Synthesis of 131I-labeled glucose-conjugated inhibitors of 06-methylguanine-DNA methyltransferase (MGMT) and comparison with nonconjugated inhibitors as potential tools for in vivo MGMT imaging

Article abstract of DOI:10.1021/jm050588q

O⁶-Substituted guanine derivatives are powerful agents used for tumor cell sensitization by inhibition of the DNA repair enzyme O ⁶-methylguanine-DNA methyltransferase (MGMT). To provide targeted accumulation of MGMT inhibitors in tu

Full text of DOI:10.1021/jm050588q

J. Med. Chem. 2006, 49, 263-272
263
Synthesis of 131I-Labeled Glucose-Conjugated Inhibitors of O6-Methylguanine-DNA
Methyltransferase (MGMT) and Comparison with Nonconjugated Inhibitors as Potential Tools
for in Vivo MGMT Imaging
Ute Mu¨hlhausen,† Ralf Schirrmacher,*,‡ Markus Piel,† Bernd Lecher,§ Manuela Briegert,§ Andrea Piee-Staffa,§
Bernd Kaina,§ and Frank Ro¨sch*,†
Institute of Nuclear Chemistry, UniVersity of Mainz, Fritz Strassmann-Weg 2, D-55128 Mainz, Department of Nuclear Medicine,
UniVersity of Mainz, Langenbeckstr. 1, D-55131 Mainz, and Department of Toxicology, UniVersity of Mainz, Obere Zahlbacher Strasse 67,
D-55131 Mainz, Germany
ReceiVed June 21, 2005
O6-Substituted guanine derivatives are powerful agents used for tumor cell sensitization by inhibition of the
DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT). To provide targeted accumulation
of MGMT inhibitors in tumor tissue as well as tools for in vivo imaging, we synthesized iodinated C8-
alkyl-linked glucose conjugates of 2-amino-6-(5-iodothenyl)-9H-purine (O6-(5-iodothenyl) guanine, ITG)
and 2-amino-6-(3-iodobenzyloxy)-9H-purine (O6-(5-iodobenzyl) guanine, IBG). These compounds have
MGMT inhibitor constants (IC50 values) of 0.8 and 0.45 µM for ITGG and IBGG, respectively, as determined
in HeLa S3 cells after 2-h incubation with inhibitor. To substantiate that the 131I-(hetero)arylmethylene
group at the O6-position of guanine is transferred to MGMT, both the glucose conjugated inhibitors ITGG
and IBGG and the corresponding nonglucose conjugated compounds ITG and IBG were labeled with iodine-
131. The radioiodinations of all compounds with [131I]I- were performed with radiochemical yields of >70%
for the destannylation of the corresponding tri-n-butylstannylated precursors. The binding ability of [131I]-
ITGG, [131]IBGG, [131I]ITG, and [131I]IBG to purified MGMT was tested. All radioactive compounds were
substrates for MGMT, as demonstrated using a competitive repair assay. The newly synthesized radioactive
inhibitors were utilized to study ex vivo biodistribution in mice, and the tumor-to-blood ratio of tissue
uptake of [131I]IBG and [131I]IBGG was determined to be 0.24 and 0.76 after 0.5 h, respectively.
Introduction
The structural requirements of O6-substituted guanines for
efficient depletion of human MGMT were demonstrated by
Moschel et al.9 They showed that O6-benzylguanine (O6-BG)
and O6-(p-chlorobenzyl)guanine are alternative substrates for
the enzyme, causing rapid depletion of MGMT in human tumor
cell extracts and intact cells.10 They synthesized and tested a
series of O6- and S6-substituted guanine derivatives for their
ability to deplete the human MGMT in cell-free extracts of
tumor cells as well as in intact cells. The order of potency for
inactivation of the enzyme was shown to be strongly dependent
on structure. In 1998, McElhinney et al. synthesized novel O6-
substituted guanine derivatives with heterocyclic moieties, and
among them was one of the most active inhibitors to date, O6-
(5-bromothenyl)guanine (O6-5-BTG), which was about 10 times
more effective than O6-BG.11
Many cytostatic drugs used in cancer chemotherapy provoke
cytotoxic effects through their covalent reaction with DNA.
Apart from platinium-containing drugs, most of these are alkyl-
ating agents that mostly target purines in DNA.1 A biologically
relevant site is the O6-position of guanine.2 However, the
therapeutic effectiveness of alkylating agents can be significantly
attenuated by the ability of tumor cells to repair DNA damage.
Enzymes responsible for removal of DNA lesions act to protect
cells from the cytotoxic effects of alkylating drugs. An important
player in alkylating drug resistance is the DNA repair protein
O6-methylguanine-DNA methyltransferase (MGMT), which
transfers alkyl groups from the O6-position of guanine to an
internal cysteine residue.3 The reaction is stoichiometric in the
sense that each MGMT molecule can accept only one alkyl
group. Following the reaction, the MGMT protein is inactivated
and degradated. Since MGMT is one of the most important
factors determining resistance to methylating and chloroethy-
lating agents, also causing acquired resistance of tumor cells
against these chemotherapeutics, strategies have been developed
to suppress MGMT activity in tumors by means of MGMT
pseudosubstrates. A great variety of different MGMT inhibitors
have been synthesized and evaluated as to their MGMT
inhibitory potency.4-7 The chemistry of these MGMT inhibitors
has been reviewed recently.8
Although most of these compounds showed encouraging in
vitro results and also no systemic side effects in patients, a phase
II trial combining O6-BG and carmustin reported no beneficial
outcome on patients with glioblastoma.12 The therapeutic failure
might be attributed to the fact that the dose level required for
effective inhibition of MGMT could not be achieved within the
tumor due to nonselective accumulation of the drug in nontumor
tissue. This lack of selectivity of MGMT inhibitors for neoplastic
tissue was presumably responsible for the systemic sensitization
of all tissues, while the increase in systemic sensitivity forced
the use of lower doses of carmustine during therapy.
* Corresponding authors. Tel: +49 6131 17 6736. Fax: +49 6131 17
2386. E-mail: schirrmacher@nuklear.klinik.uni-mainz.de (R.S.). Tel: +49
6131 39 25302. Fax: +49 6131 39 24510. E-mail: Frank.Roesch@uni-
mainz.de (F.R.).
A potential strategy for selective transport of MGMT inhibi-
tors is synthesizing glucose-conjugated MGMT antagonists
in order to target glucose transporters that are often overex-
pressed in tumor cells.13 O6-BG and O6-(4-bromothenyl) guanine
(O6-4-BTG) were linked to â-D-glucose at the N9-position of
† Institute of Nuclear Chemistry.
‡ Department of Nuclear Medicine.
§ Department of Toxicology.
10.1021/jm050588q CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/15/2005

264 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Mu¨hlhausen et al.
the guanine moiety by means of alkyl spacers of different length
and were evaluated as to their MGMT inhibition. The targeting
concept, i.e., “Will the glucose linked MGMT inhibitor be
targeted to the tumor?” still needs to be addressed.
syntheses of O6-substituted guanines in general.11 Recently, a
more convenient synthesis of 1 has been described that avoids
highly volatile trimethylamine as a reactant.25 Despite tri-
methylammonia being an excellent leaving group, the coupling
of alcohols with 1 (Scheme 1) can be difficult, depending on
the nature of the alcohol to be coupled. Generally, an excess of
the alcohol is deprotonated with sodium hydride, 1 is added,
and the mixture is stirred at room temperature.11 For the reaction
of 5-iodothenyl alcohol (2) and 1, basically the procedure of
McElhinney et al.11 was applied with additional precautions such
as shielding from light and the use of significantly longer
reaction times up to 74 h. 2-Amino-6-(5-iodothenyl)-9H-purine
(O6-(5-iodothenyl)guanine, ITG) (3) (Scheme 1, Table 5) was
obtained in 40% yield.
