Polymer-supported organotin reagent for prosthetic group labeling of biological macromolecules with radioiodine

Article abstract of DOI:10.1021/bc1004203

In this study, we investigated the use of poly-mer-bound precursor for generating a radiolabeled prosthetic group to be used for conjugate labeling of biological macromolecules. For the approach, a trialkyltin chloride in which the tin was bound to a hydrophilic PEG-based resin support via one of the alkyl groups was synthesized. This resin was then used to prepare a resin-bound trialkyltin benzoic acid, which in some cases was further derivatized on-resin by converting it to a succinimidyl ester. Exposure of the resin-bound compounds to electrophilic radioiodine (¹²⁵I) in either an aqueous or methanol solvent liberated either free radiolabeled [¹²⁵I]iodobenzoic acid or its succinimidyl ester without co-release of the resin-bound precursors. Radiochemical yield was between 35% and 75%, depending on the solvent system and precursor. As example applications for the released compounds, the amine-reactive N-succinimidyl-[¹²⁵I]iodobenzoate prosthetic group was used for conjugate radiolabeling of a peptide, tomato plant systemin, and two proteins, albumin and IgG antibody. These results demonstrate that resin-bound organotin precursors in which the compound to be labeled is tethered to the support via the tin group to be substituted can be used to produce radioiodine-labeled aromatic prosthetic groups in good specific activity without the need for HPLC purification. This solid-phase approach is potentially adaptable to kit-formulation for performing conjugate radiolabeling of biological macromolecules.

Full text of DOI:10.1021/bc1004203

ARTICLE
pubs.acs.org/bc
Polymer-Supported Organotin Reagent for Prosthetic Group Labeling
of Biological Macromolecules with Radioiodine
Andrew N. Gifford,* Sonja Kuschel, Colleen Shea, and Joanna S. Fowler
Medical Department, Brookhaven National Laboratory, Upton, New York 11973, United States
S Supporting Information
b
ABSTRACT: In this study, we investigated the use of poly-
mer-bound precursor for generating a radiolabeled prosthetic
group to be used for conjugate labeling of biological macro-
molecules. For the approach, a trialkyltin chloride in which
the tin was bound to a hydrophilic PEG-based resin support
via one of the alkyl groups was synthesized. This resin was
then used to prepare a resin-bound trialkyltin benzoic acid,
which in some cases was further derivatized on-resin by converting it to a succinimidyl ester. Exposure of the resin-bound
compounds to electrophilic radioiodine (125I) in either an aqueous or methanol solvent liberated either free radiolabeled
[
125I]iodobenzoic acid or its succinimidyl ester without co-release of the resin-bound precursors. Radiochemical yield was between
35% and 75%, depending on the solvent system and precursor. As example applications for the released compounds, the amine-
reactive N-succinimidyl-[125I]iodobenzoate prosthetic group was used for conjugate radiolabeling of a peptide, tomato plant
systemin, and two proteins, albumin and IgG antibody. These results demonstrate that resin-bound organotin precursors in which
the compound to be labeled is tethered to the support via the tin group to be substituted can be used to produce radioiodine-labeled
aromatic prosthetic groups in good specific activity without the need for HPLC purification. This solid-phase approach is potentially
adaptable to kit-formulation for performing conjugate radiolabeling of biological macromolecules.
’ INTRODUCTION
conjugate from any unreacted radiolabeled prosthetic group re-
maining in solution to give the final purified product.
Peptides, proteins, and other biological macromolecules tagged
with radioiodine have widespread uses in the biological sciences.
The most common method of labeling these molecules with
radioiodine is by the direct reaction of electrophilic radioiodine
with tyrosine or other phenolic sites on the molecule.1,2 However,
this direct approach has some disadvantages, including the fact that
the molecule must be exposed to oxidative conditions in order to
produce the electrophilic radioiodine from its sodium salt. This can
oxidize methionine and other sensitive residues on the protein or
peptide and lead to a reduction or loss its biological activity.3,4
Additionally, in the case of labeled macromolecules destined for in
vivo applications, a further disadvantage is the fact that radioiodine
on tyrosine rings is unstable in biological fluids and free iodine
slowly accumulates in the thyroid and stomach.5,6 Due to these
limitations, an alternative method of labeling biological macromo-
lecules with radioiodine that has been developed is conjugation of
the macromolecule with a small preradioiodinated prosthetic
group. Although a variety of prosthetic groups have been developed
for this purpose,7 one of the most commonly used is N-succinimi-
dyl iodobenzoate.8,9 In this molecule, the radioiodine is attached
directly onto a nonactivated aromatic ring and is resistant to
deiodination in vivo.9,10 The succinimidyl ester on the molecule
allows it to be readily conjugated to lysine amine groups on peptide
or proteins under mild aqueous, nonoxidative conditions. Follow-
ing the conjugation reaction, a gel filtration or ultrafiltration step
is typically used to separate the radiolabeled macromolecular
The most effective method for labeling nonactivated aromatic
compounds with radioiodine is via radioiododestannylation of
the corresponding organotin precursor.11 This approach has
similarly been used to produce radiolabeled N-succinimidyl-
iodobenzoate.8,9 The organotin precursors have good stability
in storage and are stable to air and moisture, and the radio-
iododestannylation reaction itself proceeds readily and in high
