A p-[18F]fluoroethoxyphenyl bicyclic nucleoside analogue as a potential positron emission tomography imaging agent for varicella-zoster virus thymidine kinase gene expression

Article abstract of DOI:10.1021/jm700971p

We recently reported a new positron emission tomography (PET) reporter gene, namely, varicella-zoster virus thymidine kinase (VZV-tk) in combination, with carbon-11 or fluorine-18 labeled m-alkoxyphenyl bicyclic nucleoside analogues (BCNAs) as PET reporte

Full text of DOI:10.1021/jm700971p

J. Med. Chem. 2007, 50, 6627–6637
6627
A p-[18F]Fluoroethoxyphenyl Bicyclic Nucleoside Analogue as a Potential Positron Emission
Tomography Imaging Agent for Varicella-Zoster Virus Thymidine Kinase Gene Expression
Satish K. Chitneni,† Christophe M. Deroose,3,‡ Jan Balzarini,§ Rik Gijsbers, Sofie Celen,† Zeger Debyser, Luc Mortelmans,‡
Alfons M. Verbruggen,† and Guy M. Bormans*,†
Molecular Small Animal Imaging Center (MoSAIC), Laboratory for RadiopharmacysFaculty of Pharmaceutical Sciences, Department of
Nuclear Medicine, Rega Institute for Medical Research, and DiVision of Molecular Medicine, Katholieke UniVersiteit LeuVen, LeuVen, Belgium
ReceiVed August 6, 2007
We recently reported a new positron emission tomography (PET) reporter gene, namely, varicella-zoster
virus thymidine kinase (VZV-tk) in combination, with carbon-11 or fluorine-18 labeled m-alkoxyphenyl
bicyclic nucleoside analogues (BCNAs) as PET reporter probes. We now report the synthesis and evaluation
of p-alkoxyphenyl-BCNA tracers ([11C]-4 and [18F]-5), which are found to be superior to the m-alkoxyphenyl-
BCNA tracers. In particular, the fluorine-18 labeled tracer ([18F]-5, IC50 of 5 is 4.2 µM) shows a higher
accumulation in VZV-tk expressing cells than the previously reported m-methoxyphenyl BCNA. [11C]-4
and [18F]-5 were synthesized by heating the phenol precursor 3 with 11CH3I and 18FCH2CH2Br, respectively,
as alkylating agents. In vitro evaluation of [11C]-4 and [18F]-5 in 293T cells showed about 14- and 54-fold
higher uptake, respectively, into VZV-tk gene-transduced cells compared to control cells. LC-MS analysis
confirmed the formation of monophosphate derivative of 5 upon catalysis by VZV TK. In vivo studies of
this new reporter gene/probe system are in progress.
Chart 1
Introduction
Development of reporter gene systems to monitor gene
expression and cell transplantation is the fastest-growing ap-
plication of positron emission tomography (PETa) technology
for small-animal research.1 Herpes simplex virus type 1 thy-
midine kinase (HSV1-tk) and its mutant HSV1-sr39tk, which
shows improved sensitivity to certain antiherpes compounds and
HSV-1-tk substrates, are the most extensively studied PET
reporter genes (PRGs) in combination with several radiolabeled
acycloguanosine and pyrimidine nucleoside analogues as PET
reporter probes (PRPs).2–5 Two receptor PRGs (for the dopamine
D2 receptor and the somatostatin receptor, respectively) and
one transporter PRG (for the sodium iodide symporter) have
been well characterized. However, these PRGs do not meet the
requirements for an ideal system because of endogenous
expression in certain mammalian tissues that leads to increased
background signal or because their endogenous ligands can bind
to the receptors, changing the biology of the cells expressing
PRGs.1,6,7
* To whom correspondence should be addressed. Address: Onderwijs
& Navorsing 2, Herestraat 49, Bus 821, BE-3000 Leuven, Belgium. Phone:
+32 16 330447. Fax: +32 16 330449. E-mail: guy.bormans@
pharm.kuleuven.be.
3
Molecular Small Animal Imaging Center.
Laboratory for Radiopharmacy.
Department of Nuclear Medicine.
Rega Institute for Medical Research.
To date, the majority of applications of radionuclide based
PRG/PRP technologies have used the HSV1-tk reporter gene,
but other herpes virus families, such as varicella-zoster virus
(VZV), encode for tk genes (note: tk refers to the gene, and
TK refers to the protein) that have different substrate require-
ments and therefore offer additional windows of opportunity
for gene expression imaging.8 Recently, a new class of antiviral
compounds, called bicyclic furopyrimidine deoxynucleoside
analogues (BCNAs), was developed by McGuigan et al.9
BCNAs are highly potent and selective inhibitors of VZV
replication.10 The specificity of BCNAs makes them suitable
candidates as PRPs for the VZV-tk reporter gene. Hence, we
have synthesized 11C or 18F labeled m-alkoxyphenyl-BCNAs 1
and 2 (Chart 1) and studied their biodistribution in normal mice
and uptake in VZV-tk expressing human embryonic kidney cells
(293T cells).11 In vitro enzymatic assays using the purified
recombinant VZV TK enzyme and nonradioactive 1 or 2 showed
that compound 1 has good affinity for the enzyme, with a 50%
inhibitory concentration value (IC50) of 4.8 µM. However,
compound 2 has poorer affinity for VZV TK (IC50 ) 53 µM),
†
‡
§
Division of Molecular Medicine.
aAbbreviations: VZV, varicella-zoster virus; tk, thymidine kinase gene;
TK, thymidine kinase enzyme; BCNAs, bicyclic nucleoside analogues;
HSV1, herpes simplex virus type 1; PET, positron emission tomography;
m, meta; p, para; [18F]-FHBG, 9-(4-[18F]fluoro-3-hydroxymethylbutyl)gua-
nine; BBB, blood-brain barrier; ClogP, calculated log partition coefficient;
CC50, 50% cytostatic concentration; 293T cells, human embryonic kidney
cells; DMF, dimethylformamide; RP-HPLC, reversed-phase high-perfor-
mance liquid chromatography; k, retention factor; HRMS, high-resolution
mass spectrometry; LC-MS, liquid chromatography coupled to mass
spectrometry; Da, daltons; m/z, mass-to-charge ratio; SPE, solid-phase
extraction; [18F]FEtBr, [18F]fluoroethyl bromide; [11C]MeI, [11C]methyl
iodide; MBq, megabecquerel; kBq, kilobecquerel; IC50, 50% inhibitory
concentration; %ID, percent of injected dose; pi, postinjection; SUV,
standard uptake value; LV, lentiviral vector; cDNA, complementary DNA;
EMCV IRES, encephalomyocarditis virus internal ribosome entry sequence;
LV-VZVTK-I-P, lentiviral vector encoding the VZV-tk gene and a
puromycin resistance gene (pac, puromycin-N-acetyltransferase); NaI(Tl),
thallium doped sodium iodide; ES, electrospray ionization; EOS, end of
synthesis; EtOH, ethanol; dThd, deoxythymidine; PBS, phosphate-buffered
saline.
10.1021/jm700971p CCC: $37.00
2007 American Chemical Society
Published on Web 11/30/2007

6628 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
Chitneni et al.
Table 1. Affinity of the Synthesized Nonradioactive Compounds for the
Enzymes VZV TK, HSV-1 TK, and Cytosolic TK-1
following the latter method. However, we obtained mainly a
disubstituted analogue with 4-hydroxyphenyl moiety on 5 and
6 positions of the furanopyrimidine base, similar to the earlier
report by Luoni et al.13 Furthermore, it was very difficult to
achieve separation of 3 from the 5,6-disubstituted side product.
Therefore, we have synthesized O-tert-butyldimethylsilyl (TB-
DMS) protected 4-hydroxyphenylacetylene (10, Scheme 1)
starting from (4-bromophenoxyl)-tert-butyldimethylsilane (8),
following a reported procedure.14 Initially, compound 8 was
treated with trimethylsilyl protected acetylene in dioxane in the
presence of catalytic amounts of PdCl2(PhCN)2, CuI, and tri(tert-
butyl)phosphane to give compound 9 in 79% yield. Treatment
of 9 with AgNO3 as catalyst in acetone containing 100 equiv
of water at room temperature (RT) resulted in chemoselective
protiodesilylation of the trimethylsilylethynyl group to give
compound 10 in 62% yield. Compound 10 was then coupled to
5-iodo-2′-deoxyuridine using Pd and cuprous ions, followed by
in situ cyclization in triethylamine and methanol at reflux to
obtain the fluorescent O-tert-butyldimethylsilyl protected com-
pound 12 in moderate yield (46%) (Scheme 2). The deprotection
of the phenol group was achieved by treating compound 12 with
tetrabutylammonium fluoride (TBAF) in THF to give p-
hydroxyphenyl-BCNA (3) in 73% yield. The reference O-methyl
derivative 4 was synthesized in 36% yield in the same way as
12 but starting from 4-methoxyphenylacetylene instead of 10.
