Synthesis and evaluation of a monoreactive DOTA derivative for indium-111-based residualizing label to estimate protein pharmacokinetics

Article abstract of DOI:10.1211/002235702320266226

The purpose of this study was to develop an indium-111 (¹¹¹In)-based residualizing label for estimating the pharmacokinetics of proteins. 1,4,7,10-Tetraazacyclododecane-N,N′,N″,N?-tetraacetic acid (DOTA), which produced a highly stable and hydrophilic ¹¹¹In chelate, was selected as the chelating site, and the monoreactive DOTA derivative with a tetrafluorophenyl group as the protein binding site (mDOTA) was designed to avoid cross-linkings of proteins. mDOTA was synthesized with an overall yield of 11%. The stability in murine plasma, the radioactivity retention in the catabolic sites of proteins and the radiochemical yields of ¹¹¹In-labelled proteins via mDOTA were investigated using human serum albumin (HSA), galactosyl-neoglycoalbumin (NGA) and cytochrome c (cyt c) as model proteins. ¹¹¹In-labelled HSA via mDOTA was highly stable for 5 days after incubation in murine plasma. Long retention of radioactivity in the catabolic sites was observed after injection of ¹¹¹In-DOTA-NGA in mice, due to the slow elimination of the radiometabolite from the lysosome. At a chelator concentration of 42.2 μM, ¹¹¹In-DOTA-cyt c was produced with over 91 % radiochemical yield. On the other hand, ¹¹¹In-DOTA-lysine and ¹¹¹In-DOTA were obtained with high radiochemical yields at lower chelator concentrations. These findings indicated that mDOTA would be an appropriate ¹¹¹In-labelling agent for estimating protein pharmacokinetics. These findings also suggested that the introduction of a protein binding site at a position distal from the unmodified DOTA structure would be preferable to preparing ¹¹¹In-DOTA-labelled proteins with higher specific activity.

Full text of DOI:10.1211/002235702320266226

Journal of
JPP 2002, 54: 1073–1081
# 2002 The Authors
Received February 20, 2002
Accepted May 9, 2002
ISSN 0022-3573
Synthesis and evaluation of a monoreactive DOTA
derivative for indium-111-based residualizing label to
estimate protein pharmacokinetics
Takahiro Mukai, Shinji Namba, Yasushi Arano, Masahiro Ono, Yasushi
Fujioka, Tomoya Uehara, Kazuma Ogawa, Junji Konishi and Hideo Saji
Abstract
The purpose of this study was to develop an indium-111 (111In)-based residualizing label for
estimating the pharmacokinetics of proteins. 1,4,7,10-Tetraazacyclododecane-N,N«,N««,N«««-tetraacetic
acid (DOTA), which produced a highly stable and hydrophilic 111In chelate, was selected as the
chelating site, and the monoreactive DOTA derivative with a tetrauorophenyl group as the protein
binding site (mDOTA) was designed to avoid cross-linkings of proteins. mDOTA was synthesized with
an overall yield of 11%. The stability in murine plasma, the radioactivity retention in the catabolic
sites of proteins and the radiochemical yields of 111In-labelled proteins via mDOTA were investigated
using human serum albumin (HSA), galactosyl-neoglycoalbumin (NGA) and cytochrome c (cyt c) as
model proteins. 111In-labelled HSA via mDOTA was highly stable for 5 days after incubation in murine
plasma. Long retention of radioactivity in the catabolic sites was observed after injection of 111In-
DOTA-NGA in mice, due to the slow elimination of the radiometabolite from the lysosome. At a
Department of Nuclear Medicine
and Diagnostic Imaging,
Graduate School of Medicine,
Kyoto University, Shogoin
Kawahara-cho, Sakyo-ku, Kyoto
606-8507, Japan
chelator concentration of 42.2
,
111In-DOTA-cyt c was produced with over 91% radiochemical yield.
M
On the other hand, 111In-DOTA-lysine and 111In-DOTA were obtained with high radiochemical yields
at lower chelator concentrations. These Ž ndings indicated that mDOTA would be an appropriate
111In-labelling agent for estimating protein pharmacokinetics. These Ž ndings also suggested that the
introduction of a protein binding site at a position distal from the unmodiŽ ed DOTA structure would
be preferable to preparing 111In-DOTA-labelled proteins with higher speciŽ c activity.
Takahiro Mukai, Junji Konishi
Department of Patho-Functional
Bioanalysis, Graduate School of
Pharmaceutical Sciences, Kyoto
University, Yoshida
Shimoadachi-cho, Sakyo-ku,
Kyoto 606-8501, Japan
Introduction
Shinji Namba, Masahiro Ono,
Yasushi Fujioka, Kazuma
Ogawa, Hideo Saji
To assess the pharmaceutical applications of newly developed or chemically modi®ed
proteins, their in-vivo fate has to be investigated and evaluated quantitatively.
Radiolabelling techniques have been used widely for these purposes because of their
high sensitivity. In addition, single-photon emission computed tomography (SPECT)
has become available to pursue the real-time localization of the radioactivity in living
small laboratory animals (Weber et al 1994 ; Ishizu et al 1995). This procedure enables
a real-time trace of the in-vivo behaviour of proteins in an animal by external
measurement of the radioactivity. For the accurate estimation of protein pharma-
cokinetics, long residence times of the radioactivity at the catabolic sites are required.
14C or 125I labels with saccharide units have been developed for these purposes due to the
generation of radiometabolites with limited ability to diŒuse through the lysosomal
membrane (Pittman et al 1979 ; Thorpe et al 1993). These radiolabelling reagents are
called residualizing labels.
