Biodistribution and imaging of 1-(2-deoxy -β-D-ribofuranosyl)-2,4- difluoro-5-[123/125I]iodobenzene (dRF[123/125I]IB), a nonpolar thymidine-mimetic nucleoside, in rats and tumor-bearing mice

Article abstract of DOI:10.1080/15257770903051072

1-(2-Deoxy - D-ribofuranosyl)-2,4-difluoro-5-iodobenzene (dRFIB) is a putative bioisostere of iododeoxyuridine (IUdR). The advantages of dRFIB over IUdR for in vivo studies include resistance to both phosphorolytic cleavage of the nucleoside bond and de-i

Full text of DOI:10.1080/15257770903051072

Nucleosides, Nucleotides and Nucleic Acids, 28:379–393, 2009
Copyright ꢀC Taylor & Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770903051072
BIODISTRIBUTION AND IMAGING OF 1-(2-DEOXY-β-D-
RIBOFURANOSYL)-2,4-DIFLUORO-5-[123/125I]IODOBENZENE
(dRF[123/125I]IB), A NONPOLAR THYMIDINE-MIMETIC NUCLEOSIDE,
IN RATS AND TUMOR-BEARING MICE
Anke Stahlschmidt,1 Panteha Khalili,2 William Sun,2,3
Hans-Ju¨ rgen Machulla,1 Edward E. Knaus,2 and Leonard I. Wiebe2,3
1PET Center, Radiopharmacy, Eberhardt-Karls Universita¨t Tu¨bingen, Germany
2Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Germany
3PET Centre, Cross Cancer Institute, Edmonton, Canada
1-(2-Deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-iodobenzene (dRFIB) is a putative bioisostere of
2
iododeoxyuridine (IUdR). The advantages of dRFIB over IUdR for in vivo studies include resistance
to both phosphorolytic cleavage of the nucleoside bond and de-iodination. dRFIB was radioiodinated
(dRF123/125IB) by copper-catalyzed exchange using commercial sodium [123/125I]iodide. The in vivo
biodistribution of dRF[125I]IB in BALBc mice and imaging of dRF[123I]IB in Sprague-Dawley rats
are reported. In vivo data for rats show rapid clearance of radioactivity from blood (>95%ID in
15 minutes), extensive excretion in urine (56%ID/24 hours), concentration in the hepatobiliary-
small intestine system and very little fecal excretion (∼3%ID/24 hours). Pharmacokinetic data for
dRF[125I]IB (i.v. 48.7 ug/kg) in rats (t1/2[h] = 0.51 0.14, AUCinf[µg.min/mL] = 3.7 0.4,
Cl[L/kg/h] = 0.75 0.12, Vss[L/kg] = 0.96 0.18) confirm previously reported dose-dependent
pharmacokinetics. Scintigraphic images of rats dosed with dRF[123I]I were compatible with rapid
soft-tissue clearance and extensive accumulation of radioactivity in bladder/urine and liver/small
intestine. In tumor-bearing mice, thyroid and stomach radioactivity was indicative of moderate
deiodination. An unidentified polar radioactive metabolite was detected in serum.
Received 10 December 2008; accepted 19 March 2009.
This article is published as part of the Morris J. Robins 70th birthday issue.
The authors wish to thank Dr. Douglas Abrams and Mr. Dwayne Rakowski, Edmonton Radiophar-
macy Centre, for assisting with the imaging protocols; Dr. Terrance Riauka, Cross Cancer Institute, for
assistance with processing the gamma camera images; Mr. Hans-Jo¨rg Rahm, Tu¨bingen University, for
assistance with graphics; and Ms. Jennifer Yang for reviewing the kinetic analyses. L.I.W. was the recipient
of an Alexander von Humboldt fellowship during part of this work. This project was carried out with
partial support from the Canadian Institutes of Health Research (CIHR grant 14480).
Address correspondence to Leonard I Wiebe, 1807 Cross Cancer Institute, 11560 University Ave,
Edmonton T6G 1Z2, Canada. E-mail: Leonard.wiebe@ualberta.ca
379

380
A. Stahlschmidt et al.
Keywords Thymidine mimic; difluorophenyl deoxyribosides; dRF[123/125I]IB; biodis-
tribution and pharmacokinetics; scintigraphic imaging; thymidine kinase; nucleoside
transport
INTRODUCTION
Single photon emission computed tomography (SPECT) and positron
emission tomography (PET) are invaluable imaging-based techniques for
the early diagnosis of cancer and for cancer treatment planning and moni-
toring. Specific objectives of nucleoside-based imaging include monitoring
tumor cell proliferation and imaging the expression of Herpes simplex
type-1 thymidine kinase (HSV-TK) after HSV-tk transgene gene therapy.
Thus, metabolic trapping of radiolabelled nucleosides that are substrates for
mammalian thymidine kinase-1 (TK1) reflects cell proliferation on the prin-
ciple that TK expression correlates with the number of cells in the S-phase
of the cell cycle,[1] that is, accumulation of TK substrate radioactivity in a re-
gion/tissue of interest mirrors cell cycling in that region. Many nucleosides
labelled with gamma- or positron-emitting radionuclides have been pro-
posed and used for cell proliferation imaging, including 11C-thymidine (11C-
TdR)[2] and 5-[18F]fluoro-2ꢁ-deoxyuridine ([18F]FUdR).[3] 11C-TdR was the
first nucleoside to be used clinically for oncological PET imaging.[4] Today,
3ꢁ-[18F]fluoro-3ꢁ-deoxythymidine (FLT) is the most commonly used radio-
pharmaceutical for cell proliferation assessments in clinical oncology. FLT is
a good thymidine kinase substrate and at the same time offers the advantage
of resistance to phosphorolytic cleavage of the nucleoside bond, which
reduces the number of metabolites and thereby simplifies image analysis
and interpretation.[5–7] Radioiodinated nucleosides used for image-based
monitoring of HSV-TK expression in herpesvirus infections[8–10] and for
HSV-1 TK based transgene expression studies[11–14] must be a substrates
for HSV-1 TK, but not for mammalian TK1. The design of radiolabelled
nucleosides for imaging-based cell proliferation and viral TK-expression
monitoring,[15,16] and clinical and preclinical applications of HSV-TK based
transgene imaging[17,18] are the subject of numerous reviews.
Synthetic C-nucleosides represent one approach to designing better
diagnostic and therapeutic nucleosides. Early interest in C-nucleosides
was based on pseudouridine (ψ-uridine), a natural product[19] occurring
ubiquitously as a minor component in various tRNAs. Chemically, the
ψ-uridines are comprised of a uracil base coupled via C-5 to the ribosyl
moiety. Although these compounds are essentially devoid of biological
activity (ψ-isocytidine has growth inhibitory effects[20]), some, like 2ꢁ-deoxy-
l-methyl-ψ-uridine (ψ-thymidine; a C-nucleoside analogue of thymidine;[21]
Figure 1), have been utilized to develop strategies to expand the genetic