The phenol precursor 3 was used as starting material for the
synthesis of nonradioactive reference compounds 5 and 6
(Scheme 3). Fluoroalkylation of the phenol precursor (3) was
achieved by heating with fluoroethyl bromide or fluorohexyl
tosylate (15) in the presence of potassium carbonate as a base,
followed by purification using semipreparative reversed-phase
high-performance liquid chromatography (RP-HPLC). Fluoro-
hexyl tosylate (15) was synthesized by tosylation of 1,6-
hexanediol to yield 1,6-hexanediol bis(p-toluenesulfonate) (14)
in 83% yield,15 and subsequent fluorination using 1 equiv of
TBAF in acetonitrile gave a 70% yield (Scheme 4).
a
IC50 (µM)
compd
R
VZV TK
HSV-1 TK
cytosolic TK-1
3
4
5
6
7
4-OH
4-OCH3
4-O(CH2)2F
4-O(CH2)6F
3-O(CH2)6F
0.4 ( 0.0
1.5 ( 0.2
4.2 ( 0.6
4.1 ( 0.1
>500
>500
>500
>500
>500
>500
>500
>500
>500
>500
>500
a 50% inhibitory concentration or compound concentration required to
inhibit nucleoside kinase-catalyzed phosphorylation of 1 µM [CH3-3H]dThd
by 50%.
which was evident in further cell uptake experiments using the
radiolabeled tracer [18F]-2. The maximum accumulation of
[11C]-1 and [18F]-2 in 293T cells was respectively 53- and 4.5-
fold higher in VZV-tk expressing cells compared to control cells.
The accumulation of these tracers in VZV-tk expressing cells
suggests that they are specifically phosphorylated by the VZV
TK enzyme present in these cells. This system thus constitutes
a new combination of a PRG (VZV-tk) and PRP ([11C]-1).
However, the short half-life (t1/2) of carbon-11 (20.4 min) limits
the in vivo imaging window to a maximum of about 100 min
postinjection (pi), after which insufficient counts are available
for statistically meaningful measurements.12
2′-Deoxy-2′-[18F]fluoro-5-fluoro-1-ꢀ-D-arabinofuranosylu-
racil ([18F]-FFAU) and 9-(4-[18F]fluoro-3-hydroxymethylbu-
tyl)guanine ([18F]-FHBG) are respectively the most sensitive
([18F]-FFAU) and most extensively studied ([18F]-FHBG) PRPs
for imaging HSV1-tk gene expression in animals. They both
contain fluorine-18 (t1/2 ) 110 min) as radionuclide, and PET
imaging with [18F]-FHBG has been performed up to more than
2 h pi, by which time the tracer was cleared from the plasma
and excreted into the intestines and the urinary bladder, resulting
in low background radiation from nontarget organs.2,3
For in vivo imaging of the newly developed VZV-tk reporter
gene, it can also be assumed that a fluorine-18 labeled PRP
must be preferred. However, as described above, [18F]-2 has
lower affinity for VZV TK and accordingly shows less ac-
cumulation in VZV-tk expressing cells. The lower affinity of
[18F]-2 compared to [11C]-1 may indicate that the affinity of
VZV TK for these m-alkoxyphenyl-BCNAs decreases as the
alkyl chain length increases. In order to examine this hypothesis,
we have synthesized m-fluorohexoxy derivative 7 (Scheme 3),
and it indeed shows no measurable affinity even at the highest
concentration studied (500 µM, Table 1).
Radiochemistry. The radiosynthesis of 4 and 5 was carried
out using compound 3 as precursor and [11C]methyl iodide
([11C]MeI) or [18F]fluoroethyl bromide ([18F]FEtBr) as labeling
agents, which were synthesized following reported procedures16,17
with some modifications. Briefly, [11C]-4 was synthesized by
heating 0.2 mg of the phenol precursor 3 with [11C]MeI in 0.2
mL of DMF in the presence of about 1 mg of Cs2CO3 at 90 °C
for 8 min. Similarly, [18F]-5 was prepared by heating the
precursor 3 (0.2 mg) with [18F]FEtBr in DMF (0.2 mL) in the
presence of about 1 mg of Cs2CO3 at 90 °C for 15 min. For
both tracers, the synthesis was followed by purification using
semipreparative RP-HPLC. The radiochemical purity of both
HPLC purified tracers was examined using HPLC on an
analytical RP C18 column and was always found to be >99%.
The identity of the tracers was confirmed by coelution with
authentic nonradioactive compounds after coinjection on ana-
lytical RP-HPLC, employing 30% acetonitrile in 0.05 M sodium
acetate buffer (pH 5.5) as mobile phase at a flow rate of 1 mL/
min (Figure 1; [11C]-4, retention factor k ) 1.19; [18F]-5, k )
1.71). The radiochemical yields were 33 ( 1% and 37 ( 2%
for [11C]-4 and [18F]-5, respectively (relative to starting
radioactivity of [11C]MeI and [18F]FEtBr, respectively, n ) 2),
and are comparable to the yields of m-alkoxyphenyl-BCNA
tracers [11C]-1 and [18F]-2.
In our search for radiolabeled BCNAs with higher affinity
for VZV TK, we have now synthesized stable and radiolabeled
p-alkoxyphenyl-BCNAs 4 and 5 (Scheme 3). This paper reports
their synthesis, their in vitro affinity for VZV TK, their
biodistribution, and a study of the in vivo stability of [18F]-5 in
normal mice. Finally, a cell uptake study with radiolabeled 4
and 5 was carried out in VZV-tk expressing 293T cells.
Results and Discussion
Chemistry. The phenolic precursor 3 can be synthesized
following the procedure described by McGuigan et al.,10 which
involves a Sonogashira coupling of 4-hydroxyphenylacetylene
to 5-iodo-2′-deoxyuridine under cocatalysis of Pd and cuprous
ions, followed by copper(I)-promoted in situ cyclization of the
coupled product. Alternatively, it can also be synthesized using
a coupling reaction between 4-iodophenol and 5-ethynyl-2′-
deoxyuridine in the presence of the above-mentioned catalysts.
Since the starting material 4-hydroxyphenyl acetylene is less
commercially accessible, we have tried the synthesis of 3
Partition Coefficient. The lipophilicity of the RP-HPLC
purified radiolabeled products was determined by partitioning
between 1-octanol and 0.025 M phosphate buffer, pH 7.4 (n )

p-[18F]Fluoroethoxyphenyl BCNA for VZV-tk imaging
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6629
Scheme 1. Synthesis of O-TBDMS Protected 4-Hydroxyphenyl Acetylene (10)a
i
t
a Reaction conditions: (a) Me3SiCCH, PdCl2(PhCN)2, Pr2NH, Bu3P, CuI, dioxane, RT, 4 h; (b) AgNO3, H2O, acetone, RT, 44 h.
Scheme 2. Synthesis of Phenol Precursor 3 for Radiolabelinga
i
a Reaction conditions: (a) Pd(PPh3)4, Pr2EtN, CuI, DMF, RT, 19 h; (b) Et3N/MeOH, CuI, reflux, 4 h; (c) TBAF, THF, RT, 1 h.
Scheme 3. Synthesis of [11C]-4 and [18F]-5 and Fluorinated
derivatives 5 and 6 is comparable to or slightly better than that
Reference Compoundsa
of the previously reported m-methoxyphenyl-BCNA 1 (IC50
)
4.8 µM).11 In contrast to their affinity for VZV TK, the test
compounds are not measurably phosphorylated by other nucleo-
side kinases examined (HSV-1 TK, cytosolic TK-1), with IC50
>500 µM (Table 1), which proves the specificity of these
compounds for VZV TK.
Biodistribution in Normal Mice. The results of in vivo
biodistribution studies of the radiolabeled tracers ([11C]-4 and
[18F]-5) in male NMRI mice are presented in Tables 2 and 3
and are similar to those of the m-alkoxyphenyl-BCNA tracers
[11C]-1 and [18F]-2. Blood clearance was rapid for both tracers
(e1.5% of injected dose (ID) in blood at 60 min postinjection
(pi)). The compounds were cleared by the hepatobiliary system
(for [11C]-4 46.7% and for [18F]-5 44.3% of ID in intestines at
60 min pi) and to a lesser extent into the urine (for [11C]-4
34.3% and for [18F]-5 34.8% of ID in urine at 60 min pi).
Except for liver (4.3–9.9% ID) and intestines (44.3–46.7%
ID), the radioactivity in other major organs was negligible at
60 min after injection of the tracers. A significant amount of
radioactivity still remained in the carcass even after 60 min pi
for [18F]-5 (11.9% ID). At 2 min pi, the highest concentration
of radioactivity was found in the kidneys with standard uptake
value (SUV) of 12.3 ( 1.1 for [11C]-4 and 13.1 ( 7.9 for [18F]-
5. This was succeeded by liver with SUV values of e5.4 for
both tracers. At 60 min pi, these values were 1.9 or less for all
organs, indicating good clearance of the tracers. As was the
case for [11C]-1 and [18F]-2, negligible brain uptake was
observed for [11C]-4 and [18F]-5 at both 2 and 60 min pi. Hansch
and co-workers have suggested that the ideal lipophilicity of
neutral molecules for passive diffusion into the central nervous
system is log Poct of about 2.18,19 However, Dischino et al. has
a Reaction conditions: (a) 11CH3I/Br(CH2)219/18F/TsO(CH2)6F, base,
DMF, ∆.
6). The log of the partition coefficient (log Poct) values of [11C]-
4 and [18F]-5 were found to be 1.3 and 1.2, respectively.
