small
vial and counted in the gamma counter. A portion of each
sample (400 µL) was pipetted to an ultrafiltration unit
(Ultrafree-MC 5000 Nominal Molecular Weight Limit filter
unit, Millipore, Bedford, MA) and centrifuged at room tem-
perature (22 ( 1 °C) or in an incubator after a 15 min
equilibration at 37 ( 1 °C (5a ). Filtrate (25 µL) was pipetted
to a small vial and counted in the gamma counter. The percent
of the agent unbound was calculated by comparing the counts
per microliter in the filtrate to that in the original solution.
The error in the binding data is (10%.
Ra t P h a r m a cology Stu d ies. All studies involving labora-
tory animals were approved by the Institutional Animal Care
and Use Committee (IACUC) at EPIX Medical, Inc. and were
in accordance with the guidelines in the “Guide to the Care
and Use of Laboratory Animals” published by the National
Research Council and the Institute of Laboratory Animal
Resources.
Unanesthetized male Sprague-Dawley rats (150-250 g,
Harlan Sprague Dawley, Indianapolis, IN; n ) 2-5) were
injected via tail vein catheter with 153Gd-labeled chelate at a
dose of 0.1 mmol/kg. Plasma samples were obtained by
centrifugation of freshly heparinized blood at 2000g for up to
20 min. The plasma supernatant was pipetted away from the
pellet fraction and either frozen or analyzed within 24 h of
collection. Concentrations were determined by inductively
coupled plasma (ICP) analysis (Galbraith Laboratories), and
plasma water T1 relaxation times were obtained as described
below.
Ack n ow led gm en t. We thank Nathan Ekborg for
assistance in obtaining the mass spectrometry data.
Refer en ces
(1) Yucel, E. K.; Lauffer, R. B. Blood pool agents for MRA. Semin.
Interventional Radiol. 1998, 15 (2), 215-222.
(2) Debatin, J . F.; Hany, T. F. MR-based assessment of vascular
morphology and function. Eur. Radiol. 1998, 8 (4), 528-539.
(3) Baum, R. A. Peripheral vascular diagnosis using magnetic
resonance angiography. Semin. Interventional Radiol. 1998, 15
(2), 135-148.
Plasma concentration data were analyzed with the J MP
statistical package (SAS Institute Inc., Cary, NC) using the

3474 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 16
McMurry et al.
(4) Schmiedl, U.; Ogan, M.; Paajanen, H.; Marotti, M.; Crooks, L.
E.; Brito, A. C.; Brasch, R. C. Albumin labeled with gadolinium-
DTPA as an intravascular, blood pool-enhancing agent for MR
imaging, biodistribution and imaging studies. Radiology 1987,
162 (1), 205-210.
(21) Brechbiel, M. W.; Gansow, O. A. Backbone-Substituted DTPA
Ligands for 90Y Radioimmunotherapy. Bioconjugate Chem.
1991, 2, 187-194.
(22) Caravan, P.; Cloutier, N. J .; Greenfield, M. T.; McDermid, S.
A.; Dunham, S. U.; Bulte, J . W. M.; Amedio, J . C., J r.; Looby, R.
J .; Supkowski, R. M.; Horrocks, W. DeW., J r.; McMurry, T. J .;
Lauffer, R. B. The interaction of MS-325 with Human Serum
Albumin and Its Effects on Proton Relaxation Rates. J . Am.
Chem. Soc. 2002, 124 (12), 3152-3162.
(23) Schmitt-Willich, H.; Brehm, M.; Ewers, C.; et al. Synthesis and
Physicochemical Characterization of a new Gadolinium Che-
late: The Liver-Specific Magnetic Resonance Imaging Contrast
Agent Gd-EOB-DTPA. Inorg. Chem. 1999, 38, 1134-1144.
(24) Beaucage, S. L.; Radhakrishnan, P. I. Tetrahedron 1991, 2 (48),
2223-2311.
(25) He, X. M.; Carter, D. C. Atomic structure and chemistry of
human serum albumin. Nature 1992, 358 (6383), 209-215
[published erratum appears in Nature 1993, 364 (6435), 362].
(5) Schuhmann-Giampieri, G.; Schmitt-Willich, H.; Frenzel, T. In
Vivo and In Vitro Evaluation of Gd-DTPA-Polylysine as
a
Macromolecular Contrast Agent for Magnetic Resonance Imag-
ing. Invest. Radiol. 1991, 26, 969-974.
(6) Kellar, K.; Fujii, D.; Gunther, W.; et al. ‘NC100150’ a preparation
of iron oxide nanoparticles ideal for positive-contrast MR an-
giography. Magn. Reson. Mater. Phys., Biol., Med. 1999, 8, 207-
213.
(7) J ohansson, L.; Nolan, M.; Taniuchi, M.; et al. High-Resolution
Magnetic Resonance Coronary Angiography of the Entire Heart
using a New Blood-Pool Agent NC100150 Injection: Comparison
with X-ray Angiography. Circulation 1998, 98 (17S), I-514.
(8) Dong, Q.; Hurst, D. R.; Weinmann, H. J .; Chenevert, T. L.;
Londy, F. J .; Prince, M. R. Magnetic Resonance Angiograaphy
with Gadomer-17: An original animal investigation. Invest.
Radiol. 1998, 33 (9), 699-708.
(9) Lauffer, R. B. Paramagnetic Metal Complexes as Water Proton
Relaxation Agents for NMR Imaging: Theory and Design. Chem.
Rev. 1987, 87, 901-927 and references therein.
(10) Caravan, P.; et al. Gadolinium(III) Chelates as MRI Contrast
Agents: Structure, Dynamics, and Applications. Chem. Rev.
