Gd-DTPA L-cystine bisamide copolymers as novel biodegradable macromolecular contrast agents for MR blood pool imaging

Article abstract of DOI:10.1007/s11095-006-0024-0

Purpose. The purpose of this study was to synthesize biodegradable Gd-DTPA l-cystine bisamide copolymers (GCAC) as safe and effective, macromolecular contrast agents for magnetic resonance imaging (MRI) and to evaluate their biodegradability and efficacy in MR blood pool imaging in an animal model. Methods. Three new biodegradable GCAC with different substituents at the cystine bisamide [R = H (GCAC), CH₂CH₂CH₃ (Gd-DTPA l-cystine bispropyl amide copolymers, GCPC), and CH(CH₃)₂ (Gd-DTPA cystine bisisopropyl copolymers, GCIC)] were prepared by the condensation copolymerization of diethylenetriamine pentaacetic acid (DTPA) dianhydride with cystine bisamide or bisalkyl amides, followed by complexation with gadolinium triacetate. The degradability of the agents was studied in vitro by incubation in 15 μM cysteine and in vivo with Sprague-Dawley rats. The kinetics of in vivo contrast enhancement was investigated in Sprague-Dawley rats on a Siemens Trio 3 T scanner. Results. The apparent molecular weight of the polydisulfide Gd(III) chelates ranged from 22 to 25 kDa. The longitudinal (T ₁) relaxivities of GCAC, GCPC, and GCIC were 4.37, 5.28, and 5.56 mM^-1 s^-1 at 3 T, respectively. The polymeric ligands and polymeric Gd(III) chelates readily degraded into smaller molecules in incubation with 15 μM cysteine via disulfide-thiol exchange reactions. The in vitro degradation rates of both the polymeric ligands and macromolecular Gd(III) chelates decreased as the steric effect around the disulfide bonds increased. The agents readily degraded in vivo, and the catabolic degradation products were detected in rat urine samples collected after intravenous injection. The agents showed strong contrast enhancement in the blood pool, major organs, and tissues at a dose of 0.1 mmol Gd/kg. The difference of their in vitro degradability did not significantly alter the kinetics of in vivo contrast enhancement of the agents. Conclusion. These novel GCAC are promising contrast agents for cardiovascular and tumor MRI, which are later cleaved into low molecular weight Gd(III) chelates and rapidly cleared from the body.

Full text of DOI:10.1007/s11095-006-0024-0

#
Pharmaceutical Research, Vol. 23, No. 6, June 2006 ( 2006)
DOI: 10.1007/s11095-006-0024-0
Research Paper
Gd-DTPA L-Cystine Bisamide Copolymers as Novel Biodegradable
Macromolecular Contrast Agents for MR Blood Pool Imaging
Todd L. Kaneshiro,1 Tianyi Ke,1 Eun-Kee Jeong,2 Dennis L. Parker,2 and Zheng-Rong Lu1,3
Received November 29, 2005; accepted January 18, 2006
Purpose. The purpose of this study was to synthesize biodegradable Gd-DTPA L-cystine bisamide
copolymers (GCAC) as safe and effective, macromolecular contrast agents for magnetic resonance
imaging (MRI) and to evaluate their biodegradability and efficacy in MR blood pool imaging in an
animal model.
Methods. Three new biodegradable GCAC with different substituents at the cystine bisamide [R = H
(GCAC), CH2CH2CH3 (Gd-DTPA L-cystine bispropyl amide copolymers, GCPC), and CH(CH3)2 (Gd-
DTPA cystine bisisopropyl copolymers, GCIC)] were prepared by the condensation copolymerization of
diethylenetriamine pentaacetic acid (DTPA) dianhydride with cystine bisamide or bisalkyl amides,
followed by complexation with gadolinium triacetate. The degradability of the agents was studied in
vitro by incubation in 15 mM cysteine and in vivo with Sprague-Dawley rats. The kinetics of in vivo
contrast enhancement was investigated in Sprague-Dawley rats on a Siemens Trio 3 T scanner.
Results. The apparent molecular weight of the polydisulfide Gd(III) chelates ranged from 22 to 25 kDa.
The longitudinal (T1) relaxivities of GCAC, GCPC, and GCIC were 4.37, 5.28, and 5.56 mMj1 sj1
at 3 T, respectively. The polymeric ligands and polymeric Gd(III) chelates readily degraded into smaller
molecules in incubation with 15 mM cysteine via disulfideYthiol exchange reactions. The in vitro
degradation rates of both the polymeric ligands and macromolecular Gd(III) chelates decreased as the
steric effect around the disulfide bonds increased. The agents readily degraded in vivo, and the catabolic
degradation products were detected in rat urine samples collected after intravenous injection. The agents
showed strong contrast enhancement in the blood pool, major organs, and tissues at a dose of 0.1 mmol
Gd/kg. The difference of their in vitro degradability did not significantly alter the kinetics of in vivo
contrast enhancement of the agents.
Conclusion. These novel GCAC are promising contrast agents for cardiovascular and tumor MRI, which
are later cleaved into low molecular weight Gd(III) chelates and rapidly cleared from the body.
KEY WORDS: biodegradable macromolecular contrast agent; blood pool imaging; Gd-DTPA L-cystine
bisamide copolymers; magnetic resonance imaging; polydisulfide.
1 Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, 421 Wakara Way, Suite 318, Salt Lake City,
Utah 84108, USA.