In the context of cancer therapy, a noninvasive method for
determining the MGMT status of a given tumor might be helpful
for individualized cancer therapy. Thus, if the tumor MGMT
level is high, cancer therapy using O6-alkylating chemothera-
peutics would not be advisable and these therapeutics should
be replaced by alternative therapeutic strategies. Efforts during
the past several years to synthesize radioactively labeled MGMT
inhibitors for the in vivo mapping of MGMT concentrations in
humans have been unable to quantify the MGMT level.14-19 It
has been demonstrated recently with 131I-iodinated O6-BG that
a tracer with significantly improved selectivity is required for
the in vivo mapping of MGMT.20 The major drawback was a
low tumor-to-blood ratio that hindered the noninvasive deter-
mination of tumor MGMT status due to high background
radioactivity. Therefore, a tracer with in vivo pharmacological
properties imparting greater blood clearance and improved tumor
targeting is required. Glucosidation is a general strategy for
improving both drug solubility and targeting via glucose
transporters. Thus, conjugation of MGMT inhibitors with
glucose might facilitate targeted depletion of MGMT activity
in neoplastic tissue, while the uptake in normal tissue might be
minimized. A 18F-labeled MGMT-glucose conjugate based on
2-amino-6-(2-[18F]fluoropyridine-4-ylmethoxy)-9H-purine has
already been synthesized, but low radiochemical yields and a
suboptimal IC50 value of 1.8 µM make this compound ineffec-
tive as a tracer for quantifying MGMT in vivo.21 It has also
been demonstrated recently that iodinated glucose conjugates,
namely, 2-amino-6-(3-iodobenzyloxy)-9-(octyl-â-D-glucosyl)-
purine (IBGG) and 2-amino-6-(5-iodothiophen-2-ylmethoxy)-
9-(octyl-â-D-glucosyl)purine (ITGG), displayed better in vitro
characteristics than the corresponding fluorinated compounds.22
To evaluate radioiodinated MGMT inhibitor-glucose conju-
gates, we synthesized iodine-131-labeled derivatives of O6-BG
and O6-(5-iodothenyl)guanine (ITG) and their corresponding
glucose conjugates. Here we report on the synthesis strategy
and properties of the compounds, and we verify that these
inhibitors are active by showing transfer of the radioactively
labeled 131I-(hetero)arylmethylene group to MGMT. In biodis-
tribution studies with the iodobenzyl derivatives in nude mice
containing a tumor xenograft, we also determined the tumor-
to-blood ratio, giving evidence that the radioactively labeled
MGMT ligands are suitable for in vivo imaging of MGMT
status.
For the synthesis of the glucose-conjugated derivative ITGG
(6) for in vitro and in vivo evaluation, the reactive benzoyl-
protected glucose-linker 8-bromooctyl 2,3,4,6-tetra-O-benzoyl-
â-D-glucopyranoside 4 had to be prepared (Scheme 2).13
Preparation of 4 was optimized with boron trifluoride etherate
as a catalyst,30 resulting in a 57% yield. As starting material
for this reaction, 2,2,2-trichloroacetimidate 2,3,4,6-tetra-O-
benzoyl-â-D-glucopyranoside (18) was needed, which was
previously synthesized from 2,3,4,6-tetra-O-benzoyl-â-D-glu-
copyranoside (17) (Scheme 2). The synthesis of the benzoyl-
protected ITGG derivative 5 (ITGGG) was carried out at room
temperature, with light exclusion and by increasing the reaction
time up to 23 h, in 53% yield. Cleavage of the benzoyl-
protecting groups was achieved by transesterfication under
Ze´mplen conditions (0.1 M sodium methanolate) to give the
nonradioactive standard compound ITGG (6) for biological
evaluation in 56% yield (Scheme 1, Table 5).
A common method for radioiodination is the radioiodode-
stannylation of the corresponding tri-n-butylstannylated precur-
sor. A first attempt to obtain 2-amino-9-(octyl-â-D-glucopyra-
noside)-6-(5-tri-n-butylstannylthenyl)purine (SnTGG) (8) started
from the iodinated compound 6, which was dissolved in
triethylamine and treated with hexabutylditin and tetrakis-
(triphenylphosphine)palladium(0) according to a similar litera-
ture procedure.26 HPLC workup did not yield compound 8,
probably due to the insolubility of 6 in triethylamine. Another
procedure for stannylation uses dichlorobis(triphenylphosphine)-
palladium(II) as a catalyst and dioxane as a solvent.19 Although
ITGG (6) was readily soluble in dioxane, the relatively harsh
reaction conditions of 104 °C led to an unwanted deiodination
of 6 without formation of the tri-n-butylstannylated precursor.
As an alternative, the more stable bromo derivative13 7 (Table
5) was reacted with hexabutylditin and dichlorobis(triph-
enylphosphine)palladium(II) as a catalyst (Scheme 1, step E).
The desired product SnTGG (8) (Table 5) was formed and could
be isolated by column chromatography in 74% yield.
Results and Discussion
Chemistry. The syntheses of the nonradioactive standard
compound 2-amino-6-(5-iodothenyl)-9-(octyl-â-D-glucosyl)pu-
rine (ITGG) (6) (Scheme 1, Table 5) for determining in vitro
and in vivo IC50 values and the tri-n-butylstannylated precursor
2-amino-9-(octyl-â-D-glucosyl)-6-(5-tri-n-butylstannylthenyl)-
purine (SnTGG) (8) (Scheme 1, Table 5) for the radioiodination
were performed starting from the trimethylammonia precursor
2-aminopurin-6-yltrimethylamonium chloride24 (1) and 5-iodo-
thenyl alcohol (2) (Schemes 1 and 2). D’Auria et al. synthesized
5-iodo-2-thiophenecarbaldehyde by reacting thenyl alcohol with
yellow mercury oxide and iodine.23 As an intermediate product
they obtained 5-iodothenyl alcohol.
For comparing the in vitro and in vivo properties of glucose-
conjugated inhibitor ITGG (6) to its nonglycosylated form ITG
(3), radioiodination of ITG (3) from the corresponding precursor
SnTGSi (10) (Table 5) had to be performed. According to a
modified literature procedure19 the first step of preparing 10
was the protection of ITG (3) at the N9-position of guanine with
trimethylsilylethoxymethyl chloride (Scheme 1). For the prepa-
ration of this protected ITG analogue 9, LiH as a base was used
instead of potassium tert-butoxide for the deprotonation of the
N9-proton,19 yielding the N9-isomer exclusively in 36% yield.
With the silyl-protected derivative 9 the stannylation reaction
was performed as described.19 To obtain larger amounts of
SnTGSi (10), we reacted ITGSi (9) with hexabutylditin and
palladium(II) as a catalyst on a larger scale, and column
For coupling 2 to the O6-position of guanine, an adequate
leaving group at the O6-position of guanine is needed. The
trimethylammonium salt 1, which was first described by Kilburis
and Lister,24 has been proven to be a suitable precursor for the

131I-Labeled Glucose-Conjugated Inhibitors MGMT
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 265
Scheme 1. Syntheses of MGMT Inhibitors and Corresponding Glucose Derivatives
the radioactive labeling with 131I, which is unproblematic due
to the long half-life of 8 d and a reaction time for the
deprotection of only 25 min.
Scheme 2. Syntheses of 5-Iodothenyl Alcohol and the Glucose
Linker
The preparations of the 3-iodobenzyl derivatives IBG (11)
and its glucose-conjugated analogue IBGG (13) were performed
similarly as described for the ITG derivatives 3 and 6 (Scheme
1, Table 5). 3-Iodobenzyl alcohol is commercially available and
IBG (11) was synthesized as previously described13 by introduc-
ing optimizations such as a longer reaction time for the
alcoholysis, which resulted in better yields (97% vs 60%13).
The conjugation of IBG (11) with the glucose linker 4,
performed analogously to the synthesis of ITGGG (5), resulted
in low yields of IBGGG (12) of only 15%. The deprotection of
the glucose unit with sodium methanolate (0.1 M) yielded IBGG
(13) in 64%. In contrast to ITGG (6), the stannylation of IBGG
(13) did not cause any problems and SnBGG (14) was obtained
in 97% yield. The preparation of the corresponding tri-n-
butylstannylated precursor SnBGSi (16) was performed in the
same way as the synthesis of SnTGSi (10). After protection
with trimethylsilylethoxymethyl chloride at the N9-position, the
final stannylation of IBGSi (15) yielded SnBGSi (16) in 45%.
chromatography on silica gel was used for isolation instead of
preparative TLC.19 Column chromatography started with 100%
of n-hexane to remove nonpolar impurities and continued with
n-hexane:ethyl acetate (3:2) to elute SnTGSi (10) in 18% yield.