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
381
FIGURE 1 Chemical structures of the pyrimidine nucleosides thymidine (TdR) and iododeoxyuridine
(IUdR), and the difluorophenyl deoxyriboside thymidine mimics, dRFMB, dRFIB and dRFFMB, and
pseudothymidine (pseudo-TdR).
alphabet.[22] A study of the metabolism of radiolabeled ψ-thymidine in mice
demonstrated that this C-nucleoside is not hydrolyzed, but also failed to find
evidence that it was utilized metabolically in vivo.[23]
Other novel C-nucleoside thymidine analogues include those in
which a difluorophenyl ‘nucleobase’ is coupled to deoxyribose to form
nonpolar, lipophilic nucleoside mimics[24,25] that are phosphorylase
resistant[26] in comparison to TdR and other natural nucleosides which
undergo rapid degradation in blood.[27] When triphosphorylated, these
novel C-nucleosides, in which thymine is replaced by a bioisosteric
5-methyl-2,4-difluorophenyl moiety (Figure 1), are inserted into DNA
opposite adenine with high fidelity and undergo efficient incorpora-
tion into DNA fragments.[28] 1-(2-Deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-
iodobenzene (dRFIB), a bioisostere of iododeoxyuridine (IUdR; Figure 1),
is a weak TK substrate; its phosphorylation rate is only 10% of TdR, and
its phosphorylation is competitively inhibited by TdR.[29] Dose-dependent
pharmacokinetics of dRFIB at high doses have been reported, with more
rapid clearance and increased urinary excretion as doses decreased from
55 mg/kg to 5 mg/kg. Unchanged dRFIB and dRFIB-glucuronides are
excreted in urine.[26,30] A second difluorophenyl nucleoside, 1-(2-deoxy-β-
D-ribofuranosyl)-2,4-difluoro-5-[18F]fluoromethylbenzene (dRFFMB), pu-
tatively bioisosteric to FUdR or TdR, has been reported to undergo rapid
defluorination in vivo.[31] The potential of these C-nucleoside thymidine
mimics for use as cell proliferation imaging agents is therefore uncertain.
To further expand the knowledge base for this class of thymidine
mimics, preliminary pharmacokinetic data for radioiodinated dRFIB at
radiotracer doses of ∼50 µg/kg in rats are now reported, together with

382
A. Stahlschmidt et al.
scintigraphic images in rats and whole-body biodistribution data in tumor-
bearing mice.
EXPERIMENTAL
Radiolabelling
1-(2-Deoxy-ß-D-ribofuranosyl-2,4-difluoro-5-[123/125I]iodobenzene
(dRF123/125IB) was radiolabelled with either [123I]iodide or [125I]iodide
via isotope exchange as described elsewhere.[32] In a typical reaction, the
precursor (dRFIB; ∼1 mg) was dissolved in methanol (1 mL), and an
aliquot (100 µL; ∼100 µg dRFIB) was added to aqueous CuSO4 solution
(0.36 mM; 100 µL) mixed with aqueous (NH4)2SO4 buffer (3.6 mM;
100 µL). [123I]Iodide (40 MBq in 20 µL 0.1 M NaOH) was added, then
the reaction vial was closed, a syringe filled with activated carbon to trap
volatile radioiodine was attached for ventilation, and then the vial was
heated at 120◦C for 1 hour. The resulting solution was passed through a
C-18 cartridge pre-activated with methanol (10 mL) and washed with water
(10 mL). Following a water flush (10 mL), elution with methanol (4 mL)
afforded the labelled product (dRF[123/125I]IB) with radiochemical purity
>98% by radio-HPLC assay.
Animals were handled in accordance with guidelines established by the
Canadian Council of Animal Care. All experimental and surgical protocols
were approved by the University of Alberta Health Sciences Animal Welfare
Committee.
Pharmacokinetics
Adult male Sprague Dawley rats (250 g; Health Sciences Laboratory
Animal Services, University of Alberta) were fitted with an indwelling jugular
vein cannula 1 day before intravenous (i.v.) dosing with dRF[125I]IB (46.8
µg/kg). The animals were dosed via the catheter. After injection via the
catheter, it was flushed in situ with blood (150 µL, previously taken from the
rat) and then with saline (300 µL) to avoid contamination of subsequent
blood samples taken via this catheter.[33] Blood samples (250 µL) from
the first animal were withdrawn 1, 5, 10, 15, 30, 45, 60, 90, 120, 150, and
180 minutes after injection, to establish appropriate sampling times. On
this basis, samples were withdrawn 1, 3, 5, 7, 10, 15, 30, 45, 60, 90, 120,
180, and 1440 minutes after injection. The catheter was flushed with saline
(300 µL) after each withdrawal to clear residual blood and radioactivity,
and to maintain hydration of the animal. Blood samples were stored on ice
prior to work up. Aliquots (50 µL) of each blood sample were radioassayed
(total radioactivity) and larger aliquots (200 µL) were added to ice-cold
acetonitrile (650 µL) to precipitate proteins and macromolecules. The