2 Department of Radiology, University of Utah, Salt Lake City, Utah 84108, USA.
3 To whom correspondence should be addressed. (e-mail: zhengrong.lu@utah.edu)
ABBREVIATIONS: DCAC, DTPA L-cystine bisamide copolymers; DCC, dicyclohexylcarbodiimide; DCIC, DTPA L bisisopropyl amide
copolymers; DCPC, DTPA L-cystine bispropyl amide copolymers; DCU, dicyclohexylurea; DI, deionized; DMSO, dimethylsulfoxide; DTPA
dianhydride, diethylenetriamine penta acetic acid dianhydride; ESI-MS, electrospray ionization mass spectrometry; GCAC, Gd-DTPA L-
cystine bisamide copolymers; GCIC, Gd-DTPA L-cystine bisisopropyl amide copolymers; GCPC, Gd-DTPA L-cystine bispropyl amide
copolymers; GDCC, Gd-DTPA cystamine copolymers; GDCEP, Gd-DTPA cystine diethyl ester copolymers; GDCP, Gd-DTPA cystine
copolymers; Gd-(DTPA-BMA), Gd-(DTPA-bismethyl amide); Gd(OAc)3, gadolinium triacetate; HPMA, poly[N-(2-hydroxypropyl)metha-
crylamide]; ICP-OES, inductively coupled argon plasma optical emission spectrometer; MALDI-TOF, matrix-assisted laser desorption ioni-
zation time of flight; MRI, magnetic resonance imaging; MWCO, molecular weight cutoff; PBS, phosphate-buffered saline; r1, longitudinal
relaxivity; r2, transverse relaxivity; SEC, size exclusion chromatography; T1, longitudinal; T2, transverse; TEA, triethylamine; TFA, trifluor-
oacetic acid; THF, tetrahydrofuran; 1/T1, proton longitudinal relaxation rate; 1/T2, proton transverse relaxation rate.
#
1285
0724-8741/06/0600-1285/0
2006 Springer Science + Business Media, Inc.

1286
Kaneshiro et al.
INTRODUCTION
the amide and substituted alkyl amides have on the
physicochemical properties and in vivo contrast enhancement
Magnetic resonance imaging (MRI) is a medical imaging of the agents. The macromolecular Gd(III) chelates with
modality that is routinely used to identify abnormal tissues. substituted cystine bisamides exhibited slightly higher relax-
The high spatial and temporal resolution of MRI provides ivities and a slower in vitro degradation rate than that of
both anatomical and physiological information (1,2). Para- nonsubstituted cystine bisamide Gd(III) complexes. In vivo
magnetic gadolinium [Gd(III)] complexes, or Gd(III) che- degradation was studied in rats and showed that the catabolic
lates, can significantly increase the proton longitudinal (1/T1) products, which were low molecular weight Gd(III) chelates,
and transverse (1/T2) relaxation rates of surrounding water were detected in the urine samples. MRI studies in rats
molecules (3,4). Stable Gd(III) chelates, e.g., Gd-(DTPA- showed that these newly prepared biodegradable, high
bismethyl amide) [Gd-(DTPA-BMA)], are used as low molecular weight contrast agents significantly improved con-
molecular weight MRI contrast agents to improve the trast enhancement of the blood pool, liver, and kidneys.
diagnostic accuracy by enhancing the image contrast between
normal and diseased tissues. Approximately 30% of the MRI
examinations utilize these contrast agents (3).
MATERIALS AND METHODS
The most common clinical MRI contrast agents are Gd-
diethylenetriamine pentaacetic acid (Gd-DTPA), Gd-DOTA,
All reagents were used without further purification unless
otherwise stated. All reactions were performed at room
or their derivatives (4). However, their clinical application is
temperature under ambient pressure unless otherwise stated.
limited because of their transient tissue retention and unfa-
Cystine was purchased from Aldrich Chemical Co. (Milwau-
vorable pharmacokinetics, which includes rapid extravasation
from the cardiovasculature and undiscriminable distribution
Peninsula Laboratories (San Carlos, CA, USA). Isopropyl
into the extracellular space (5,6). As a result, macromolecular
kee, WI, USA). Cystine bisamide was purchased from
amine, propyl amine, and N-hydroxysuccinimide (NHS) were
purchased from Lancaster Synthesis Inc. (Pelham, NH,
USA). Gadolinium triacetate (Gd(OAc)3) and di-tert-butyl
dicarbonate (t-BOC) were purchased from Alfa Aesar (Ward
Hill, MA, USA). DTPA was purchased from J.T. Baker
Gd(III) chelates have been developed as blood pool MRI
contrast agents to modify in vivo pharmacokinetics (5Y9).
These blood pool contrast agents have a prolonged blood
retention time or plasma half-life that permits a broader
contrast enhancement time window for effective MRI of the
(Philipsburg, NJ, USA). Dicyclohexylcarbodiimide (DCC)
vasculature. These agents also exhibit greater image contrast
was purchased from Pierce (Rockford, IL, USA). (t-BOC)2-
cystine (18) and DTPA dianhydride (19) were synthesized
because they have higher proton relaxation rates, which is
caused by the larger number of Gd(III) chelates in a single
according to the literature. Spectra/Pore regenerated cellu-
macromolecule. These high molecular weight agents are often
prepared by conjugation of a Gd(III) chelate, such as Gd-
kDa] were purchased from Spectrum Laboratories (Rancho
DTPA, to polymers (7,9) dendrimers (5,8), or proteins (10) or
lose membranes [molecular weight cutoff (MWCO) = 25
Dominguez, CA, USA). PD-10 desalting columns were
by the copolymerization of Gd-DTPA with alkyldiamines
purchased from Amersham Bioscience (Uppsala, Sweden).
(6,11). One limitation of macromolecular contrast agents for
clinical applications, however, is their slow excretion, which
1H NMR spectra were acquired on a Varian INOVA 400
may result in the release of toxic Gd3+ ions from the chelates
by metabolism and accumulation of the ions in the bone
at 400 MHz at 25-C. Molecular weight was determined by
size exclusion chromatography (SEC) on an AKTA FPLC
system with a Superdex 200 (10/300 GL) column equipped
(9,12).
with UV and refractive index detectors (Amersham
Biosciences Corp., Piscataway, NJ, USA), calibrated with
poly[N-(2-hydroxypropyl)methacrylamide] (HPMA) stan-
dards with a molecular weight range of 14Y300 kDa. Electro-
spray ionization mass spectrometry (ESI-MS) spectra were
acquired on a PE Sciex API III Mass Spectrometer at the
University of Utah Mass Spectrometry and Proteonomic
Core Facility (Salt Lake City, UT, USA). The Gd metal
content was determined using an inductively coupled argon
plasma optical emission spectrometer (ICP-OES) (Optima
3100XL; Perkin Elmer, Norwalk, CT, USA). The T1 relax-
ivity measurements for the polymeric Gd(III) chelates were
performed on a Siemens Trio 3 T MRI scanner utilizing an
inversion recovery sequence; the T2 relaxivity measurements
for the macromolecular Gd(III) chelates were performed
utilizing a spin-echo recovery sequence. T1-weighted MRI
was acquired on a Siemens Trio 3 T MRI scanner.