yield.11 To ensure consumption of the radioiodine in radioiodo-
destannylation reactions, and thus a good overall radiochemical
yield, the organotin precursor is typically present in an amount
between 1 and 2 orders of magnitude excess over the concentra-
tion of radioiodine. However, in the case of prosthetic groups
used for conjugate radiolabeling of macromolecules, a conse-
quence of the presence of a large excess of organotin compounds
in the reaction mixture is that an HPLC or equivalent purification
step is normally needed following the radiolabeling reaction to
separate the final radiolabeled product from unreacted organotin
precursors. This is necessary in order to avoid unnecessary
blockade of amine sites on the biomolecule by the amine-reactive
organotin precursors. Complete removal of tin precursors and
byproducts is also especially important where the products may
Received: September 22, 2010
Revised:
January 11, 2011
Published: February 10, 2011
r
2011 American Chemical Society
406
dx.doi.org/10.1021/bc1004203 Bioconjugate Chem. 2011, 22, 406–412
|

Bioconjugate Chemistry
ARTICLE
have potential for use in clinical studies, since organotin is highly
toxic to humans. However, HPLC purifying the radiolabled
prosthetic group prior to conjugation to biological protein or
macromolecules to remove organotin precursors and byproducts
introduces extra complexity into the labeling process and creates
significant volumes of radioactive waste. Additionally, HPLC has
limitations in effectively removing toxic organotin byproducts
from solution-phase reactions. Thus, trialkyltin-aromatic pre-
cursors are strongly retained by reverse-phase material, and
hence, HPLC columns must be washed well between runs, while
cleaved trialkyltin salt byproducts are prone to absorption onto
exposed silanol groups on the silica resulting in broad peaks that
can potentially overlap with product peaks (ref 12 and figures
given in the Supporting Information).
In the current study, we investigated the use of a solid phase
approach to labeling the radioiodinated prosthetic group with the
goal of avoiding the need for postlabeling purification steps prior
to conjugation to the biological macromolecule. In this solid-
phase approach, the organotin precursor to be iodinated is
tethered to a resin support via the alkyl tin group to be
substituted. Following the iodination, the radioiodinated product
is released free into the solvent, whereas unreacted tin precursors
and byproducts remain bound to the insoluble resin beads,
enabling the latter to be easily separated from the radioiodinated
product by a simple filtration or extraction step without HPLC.
Polymer-bound tin compounds have been previously exam-
ined for catalytic applications13,14 and for preparation of cold
(unlabeled) iodinated compounds.15 However, there have only
been limited prior studies reported in the literature in which
polymer-bound organotin precursors have been used to produce
radiolabeled compounds, namely, metataiodobenzylguanidine16,17
and Congo Red.18 In analogous approaches, organogermanium19
and organoborane20 precursors that were linked to a solid support
have been examined for their potential to produce radio-iodinated
compounds, but these precursors have reduced reactivity to radio-
iodine compared to organotin compounds.11 With respect to
prosthetic groups for conjugate radioiodine labeling of macromo-
lecules, Donovan et al.21,22 have reported a solution-phase analogue
of the solid-phase labeling strategy in which a perfluoroakyl tin
group was used as the leaving group. The high affinity of the
perfluoroalkyl moiety for fluorous solid-phase extraction media
enabled its easy removal from the reaction mix. However, despite
these latter variants, as far as we are aware, a prosthetic group for
radiolabeling macromolecules in which the organotin precursor is
covalently bound to the polymer support has not yet been
examined.
Acrylamide-Functionalized Resin (1). To a suspension of
amino NovaPEG resin (4.5 g; loading 0.66 mmol/g; EMD
Chemicals Inc., Gibbstown, NJ) in DMF (50 mL) was added
acryloyl chloride (1.05 mL, 13 mmol) and TEA (1.85 mL, 13
mmol). The mixture was allowed to react for 2 h at room tem-
perature with gentle shaking. The solvent was then removed and
the resin washed twice with DMF (10 mL), once with metha-
nol (10 mL), and then evaporated to dryness. A ninhydrin test
performed on samples of the resin was negative following the
reaction, indicating complete coupling of the acryloyl chloride to
the resin amines.
Dibutyltin Dihydride. Sodium borohydride (2.2 g, 58 mmol)
was dissolved in ice-cold water (45 mL) with continuous
stirring and nitrogen bubbling. To this mixture was added
Bu2SnCl2 (3 g, 10 mmol) in ether. Stirring and nitrogen bubbling
was continued and the mixture allowed to warm to room tem-
perature over 15 min. The ether layer was then separated, washed
with water (2 ꢀ 15 mL), dried over MgSO4, and evaporated to
yield Bu2SnH2 (1.5 g) as a colorless oil. Since alkyltin hydrides
have poor stability in storage, the Bu2SnH2 was used directly in
the next step without further characterization.
Dibutyltin Chloride-Functionalized Resin (2). To an oven-
dried Schlenk flask under nitrogen was added 4.35 g of the
acrylamide functionalized resin and THF (50 mL). Once the
resin was fully swollen, Bu2SnCl2 (2 g, 6.6 mmol), Bu2SnH2 (1.5
g, 6.4 mmol), and 2,20-azobisisobutyronitrile (AIBN; 58 mg, 0.35
mmol) were added. The mixture was allowed to react overnight
at room temperature with gentle shaking and irradiation from a
halogen light source (150 W). During this process, the resin was
observed to change appearance from yellow-orange color to
more of an opaque off-white color. Following the reaction, the
resin was washed with THF, butanol, and methanol and evapo-
rated to dryness to give a yellow-white powder (5.05 g).