The deprotection to ITG at the N9-position was performed after
Radioactive Labeling. Both electrophilic and nucleophilic
aromatic substitutions are common methods for obtaining
radioiodinated compounds. The nucleophilic approach requires

266 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Mu¨hlhausen et al.
high temperatures of about 140 °C for the exchange of bromine
with radioiodine, even when a copper(I)-salt is used as a
catalyst.27 For the synthesis of [131I]ITGG ([131I]6) the electro-
philic labeling method was chosen to avoid possible decomposi-
tion and deiodination at higher temperatures. The electrophilic
131I-iodination of the corresponding precursor SnTGG (8) was
performed by regioselective destannylation under mild reaction
conditions using [131I]iodide and chloramine-T (CAT) or Io-
dogen as oxidizing agents. These oxidizing agents form elec-
trophilic iodine species for subsequent reaction with the aromatic
or heteroaromatic system. A first attempt to label SnTGG (8)
with CAT under acidic conditions (2 N HCl) resulted in an
almost quantitative cleavage of the 131I-iodinated iodothenyl
alcohol at the O6-position of guanine. Interestingly, contrasting
reports about the acid sensitivity of O6-substituted guanines were
found in the literature. Safadi et al. reported on the instability
of O6-benzylguanine to acidic conditions,28 whereas Vaidy-
anathan et al.19 did not detect any cleavage of the 3-iodobenzyl
group from the O6-position of guanine using TFA for the
deprotection of the N9-position, which is explained to be due
to a lack of water. Furthermore, they did not detect any
decomposition of [131I]IBG ([131I]11) during HPLC purification
using 0.1% TFA in water as a cosolvent in contrast to our
results. Our next approach using CAT, [131I]I-, and a phosphate
buffer at pH 7.0 yielded the desired product [131I]6 (Scheme 1)
in radiochemical yields (RCY) of up to 93% and a radiochemical
purity of >99.9% after HPLC separation. The solvent of the
HPLC fractions containing [131I]6 could be easily removed using
a micro rotary evaporator under reduced pressure.
The radioiodination of SnTGSi (10) to [131I]ITGSi ([131I]9)
(Scheme 1) was performed similarly using the CAT/phosphate
buffer system with radiochemical yields of 88%. Obviously,
the amount of CAT used in the synthesis plays a critical role in
obtaining high radiochemical yields. A relatively low concentra-
tion of CAT (2.5 µL, c ) 1.5 mg CAT/mL) prevented cleavage
of the O6-ethereal linkage that occurred to some extent at higher
CAT concentrations. The deprotection19 of [131I]9 to [131I]ITG
([131I]3) (Scheme 1) was performed in a single step without
HPLC separation of the intermediate [131I]9. Removal of the
trimethylsilylethoxymethyl protecting group was carried out with
tetrabutylammonium fluoride solution in THF instead of TFA,19
due to the instability of [131I]3 to acidic conditions as observed
in our laboratory. After the deprotection, isolation was performed
by reverse-phase HPLC and [131I]ITG ([131I]3) was obtained in
an overall RCY of 71% and a radiochemical purity of >99%.
In our experience, it is therefore not necessary to isolate [131I]9
previous to the deprotection step, making the synthesis more
convenient and minimizing the handling of radioactivity. Using
normal-phase HPLC and the single-step approach, Vaidyanathan
et al.19 detected an unlabeled side product coeluting with [131I]-
11 which was attributed to the unreacted and deprotected at
the N9-position labeling precursor. This observation could not
be confirmed in our laboratory, probably because reverse-phase
HPLC was used for purification, preventing coelution.
Figure 1. Radiochemical purity of [131I]IBG, [131I]IBGG, [131I]ITG,
and [131I]ITGG in physiological saline containing 1% DMSO (A) and
of [131I]ITG and [131I]ITGG in phosphate buffer containing 1% DMSO
(B).
the O6-benzyl group was observed during the synthesis of the
silyl-protected [131I]IBGSi ([131I]15). The compound was ob-
tained in 95% RCY and was successfully deprotected to yield
final [131I]11 in 82% RCY and a radiochemical purity of
>99.4%.
For further studies, such as in vitro testing in cell cultures
and animal studies, the radioiodinated compounds had to be
dissolved in physiological saline or in a suitable buffer system.
All radioiodinated compounds are sparingly soluble in water.
Therefore, we added 1% DMSO to the physiological saline as
a cosolvent. Unexpectedly, [131I]ITG and [131I]ITGG ([131I]3,
[
131I]6) displayed decomposition in this solvent system, in
contrast to the 131I-iodobenzylated compounds [131I]11 and [131I]-
13 (Figure 1A). No decomposition of the iodothenyl compounds
[
131I]3 and [131I]6 could be observed after 1-h incubation using
phosphate buffer (pH 7.0) with 1% DMSO as a cosolvent
(Figure 1 B), which was sufficient to start the in vivo and in
vitro studies. Different solvents for the longtime storage of
radioiodinated compounds, such as physiological saline/1%
DMSO (Figure 1A), phosphate buffer/1% DMSO (Figure 1B),
and pure ethanol were investigated. Among them, pure ethanol
was proven to be the most suitable. All compounds were stable
at 5 °C, and even after 1 month a radiochemical purity of >99%
was observed. Removing the ethanol directly prior to the
application by evaporation and dissolving the radioiodinated
compound in phosphate buffer or physiological saline with 1%
DMSO or diluting the ethanolic solution directly to an appropri-
ate concentration of ethanol for the given purpose make these
compounds suitable for in vitro and in vivo applications.
MGMT Inhibition by ITG (3), ITGG (6), IBG (11), and
IBGG (13). We determined the compounds’ ability to inactivate
MGMT in HeLa S3 cell extracts.22 As shown in Table 1, the
For the radioiodination of SnBGG (14) to [131I]IBGG ([131I]-
13) (Scheme 1), the CAT/phosphate buffer system at pH 7 was
used. Optimization of the reaction conditions such as the amount
of CAT and reaction time gave [131I]13 in 89% RCY with a
radiochemical purity of >99.2%. The specific activity of [131I]-
13 was 38 GBq/µmol, as determined by HPLC using a UV
calibration curve.
The radioactive synthesis of [131I]IBG ([131I]11) was per-
formed analogously to that of [131I]ITG ([131I]3). Using a small
amount of CAT (2.5 µL, c ) 1.5 mg/mL) only, no cleavage of

131I-Labeled Glucose-Conjugated Inhibitors MGMT
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 267
Table 1. In Vitro IC50 Values of Selected MGMT Inhibitors22
conjugate
IC50/µM
conjugate
IC50/µM
O6-BG
IBG
0.3
0.1
ITG
ITGG
0.75
0.8
IBGG
0.45
effective dose to produce 50% inactivation of MGMT (IC50)
for the nonglucose conjugated ITG (3) and IBG (11) was
comparable or even better than the IC50 value of O6-BG, an
inhibitor used in phase II clinical trials.12 To obtain tumor-
targeting, the glucose-conjugated derivatives were synthesized,
and their ability to inhibit MGMT was evaluated. Glucose
conjugation reduced the inhibition of MGMT in the in vitro
experiments by a factor of 4.5 for IBGG (13) when compared
to IBG (11). For ITG (3) and the corresponding ITGG (6), the
glucose conjugation did not lead to significantly different IC50
values for both compounds, which seems to be exceptional
because the aforementioned factor of 4.5 for the difference in
IC50 values was confirmed previously by Reinhard et al.13 Most
importantly, the IC50 values of all compounds were in the lower
micromolar range and therefore suitable for the validation of
the proposed targeting concept via glucose transporters. Fur-
thermore, these novel radioiodinated glucose conjugates might
thus serve as potential tracers for the quantification of MGMT
in neoplastic tissue, if the appropriate iodine isotope (e.g. [123I]-
iodine or [124I]iodine) is used for labeling.
Binding of [131I]ITG ([131I]3), [131I]ITGG ([131I]6), [131I]-
IBG ([131I]11), and [131I]IBGG ([131I]13) to Purified MGMT.
To determine the effect of accumulation of the glucose
conjugates via overexpressed glucose transporters present on
the cell surface of various tumors as well as to determine
noninvasively the MGMT level in neoplastic tissue, it is essential
that the radiolabel is transferred to MGMT. This is described
for [131I]IBG ([131I]11)19 but is not proven for the new
compounds [131I]IBGG ([131I]13), [131I]ITG ([131I]3), and [131I]-
ITGG ([131I]6). To determine the transfer rate of radioactivity
to the purified MGMT, the radiolabeled inhibitor was incubated
with purified MGMT (hMGMT), by adding increasing concen-
trations of the unlabeled compound. At a concentration of 1
µM of unlabeled IBG, 97% of [131I]IBG was bound to MGMT.