To resolve the issue of macromolecular contrast agents
limiting their clinical development, we have recently designed
and developed polydisulfide-based macromolecular Gd(III)
chelates as biodegradable MRI contrast agents (13Y17). These
agents provide superior blood pool contrast enhancement
within vasculature and readily degrade into low molecular
weight species that can rapidly excrete from the body (16).
Because the polymeric backbone is constructed upon disul-
fide bonds, they can degrade via reduction with endogenous
plasma free thiols, including glutathione and cysteine.
Polydisulfide Gd(III) chelates with different structural
modifications around the disulfide bonds have been prepared
and evaluated. The preliminary results have shown that the
introduction of functional groups, like carboxylic acid in Gd-
DTPA cystine copolymers (GDCP) and ethyl carboxylate
Gd-DTPA cystine diethyl ester copolymers (GDCEP),
improved in vivo contrast enhancement as compared with
Gd-DTPA cystamine copolymers (GDCC) (14).
Here, we report the synthesis and evaluation of three Gd- Synthesis of Cystine Bispropyl Amide
DTPA L-cystine bisamide copolymers (GCAC) as biodegrad-
able macromolecular MRI contrast agents. These novel
(t-BOC-L-Cys)2 (10.0 g, 22.8 mmol) and NHS (5.84 g,
macromolecular MRI contrast agents were prepared in high 50.7 mmol) were dissolved in tetrahydrofuran (THF; 200 mL).
yield and purity. In addition, we investigated the effects that DCC (10.1 g, 49.1 mmol) was added to the solution, and the

Gd-DTPA L-Cystine Bisamide Copolymers
1287
mixture was stirred overnight at room temperature. N,N0- additional 0.50 mL of DMSO was added, and the reaction
Dicyclohexylurea (DCU) precipitated out of the THF was stirred overnight at room temperature. The solution was
solution and was removed by filtration. Propyl amine diluted with methanol, and DTPA L-cystine bisisopropyl
(10.0 mL, 116 mmol) was added to the filtrate while stirring amide copolymers (DCIC) were precipitated in acetone.
at j5-C. Additional DCU precipitated out within 5 min and DCIC was dissolved in DI H2O at pH 7 and dialyzed against
was removed by vacuum filtration. The solvent was removed DI H2O for approximately 30 h using a regenerated cellulose
in vacuo to result in the crude product. The residue was membrane (MWCO = 25 kDa). The polymeric ligand was
dissolved in chloroform (100 mL) and washed three times obtained after concentrating to dryness in vacuo. The
with deionized (DI) H2O at pH 10. The organic layer was molecular weight of DCIC was determined by SEC. DCIC
dried over anhydrous Na2SO4, filtered, and then concentrat- was mixed with a large excess of Gd(OAc)3 in 10 mL of DI
ed in vacuo to dryness. The crude product was washed with H2O at pH 5.5 while stirring at room temperature for 1 h.
ethyl acetate (100 mL), filtered, and then washed again with Free Gd3+ ions were removed by SEC using a PD-10
diethyl ether to yield (t-BOC)2-cystine bispropyl amide desalting column, eluted with DI H2O. The solution was
(71%) as a white solid. 1H NMR (CDCl3, ppm): 7.64 (d, concentrated in vacuo to dryness to give Gd-DTPA L-cystine
2H, t-BOC-NHCH), 5.55 (br, 2H, CONHCH2CH2CH3), 4.79 bisisopropyl copolymers (GCIC). The Gd content of GCIC
[q, 2H, NHCH(CO)CH2], 3.18Y3.22 (m, 8H, NHCH2CH2CH3 was measured using ICP-OES; molecular weight was
+ SCH2CH), 2.90 (m, 4H, NHCH2CH2CH3), 1.40 [s, 18H, determined by SEC.
OC(CH3)3], 0.90 (t, 6H NHCH2CH2CH3). ESI-MS (m/z, M +
H+): 523.3 (observed), 523.2 (calculated).
t-BOC protection was removed by dissolving the above
Gd-DTPA L-Cystine Bisamide Copolymers
product (1.54 g, 2.94 mmol) in trifluoroacetic acid (TFA) (3.0
mL) for 30 min while stirring at room temperature. The so-
lution was then concentrated in vacuo to give a viscous oil.
The product was precipitated out in diethyl ether and dis-
solved in DI H2O. The aqueous phase was separated and
concentrated to dryness in vacuo to yield L-cystine bispropyl
DTPA L-cystine bisamide copolymers (DCAC): yield =
67.5%; Mn = 30.8 kDa, Mw = 34.5 kDa; 1H NMR (D2O,
ppm): 4.20 (t, 2H, NHCHCO), 3.77 (s, 6H, NCH2COOH),
3.58 (s, 4H, COCH2N), 3.20 (d, 4H, SCH2CH), 2.87, (d, 8H,
NCH2CH2N). GCAC: yield = 81%; Mn = 14.4 kDa, Mw
=
22.3 kDa. The Gd content in GCAC was 169 mg Gd/g
copolymer.
1
amide (41%) as a colorless salt. H NMR (D2O, ppm): 4.15
[t, 2H, CO(NH)CHCH2], 3.22Y2.96 (m, 8H, SCH2CH +
NHCH2CH2CH3), 1.40 (m, 4H, CH2CH2CH3), 0.26 [t, 6H,
CH(CH3)2]. ESI-MS (m/z, M + H+): 323.3 (observed), 323.2
(calculated).