Affinity of Synthesized Compounds for Kinase Enzymes. The
nonradioactive compounds were evaluated for their affinity for
purified recombinant VZV TK, HSV-1 TK, and human cytosolic
TK-1, and the results are expressed as 50% inhibitory concen-
tration (IC50) values in Table 1. The phenolic precursor 3
displayed the highest affinity for VZV TK, with an IC50 value
of 0.4 µM. The O-methyl and O-fluoroethyl derivatives 4 and
5 also exhibited pronounced affinity with IC50 values of 1.5
and 4.2 µM, respectively. Unlike m-fluorohexoxyphenyl-BCNA
(7), the p-fluorohexoxyphenyl-BCNA 6 also showed good
affinity for VZV TK (IC50 ) 4.1 µM), indicating that the affinity
of the para-substituted compounds (4-6) for VZV TK does not
decrease as the alkyl chain length increases and/or that the
presence of a fluorine atom does not affect the affinity for the
enzyme for the p-alkoxyphenyl compounds in contrast to the m-
alkoxyphenyl derivatives. The affinity of the p-fluoroalkoxy

6630 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
Scheme 4. Synthesis of Fluorohexyl Tosylate 15 a
Chitneni et al.
a Reaction conditions: (a) tosyl chloride, Et3N, Et3N·HCl, CH3CN, 0–5 °C, 1 h; (b) TBAF, CH3CN, 100 °C, 6 h.
Figure 1. HPLC chromatograms showing the coelution of [11C]-4 (A) or [18F]-5 (B) after coinjection with respective authentic nonradioactive
compounds.
Table 2. Biodistribution of [11C]-4 in Normal Mice at 2 and 60 min pia
%IDb SUVc
657 Da.21 We and several others have been considering these
as upper range limits for prediction of BBB penetration of
radiotracers.22,23 In the present study, the log Poct values of [11C]-
4 and [18F]-5 were 1.3 and 1.2, respectively, both compounds
are neutral at physiological pH, and their molecular masses are
358 and 390 Da, respectively. Despite fulfilling the above-
mentioned theoretical requirements, [11C]-4 and [18F]-5 were
not taken up in the brain. We initially hypothesized that the
presence of a polar sugar moiety might be detrimental for BBB
penetration of these tracers, irrespective of their overall lipo-
philicity. However, biodistribution studies using 11C-labeled
BCNAs, which do not have a sugar moiety or where the 2′-
deoxyribose sugar in [11C]-1 was replaced with a (2-hydroxy-
ethoxy)methyl group (present in the antiherpes drug acyclovir),
suggest that the compounds based on the furopyrimidine base
system are unable to diffuse over the BBB irrespective of the
presence or absence of a sugar moiety because these tracers
were also not taken up in the brain.24 Hence, similar to [11C]-
1, the newly synthesized tracers ([11C]-4 and [18F]-5) can be
useful for imaging VZV infection as well as VZV-tk reporter
gene expression in vivo outside the brain only.
Distribution of [18F]-5 between Whole Blood and Pla-
sma. At 2 min after injection of [18F]-5, the ratio of radioactivity
concentration (%ID/g) in plasma to the radioactivity concentra-
tion in whole blood was 1.39 decreased slightly to 1.34 at 10
min pi. At 30 min pi the %ID/g values for plasma and whole
blood were almost the same (ratio 1.02). The compound seems
to diffuse through the cell membrane of the red blood cells as
a function of time pi.
Biostability of [18F]-5. In vivo stability of [18F]-5 after
intravenous (iv) injection in normal mice was assessed by RP-
HPLC analysis of plasma at 2, 10, and 30 min pi (n ) 2 per
time point) and of a urine sample at 1 h pi. Samples were
analyzed by employing a novel Oasis HLB (hydrophilic–lipo-
philic balanced) column, without prior precipitation of proteins
or removal of other biological matrixes, as described earlier.11
At 2 min after injection of [18F]-5, 94.7 ( 0.3% of the recovered
radioactivity in plasma was present as intact tracer and this was
at a level of 82.0 ( 2.1% at 10 min pi. However, at 30 min pi
organ
2 min
60 min
2 min
60 min
urine
1.4 ( 1.8 34.3 ( 3.1
kidneys
liver
spleen + pancreas
lungs
22.9 ( 1.2
20.3 ( 1.2
1.5 ( 0.2
1.2 ( 0.2
0.5 ( 0.0
1.2 ( 0.2
1.4 ( 1.2 12.3 ( 1.1 0.7 ( 0.6
9.9 ( 2.4
0.4 ( 0.1
0.1 ( 0.0
0.0 ( 0.0
1.0 ( 0.3
3.9 ( 0.3 1.9 ( 0.5
1.5 ( 0.3 0.4 ( 0.0
1.3 ( 0.2 0.1 ( 0.0
0.9 ( 0.1 0.1 ( 0.0
heart
stomach
intestines
brain
blood
carcass
16.8 ( 1.8 46.7 ( 0.5
0.0 ( 0.0
7.1 ( 0.7
29.5 ( 1.3
0.0 ( 0.0
0.6 ( 0.1
5.8 ( 0.5
0.1 ( 0.0 0.0 ( 0.0
1.0 ( 0.1 0.1 ( 0.0
a Data are expressed as the mean ( SD; n ) 3 per time point.
b Percentage of injected dose calculated as cpm in organ/total cpm recovered.
c Standard uptake values calculated as (radioactivity in cpm in organ/weight
of the organ (g))/(total counts recovered/body weight (g)).
Table 3. Biodistribution of [18F]-5 in Normal Mice at 2 and 60 min pia
%IDb
SUVc
organ
2 min
60 min
2 min
60 min
urine
kidneys
liver
spleen + pancreas
lungs
11.8 ( 8.1
20.1 ( 10.2
24.7 ( 3.0
1.4 ( 0.1
0.7 ( 0.3
0.3 ( 0.3
1.2 ( 0.4
17.7 ( 3.3
0.0 ( 0.0
4.8 ( 0.8
18.3 ( 2.9
34.8 ( 0.3
1.1 ( 0.0 13.1 ( 7.9 0.8 ( 0.1
4.3 ( 0.8
1.1 ( 0.1
0.2 ( 0.0
0.1 ( 0.0
1.1 ( 0.2
44.3 ( 0.1
0.0 ( 0.0
1.5 ( 0.2
11.9 ( 0.7
5.4 ( 0.6 0.8 ( 0.1
1.7 ( 0.1 1.3 ( 0.2
1.3 ( 0.5 0.3 ( 0.1
1.0 ( 0.9 0.4 ( 0.1
heart
stomach
intestines
brain
blood
carcass
0.1 ( 0.1 0.0 ( 0.0
0.7 ( 0.1 0.2 ( 0.0
a Data are expressed as mean ( SD; n ) 3 per time point. b Percentage
of injected dose calculated as cpm in organ/total cpm recovered. c Standard
uptake values calculated as (radioactivity in cpm in organ/weight of the
organ (g))/(total counts recovered/body weight (g)).
observed that (11C-labeled) compounds with log Poct values
between 0.9 and 2.5 were found to pass freely across the
blood-brain barrier (BBB).20 Further, Levin has suggested that
the molecular weight limit for a significant entry of a compound
into the brain (regardless of lipophilicity) is between 400 and

p-[18F]Fluoroethoxyphenyl BCNA for VZV-tk imaging
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6631
Table 4. In Vitro Incorporation of [11C]-4 and [18F]-5 in VZV-tk Gene-Transduced and Control 293T Cells at Time of Maximal uptake (0.5 and 3 h,
Respectively)
radioactivity in LV-VZVTK-I-P
transduced cellsa
radioactivity in
control cellsa
expt
compd
mean ( SD
28.8 ( 2.6
mean ( SD
2.1 ( 0.5
uptake ratiob
13.9
1
2
3
1
2
3
[
11C]-4
29.6
25.9
30.8
67.5
79.9
82.5
1.8
2.6
1.8
1.8
1.3
1.2
[
18F]-5
76.6 ( 8.0
1.4 ( 0.3
53.5
a Data are expressed as % tracer/mg protein. b p < 0.0005.
only 40.6 ( 4.3% of the recovered radioactivity was in the form
of unchanged tracer. A polar metabolite, which eluted in the
first wash fraction of the Oasis column (fraction no. 1), was
observed during these analyses. The recoveries of HPLC/Oasis
column injected radioactivity during these analyses were found
to be high with a mean recovery of 91.5 ( 10% (n ) 5). A
urine sample that was collected from a mouse at 1 h pi of [18F]-
5 was also analyzed following the same procedure as that for
plasma analysis. Although urine analysis does not require solid-
phase extraction (SPE) or workup of the sample prior to the
HPLC analysis, use of same online SPE method using an Oasis
column for plasma and urine samples enabled us to make a
better comparison of the metabolite profile of [18F]-5 between
both samples. On the basis of this analysis, 75.1% of the total
radioactivity found in the urine corresponds to [18F]-5 and 24.9%
of the radioactivity was in the form of polar metabolite that
was also present in plasma samples. This indicates the metabo-
lism of [18F]-5 to a polar metabolite in vivo in mice. However,
it is important to note here that the metabolism of a radiotracer
may differ significantly between different species.25 Therefore,
prior to its use in human studies, metabolite analysis of [18F]-5
in other animal species such as rats and monkeys is required in
order to confirm its in vivo metabolism. In general, BCNAs
are not susceptible to breakdown by the nucleoside catabolic
enzyme thymidine phosphorylase, which converts many pyri-
midine nucleosides into the inactive free base as shown before.26
On this basis and in view of the very high hydrophilicity of the
polar metabolite that was observed in plasma as well as in urine,
it is unlikely that this metabolite is the free base of [18F]-5.