Elemental analysis (ICP-MS) indicated a tin content of the resin
of 5.5% and a chloride content of 2.5%.
Polymer-Bound 3-(Dibutylstannyl)benzoic Acid (4). To
an oven-dried Schlenk flask under nitrogen was added the
dibutyltin chloride-functionalized resin (1.3 g) and THF (10
mL). Once the resin was fully swollen, the solvent was aspirated
off to just above the level of the resin and 3-(ethoxycarbonyl)-
phenylzinc (0.5 M in THF; 6 mL) added. The mixture was
allowed to react overnight at room temperature with gentle
shaking. The resin was then washed with THF (2 ꢀ 10 mL),
butanol (1 ꢀ 10 mL), and water (2 ꢀ 10 mL). Following
washing, the resin was transferred to a plastic 50 mL tube and 1 M
aq. NaOH (20 mL) added. The tube was then incubated with
gentle shaking at 55 °C for 4 h. The resin was washed with water
(2 ꢀ 10 mL) and methanol (1 ꢀ 10 mL) and evaporated to
dryness.
’ EXPERIMENTAL PROCEDURES
General. Organometallic reagents were handled under a
nitrogen atmosphere using oven-dried glassware and anhydrous
solvents. Elemental analysis on resin samples was conducted by
To test for formation of polymer-bound 3-(dibutylstannyl)-
benzoic acid, a sample of the resin (1 mg) was transferred to a
glass vial and mixed with methanol (0.1 mL) containing 0.1%
acetic acid. To this, N-chlorosuccinimide (100 μg, 0.75 μmol in
10 μL methanol) and either NaI (100 μg, 0.67 μmol in 10 μL
water) or Na[125I] (6 μCi in 1 μL of 0.01 M NaOH) added. A
yellow color from molecular iodine formed in the tube contain-
ing NaI. The mixture was allowed to react for 30 min at room
temperature with slow rotation of the vial. The solvent contain-
ing the released iodobenzoic acid was then separated from the
resin and analyzed on HPLC. A primary UV and radioactive peak
eluting at 12.8 min (conditions A) was observed. Co-elution
of both the UV and radioactive peaks with an authentic
1
Robertson Microlit Laboratories (Madison NJ). H and 13C
NMR spectra were recorded on a Bruker 400 MHz NMR. Mass
spectrometry data was obtained using a Agilent 6300 ion trap
LC/MS. RadioHPLC was performed using a Knauer HPLC
system. HPLC chromatographic conditions (A) were as follows:
C-18 Microsorb-MV column 50 mm ꢀ 4.7 mm, flow rate of 2
mL/min, gradient of 10% MeCN with 0.1% TFA to 60% MeCN
with 0.1% TFA over 25 min, UV detection at either 254 or 210
nm, and radioactivity detection using an Ludlum sodium iodide
detector placed over the effluent line from the UV detector.
407
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412

Bioconjugate Chemistry
ARTICLE
3-iodobenzoic acid chromatographic standard was used to con-
firm identity. ESI-MS analysis from incubation of resin (4) with
excess cold iodine is given in the Supporting Information.
The degree of loading of the resin by the alkyltin benzoic acid
was determined by incubating a preweighed sample of the resin
overnight in neat TFA to cleave the tin-aromatic bond, followed
by quantitation of the released benzoic acid using HPLC. Using
this approach, a loading value of 0.32 mmol/g was obtained, rep-
resenting about half the initial substitution of the resin. Due to
the prior incubation step in aqueous sodium hydroxide the
remaining unoccupied tin sites on the resin are likely to be predo-
minately in the form of the alkyltin oxide or hydroxide.
Polymer-Bound N-Succinimidyl 3-(Dibutylstannyl)benzoate
(5). To a suspension of polymer-bound 3-(dibutylstannyl)benzoic
acid 4 (5 mg) in DMF (0.2 mL) in a glass vial was added
4-dimethylaminopyridine (DMAP; 4 mg, 33 μmol), N,N,N0N0-
tetramethyl-O-(N-succinimidyl)uronium tetrafluoroborate (TSTU;
10 mg, 33 μmol) and benzoic acid (1 mg, 8 μmol). The purpose of
the benzoic acid was to cap any free amines present in the resin,
which may have developed during the prior alkaline hydrolysis step.
The mixture was allowed to react for 4 h at room temperature with
slow rotation of the vial. The resin was then washed twice with
DMF and stored in toluene at -20 °C.
3-[125I]iodobenzoate 7 was stored at -20 °C until needed for
coupling reactions. The N-succinimidyl 3-[125I]iodobenzoate
was stable when stored in this manner over several weeks. Co-
elution of the radioactive peak with a N-succinimidyl 3-iodo-
benzoate chromatographic standard on HPLC was used to
confirm identity.