Gd-DTPA L-Cystine Bispropyl Amide Copolymer
DTPA L-cystine bispropyl copolymers (DCPC): yield =
60.2%; Mn = 24.1 kDa, Mw = 33.2 kDa; 1H NMR (D2O,
ppm): 4.55 (t, 2H, NHCHCO), 3.80 [m, 2H, NHCH(CH3)2],
3.66 (s, 6H, COCH2N + NCH2COOH), 3.55 (s, 4H, NC
Synthesis of L-Cystine Bisisopropyl Amide
(t-BOC-L-Cys)2-bisisopropyl amide was synthesized sim- H2COOH), 3.22 (d, 4H, SCH2CH), 3.03 (m, 2H, NHC
ilarly as (t-BOC-L-Cys)2-bispropyl amide. Yield: 35%. 1H H2CH2CH3), 2.87, (d, 8H, NCH2CH2N), 1.37 (m, 4H,
NMR (CDCl3, ppm): 7.34 (d, 2H, t-BOC-NHCH), 5.57 [d, NHCH2CH2CH3), 0.78 (t, 6H, NHCH2CH2CH3). GCPC:
2H, CONHCH(CH3)2], 4.70 [m, 2H, NHCH(CO)CH2], 4.12 yield = 73%; Mn = 17.1 kDa, Mw = 22.3 kDa. The Gd
[m, 2H, NHCH(CH3)2], 2.90 (m, 4H, SCH2CH), 1.48 [s, 18H, content in GCPC was 169 mg Gd/g copolymer.
OC(CH3)3], 1.19 [d, 12H, CH(CH3)2]. ESI-MS (m/z, M +
H+): 523.3 (observed), 523.2 (calculated).
Gd-DTPA L-Cystine Bisisopropyl Copolymers
L -Cystine bisisopropyl amide was similarly prepared and
1
purified as L-cystine bispropyl amide. Yield: 71%. H-NMR
(D2O, ppm): 4.08 [t, 2H, CO(NH)CHCH2], 3.83 [m, 2H,
NHCH(CH3)2], 3.12 (m, 4H, SCH2CH), 1.05 [d, 12H,
CH(CH3)2]. ESI-MS (m/z, M + H+): 323.3 (observed), 323.2
(calculated).
DCIC: yield = 60.0 %; Mn = 26.4 kDa, Mw = 36.7 kDa;
1H NMR (D2O, ppm): 4.55 (t, 2H, NHCHCO), 3.80 [m, 2H,
NHCH(CH3)2], 3.66 (s, 6H, COCH2N + NCH2COOH), 3.54
(s, 4H, NCH2COOH), 3.21 (d, 4H, SCH2CH), 2.86 (m, 8H,
NCH2CH2N), 1.01 [d, 12H, CH(CH3)2]. GCIC: yield = 71%;
Mn = 16.6 kDa, Mw = 25.0 kDa. The Gd content was 162 mg
Gd/g.
Synthesis of Biodegradable Paramagnetic Copolymers
The general procedure for the synthesis of Gd-DTPA Relaxivity Measurement of Gd-Containing Copolymers
L-cystine bisamide and substituted bisamide copolymers is
described below using L-cystine bisisopropyl amide as an
Longitudinal relaxivity (r1) was determined by measur-
example. L-Cystine bisisopropyl amide (0.815 g, 2.53 mmol) ing the T1 relaxation times for three different concentrations
was dissolved in 0.30 mL triethylamine (TEA) and 0.70 mL of GCAC (22.3 kDa), GCPC (22.3 kDa), and GCIC (25.0
of anhydrous dimethyl sulfoxide (DMSO) while stirring in an kDa), as well as a water reference on a Siemens Trio 3T MRI
ice-water bath at 5-C. DTPA dianhydride (0.904 g, 2.53 Scanner at room temperature using an inversion recovery
mmol) was added in portions over a 25 min period. The prepared turbo spin-echo imaging pulse sequence. The
reaction mixture turned brown and viscous and then solidi- inversion times were as follows: inversion time (TI) = 24,
fied after 40 min. The mixture was removed from the ice- 50, 75, 100, 150, 200, 300, 400, 600, 800, 1200, and 1600 ms;
water bath and allowed to settle to room temperature. An repetition time (TR) = 5000 ms; echo time (TE) = 15.0 ms.

1288
Kaneshiro et al.
The net magnetization amplitude data from the region of Images were obtained on a wrist coil using a 3D FLASH pulse
interest (ROI) of each sample were fit using a Marquardt- sequence. The imaging parameters were TE = 1.64 ms, TR =
Levenberg algorithm for multiparametric nonlinear regres- 4.30 ms, 19- flip angle, 3D acquisition with 80 slices/slab,
sion analysis (20). T1 and M0 were calculated from these data and 0.5-mm coronal slice thickness. In vivo contrast en-
and r1 was determined from the slope of 1/T1 vs. [Gd3+] plot. hancement in the major organs and tissues, including the
Transverse relaxivity (r2) was determined by measuring the heart, kidneys, liver, bladder, and muscle periphery, was
T2 relaxation times for four different concentrations of analyzed using Osirix software package (http://homepage.-
GCAC, GCPC, and GCIC, as well as a ultrapure H2O mac.com/rossetantoine/osirix/). Signal intensities were mea-
reference on a Siemens Trio 3T MRI scanner at room sured in different ROIs at each time point and were averaged
temperature using a turbo spin-echo imaging sequence with from three different Sprague<Dawley rats. A one-way
turbo factor 3. The following times were as follows: TE = 12, ANOVA statistical analysis was performed to determine
24, 35, 47, 59, 71, 83, 94, and 106 ms; TR = 3000 ms. T2 values statistical significance (p < 0.05) of the differences in the
were calculated on a semilog plot of the net magnetization signal intensities of the contrast agents in the ROIs.
amplitude data from the ROI; r2 was determined from the
slope of 1/T2 vs. [Gd3+] plot.
RESULTS
In Vitro Degradation of Copolymers
Synthesis of Gd-DTPA L-Cystine Bisamide Copolymers
The degradation of each copolymer, with and without
Gd(III), was studied in the presence of L-cysteine at 15 mM,
Three new polydisulfide Gd(III) chelates with different
which is the free thiol concentration in the plasma. The amide substituents at the cystine carboxylic groups were
copolymers (420 mM, monomeric repeat unit concentration) synthesized as novel biodegradable macromolecular MRI
and 15 mM L-cysteine in phosphate-buffered saline (PBS; pH contrast agents. Alkyl-substituted cystine amides, L-cystine
7.4) were incubated at 37-C. The molecular weight distribu- bispropyl amide and L-cystine bisisopropyl amide, were first
tions were immediately evaluated by SEC. The molecular prepared by the reaction of (t-BOC-Cys)2 bis(N-hydroxysuc-
weights of the remaining copolymers were calculated using a cinimide) ester with the alkyl amines, followed by t-BOC
time window, which was from the time point when the deprotection with TFA (Fig. 1). The monomers were
copolymers first appeared to 36 mins, when the monomeric extensively purified to ensure high purity because monomer
degradation products began to elute. The molecular weight purity was crucial for the molecular weight of the copolymers.