Uptake of [18F]-5 in Bone and Muscle. The uptake of [18F]-
5 in bone as well as muscle was determined at 2, 10, or 30 min
pi of the tracer (n ) 2 per time point). The %ID/g value for
bones was 0.57, 0.66, and 1.25 at 2, 10, and 30 min, respectively.
At the same time points the %ID/g values for blood were
respectively 1.57, 1.42, and 0.63. Thus, at 30 min pi the bone-
to-blood ratio was about 2, indicating the accumulation of 18F
activity by bone. Bone uptake of a radiotracer could be attributed
to the accumulation of the tracer in the bone marrow and/or
free [18F]fluoride in bone matrix after defluorination. Therefore,
further studies are required in order to differentiate between bone
and marrow uptake of [18F]-5. Furthermore, a significant
difference was observed between mice and monkeys with regard
to defluorination for the widely used HSV1-tk reporter probe
[18F]FHBG. For this tracer ([18F]FHBG), a high uptake of [18F]
in bone was detected in mice but not in monkeys.27 On the
other hand, the uptake in muscle was minimal and decreased
with time (0.35 and 0.25 %ID/g at 2 and 30 min pi, respec-
tively).
acetyl-transferase) linked by the encephalomyocarditis virus
internal ribosome entry sequence (EMCV IRES), and control
cells. The LVs were produced,28 and cell cultures were
transduced as described earlier.11 Cell lines were denominated
LV-VZVTK-I-P or control cells and incubated with [11C]-4 or
[18F]-5 at 37 °C for different time intervals. It was found that
a maximal uptake was observed after 0.5 h of incubation for
[11C]-4 and 3 h of incubation for [18F]-5. Further evaluation of
cell uptake was done at the respective times of maximal uptake.
As shown in Table 4, uptake of [11C]-4 and [18F]-5 in the LV-
VZVTK-I-P cells was 13.9- and 53.5-fold higher, respectively,
than in control cells (n ) 3; p < 0.0005 for both, unpaired
bidirectional Student t test). Although the affinity of compound
4 for VZV TK is better, the uptake of [11C]-4 was less than
that of [18F]-5. The uptake ratio of [18F]-5 is the same as that
of the previously reported [11C]-1, and it is 12-fold higher than
that of the m-fluoroethoxyphenyl-BCNA [18F]-2, in the same
cell line. Moreover, the accumulation of both [11C]-4 (28.8%/
mg protein) and [18F]-5 (76.6%/mg protein) is higher than that
of [11C]-1 (12%/mg protein), indicating that the p-alkoxyphenyl-
BCNA tracers accumulate better than m-alkoxyphenyl-BCNA
tracers.
HPLC Analysis of Cell Lysate. The specific accumulation
of [11C]-4 and [18F]-5 in LV-VZVTK-I-P-transduced cells
suggests their phosphorylation by the VZV TK enzyme that is
present in these cells, in which case the phosphorylated
metabolites will be more polar than the parent tracer. In order
to evaluate this, the cell lysate that was obtained after incubation
of transduced cells with [18F]-5 for 90 and 180 min (as described
in the previous section) was analyzed using RP-HPLC and an
Oasis column. As expected, the HPLC profile of the cell lysate
(VZV-tk+ cells) at both time points was similar to that of
previously reported [11C]-1, with recovered radioactivity pre-
dominantly present (180 min postincubation) among two
hydrophilic components besides the residual parent tracer. The
first component was too polar to be retained on the Oasis column
and was therefore eluted in the first 3 mL of water rinse (fraction
no. 1). The second component eluted between 3 and 4 min
(fraction no. 6), while the parent tracer eluted at 8.2 min (in
fraction no. 11, as judged from the UV absorption/peak of the
coinjected nonradioactive compound 5). On the basis of reported
data on phosphorylation of BCNAs,29 we assume that the first
compound (metabolite b) is the diphosphorylated derivative and
the second is the monophosphorylated metabolite (metabolite
a), estimated as 17% and 32% of recovered radioactivity,
respectively, at 90 min postincubation. At 180 min, the
percentage of diphosphorylated metabolite increased to 56%,
whereas the monophosphate derivative remained at a similar
level (30%; Figure 2). These results are as expected and suggest
that [18F]-5 undergoes monophosphorylation and then diphos-
phorylation with longer incubation times. This further supports
our previous report on [11C]-1 in which we also described two
similar putative mono- and diphosphorylated metabolites.
In Vitro Evaluation of the Tracers. In vitro evaluation of
[11C]-4 and [18F]-5 was performed in 293T (human embryonic
kidney) cells transduced with human immunodeficiency virus
(HIV) type 1 derived lentiviral vector (LV), encoding the VZV-
tk gene and a puromycin resistance gene (pac, puromycin-N-

6632 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
Chitneni et al.
earlier,29 it was done by comparing the retention times with
those of chemically synthesized reference phosphorylated
compounds on RP-HPLC. In the present study we have achieved
the direct identity confirmation of the monophosphate formed
by the VZV TK mediated catalysis for the first time for this
class of compounds.
Conclusion
We have synthesized and evaluated p-alkoxyphenyl-BCNA
tracers as substrates for the enzyme VZV TK and compared
them to previously reported m-alkoxyphenyl-BCNAs. In vitro
affinity data show that precursor 3 has the highest affinity for
VZV TK, followed by the methyl ether 4. Unlike m-fluoro-
alkoxyphenyl-BCNAs, the ethoxy- and hexoxy-substituted
compounds 5 and 6 retained good affinity for VZV TK,
indicating that an increase in carbon chain length at the para
position is well-tolerated with respect to enzymatic affinity. The
loss of enzymatic affinity for the m-fluoroalkoxyphenyl-BCNAs
might be due to the difference in geometry and steric hindrances
of the side chain compared to the p-alkoxyphenyl-BCNAs.
Although the presence of a long alkyl or alkylaryl side chain is
a requirement for the high anti-VZV activity of BCNAs, this
could not be modeled into the recently resolved VZV TK
structure model without the steric clash with the protein.30
BCNAs might therefore adopt a different binding mode in VZV
TK compared with conventional nucleosides30 or may induce
a change in the three-dimensional structure of the VZV TK
protein upon binding. The present work focused on synthesizing
fluorine-18 labeled BCNAs that have good affinity for the VZV-
tk reporter gene and resulted in the identification of [18F]-5,
which showed an equal uptake ratio and better accumulation
than the previously reported [11C]-1. The radiochemical syn-
thesis produced [11C]-4 and [18F]-5 in amounts and purity
suitable for PET studies. This study also helps to understand
the structure–activity relationship (SAR) of alkoxyphenyl-BCNA
tracers as substrates for the VZV-tk reporter gene. LC-MS
analysis supports the formation of phosphorylated metabolites
of BCNAs in the presence of VZV TK. In view of its good
affinity for VZV TK, the long half-life of 18F as label compared
to 11C, and its higher accumulation rate, [18F]-5 is the most
promising agent for in vivo imaging of VZV-tk. RP-HPLC
analysis of plasma and urine samples of mice injected with [18F]-
5 revealed the conversion of this tracer into a more polar
metabolite. In vivo studies are in progress in mice bearing VZV-
tk expressing xenografts.
Figure 2. Metabolism of [18F]-5 in VZV-tk gene-transduced and
control 293T cells after 3 h of incubation.
However, [18F]-5 is converted to the diphosphorylated metabo-
lite (metabolite b) at a higher rate than [11C]-1 (56% for [18F]-
5 against 27% for [11C]-1), presumably because of the longer
incubation time for [18F]-5 (3 h for [18F]-5 against 30 min for
[11C]-1). As mentioned above, this indicates a progressive
conversion of mono- to diphosphorylated metabolite with longer
incubation time. Analysis of the cell lysate from control cells
showed no significant presence of metabolite a or b, and g90%
of the recovered radioactivity corresponded to the parent tracer
at 90 and 180 min time points (Figure 2).
Identification of Phosphorylated Metabolite(s) Using
Liquid Chromatography Coupled to Mass Spectrometry
(LC-MS). Because the control cells only differ from the LV-
VZVTK-I-P-transduced cells in their ability to synthesize the
VZV TK enzyme, this further supports the hypothesis that the
metabolites observed after incubation with the LV-VZVTK-I-
P-transduced cells are indeed the mono- and diphosphorylated
derivatives of [18F]-5 and [11C]-1. However, in order to achieve
accurate identification of these metabolites, we have carried out
LC-MS analysis of the incubation reaction mixture (Figure 3).
First, the nonradioactive p-fluoroethoxyphenyl-BCNA (5, 100
µmol) was incubated with purified VZV TK enzyme, similar
to the affinity tests but after 5-fold upscale of the ingredients
and in the absence of the competitor [CH3-3H]dThd. After
incubation for 2 h at 37 °C the reaction was terminated by
adding methanol (final concentration of ∼66%) followed by
centrifugation. The incubation mixture was then analyzed using
LC-MS, which showed a peak at 13.86 min (k ) 8.8) with an
accurate mass of 389.1152 Da corresponding to the theoretical
molecular ion mass of the parent compound 5 (Figure 4A).