Conjugation of N-Succinimidyl 3-[125I]Iodobenzoate (7)
to a Plant Peptide. An aliquot of the N-succinimidyl 3-[125I-
[iodobenzoate 7 (4 μCi, 148 KBq) in toluene was added to a vial
containing a trace (1 μL) of acetic acid to stabilize the activated
ester group and evaporated to dryness. To the vial was then
added tomato plant systemin peptide (100 μg in 10 μL water)
and bicarbonate buffer (10 μL of 100 mM aq solution, pH 8).
The mixture was allowed to react for 1 h at room temperature
and then analyzed using HPLC. Co-elution of the radioactive
peaks (tR 7.5-9.5 min; HPLC conditions (A) with a systemin-
iodobenzamide chromatographic standard was used to confirm
identity.
Conjugation of N-Succinimidyl 3-[125I]Iodobenzoate (7)
to Proteins. An aliquot of the N-succinimidyl 3-[125I-
[iodobenzoate 7 (5 μCi, 185 KBq) in toluene was evaporated to
dryness in the presence of a trace (1 μL) of acetic acid . To the vial
was then added bicarbonate buffer (50 μL of 100 mM aq solution,
pH 8) and the protein to be labeled (bovine serum albumin or IgG;
200 μg in 10 μL water). The mixture was allowed to react for 2 h at
rt. A sample of the mixture was then transferred to ultrafiltration
tubes (Centricon, 3500 MW cutoff) together with 2 mL of aq
bicarbonate buffer containing bovine serum albumin (100 μg/mL)
and benzoic acid (2 mg/mL). The purpose of the latter two
additives was to inhibit any nonspecific absorption of either the
To test for formation of polymer-bound N-succinimidyl 3-
(dibutylstannyl)benzoate, a sample of the resin was reacted with
either NaI or Na[125I] and N-chlorosuccinimide and the solvent
containing the released N-succinimidyl 3-iodobenzoate analyzed
by HPLC. A primary peak eluting at 15.6 min (conditions A) was
observed. Co-elution of both the UV and radioactive peaks with
an N-succinimidyl 3-iodobenzoate chromatographic standard
was used to confirm identity.
[
125I]-labeled protein or unreacted [125I]iodobenzoic acid to the
Typical Radioiodination Procedure for 3-[125I]Iodobenzoic
Acid (6). Between 0.5 and 1 mg of the resin-bound tin precursor
was placed in a 1.5 mL glass tube to which was added methanol (1
mL) containing sodium acetate (100 mM) and acetic acid (16
mM). The tube was slowly rotated for 1-2 h to allow the resin to
equilibrate with the solvent and wash out any impurities present in
the resin. The solvent was then aspirated off and the resin washed
briefly with fresh solvent (1 mL). The resin was then resuspended
in fresh methanol (100 μL) containing sodium acetate and acetic
acid. N-Chlorosuccinimide (10 μg, 0.075 μmol in 1 μL of metha-
nol) and Na125I (10-200 μCi, 0.37-7.4 MBq, in 1 μL of 0.01 M
NaOH) was added to the tube. The tube was sealed and the tube
placed at 45° angle with slow rotation (10 rpm) to ensure efficient
mixing of the resin beads with the small amount of solvent in the
bottom of the tube. After allowing the mixture to react for 50 min at
room temperature, the solvent containing the released radioiodi-
nated compounds was separated from the resin beads and trans-
ferred to a new vial. A sample of the mixture was injected into the
HPLC for analysis. Co-elution of the radioactive peak with a
3-iodobenzoic acid chromatographic standard was used to confirm
identity.
ultrafiltration filters. The tubes were centrifuged at 6000 ꢀ g for 1 h
after which an additional 2 mL of water was added to the top of the
filter assembly and the tubes recentrifuged. Following the second
centrifugation, theretentate(approximately0.2mL) andcombined
filtrate (3.8 mL) were withdawn from the spin columns, transferred
to separate tubes, and counted in a gamma counter for 125I.
Chromatographic Standards. Synthesis and characteriza-
tion of N-succinimidyl 3-iodobenzoate and systemin-iodoben-
zamide conjugate chromatographic standards are given in the
Supporting Information.