reduction of copolymers in the incubation was calculated as
Figure 2 shows the synthetic procedure for paramagnetic
the percentage of the weight-average molecular weight of the copolymers of Gd-DTPA and cystine bisamides. Macromol-
remaining copolymers at various time points of the incuba- ecular ligands, DCAC, DCPC, and DCIC, were first synthe-
tion to their original values.
sized by condensation copolymerization of equimolar
amounts of DTPA dianhydride and corresponding L-cystine
bisamides. The reaction was performed in a mixture of TEA
and anhydrous DMSO. TEA acted as a base to neutralize the
In Vivo Metabolism
GCAC, GCPC, and GCIC with average molecular salts of the monomers and as a solvent to increase the
weights (Mw) of 22.3, 22.3, and 25.0 kDa, respectively, were solubility of the copolymers. The macromolecular Gd(III)
prepared with similar methods described above for in vivo chelates were prepared in relatively high yield by complex-
metabolism and MRI studies. The agents were intravenously ation of the polymeric ligands with Gd(OAc)3. The molec-
injected into male Sprague<Dawley rats (200Y260 g) at a dose ular weight of the paramagnetic Gd(III) chelates can be as
of 0.1 mmol Gd/kg body weight via the tail vein. The rats high as 55 kDa. The molecular weights, Gd content, and T1
were kept in metabolic cages, and their urine samples were and T2 relaxivities of the paramagnetic copolymers used in
collected at 8 and 24 h postinjection; samples were analyzed this study are listed in Table I.
with both positive- and negative-charge-labeled MALDI-
After complexation with Gd3+, there was a significant
TOF mass spectrometry on an a-cyano-4-hydroxycinnamic decrease in the apparent molecular weights of the
acid matrix.
copolymers (Table I). The polymeric ligands were anionic
and had a large hydrodynamic volume as a result of the free
carboxylic groups of DTPA. The complexation with Gd3+ ions
formed neutral macromolecular Gd(III) chelates and resulted
Contrast-Enhanced Magnetic Resonance Imaging in Rats
In vivo magnetic resonance imaging (MRI) contrast in a significant reduction of the copolymers’ hydrodynamic
enhancement of the agents was evaluated in male Sprague< volume. The apparent molecular weights of the
Dawley rats (200Y260 g). The animals were cared for under macromolecular Gd(III) chelates, GCAC, GCPC, and
an approved protocol by the University of Utah Institutional GCIC, decreased by approximately 30% as compared with
Animal Care and Use Committee. The rats were anesthetized the average molecular weight of the corresponding polymeric
by intraperitoneal administration of xylazine (12 mg/kg) and ligands.
ketamine (80 mg/kg). The contrast agents were administered
The Gd content of GCAC, GCPC, and GCIC was 16.9%,
at a dose of 0.1 mmol Gd/kg via intravenous tail vein injection; 16.9%, and 16.2%, respectively, as calculated by ICP-OES.
each agent was investigated in a group of three rats. MR Elemental metal analysis showed that the Gd content for the
images were obtained on a Siemens Trio 3 T MRI scanner polymeric chelates was lower than the calculated content,
before injection and at 2, 5, 10, 15, and 30 min postinjection. which was the theoretical Gd composition calculated from the

Gd-DTPA L-Cystine Bisamide Copolymers
1289
Fig. 1. Synthetic scheme of cystine bispropyl amide and L-cystine bisisopropyl amide.
structure of the repeat units of the copolymers. The low SD) shows the percentage of the molecular weight reduction
measured Gd content might be attributed to the association of the remaining polymeric Gd(III) chelates at various time
of water molecules to the hydrophilic polymers. The Gd points in the incubation with 15 mM cysteine at 37-C. The
cystine bisamide copolymers were stable in the solid state, degradation rate of the polymeric chelates decreased in the
and their molecular weight distributions did not change after order of GCAC, GCPC, and GCIC as the steric hindrance
approximately 6 months in cold storage.
around the disulfide bonds increased. All macromolecular
Gd(III) chelates completely degraded into low molecular
weight oligomers after 75 min.
T1 and T2 Relaxivities of the Paramagnetic Copolymers
The degradation products, as a result of the disulfideY
The T1 relaxivities of the macromolecular Gd(III) thiol exchange reaction with cysteine, were also identified in
chelates GCAC, GCPC, and GCIC at 3 T were 4.37, 5.28, the MALDI-TOF mass spectra of the incubation mixtures,
and 5.56 mMj1 sj1, respectively. These values were similar
to the clinical contrast agent Gd-(DTPA-BMA) (4.62 mMj1
sj1) and the previously reported GDCP (5.43 mMj1 sj1) and
GDCEP (5.86 mMj1 sj1) at 3 T (14). The T2 relaxivities of
GCAC, GCPC, and GCIC at 3 T were 6.21, 6.08, and 6.54
mMj1 sj1, respectively, and were similar to our previously
reported GDCC (13). The relaxivities of the agents were
lower than those of other reported macromolecular Gd(III)
chelates because of the local flexibility of the disulfide bonds.
In Vitro Degradation of the Copolymers
Figure 3 (N = 3, mean T SD) shows the molecular weight
reduction of the remaining polymeric ligands from their
original values at various time points in the incubation with
15 mM cysteine in PBS at 37-C. The disulfideYthiol exchange
reaction between the polydisulfide backbone and cysteine
resulted in molecular weight decreases with increasing
incubation time. The rates of molecular weight reduction
varied among the copolymers. DCIC with bulky isopropyl
groups degraded much slower than DCPC and DCAC. All
polymeric ligands completely degraded into low molecular
weight species within 24 h. The steric effect around the
disulfide bonds significantly affected the degradation rate of
the copolymers in the presence of cysteine at the endogenous
free thiol concentration.
The polymeric Gd(III) chelates were more susceptible to
the disulfideYthiol exchange reaction with cysteine than the
corresponding polymeric ligands. Figure 4 (N = 3, mean T
Fig. 2. Synthetic scheme for the Gd-DTPA L-cystine bisamide
copolymers (GCAC) and alkyl-substituted GCAC.

