Content uploaded by Qi-Huang Zheng
Author content
All content in this area was uploaded by Qi-Huang Zheng on Sep 19, 2014
Content may be subject to copyright.
1
RADIOPHARMACEUTICAL CHEMISTRY: IODINATION TECHNIQUES
INTRODUCTION
The labeling of compounds with radioiodine has a long and varied history in biomedical
research and the practice of nuclear medicine. In its simplest radiochemical form, sodium
[131I]iodide (t½ = 8.05 d) has had tremendous impact on the diagnostic evaluation of thyroid
function in vivo as well as in the treatment of hyperthyroidism and thyroid cancer. Dramatic
improvement in diagnostic image quality and marked reduction in patient absorbed radiation
dose were achieved when 123I (t½ = 13.3 h) was introduced into clinical practice. However, the
application of any radioisotope of iodine is generally limited to the thyroid gland as long as the
chemical form is restricted to the iodide (I-) anion.
Fortunately, the chemistry of iodine is sufficiently well understood to enable a wide
variety of chemical and biologic compounds to be labeled with this versatile tracer. By tagging
different carrier molecules, the diagnostic and/or therapeutic potential of radioiodine can be
extended to many other organ systems. Unlike with [99mTc]technetium, one can develop and
synthesize iodinated compounds at the macro scale using stable [127I]iodine. Chemical, physical,
and pharmacologic properties of these compounds can then be characterized without the
constraints inherent in the use of radioactive materials. When it is time to synthesize the
compounds at the micro scale, perform in vitro binding assays, or perform in vivo biodistribution
studies in animals, the radioisotope 125I (t½ = 60.2 d) can be used as the tracer. Similarly,
compounds that show promise as diagnostic agents can then be labeled with a more suitable
imaging isotope, 123I for SPECT imaging, or 124I (t½ = 4.15 d) for PET imaging, and those that
offer radiotherapeutic potential can be labeled with 131I.
2
Tracer theory dictates that all radioisotopes of iodine should behave in an identical
manner. However, it is overly simplistic to assume that one can switch back and forth between
the various radioisotopes of iodine without encountering some difficulty. Half-life differences
are most obvious. Lengthy labeling and purification methods or the evaluation of biologic
processes that take days to complete preclude the use of short half-life tracers like 123I. Chemical
purity differences, arising from the various target processing and isotope recovery procedures,
drastically affect the radiolabeling kinetics of reactions. Cost and availability differences may
dictate the appropriate radiotracer to use, especially if the labeling efficiency is low for a given
method. Lastly, the shelf-life of compounds susceptible to radiolytic decomposition might be
difficult to maintain if beta-emitting 131I or positron-emitting 124I is used as the label.
The tremendous versatility of radioiodine as a tracer for biologic molecules is also one of
its liabilities. There is no single radiolabeling method that is the best, the easiest, or the most
universally used. The type of molecule to be labeled may suggest a labeling approach. For
example, a compound that already contains a stable iodine atom covalently attached to an
aromatic ring is an ideal candidate for one of the isotope exchange labeling methods. A protein
molecule with an abundance of tyrosine residues is a good candidate for the addition of
radioiodine oxidized by chloramine-T. Long-chain molecules with carbon-to-carbon double
bonds might be radiolabeled at the site of the double bond using either iodine monochloride or
elemental radioiodine. However, even a casual literature search reveals that, shortly after a
specific radiolabeling method is introduced for a given molecule, other articles appear that
describe various modifications to the new method, each purporting to be easier, faster, better, or
more reproducible than the original. The new investigator wishing to use the same radiolabeled
3
molecule in a different research project is faced with a difficult question: "Which method should
I use to label this molecule?"
The purpose of this chapter is to introduce the main strategies commonly used to
radioiodinate molecules of interest in nuclear medicine research. For details of specific labeling
procedures, the reader is directed to delve into the scientific literature (and make a few phone
calls) before attempting to insert radioiodine into anything other than a shielded container.
PROTEIN IODINATION
Protein molecules labeled with radioactive iodine have long been used in clinical nuclear
medicine and basic biomedical research. One of the first lung scanning agents was macro-
aggregated human serum albumin labeled with 131I. This was a crude preparation in that the
albumin molecules were deliberately heat-denatured to form insoluble particles of sufficient size
to become trapped in the pulmonary vasculature after intravenous injection. Whatever chemical
damage that may have been caused by the radiolabeling procedure was inconsequential when
compared to the denaturation.
Non-denatured human serum albumin labeled with 125I (RISA) is still used clinically for
the determination of plasma volume. Because the intent now is to have the radiopharmaceutical
agent remain in circulation for sufficient time to perform the plasma volume determination,
chemical damage caused by the labeling process becomes much more important. Since plasma
volume is calculated according to the isotope dilution principle, alterations to the soluble protein
that cause accelerated plasma clearance of the tracer lead to lower than expected blood
concentrations and falsely high plasma volume results. In addition, the long physical half-life of
125I requires that greater care be given to maintaining the in vitro stability of the product.
4
Fortunately, this agent is still available commercially, so investigators rarely need to label it
themselves. It is, however, a good model protein on which to test various radioiodination
methods.