Similarly, the single ion mass (468.67 ( 0.2 Da, mass
corresponding to the monophosphate derivative of 5) chromato-
gram showed a relatively polar peak at 10.20 min (k ) 6.2,
Figure 3B) and the accurate mass of this compound was
determined to be 469.0801 Da (Figure 4B) corresponding to
the theoretical molecular ion mass of the monophosphate
derivative, the structure of which is depicted in Figure 5. This
experiment confirms the conversion of BCNAs to their phos-
phate derivatives by the VZV TK enzyme and corroborates that
the hydrophilic metabolites observed with [18F]-5 and [11C]-1
in VZV-tk expressing cells are indeed the phosphorylated
metabolites. The formation of diphosphate derivative of 5 was
not observed in these experiments (as evident from RP-HPLC
and LC-MS analyses) probably because of the relatively short
incubation period. However, this indicates that after phospho-
rylation, the BCNA monophosphate dissociates from the binding
pocket of the VZV TK enzyme before the second phosphory-
lation step can occur. Although the conversion of BCNAs to
their mono- and diphosphates by the VZV TK was demonstrated
Experimental Section
Instruments and General Conditions. For ascending thin layer
chromatography (TLC), precoated aluminum backed plates (silica
gel 60 with fluorescent indicator, 0.2 mm thickness), supplied by
Macherey-Nagel (Düren, Germany), were used and visualization
1
of the spots was performed under UV light (254 or 366 nm). H
NMR spectra were recorded on a Gemini 300 MHz spectrometer
(Varian, Palo Alto, CA) using DMSO-d6 or CDCl3 as solvent.
HPLC purification and analysis were performed on an HPLC system
consisting of a Merck Hitachi L6200 intelligent pump (Hitachi,
Tokyo, Japan) or a Waters 600 pump (Waters Corporation, Milford,
MA) connected to a UV spectrometer (Waters 2487 dual λ
absorbance detector) set at 254 nm. For the analysis of radiolabeled
compounds, the HPLC eluate after passage through the UV detector
was led over a 3 in. NaI(Tl) scintillation detector connected to a
single channel analyzer (Medi-Laboratory Select, Mechelen, Bel-
gium). The radioactivity measurements during biodistribution
studies, biostability analyses, and cell-uptake studies were done
using an automatic γ counter (with a 3 in. NaI(Tl) well crystal)
coupled to a multichannel analyzer (Wallac 1480 Wizard 3”,

p-[18F]Fluoroethoxyphenyl BCNA for VZV-tk imaging
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6633
Figure 3. Single ion mass chromatograms for the mass corresponding to unreacted p-fluoroethoxyphenyl-BCNA (A; tR ) 13.86 min, k ) 8.8) and
its monophosphate derivative (B; tR ) 10.20 min, k ) 6.2) and the total ion current (TIC) chromatogram (C) of the phosphorylation reaction
mixture.
Wallac, Turku, Finland). Mass spectra were recorded on a time-
of-flight mass spectrometer (LCT, Micromass, Manchester, U.K.)
equipped with an orthogonal electrospray ionization (ESI) interface
and operating in positive mode (ES+). Samples were infused in
acetonitrile/water using a Harvard 22 syringe pump (Harvard
Instruments, MA). Accurate mass determination was done by
coinfusion with a 10 µg/mL solution of Kryptofix 222 as an internal
calibration standard. Acquisition and processing of data were done
using Masslynx software (version 3.5, Waters). LC-MS analysis
was performed on a system consisting of a Waters Alliance 2690
separation module connected to an XTerra MS C18 column (3.5
µm, 2.1 mm × 50 mm; Waters) eluted with gradient mixtures of
0.01 M ammonium acetate (pH 7.5)-acetonitrile (100:0 v/v at 0
min, linear gradient to 50:50 v/v in 6 min, isocratic elution with
50:50 v/v till 10 min) at a flow rate of 0.2 mL/min. The HPLC
eluate was led through a UV spectrometer (Waters PDA 996) into
the time-of-flight mass spectrometer (LCT, Micromass) with an ESI
interface operating in negative mode (ES-). For these experiments
a 10 µg/mL solution of uridine monophosphate was coinfused with
the HPLC eluate into the mass spectrometer as an internal
calibration mass. All reagents and solvents were obtained com-
mercially from Acros Organics (Geel, Belgium), Aldrich, Fluka,
Sigma (Sigma-Aldrich, Bornem, Belgium), or Fischer Bioblock
Scientific (Tournai, Belgium) and used as supplied.
purified using column chromatography on silica gel eluted with
hexane to give a pale-yellow liquid (2.24 g, 79%).
1-[(tert-Butyldimethylsilyl)oxy]-4-ethynylbenzene (10). Water
(13 mL, 0.74 mol) and silver nitrate (125 mg, 0.74 mmol) were
added to a solution of 9 (2.24 g, 7.35 mmol) in acetone (55 mL),
and the resulting mixture was stirred in the dark at RT for 44 h. It
was then poured into a saturated aqueous NaCl solution (88 mL)
and extracted with diethyl ether (6 × 40 mL). The organic extract
was washed with brine (3 × 25 mL), dried, and concentrated under
reduced pressure. The residue was purified using column chroma-
tography on silica gel with hexane as eluent to give a colorless
liquid (1.1 g, 62%).
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(4-tert-butyldimethylsilylo-
xyphenyl)-2,3-dihydrofuro[2,3-d]pyrimidin-2(3H)-one (12). To
a stirred solution of 10 (1.1 g, 4.58 mmol) in DMF (15 mL) under
nitrogen at RT were sequentially added 5-iodo-2′-deoxyuridine (11,
1.48 g, 4.17 mmol), N,N-diisopropylethylamine (1.54 mL, 8.7
mmol), tetrakis(triphenylphosphine)palladium(0) (0.488 g, 0.417
mmol), and copper(I) iodide (0.160 g, 0.83 mmol). The resulting
mixture was stirred for 19 h after which TLC showed complete
conversion of the starting material. Copper(I) iodide (0.148 g),
methanol (37 mL), and triethylamine (27.7 mL) were then added
to the reaction mixture, and this was heated under reflux for 4 h.
The solvents were evaporated under reduced pressure, and the
resulting residue was stirred with Amberlite IRA-400 (HCO3-) in
CH3OH/CH2Cl2 (1:1) (15 mL) for 1 h. The resin was filtered and
washed several times with methanol, and the combined filtrates
were evaporated to dryness. The crude product was purified using
column chromatography on silica gel eluted with gradient mixtures
of CH2Cl2 and CH3OH (up to 8%). Appropriate fractions were
combined, and the solvent was removed under reduced pressure to
yield a white solid (0.88 g, 46%).
{[4-(tert-Butyldimethylsilyloxy)phenyl]ethynyl}trimethylsil-
ane (9). To a stirred solution of bis(benzonitrile)palladium(II)
chloride (PdCl2(PhCN)2) (107 mg, 0.28 mmol) and copper(I) iodide
(35 mg, 0.38 mmol) in dioxane (3.5 mL) were sequentially added
tri(tert-butyl)phosphane (139 µL, 0.56 mmol), (4-bromophenoxyl)-
tert-butyldimethylsilane (8) (2.67 g, 9.3 mmol in 3.5 mL of
dioxane), (trimethylsilyl)acetylene (1.54 mL, 11.15 mmol), and
diisopropylamine (0.95 mL, 11.12 mmol) under nitrogen at RT.
The resulting mixture was stirred at RT for 4 h, and the reaction
mixture was diluted with EtOAc (70 mL) and filtered. The filtrate
was concentrated under reduced pressure and the residue was
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(4-hydroxyphenyl)-2,3-dihy-
drofuro[2,3-d]pyrimidin-2(3H)-one (3). To a stirred solution of
12 (0.77 g, 1.68 mmol) in THF (100 mL) under nitrogen was slowly

6634 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
Chitneni et al.
Figure 4. Accurate mass spectra corresponding to the unreacted p-fluoroethoxyphenyl-BCNA (A) and its monophosphate derivative (B) after
summing the respective peaks in Figure 3 and using uridine monophosphate (m/z 323.0280) as an internal lock mass.
added tetrabutylammonium fluoride (5.04 mL, 1 M solution in
THF). The resulting mixture was stirred at RT for 1 h after which
TLC showed complete conversion of the starting material. The crude
mixture was filtered. The residue was washed with THF and purified
using column chromatography on silica gel eluted with gradient
mixtures of CH2Cl2 and CH3OH (up to 12%). Appropriate fractions
were combined and the solvent was removed under reduced pressure
to yield the product as a yellow solid (0.48 g, 73%).
1,6-Hexanediol Bis(p-toluenesulfonate) (14). p-Toluenesulfonyl
chloride (8 g, 42.25 mmol) in acetonitrile (20 mL) was added
dropwise to a solution of 1,6-hexanediol (2.0 g, 16.9 mmol), Et3N
(9.5 mL, 67.6 mmol), and Et3N·HCl (0.32 g, 3.4 mmol) in
acetonitrile (17 mL) at 0–5 °C, and the mixture was stirred for 1 h.