’ RESULTS AND DISCUSSION
Synthesis of Polymer-Bound 3-(Dibutylstannyl)benzoic
Acid and N-Succinimidyl 3-(Dibutylstannyl)benzoate. The
synthetic scheme for producing polymer-bound 3-(dibutyl-
stannyl)benzoic acid 4 and its succinimidyl ester derivative 5
is shown in Scheme 1. For the resin support, we employed a
PEG-based support (NovaPEG) rather than the more tradi-
tional polystyrene based resins (e.g., Merrifield resin) due to the
fact the PEG-based resins are able to swell in water and
alcoholic solvents, as well as organic solvents. This is in contrast
to the polystyrene-based resins which show little to no swelling
in water and alcohol solvents due to their strongly hydrophobic
backbone. Good swelling in alcohol was considered advanta-
geous for the radioiododestannylation reaction, since this
reaction is normally performed in a methanol solvent. Creation
of the polymer-bound dibutyltin chloride 2 was performed
using an adaptation of the methods described by Newmann et
al.13 and Zhu et al.15 A terminal alkene group was introduced
onto the amino-PEG resin, and this was then stannylated using
Typical Radioiodination Procedure for N-Succinimidyl 3-
[
[
125I]Iodobenzoate (7). Procedure was as described for 3-
125I]iodobenzoic acid except that preincubation and radioiodi-
nation were performed in methanol-containing acetic acid (0.1%;
16 mM) only. Following the radioiodination, sodium metabi-
sulfite (100 μg in 10 μL water) and 0.4 mL of either water or aq
phosphate buffer (pH 6) was added to the vial and the radio-
iodinated N-succinimidyl 3-[125I]iodobenzoate 7 extracted into
toluene (2 ꢀ 0.5 mL). After drying over a few crystals of MgSO4,
the toluene extract containing the radioiodinated N-succinimidyl
408
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412

Bioconjugate Chemistry
ARTICLE
Scheme 1. Preparation of Polymer-Bound 3-(Dibutylstannyl)benzoic Acid (4) and Its Succinimidyl Ester (5)a
a Reagents and conditions: (a) acryloyl chloride, TEA, DMF, room temperature, 2 h; (b) Bu2SnCl2, Bu2SnH2, AIBN, THF, light, room temperature,
overnight; (c) 3-(ethoxycarbonyl)phenylzinc, THF, room temperature, overnight; (d) aq NaOH, 55 °C, 4 h; (e) TSTU, DMAP, benzoic acid, DMF,
room temperature, 4 h.
Bu2SnHCl together with AIBN and light irradiation. The
Bu2SnHCl was prepared in situ from a 1:1 mixture of Bu2SnH2
and Bu2SnCl2.
To enable the released radiolabeled benzoic acid to couple to
proteins, we chose a succinimidyl ester as the activating group.
This activating group has a combination of reasonable water
solubility when at tracer levels and good reactivity to amine
groups. Although the succinimidyl ester group could in principle
be added once the radiolabeled benzoic acid had been released
from the resin in solution, we chose instead to convert the
benzoic acid to an activated ester directly on the resin, since this
would simplify radiolabeling of biological macromolecules
using the solid-phase approach and would facilitate enabling
the resin to be developed in a kit form for protein radiolabel-
ing. However, a drawback of this preactivation approach is that
resin is likely to be less stable in storage when in the activated
state, and also care has to be taken for the succinimidyl ester
not to be prematurely hydrolyzed during incubation of the
resin in radioiodine. For the activation of the resin-bound
alkyltin benzoic acid, we initially used hydoxysuccinimide in
the presence of DIC and base. However, only a partial con-
version to the succinimidyl ester 5 was obtained using this
approach, despite several variations in reaction conditions.
Subsequently, we used the uronium-based activating agent, N,
N,N0N0-tetramethyl-O-(N-succinimidyl)uronium tetrafluoro-
borate (TSTU) to produce the succinimidyl ester 5. This gave
a more satisfactory conversion. Following conversion of the resin to
the succinimidyl ester, the final polymer-bound N-succinimidyl
3-(dibutylstannyl)benzoate 5 was stored in toluene at -20 °C until
needed for radioiodinations.
Radioiodination of Polymer-Bound Precursors. Initial
tests to optimize reaction conditions for the release of [125I]-
labeled compounds following incubation of the resin with
electrophilic radioiodine were conducted using the polymer-
bound alkyltin benzoic acid resin 4 (Scheme 2a). Figure 1 shows
an HPLC analysis of the solvent after incubating polymer-bound
alkyltin benzoic acid resin 4 with radioiodine in a methanol
solvent containing acetic acid and sodium acetate. Most of the
free radioiodine, which elutes close to the void volume, has
disappeared, and a large peak from 3-[125I]iodobenzoic acid
released into the solution is apparent. The absence of significant
peaks in the UV trace suggested a good chemical purity for
Although characterization of organotin bound to polystyrene-
based resins has been previously performed using gel-phase 13C
NMR, we had limited success using this approach with the PEG-
based resin used as the polymer support in the current study. This
was due the fact that large background peaks from the PEG resin
made it difficult to distinguish peaks coming from the terminal
functionalization of the resin. Similar problems were encountered
when trying to analyze the resin using IR. We thus relied on ele-
mental analysis of tin and chloride to determine the degree of
stannylation of the resin. Following tethering of aromatic com-
pounds to the resin, the progress of reactions was monitored by
cleaving off the compounds from resin samples using molecular
iodine and assaying the released products with HPLC.
Once the polymer-bound chloride was synthesized, it was used
to generate a resin-bound dibutyltin ethyl benzoate 3 via transme-
talation of a commercially available zinc precursor. In order to
cleave the ethyl ester and generate a free carboxylic acid on the
molecule, necessary for producing a reactive functional group for
subsequent conjugation to amine-containing biomolecules, the
resin was incubated in aqueous sodium hydroxide to give 4. HPLC
analysis on samples of the reaction products that had been cleaved
off the resin using iodine indicated a > 95% hydrolysis of the ester.
Interestingly, in parallel experiments in which we used Merrifield
resin rather than NovaPEG as the polymer support, treatment of
the polymer-bound dibutyltin ethyl benzoate with aqueous sodium
hydroxide did not produce cleavage of the ester, even with over-
night incubation in the hydroxide. PEGA resin, which consists of a
polystyrene backbone on which PEG is grafted, was also found
to be unsatisfactory as the polymer support since incubation in
sodium hydroxide caused it to become brittle and fragile. The
NovaPEG resin by contrast appeared unaffected. Following
the hydrolysis step, the resin was washed, dried, and stored in a
darkened desiccator at room temperature until needed for
radioiodinations or for conversion to an activated ester. The
resin showed good stability under these conditions when stored
over a period of several months.