Monoclonal antibodies are commonly labeled with radioiodine in the research laboratory
to help determine the antibody's specificity for and binding affinity to a given antigen. Whether
being developed for use as a radioimmunoassay reagent for in vitro testing or for in vivo use as a
diagnostic or therapeutic agent, even greater care and testing is required to maintain normal
immunoreactivity of the protein following radiolabeling and purification. However, no matter
how gentle the method and how high the labeling efficiency, if the location of iodine attachment
is at tyrosine or histidine residues near the antigen binding sites, the final product would probably
be worthless. In addition, high specific activity now becomes important to achieve because of
the necessity to detect very low concentrations of specific antigen.
Direct Addition of Radioiodine
Keeping the previously mentioned concerns in mind, let us now look at a few of the
methods generally used to attach radioiodine directly to protein molecules. As has already been
stated, the primary site for iodine addition to a protein is on the activated phenolic ring of
tyrosine amino acid residues (Fig. 1). If the pH exceeds 8.5, the secondary site on the imidazole
ring of histidine is favored. To achieve iodination, the radioactive iodide anion (I-) must first be
oxidized to I+ while in the presence of the soluble protein. In its reduced form, the iodide anion
will not attack and label the protein molecule. At the pH normally used for protein iodination,
pH 7 to pH 9, the oxidized I+ electrophilic species forms unstable complexes with water. The
reactive iodine species, therefore, is thought to be the hydrated iodonium ion, H2OI+, and/or
hypoiodous acid, HOI. With tyrosine, substitution of a hydrogen ion with the reactive iodonium
5
ion occurs ortho- to the phenolic hydroxyl group. With histidine, substitution occurs at the 2-
position of the five-member imidazole ring. Following the desired incubation period, residual
reactive I+ species are reduced back to the I- form and removed from the product solution by
passage through either an anion exchange resin column or a gel filtration column. In this
manner, high radiochemical purity can be obtained even if the labeling efficiency is low.
Soluble Oxidants. A variety of oxidizing agents have been used to convert the iodide
anion into a reactive I+ labeling species. Perhaps the most frequently used agent is chloramine-
T (Fig. 2A), the sodium salt of N-monochloro-p-toluene-sulfonamide.6 In aqueous solution,
chloramine-T slowly breaks down to form hypochlorous acid, HOCl, the specific oxidant.
Chloramine-T is a very effective oxidizing agent for radioiodine in that only a few micrograms
are required to achieve nearly quantitative iodinations. However, if the incubation times are
longer than a few seconds, it is also quite capable of causing significant damage to the protein
molecules, such as thiol oxidation, chlorination of aromatic rings and/or primary amines, and
peptide bond cleavage.21 Typically, the protein solution is mixed with radioactive iodide in a
slightly alkaline buffer (pH 7.5). A freshly prepared solution of chloramine-T is added to this
mixture. To stop the reaction, a slight molar excess of reducing agent such as sodium
metabisulfite (Na2S2O5) is immediately added to the mixture to inactivate the chloramine-T. It is
important to remember that the reducing agent used to quench the oxidation can also cause
cleavage of the disulfide bridges within the protein molecules themselves and alter the tertiary
structure. So while it is relatively easy to obtain radioiodination efficiencies greater than 90%
with chloramine-T, it is often even easier to end up with a chemically damaged product.
6
One strategy used to reduce the oxidative damage to proteins caused by chloramine-T was
to develop soluble N-chloro-secondary amines with lower chlorine potential.11 As one example,
N-chloro-morpholine was found to produce higher radioiodination yields and less degradation
than chloramine-T when reacted with L-tyrosine or leucine enkephalin (Leu-Gly-Gly-Phe-Leu).8
However, due to its own instability, this milder oxidizing reagent had to be prepared in situ prior
to introduction of the substrate. Instead, penta-O-acetyl-N-chloro-N-methylglucamine (NCMGE)
(Fig. 3) was found to be a stable, water-soluble reagent which could be added directly to the
buffered radioiodide and peptide reaction mixture. Comparative studies revealed that the lower
chlorine potential of NCMGE afforded higher radiochemical yield and less decomposition to
model amino acids and small peptides than did chloramine-T.22 N-chlorosuccinimide (NCS)
(Fig. 3) is a secondary amine oxidizing reagent which has been used extensively to radioiodinate
a number of activated arylstannane derivatives as a prelude to conjugate protein labeling.27 It is
not used, however, to radiolabel proteins directly because of its instability in aqueous media.30
Solid-State Oxidants. A more popular strategy employed to minimize the chemical
damage caused by the chloramine-T + metabisulfite method has been to remove the offending
reagents from the substrate solution itself. The simplest has been the covalent attachment of
chloramine-T to the surface of large polystyrene beads that can be easily removed from the
mixture with tweezers to stop the oxidative reaction (IODO-BEADS, Pierce Biotechnology,
Inc., Rockford, IL). The rate of radiolabeling is slower with this solid-state method because the
appropriate oxidizing environment for protein iodination exists only near the surface of the
beads, not throughout the incubation solution. The rate of oxidative protein damage is also
7
slower, but not eliminated. Reductive damage caused by metabisulfite is eliminated, however,
because none is used to stop the reaction. Convenience and commercial availability make the
IODO-BEAD method a popular choice for protein radioiodination.
Another popular solid-state alternative is IODO-GEN, also from Pierce Biotechnology,
Inc. Chemically known as 1,3,4,6-Tetrachloro-3α,6α-diphenyl-glycouril (Fig. 2B), it is basically
a water-insoluble equivalent of chloramine-T. By dissolving IODO-GEN in chloroform or
methylene chloride, one can plate aliquots of any amount on the interior walls of test tubes
merely by evaporating the solvent under a stream of nitrogen gas. The pre-coated tubes can be
stored in a desiccator until time of use. Radiolabeling is accomplished by adding the buffered
protein solution to the test tube, followed by the desired amount of sodium radioiodide solution.