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
383
latter samples were centrifuged and the supernatant was filtered through a
0.45 µm syringe filter. Aliquots of the resulting serum filtrates (50 µL) were
taken for radioassay, and second aliquots (100 µL) of serum were analyzed
by HPLC. Urine and feces of rats housed in metabolic cages were collected
for 24 hours and radioassayed. The volume of collected urine was measured;
three aliquots (50 µL each) were taken for radioassay and two aliquots (25
µL each) were subjected to HPLC analysis. Pharmacokinetic parameters
were estimated from blood radioactivity data using an industry standard
pharmacokinetics package (WinNonlin version 1.1; Pharsight Corporation,
USA).
Imaging
dRF[123I]IB was injected into two rats via their tail veins. One rat (250 g)
received 3.9 MBq of dRF[123I]IB (11.7 µg; 107 MBq/µmol), the second rat
(600 g) received a dose of 3.6 MBq. The rats were immobilized under light
isoflurane anesthesia during imaging. A dynamic scan (Siemens Orbiter
Digitrac 3700, LEAP//PH collimator, Odyssey FX Image software, V8.5.6
3/28/99 processor) was performed over the first 20 minutes (32 frames/30
sec) and static images (20 minutes acquisitions) were acquired after 30, 70,
100 and 130 minutes for the 250 g rat and 30, 60, 90, and 120 minutes for
larger rat.
Biodistribution
EMT6 tumor cells in suspension were injected subcutaneously under the
left front leg of male Balb/C mice (20–24 g; Health Sciences Laboratory Ani-
mal Services, University of Alberta). After the tumors grew to an appropriate
size (0.7–1 cc as estimated by calliper measurement), dRF[125I]IB (46.8
µg/kg) was injected via the tail vein. At preselected times after dosing, the
mice (n = 3 per time point) were euthanized, tissues and/or organs were
removed by dissection, and their weights were recorded prior to radioassay.
Blood and Urine Analysis
Total radioactivity per aliquot was determined by counting using a
well-type gamma scintillation counter. HPLC analyses were carried out
on a reverse phase Radial-Pak cartridge column (8NVC184µ, Nova-Pak,
C-18; Waters, Canada) with a guard column of the same material, using
acetonitrile:water 45:55 v/v at a flow of 1 mL/min. The retention time for
authentic dRFIB was 12.5 minutes; iodide eluted as a broad peak just before
dRFIB. Eluant fractions corresponding to UV absorption peaks (273 nm)
were collected for radiometry.

384
A. Stahlschmidt et al.
FIGURE 2 Clearance of radioactivity from blood, after bolus i.v. injection of dRF[125I]IB, as a function
of time, in rats (ꢀ) and mice (•). Data are means, n = 3 S.D.
RESULTS
The amount of radioactivity of dRF[125I]IB in rat blood decreased
rapidly over the first 5 minutes following injection, to <1% of the injected
dose (%ID). About 56% of the administered radioactivity was recovered in
urine over 24 hours, and another 3% of the injected dose was present in
0–24 hour faeces. The mean blood levels of dRF[125I]IB, expressed as a
percentage of injected dose/mL (%ID/mL), are plotted in Figure 2 as a
function of time after injection.
Blood samples were analyzed by radio-HPLC to quantify unchanged
dRF[125I]IB as a percentage of total radioactivity. These data, presented
as %ID/mL blood in Figure 3, show that unchanged dRF[125I]IB, as a
fraction of total blood radioactivity, decreased rapidly in the first hour, from
approximately 87% down to 11%, then increased at late intervals (24 hours)
to the fractional content present in the early fractions. That is, dRF[125I]IB
in the last fraction accounted for 78% of the total radioactivity in this blood
sample. This effect may be due to hepatobiliary recirculation, as postulated
for high dose dRFIB (53.4 mg/kg).[26] The radioactive metabolites observed
as HPLC peaks in the chromatograms were all more polar than dRF[125I]I as
indicated by their shorter retention times. Two radioactive metabolites were
found, but not unequivocally identified; one was thought to be radioiodide,