409
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412

Bioconjugate Chemistry
ARTICLE
Scheme 2. Radioiodination of Polymer-Bound Organotin Precursors and Products Formed (a) Polymer-Bound 3--
(Dibutylstannyl)benzoic Acid (4), (b) Polymer-Bound N-Succinimidyl 3-(Dibutylstannyl)benzoate (5)
Table 1. Radiochemical Yields and Purities for Production of
3-[125I]Iodobenzoic Acid (6) from Incubation of the Poly-
mer-Bound Precursor with Na[125I] and N-Chlorosuccini-
mide in Different Solvent Systems
radiochemical
yield (%)
radiochemical
purity (%)
solvent
MeOH, 16 mM HAc
MeOH, 16 mM HAc, 100 mM NaAca
51.8
59.2
74.8
68.3
66.8
65.5
91.9
95.0
93.4
93.7
94.5
80.5
MeOH, 100 mM NaAc
Water, 100 mM NaPO4, pH 6.2
Water, 100 mM NaPO4, pH 7.0
Water, 100 mM NaPO4, pH 8.0
a HPLC shown in Figure 1.
with the methanol solvent, despite the fact that tributyltin
compounds are normally insoluble in the former. This was
presumably a consequence of conjugation of the alkyl tin to
the hydrophilic PEG support. For the methanol-based sol-
vents, best yields (75%) were obtained with a relatively basic
rather than acidic solvent. In the case of the aqueous phos-
phate-buffered solvent, similar radiochemical yields were
obtained at the three different pH values tested. However, at
the most alkaline pH tested, a free radioiodine peak eluting
close to the void volume in the HPLC chromatogram was
observed, which hence reduced the overall radiochemical
purity of the released 3-[125I]iodobenzoic acid.
In the case of the succinimidyl ester-activated resin 5, a mildly
acidic reaction medium was chosen as the radioiodination
solvent, despite a lower total yield of radioiodinated products
compared with a more basic solvent (Scheme 2b). This was to
avoid premature loss of the activated ester group under basic
conditions. However, because of the potential for cleavage of the
tin-aromatic bond under strongly acidic conditions, the level of
acetic acid in the solvent was minimized. For the iodination
reaction, we thus chose 0.1% (16 mM) acetic acid in methanol as
providing a suitable compromise between providing sufficient
acid to neutralize sodium hydroxide in which the Na[125I] is
Figure 1. HPLC analysis of the reaction medium following incubation
of polymer-bound 3-(dibutylstannyl)benzoic acid (4) with Na[125I] and
N-chlorosuccinimide. HPLC conditions A.
the released radioactive compound. Radiochemical yield for
3-[125I]iodobenzoic acid using a methanol with acetic acid
solvent typically ranged from 50% to 65% over different runs,
with the remaining radioactivity found to be mostly absorbed to
the resin.
A comparison of the effect of several different solvent condi-
tions on radiochemical yield and radiochemical purity of the
released 3-[125I]iodobenzoic acid is shown in Table 1. Both
methanol-only and fully aqueous solvent systems were exam-
ined. Acid versus basic conditions were adjusted using acetic
acid and sodium acetate for the methanol solvent and sodium
phosphate for the aqueous solvent. Interestingly, a similar
radiochemical yield was obtained using an aqueous solvent as
410
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412

Bioconjugate Chemistry
ARTICLE
Table 2. Labeling Yields, as Indicated by the Percentage of
Radioactivity Retained after Ultrafiltration of the Reaction
Mixture, from Incubation of N-Succinimidyl 3-[125I]Iodo-
benzoate (7) or Unactivated 3-[125I]Iodobenzoic Acid (6)
with Bovine Serum Albumin or IgG Protein
radiolabeling
group
radioactivity
retained (%)
protein
N-succinimidyl 3-[125I]iodobenzoate
N-succinimidyl 3-[125I]iodobenzoate
3-[125I]iodobenzoic acid (control)
Bovine serum albumin
Goat IgG
Bovine serum albumin
67.6
67.3
2.5
Figure 2. HPLC analysis of the reaction medium following incubation
of polymer-bound N-succinimidyl 3-(dibutylstannyl)benzoate (5) with
Na[125I] and N-chlorosuccinimide. HPLC conditions A.