Department of Molecular
Imaging and Radiotherapy,
Graduate School of
Pharmaceutical Sciences, Chiba
University, 1-33 Yayoi-cho,
Inage-ku, Chiba 263-8522, Japan
Yasushi Arano, Tomoya Uehara
Correspondence: T. Mukai,
Department of Patho-Functional
Bioanalysis, Graduate School of
Pharmaceutical Sciences, Kyoto
University, Yoshida
Shimoadachi-cho, Sakyo-ku,
Kyoto 606-8501, Japan. E-mail:
tmukai!kuhp.kyoto-u.ac.jp
Recently, a metallic radionuclide, indium-111 (111In) was used as a residualizing label
after the reaction of proteins with bifunctional chelating agents (BCAs) (Mukai et al
1999a ; Nishikawa et al 1999). 111In emits gamma rays of adequate energies (173 and
247 keV) suitable for both radioactivity measuring with a gamma counter and external
Acknowledgement: We are
grateful to Nihon Medi-Physics
Co. Ltd, Takarazuka, Japan for
its kind gift of 111InCl3.
.
imaging using SPECT for small animals. The physical half-life (2 8 days) of this
radionuclide is su cient for pursuing the pharmacokinetics of proteins with a long
plasma half-life but not too long for the disposal of radioactive materials. In addition,
1073

1074
Takahiro Mukai et al
111In-labelled proteins are prepared with high radiochemical ment mass spectra (FAB-MS) were obtained with a JMS-
yields by a simple addition of 111In solution to the conjugate HX HX 110 A model (JEOL Ltd, Tokyo, Japan). Size-
solution, because of the rapid and high yield complexation exclusion HPLC (SE-HPLC) was performed using a 5
}
reaction between chelator-conjugated proteins and 111In Diol-300 column (7 5 600 mm, Nacalai Tesque Inc.,
. ¬
(Hnatowich et al 1983). Cyclic diethylenetriaminepenta- Kyoto, Japan), eluted with 0 1 phosphate buŒer (pH 6 8)
. m
.
acetic anhydride (cDTPA) is used as the BCA to label at a ¯ ow rate of 1 0 mL min . Cellulose acetate electro-
1
-
.
proteins with 111In, due to its simple conjugation reaction phoresis (CAE) strips were run in veronal buŒer (pH 8 6,
.
I
0 06) at a constant current of 0 8 mA for 40 min. TLC
¯ .
.
with proteins and its ready availability from commercial
sources. Previous studies suggested that 111In-labelled prot- analyses were performed with silica plates (Merck Art
eins via cDTPA exhibited relatively long residence of the 5553) with 10% aqueous ammonium acetate±methanol (3:
111
1
-
radioactivity in the catabolic sites (Mukai et al 1999a ; 1) as the developing solvent. InCl3 (74 MBq mL ) was
Nishikawa et al 1999). However, formation of inter- and supplied by Nihon Medi-Physics (Takarazuka, Japan).
intramolecular cross-linkings is unavoidable during the HSA and cyt c were purchased from Sigma Chemical Co.
conjugation reaction since cDTPA possesses two binding (St Louis, MO). NGA (44 galactoses per HSA) was pre-
sites for proteins (Maisano et al 1992 ; Reilly et al 1992 ; pared by the method reported previously (Mukai et al
Arano et al 1996), which cause gradual release of 111In from 1999b). cDTPA was obtained from Dojindo Laboratories
DTPA-conjugated proteins to transferrin while circulating (Kumamoto, Japan). All other chemicals were of reagent
in plasma (Arano et al 1996).
The purpose of this study was to develop a new 111In-
labelling agent suitable for estimating protein pharmacoki-
grade.
N N« N«« N«««
-te-
netics. 1,4,7,10-Tetraazacyclododecane-
,
,
,
Synthesis of mDOTA
traacetic acid (DOTA), which produced a highly stable and
hydrophilic 111In chelate (Wu et al 1992), was selected as the
chelating site, and the monoreactive DOTA derivative with
a tetra¯ uorophenyl group as the protein binding site
(mDOTA) was designed and synthesized to avoid cross-
linking of proteins. The plasma stability, radioactivity
retention in the catabolic site and radiochemical yields of
111In-DOTA-proteins wereinvestigated using human serum
albumin (HSA), galactosyl-neoglycoalbumin (NGA) and
cytochrome c (cyt c), respectively, as model proteins. For
estimation of plasma stability of 111In-labelled proteins via
mDOTA, the parental proteins should be stable in plasma
so as not to generate any radiometabolites derived from
parental proteins. Thus, we used a typical serum protein,
HSA, as a model. NGA is a useful polypeptide to estimate
the behaviour of radiometabolites generated after lyso-
somal proteolysis in hepatocytes (Arano et al 1994a, 1995 ;
Mukai et al 1999a, b), and so this protein was used for
evaluation of the residence time of the radioactivity derived
from 111In-DOTA labels in the catabolic site. In addition,
the number of chelators attached per molecule of protein
plays a critical role in radiochemical yields with 111In. Since
the exact number of chelators attached to cyt c was assessed
by mass spectrometry (Lewis et al 1994), this protein was
1,4,7,10-Tetraazacyclododecane-
«
N,N ,N
««-triacetic acid
tri-
To Cyclem (1 66 g, 9 64 mmol) (Figure 1) dissolved in
-butyl ester (1)
tert
.
.
N N
, -diisopropylethylamine
-butylbromoacetate(4 24 mL,
acetonitrile was added
tert
.
.
.
(5 04 mL,28 9 mmol)and
.
28 9 mmol), successively, while keeping the reaction tem-
perature below 0 C. After 13 h stirring at room tempera-
°
ture, the solution was concentrated in-vacuo. Chloroform
was added to the residue, and the organic phase was washed
with saturated aqueous NaCl. After drying the organic
phase over CaSO4, the solution was evaporated and the
tert
residue was chromatographed on silica gel (CHCl3±
1
-
BuOH 5:1) to aŒord (2 02 g, 40 8% ). H NMR (CDCl3):
1
.