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
385
FIGURE 3 dRF[125I]IB as%ID/mL in rat blood after i.v. injection of dRF[125I]IB via a catheter. Data
are means, n = 3 S.D.
based on HPLC elution volume. An in vitro study incubation in blood
showed about 7% deiodination.
Analysis of the total radioactivity data (Figure 2) provided an elimination
half-life (t1/2 = 0.51 hours) that was relatively long compared to the t1/2 at
higher doses; the distribution volume (Vss = 0.96 0.18 L/kg) was greater
than the average total body water in rats, and clearance (Cl = 0.75 0.12
L/kg/h) was low at this tracer dose compared with literature data for much
larger doses of the unlabelled compound.[26] Pharmacokinetic analysis of
dRF[125I]IB data in the rat indicated a very short distribution half-life (0.04
hours), a large distribution volume (∼50 L/kg) and high clearance (∼20
L/kg/h), but these values are only rough approximations at best, since the
raw data could not be accurately modelled.
For both total radioactivity and dRF[125I]IB data, the large volumes of
distribution imply rapid and extensive intracellular distribution and are
compatible with the very rapid distribution phase observed. Literature data
for dRFIB are based on doses up to 53.6 mg/kg, about three orders of
magnitude greater than the dose used in this study (48.7 µg/kg), provide
an insight into the impact of dose on distribution kinetics. Selected phar-
macokinetic data from the literature and from this study are summarized in
Table 1.
Literature data for the intermediate doses (5 and 15 mg/kg) were ob-
tained using a two-compartment model, whereas the data for the radioactive
compound and 53.6 mg/kg dose were modelled using a noncompartmental
model. Data obtained for the radioactive compound (low dose dRF[125I]IB)

386
A. Stahlschmidt et al.
TABLE 1 Pharmacokinetic data for single intravenous doses of dRFIB[21,25] and tracer kinetics for
dRF[125I]IB, in rats (n = 3 SD)
Rat Total 125I
48.7 ug/kg
Rat dRFIB[30]
5 mg/kg
Rat dRFIB[30]
15 mg/kg
Rat dRFIB[26]
53.6 mg/kg
PK Parameter
t1/2 [h]
AUCinf
0.51 0.14
3.70 0.40
0.14 0.01
174 10
0.13 0.01
472 60
0.35 0.27
2872 149
[µg.min/mL]
Cl [L/kg/h]
Vss [L/kg]
0.75 0.12
0.96 0.18
1.72 0.06
0.30 0.03
1.93 0.23
0.33 0.02
1.23 0.17
0.38 0.06
could not be modeled satisfactorily due to the very rapid early clearance and
undulating blood levels.
Blood level data (Figure 2) for dRF[125I]IB in mice show more gradual
clearance than for rats, with 5% ID/g in blood after 15 minutes and 1%ID/g
after 2 hours. After 24 hours, concentrations in blood were virtually identical
in rats and mice. Accumulation of radioactivity in tumor reached a maxi-
mum of approximately 5% ID/g after 1 hour, with the highest tumor:blood
ratio (3.5) occurring 2 hours after dosing. The relatively small amounts of
radioactivity in 0–24 hours feces, when combined with gut, liver, and bile
data, support a model of hepatic uptake and metabolic degradation leading
predominantly to water-soluble metabolites which are excreted in the urine
(56%ID in total over 24 hours). Liver, thyroid, and tumor radioactivity levels
increased during the first hour following injection, whereas other tissues
like lung, kidney, and tail (injection site) essentially followed blood level
changes (Figure 4). Total thyroid radioactivity levels were approximately
5%ID between 15 and 240 minutes, and dropped to <1%ID after 24 hours.
Deiodination in vivo, based on combined thyroid and stomach radioactivity,
represented approximately 15% of the injected dose in mice (Figure 4), in
line with the deiodination observed during an incubation of dRF[125I]IB in
fresh whole blood in vitro (∼7%). Biodistribution data, expressed as%ID
per whole organ are provided in Figure 4.
Two static images obtained for rat #1 (weight 250 g) are shown in
Figure 5. Tissue perfusion by dRF[125I]IB was evident in the dynamic scan
(images not shown). Uptake by the liver and accumulation in the intestine,
visible almost from the start of imaging, was indicative of hepatobiliary
clearance. The first static image (70 minutes; Figure 5, right side image)
showed accumulation in the stomach and thyroid (deiodination), bladder
(renal clearance) and intestine, as well as major blood vessels and heart
(blood radioactivity). In later static images (130 minutes; Figure 4) the
background radioactivity decreased, while radioactivity accumulation in
stomach and thyroid increased. Radioactivity in the bladder was high
throughout, whereas it was low-moderate in the thyroid during the entire
imaging procedure. In contrast to the biodistribution study in mice, region

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
387
FIGURE 4 Biodistribution of dRF[125I]IB (46.8 µg/kg) in mice after i.v. bolus into tail vein. Data are
mean%ID/organ values, n = 3 S.D.
of interest measurements (rat images) revealed slight increases in thyroid
towards the end of the imaging procedure. Although the dose/g to the
second rat was only 40% of that to the first rat, and although the second rat
was much larger and older, only minor qualitative differences were observed
in the images of these two animals, mainly showing much less renal/bladder
radioactivity in the larger rat (not shown).
FIGURE 5 Static scintigraphic images of a rat (250 g), dosed with 3.9 MBq of dRF[123I]IB via bolus tail
vein injection. The images were taken 70 (left) and 130 (right) minutes after injection under isoflurane
anaesthesia. Acquisition time was 20 minutes per image. Both images depict radioactivity in bladder, liver-
intestine-stomach and thyroid in addition to generalized low-level radioactivity. Radioactivity in blood
appears to be provide above background image density in the region of the heart and aorta, and an
outline of the entire animal. Since dRF[123I]IB is very lipophilic, and given the rapid clearance from
blood in the rat (Figure 2), nonspecific partitioning from blood into soft tissues may also explain the
general high background observed.