Oxidation and protein labeling takes place along the walls of the vessel. As with IODO-BEADS,
the rate of reaction is slower than that with soluble chloramine-T and labeling efficiencies are
somewhat lower for the same reasons. However, the reaction is easily terminated merely by
decanting the incubation solution from the tube directly onto a gel purification column. Again,
no reducing agent is needed and minimal damage to the protein is observed. It has been reported,
however, that as much as 10-20% of the initial iodine radioactivity is consumed by self-
radioiodination of the reagent coating itself.24 Such loss would reduce overall yield of the
technique, but not affect radiochemical purity of the labeled protein. In addition, the rate of
iodide oxidation by IODO-GEN has been observed to vary according to the type of surface to
which the reagent was coated.3 Polypropylene test tubes resulted in the lowest rate of oxidation,
followed by borosilicate glass, with polar soda-lime glass giving the highest rate of oxidation.
Pierce Biotechnology, Inc. has recently introduced IODO-GEN pre-coated glass reaction tubes
8
into its product line (2002), thereby providing the user with more consistency in the amount and
uniformity of the chemical coating.
A modified IODO-GEN method has been described that introduces the water-insoluble
reagent to the labeling mixture in a novel manner.20 Rather than coat the inside surface of a
reaction test tube, a suspension of freshly prepared ~ 3-µm diameter IODO-GEN particles is
added to the labeling mixture. This stock suspension is prepared immediately prior to
radiolabeling by adding 100 µL of a 23 mM solution of IODO-GEN in acetone to 4 mL of
phosphate buffered saline, pH 8. Higher labeling efficiencies are reported for this method
compared to the conventional method. As mentioned, the reaction is stopped and the particulate
reagent removed from the final product merely by pouring the incubation mixture through a gel
purification column.
Yet another variant is to pre-coat the protein molecules in a larger aqueous reaction
solution with as little as 35 μg of IODO-GEN, introduced as a 10 mg/mL solution in
acetonitrile.25 By apparent precipitation of the reagent to the apolar regions of the monoclonal
antibody, an even closer proximity was established between the oxidant and the target protein
molecule. In only 3 minutes, a radiochemical yield of >85% was reported for high level I-131
radiolabeling with no loss of immunoreactivity. Following labeling, the IODO-GEN was
effectively removed from the protein by ascorbic acid reduction to the 3α,6α-diphenylglycouril
compound and the entire mixture subjected to gel column purification.
Enzymatic Oxidation. Another popular mild radioiodination technique that has been
frequently referenced used specific oxidative enzymes and nanomolar concentrations of hydrogen
peroxide (H2O2) instead of the harsher hypochlorous acid liberated by chloramine-T to oxidize
9
iodine. Hydrogen peroxide by itself can oxidize radioiodide and affect labeling, but only at
concentrations that also damage the protein. However, the enzyme lactoperoxidase catalyzes the
peroxide oxidation of iodide and permits extremely low H2O2 concentrations to be used. To
control the peroxide concentration, a very novel solid-state system (Enzymobeads) was
developed by Bio-Rad Laboratories of Richmond, CA. Small, insoluble resin beads were
covalently coated with a blend of two enzymes: lactoperoxidase and glucose oxidase. When a
buffered solution of protein and radioiodide was added to a suspension of the enzyme-coated
beads, no labeling occurred. However, upon the addition of a small quantity of glucose, an
interesting chain of events was initiated. The glucose oxidase enzyme used the glucose to
generate a small, steady amount of hydrogen peroxide at the surface of the beads. The
immobilized lactoperoxidase, in turn, catalyzed the chemical peroxide oxidation of iodide in the
solution. The oxidized radioiodine species then interacted to label the protein molecules at the
usual tyrosine positions. To stop the reaction, one merely centrifuged the beads to the bottom of
the test tube or decanted the entire mixture onto the top of a gel filtration column. With proper
column elution, the desired labeled protein molecules eluted first, the unreacted radioiodine was
retained within the gel, and the beads remained at the top of the column. Unfortunately, this
product was discontinued in the early 1990’s and, per Bio-Rad Laboratories, the technology was
not transferred to another commercial vendor.
Indirect Addition of Pre-labeled Prosthetic Group
There are times when it is not appropriate to radiolabel proteins by direct electrophilic
addition to tyrosine residues. Perhaps the number of residues is few; they may be buried within
the tertiary structure and not readily available; or they may be located at or near the active
binding site of the molecule and cannot be altered by iodination. Therefore, a number of
10
different labeling procedures have been developed to radioiodinate protein molecules at sites
other than tyrosine and histidine. The most common alternative approach has been to perform
classical chloramine-T radioiodination of a small tyrosine-like molecule and then to attach this
pre-labeled molecule to the free amino function of the lysine moieties in the protein molecule.
Overall radiolabeling efficiencies are lower with this approach for the simple reason that two
separate labeling reactions must be used. However, a more viable end-product is obtainable
because the protein is not exposed to the harsh oxidative conditions of the chloramine-T reaction.
The conjugation of the pre-labeled small molecule to the protein is a milder reaction.