1 41±1 53 (27H, s), 2 85±2 95 (12H, t), 3 07±3 14 (4H, t),
.
.
3 27±3 29 (2H, s), 3 36±3 39 (4H, s). FAB-MS calcd for
.
.
.
.
.
.
C26H50N4O6 (M H): m z 515, found 515. Anal. calcd for
.
.
.
}
-
}
(C26H50N4O6[3 5CHCl3) C, H, N.
-(Carbobenzoxymethyl )-1,4,7,10-
N
tetraazacyclododecane-
butyl ester (2)
« «« «««-triacetic acid tri-
N ,N ,N
tert-
.
.
used as the model protein to evaluate the radiochemical To a suspension of 60% NaH (0 121 g, 2 38 mmol) in dry
1
yield of 111In-DOTA-labelled proteins. The ability of dimethylformamide (DMF), was added compound
mDOTA as the 111In-labelling agent to estimate the protein (1 08 g, 1 58 mmol) in dry DMF to maintain the reaction
.
.
pharmacokinetics was evaluated in comparison with temperature below
cDTPA.
® °
15 C. After 2 h stirring at room
. .
temperature, benzyl bromoacetate (1 24 mL, 6 33 mmol)
°
was added drop-wise below 0 C, and the reaction mixture
was stirred for 27 h at room temperature. A 20-mL solution
of 5% citric acid was then added to the reaction mixture
¬
before extraction with chloroform (3 20 mL). The organic
phase was dried and evaporated, and the residue was
Materials and Methods
Materials
tert
-BuOH 5:1) to
Proton nuclear magnetic resonance (1H NMR) spectra yield (0 597 g, 42 8% ). H NMR (CDCl3): 1 24±1 49
chromatographed on silica gel (CHCl3±
1
2
were recorded on a Bruker AC-200 spectrometer, and the (27H, m), 2 15±3 54 (24H, m), 5 13 (2H, d), 7 33±7 42
.
.
.
.
.
chemical shifts were reported in ppm down®eld from an (5H, m). FAB-MS calcd for C35H58N4O8 (M Na): m z
.
.
.
.
}
-
internal tetramethylsilane standard. Fast atom bombard- 685, found 685.

1075
A residualizing label for protein pharmacokinetics
tert
-butyl bromoacetate,
Pd C; d, 2,3,5,6-tetra¯uorophenol, 1,3-dicyclohexylcarbodiimide; e, tri¯ uoroacetic acid, anisole.
N N
, -diisopropylethylamine; b, benzyl bromoacetate, NaH; c, H2,
Figure 1 Synthesis of mDOTA. Reagents: a,
}
-(Carboxymethyl )-1,4,7,10-tetraazacyclododecane-
N
(3-dimethylaminopropyl)carbodiimide
hydrochloride
«
N ,N ,N
Bu) (3)
«« «««-triacetic acid tri-
-butyl ester (DOTA(
tert tert
-
.
(0 131 g, 0 681 mmol) in chloroform below 0 C. After 13 h
stirring at room temperature, the reaction mixture was
.
°
$
2
A mixture of compound (0 597 g, 0 901 mmol) and 10% washed with 5% citric acid, dried and evaporated. The
palladium on carbon (0 4 g)in methanol was hydrogenated residue was chromatographed on silica gel (CHCl3±MeOH
.
.
.
tert
.
at room temperature for 3 h. The catalyst was removed by 8:1) to produce DOTA( -Bu)3-Boc-Lys-OMe (63 0 mg,
®ltration, and the ®ltrate was evaporated to produce
1
33 9% ). 1H NMR (CDCl3 ):
3
(0 439 g, 85 1% ). H NMR (CDCl3): 1 46±1 47 (27H, m), 2 05±3 39 (24H, m), 3 72 (3H, s), 4 17±4 24 (1H, m), 5 19±
.
. . -
1 25±1 78(36H 8H, m),
.
.
.
.
.
.
.
.
.
.
1 91±3 65 (24H, m). FAB-MS calcd for C28H52N4O8 5 25 (1H, br). FAB-MS calcd for C40H74N6O11 (M Na):
.
(M Na): m z 595, found 595.
.
.
-
}
}
-
m z 837, found 837.
tert
was dissolved in ethanol (2 mL), and 2 NaOH (4 mL)
. .
-Bu)3-Boc-Lys-OMe (62 0 mg, 76 0 mol)
DOTA(
m
-[Carbo(2,3,5,6-tetra¯ uorophenoxyl)methyl ]-1,4,7,10-
«
N
tetraazacyclododecane-
was added. After 4 h stirring at room temperature, the
«« «««-triacetic acid tri-
-
tert
N ,N ,N
.
m
solution was acidi®ed to pH 3 5 with 4 H2SO4 before
extraction with chloroform. The organic phase was dried
and evaporated. A mixed solution of tri¯ uoroacetic acid
and anisole was added to the residue, and the reaction
solution was stirred for 30 h at room temperature. After
removal of tri¯ uoroacetic acid in-vacuo, dry ether was
butyl ester (4)
To a mixture of compound (0 100 g, 0 175 mmol) and
3
.
.
2,3,5,6-tetra¯ uorophenol (29 0 mg, 0 175 mmol) in dry
.
chloroform, was added 1,3-dicyclohexylcarbodiimide
.
.
(36 0 mg, 0 175 mmol) in dry chloroform to maintain the
reaction temperature below 0 C. After 9 h stirring at room
.
°
added to produce DOTA-lysine (22 0 mg, 53 7% ). 1H
.
.
. . . .