At concentrations higher than 1 µM, a decrease of input activity
which bound to MGMT was observed (Figure 2A). For [131I]-
IBGG, only 20% bound to MGMT at a concentration of 1 µM
of IBGG. The bound activity also decreased with increasing
concentrations of unlabeled IBGG (Figure 2A). This difference
of the transfer of the radiolabeled groups of the two compounds
under the same in vitro conditions reflects their different abilities
to inhibit MGMT. IBGG is not as potent as IBG and, therefore,
only 20% of [131I]IBGG binds to MGMT versus 97% of [131I]-
IBG at a concentration of 1 µM unlabeled competitor. The
observed binding of both compounds to purified MGMT was
specific. This was proven by the fact that only a small amount
of radioactivity was bound to bovine serum albumin (BSA),
which was included as a negative control, and by the competition
with the unlabeled compound.
Figure 2. Binding of [131I]IBG, [131I]IBGG (A) and [131I]ITG, [131I]-
ITGG (B) to purified human MGMT as a function of unlabeled inhibitor
concentration, respectively. As a control for unspecific binding, the
assays were performed with BSA as well. The binding activity (% of
input radioactivity that was found to be bound to MGMT) was 90%
for [131I]IBG, 18% for [131I]IBGG, 80% for [131I]ITG and 63% for
[
131I]ITGG.
MGMT. With these assays it could be demonstrated that the
radiolabeled group of each inhibitor is transferred specifically
to MGMT, which was a basic requirement for our further
investigations.
In Vivo Studies. For the non-glucose-conjugated compound
[
131I]IBG, the tissue distribution of 131I has been described
recently.20 Although the uptake of 131I into the tumor (TE-671
rhabdomyosarcoma xenograft) was lower than that into several
normal tissues, this does not necessarily pose a problem for the
intended use of [131I]IBG as a tracer to quantify MGMT in
tumors. But nevertheless it would be highly desirable to provide
a tracer that is more selective for the tumor, more metabolically
stable, and is cleared more rapidly from the blood. To possibly
achieve this, we conjugated IBG with a glucose linker.
In these studies, we compared the biodistribution of the 131I-
labeled compounds [131I]IBG and [131I]IBGG in blood, tumor,
and thyroid in nude mice bearing subcutaneous HeLa S3
xenografts. The distribution pattern in the selected tissues of
For [131I]ITG the maximum binding to MGMT was 80% at
a concentration of 0.1 µM of ITG, and the activity bound to
hMGMT was also decreasing with increasing concentration of
the unlabeled compound (Figure 2B). At the same concentration
(0.1 µM), 58% of [131I]ITGG bound to MGMT, decreasing also
with increasing concentration of ITGG (Figure 2B). The
difference in the percentage of bound radioactivity for these
two compounds is not so pronounced, reflecting that their ability
to inhibit MGMT is not as different as for IBG and IBGG,
respectively. In this case, too, the binding was specific to
[
131I]IBG and [131I]IBGG measured 0.5, 1, and 4 h postinjection
is shown in Figure 3 and Tables 2 and 3.
An important route of metabolism for all iodinated com-
pounds is deiodination, which leads to free iodine, which can
be accumulating in the thyroid. Comparing the uptake of [131I]-
iodide into the thyroid, it is interesting to note that this uptake

268 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Mu¨hlhausen et al.
buffer, which guaranteed high stability at 5 °C over a long
period. In vitro testing with the unlabeled compounds IBG (11),
IBGG (13), ITG (3), and ITGG (6) revealed IC50 values for the
glucose conjugates slightly lower than the corresponding non-
conjugated derivatives. Nevertheless, they are still eligible for
the purpose of investigation of tissue and tumor targeting and
determining MGMT status noninvasively. For all 131I-iodinated
compounds, we were able to show that the radioactively labeled
(hetero)arylmethylene group is transferred to MGMT, which is
a basic requirement for their use as potential tracers for the
investigation of mechanisms of targeting, action, and inactivation
of MGMT in situ. Further experiments with mice bearing tumors
are being planned for glucose conjugates [131I]IBGG, [131I]ITGG
and their nonconjugated counterparts [131I]IBG and [131I]ITG
in order to elucidate tumor uptake, protein binding, and organ
uptake as well as blood clearance. Initial studies with [131I]IBG
and [131I]IBGG administered to nude mice bearing HeLa S3
xenografts showed less deiodination and a more rapid clearance
from the blood of the glucose-conjugated tracer. This resulted
in a significantly better tumor-to-blood ratio of radioactivity (as
determined 30 min postinjection). Therefore, the glucose-con-
jugate IBGG, when labeled with an iodine isotope suitable for
a 3D-imaging technique such as single-photon emission com-
puted tomography (SPECT) or positron emission tomography
(PET), might become a more suitable tracer for determination
of MGMT status of tumors than the corresponding IBG.
Figure 3. Comparison of 131I uptake in selected tissues in nude mice
bearing HeLa S3 xenografts after treatment with nca [131I]IBG or nca
[
131I]IBGG, as measured 0.5 h after iv administration of the compounds.
Table 2. Tissue Distribution of No-Carrier-Added [131I]IBG in Nude
Mice Bearing HeLa S3 Xenografts
% ID/g (mean ( SD; n ) 3-7)
tissue
0.5 h
1 h
4 h
tumor
blood
thyroid
0.51 ( 0.06
2.16 ( 0.24
4.01 ( 0.34
0.83 ( 0.04
1.81 ( 0.12
6.55 ( 0.42
0.44 ( 0.03
0.98 ( 0.06
5.95 ( 1.36
Table 3. Tissue Distribution of No-Carrier-Added [131I]IBGG in Nude
Mice Bearing HeLa S3 Xenografts
% ID/g (mean ( SD; n ) 4-10)
tissue
0.5 h
1 h
4 h
tumor
blood
thyroid
0.54 ( 0.10
0.71 ( 0.08
1.16 ( 0.14
0.34 ( 0.02
0.90 ( 0.13
1.57 ( 0.22
0.17 ( 0.01
0.83 ( 0.26
2.83 ( 0.57
Experimental Section
Chemistry. Materials and Methods. The chemicals were
analytical grade or better and used without further purification.
Sodium [131I]iodide (IBS30) was purchased from Amersham
Bioscience (Braunschweig, Germany). Analytical TLC was per-
formed on aluminum-backed sheets (Silica gel 60 F254) and normal-
phase column chromatography was performed using silica gel 60,
both from Merck (Darmstadt, Germany). High performance liquid
chromatography (HPLC) was performed using two different
systems: (1) a Waters 1525 binary HPLC pump, a Waters 2487
dual wavelength absorbance detector (254 nm), a NaI(Tl) radio-
activity-detector M2102 (Messelektronik Dresden), injection valve
from Rheodyne (type 8125), and analysis of the HPLC data with
Breeze software (Waters), and (2) a Dionex P680A HPLC pump,
a Dionex UVD170U UV/vis detector (254 nm), a GABI NaI(Tl)
radioactivity detector with GINA STAR software (Raytest), a
Dionex injection valve P680, and analysis of HPLC data with CM-
PCS-1 software (Dionex). Reverse-phase HPLC was carried out
using a LiChrospher 100 RP 18 5 µm EC (250 mm × 4 mm)
column from CS-Chromatographie Service GmbH (Langerwehe,
Germany). For the separation of 131I-iodinated compounds [131I]3,
Table 4. Tumor to Blood Ratio of No-Carrier-Added [131I]IBG and
[
131I]IBGG in Nude Mice Bearing HeLa S3 Xenografts
tumor/blood ratio
0.5 h
1 h
4 h
[
[
131I]IBGG
131I]IBG
0.76
0.24
0.38
0.46
0.20
0.45
is at all time points much less for the glucose conjugate [131I]-
IBGG compared to [131I]IBG (Figure 3), e.g. 1.57% ( 0.22%
ID/g compared to 6.55% ( 0.42% ID/g at 1 h postinjection.
Because deiodination of the molecule in the body is not
desirable, the glucose conjugate appears to be more suitable as
a tracer. Another desirable property for a potential tracer is rapid
blood clearance. As anticipated, blood clearance was found to
occur faster for [131I]IBGG than for [131I]IBG. The uptake of
[
131I]IBGG into the tumor reached a maximum of 0.54% (
[
131I]6, [131I]11, and [131I]13, two different conditions were used:
0.10% ID/g at 0.5 h postinjection, decreasing at 1 and 4 h.