Bolton and Hunter Reagent. An effective acylating agent for coupling to free amino
groups, N-succinimidyl 3-(4-hydroxyphenyl) propionate is commercially available (Fig. 2C).2 It
is radioiodinated using a rapid, sequential addition of buffered radioiodide, chloramine-T,
sodium metabisulfite, carrier sodium iodide, and benzene. The benzene is used to extract the
radiolabeled reagent since it hydrolyzes rapidly in the presence of water. Following extraction
and evaporation of the benzene, the desired protein solution is added and the mixture allowed to
react for 15 to 30 minutes in an ice bath. To prevent subsequent conjugation with other carrier
proteins, glycine solution may be added to consume any unreacted reagent prior to final product
purification with gel column chromatography.
Wood Reagent. Methyl-p-hydroxybenzimidate⋅HCl offers several advantages over the
Bolton-Hunter reagent. It is a water-soluble compound, which is moderately stable at pH 9.5 and
37°C (Fig. 2D). It too is labeled with radioiodine using a more standard chloramine-T method,
and couples directly to the epsilon-amino groups of the lysine residues. Because this reagent
11
maintains a positive charge, the labeled protein can also keep its native charge because the
epsilon-amino group of lysine is protonated at physiologic pH.29
Organometallic Intermediates. Native tyrosine and tyrosine-like prosthetic
intermediates are easily radioiodinated in vitro due to the activated nature of the phenolic residue.
However, they are just as easily dehalogenated in vivo, depending on the biodistribution of the
labeled protein.10 In one of the first attempts to circumvent this potential problem, a para-
iodobenzoate derivative was developed for conjugation to the lysine residues of antibody
molecules (Fig. 3).27 The precursor compound, N-succinimidyl 4-tri-n-butylstannylbenzoate, is
first radioiodinated in methanol using N-chlorosuccinimide as the oxidant. After quenching the
reaction with sodium thiosulfate (Na2S2O3), the methanol solvent is carefully evaporated with a
stream of nitrogen gas. A buffered solution of the protein to be labeled is then added to the
reaction vial for short room temperature incubation. The entire reaction mixture is then purified
with Sephadex gel column filtration. Animal biodistribution studies support the hypothesis that
less in vivo deiodination of the labeled antibodies occurs as a result of the different position of
the radioiodine atom on the bridging molecule.
The use of tri-n-butyl-tin to effect site-specific halogenation of small molecules has led to
widespread use of the arylstannanes as reagents of choice for synthesis of a variety of iodinated
and radioiodinated conjugates. In essence, one end of the molecule can be designed to enhance
the in vitro labeling efficiency and/or to test the in vivo stability of the radiolabel to metabolic
degradation, while the other end of the conjugate molecule can be modified to direct protein
radiolabeling at sites other than the commonly used lysine residues as previously mentioned. For
example, conjugates can be prepared which differ only in the position of the radioiodine atom on
12
the aromatic ring. Differing length linkers can be placed between the radiolabeled aromatic ring
and the protein coupling moiety to study the effect of spacing on radiolabeled protein stability.
Likewise, identical conjugates can be selectively attached to the cysteine residues of proteins
instead of the lysine residues by using N-hydroxymaleimide to create the coupler portion of the
conjugate molecule instead of N-hydroxysuccinimide. For an exhaustive discussion of the state-
of-the-art in radiohalogenation of proteins, specifically conjugate labeling, the reader is strongly
encouraged to reference the excellent review article by DS Wilbur.26
ISOTOPE EXCHANGE REACTIONS
There are many small organic molecules that are of interest as tracers for nuclear
medicine. Several of them can or do contain an iodine atom covalently bound to a known
position on the molecule, as opposed to on a tyrosine or histidine moiety somewhere on a protein
molecule. However, directing a radioiodine atom to attach itself to a specific position other than
ortho- to a phenolic hydroxyl group is not an easy task. Accordingly, a different strategy is
generally employed to radioiodinate these small molecules of known chemical structure. That
strategy is iodine for iodine isotopic exchange.
Theoretical Labeling Efficiency
In principle, if equal molar quantities of an iodinated molecule [R-127I] and free
radioiodide ions [123I-] were incubated under appropriate conditions, at equilibrium one would
find the same chemical composition as before, but 50% of the radiotracer would now be
covalently attached to the molecule, and 50% of the tracer would still be in the free iodide form.
Maximum % labeling efficiency, assuming equilibrium is achieved, is 50% according to the
following equation:
13
% LE = [RI-127] / [I-123] x 100%
[[RI-127] / [I-123]] + 1 ,
where [ ] represents the initial molar concentration of each species.
Fortunately, we typically radiolabel with carrier-free, or at least no-carrier-added
radioiodide solutions. To illustrate this effect, if 1 mg of sodium hippurate (3.06 µmol) were
incubated with 1 mCi of carrier-free sodium [123I] iodide (4.25 pmol), the molar ratio of hippuran
to radioiodide would be 720,000 to 1. Therefore, the labeling efficiency possible at equilibrium
would be 720,000 / 720,001 = 99.9999%. Obviously, excellent radiolabeling efficiencies are
possible with radioiodide for iodine exchange reactions, even if some carrier iodide is present.
The specific activity, however, of the final product is limited in the previous example to 1
mCi/mg, or 327 mCi/mmol.