1 15±1 48 (6H, m), 1 75±1 82 (2H, m),
temperature, the reaction mixture was ®ltered, and the
®ltrate was concentrated in-vacuo. The residue was chro-
matographed on silica gel (CHCl3±MeOH 16:1)to produce
NMR (D2O):
.
.
.
2 94±3 15 (16H, m), 3 62±3 81 (8H, m), 4 90±4 97 (1H, m),
5 31±5 36 (1H, br). FAB-MS calcd for C22H40N6O9
.
.
.
.
.
(0 117 g, 93 0% ). 1H-NMR (CDCl3):
1 24±1 46
4
(27H, m), 2 16±3 98 (24H, m), 7 08±7 19 (1H, m). FAB-
.
.
.
.
}
(M H): m z 533, found 533.
-
.
MS calcd for C34H52N4O8F4 (M Na): m z 743, found 743.
.
.
.
}
-
Conjugation of mDOTA with proteins
}
Anal. calcd for (C34H52N4O8F4[5 2CHCl3) C, H, N.
1
To a solution of HSA, NGA and cyt c (9 mg mL ) in
-
. m
.
500 L of borate-buŒered saline (0 1 , pH 9 0), 10, 30 and
20 molar excess of mDOTA in 5 L of dimethyl sulfoxide
(DMSO) was added, respectively. Previous studies sug-
gested that DMSO at this concentration had no eŒect on
the biological activity of proteins (Hartikka et al 1989 ;
Nishikawa et al 1999). After incubating for 13 h at room
temperature, the conjugate was puri®ed by the centrifuged
column procedure using Sephadex G-50 (Pharmacia Bio-
-[Carbo(2,3,5,6-tetra¯ uorophenoxyl)methyl ]-1,4,7,10-
«« «««-triacetic acid
N
tetraazacyclododecane-
(mDOTA, 5)
«
N ,N ,N
4
Compound (0 104 g, 0 144 mmol) was stirred in a mixed
solution of tri¯ uoroacetic acid (2 mL) and anisole (0 1 mL)
.
.
.
for 30 h at room temperature. After removal of tri¯ uoro-
acetic acid in-vacuo, the residue was treated with dry ether
1
to precipitate (63 0 mg, 79 0% ). H NMR ((CD3)2SO):
5
2 98±3 80 (24H, m), 7 20±7 28 (1H, m). FAB-MS calcd for
.
.
. m
tech, Tokyo, Japan), equilibrated with 0 5 acetate buŒer
.
.
C22H28N4O8F4 (M Na): m z 575, found 575.
.
.
.
(pH 6 0). Conjugation of cDTPA with proteins was per-
}
-
formed as described previously (Arano et al 1994a). The
average number of DOTA or DTPA chelates incorporated
into cyt c was found to be 2, as determined by FAB-MS.
Synthesis of DOTA-lysine
tert
To a mixture of DOTA( -Bu)3 (0 130 g, 0 227 mmol),
-butoxycarbonyl- -lysine methyl ester (Boc-Lys-
.
.
111In labelling of proteins
N tert
l
-
.
.
m
.
OMe; 0 136 g, 0 681 mmol) and 1-hydroxybenzotriazole To 7 L of acetate buŒer (3 , pH 6 0), was added 35 L of
.
(92 0 mg, 0 681 mmol)in chloroform, was added 1-ethyl-3- 111InCl3. After the mixture was allowed to stand for 5 min
.

1076
at room temperature, a 26- L solution of DOTA- or
Takahiro Mukai et al
Subcellular distribution of radiometabolites in
the liver
1
DTPA-conjugated proteins (3 0±10 mg mL ) was added.
-
.
After incubating for 1 5 h at 37 C, the conjugate was
.
puri®ed by the centrifuged column procedure using Sepha-
°
The subcellular distribution of radioactivity in the liver was
investigated by density-gradient centrifugation. At 1 and
24 h post-injection of 111In-DOTA-NGA, liver homogenate
was prepared according to the method described previously
(Arano et al 1994a ; Mukai et al 1999b). The isolated
supernatant was then layered on the isotonic Percoll (Phar-
. m
buŒer (pH 7 4). The radiochemical purity of 111In-labelled
dex G-50, equilibrated and eluted with 0 1 phosphate
.
proteins was determined by SE-HPLC, CAE and TLC.
For 111In labelling of bovine apotransferrin (Nacalai
Tesque Inc.), a 40- L solution of 111InCl3 was mixed with a
1
-
.
macia Biotech) at a density of 1 08 g mL . After centri-
m
40- L solution of 1 sodium acetate, and the resulting
111In-acetate solution (50 L) was added to an apotrans-
g
°
fugation at 20000 at 4 C for 90 min (RP 30 rotor Hitachi
Co. Ltd, Tokyo, Japan), the gradients were collected in 14
fractions, and the activity of lysosomal enzyme ( -galacto-
sidase), the density and radioactivity of respective fractions
were analysed.
1
-
ferrin solution (150 L; 5 mg mL ) in 0 1 phosphate
. m
buŒer (pH 7 4). After incubating for 10 min at 37 C, a 20-
.
L solution of 2 m EDTA was added. After a 30-min
°
m
incubation, 111In-labelled transferrin was puri®ed by the
centrifuged gel-®ltration column.
Radiochemical yields of 111In-labelled cyt c
The chelator concentrations of DOTA- and DTPA-cyt c
.
Plasma stability of 111In-labelled HSA
.
.
.
(2 0 each chelator per cyt c) were adjusted to 5 6, 16 9, 28 1,
m
111In-DOTA-HSA and 111In-DTPA-HSA were diluted 20
fold with freshly prepared murine plasma containing 0 1%
. .
42 2 and 84 4
7
. m .
with 0 5 acetate buŒer (pH 6 0). To
m .
L of acetate buŒer (3 , pH 6 0), was added 35 L of
.