The reaction was stopped by addition of water, and the solvents
were evaporated under reduced pressure. The residue was then
dissolved in ethyl acetate (100 mL) and extracted with water (2 ×
100 mL). The organic phase was dried (MgSO4) and concentrated.
The crude product was purified using column chromatography on
silica gel eluted with gradient mixtures of hexane and EtOAc (up
to 30%) to yield compound 12 as a white solid (6 g, 83%).
1-Fluoro-6-(p-toluenesulfonyl)hexanol (15). Tetrabutylammo-
nium fluoride (4.6 mL, 1 M solution in THF) was added to a
solution of 14 (2.0 g, 4.6 mmol) in dry acetonitrile (50 mL), and
the reaction mixture was stirred at 100 °C under reflux for 6 h.

p-[18F]Fluoroethoxyphenyl BCNA for VZV-tk imaging
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6635
µm; 10 mm × 250 mm; Alltech) that was eluted with 0.05 M
sodium acetate buffer (pH 5.5)/EtOH (65:35, v/v) at a flow rate of
2.5 mL/min ([11C]-4, k ) 0.9). The purity of the labeled tracer
was analyzed using RP-HPLC on an XTerra RP C18 column (5
µm, 4.6 mm × 250 mm; Waters), eluted with 0.05 M sodium acetate
buffer (pH 5.5)-acetonitrile (70:30, v/v) at a flow rate of 1 mL/
min (k ) 1.2). [11C]-4 was synthesized in 33 ( 1% radiochemical
yield (relative to starting [11C]MeI, nondecay-corrected, n ) 2),
and the average specific radioactivity was 98.6 GBq/µmol (2.66
Ci/µmol) at the end of synthesis (EOS).
Production of [18F]Fluoride, [18F]FEtBr, and [18F]-5. [18F]F-
was produced by an [18O(p,n)18F] reaction by irradiation of 0.5
mL of 97% enriched H218O (Rotem HYOX18, Rotem Industries,
Beer Sheva, Israel) in a niobium target using 18 MeV protons and
separated from [18O]H2O using a SepPak Light Accell plus QMA
anion exchange cartridge (Waters), which was preconditioned by
successive treatments with 0.5 M K2CO3 solution (10 mL) and water
(2 × 10 mL). The [18F]F- was then eluted from the cartridge into
a reaction vial with a solution containing 2.47 mg of potassium
carbonate and 27.9 mg of Kryptofix 222 in 0.75 mL of
H2O-CH3CN (5:95 v/v).
Figure 5. Proposed chemical structure of p-fluoroethoxyphenyl-BCNA
monophosphate.
The solvent was evaporated under reduced pressure and the residue
was purified using column chromatography on silica gel eluted with
gradient mixtures of hexane and EtOAc (up to 15%) to yield
compound 15 as a pale-yellow oil (0.88 g, 70%).
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(4-methoxyphenyl)-2,3-dihy-
drofuro[2,3-d]pyrimidin-2(3H)-one (4). The same procedure was
used as described for compound 12 but using 2.82 mmol of 11
and 7.57 mmol of 4-methoxyphenylacetylene. The product was
purified using column chromatography on silica gel eluted with
gradient mixtures of CH2Cl2 and CH3OH (up to 10%) to yield
compound 4 as a white solid (0.37 g, 36%).
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(4-(1-fluoroethoxy)phenyl)-
2,3-dihydrofuro[2,3-d]pyrimidin-2(3H)-one (5). To a solution of
3 (50 mg, 0.145 mmol) in DMF (10 mL) was added K2CO3 (41
mg, 0.29 mmol) and 1-bromo-2-fluoroethane (27.6 µL, 0.22 mmol)
at RT. The reaction mixture was stirred at 70 °C for 2 h after which
a more lipophilic and highly fluorescent spot appeared on TLC.
The product was purified by HPLC using a C-18 semipreparative
column (Econosphere, 10 mm × 250 mm; Alltech, Deerfield, IL)
eluted with water-acetonitrile (75:25, v/v) at a flow rate of 2.5
mL/min (k ) 3.1). After evaporation of the solvent, the product
was obtained as a white solid (30 mg, 53%).
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(4-(1-fluorohexyloxy)phen-
yl)-2,3-dihydrofuro[2,3-d]pyrimidin-2(3H)-one (6). The same
procedure was used as described for compound 5 but using 1-fluoro-
6-(p-toluenesulfonyl)hexanol (15) instead of 1-bromo-2-fluoroet-
hane. The product was purified by HPLC using a C-18 semiprepar-
ative column (Econosphere, 10 mm × 250 mm; Alltech) eluted
with 40% acetonitrile in water at a flow rate of 3 mL/min (k )
3.5). After evaporation of the solvent, the product was obtained as
a white solid (50 mg, 77%).
After evaporation of the solvent, [18F]F- was further dried by
azeotropic distillation of traces of water using acetonitrile. 2-Bro-
moethyl triflate (BrCH2CH2OTf) (5 µL) in o-dichlorobenzene (0.7
mL) was added into the vial containing [18F]F-. The reaction vial
was then heated to 110 °C, and the resulting [18F]FEtBr was distilled
using a flow of helium (3–4 mL/min) into another vial containing
a solution of 3 (0.2 mg) and Cs2CO3 (1 mg) in anhydrous DMF
(0.2 mL). After maximum radioactivity was distilled into the
solution, the reaction mixture was heated at 90 °C for 15 min.
Volatile impurities and unreacted [18F]FEtBr were evaporated by
flushing the mixture with helium at 50 °C until the radioactivity in
the vial remained at a stable level. The crude mixture was then
diluted with 1.6 mL of water and purified in the same manner as
for [11C]-4. The product was collected at 10.4 min (k ) 0.7), while
the precursor eluted at 6.4 min (k ) 0.1). The purity of the labeled
tracer was analyzed in the same manner as for [11C]-4 (k ) 1.7).
The mean decay-corrected radiochemical yield was 37 ( 2%
(relative to starting [18F]FEtBr, n ) 2). Starting from [18F]FEtBr,
the synthesis time to obtain the pure product was about 60 min.
The average specific activity was found to be 55.4 GBq/µmol (1.50
Ci/µmol) at EOS.
Partition Coefficient Determination. An aliquot of 25 µL of
RP-HPLC isolated labeled product containing approximately 550
kBq of [11C]-4 or 185 kBq of [18F]-5 was added to a tube containing
a mixture of 1-octanol and 0.025 M phosphate buffer, pH 7.4 (2
mL each). The test tube was vortexed at room temperature for 2
min followed by centrifugation at 3000 rpm (1837 g) for 5 min
(Eppendorf centrifuge 5810, Eppendorf, Westbury, NY). Aliquots
of about 60 and 500 µL were drawn from the 1-octanol and aqueous
phases, respectively, taking care to avoid cross-contamination
between the two phases, and weighed. The radioactivity in the
aliquots was counted using an automatic γ counter. After correction
for density and the mass difference between the two phases, the
partition coefficient (P) was calculated as the ratio (radioactivity
(cpm/mL) in 1-octanol)/(radioactivity (cpm/mL) in phosphate
buffer, pH 7.4).
3-(2′-Deoxy-ꢀ-D-ribofuranosyl)-6-(3-(1-fluorohexyloxyphenyl)-
2,3-dihydrofuro[2,3-d]pyrimidin-2(3H)-one (7). The same
procedure was used as described for compound 5 but using 3-
hydroxyphenyl BCNA as a starting material and 1-fluoro-6-(p-
toluenesulfonyl)hexanol (15) instead of 1-bromo-2-fluoroethane.
The product was purified by HPLC using a C-18 semipreparative
column (Econosphere, 10 mm × 250 mm; Alltech) eluted with 45%
acetonitrile in water at a flow rate of 3 mL/min (k ) 2.0). After
evaporation of the solvent, the product was obtained as a white
solid (40 mg, 62%).