This is even less than the uptake of [131I]IBG into the tumor
(maximum 0.83% ( 0.04% ID/g 1 h postinjection). Since
clearance from blood was more rapid for the glucose conjugate,
the tumor to blood radioactivity ratio was clearly better for [131I]-
IBGG (0.76 and 0.24 for [131I]IBGG and [131I]IBG at 0.5 h post-
injection, respectively; Table 4). The high tumor-to-blood ratio
is a very important factor for the potential use of an inhibitor
as a tracer for the in vivo imaging of MGMT in tumor tissue.
(I) isocratic elution with acetonitrile and water 33:67 at a flow of
1 mL/min and (II) gradient elution using acetonitrile and water
(starting with 10% acetonitrile the gradient was increased to 100%
acetonitrile over 20 min and held there for 5 min at a flow of 1
mL/min). Radio-TLCs were analyzed using an Instant Imager
(Packard Canberra). Activity of tissue samples was counted in a
high-purity germanium detector system (Ortec) using GammaVision
5.0 software (Ortec) for analysis. NMR spectra (1H, 400 MHz; 13C,
100 MHz; 119Sn, 149 MHz) were obtained on a DRX400 spec-
trometer (Bruker Analytik GmbH). Chemical shifts are reported in
parts per million, and solvent peaks were referenced appropriately.
For identifying the abbreviations used in characterization of the
compounds cf. Figure 4. FD mass spectra were obtained on a
MAT90 spectrometer (Finnigan) and ESI mass spectra on a
Navigator instrument (ThermoQuest).
Conclusions
We report on the synthesis of four inhibitors for MGMT,
two of them conjugated with glucose via an alkyl spacer. Their
no-carrier-added (nca) radioiodination with 131I using the tri-
n-butylstannylated precursors was performed with good to
excellent yields. The stability in aqueous solution, which is a
prerequisite for further investigations, was examined for all
labeled 131I-iodinated compounds extensively. Absolute ethanol
was found to be superior to physiological saline and phosphate
5-Iodothenyl Alcohol (2). 5-Iodothenyl alcohol was prepared
similarly to a previously described procedure.23 A solution of thenyl
alcohol (5.10 g, 44.5 mmol) in toluene (30 mL) was treated
alternately with portions of iodine (12.01 g, 47.3 mmol) and yellow
mercuric oxide (10.02 g, 46.2 mmol). The resulting mixture was

131I-Labeled Glucose-Conjugated Inhibitors MGMT
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 269
Table 5. Synthesized Compounds
stirred for 3 h, filtered, and adsorbed on silica gel. Silica gel
chromatography using 6:2 n-hexane:ethyl acetate afforded 2 (10.10
g, 42.1 mmol) in 95% yield as a yellow liquid.
2,2,2-Tricholoroacetimidate 2,3,4,6-Tetra-O-benzoyl-r-D-glu-
copyranoside (18). 2,3,4,6-Tetra-O-benzoyl-R-D-glucopyranoside
(17) (1.90 g, 3.2 mmol) was dissolved in dry dichloromethane (19
mL) under an atmosphere of argon. The solution was cooled to 0
°C, and trichloroacetonitrile (1.37 g, 9.5 mmol) and DBU (48 µL,
0.3 mmol) were added. The mixture was stirred for 2 h until no
2,3,4,6-tetra-O-benzoyl-R-D-glucopyranoside (17) could be detected
(TLC, 3:1, n-hexane:ethyl acetate). The solvent was evaporated and
the product dried under vacuo to give 2,2,2-trichloroacetimidate
2,3,4,6-tetra-O-benzoyl-R-D-glucopyranoside (18) (2.2 g, 3.0 mmol)
in 95% yield as a foamy yellow solid. The product is sensitive to
air and should be stored under argon at -20 °C.
2,3,4,6-Tetra-O-benzoyl-r-D-glucopyranoside (17). 2,3,4,6-
Tetra-O-benzoyl-R-D-glucopyranosyl bromide (10.01 g, 15.2 mmol)
was dissolved in acetone (35 mL) and water was added until the
solution remained cloudy. After addition of sodium iodide (0.45 g,
3.0 mmol) the mixture was stirred at room temperature for 80 h
before removing the solvents in vacuo. The residue was resuspended
in water (25 mL) and the resulting water phase extracted with ethyl
acetate (3 × 100 mL). The mixture was concentrated in vacuo and
purified on a silica gel column using 5:1 n-hexane:ethyl acetate to
give 2,3,4,6-tetra-O-benzoyl-R-D-glucopyranoside (17) (6.90 g, 11.6
mmol) in 76% yield as a foamy solid.
8-Bromooctyl 2,3,4,6-Tetra-O-benzoyl-â-D-glucopyranoside
(4). The synthesis of this compound was performed similarly to a
Figure 4. Abbreviations for NMR spectra.

270 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Mu¨hlhausen et al.
previously described procedure.30 To a solution of 2,2,2-trichloro-
acetimidate 2,3,4,6-tetra-O-benzoyl-R-D-glucopyranoside (18) (3.31
g, 4.5 mmol) and 1-bromooctanol (0.58 g, 2.8 mmol) in dry
dichloromethane (3.5 mL) was added boron trifluoride etherate (0.34
mL, 4.5 mmol) under argon at 0 °C. The mixture was allowed to
warm to room temperature, stirred for 4 h, and poured into an ice-
cold saturated aqueous sodium bicarbonate solution. The aqueous
layer was extracted with diethyl ether (3 × 50 mL) and washed
with water (3 × 60 mL). After adsorbing on silica gel, column
chromatography (2:1 n-hexane:ethyl acetate) gave 8-bromooctyl
2,3,4,6-tetra-O-benzoyl-â-D-glucopyranoside (4) (2.03 g, 2.6 mmol)
in 57% yield as an oil.
2-Amino-6-(5-iodothenyl)-9H-purine (3; ITG). To a solution
of 5-iodothenyl alcohol (2) (4.51 g, 18.7 mmol) in dry DMSO (2.1
mL) under argon in the dark was carefully added sodium hydride
(0.30 g, 60% in mineral oil, 7.5 mmol). After stirring for 1 h at
room temperature, (2-aminopurin-6-yl)trimethylammonium chloride
125 (0.77 g, 3.4 mmol) was added and the mixture was stirred for
another 73 h. The reaction was terminated by the addition of acetic
acid (0.58 mL) and diethyl ether (101 mL). After 2 d at 5 °C the
resulting precipitate was collected, washed with diethyl ether (20
mL) and water (15 mL), and recrystallized from methanol.
2-Amino-6-(5-iodothenyl)-9H-purine (3) was obtained as a brown-
ish solid (0.51 g, 1.3 mmol) in 40% yield.
2-Amino-6-(5-iodothenyl)-9-(octyltetra-O-benzoyl-â-D-glu-
copyranoside)purine (5). Compound 3 (0.41 g, 1.1 mmol) was
dissolved in dry DMF (10 mL) containing 4-Å molecular sieves
under argon in the dark, and lithium hydride (8 mg, 0.1 mmol)
was added. This mixture was stirred for 1 h at room temperature
before adding 4 (0.81 g, 1.03 mmol) in dry DMF (6.5 mL). After
stirring for 23 h the reaction mixture was adsorbed on silica gel
and purified by silica gel chromatography (1:1 n-hexane:acetone)
to give 5 (0.61 g, 0.56 mmol) in a yield of 53% as a yellow oil.
2-Amino-6-(5-iodothenyl)-9-(octyl-â-D-glucopyranoside)pu-
rine (6; ITGG). 2-Amino-6-(5-iodothenyl)-9-(octyl-â-D-glucopy-
ranoside)purine (6) was prepared similar to a previously described
procedure.13 2-Amino-6-(5-iodothenyl)-9-(octyltetra-O-benzoyl-â-
D-glucopyranoside)purine (5) (0.59 g, 0.55 mmol) was dissolved
in dry methanol (13 mL) and treated with a freshly prepared sodium
methanolate solution (0.45 mL, 0.1 M). After stirring for 4 h at
room temperature the solution was neutralized with Dowex (H+
50WX2, 100-200 mesh) ion exchanger. After filtration, the solvent
was evaporated and the product purified by column chromatography
(5:1 chloroform:methanol) to give ITGG (6) (0.21 g, 0.30 mmol)
in 56% yield as a foamy solid.