111InCl3, and the solution was incubated for 5 min before
adding to 26 L of each chelator-conjugated cyt c. After
sodium azide. After 1, 2, 3 and 5 days of incubation at
111
3
+
°
37 C, the samples were analysed by CAE. Since In ,
released from 111In-labelled proteins in the circulation,
forms a strong complex with transferrin (a serum iron
transport protein) (Himmelsbach & Wahl 1989 ; Reilly et al
1992 ; Claessens et al 1995), 111In-labelled transferrin was
used as an analytical reference.
.
°
incubating for 1 5 h at 37 C, EDTA was added to a 100
molar excess over each protein molecule, and the reaction
mixture was gently agitated for 30 min at room tempera-
ture. The radiochemical yields of 111In-labelled proteins
were determined by CAE. DOTA-lysine and DOTA were
labelled with 111In in the same manner, and the radio-
chemical yields were determined by TLC.
Biodistribution of 111In-labelled NGA
Statistical analysis
Animal studies were conducted in accordance with our
institutional guidelines and were approved by Kyoto Uni-
versity Animal Care Committee. Biodistribution studies
were performed by the intravenous administration of 111In-
labelled NGA (9 g) to 6-week-old male ddY mice (27±
30 g). At 10 and 30 min, 1, 3, 6 and 24 h post-injection,
mice were killed by decapitation and samples of blood were
collected. Liver, kidney, spleen, stomach, intestine and
lung were excised and weighed, and the radioactivity counts
were determined with an auto well gamma counter (ARC
2000, Aloka, Tokyo, Japan).
Percentages of HSA-bound radioactivity after incubation
of 111In-labelled HSA in murine plasma were expressed as
³
the mean s.d. of three experiments, and the statistical
signi®cance of the diŒerence between 111In-DOTA-HSA
and 111In-DTPA-HSA was calculated using the Mann±
U
Whitney -test. Experimental results regarding mouse
distribution of 111In-labelled NGA were given as the
³
mean s.d. of ®ve experiments, and the statistical analysis
was performed by applying the Mann±Whitney -test.
111In-complexation yields of cyt c conjugates were expressed
U
³
as the mean s.d. of three experiments, and the statistical
signi®cance of the diŒerence between 111In-DOTA-cyt c
and 111In-DTPA-cyt c was estimated using the Mann±
IdentiŽ cation of radiometabolites of 111In-
labelled NGA
U
Whitney -test.
signi®cant.
P
.
0 05 was considered to be statistically
!
To identify the radiometabolites, 111In-DOTA-NGA was
injected intravenously into mice. At 1 and 24 h post-
injection, liver homogenate was prepared according to the
method described previously (Arano et al 1994b). Super-
Results and Discussion
N
natants were separated from the pellets, ®ltered through a In a previous study, Lewis et al (1994) synthesized the
-
.
polycarbonate membrane with a pore diameter of 0 45 m, hydroxysulfosuccinimide ester of DOTA in the presence of
and analysed by SE-HPLC, CAE and TLC. Each analysis water-soluble carbodiimide to prepare 111In-labelled pro-
was also carried out by co-chromatography with 111In- teins. This procedure may be applicable to assess 111In-
DOTA-lysine and 111In-DOTA.
DOTA as a residualizing label for protein pharmaco-

1077
A residualizing label for protein pharmacokinetics
Figure 2 Stability of 111In-labelled HSA after incubation in murine
plasma at 37°C. E, 111In-DOTA-HSA; D, 111In-DTPA-HSA. Each
!
P
.
0 05
value represents the mean³s.d. of three experiments. *
111
compared with In-DOTA-HSA (Mann±Whitney -test).
U
kinetics. However, since the active ester was not isolated
from the reaction solution before conjugation reactions
with proteins, a small amount of multiple activation of
DOTAwas observed in the preparations. When the DOTA-
N
antibody conjugates wereprepared by using the -hydroxy-
sulfosuccinimide ester, approximately 6% of the antibody
was converted to dimers (Lewis et al 1994). The inter-
molecular cross-linked species such as antibody dimers
could be removed from the protein conjugates during
puri®cation, as they described. However, the products of
intramolecular cross-linking were di cult to detect and
remove, and would impair the 111In-chelate stability, as was
observed in cDTPA radiolabelling of proteins (Arano et al
1996). To estimate the ability of DOTA as an 111In-based
residualizing label of proteins, we developed a synthetic
procedure for a monoreactive DOTA derivative to avoid
cross-linking of proteins.
Monoreactive DOTA was prepared by protecting one
carboxylate with a reagent that can be removed under
conditions diŒerent from the protecting groups of the other
three carboxylates in the DOTA skeleton. We selected the
Figure 3 Cellulose acetate electrophoresis (CAE) pro®les at 5 days
after incubation of 111In-DOTA-HSA (A) and 111In-DTPA-HSA (B)
in murine plasma at 37°C. The peak of 111In-labelled apotransferrin
.
was observed at 0 5±1 cm anode from the origin (C).
tert
benzyl-protecting group for one carboxylate and the
-
butyl-protecting group for the rest of the three carboxylates
2
(compound ). The precursor of the monoreactive DOTA,
.
3,
tert
N
-butyl esters. However, -hydroxy-
compound was produced in relatively high yield (85 1% ) deprotection of the
2
by catalytic hydrogenation of compound . This compound succinimide ester underwent decomposition during the
is a monocarboxylic acid derivative of DOTA with the deprotection reaction. Therefore, we selected a tetra¯ uoro-
remainder of the three carboxylates being protected with phenyl ester as a protein binding site. According to the
tert
acid-labile
-butyl ester. The high solubility of this procedure outlined in Figure 1, mDOTA was synthesized
compound in a variety of organic solvents would render with an overall yield of 11% , and the chemical purity of
the reagent versatile for introducing a DOTA skeleton to a mDOTA synthesized was found to be over 99% , as de-
variety of peptides by either liquid or solid synthesis for termined by TLC.
diagnostic and therapeutic applications in nuclear medi-
cine.