388
A. Stahlschmidt et al.
DISCUSSION
Substantial in vitro biochemical DNA-level interaction of dRFMB as a
thymidine mimic has been reported.[28] A second member of this class
of thymidine mimics, dRFIB, has been reported to exhibit pyrimidine
nucleoside-like properties at the cellular level, as both a cell membrane
permeant through interaction with the human concentrative nucleoside
transporter 1 (hCNT1)[34] and as a relatively weak substrate for phos-
phorylation by nucleoside kinases.[29] Whole-body pharmacokinetic and
metabolic studies at high doses pointed to rapid blood clearance, substantial
hepatobiliary recycling with extensive renal elimination, dose-dependent
pharmacokinetics and minimal deiodination.[26,30]
The current biological evaluation of low radiotracer doses of
dRF[125I]IB in rats showed rapid clearance from blood, with serum levels
of dRF[125I]IB decreasing after i.v. bolus injection to a level below 1% of
the injected dose within the first 5 minutes (Figure 2). This rapid distri-
bution phase is to be expected given the small chemical dose (47 µg/kg)
and the high lipophilicity of dRF[125I]IB. High lipophilicity contributes
to rapid extravasation and dissociation into lipoidal tissues. As shown by
the pharmacokinetic analysis (Table 1) this gives rise to the relatively
large biodistribution volume (Vss = 0.96
0.18 L/kg) observed. Once
this fast distribution phase is complete (<5 minutes), the model changes
completely, that is, protracted blood clearance (Cl = 0.75 0.12 L/kg/h)
of the remaining traces of radioactivity as anticipated for a compound that
is highly protein bound. However, the pharmacokinetic model is compli-
cated by periodic increases in the concentration of total radioactivity and
dRF[125I]IB in blood. Comparative pharmacokinetic data are not available
for ψ-thymidine, but it appears that in contrast to dRF[125I]IB, it is not
cleared from blood, with between ∼4 and ∼6% of ID in blood at times
of 5–240 minutes after injection.[23] These differences in the distribution
phases may be due in part to their lipophilicity, since lipophilic compounds
are known to partition rapidly from blood into lipoidal elements includ-
ing fatty tissues and the phospholipid structures of cell membranes.[35]
dRF[125I]IB has a calculated partition coefficient (CLogP; Calculated using
ChemDraw Ultra10, CambridgeSoft, USA) of 2.1534, about four orders
of magnitude greater than ψ-thymidine (CLogP −2.5976) and two orders
greater than IUdR (CLogP -0.4248). The CLogP values are derived from
structural models and may vary somewhat from experimentally-derived
values. For example, dRFIB has a CLogP of 2.1534, compared to a reported
experimental value of 2.87.[26] Nonetheless, the CLogP values provide a
good basis for the purposes of this work.
Just over half (56%ID) of the dose of radioactivity was present in urine
collected over 24 hours, and analysis of the urine showed that only about 2%
of this radioactivity was due to unchanged dRF[125I]IB. Urinary metabolites

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
389
were not identified, but are likely to include glucuronide and sulphate
conjugates, which were detected at higher doses.[30] A comparison of dRFIB
literature data with the total radioactivity data obtained in this study shows
that clearance decreased by one-third when the dose was reduced from
53.4 mg/kg to 46.8 µg/kg, and at the same time, the distribution volume
increased by a factor of almost 3. The elimination half-life for the lower dose
was higher by a factor of 3.4 compared to the higher dose. Unfortunately,
pharmacokinetic analysis of dRF[125I]IB data was not possible because of the
complex, cyclic nature of changes in the concentration of dRF[125I]IB in the
blood as a function of time. By curtailing analysis at the 180 minutes interval,
some pharmacokinetic estimates were possible, but poor fit precluded
reliable estimation of statistical parameters. Total radioactivity data could be
modeled, but these data reflect not only the nucleoside, but also radioactive
metabolites.
In the earlier high dose studies, very little free dRFIB was found in bile;
the hepatic capacity for glucuronidation over 8 hours was constant, between
150 and 200 µg, for doses ranging from 5 to 54 mg/kg, indicating saturated
capacity at these doses. For the relatively small doses used in the current
study, this glucuronidation capacity would be adequate to conjugate all of
the dRFIB taken up by liver. These water-soluble metabolites (conjugates)
would be renally cleared after re-entering the blood stream and thereby
accounting for the high urinary excretion observed. In a competing path-
way, the conjugates would be excreted in bile, hydrolyzed in the intestines
and re-absorbed as the highly lipophilic (logP 2.87[26]) dRF[125I]IB. The
recycling pathway, together with re-distribution of dRF[125I]IB from tissues
back into the bloodstream, would account for the high fraction of dRF125IB
present in late serum samples, the low fecal excretion and the highly
complex dRF[125I]IB pharmacokinetics observed at the radiotracer dose.
Biodistribution studies with dRF[125I]IB in tumor-bearing mice showed
that most radioactivity was present in the bladder, bile, and stomach,
further supporting a model of hepatobiliary recycling leading to the for-
mation of water-soluble metabolites such as the glucuronide conjugates,
and consequent renal elimination. The distribution phase and blood clear-
ance in mice were slower than in the rat, but concentrations in blood
reached corresponding rat levels within 24 hours (Figure 2). However,
this prolonged distribution phase, with ∼1.5% ID in blood after 2 hours
and <0.5% at 4 hours, is still very fast in comparison to radioiodinated
iododeoxyuridine ([125I]IUdR), which retained 11.2% of the injected dose
in blood after 2 hours and 3.3% after 6 hours.[36] Biodistribution in
mice was characterized by persistent stomach and thyroid radioactivity. The
actual amounts in thyroid and stomach are indicative of deiodination; quite
surprisingly, thyroid retained about 5% of the ID throughout the study,
similar to the values reported for [125I]IUdR in the same animal/tumor