2-Amino-9-(octyl-â-D-glucopyranoside)-6-(5-tri-n-butylstan-
nylthenyl)purine (8; SnTGG). The preparation of the tin precursor
8 was carried out with some modifications of a procedure described
by Vaidyanathan et al.19 Compound 713 (108 mg, 0.17 mmol) was
dissolved in dioxane (5.3 mL) under argon. Hexabutylditin (649
mg, 1.1 mmol) and dichlorobis(triphenylphosphine)palladium(II)
(61 mg, 0.09 mmol) were added, and the mixture was heated under
reflux for 4 h. After cooling the reaction mixture was filtered
(porewidth 1.0 µm) and concentrated in vacuo. Purification by silica
gel chromatography (4.5:1 chloroform:methanol) gave 8 (107 mg,
0.13 mmol) in 74% as a white foamy solid.
2-Amino-6-(5-iodothenyl)-9-(2-(trimethylsilanyl)ethoxymeth-
yl)purine (9; ITGSi). 2-Amino-6-(5-iodothenyl)-9H-purine (3) (440
mg, 1.18 mmol) was dissolved in dry DMF (6.0 mL) under argon
in the dark. Lithium hydride (93 mg, 11.8 mmol) was added slowly
and the mixture was stirred for 1 h at room temperature. After
cooling to 0 °C, 2-trimethylsilylethoxymethyl chloride (198 mg,
1.19 mmol) was added dropwise and the suspension was stirred
for another 24 h while warming up to room temperature. After
addition of ice-cold water (60 mL), the crude product was extracted
with ethyl acetate (3 × 50 mL) and the organic phase was dried
over Na2SO4. Final purification was performed by column chro-
matography (3:2 n-hexane:ethyl acetate) after adsorbing the crude
product on silica gel to give 9 (210 mg, 0.42 mmol) in 36% yield
as a yellow oil.
2-Amino-9-(2-(trimethylsilanyl)ethoxymethyl)-6-(5-tri-n-bu-
tylstannylthenyl)purine (10; SnTGSi). 2-Amino-6-(5-iodothenyl)-
9-(2-trimethylsilanylethoxymethyl)purine (100 mg, 0.21 mmol) was
dissolved in dioxane (6.0 mL) under argon. Hexabutylditin (742
mg, 1.28 mmol) and dichlorobis(triphenylphosphine)palladium(II)
(70 mg, 0.10 mmol) were added, and the resulting suspension was
heated under reflux for 140 min. After cooling to room temperature,
the crude reaction mixture was adsorbed on silica gel. Final workup
was achieved by column chromatography. To remove lipophilic
impurities, the first elution occurred with n-hexane, and for isolating
the product, n-hexane:ethyl acetate (3:2) was used as an eluent to
give 10 (24 mg, 0.02 mmol) in 18% yield as a yellow oil.
2-Amino-6-(3-iodobenzyloxy)-9H-purine (11; IBG). The prepa-
ration of 2-amino-6-(3-iodobenzyloxy)-9H-purine followed the
synthesis of ITG (3), using 3-iodobenzyl alcohol (5.01 g, 21.4
mmol), dry DMSO (2.5 mL), sodium hydride (0.35 g, 60% on
mineral oil, 8.8 mmol), and 1 (0.90 g, 3.9 mmol). After washing
the precipitate with diethyl ether (25 mL) and water (20 mL), 11
(1.41 g, 3.8 mmol) was obtained as a white solid in 97% yield.
2-Amino-6-(3-iodobenzyloxy)-9-(octyltetra-O-benzoyl-â-D-glu-
copyranoside)purine (12). The preparation of 12 followed the
synthesis of 5, using 2-amino-6-(3-iodobenzyloxy)-9H-purine (0.52
g, 1.4 mmol) in dry DMF (13.3 mL), lithium hydride (11 mg, 1.3
mmol), and 8-bromooctyl 2,3,4,6-tetra-O-benzoyl-â-D-glucopyra-
noside (4) (1.10 g, 1.4 mmol) in dry DMF (8.6 mL) to give 12
(230 mg, 0.21 mmol) in 15% yield as a foamy yellow solid.
2-Amino-6-(3-iodobenzyloxy)-9-(octyl-â-D-glucopyranoside)-
purine (13; IBGG). The deprotection of 12 was performed as
described for 5. The protected sugar-conjugate 12 (180 mg, 0.17
mmol) was dissolved in dry methanol (5 mL) and treated with
freshly prepared sodium methanolate solution (0.15 mL, 0.1 M).
Purification by column chromatography (5:1 chloroform:methanol)
gave 13 (70 mg, 0.11 mmol) in 64% yield as a foamy white solid.
2-Amino-9-(octyl-â-D-glucopyranoside)-6-(3-tri-n-butylstan-
nylbenzyloxy)purine (14; SnBGG). 13 (87 mg, 0.13 mmol) was
reacted with hexabutylditin (487 mg, 0.83 mmol) and dichlorobis-
(triphenylphosphine)palladium(II) (46 mg, 0.07 mmol) in dioxane
(3.9 mL) under reflux. After purification by column chromatography
(4.5:1 chloroform:methanol), 14 (110 mg, 0.13 mmol) was obtained
in 97% yield as a yellow oil.
2-Amino-6-(3-iodobenzyloxy)-9-(2-(trimethylsilanyl)ethoxym-
ethyl)purine (15; IBGSi). The N9-protected compound 15 was
prepared as described for 9, using 11 (440 mg, 1.09 mmol) in dry
DMF (5.5 mL), lithium hydride (86 mg, 10.9 mmol), and trimeth-
ylsilylethoxymethyl chloride (183 mg, 1.09 mmol). After purifica-
tion by column chromatography (3:2 n-hexane:ethyl acetate), 15
(190 mg, 0.38 mmol) was obtained in 35% yield as a yellow oil.
2-Amino-9-(2-(trimethylsilanyl)ethoxymethyl)-6-(3-tri-n-bu-
tylstannylbenzyloxy)purine (16; SnBGSi). The tri-n-butylstan-
nylated precursor 16 was prepared as described for 10, using IBGSi
(100 mg, 0.20 mmol), hexabutylditin (742 mg, 1.28 mmol), and
dichlorobis(triphenylphosphine)palladium(II) (70 mg, 0.10 mmol)
in dioxane (6.0 mL). After reaction and purification by column
chromatography (starting with n-hexane, then 3:2 n-hexane:ethyl
acetate), 16 (59 mg, 0.09 mmol) was obtained in 45% yield as a
white crystalline solid.
2-Amino-6-(5-[131I]iodothenyl)-9H-purine ([131I]3, [131I]ITG).
To a solution of precursor 10 in absolute ethanol (25 µL, 1 µg/µL)
was added 10 µL of [131I]I- (20 MBq) diluted in water followed
by 10 µL of phosphate buffer (pH 7.0) and 2.5 µL of CAT (1.5
mg/mL in phosphate buffer, pH 7.0). The reaction was allowed to
proceed for 5 min at room temperature. Monitoring by radio-TLC
(4:1 toluene:methanol) showed a RCY of 88.0% ( 7.3% for [131I]-
ITGSi after 5 min. For removing the 2-(trimethylsilanyl)ethoxy-
methyl group, the solvent was removed in vacuo, and after addition
of 50 µL of tetrabutylammonium fluoride (1 M in THF), the mixture
was heated to 68 °C for 25 min.19 The solvent was evaporated,
and the residue was dissolved in absolute ethanol (30 µL) and
injected on to a reverse-phase HPLC (condition I). Elution of [131I]-
ITG occurred at tR ) 9.4 min. Isolation afforded the labeled
compound with an overall RCY of 71.5% ( 5.0% and a radio-

131I-Labeled Glucose-Conjugated Inhibitors MGMT
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 271
chemical purity of >99%. For in vitro and in vivo testing, the HPLC
fraction was evaporated and [131I]3 was dissolved in phosphate
buffer (pH 7.0; 3 mL) with 1% of DMSO.
measured, and the results are expressed as the percentage of input
activity. The assay was performed twice for each concentration and
twice for all 131I-iodinated compounds.
2-Amino-6-(4-[131I]iodothenyl)-9-(octyl-â-D-glucopyranoside)-
purine ([131I]6, [131I]ITGG). To a solution of SnTGG (8) in
absolute ethanol (25 µL, 1 µg/ µL) was added 10 µL of [131I]I-
(19 MBq) diluted in water. This mixture was treated with 10 µL
of phosphate buffer (pH 7.0) and 7.5 µL of CAT [1.5 mg/mL in
phosphate buffer (pH 7.0)], and the reaction was allowed to proceed
for 5 min at room temperature. The reaction mixture was injected
on to a reverse-phase HPLC (condition II) where [131I]ITGG eluted
at tR ) 10.9 min with a RCY of 93.6% ( 0.8% and a radiochemical
purity of >99.9%. For further studies the solvent of the HPLC
fraction was evaporated and the activity was dissolved in phosphate
buffer (pH 7.0; 3 mL) with 1% of DMSO.