1290
Kaneshiro et al.
Table I. Physicochemical Parameters of the Copolymers
Mn
(kDa)
Mw
(kDa)
Gd content
(calculated)
Copolymers
r1 (mMj1 sj1
)
r2 (mMj1 sj1
)
Yield (%)
DCAC
GCAC
DCPC
GCPC
DCIC
GCIC
30.8
14.1
24.1
17.1
26.4
16.6
34.5
22.3
33.2
22.3
36.7
25.0
Y
16.9% (20.4%)
Y
4.37
Y
5.28
Y
Y
6.21
Y
6.08
Y
67.5
81.0
60.2
73.0
60.0
71.0
Y
16.9% (18.4%)
Y
16.2% (18.4%)
5.56
6.53
Apparent Mn and Mw were determined by size exclusion chromatography, whereas Gd3+ content was determined by inductively coupled
argon plasma optical emission spectrometer analysis, longitudinal and transverse relaxivities (determined at 3 T), and reaction yield.
DCAC = DTPA L-cystine bisamide copolymers; GCAC = Gd-DTPA L-cystine bisamide copolymers; DCPC = DTPA L-cystine bispropyl
amide copolymers; GCPC = Gd-DTPA L-cystine bispropyl amide copolymers; DCIC = DTPA L-cystine bisisopropyl amide copolymers and
GCIC = Gd-DTPA L-cystine bisisopropyl amide copolymers.
which verified the degradation reaction mechanism (Fig. 5). agents. This metabolite was also observed in the urine
The degradation products had molecular weights that were samples of the other polydisulfide agents (14). For GCIC,
indicative of a single repeat unit, Gd-DTPA bis(cysteine the peak with a mass (m/z) of approximately 1589.36
amide), and up to three repeat units. The mass of degrada- corresponds to two GCIC monomeric repeat units with two
tion products with one or two repeat units for GCAC was of its isopropyl amide groups hydrolyzed [834.18 ꢀ 2 j 2
around 753.0 or 1501.2 (m/z) as observed in the mass C3H8N (58.07) + 2 OH + 2 H + H+], and the peak at 1707.29
spectrum (not shown). The in vitro degradation products of (m/z) corresponds to two GCIC monomeric repeat units plus
GCPC and GCIC had the same masses, and the degradation a K+ ion (834.18 ꢀ 2 + K+). These additional dimeric
products with one, two, and three repeat units were identified degradation products have been observed for GCIC probably
with the mass of 835.2, 1668.4, and 2502.6 (m/z), respectively. because the steric hindrance of the isopropyl groups around
the disulfide spacer slows polymer degradation. Further inv-
In Vivo Metabolism in Rats
estigation of the unknown metabolites for both GCIC and
GCPC is planned for future studies.
Metabolites were also identified in the negative charge
Catabolic degradation products were identified by
MALDI-TOF mass spectrometry in the urine samples labeled mass spectra of the 8 h urine samples with the agents.
collected 8 h postinjection with all three agents. No major Both GCIC (Fig. 6b) and GCPC (spectrum not included) had
metabolite was found in the 24 h urine samples. Figure 6 major metabolites with similar masses (m/z) at 589.04,
shows the positive- and negative-charge-labeled mass spectra 734.08, and 911.04. The masses (m/z) of the major metabo-
of the 8-h urine samples of GCIC. The major metabolites for lites of GCAC (spectrum not included) were approximately
GCIC had a mass (m/z) of approximately 718.26, 875.15, 589.06, 707.09, and 734.10, which were all observed in our
913.11, 988.02, 1589.36, and 1707.29 at 8 h postinjection in the previous study on GDCP and GDCEP (14). Two unknown
positive-charge-labeled mass spectrum (Fig. 6a). Three metabolites were observed for GCIC, 531.97 (m/z) and
unknown peaks were present with masses (m/z) of 895.09,
1215.02, and 1397.51. Less metabolites were identified in the
positive-charge-labeled mass spectra of GCPC and GCAC
(spectra not shown). The major metabolites for GCPC had a
mass (m/z) of 718.25, 875.15, 913.12, and 988.02, whereas only
one major metabolite with a mass (m/z) of 988.02 was
identified for GCAC. One unknown peak with a mass (m/z)
of 1214.04 was observed for GCPC.
The Gd isotope pattern was observed in every major peak
except for a mass (m/z) at 718 for GCIC and GCPC. The
peaks at approximately 718 for both GCPC and GCIC
correspond to the ligands (679.26 + K+) of the repeat units
in these agents. The compounds might be formed by the
dissociation of Gd3+ ions from the chelates at low pH in
the process of sample preparation for mass spectrometry.
a-Cyano-4-hydroxycinnamic acid was used as the MALDI-
TOF matrix, and the pH of the samples was approximately
1.2.
The peaks around 875.15 (m/z) and 913.11 (m/z) for
Fig. 3. Percent molecular weight reduction of the remaining
GCIC and GCPC represent their monomeric repeat units
with a K+ ion (834.18 + 2H + K+) and two K+ ions (834.18 +
H + 2K+), respectively. The metabolite with a mass 988.02
polymeric ligands from their original values in 15 mM cysteine
[phosphate-buffered saline (PBS) and 37-C] at varying incubation
time (in minutes). Mw was determined by size exclusion chromatog-
(m/z) was also observed in the mass spectra of all three raphy (SEC) (N = 3, mean T SD).

Gd-DTPA L-Cystine Bisamide Copolymers
1291
were readily degraded when incubated with 15 mM cysteine
under physiological conditions (Fig. 4).
The degradation rates of the polymeric ligands were
slower than the macromolecular Gd(III) chelates as a result
of the ligands’ negative charges; negative charges in the
polymeric ligands may inhibit the reaction of disulfide bonds
with negatively charged cysteine as a result of charge
repulsion at pH 7.4 (21). The alkyl substituents had a similar
steric effect on the degradation of the polymeric Gd(III)
chelates as the ligands. GCIC, which contain isopropyl
groups, degraded slower than GCPC and GCAC when
incubated with cysteine (Fig. 4). The degradation rate of
the Gd-DTPA copolymers with cystine bisamide and substi-
tuted bisamides was also much faster than the previously
reported GDCC, GDCP, and GDCEP under the same
reaction conditions (14). Apparently, substituting the carbox-
ylic groups in L-cystine with amide groups renders the
disulfide bonds more susceptible to degradation.