shipped and thus preserving the succinimidyl ester group, but at
the same time minimizing the potential for concomitant acid-
olysis of the tin-aromatic bond. HPLC analysis of the reac-
tion solution following incubation of the resin in radioiodine and
with N-chlorosuccinimide oxidant is shown in Figure 2. In the
radioactivity trace, a large peak from released N-succinimidyl
3-[125I]iodobenzoate 7 was apparent. Smaller peaks from un-
esterified 3-[125I]iodobenzoic acid and a late-running peak from
a radiochemical impurity (at 22 min) were also present. The
identity of the late-running impurity was not established, but
it was observed to remain unchanged following hydrolysis of
the released N-succinimidyl 3-[125I]iodobenzoate with bicar-
bonate buffer. Some trace levels of chemical impurities were
observed on the UV trace in different runs but were at or close
to the detection limits of the UV and were not further
characterized. Median radiochemical yield for N-succinimi-
dyl 3-[125I]iodobenzoate over several runs was 35% (range
28-36%) at low levels of radioiodine (5-10 μCi, 0.19-0.37
MBq), but was slightly higher at 44% for a run performed with
a higher level (1 mCi, 37 MBq) of radioiodine (Supporting
Information). A slightly lower radiochemical yield of the
succinimidyl ester compared to the free acid can be explained
by the electron withdrawing effect of the activated ester
reducing the susceptibility of the aromatic tin to electrophilic
attack, combined with the fact that not all of the free acid sites
on the resin were completely converted to the succinimidyl
ester by the TSTU activating agent. Following the radio-
iodination reaction, the released N-succinimidyl 3-[125I]-
iodobenzoate was extracted into toluene and was stable in
this formulation when stored in the freezer.
Figure 3. HPLC analysis of the reaction medium following incubation
of N-succinimidyl 3-[125I]iodobenzoate (7) with tomato plant systemin
peptide. Unlabeled iodobenzoic acid (10 μg) included as a chromato-
graphic standard.
was performed in an aqueous bicarbonate buffer followed by
ultrafiltration through a MW 3500 cutoff membrane to separate
the radiolabeled protein from unbound [125I]]iodobenzoate.
Using this approach, yields of 67% relative to the added N-
succinimidyl 3-[125I]iodobenzoate radioactivity were obtained
for both proteins. Unactivated 3-[125I]iodobenzoic acid was not
significantly retained by the filters. For conjugation to tomato
plant systemin, yields of 69% and 92% for the combined peaks
and relative to the added N-succinimidyl 3-[125I]iodobenzoate
radioactivity were obtained in two separate runs, as determined
by HPLC analysis of the peptide conjugate. Several individual
[
125I] peaks could be resolved in the HPLC chromatogram. This
probably reflects labeling of different amine residues on the
peptide, since this water-soluble peptide contains 3 lysine amine
groups and one terminal amino group,23 although the possibility
of partial conjugation to co-eluting truncated impurities that may
have been present in the original peptide sample also exists.
Conjugation of N-Succinimidyl 3-[125I]Iodobenzoate to
Biomolecules. To test the ability of N-succinimidyl 3-[125I]-
iodobenzoate to couple an amine-containing biological macromo-
lecules, its conjugation to the proteins bovine serum albumin (BSA)
and goat antimouse IgG (Table 2), to the peptide, tomato plant
systemin (Figure 3), and to an aminoglycoside compound (Support-
ing Information) was examined. Conjugation to BSA and IgG
’ CONCLUSIONS
In summary, we have examined an approach for conjugate
prosthetic group labeling of protein and peptide labeling in which
the organotin precursor for the radioiodinated prosthetic group
411
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412

Bioconjugate Chemistry
ARTICLE
is tethered to a solid support via the tin leaving group. Exposing
the polymer-bound organotin precursor to electrophilic radio-
iodine releases the radiolabeled prosthetic group into solution,
with the unreacted organotin precursor and organotin bypro-
ducts remaining bound to the resin. This approach thus simplifies the
process of radiolabeling of proteins and peptides via the conjugate
radiolabeled group approach by reducing the need for postiodina-
tion purification of the prosthetic group prior to conjugating it to
biological macromolecules. By avoiding the need for HPLC, the
quantity of radioactive waste generated in the labeling process is also
greatly reduced. This solid-phase approach may have additional
applications for development of a simple kit formulation for radio-
chemical labeling of protein and peptide molecules or for develop-
ment of automated radiolabeling processes.
(8) Zalutsky, M. R., and Narula, A. S. (1987) A method for the
radiohalogenation of proteins resulting in decreased thyroid uptake of
radioiodine. Appl. Radiat. Isot. 38, 1051–1055.
(9) Wilbur, D. S., Hadley, S. W., Hylarides, M. D., Abrams, P. G.,
Beaumier, P. A., Morgan, A. C., Reno, J. M., and Fritzberg, A. R. (1989)
Development of a stable radioiodinating reagent to label monoclonal-
antibodies for radiotherapy of cancer. J. Nucl. Med. 30, 216–226.
(10) Zalutsky, M. R., and Narula, A. S. (1988) Radiohalogenation of
a monoclonal-antibody using an N-succinimidyl 3-(tri-n-butylstan-
nyl)benzoate intermediate. Cancer Res. 48, 1446–1450.
(11) Moerlein, S. M., Mathis, C. A., and Yano, Y. (1987) Compara-
tive evaluation of electrophilic aromatic iododemetallation techniques
for labeling radiopharmaceuticals with iodine-122. Int. J. Rad. Appl.
Instrum. A 38, 85–90.
(12) Harrington, C. F., Eigendorf, G. K., and Cullen, W. R. (1996)
The use of high-performance liquid chromatography for the speciation
of organotin compounds. Appl. Organomet. Chem. 10, 339–362.
(13) Neumann, W. P. (1992) Tin for organic-synthesis 0.6. the new
role of organotin reagents in organic-synthesis 0.1. stannyl groups as
leaving groups in electrophilic substitutions 0.2. polymer-supported
organotin reagents for organic-synthesis. J. Organomet. Chem. 437, 23–
39.