1999, 99 (9), 2293-2352 and references therein.
(26) Hamm, B.; Staks, T.; Muhler, A.; et al. Phase
I Clinical
Evaluation of Gd-EOB-DTPA as a Hepatobiliary MR Contrast
Agent: Safety, Pharmacokineics, and MR Imaging. Radiology
1995, 195, 785-792.
(27) Klaassen, C. D.; Watkins, J . B. I. Am. Soc. Pharm. Exp. Ther.
1984, 36, 1.
(28) Adzamli, K.; et al. Development of a Novel Nonaromatic Small-
Molecule MR Contrast Agent for the Blood Pool. Acad. Radiol.
1998, 5 (Suppl. 1), S210-S213.
(29) Wallace, R.; et al. Synthesis and Preliminary Evaluation of MP-
2269: A Novel, Nonaromatic Small-Molecule Blood-Pool MR
Contrast Agent. Magn. Reson. Med. 1998, 40, 733-730.
(30) McMurry, T. J .; et al. PCT Int. Appl WO9623526, 1996; Chem.
Abstr. 1996, 125, 230822.
(11) Schuhmann-Giampieri, G.; Schmitt-Willich, H.; Press, W.; et al.
Preclinical Evaluation of Gd-EOB-DTPA as a Contrast Agent
in MR Imaging of the Hepatobiliary System. Radiology 1992,
183, 59-64.
(12) Scott, D. M.; Lauffer, R. B. 1-(p-n-Butylbenzyl)DTPA for Mag-
netic Resonance Imaging. U.S. Patent 5,582,814, 1996.
(13) Parmelee, D. J .; Walovitch, R. C.; Ouellet, H. S.; Lauffer, R. B.
Preclinical evaluation of the pharmacokinetics, biodistribution,
and elimination of MS-325, a blood pool agent for magnetic
resonance imaging. Invest. Radiol. 1997, 32 (12), 741-747.
(14) Lauffer, R. B.; Parmelee, D. J .; Dunham, S. U.; Ouellet, H. S.;
Dolan, R. P.; Witte, S.; McMurry, T. J .; Walovitch, R. C. MS-
325: Albumin-targeted contrast agent for MR angiography.
Radiology 1998, 207 (2), 529-538.
(31) Sieving, P.; Watson, S.; Rocklage, S. Preparation and Charac-
terization of Paramagnetic Polychelates and Their Protein
Conjugates. Bioconjugate Chem. 1989, 65-81.
(32) Spanoghe, M.; et al. Proton Relaxation Enhancement by Means
of Serum Albumin and Poly-L-lysine Labeled with DTPA-Gd3+
:
Relaxivities as a Function of Molecular Weight and Conjugation
Efficiency. Magnetic Resonance Imaging; Pergamon Press: Elms-
ford, NY, 1992; pp 913-917.
(15) Brasch, R. C. Rationale and applications for macromolecular
gadolinium-based contrast agents. Magn. Reson. Med. 1991, 22
(2), 282-287.
(33) Frank, H.; Weissleder, R.; Bogdanov, A.; Brady, T. Am. J .
Roentgenol. 1994, 1041-1046.
(34) Weiner, E.; Brechbiel, M.; Brothers, H.; Magin, R.; Gansow, O.;
Tomalia, D.; Lauterbur, P. Dendrimer-Based Metal Chelates:
A New Class of Magnetic Resonance Imaging Contrast Agents.
Magn. Reson. Med. 1994, 31, 1-8.
(35) Sajiki, H.; Ong, K. Y.; Nadler, S. T.; Wages, H. E.; McMurry, T.
J . Synthesis of enantiomerically pure 1-(R)- and 1-(S)-hydroxy-
methyl-DTPA penta-t-butylesters via chiral amino alcohols
Synth. Commun. 1996, 26 (13), 2511-2522.
(36) Amedio, J . C., J r.; Bernard, P. J .; Fountain, M.; Van Wagenen,
G., J r. A practical manufacturing synthesis of 1-(R)-hydroxy-
methyl-DTPA: an important intermediate in the synthesis of
MRI contrast agents. Synth. Commun. 1999, 29 (14), 2377-2391.
(16) Schmiedl, U. Comparison of the Contrast Enhancing Properties
of Albumin-(Gd-DTPA) and Gd-DTPA at 2.0T: An Experimental
Study in Rats. Am. J . Roentgenol. 1986, 147, 1263-1270.
(17) Grist, T. M.; Korosec, F. R.; Peters, D. C.; Witte, S.; Walovitch,
R. C.; Dolan, R. P.; Bridson, W. E.; Yucel, E. K.; Mistretta, C. A.
Steady-state and dynamic MR angiography with 5a : Initial
experience in humans. Radiology 1998, 207 (2), 539-544.
(18) O’Donnell, M. J .; Polt, R. L. A mild and efficient route to Schiff
base derivatives of amino acids. J . Org. Chem. 1982, 47 (13),
2663-2666.
(19) Kumar, K.; et al. True Tracer Radiolabeling of Gadolinium
Complex of 10-(2-Hydroxypropyl)-1,4,7,10-tetraazacyclodode-
cane-1,4,7-triacetic acid (HP-DO3A). J . Labelled Compd. Ra-
diopharm. 1993, 33 (6), 473-482.
(20) Brechbiel, M. W.; et al. Synthesis of 1-(p-isothiocyanatobenzyl)
derivatives of DTPA and EDTA. Antibody labeling and tumor-
imaging studies. Inorg. Chem. 1986, 25, 2772-2781.
J M0102351