The identification of catabolic degradation products in rat
urine samples demonstrated that the novel GCAC were
cleaved into smaller Gd(III) chelate moieties, which were
readily excreted via renal filtration. Analyses of degradation
products within urine samples suggest that in vivo degradation
processes are more complicated than in vitro disulfideYthiol
exchange reactions. The complicated MALDI-TOF spectra
of the in vivo metabolism of GCAC, GCPC, and GCIC may
be the result of other degradative processes [e.g., enzymatic
(22) reactions and oxidation] within the blood. Although
there were a few unknown peaks, the mass of these
catabolites is above the Gd-DTPA chelate (584.04 Da),
indicating that the Gd(III)-DTPA complexes were intact
and stable in the plasma. The absence of major metabolic
peaks at 24 h postinjection for both the positive- and
Fig. 4. Percent molecular weight reduction of the remaining macro-
molecular Gd(III) complexes from their original values in 15 mM
cysteine (PBS and 37-C) at varying incubation time (in minutes). Mw
was determined by SEC (N = 3, mean T SD).
631.03 (m/z), and one unknown metabolite was found in
GCPC, 531.95 (m/z).
The metabolite with the mass (m/z) of 911 for GCIC and
GCPC could be the same metabolite with the mass (m/z) of
913 identified in the positive-charge-labeled mass spectra of
the agents. The structures of other metabolites are unknown,
and further study is on going for accurate characterization of
these metabolites. Furthermore, no metabolites were ob-
served in the negative charge labeled mass spectra for all
three agents at 24 h postinjection.
In Vivo Contrast Enhancement in Rats
Figure 7 shows the 2D coronal rat MR images cross the
heart and liver before and at 2, 5, 10, 15, and 30 min after
injection of the agents at a dose of 0.1 mmol Gd/kg. Strong
contrast enhancement was observed within the heart and
blood vessels, differentiating these structures from the
surrounding tissues at 2 min postinjection. The contrast
enhancement then gradually decreased but was still visible
at 30 min postinjection. Significant contrast enhancement was
also observed in the liver and kidneys, with similar dynamic
changes as observed in the heart. Contrast enhancement in
the urinary bladder gradually increased over time, indicating
the accumulation of low molecular weight Gd(III) chelates.
Statistical analysis of MR signal intensities revealed no
significant difference (p > 0.05) between the contrast
enhancement in the organs and tissues (heart, liver, kidney,
bladder, and muscle periphery) for GCAC, GCPC, and
GCIC at all time points post injection.
DISCUSSION
Biodegradable, macromolecular MRI contrast agents
were designed and prepared to facilitate the excretion of
low molecular weight Gd(III) chelates after the polydisulfide
backbone was cleaved via thiolYdisulfide exchange reactions
with the endogenous thiols in the body. These novel agents
Fig. 5. Degradation reaction of GCAC with cysteine. In vitro
degradation products were identified by matrix-assisted laser desorp-
tion ionization time of flight (MALDI-TOF) analysis.

1292
Kaneshiro et al.
Fig. 6. Positive-charge-labeled (a, top) and negative-charge-labeled (b, bottom) MALDI-TOF spectra of rat urine
samples collected at 8 h postinjection of Gd-DTPA cystine bisisopropyl amide copolymers (GCIC) at 0.10 mmol Gd/kg.

Gd-DTPA L-Cystine Bisamide Copolymers
1293
Fig. 7. 2D coronal magnetic resonance images at precontrast and 2, 5, 10, 15, and 30 min postinjection of
GCAC (A), GCPC (B), and GCIC (C).
negative-charge-labeled MALDI-TOF spectra indicates that
These newly synthesized agents have a relatively fast
the majority of the macromolecular Gd(III) chelates were degradation rate and short enhancement window for vascular
excreted and cleared from the body within a short time imaging as compared with our previously prepared MRI
period. These results were similar to the clearance of GDCC, contrast agents, GDCP and PEG-g-GDCP (14,15,17). The
which showed that most of the agents were excreted in the steric effect of the alkyl groups was not observed during in
urine sample within the first 8 h after injection (16).
vivo MRI because of their rapid degradation. On the other
All three biodegradable macromolecular agents of hand, GCAC, GCPC, and GCIC are neutral and are
similar molecular weights resulted in more significant con- expected to have lower osmolality than the negatively
trast enhancement in the blood pool of rats than a low charged GDCP and PEG-g-GDCP. Therefore, these neutral
molecular weight clinical agent, Gd-(DTPA-BMA), which copolymers show potential as MRI contrast agents for
was reported previously (16). The heart, liver, kidneys, and patients with a history of renal malfunction.
blood vessels were clearly visualized with these new contrast
agents. As shown in the dynamic contrast-enhanced MR
images, the enhancement in the heart and liver was still CONCLUSION
visible at 30 min postinjection (Fig. 7). The enhancement
window should be long enough for most clinical MRI
Biodegradable macromolecular Gd(III) chelates, (Gd-
procedures of cancer imaging and angiography. All three DTPA) cystine bisamide copolymers, can be prepared in
agents resulted in similar enhancement kinetics in MR blood relatively high molecular weight. The paramagnetic copoly-
pool imaging in rats. Although the steric effect in the Gd- mers rapidly degraded into low molecular weight species in
DTPA cystamine bisamide copolymers resulted in slight the presence of cysteine via the disulfideYthiol exchange
differences in the in vitro degradation rates among the reaction. The in vitro degradation rate was reduced with the
agents, they all had similar enhancement kinetics. This is substitution of bulky alkyl groups in cystine bisamides. These
attributed to their rapid degradation, which was not able to agents were readily degraded and excreted via renal filtra-
differentiate the effect of the structural differences on in vivo tion, which was confirmed by the detection of catabolic
contrast enhancement. A similar phenomenon was also products in rat urine samples. The agents resulted in
observed for another rapid degrading polydisulfide agent, significant MRI contrast enhancement of the heart, vascula-
GDCC. GDCC with a high molecular weight (60 kDa) and a ture, liver, and kidneys in rats. No significant difference was
low molecular weight (18 kDa) resulted in similar enhance- observed in the dynamic contrast enhancement among the
ment kinetics in rats (16).
macromolecular Gd(III) chelates. These novel agents are

1294
Kaneshiro et al.