In Vivo Studies. Biodistribution of [131I]IBG and [131I]IBGG
in Nude Mice Bearing Mex(+) Xenografts. All biodistribution
studies were performed on male athymic nude mice (Charles River;
4-6 weeks old), which had access to food and water ad libitum.
Ten to 13 days prior to the study they were injected subcutaneously
with Mex(+) HeLa S3 cells in the right side (106 cells in 50 µL of
PBS). For the injection of the 131I-iodinated compounds via the
tail vein the mice were anesthetized with Ketanest S. No-carrier-
added [131I]IBG or [131I]IBGG (400 kBq/mouse) was injected in a
volume of 250 µL of physiological saline containing 1% of DMSO.
The accumulation of activity in different tissues was evaluated 0.5,
1, and 4 h postinjection of the tracer. For each time-point four to
eight mice were used. The animals were euthanized by cervical
dislocation, and the tissues of interest were removed, dried of blood,
weighed, and homogenized in potassium hydroxide solution. For
each sample the activity accumulated in the respective organ was
counted in a high-purity germanium detector along with injection
standards. The tissue uptake of 131I-activity is expressed as the
percentage of injected dose per gram tissue (% ID/g).
2-Amino-6-(3-[131I]iodobenzyloxy)-9H-purine ([131I]11, [131I]-
IBG). The radiochemical synthesis of this compound was ac-
complished as described for [131I]ITG. The initial reaction afforded
[
131I]IBGSi ([131I]15) after 5 min with a RCY of 95.5% ( 1%.
After deprotection with tetrabutylammoniun fluoride (1 M in THF)
[
131I]IBG could be isolated by reverse-phase HPLC (condition I)
at tR ) 11.6 min in 82% ( 6.9% radiochemical yield with a
radiochemical purity of >99.4%. For further studies the HPLC
fraction was evaporated under vacuo and the activity reconstituted
in physiological saline (3 mL) containing 1% of DMSO.
Acknowledgment. The authors wish to thank Dirk Bier,
Marcus H. Holschbach, H. H. Coenen, Daniela Stark, Sabine
Ho¨hnemann, Georg Nagel, Esther Schirrmacher, and Gerhard
Fritz for their support. This project was supported by DFG
Ka724/13-1 and Ka724/14-1 and SFB432/B7.
2-Amino-6-(3-[131I]iodobenzyloxy)-9-(octyl-â-D-glucopyra-
noside)purine ([131I]13, [131I]IBGG). To a solution of SnBGG in
absolute ethanol (25 µL, 1 µg/µL) were added 10 µL of [131I]I-
(20 MBq) diluted in water, and after that 10 µL of phosphate buffer
(pH 7.0) and 10 µL of CAT [1.5 mg/mL in phosphate buffer (pH
7.0)]. The reaction was allowed to proceed for 5 min at room
temperature before injecting the mixture on to a reverse-phase
HPLC (condition II). [131I]IBGG could be isolated at tR ) 11.1
min in 89.1% ( 0.9% radiochemical yield with a radiochemical
purity of >99.2%. For further studies, the HPLC fraction was
evaporated under vacuo and [131I]13 was dissolved in physiological
saline (3 mL) containing 1% of DMSO.
Supporting Information Available: NMR, HPLC, and MS
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) Tew, K.; Colvin, M.; Chabner, B. A. Alkylating agents. In Cancer
chemotherapy: principles and practise, 2nd ed.; Chabner, B. A.,
Longo, D. L., Eds.; Lippincott, J. B.: Philadelphia, 1995.
(2) (a) Dolan, M. E.; Mitchell, R. B.; Mummert, C.; Moschel, R. C.;
Pegg, A. E. Effect of O6-benzylguanine analogues on sensitivity of
human tumor cells to the cytotoxic effects of alkylating agents.
Cancer Res. 1991, 51, 3367-3373. (b) Loveless, A. Possible
relevance of O6-alkylation of deoxyguanosine to the mutagenicity
and carcinogenicity of nitrosamines and nitrosamides. Nature 1969,
223, 206-207. (c) Singer, B. N-nitroso alkylating agents: Formation
and persistence of alkyl derivatives in mammalian nucleic acids as
contributing factors in carcinogenesis. J. Nat. Cancer Inst. 1979, 62,
1329-1339. (d) Saffhill, R.; Margison, G. P.; O’Connor, P. J.
Mechanisms of carcinogenesis induced by alkylating agents. Biochim.
Biophys. Acta 1985, 823, 111-145.
In Vitro Studies. Preparation of Cell Extracts. Exponentially
growing HeLa S3 cells expressing MGMT extracts were prepared
as described previously.22 The HeLa S3 cells used by us expressed
MGMT at a level of 588 ( 86 fmol/mg protein (mean of 13
determinations). As a negative background control an extract of
HeLa MR cells deficient of MGMT served in each assay.
Determination of MGMT Activity. The inactivation of MGMT
in cell extracts was determined as described previously.22 After 30
min incubation of cell extracts with different concentrations of the
inhibitors IBG, IBGG, ITG, and ITGG at 37 °C the MGMT activity
was measured by assaying the loss of [3H]-O6-methylguanine from
a [3H]methylated calf thymus DNA substrate as reported.22 Each
assay was performed at least three times.
(3) Pegg, A. E.; Byers, T. L.; Repair of DNA containing O6-alkylguanine.
FASEB J. 1992, 6, 2302-2310.
In Vitro Binding of the 131I-Iodinated Inhibitors by MGMT.
Purified human MGMT was prepared as described from Escherichia
coli transformed with MGMT expression vector according to
standard techniques. The activity of the purified MGMT was
assayed as described above. The following protocol was the same
for all 131I-iodinated compounds. Similar to a previously reported
method,19 nearly identical activities of the 131I-labeled compounds
were added in the presence or absence of increasing amounts of
the corresponding nonradioactive IBG, IBGG, ITG, or ITGG,
respectively, to 10 µg of purified MGMT. The inhibitor compounds
were dissolved and stored at -20 °C in ethanol. Immediately before
the experiment, they were diluted 1:100 with distilled water and
were added to the reaction mixture. The final ethanol concentration
in the assay did not exceed 0.1%. To control for unspecific binding,
the assay was performed with BSA instead of MGMT. The
incubation was done in 100 µL of 50 mM Tris-Cl (pH 7.4), 5 mM
DTT, and 0.1 mM EDTA. After incubation for 30 min at 37 °C,
precipitation of the protein was accomplished by adding 200 µg of
carrier BSA and 1 mL of cold TCA (12%). The precipitated proteins
were collected on GF/C filters (Whatman) and washed extensively
with cold TCA (5%). The activity remaining on the filter was
(4) McElhinney, R. S.; McMurry, T. B. H.; Margison, G. P. O6-
Alkylguanine-DNA alkyltransferase inactivation in cancer chemo-
therapy. Mini ReV. Med. Chem. 2003, 3, 471-485.
(5) Christmann, M.; Tomicic, M. T.; Roos, W.; Kaina, B. Mechanisms
of human DNA repair: An update. Toxicology 2003, 193, 3-34.
(6) Erickson, L. C. The role of O-6 methylguanine DNA methyltrans-
ferase (MGMT) in drug resistance and strategies for its inhibition.
Seminars Cancer Biol. 1991, 2, 257-265.
(7) D’Incalci, M.; Citti, L.; Taverna, P.; Cata Pano, C. V. Importance of
the DNA repair enzyme O6-alkylguanine alkyltransferase (AT) in
cancer chemotherapy. Cancer Treat. ReV. 1988, 15, 279-292.
(8) Schirrmacher, R.; Schirrmacher, E.; Mu¨hlhausen, U.; Kaina, B.;
Wa¨ngler, B. Synthetic strategies towards O6-substituted guanine
derivatives and their application in medicine. Curr. Org. Synth. 2005,
2, 547-562.
(9) Moschel, R. C.; McDougall, M. G.; Dolan, M. E.; Stine, L.; Pegg,
A. E. Structural features of substituted purine derivatives compatible
with depletion of human O6-alkylguanine DNA alkyltransferase. J.
Med. Chem. 1992, 35, 4486-4491.