The stability of 111In-labelled proteins in murine plasma
was investigated using HSA as a model protein (Figure 2).
We attempted to convert the free carboxylate of com- Figures 3A and 3B show typical CAE pro®les at 5 days
pound to -hydroxysuccinimide ester and subsequent after incubation of 111In-DOTA-HSA and 111In-DTPA-
3
N

1078
Takahiro Mukai et al
cross-linking during cDTPA conjugation reactions (Arano
et al 1996). Therefore, the 111In-labelling method using
cDTPA would be unsuitable for pursuing the pharma-
cokinetics of proteins with long biological half-lives of a
few days.
To assess the protein pharmacokinetics accurately, long
retention of the radiometabolites at the catabolic sites is
required (Thorpe et al 1993). Our previous ®ndings sug-
gested that the 111In-DTPA-labelled proteins, without in-
ducing cross-linking, exhibited long residence of the radio-
activity in the catabolic sites (Arano et al 1994a ; Mukai et
al 1999a). For further evaluation of the 111In-DOTA label
as a residualizing label, the residence time of the radio-
activity in the catabolic site was investigated after admini-
stration of 111In-DOTA-labelled proteins and compared
with 111In-DTPA-labelled proteins using NGA as a model
protein. Use of NGA provides explicit information re-
garding the fate of radioactivity after lysosomal proteolysis
in hepatocytes without being aŒected by transchelation of
radiolabels in plasma or a redistribution of radiometa-
bolites generated elsewhere in the body to the liver (Arano
et al 1994a, 1995 ; Mukai et al 1999a, 1999b). Figure 4
shows the radioactivity pro®les in the liver and blood after
injection of 111In-DOTA-NGA and 111In-DTPA-NGA into
mice. Approximately 86% of the injected radioactivity
accumulated in the liver at 10 min post-injection of 111In-
DOTA-NGA, while only low radioactivity existed in the
blood at this post-injection time (Figure 4). Elimination of
radioactivity from the liver was slow, with 61% of the
injected radioactivity remaining even at 24 h post-injection.
No signi®cant diŒerence was seen in the hepatic radio-
activity levels at any post-injection times between 111In-
DOTA-NGA and 111In-DTPA-NGA. Low radioactivity,
.
amounting to less than 4 0% of that injected, was detected
Figure 4 Radioactivitypro®les in the liver and blood after injection
of 111In-labelled NGA into mice. E, 111In-DOTA-NGA; D, 111In-
DTPA-NGA. Each value represents the mean³s.d. of ®ve experi-
ments. No signi®cantdiŒerencebetween111In-DOTA-NGAand 111In-
DTPA-NGA was seen in the hepatic and blood radioactivity levels at
in kidney, spleen, stomach, intestine and lung after injection
of 111In-labelled NGA (data not shown). The high plasma
stability and long retention of radioactivity at the catabolic
sites indicated that the 111In-labelling procedure using
mDOTA as a BCA would be applicable for estimating
protein pharmacokinetics.
U
any post-injection time (Mann±Whitney -test).
To determine the radiometabolites trapped in the cata-
bolic site, analyses of the supernatant of the liver homo-
HSA, respectively. The CAE pro®le of 111In-labelled apo- genate at 1 and 24 h post-injection of 111In-DOTA-NGA
transferrin is also illustrated in Figure 3C. The radioactivity were performed (Figure 5). Extraction e ciency of the
associated with HSA fractions of 111In-DOTA-HSA was radioactivity from each homogenate was more than 93% .
unchanged for 5 days, indicating high stability of 111In- On the basis of the previous ®ndings that 111In-DTPA-
DOTA labels. For the assessment of protein pharma- NGA was degraded to 111In-DTPA-lysine in the liver
cokinetics, high stability of radiolabels is required. The (Arano et al 1994c ; Franano et al 1994), we synthesized
111In-labelling procedure using mDOTA would satisfy this 111In-DOTA-lysine as an authentic standard. The SE-
requirement although one of the four carboxylates in the HPLC pro®le of each supernatant depicted a single radio-
.
DOTA molecule was utilized for protein conjugation. On activity peak at 28 5 min, a retention time that was rep-
the other hand, 111In-DTPA-HSA released the radioac- resentative of small-molecular-weight compounds such as
tivity, and approximately 20% of the initial radioactivity 111In-DOTA-lysine and 111In-DOTA. Under similar con-
was detected in the transferrin fraction at 5 days. These ditions, 111In-DOTA-NGA was eluted at 17 min. CAE
.
observations recon®rmed the involvement of transchela- analyses showed the major radioactivity peak at 0 5±1 cm
tion of 111In from 111In-DTPA-labelled proteins to trans- cathode from the origin, which was identical to 111In-
ferrin in plasma (Himmelsbach & Wahl 1989 ; Reilly et al DOTA-lysine (Figure 5B). TLC analyses registered the
1992 ; Claessens et al 1995). The low stability of 111In- major radioactivity peak with an Rf value similar to that of
DTPA-labelled proteins is not attributable to low stability 111In-DOTA-lysine (Figure 5C). In every analytical system,
of 111In-DTPA chelate, but to formation of intramolecular each sample showed a single radioactivity peak even when

1079
A residualizing label for protein pharmacokinetics
Figure 6 Percoll density gradient pro®les of liver homogenate at 1 h
(A) and 24 h (B) post-injection of 111In-DOTA-NGA into mice.
A single radioactivity peak (E) that coincided with -galactosidase
1
-
.
activity (_) was detected at a density (D) of ca. 1 13 g mL
.