390
A. Stahlschmidt et al.
model.[36] Radioactivity levels in kidney, liver, and gut were consistent
with the hepatobiliary recycling model evident in the rat. The retention
of dRF[125I]IB in tumor was low, peaking at ∼5% ID at 1–2 hours, then
dropping to half at 4 hours and to negligible levels at 24 hours. Tumor
uptake was not investigated further, but given that dRF[125I]IB is a poor
substrate for mammalian thymidine kinases, and in line with the uptake and
clearance data, it is postulated that a combination of decreased perfusion
and concomitant slow back-diffusion account for this apparent uptake by
tumor tissue. The tumor uptake of [125I]IUdR, on the other hand, was
steady at ∼5% ID over the entire study period, indicative of metabolic
trapping.[36]
The SPECT images obtained after administration of dRF[123I]IB to rats
represent preliminary experiments involving only two animals. The animals
were anaesthetized during the imaging procedure, making it difficult to
compare the kinetic information in these images with the data from phar-
macokinetic studies, as the anaesthetic may alter the biodistribution in the
imaging studies.[37] However, the levels of radioactivity in the intestines of
both animals support the model of hepatobiliary metabolism. Radioactivity
in the stomach and thyroid are indicative of moderate deiodination.
The moderate deiodination is in stark contrast to defluorination data
for 1-(2-deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-[18F]fluoromethylbenzene
(dRFFMB), which was rapidly and extensively defluorinated in vivo,
yielding high radioactivity concentrations in bone.[31] 5-Fluoromethyl-, 5-
difluoromethyl- and 5-trifluoromethyl-2ꢁ-deoxyuridines are known to un-
dergo rapid defluorination under alkaline conditions[38–40] in which the
electronegative properties of the uracil ring facilitate solvolysis of the fluoro-
methyl bond. A similar mechanism may be operative for the difluorophenyl
moiety, possibly accounting for the facile defluorination of the fluoromethyl
substituent of dRFFMB. In contrast, dRFIB is not readily deiodinated. The
deiodination mechanism for IUdR and iodouracil, which readily undergo
alkaline hydrolysis,[41] is quite different than the case for dRFIB, the former
being a Michael system which facilitates C-5 dehalogenation by directing
uracil-ring electrons to C-5, whereas the 5-iodo-difluorophenyl moiety is
fully aromatic and therefore more resistant to cleavage of the C-I bond.
The active species responsible for the tissue radioactivity concentrations
observed have not been identified, but dRFIB-monophosphate is an un-
likely anabolite based on in vitro phosphorylation studies with TK1 and
TK2.[29] The low potential for in vitro phosphorylation can be inferred
from kcat/Km, the second order rate constant which indicates the catalytic
efficiency of the enzyme; kcat is the first order rate constant, that is, the
turnover number of the enzyme, which defines the maximum number of
substrate molecules converted to product per unit of time (s−1), and Km
represents the substrate concentration ([S]; µM) at half Vmax. kcat/Km

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
391
values for dRFIB were only 4% and 44% of TdR values for TK1 and TK2,
respectively.[29] This low phosphorylation potential was not predictive of the
apparent uptake of dRF[125I]IB by tumor, but was compatible with the ab-
sence of appreciable uptake into bone marrow, where high phosphorylation
reflects intensive cell cycling, seen in scintigraphic images of rats (Figure 5).
High marrow uptake is expected for proliferation markers, and is observed
in PET images using the proliferation marker FLT.[5]
SUMMARY
dRFIB demonstrated complex pharmacokinetics at radiotracer doses.
Rapid blood clearance with persistent and even increasing levels of dRFIB in
late blood samples, extensive urinary excretion, and minimal fecal excretion
support a model of extensive hepatobiliary recycling with minimal deiodina-
tion. Although these observations appear to preclude a future for dRFIB as
a cell proliferation imaging agent, additional studies are required to further
ascertain its potential to function as a marker for cellular expression of a
tk-transgene, for example, as a substrate of herpes simplex type-1 thymidine
kinase (HSV-1 TK).
ABBREVIATIONS
dRFIB
dRFFMB
1-(2-deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-iodobenzene
1-(2-deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-
[18F]fluoromethylbenzene
dRFMB
TdR
1-(2-deoxy-β-D-ribofuranosyl)-2,4-difluoro-5-methylbenzene
thymidine
IUdR
iododeoxyuridine
hCNT1
human concentrative nucleoside transporter 1
HSV-1 TK herpes simplex type-1 thymidine kinase
TK
%ID
mammalian thymidine kinase type-1
percentage of the injected dose
REFERENCES
1. Kornberg, A.; Baker, T.A., eds. Biosynthesis of DNA precursors, in DNA Replication 2nd Ed., Freeman,
W.H., New York, 1992, pp. 53–100.
2. Christman, D.; Crawford, E.J.; Friedkin, M.; Wolf, A.P. Detection of DNA synthesis in intact organisms
with positron-emitting [methyl-11C]thymidine. Proc. Nat. Acad. Sci. USA, 1972, 988–992.
3. Abrams, D.N.; Knaus, EE, McQuarrie, S.A.; Wiebe, L.I. 18F-5-Fluoro-2ꢁ-deoxyuridine as a radiophar-
maceutical for diagnostic oncology, in The Chemistry of Radiopharmaceuticals, eds. N.D. Heindel, H.D.
Burns, T. Honda, and L.W. Brady, MASSON, New York, 1976, pp. 205–213,
4. Martiat, P.; Ferrant, A.; Labar, D.; Cogneau, M.; Bol, A.; et al. In vivo measurement of carbon-11
thymidine uptake in non-Hodgkin’s lymphoma using positron emission tomography. J. Nucl. Med.
1988, 29, 1633–1637.
5. Shields, A.F.; Grierson, J.R.; Dohmen, B.M.; Machulla, H-J.; Stayanoff, J.C.; et al. Imaging prolifera-
tion in vivo with [F-18]FLT and positron emission tomography. Nature Med. 1998, 4, 1334–1336.