Strategies for Exchange Labeling
A discussion of methods for region-specific placement of stable iodine atoms on small
organic molecules is well beyond the scope of this chapter . That is the domain of creative
synthetic organic chemists. Instead, the intent here is to discuss a few of the methods that have
been developed to speed up the labeling process and enable equilibrium conditions to be reached
in a timely manner. It is easy to describe the general principles of isotope exchange reactions. It
is difficult, if not impossible, to describe the specific mechanism(s) by which that atom for atom
exchange actually occurs because radiolabeling is never performed under conditions in which
reagent concentrations are even remotely equimolar. Furthermore, the extremely low chemical
concentration of radioiodide that is used (nanomolar range) makes the desired reaction highly
susceptible to unpredictable interference by trace contaminants present in the reaction mixture.
14
Solution Phase Exchange Reaction. One of the easiest reaction parameters to control is
the temperature. Since rate of reaction is directly proportional to temperature, one can increase
this rate for thermally stable molecules and reach equilibrium conditions faster merely by
conducting the exchange reaction in a boiling water bath or in an autoclave. Reactions in non-
aqueous solvents can be done under reflux conditions or in sealed tubes. Likewise, high
temperature mixed halogen exchange reactions (*I for Br) have been reported at ~90% yields
with the reagents dissolved in molten (160 ºC) ammonium acetate (m.p.= 114 ºC).4 Microwave
ovens have also been used to increase the rate of heating and shorten the required reaction time
by affecting the polarity of the solvent and enhancing the reactivity of the reagents.13
Radioiodine exchange reactions are generally required to be performed in an acidic
environment. At low pH and in the presence of oxygen, iodide can oxidize to form I+I- and
become more reactive to exchange. Since commercial reductant-free (labeling grade) radioiodide
is most commonly shipped in dilute sodium hydroxide solution, one needs to be aware of the
amount of base in the reaction mixture (in addition to the quantity of radioactivity) so that
sufficient acidity is achieved for optimal exchange labeling. However, acidic pH increases the
volatility of radioiodine as hydriodic acid and thus can create contamination problems if the
solution is not neutralized prior to removal from the reaction vessel. As with any radioiodination
procedure, exchange labeling should always be performed in a charcoal filtered isotope hood.
In the earlier discussion of protein radioiodination, emphasis was placed on labeling
methods that were the least damaging to the protein molecule itself. Labeling efficiency was not
as important because it is relatively easy to remove the unreacted radioiodide from the final
product by gel filtration. In the case of small organic molecules, however, final product
15
purification to remove free iodide is not as simple. Molecular size differences are too small for
gel filtration. While anion exchange resins can be used to remove free iodide, a large portion of
the labeled molecule may also be retained depending on its ionic state or non-specific affinity for
the resin matrix. Therefore, the emphasis for isotope exchange reactions has been to create
reaction conditions that lead to maximum, clinically-acceptable radiochemical purities, typically
>95%.
Sodium ortho-iodohippurate (OIH) (Fig. 4A) is an example of a clinically used aryl
iodide that has been exchange labeled with both 131I and 123I. One of the earliest methods
developed was to add crystals of OIH to dried radioiodide, seal the vessel, and apply heat until
the crystals melted. Exchange labeling occurred in the molten state, but at efficiencies of only
25% to 50%. High boiling solvents such as acetamide have also been used to dissolve the
reactants at high temperature in an attempt to increase the labeling yield. Commercial production
used reflux in acidic ethanol to facilitate the reaction, achieving higher labeling efficiencies, but
final product purification was still required. More recently, various copper salts have been used
as apparent exchange catalysts in aqueous kit methods, which require less than 30 minutes in a
boiling water bath or an autoclave.6
N-isopropyl-p-iodoamphetamine (IAMP) (Fig. 4B) is another aryl iodide that has proven
difficult to label quantitatively with the isotope exchange technique.28 It was originally prepared
at 70% to 80% yield by sealing dry IAMP with neutralized radioiodide in a Pyrex tube and
heating to 150 °C for 30 minutes. An elaborate purification procedure was required, however, to
extract the labeled product prior to radiopharmaceutical formulation. Higher yields of 85% to
95% were reported when the reaction was performed under reflux conditions in glacial acetic
16
acid. Yet another method combined Cu(II) nitrate with refluxing acetic acid, but the reaction was
still not quantitative.
N,N,N'-trimethyl-[2-hydroxy-3-methyl-5-iodobenzyl]-1,3-propanediamine (HIPDM)
(Fig. 4C) represents a chemically modified aryl iodide that is easily radiolabeled in aqueous
solution.14 The presence of a phenolic hydroxyl group at position 2 of the aromatic ring serves
to activate the ring and allow for rapid radioiodine exchange at only 100 °C. Kinetic and
mechanistic studies have shown that the optimal pH for labeling is around 3, and that the reaction
is virtually complete within 5 minutes at an HIPDM concentration of 2 mg/mL. The presence of
chloride ions does not interfere with the iodine exchange. Labeling yields decrease quite
dramatically at pH >6 or in the presence of 20 µg/mL of sodium bisulfite. These observations
strongly suggest that the exchange mechanism for HIPDM involves the formation of an active I+
or iodine free radical.