(propylene imine) dendrimers for magnetic resonance imaging.
Macromolecules 37:3084Y3091 (2004).
promising safe and effective, biodegradable macromolecular
MRI contrast agents for blood pool imaging, including cancer
imaging and angiography.
10. U. Schmiedl, M. Ogan, H. Paajanen, M. Marotti, L. E. Crooks,
A. C. Brito, and R. C. Brasch. Albumin labeled with Gd-DTPA
as an intravascular, blood pool-enhancing agent for MR imaging:
biodistribution and imaging studies. Radiology 162:205Y210
(1987).
ACKNOWLEDGMENTS
11. E. Toth, L. Helm, K. E. Kellar, and A. E. Merbach. Gd(DTPA-
bisamide)alkyl copolymers: a hint for very high relaxivity MRI
contrast agents. Chem. Eur. J. 5:1202Y1211 (1999).
12. R. Rebizak, M. Schaefer, and E. Dellacherie. Macromolecular
contrast agents for magnetic resonance imaging: influence of
polymer content in ligand on the paramagnetic properties. Eur.
J. Pharm. Sci. 7:243Y248 (1998).
This research was supported in part by the NIH grants
R01 EB00489 and R33 CA095873. We greatly appreciate Dr.
Yongen Sun and Ms. Melody Johnson for their technical
help.
13. Z.-R. Lu, D. L. Parker, K. C. Goodrich, X. Wang, J. G. Dalle,
and H. R. Buswell. Extracellular biodegradable macromolecular
gadolinium(III) complexes for MRI. Magn. Reson. Med.
51:27Y34 (2004).
14. Y. Zong, X. Wang, K. G. Goodrich, A. M. Mohs, D. L. Parker,
and Z.-R. Lu. Contrast-enhanced MRI with new biodegradable
macromolecular Gd(III) complexes in tumor-bearing mice.
Magn. Reson. Med. 53:835Y842 (2005).
REFERENCES
1. Z.-P. Liang and P. C. Lauterbur. Principles of Magnetic
Resonance Imaging, IEEE Press, New York, 1999.
2. M. T. Vlaardingerbroek and J. A. den Boer. Magnetic Reso-
nance Imaging, 3rd ed., Springer, Berlin Heidelberg New York,
2003.
3. P. Caravan, J. J. Ellison, T. J. McMurry, and R. B. Lauffer. 15. A. M. Mohs, X. Wang, K. C. Goodrich, Y. Zong, D. L. Parker,
Gadolinium(III) chelates as MRI contrast agents: structure,
dynamics, and applications. Chem. Rev. 20:2293Y2352 (1999).
´
4. A. E. Merbach and E. To´th. The Chemistry of Contrast Agents in
and Z.-R. Lu. PEG-g-poly(GdDTPA-co-L-cystine): a biodegrad-
able macromolecular blood pool contrast agent for MR imaging.
Bioconj. Chem. 15:1424Y1430 (2004).
Medical Magnetic Resonance Imaging, Wiley, New York, 2001.
5. G. Schuhmann-Giampieri, H. Schmitt-Willich, T. Franzel, W.-R.
Press, and H.-J. Weinmann. In vivo and in vitro evaluation of
16. X. Wang, Y. Feng, T. Ke, M. Schabel, and Z.-R. Lu. Pharmaco-
kinetics and tissue retention of (Gd-DTPA)-cystamine copolymers,
a biodegradable macromolecular magnetic resonance imaging
contrast agent. Pharm. Res. 22:596Y602 (2005).
Gd-DTPA-polylysine as
a macromolecular contrast agent
for magnetic resonance imaging. Invest. Radiol. 26:969Y974 17. A. M. Mohs, Y. Zong, J. Guo, D. L. Parker, and Z.-R. Lu. PEG-
(1991).
g-Poly(GdDTPA-co-L-cystine): effect of PEG chain length on
in vivo contrast enhancement in MRI. Biomacromolecules
6:2305Y2311 (2005).
6. D. L. Ladd, R. Hollister, X. Peng, D. Wei, G. Wu, D. Delecki,
R. A. Snow, J. L. Toner, K. Kellar, J. Eck, V. C. Desai,
G. Raymond, L. B. Kinter, T. S. Desser, and D. L. Rubin.
Polymeric gadolinium chelate magnetic resonance imaging con-
18. M. Bodanszky, A. Bodanszky. The Practice of Peptide Synthesis,
Springer, Berlin Heidelberg New York, 1985.
trast agents: design, synthesis, and properties. Bioconj. Chem. 19. D. J. Hnatowich, B. Friedman, B. Clancy, and M. Novak.
Labeling of preformed liposomes with Ga-67 and Tc-99 m by
10:361Y370 (1999).
7. R. Weissleder, A. Bogdanov, C. H. Tung, and H.-J. Weinmann.
chelation. J. Nucl. Med. 22:810Y814 (1981).
Size optimization of synthetic graft copolymers for in vivo 20. B. Fichtlscherer and A. Mulsch. MR imaging of nitrosylYiron
angiogenesis imaging. Bioconj. Chem. 12:213Y219 (2001).
8. H. Kobayashi, S. Kawamoto, S.-K. Jo, H.-L. Bryant, W. W.
complexes: experimental study in rats. Radiology 216:225Y231
(2000).
Brechbiel, and R. A. Star. Macromolecular MRI contrast agents 21. D. J. Hupe and D. Wu. Effect of charged substituents on rates of
with small dendrimers: pharmacokinetic differences between
the thiol Y disulfide interchange reaction. J. Org. Chem.
45:3100Y3103 (1980).
9. S. Langereis, Q. G. D. Lussanet, M. H. P. Genderen, W. H. 22. S. E. Moriarty-Craige and D. P. Jones. Extracellular thiols and
sizes and cores. Bioconj. Chem. 14:388Y394 (2003).
Backes, and E. W. Meijer. Multivalent contrast agents based on
gadolinium-diethylenetriaminepentaacetic acid-terminated poly
thiol/disulfide redox in metabolism. Annu. Rev. Nutr. 24:481Y509
(2004).