392
A. Stahlschmidt et al.
6. Shields, A.; Lim, K.; Grierson, J.; Link, J.; Krohn, K.A. Utilization of labeled thymidine in DNA
synthesis: studies for PET. J. Nucl. Med. 1990, 31, 337–342.
7. Grierson, J.R.; Schwartz, J.L.; Muz, M.; Jordan, R.; Krohn, K.A. Metabolism of 3ꢁ-deoxy-3ꢁ-[F-
18]fluorothymidine in proliferating A549 cells: validations for positron emission tomography. Nucl.
Med. Biol. 2004, 31, 829–837.
8. Samuel, J.; Knaus, E.E.; Wiebe, L.I.; Tyrrell, D.L. Synthesis of 131I, 125I and 82Br labelled (E)-5-(2-
halovinyl)-2ꢁ-deoxyuridines. Int. J. Appl. Radiat. Isot. 1984, 35, 1049–1052.
9. Samuel, J. Nucleosides for Herpes encephalitis diagnosis. Ph.D. thesis, University of Alberta, 1985.
10. Tovell, D.R.; Samuel, J.; Mercer, J.R.; Misra, H.K.; Xu, L.; et al. The in vitro evaluation of nucleoside
analogues as probes for use in the non-invasive diagnosis of Herpes Simplex encephalitis. Drug Design
Del. 1988, 3, 213–221.
11. Balzarini, J.; Morin, K.W.; Knaus, E.E.; Wiebe, L.I.; De Clercq, E. Novel (E)-5-(2-iodovinyl)-2ꢁ-
deoxyuridine derivatives as potential cytostatic agents against Herpes Simplex Virus-thymidine
kinase gene-transfected tumors. Gene Ther. 1995, 2, 317–322.
12. Tjuvajev, J.G.; Stockhammer, G.; Desai, R.; Uehara, H.; Watanabe, K.; et al. Imaging the expression
of transfected genes in vivo. Cancer Res. 1995, 55, 6126–6132.
13. Morin, K.W.; Wiebe, L.I.; Knaus, E.E. Non-invasive scintigraphic monitoring of gene expression in a
HSV-1 thymidine kinase gene therapy model. Nucl. Med. Commun. 1997, 18, 599–605.
14. Haberkorn, U.; Mier, W.; Eisenhut, M. Scintigraphic imaging of gene expression and gene transfer.
Current Med. Chem. 2005, 12, 779–794.
15. Wiebe, L.I. Applications of nucleoside-based molecular probes for the in vivo assessment of
tumour biochemistry using positron emission tomography (PET). Brazil Arch. Biol. Technol. 2007,
50, 445–459.
16. Wiebe, L.I.; Morin, K.W.; Knaus, E.E. Radiopharmaceuticals to monitor gene transfer. Quart. J. Nucl.
Med. 1997, 41, 79–89.
17. Inubushi, M.; Tamaki, N. Radionuclide reporter gene imaging for cardiac gene therapy. Eur. J. Nucl.
Med. Mol. Imaging, 2007, 34(Suppl. 1), S27–S33.
18. Kuruppu, D.; Dorfman, J.D.; Tanabe, K.K. HSV-1 viral oncolysis and molecular imaging with PET.
Curr. Cancer Drug Targets 2007, 7, 175–180.
19. Cohn, W.E. Pseudouridine, a carbon-carbon linked ribonucleoside in ribonucleic acids: isolation,
structure, and chemical characteristics. J. Biol. Chem. 1960, 235, 1488–1498.
20. Chu, C.K.; Reichman, U.; Watanabe, K.A.; Fox, J.J. Nucleosides. CIX. 2ꢁ-Deoxy-ꢀ-isocytidine,
2ꢁ-deoxy-ꢀ-uridine, and 2ꢁ-deoxy-1-methyl-ꢀ-uridine. Isosters of deoxycytidine, deoxyuridine and
thymidine. J. Heterocyclic Chem. 1977, 14, 1119–1121.
21. Matsuda, A.; Pankiewicz, K.; Marcus, B. K.; Watanabe, K. A.; Fox, J. J. Synthesis of 3-
methylpseudouridine and 2ꢁ-deoxy-3-methylpseudouridine. Carbohyd. Res. 1982, 100, 297–302.
22. Havemann, S.A.; Hoshika, S.; Hutter, D.; Benner, S.A. Incorporation of multiple sequential
pseudothymidines by DNA polymerases and their impact on DNA Duplex Structure. Nucleosides
Nucleotides Nucleic Acids 2008, 27, 261–278.
23. Grierson, J.R.; Shields, A.F.; Zheng, M.; Kozawa, S.M.; Courter, J.H. Radiosyntheses of labeled
pseudothymidine∗ ([C-11]- and [H-3 methyl) and its biodistribution and metabolism in normal and
tumored mice. Nucl. Med. Biol. 1995, 22, 671–678.
24. Schweitzer, B.A.; Kool, E.T. Aromatic nonpolar nucleosides as hydrophobic isosteres of pyrimidine
and purine nucleosides. J. Org. Chem. 1994, 59, 7238–7242.
25. Wang, Z-X.; Duan, W.; Wiebe, L.I.; Balzarini, J.; De Clercq, E.; Knaus, E.E. Synthesis of 1-(2-deoxy-b-
D-ribofuranosyl)-2,4-difluoro-5-substituted-benzene thymidine mimics, some related a-anomers; and
their evaluation as antiviral and anticancer agents. Nucleosides Nucleotides Nucleic Acids 2001, 20, 11–40.
26. Khalili, P.; Naimi, E.; Knaus, E.E.; Wiebe, L.I. Pharmacokinetics and metabolism of the novel
synthetic C-nucleoside, 1-(2-deoxy-b-D-ribofuranosyl)-2,4-difluoro-5-iodobenzene, a potential mimic
of 5-iodo-2ꢁ-deoxyuridine. Biopharm. Drug Dispos. 2002, 23, 105–113.
27. Shaw, T.; MacPhee, D.G. Rapid and complete degradation of thymidine by human peripheral blood
platelets: implications for genotoxicity assays. Mutation Res. 1986, 163, 75–80.
28. Kool, E.T.; Sintim, H.O. The difluorotoluene debate-a decade later. Chem. Commun. 2006,
3665–3675.
29. Al-Madhoun, A.S.; Eriksson, S.; Wang, Z.X.; Naimi, E.; Knaus, E.E.; Wiebe, L.I. Phosphorylation
of isocarbostyril- and difluorophenyl-nucleoside thymidine mimics by the human deoxynucleoside
kinases. Nucleosides Nucleotides Nucleic Acids 2004, 23, 1865–1874.