2001). These ®ndings indicated that transchelation of 111In
from 111In-DOTA chelate to biomolecules inside the cells
was negligible.
The subcellular distribution of radioactivity in liver
homogenates at 1 and 24 h post-injection of 111In-DOTA-
NGA was investigated by Percoll density gradient centri-
fugation (Figure 6). Each liver homogenate demonstrated
Figure 5 Chromatographic analyses of liver homogenates after
injection of 111In-DOTA-NGA into mice. E, liver homogenateat 1 h;
D, liver homogenate at 24 h; _, 111In-DOTA-lysine. Supernatants
were analysed by SE-HPLC (A), CAE (B) and TLC (C). The parental
111In-DOTA-NGA (dotted line) was eluted at 17 min on SE-HPLC
(A). The peak of 111In-DOTA was observed at 1 cm anode on CAE or
1
-
a single radioactivity peak at a density of ca. 1 13 g mL ,
.
which correlated well with the lysosomal enzyme ( -galac-
tosidase) activity pro®le. Thus, the long residence times of
the radioactivity in the catabolic site of 111In-DOTA-
labelled proteins would be attributable to the slow elim-
ination rate of the ®nal radiometabolite, 111In-DOTA-
lysine, from the lysosomal compartment.
.
Rf value of ca. 0 2 on TLC.
To pursue the real-time localization of the radioactivity
in small animals by SPECT, a high dose of radioactivity
co-chromatographed with 111In-DOTA-lysine (data not has to be administered into the animals. Thus, cyt c was
shown). These results are supported by the recent ®nding used as a model protein to evaluate the radiochemical yield
of Tsai and coworkers who demonstrated a generation of of the 111In-DOTA-labelled proteins, since the number of
this radiometabolite in the kidney following administration chelators introduced per molecule of cyt c was assessed by
of 111In-DOTA-conjugated antibody fragments (Tsai et al mass spectrometry (Lewis et al 1994). 111In-complexation

1080
Takahiro Mukai et al
introduction of a protein binding site at a position distal
from an intact DOTA structure would be preferable to
prepare 111In-labelled proteins of high speci®c activity for
SPECT studies.
References
Arano, Y., Mukai, T., Uezono, T., Wakisaka, K., Motonari, H.,
Akizawa, H., Taoka, Y., Yokoyama, A. (1994a) A biological
method to evaluatebifunctionalchelatingagentsto label antibodies
J Nucl Med
. . . 35: 890±898
with metallic radionuclides.
Arano, Y., Inoue, T., Mukai, T., Wakisaka, K., Sakahara, H.,
Konishi, J., Yokoyama, A. (1994b) Discriminated release of a
hippurate-like radiometal chelate in nontarget tissues for target-
selectiveradioactivitylocalizationusing pH-dependentdissociation
J Nucl Med
. . . 35: 326±333
of reduced antibody.
Arano, Y., Mukai, T., Uezono, T., Motonari, H., Wakisaka, K.,
Yokoyama, A. (1994c) Biological comparison of DTPA and SCN-
benzyl-EDTAas chelatingagentsfor indium labeling of antibodies.
Figure 7 111In-complexation yields of chelator-conjugates. E,
DOTA-cyt c; D, DTPA-cyt c; _, DOTA-lysine; ^, DOTA. Radio-
labelling was performed by incubating chelator conjugates with 111In-
J Labelled Compd Radiopharm
. . . 35: 381±383
Arano, Y., Mukai, T., Akizawa, H., Uezono, T., Motonari, H.,
Wakisaka, K., Kairiyama, C., Yokoyama, A. (1995) Radiolabeled
metabolites of proteins play a critical role in radioactivity elim-
.
acetate complexes for 1 5 h at 37°C. Each value represents the
mean³s.d. of three experiments. The radiochemical yield of 111In-
DTPA-cyt c was signi®cantly higher than that of 111In-DOTA-cyt c at
!
Nucl Med Biol
. . . 22: 555±564
P
. U
0 05, Mann±Whitney -test).
ination from the liver.
any concentration (
Arano, Y., Uezono, T., Akizawa, H., Ono, M., Wakisaka, K.,
Nakayama, M., Sakahara, H., Konishi, J., Yokoyama, A. (1996)
Reassessment of diethylenetriaminepentaacetic acid (DTPA) as a
chelating agent for indium-111 labeling of polypeptides using a
.
yields of DOTA-cyt c and DTPA-cyt c (2 0 chelates per cyt
c) are summarized in Figure 7. At chelator concentrations
J Med Chem
. . .
newly synthesized monoreactive DTPA derivative.
39: 3451±3460
of 84 4 and 42 2
, DOTA-cyt c generated 111In-labelled
.
.
m
Claessens, R. A. M. J., Koenders, E. B., Boerman, O. C., Oyen,
W. J. G., Borm, G. F., van der Meer, J. W. M., Corstens, F. H. M.
(1995) Dissociation of indium from indium-111-labelleddiethylene
triamine penta-acetic acid conjugated non-speci®c polyclonal hu-
1
-
.
cyt c of 81 4 and 159 MBq mg speci®c activity, respect-
ively, with over 91% radiochemical yields. However, sig-
ni®cant decreases in the radiochemical yields were observed
with a decrease in the chelator concentrations. On the other
hand, DTPA-cyt c produced 111In-labelled cyt c of
Eur J Nucl Med
. . . .
man immunoglobulin G in in¯ ammatory foci.
22: 212±219
1
-
1 20 GBq mg speci®c activity, with over 97% radio-
.