(10) Dolan, M. E.; Moschel, R. C.; Pegg, A. E. Depletion of mammalian
O6-alkylguanine-DNA alkyltransferase activity by O6-benzylguanine
provides a means to evaluate the role of this protein in protection
against carcinogenic and therapeutic alkylating agents. Proc. Natl.
Acad. Sci. U.S.A. 1990, 87, 5368-5372.

272 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Mu¨hlhausen et al.
(11) McElhinney, R. S.; Bonnellys, D. J.; McCormick, J. E.; Kelly, J.;
Eatson, A. J.; Rafferty, J. A.; Elder, R. H.; Middleton, M. R.;
Willington, M. A.; McMurry, T. B. Inactivation of O6-alkylguanine-
DNA alkyltransferase. 1. Novel O6-(hetarylmethyl)guanines having
basic rings in the side chain. J. Med. Chem. 1998, 41, 5265-5271.
(12) Quinn, J. A.; Pluda, J.; Dolan, M. E.; Delaney, S.; Kaplan, R.; Rich,
J. N.; Friedman, A. H.; Reardon, D. A.; Sampson, J. H.; Colvin, O.
M.; Haglund, M. M.; Pegg, A. E.; Moschel, R. C.; McLendon, R.
E.; Provenzale, J. H.; Gururangan, S.; Tourt-Uhlig, S.; Herndon, J.
E.; Bigner, D. D.; Friedman, H. S. Phase II trial of carmustine plus
O6-benzylguanine for patients with nitrosourea-resistant recurrent
or progressive malignant glioma. J. Clin. Oncol. 2002, 20, 2277-
2283.
(13) Reinhard, J.; Hull, W. E.; von der Lieth, C.-W.; Eichhorn, U.; Kliem,
H.-C.; Kaina, B.; Wiessler, M. Monosaccharide-linked inhibitors of
O6-methylguanine-DNA methyltransferase (MGMT): Synthesis, mo-
lecular modeling, and structure-activity relationship. J. Med. Chem.
2001, 44, 4050-4061.
(14) Schirrmacher, R.; Nesseler, E.; Hamkens, W.; Eichhorn, U.; Schreck-
enberger, M.; Kaina, B.; Ro¨sch, F. An approach to the evaluation of
the activity of the DNA repair enzyme O6-methylguanine DNA
methyl-transferase in tumor tissue in vivo: Syntheses of 6-benzyloxy-
9-(2-[18F]fluoroethyl)-9H-purin-2-yl-amine and 6-benzyloxy-1-(2-
[18F]fluoroethyl)7H-purin-2-yl-amine. Appl. Radiat. Isot. 2002, 56,
511-517.
(15) Zheng, Q.-H.; Liu, X.; Fei, X.; Wang, J.-Q.; Ohannesian, D. W.;
Erickson, L.; Stone, K. L.; Martinez, T. D.; Miller, K. D.; Hutchins,
G. D. Synthesis of radiolabeled O6-benzylguanine derivatives as new
potential PET tumor imaging agents for the DNA repair protein O6-
alkylguanine-DNA alkyltransferase. J. Lab. Compds. Radiopharm.
2002, 45, 1239-1252.
(16) Liu, X.; Zheng, Q.-H.; Fei, X.; Wang, J.-Q.; Ohannesian, D. W.;
Erickson, L. C.; Stone, K. L.; Hutchins, G. D. Synthesis and
preliminary biological evaluation of 6-O-[11C]-[(methoxymethyl)]-
benzylguanines, new potential PET breast cancer imaging agents for
the DNA repair protein AGT. Bioorg. Med. Chem. Lett. 2003, 13,
641-644.
(17) Liu, X.; Zheng, Q.-H.; Hutchins, G. D.; Fei, X.; Erickson, L. C.;
Miller, K. D.; Mock, B. H.; Glick-Wilson, B. E.; Winkle, W. L.;
Stone, K. L.; Carlson, K. A. A convenient procedure for the synthesis
of O6-benzylguanine derivatives by phase transfer catalysis. Synth.
Commun. 2003, 33, 941-952.
(18) Zheng, Q.-H.; Liu, X.; Fei, X.; Wang, J.-Q.; Ohannesian, D. W.;
Erickson, L. C.; Stone, K. L.; Hutchins, G. D. Synthesis and
preliminary biological evaluation of radiolabeled O6-benzylguanine
derivatives, new potential PET imaging agents for the DNA repair
protein O6-alkyl-guanine-DNA alkyltransferase in breast cancer. Nucl.
Med. Biol. 2003, 30, 405-415.
(19) Vaidyanathan, G.; Affleck, D. J.; Cavazos, C. M.; Johnson, S. P.;
Shankar, S.; Friedman, H. S.; Colvin, M. O.; Zalutsky, M. R.
Radiolabeled guanine derivatives for the in vivo mapping of O6-
alkylguanine-DNA alkyltransferase: 6-(4-[18F]Fluoro-benzyloxy)-9H-
purin-2-ylamine and 6-(3-[131I]iodo-benzyloxy)-9H-purin-2-ylamine.
Bioconjugate Chem. 2000, 6, 868-875.
(20) Vaidyanathan, G.; Affleck, D. J.; Norman, J.; Welsh, P.; Liu, W.;
Johnson, S. P.; Friedman, H. S.; Zalutsky, M. R. O6-3-[131I]-
Iodobenzylguanine: Improved synthesis and further evaluation of a
potential agent for imaging of alkylguanine-DNA alkyltransferase.
Bioconjugate Chem. 2004, 15, 402-408.
(21) Schirrmacher, R.; Mu¨hlhausen, U.; Wa¨ngler, B.; Schirrmacher, E.;
Reinhard, J.; Nagel, G.; Kaina, B.; Piel, M.; Wieâler, M.; Ro¨sch, F.
Synthesis of 2-amino-6-(2-[18F]fluoro-pyridine-4-ylmethoxy)-9-(oc-
tyl-â-D-glucosyl)-purine: A novel radioligand for positron emission
tomography studies of the O6-methylguanine-DNA methyltransferase
(MGMT) status of tumour tissue. Tetrahedron Lett. 2002, 43, 6301-
6304.
(22) Kaina, B.; Mu¨hlhausen, U.; Piee-Staffa, A.; Christmann, M.; Garcia-
Boy, R.; Ro¨sch, F.; Schirrmacher, R. Inhibition of O6-methylguanine-
DNA methyltransferase (MGMT) by glucose conjugated inhibitors:
Comparison with non-conjugated inhibitors and effect on fotemustin-
induced cell death. J. Pharmacol. Exp. Ther. 2004, 311, 585-593.
(23) D’Auria, M.; De Mico, A.; D’Onofrio, F.; Piancatelli, G. Synthesis
of naturally occurring bithiophenes: A photochemical approach. J.
Org. Chem. 1987, 52, 5243-5247.
(24) Kilburis, J.; Lister, J. H. Nucleophilic Displacement of the trimethyl-
ammonia-group: A new route to fluoropyridines. J. Chem. Soc. (C)
1971, 3942-3947.
(25) Schirrmacher, R.; Wa¨ngler, B.; Schirrmacher, E.; August, T.; Ro¨sch,
F. Dimethylpyridin-4-ylamine-catalysed alcoholysis of 2-amino-N,
N,N-trimethyl-9H-purin-6-ylammonium chloride: An effective route
to O6-substituted guanine derivatives from alcohols with poor
nucleophilicity. Synthesis 2002, 4, 538-542.
(26) Chumpradit, S.; Kung, M. P.; Billings, J.; Mach, R.; Kung, H. F.
Fluorinated and iodinated dopamine agents: D2 imaging agents for
PET and SPECT. J. Med. Chem. 1993, 36, 221-228.
(27) Moerlein, S. M. Regiospecific aromatic radioiodination via no-carrier-
added copper(I) chloride-assisted iododebromination. Radiochim. Acta
1990, 50, 55-61.
(28) Safadi, M.; Bindra, D. S.; Williams, T.; Moschel, R. C.; Stella, V.
Kinetics and mechanism of acid-catalysed hydrolysis of O6-ben-
zylguanine. Int. J. Pharm. 1993, 90, 239-246.
(29) Amersham Biosciences. Radiochemical DecompositionsA Guide to
the Properties and Storage of Radiolabeled Compound, p 11.
(30) Lindhorst, T. K. Essentials of Carbohydrate Chemistry and Bio-
chemistry; Wiley-VCH: Weinheim, 2000; p 89.
JM050588Q