Solid Phase Exchange Reaction. An alternative to isotopic exchange within molten
substrate is the incubation of dried radioiodide and substrate held at a temperature just below its
melting point. Radioiodinated [131I] meta-iodobenzylguanidine (mIBG) (Fig. 4D) has been
shown to be a valuable agent for both the visualization and potential treatment of
pheochromocytoma and neuroblastoma. When labeled with 123I, the superior imaging
characteristics of the label extend the compound's usefulness to the visualization of myocardial
sympathetic innervation. However, being another non-activated aryl iodide, mIBG has proven to
be somewhat difficult to radiolabel routinely. A number of factors have been identified that
influence the solid-state iodine exchange reaction, but the exact mechanism has yet to be
17
determined. Once it is fully understood, mIBG radiolabeling should become at least as reliable
as HIPDM.
Meanwhile, the following characterizations of the ammonium sulfate facilitated solid-
state exchange reaction for mIBG can be made.15 Ammonium sulfate is a necessary adjunct to
ensure that the mixture's pH slowly becomes acidic (pH ~3) prior to complete solvent
evaporation; the pH goes down as ammonium sulfate decomposes under heating to liberate
volatile ammonia. The final reaction temperature needs to be maintained between 140 °C and
155 °C (melting point = 162 °C) for at least 60 minutes. Oxygen needs to be present and
chemical reductants need to be absent for the exchange to occur; however, continuous air flow
throughout the entire evaporation and incubation periods leads to large variations in final product
yield. Chloride ions do markedly interfere with the iodine-for-iodine exchange and must be
avoided. Finally, the exchange does not occur if the reactants are not truly dry during incubation
at the elevated temperature. Radiochemical purity typically exceeds 95% upon reconstitution of
the dried reaction matrix with acetate buffer.18
Copper Ion Catalysts. A number of attempts have been made to use metal ions as
possible catalysts to increase the rate at which isotopic exchange reactions occur. The most
common metal used in radioiodine exchange schemes has been copper. However, it is not yet
clear whether Cu(I) or Cu(II) is the necessary catalyst for high yield radioiodine exchange.19
Copper(II) sulfate has been used to facilitate the rapid radiolabeling of OIH in a 50% ethanol
solution at 121 °C. But it was not resolved whether the copper ion truly catalyzed the desired
reaction, or merely interfered with the exchange labeling of ortho-iodobenzoic acid, a chemical
impurity known to be present. Similarly, Cu(II) nitrate was used to assist in the radioiodination
18
of IAMP in glacial acetic acid. However, a 70% yield after 2 hours at 120°C does not indicate a
dramatic catalytic effect for this system.
On the other hand, a promising Cu(I) method has been proposed,17 which purports >99%
yields for a number of aryl iodides, including IAMP, OIH, HIPDM, and mIBG. The general
premise is that Cu(I) ions are generated in situ from Cu(II) sulfate by the presence of an excess of
reducing agents in oxygen-free acidic solution. These required reducing agents include Sn(II)
sulfate, 2,5-dihydroxybenzoic acid, and citric or ascorbic acid. For most of the compounds
tested, radiolabeling was completed within 40 minutes at only 100 °C. Following reaction, a
buffer solution must be added to adjust the pH and render the final formulation isotonic for
injection. A variant procedure using Cu(I) chloride in a similar reducing environment has also
been reported to yield high specific activity 5-iodo-dL-nicotine from the 5-bromo-nicotine
precursor, but only at a 55% radiochemical yield.12 A subsequent mechanistic study of Cu(I) vs.
Cu(II) catalysts appears to confirm the preference for the cuprous ion (Cu+) in aqueous
hydrochloric acid (pH= 3) with sodium metabisulfite as the concomitant reducing agent to
prevent the side-product reactions attributable to Cu(II).5 It is somewhat surprising, however,
that at 100 ºC, it required >100 minutes to achieve less than 90% exchange with either 3-
iodotyrosine or 4-iodophenylalanine. It is apparent that Cu(I) is a non-typical catalyst, in that its
optimal concentration and chemical purity are critical to reproducible, high yield radioiodination.
19
IODINATION OF NEW MOLECULES
The focus of this chapter has been on the historic methods used for radiolabeling proteins
and on the simpler “shake and bake” techniques for radioiodine isotopic exchange. As long as
monoclonal antibodies offer hope for targeted delivery of tracer to cancer cells, there will be a
continued need to refine the radiolabeling procedures used to “tag” those molecules with
diagnostic and/or therapeutic radionuclides. But as the practice of Nuclear Medicine moves
further into the realm of receptor ligands, we will require new radiolabeling procedures which
afford significantly higher specific activities than are obtainable with the “*I for an I” exchange
method. And while 124I was once regarded as an undesirable contaminant in commercially
available 123I preparations, it has re-appeared as a valuable tracer for clinical PET evaluations, in
part due to its commercial distribution potential. Accordingly, many of the techniques
highlighted here for position-selective radioiodination of protein conjugation agents are equally
applicable for developing new designer radiopharmaceuticals.
Molecular Imaging of Gene Expression.
One of the more exciting areas of clinical research is the development of new radiotracers
to monitor the localization, magnitude and duration of gene expression following gene therapy.
The nucleoside analog FIAU (2’-fluoro-2’-deoxy-1β-D-arabinofuranosyl-5-iodo-uracil is a
specific marker substrate for thymidine kinase gene expression. It has been labeled with either
131I or 124I as the radionuclide for imaging. However, a standardized radiolabeling protocol has
not yet been adopted which could help validate the molecular imaging studies currently
underway. In one method, the non-iodinated FAU precursor was merely heated to 65 ºC in the
presence of sodium [131I]iodide and IODO-GEN to give 35-41% radiochemical yields.1 In
another, FAU was dissolved in 2 M nitric acid and held at 115 ºC for 45 minutes with n.c.a.