Franano, F. N., Edwards, W. B., Welch, M. J., Duncan, J. R. (1994)
Metabolism of receptortargeted111In-DTPA-glycoproteins: identi-
®cation of 111In-DTPA- -lysine as the primary metabolic and
.
m
).
chemical yields at lower chelator concentrations (5 62
For further pursuit of the 111In-complexation reactions of
DOTA-conjugated proteins, DOTA-lysine and DOTA
were also labelled with 111In. 111In-DOTA-lysine and 111In-
DOTA were obtained with over 84% radiochemical yields
Nucl Med Biol
. . . 21: 1023±1034
excretory product.
Hartikka, M., Vihko, P., Sodervall,M., Hakalahti, L., Torniainen, P.,
Vihko, R. (1989)Radiolabellingof monoclonal antibodies: optimi-
zation of conjugation of DTPA to F(ab«)2-fragments and a novel
measurement of the degree of conjugation using Eu(III)-labelling.
.
.
m
at chelator concentrations of more than 28 1 and 10 0 ,
respectively (Figure 7). These observations suggested that
the relatively low speci®c activity of 111In-DOTA-cyt c
could be attributed not only to a steric interference of
protein molecules but also to the modi®cation of one
carboxylate of DOTA (Keire & Kobayashi 1999).
Eur J Nucl Med
. . . . 15: 157±161
Himmelsbach,M., Wahl, R. L. (1989)Studies on the metabolic fate of
111
Nucl Med Biol
. . 16: 839±845
In-labeled antibodies.
.
Hnatowich, D. J., Layne, W. W., Childs, R. L., Lanteigne, D., Davis,
M. A., Gri n, T. W., Doherty, P. W. (1983) Radioactive labeling
Science
of antibody: a simple and e cient method.
220: 613±615
Ishizu, K., Mukai, T., Yonekura, Y., Pagani, M., Fujita, T., Magata,
Y., Nishizawa, S., Tamaki, N., Shibasaki, H., Konishi, J. (1995)
Ultra-high resolution SPECT system using four pinholecollimators
Conclusions
J Nucl Med
. . . 36: 2282±2287
for small animal studies.
We developed mDOTA as a BCA to label proteins with
111In. 111In-labelled proteins using mDOTA as a BCA were
highly stable in-vitro and in-vivo. Furthermore, long resi-
dence times of the radioactivity in the catabolic site were
observed after injection of 111In-DOTA-labelled proteins,
due to the slow elimination rate of the radiometabolite
from the lysosome. Thus, mDOTA appears to be an
appropriate 111In-labelling agent for estimating protein
pharmacokinetics. The ®ndings of this study also suggested
Keire, D. A., Kobayashi, M. (1999) NMR studies of the metal-
loading kinetics and acid-base chemistryof DOTA and butylamide-
Bioconjugate Chem
. 10: 454±463
DOTA.
Lewis, M. R., Raubitschek, A., Shively, J. E. (1994) A facile, water-
soluble method for modi®cation of proteins with DOTA. Use of
elevated temperature and optimized pH to achieve high speci®c
activity and high chelate stability in radiolabeled immunocon-
Bioconjugate Chem
jugates. . 5: 565±576
Maisano, F., Gozzini, L., de Haen, C. (1992) Coupling of DTPA to

1081
A residualizing label for protein pharmacokinetics
proteins: a critical analysis of the cyclic dianhydride method in the
Bioconjugate Chem
stability of EDTA and DTPA immunoconjugates of monoclonal
Appl Radiat Isot
. 43:
case of insulin modi®cation.
. 3: 212±217
antibody 2G3 labeled with indium-111.
961±967
.
.
Mukai, T., Arano, Y., Nishida, K., Sasaki, H., Saji, H., Nakamura, J.
(1999a)In-vivo evaluationof indium-111-diethylenetriaminepenta- Thorpe, S. R., Baynes, J. W., Chroneos, Z. C. (1993) The design and
acetic acid-labelling for determining the sites and rates of protein application of residualizing labels for studies of protein catabolism.
J Pharm Pharmacol FASEB J
catabolism in mice.
.
.
. 51: 15±20 . 7: 399±405
Mukai, T., Arano, Y., Nishida, K., Sasaki, H., Akizawa, H., Ogawa, Tsai, S. W., Li, L., Williams, L. E., Anderson, A.-L., Raubitschek,
A. A., Shively,J. E. (2001)Metabolism and renal clearanceof 111In-
BioconjugateChem
K., Ono, M., Saji, H., Nakamura, J. (1999b) Species diŒerence in
radioactivityeliminationfrom liverparenchymalcellsafterinjection
labeledDOTA-conjugatedantibodyfragments.
.
Nucl Med Biol
. . . 26: 281±289
of radiolabeled proteins.
12: 264±270
Nishikawa, M., Staud, F., Takemura, S., Takakura, Y., Hashida, M.
(1999)Pharmacokineticevaluationof biodistribution data obtained
Biol Pharm Bull
. 22: 214±218
Weber, D. A., Ivanovic,M., Franceschi,D.,Strand,S.-E., Erlandsson,
K., Franceschi, M., Atkins, H. L., Coderre, J. A., Susskind, H.,
Button, T., Ljunggren, K. (1994) Pinhole SPECT: an approach to
in vivohigh resolution SPECT imaging in small laboratory animals.
with radiolabeled proteins in mice.
.
.
Pittman, R. C., Green, S. R., Attie, A. D., Steinberg, D. (1979)
Radiolabeled sucrose covalently linked to protein. A device for
quantifying degradation of plasma proteins catabolized by lysoso-
J Biol Chem
J Nucl Med
. . . 35: 342±348
Wu, C., Virzi, F., Hnatowich, D. J. (1992) Investigations of N-linked
macrocyclesfor 111In and 90Y labeling of proteins.
Nucl Med Biol
. . .
mal mechanisms.
.
.
. 254: 6876±6879
Reilly, R., Lee, N., Houle, S., Law, J., Marks, A. (1992) In vitro
19: 239±244