In Vivo Kinetics of the Nucleoside Mimic dRF[123/125I]IB
393
30. Khalili, P.; Naimi, E.; Sunm, W-Y., Knausm, E.E., Wiebe, L.I. Dose-dependent pharmacokinet-
ics of 1-(2-deoxy-b-D-ribofuranosyl)-2,4-difluoro-5-iodobenzene: A potential mimic of 5-iodo-2ꢁ-
deoxyuridine. Biopharm. Drug Dispos. 2003, 24, 385–395.
31. Issa, W.; Tochon-Danguy, H.J.; Lambert, J.; Sachinidis, J.I.; Achermann, U.; et al. Synthesis and
evaluation of a thymidine analogue for positron emission tomography study of tumor proliferation
in vivo. Nucl. Med. Biol. 2004, 31, 839–849.
32. Stahlschmidt, A.; Machulla, H-J.; Reischl, G.; Knaus, E.E.; Wiebe, L.I. Radioiodination of 1-(2-deoxy-
b-D-ribofuranosyl)-2,4-difluoro-5-iodobenzene (dRFIB), a putative thymidine mimic nucleoside for
cell proliferation studies. Appl. Radiat. Isotopes 2008, 66, 1221–1228.
33. Stypinski, D.; Wiebe, L.I.; Mercer, J.R. A rapid and simple assay to determine the blood and
urine concentrations of 1-(5-[123 125I]iodo-5-deoxyarabinofuranosyl)-2-nitroimidazole, a hypoxic
/
cell marker. J. Pharm. Biomed. Analysis 1997, 16, 1067–1073.
34. Smith, K.M.; Ng, A.M.L.; Yao, S.Y.M.; Labedz, K.A.; Knaus, E.E.; et al. Electrophysiological characteri-
zation of a recombinant human Na+-coupled nucleoside transporter (hCNT1) produced in Xenopus
oocytes. J. Physiol. (Lond.) 2004, 558, 807–823.
35. Avdeef, A. Physicochemical profiling (solubility, permeability and charge state). Curr. Top. Med. Chem.
2001, 1, 277–351.
36. Yang, X.-H.; Singh, S.; Kumar, P.; Diakur, J.; Wiebe, L.I. b-Cyclodextrin complexes of hydrolyzable
adamantanoyl-IUdR prodrugs—radioiodination and biodistribution in mice bearing implanted
KBALB tumours. Current Radiopharmaceuticals 2009, 2, 137–142.
37. Stypinski, D.D.; Wiebe, L.I.; Tam, Y.K.; McEwan, A.J.B. Effects of methoxyflurane anesthesia on the
pharmacokinetics of 125I-IAZA in Sprague-Dawley rats. Nucl. Med. Biol. 1999, 26, 659–665.
38. Nestler, H.J.; Garrett, E.R. Prediction of stability in pharmaceutical preparations XV. Kinetics of
hydrolysis of 5-trifluoromethyl-2ꢁ-deoxyuridine. J. Pharm. Sci. 1968, 57, 1117–1124.
39. Tandon, M.; Kumar, P.; Wiebe, G.; Wiebe, L.I. Detection of new metabolites of trifluridine (F3TdR)
using F NMR spectroscopy. Biochem. Pharmac. 1992, 44, 2223–2228.
40. Tandon, M.; Kumar, P.; Wiebe, L.I. α-Trifluoromethyl-β-alanyl glycine (F3MBAG): a novel mam-
malian metabolite of trifluridine (F3TdR). Biochem. Pharmac. 1994, 48, 1033–1041.
41. Garrett, E.R.; Nestler, H.J.; Somodi, A. Kinetics and mechanisms of hydrolysis of 5-halouracils. J. Org.
Chem. 1968, 33, 3460–3468.