Production of [11C]Methyl Iodide and Radiosynthesis of
[11C]-4. Carbon-11 was produced by a [14N(p,R)11C] nuclear
reaction. The target gas (a mixture of 95% N2 and 5% H2) was
irradiated using 18 MeV protons from a Cyclone 18/9 cyclotron
(Ion Beam Applications, Louvain-la-Neuve, Belgium) at a beam
current of 25 µA for about 25 min to yield [11C]CH4. This was
then reacted with vaporous I2 at 650 °C to convert it to [11C]methyl
iodide. The resulting volatile [11C]MeI was bubbled into a reaction
vial containing a solution of 3 (0.2 mg) and Cs2CO3 (1 mg) in
anhydrous DMF (0.2 mL). When maximum radioactivity was
trapped, the reaction mixture was heated at 90 °C for 8 min. Volatile
impurities and unreacted [11C]MeI were flushed out of the reaction
mixture by bubbling it with helium at 50 °C for about 1 min. The
crude reaction mixture was diluted with water and applied onto a
semipreparative HPLC column (HS HyperPrep RP C18, 100 Å, 8
Biodistribution in Normal Mice. Solutions of [11C]-4 and [18F]-
5 obtained after RP-HPLC purification were diluted using water
for injection to have the concentration of EtOH be <5%, and further
to a concentration of 37 and 3.7 MBq/mL, respectively. The
biodistribution of both tracers was determined in male NMRI mice
(weight, 37–43 g). The animal studies were performed according
to the Belgian code of practice for the care and use of animals,
after approval from the university ethics committee for animals. A
volume of 0.1 mL of the diluted tracer solution was injected into
the mice via a tail vein, under anesthesia (intraperitoneal injection
of 0.1 mL of a solution containing 3 mg of ketamine and 0.225
mg of xylazine). The mice were sacrificed by decapitation at 2 or
60 min after injection (n ) 3 at each time point). Blood was
collected in a tared tube and weighed. All organs and other body

6636 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
Chitneni et al.
parts were dissected and weighed, and their radioactivity was
counted in a γ counter. Results are expressed as %ID (cpm in organ/
total cpm recovered) or, where possible, as SUVs. SUVs were
calculated as (radioactivity in cpm in organ/weight of the organ)/
(total counts recovered/body weight). For calculation of total
radioactivity in blood, blood mass was assumed to be 7% of the
body mass.
concentrations of the test compounds. During this time period, the
enzymatic reaction proceeded linearly. Aliquots of 45 µL of the
reaction mixtures were spotted on Whatman DE-81 filter paper disks
(Whatman, Maidstone, U.K.). The filters were washed three times
for 5 min in 1 mM ammonium formate and once for 5 min in
ethanol. The radioactivity on the filters was determined by liquid
scintillation counting.
Distribution of [18F]-5 between Whole Blood and Plasma.
After intravenous administration of about 11.5 MBq [18F]-5 into
mice, the animals were sacrificed by decapitation at 2, 10, or 30
min pi (n ) 2 at each time point). Blood was collected into a
previously weighed BD vacutainer (containing 7.2 mg of K2EDTA;
BD, Franklin Lakes, NJ), and the radioactivity was measured in a
γ counter. The samples were centrifuged at 3000 rpm (1837g) for
10 min (Eppendorf centrifuge 5810, Eppendorf) to separate plasma.
The plasma was transferred to a 1.5 mL Eppendorf tube followed
by weighing and measuring of the radioactivity (γ counter) for
remaining blood as well as plasma samples. The disintegrations
per minute were converted to MBq normalized by sample weight
and the injected dose to obtain %ID/g. The ratio of [18F]-5
concentration in plasma to concentration in whole blood was
calculated by dividing the %ID/g value of plasma by the %ID/g
value of blood. After the measurement of radioactivity, the plasma
samples were placed on ice until HPLC analysis.
Cell-Uptake Studies. A lentiviral vector (LV) encoding the
cDNA of VZV-tk linked to the puromycin-N-acetyltransferase (pac)
gene through an encephalomyocarditis virus internal ribosome entry
sequence (IRES) was produced and denominated LV-VZVTK-I-
P. Human embryonic kidney cells (293T) transduced with LV-
VZVTK-I-P were maintained in cell culture medium containing 1
µg/mL puromycin. LV-VZVTK-I-P and control cells (293T) were
plated in triplicate at a density of 200 000 cells per well in 24-well
plates. After 24 h, the medium was discarded and 0.25 mL of fresh
medium with HPLC-purified tracer (1.1 MBq [11C]-4 or 185 kBq
[18F]-5 per well) was added. The cells were then incubated at 37
°C for 3 h for [18F]-5 and 0.5 h for [11C]-4. Following incubation
and removal of medium, the cells were washed three times with
0.4 mL of ice-cold phosphate-buffered saline (PBS). The cells were
then lysed with 0.25 mL of cell culture lysis reagent 1× solution
(Promega Corporation, Madison, WI) for 10 min after which the
lysate was collected, followed by a 0.125 mL rinse using the same
solution. The cell fractions (lysate and rinse) as well as the wash
fractions (medium and PBS) were collected separately for each well,
and the radioactivity was measured using a γ counter. The protein
concentration of each cell fraction was determined using the Bio-
Rad protein assay (Biorad, München, Germany) and a spectropho-
tometer (Bio Synchron-Anthos 2010, Anthos LabTec Instruments,
Austria) at 595 nm. The tracer uptake was normalized for total
protein content in the cell fraction for each individual well and
expressed as a percentage of total radioactivity per milligram of
protein.
Metabolism of [18F]-5 in Mice. The metabolic stability of [18F]-
5 was studied in normal mice by determination of the relative
amounts of the parent tracer and radiolabeled metabolites in plasma
samples at different time points and in a urine sample. The plasma
samples obtained from the above experiments at 2, 10, or 30 min
postinjection of tracer (n ) 2 per time point) were mixed with 20
µL of a 1 mg/mL solution of authentic 5 and injected onto an Oasis
HLB column (4.6 mm × 20 mm, Waters) that was preconditioned
by successive washings with acetonitrile and water. The proteins
and other biological matrices were rinsed off the Oasis column by
pumping water (6 mL), which was collected as two 3-mL fractions
(fractions 1 and 2). The outlet of the Oasis column was then
connected to an XTerra RP C18 column (5 µm, 3.9 mm × 250
mm; Waters), and both columns in series were then eluted using
0.05 M sodium acetate-acetonitrile (70:30 v/v) as the mobile phase
at a flow rate of 1 mL/min. After passage through an in-line UV
detector, the HPLC eluate was collected in 1 mL fractions and their
radioactivity as well as the activity in fractions 1 and 2 was
measured using a γ counter. Recoveries were calculated on the basis
of the amount of radioactivity (number of CPMs) injected onto the
Oasis column/HPLC and the amount of radioactivity (number of
CPMs) recovered from wash fractions as well as 1 mL eluate
fractions after HPLC analysis. A urine sample was collected at 60
min pi by manual voiding of the bladder, mixed with authentic 5,
and analyzed following the same procedure, without any pretreat-
ment or workup.
HPLC Analysis of [18F]-5 Metabolites from VZV-tk-Gene-
Transduced and Control 293T Cells. The cell lysate from the three
individual wells from above cell uptake experiments was pooled
together for both VZV-tk expressing and control cells at 90 or 180
min after incubation with [18F]-5. An amount of 0.9 mL of this
pooled cell lysate was mixed with 20 µL of a 1 mg/mL solution of
reference compound 5 and injected onto a preconditioned (as
described above) Oasis cartridge. The proteins and other biological
matrices were rinsed off the Oasis column by elution with 6 mL of
water (1 mL/min flow rate), and the eluate was collected as two
3-mL fractions (fractions 1 and 2). The outlet of the Oasis column
was then connected to an XTerra RP C18 column (5 µm, 4.6 mm
× 250 mm; Waters). The two columns in series were eluted using
0.05 M sodium acetate-acetonitrile (70:30 v/v) as the mobile phase
at a flow rate of 1 mL/min. The HPLC eluate was collected as 1
mL fractions, and the radioactivity of all fractions was measured
using a γ counter (5, k ) 1.7, fraction no. 11).
Uptake of [18F]-5 in Bone and Muscle. The bone and muscle
uptake of [18F]-5 was studied in normal mice at 2, 10, or 30 min
pi after injection of 11.5 MBq of the tracer (two animals at each
time point). After sacrifice at the specified time point, the femur
and the muscle were collected separately and weighed and their
radioactivity was measured in a γ counter. The uptake was
calculated by determining the %ID/g for the left as well as for the
right bone and muscle for all the mice at the above-mentioned time
points.
Affinity of Test Compounds for Nucleoside Kinase Enzymes
in Vitro. The 50% inhibitory concentration (IC50) of the test
compounds against phosphorylation of [CH3-3H]dThd as the natural
substrate for VZV TK, HSV-1 TK, and cytosolic TK-1 was
determined. Briefly, the activity of purified recombinant nucleoside
kinase enzymes was assayed in a 50 µL reaction mixture containing
50 mM Tris-HCl, pH 8.0, 2.5 mM MgCl2, 10 mM dithiothreitol,
2.5 mM ATP, 1.0 mg/mL bovine serum albumin, 10 mM NaF,
[CH3-3H]dThd (3.7 kBq in 5 µL, 1 µM final concentration), and 5
µL of recombinant enzyme (containing 7.75 ng of VZV TK or 226
ng of HSV-1 TK or 455 ng of TK-1 protein). The samples were
incubated at 37 °C for 30 min in the presence or absence of different
Acknowledgment. The technical assistance of Lizette van
Berckelaer and Ria Van Berwaer is gratefully acknowledged.
We thank Marva Bex, Peter Vermaelen, Humphrey Fonge, and
Christelle Terwinghe for their help during the experiments. This
work was supported by an SBO grant (Grant IWT-30 238) of
the Flemish Institute supporting Scientific-Technological Re-
search in industry (IWT), an IDO grant (Grant IDO/02/012) of
the Katholieke Universiteit Leuven, and in part the EC (FP6
Project DiMI LSHB-CT-2005-512146).
Supporting Information Available: Data on 1H NMR, HRMS,
and HPLC analysis for the synthesized intermediates and final
products. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(1) Herschman, H. R. Micro-PET imaging and small animal models of
disease. Curr. Opin. Immunol. 2003, 15, 378–384.