20
sodium [124I]iodide to give ~55% radiochemical yield.9 In yet another variation, the 5’-
stannylated FAU precursor was reacted at room temperature with sodium [124I]iodide in a
mixture of acetic acid and 30% H2O2 to give 51-54% radiochemical yield. Interestingly, when
the same radiochemists used sodium [131I]iodide from a commercial source instead to label the
same stannylated precursor under the same reaction conditions, 93% radiochemical yields were
obtained.23 Accordingly, comparison of experimental results among different research
laboratories using F*IAU as a molecular probe will be difficult to analyze since the
(radio)chemical impurities also vary among the several radiolabeling procedures used for this
single radiopharmaceutical.
Metal Salt Oxidants.
One of the more promising new techniques for no-carrier-added (n.c.a.) aromatic
radioiodination is, in reality, a re-visitation of older methods that have been used for iodination at
the macro scale. When trifluoroacetic acid (TFA) solutions of benzene and metal salt were
added sequentially to sodium [131I]iodide and stirred in a closed reaction vessel, [131I]
iodobenzene was produced with radiochemical yields of 82% and 64% using Pb(IV) acetate and
Ce(IV) triflate, respectively, as the oxidant.16 Addition of as little as 0.5 μmol of iodide carrier
to the TFA-insoluble cerium salt increased the yield to ~90%. In a relatively simple one-pot
synthesis, reactions were completed within 15 minutes at room temperature. The ortho/para
isomer distribution for all monosubstituted arenes tested was found to be independent of the
particular metal salt used and typical of electrophilic substitution. The non-inhibitory effect of
carrier iodide on the radiochemical yields further supports the conclusion that the metal salts
caused in situ oxidation of n.c.a. radioiodide followed by an electrophilic aromatic substitution.
21
If both the precursor substrate and the desired radioiodinated end-product are stable in TFA, such
a method should greatly facilitate the development of many new radiotracer molecules and
extend the usefulness and versatility of radioiodine in nuclear medicine research.
22
REFERENCES
1. Balatoni J, Finn R, Tjuvajev JG, et al. Synthesis and quality assurance of radioiodinated
2’-fluoro-2’-deoxy-1-β-D-arabinofuranosyl-5-iodo-uracil. J Labelled Compd
Radiopharm 40: 103-103, 1997.
2. Bolton AE and Hunter WM. The labelling of proteins to high specific radioactivities by
conjugation to a 125I-containing acylating agent. Biochem J 133: 529-538, 1973.
3. Boonkitticharoen V and Laohathai K. Assessing performances of Iodogen-coated
surfaces used for radioiodination of proteins. Nucl Med Commun 11(4): 295-304, 1990.
4. El-Shaboury G, Farah K and El-Tawoosy M. Nucleophilic radioiodination of 6-
bromocholesterol via non-isotopic exchange reaction in molten state. J Radioanal Nucl
Chem 249(3): 535-540, 2001.
5. Farah K and Farouk N. Copper catalyzed radioiodination of 3-iodotyrosine and 4-
iodophenyl alanine. J Labelled Compd Radiopharm 34(11): 915-926, 1997.
6. Greenwood FC, Hunter WM and Glover JS. The preparation of 131I-labelled human
growth hormone of high specific radioactivity, Biochem J 89: 114-123, 1963.
7. Hawkins L, Elliott A, Shields R, et al. A rapid quantitative method for the preparation of
123I-iodohippuric acid. Eur J Nucl Med 7: 58-61, 1982.
8. Hussain AA, Jona JA, Yamada A and Dittert LW. Chloramine-T in radiolabeling
techniques. II. A nondestructive method for radiolabeling biomolecules by halogenation.
Anal Biochem 224: 221-226, 1995.
9. Jacobs A, Bräunlich I, Graf R, et al. Quantitative kinetics of [124I]FIAU in cat and man. J
Nuc Med 42: 467-475, 2001.
23
10. Kabalka GW and Varma RS. The synthesis of radiolabeled compounds via
organometallic intermediates, Tetrahedron 45(21): 6601-6621, 1989.
11. Kaminski JJ, Bodor N and Higuchi T. N-Halo derivatives. IV. Synthesis of low chlorine
potential soft N-chloroamine systems. J Pharm Sci 65: 1733-1737, 1976.
12. Kampfer I, Sorger D, Schliebs R, et al. Radioiodination of nicotine with specific activity
high enough for mapping nicotinic acetylcholine receptors. Eur J Nucl Med 23(2): 157-
162, 1996.
13. Kumar P, Wiebe LI, Asikoglu M, et al. Microwave-assisted (radio)halogenation of
nitroimidazole-based hypoxia markers. Appl Radiat Isot 57: 697-703, 2002.
14. Lui B, Chang J, Sun JS, et al. Radioactive iodine exchange reaction of HIPDM: kinetics
and mechanism. J Nucl Med 28: 360-365, 1987.
15. Mangner TJ, Wu J-l and Wieland DM. Solid-phase exchange radioiodination of aryl
iodides: facilitation by ammonium sulfate. J Org Chem 47: 1484-1488, 1982.
16. Mennicke E, Holschbach MH and Coenen HH. Direct n.c.a. radioiodination of weakly
activated arenes using metal salts. Radiochimica Acta 88(3-4): 221-227, 2000.