(14) Hernan, A. G., Horton, P. N., Hursthouse, M. B., and Kilburn,
J. D. (2006) New and efficient synthesis of solid-supported organotin
reagents and their use in organic synthesis. J. Organomet. Chem. 691,
1466–1475.
(15) Zhu, X. Z., Blough, B. E., and Carroll, F. I. (2000) Synthesis and
reactions of a novel chlorostannane resin: coupling with functionalized
organozinc halides. Tetrahedron Lett. 41, 9219–9222.
(16) Hunter, D. H., and Zhu, X. Z. (1999) Polymer-supported
radiopharmaceuticals: [131I]MIBG and [123I]MIBG. J. Label. Compd.
Radiopharm. 42, 653–661.
(17) Vaidyanathan, G., Affleck, D. J., Alston, K. L., Zhao, X. G., Hens,
M., Hunter, D. H., Babich, J., and Zalutsky, M. R. (2007) A kit method
for the high level synthesis of [211At]MABG. Bioorg. Med. Chem. 15,
3430–3436.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures and
b
analytical characterization of organotin and peptide chromato-
graphic standards. ESI-MS analysis for an example of a solution-
phase iododestannylation reaction and from cold (unlabled)
iodination of resin (4). HPLC radiochromatograms from high-
activity radioiodination of resin (5) and from conjugation of (7)
to an aminoglycoside compound. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Corresponding author. Dr. A. N. Gifford, Medical Department,
Brookhaven National Laboratory, Upton, NY 11973. Tel 631-
344-7069, Fax 631-34-5311, e-mail gifforda@bnl.gov.
(18) Kabalka, G. W., Namboodiri, V., and Akula, M. R. (2001)
Synthesis of 123I labeled Congo Red via solid phase organic chemistry.
J. Label. Compd. Radiopharm. 44, 921–929.
(19) Spivey, A. C., Tseng, C. C., Jones, T. C., Kohler, A. D., and
Ellames, G. J. (2009) A method for parallel solid-phase synthesis of
iodinated analogues of the CB1 receptor inverse agonist Rimonabant.
Org. Lett. 11, 4760–4763.
(20) Yong, L., Yao, M. L., Green, J. F., Kelly, H., and Kabalka, G. W.
(2010) Syntheses and characterization of polymer-supported organotri-
fluoroborates: applications in radioiodination reactions. Chem. Commun.
46, 2623–2625.
(21) Donovan, A., Forbes, J., Dorff, P., Schaffer, P., Babich, J., and
Valliant, J. F. (2006) A new strategy for preparing molecular imaging and
therapy agents using fluorine-rich (fluorous) soluble supports. J. Am.
Chem. Soc. 128, 3536–3537.
’ ACKNOWLEDGMENT
Funded by the DOE office of biological and environmental
research and performed under Brookhaven Science Associates
contract No. DE-AC02-98CH1-886 with the U. S. Department
of Energy. The authors would like to thank Dr. S. Kim for
valuable advice and for collecting NMR data on the compounds.
’ REFERENCES
(1) Bolton, R. (2002) Radiohalogen incorporation into organic
systems. J. Label. Compd. Radiopharm. 45, 485–528.
(2) Adam, M. J., and Wilbur, D. S. (2005) Radiohalogens for imaging
and therapy. Chem. Soc. Rev. 34, 153–163.
(3) Yamada, A., Traboulsi, A., Dittert, L. W., and Hussain, A. A. (2000)
Chloramine-T in radiolabeling techniques - III. Radioiodination of biomo-
lecules containing thioether groups. Anal. Biochem. 277, 232–235.
(4) Crim, J. W., Garczynski, S. F., and Brown, M. R. (2002)
Approaches to radioiodination of insect neuropeptides. Peptides 23,
2045–2051.
(5) Garg, P. K., Alston, K. L., and Zalutsky, M. R. (1995) Catabolism
of radioiodinated murine monoclonal-antibody F(Ab0)2 fragment la-
beled using N-succinimidyl 3-iodobenzoate and iodogen methods.
Bioconjugate Chem. 6, 493–501.
(22) Donovan, A. C., and Valliant, J. F. (2009) Fluorous isocyanates:
convenient synthons for the preparation of radioiodinated compounds
in high effective specific activity. J. Org. Chem. 74, 8133–8138.
(23) Pearce, G., Strydom, D., Johnson, S., and Ryan, C. A. (1991) A
polypeptide from tomato leaves induces wound-inducible proteinase-
inhibitor proteins. Science 253, 895–898.
(6) Nestor, M., Persson, M., Cheng, J. P., Tolmachev, V., van
Dongen, G., Anniko, M., and Kairemo, K. (2003) Biodistribution of
the chimeric monoclonal antibody U36 radioiodinated with a closo-
dodecaborate-containing linker. Comparison with other radioiodination
methods. Bioconjugate Chem. 14, 805–810.
(7) Wilbur, D. S. (1992) Radiohalogenation of proteins - an over-
view of radionuclides, labeling methods, and reagents for conjugate
labeling. Bioconjugate Chem. 3, 433–470.
412
dx.doi.org/10.1021/bc1004203 |Bioconjugate Chem. 2011, 22, 406–412