p-[18F]Fluoroethoxyphenyl BCNA for VZV-tk imaging
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6637
(2) Gambhir, S. S.; Bauer, E.; Black, M. E.; Liang, Q.; Kokoris, M. S.;
Barrio, J. R.; Iyer, M.; Namavari, M.; Phelps, M. E.; Herschman, H. R.
A mutant herpes simplex virus type 1 thymidine kinase reporter gene
shows improved sensitivity for imaging reporter gene expression with
positron emission tomography. Proc. Natl. Acad. Sci. U.S.A. 2000,
97, 2785–2790.
(16) Zhang, M. R.; Tsuchiyama, A.; Haradahira, T.; Yoshida, Y.; Furutsuka,
K.; Suzuki, K. Development of an automated system for synthesizing
18F-labeled compounds using [18F]fluoroethyl bromide as a synthetic
precursor. Appl. Radiat. Isot. 2002, 57, 335–342.
(17) Larsen, P.; Ulin, J.; Dahlstrom, K.; Jensen, M. Synthesis of
[
11C]iodomethane by iodination of [11C]methane. Appl. Radiat. Isot.
(3) Tjuvajev, J. G.; Doubrovin, M.; Akhurst, T.; Cai, S. D.; Balatoni, J.;
Alauddin, M. M.; Finn, R.; Bornmann, W.; Thaler, H.; Conti, P. S.;
Blasberg, R. G. Comparison of radiolabeled nucleoside probes (FIAU,
FHBG, and FHPG) for PET imaging of HSV1-tk gene expression.
J. Nucl. Med. 2002, 43, 1072–1083.
(4) Alauddin, M. M.; Shahinian, A.; Park, R.; Tohme, M.; Fissekis, J. D.;
Conti, P. S. Synthesis and evaluation of 2′-deoxy-2′-F-18-fluoro-5-
fluoro-1-beta-D-arabinofuranosyluracil as a potential PET imaging
agent for suicide gene expression. J. Nucl. Med. 2004, 45, 2063–2069.
(5) Yaghoubi, S. S.; Couto, M. A.; Chen, C. C.; Polavaram, L.; Cui, G. G.;
Sen, L. Y.; Gambhir, S. S. Preclinical safety evaluation of F-18-FHBG:
a PET reporter probe for imaging herpes simplex virus type 1
thymidine kinase (HSV1-tk) or mutant HSV1-sr39tk’s expression.
J. Nucl. Med. 2006, 47, 706–715.
(6) Herschman, H. R. PET reporter genes for noninvasive imaging of gene
therapy, cell tracking and transgenic analysis. Crit. ReV. Oncol.
Hematol. 2004, 51, 191–204.
(7) Serganova, I.; Blasberg, R. Reporter gene imaging: potential impact
on therapy. Nucl. Med. Biol. 2005, 32, 763–780.
(8) Wiebe, L. I.; Knaus, E. E. Enzyme-targeted, nucleoside-based radio-
pharmaceuticals for scintigraphic monitoring of gene transfer and
expression. Curr. Pharm. Des. 2001, 7, 1893–1906.
(9) McGuigan, C.; Yarnold, C. J.; Jones, G.; Velazquez, S.; Barucki, H.;
Brancale, A.; Andrei, G.; Snoeck, R.; De Clercq, E.; Balzarini, J. Potent
and selective inhibition of varicella-zoster virus (VZV) by nucleoside
analogues with an unusual bicyclic base. J. Med. Chem. 1999, 42,
4479–4484.
(10) McGuigan, C.; Barucki, H.; Blewett, S.; Carangio, A.; Erichsen, J. T.;
Andrei, G.; Snoeck, R.; De Clercq, E.; Balzarini, J. Highly potent
and selective inhibition of varicella-zoster virus by bicyclic furopy-
rimidine nucleosides bearing an aryl side chain. J. Med. Chem. 2000,
43, 4993–4997.
(11) Chitneni, S. K.; Deroose, C. M.; Balzarini, J.; Gijsbers, R.; Celen,
S. J. L.; de Groot, T. J.; Debyser, Z.; Mortelmans, L.; Verbruggen,
A. M.; Bormans, G. M. Synthesis and preliminary evaluation of 18F
or 11C labeled bicyclic nucleoside analogues as potential probes for
imaging varicella-zoster virus thymidine kinase gene expression using
positron emission tomography. J. Med. Chem. 2007, 50, 1041–1049.
(12) Wilson, A. A.; Dasilva, J. N.; Houle, S. Synthesis of two radioflu-
orinated cocaine analogues using distilled 2-[18F]fluoroethyl bromide.
Appl. Radiat. Isot. 1995, 46, 765–770.
(13) Luoni, G.; McGuigan, C.; Andrei, G.; Snoeck, R.; De Clercq, E.;
Balzarini, J. Bicyclic nucleoside inhibitors of varicella-zoster virus:
the effect of branching in the p-alkylphenyl side chain. Bioorg. Med.
Chem. Lett. 2005, 15, 3791–3796.
(14) Carpita, A.; Mannocci, L.; Rossi, R. Silver(I)-catalysed protiodesily-
lation of 1-(trimethylsilyl)-1-alkynes. Eur. J. Org. Chem. 2005, 1859–
1864.
1997, 48, 153–157.
(18) Hansch, C.; Steward, A. R.; Anderson, S. M.; Bentley, D. The parabolic
dependence of drug action upon lipophilic character as revealed by a
study of hypnotics. J. Med. Chem. 1968, 11, 1–11.
(19) Glave, W. R.; Hansch, C. Relationship between lipophilic character
and anesthetic activity. J. Pharm. Sci. 1972, 61, 589–591.
(20) Dischino, D. D.; Welch, M. J.; Kilbourn, M. R.; Raichle, M. E.
Relationship between lipophilicity and brain extraction of C-11-labeled
radiopharmaceuticals. J. Nucl. Med. 1983, 24, 1030–1038.
(21) Levin, V. A. Relationship of octanol/water partition coefficient and
molecular weight to rat brain capillary permeability. J. Med. Chem.
1980, 23, 682–684.
(22) Vanbilloen, H. P.; Kieffer, D. M.; Cleynhens, B. J.; Bormans, G. M.;
Mortelmans, L.; Verbruggen, A. M. Evaluation of 99mTc-labeled
tropanes with alkyl substituents on the 3ꢀ-phenyl ring as potential
dopamine transporter tracers. Nucl. Med. Biol. 2006, 33, 413–418.
(23) Elsinga, P. H.; Hendrikse, N. H.; Bart, J.; van Waarde, A.; Vaalburg,
W. Positron emission tomography studies on binding of central nervous
system drugs and P-glycoprotein function in the rodent brain. Mol.
Imaging Biol. 2005, 7, 37–44.
(24) Chitneni, S. K.; Balzarini, J.; Celen, S.; Dyubankova, N.; Verbruggen,
A. M.; Bormans, G. M. Relation between structure and brain uptake
of carbon-11 labeled acyclic and furo[2,3-d]pyrimidine-derivatives of
bicyclic nucleoside analogues (BCNAs). Unpublished data.
(25) Kida, T.; Noguchi, J.; Zhang, M. R.; Suhara, T.; Suzuki, K. Metabolite
analysis of [11C]Ro15-4513 in mice, rats, monkeys and humans. Nucl.
Med. Biol. 2003, 30, 779–784.
(26) Balzarini, J.; Sienaert, R.; Liekens, S.; Van Kuilenburg, A.; Carangio,
A.; Esnouf, R.; De Clercq, E.; McGuigan, C. Lack of susceptibility
of bicyclic nucleoside analogs, highly potent inhibitors of varicella-
zoster virus, to the catabolic action of thymidine phosphorylase and
dihydropyrimidine dehydrogenase. Mol. Pharmacol. 2002, 61, 1140–
1145.
(27) Alauddin, M. M.; Shahinian, A.; Gordon, E. M.; Bading, J. R.; Conti,
P. S. Preclinical evaluation of the penciclovir analog 9-(4-[F-18] fluoro-
3-hydroxymethylbutyl)guanine for in vivo measurement of suicide
gene expression with PET. J. Nucl. Med. 2001, 42, 1682–1690.
(28) Geraerts, M.; Michiels, M.; Baekelandt, V.; Debyser, Z.; Gijsbers, R.
Upscaling of lentiviral vector production by tangential flow filtration.
J. Gene Med. 2005, 7, 1299–1310.
(29) Sienaert, R.; Naesens, L.; Brancale, A.; De Clercq, E.; McGuigan,
C.; Balzarini, J. Specific recognition of the bicyclic pyrimidine
nucleoside analogs, a new class of highly potent and selective inhibitors
of varicella-zoster virus (VZV), by the VZV-encoded thymidine kinase.
Mol. Pharmacol. 2002, 61, 249–254.
(30) Bird, L. E.; Ren, J.; Wright, A.; Leslie, K. D.; Degreve, B.; Balzarini,
J.; Stammers, D. K. Crystal structure of varicella-zoster virus thymidine
kinase. J. Biol. Chem. 2003, 278, 24680–24687.
(15) Yoshida, Y.; Sakakura, Y.; Aso, N.; Okada, S.; Tanabe, Y. Practical
and efficient methods for sulfonylation of alcohols using Ts(Ms)Cl/
Et3N and catalytic Me3N·HCl as combined base: promising alternative
to traditional pyridine. Tetrahedron 1999, 55, 2183–2192.
JM700971P