17. Mertens J and Gysemans M. Cu1+ assisted nucleophilic exchange radiohalogenation:
application and mechanistic approach. In Emran AM, editor: New trends in
radiopharmaceutical synthesis, quality assurance, and regulatory control, pp. 53-65, New
York, 1991, Plenum Press.
18. Mock BH and Weiner RE. Simplified solid-state labeling of [123I ] m-iodobenzyl-
guanidine. Appl Radiat Isot 39(9): 939-942, 1988.
24
19. Prabhakar G, Mehra KS, Ramamoorthy N, et al. Evaluation of radioiodination of meta-
iodobenzylguanidine (MIBG) catalysed by in situ generated Cu(I) and directly added
Cu(II). Appl Radiat Isot 50: 1011-1014, 1999.
20. Richardson AP, Mountford PJ, Baird AC, et al. An improved iodogen method of
labelling antibodies with 123I. Nucl Med Commun 7: 355-362, 1986.
21. Seevers RH and Counsell RE. Radioiodination techniques for small organic molecules.
Chem Rev 82: 575-590, 1982.
22. Tashtoush BM, Traboulsi AA, Dittert L and Hussain AA. Chloramine-T in radiolabeling
techniques. IV. Pento-O-acetyl-N-chloro-N-methylglucamine as an oxidizing agent in
radiolabeling techniques. Anal Biochem 288: 16-21, 2001.
23. Tjuvajev JG, Doubrovin M, Akhurst T et al. Comparison of radiolabeled nucleoside
probes (FIAU, FHBG, and FHPG) for PET imaging of HSV1-tk gene expression. J Nuc
Med 43(8): 1072-1083, 2002.
24. Unak T, Akgun Z, Yildirim Y, et al. Self-radioiodination of Iodogen. Appl Radiat Isot
54(5): 749-752, 2001.
25. Visser GW, Klok RP, Klein-Gebbinck JW, et al. Optimal quality 131I-monoclonal
antibodies on high-dose labeling in a large reaction volume and temporarily coating the
antibody with IODO-GEN. J Nuc Med 42(3): 509-519, 2001.
26. Wilbur DS. Radiohalogenation of proteins: An overview of radionuclides, labeling
methods, and reagents for conjugate labeling. Bioconjugate Chem 3(6): 433-470, 1992.
27. Wilbur DS, Hadley SW, Hylarides MD, et al: Development of a stable radioiodinating
reagent to label monoclonal antibodies for radiotherapy of cancer. J Nucl Med 30(2):
216-226, 1989.
25
28. Winchell HS, Baldwin RM and Lin TH. Development of I-123-labeled amines for brain
studies: localization of I-123 iodophenylalkyl amines in rat brain. J Nucl Med 21: 940-
946, 1980.
29. Wood FT, Wu MM and Gerhart JC. The radioactive labeling of proteins with an
iodinated amidination reagent. Anal Biochem 69: 339-349, 1975.
30. Youfeng H, Coenen HH, Petzold G and Stocklin G. A comparative study of
radioiodination of simple aromatic compounds via N-halosuccinimides and chloramine-T
in TFAA. J Labelled Compd Radiopharm 19: 807-819, 1982.
26
FIGURE LEGENDS
Fig. 1. Sample peptide illustrating common sites of protein radiolabeling.
Fig. 2. Molecular structure of common radioiodination reagents: Oxidizing agents for direct
protein labeling, Chloramine-T (A) and IODO-GEN™ (B). Pre-labeled agents for
indirect protein labeling, Bolton-Hunter reagent (C) and Wood reagent (D).
Fig. 3. Examples of N-chloro-secondary amines with lower chlorine potential than
Chloramine-T: NCMGE and NCS. An indirect protein labeling scheme is
illustrated where radioiodine is first attached to an organometallic precursor using
NCS as the oxidant. The pre-labeled conjugate is then reacted with a free amino
function on the protein.
Fig. 4. Molecular structure of aryl iodides commonly radiolabeled by isotopic exchange.
A, OIH. B, IAMP. C, HIPDM. D, mIBG.
27
Figure 1
I*
OHCH
2
HC
HN CO
HC
NH
C
CH
H
2
C
CH
2
CH
2
CH
2
CH
2
NH
2
HNC
HCO
NH
2
H
2
CHS
O
C
HO O
N N
Tyrosine
Histidine
Lysine
Cysteine
I*
28
Figure 2
H
3
C S N Na
Cl
O
O
NN
NN
O
O
Cl Cl
ClCl
CH
2
CH
2
CO
*I
HO
O
N
O
OC
OH
I*
NH
2+
Cl
-
OCH
3
A. Chloramine-T B. IODO-GEN
TM
C. Bolton-Hunter Reagent D. Wood Reagent
29
Figure 3
CH
3
N
CH2
Cl
H OOCCH3
HH3CCOO
OOCCH3
H
OOCCH3
H
CH2OOCCH3
Penta-O-acetyl-N-chloro-N-methylglucamine (NCMGE)
nBu3Sn C O
O
N
O
O
*I C O
O
N
O
O
N
O
O
Cl
N-Chlorosuccinimide (NCS)
Na*I
*I C
O
N Lysine
H
Protein
H2N Lysine Protein
30
Figure 4
I*
N
OH
O
O
H*I
N
H
N
I*
H3C
OH
NN
H
I*
NH2
NH
A. OIH B. IAMP
C. HIPDM D. mIBG
