Synthesis and in vitro evaluation of18F-labelled di- and tri(ethylene glycol) metomidate esters

Article abstract of DOI:10.1002/jlcr.1597

By replacing the alkyl chain in a metomidate ester with ¹⁸F-labelled di- or tri(ethylene glycol) chains, two ¹⁸F-labelled PET tracers, i.e. 2-(2-[¹⁸F]fluoroethoxy) ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate (1

Full text of DOI:10.1002/jlcr.1597

Research Article
Received 14 December 2008,
Revised 3 February 2009,
Accepted 4 February 2009
Published online 1 April 2009 in Wiley Interscience
(www.interscience.wiley.com) DOI: 10.1002/jlcr.1597
Synthesis and in vitro evaluation of
18F-labelled di- and tri(ethylene glycol)
metomidate esters
a
b,c
a
Ã
˚
Maria Erlandsson, Hakan Hall, and Bengt Langstrom
˚
¨
By replacing the alkyl chain in a metomidate ester with 18F-labelled di- or tri(ethylene glycol) chains, two 18F-labelled PET
tracers, i.e. 2-(2-[18F]fluoroethoxy)ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate (1) and 2-[2-(2-[18F]fluoroethoxy)-
ethoxy]ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate (2), were synthesized. Two precursors, 2-(2-bromoethoxy)ethyl
1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate and 2-[2-(2-chloroethoxy)ethoxy]ethyl 1-[(1R)-1-phenylethyl]-1H-imida-
zole-5-carboxylate, were prepared and used in one-step nucleophilic [18F]fluorination reactions using conventional and
microwave heating. Organ distribution, frozen section autoradiography and metabolite analysis were performed. The
decay-corrected radiochemical yields of 1 and 2 were 2678 and 2378%, respectively, when they were prepared using
conventional heating. By performing microwave heating, the reaction time could be decreased and the yields of analogues 1
and 2 could be increased to 57712 and 51718%, respectively. Organ distribution studies in the rat showed considerable
uptake in the lungs, adrenals and liver. Both compounds bound with low nonspecific binding (1: approx. 20–30%; 2: 2.9% or
lower) to tissue from pig and human normal and pathologic adrenals. Metabolite analyses were performed in rats after 5 and
30 min for tracer 1 (2076 and 271%) and tracer 2 (2775 and 674%). Both compounds are interesting candidates for the
detection of different types of adrenal disorders.
Keywords: n.c.a. nucleophilic 18F-fluorination; di- and tri(ethylene glycol); metomidate; analogues
made us interested to explore the one-step synthesis of two
18F-labelled analogues, 2-(2-[18F]fluoroethoxy)ethyl 1-[(1R)-1-phe-
Introduction
nylethyl]-1H-imidazole-5-carboxylate and 2-[2-(2-[18F]fluoroethoxy)
1–[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylic acid esters
(MTO) are potent inhibitors of 11-b-hydroxylase, a key enzyme
in the synthesis of cortisol and aldosterone within the adrenal
cortex.1,2 MTO analogues labelled with b1-emitting radio-
nuclides such as 11C (t1/2 = 20.3 min) and 18F (t1/2 = 109.7 min)
are thus of interest.
ethoxy]ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate.
The one-step 18F-labelling syntheses of the two analogues were
performed with both conventional and microwave heating.
The two new tracers were evaluated with regard to organ
distribution, frozen section autoradiography and metabolism to
find out whether the di- and tri(ethylene glycol) modifications
confer different biological properties.
In clinical studies, the radiotracer [11C]methyl 1-[(1R)-1-
phenylethyl]-1H-imidazole-5-carboxylate has shown high uptake
in lesions of adrenocortical origin, including adenomas, but low
uptake in lesions of nonadrenocortical origin.3–7 Recently,
Results and discussion
11C-labelled
1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylic
Chemistry
acid ester analogues have been synthesized and biologically
evaluated.8 The 18F-labelled tracer analogue 2-[18F]fluoroethyl
1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylate ([18F]FETO) has
also shown similar pharmacokinetic and pharmacodynamic
properties.9,10
To circumvent the problem of high lipophilicity that can result
from the addition of a fluoroalkyl group, we have developed
A crucial factor in the utility of radiolabelling procedures with aDepartment of Biochemistry and Organic Chemistry, Uppsala University, Box
576, Husargatan 3, BMC, S-751 23 Uppsala, Sweden
short-lived radionuclides is the time required for synthesis and
bUppsala Applied Science Lab, GE Healthcare, Box 967, S-751 09 Uppsala,
Sweden
purification. There is thus a need to find synthetic strategies that
can reduce the synthesis time. Microwave heating has been used
for the radiolabelling of different organic molecules with short-
lived radionuclides such as 11C and 18F.11 This technique not only
shortens the chemical reaction time but also reduces side
reactions, increases the yield and improves reproducibility.11
The biological properties of the lead tracer have been shown to
be relatively insensitive to the changes in the ester part.12 This
cDepartment of Public Health and Caring Sciences, Uppsala University, Uppsala
Science Park, S-751 85 Uppsala, Sweden
˚
¨
*Correspondence to: Bengt Langstrom, Department of Biochemistry and Organic
Chemistry, Uppsala University, Box 576, Husargatan 3, BMC, S-751 23 Uppsala,
Sweden.
E-mail: Bengt.Langstrom@biorg.uu.se
J. Label Compd. Radiopharm 2009, 52 278–285
Copyright r 2009 John Wiley & Sons, Ltd.

M. Erlandsson et al.
R 1
O
O
OH
O
1) TBAH
O
N
N
Br
Br
N
N
or
O
Cl
(3) R1 = (CH2)2O(CH2)2Br
Cl
O
(4) R1 = (CH2)2O(CH2)2O(CH2)2Cl
Scheme 1. Synthesis of precursor compounds 3 and 4.
two PET tracers based on the metomidate structure, which
contain di- or tri(ethylene glycol) substituents with [18F]fluorine
atoms attached at the end of the chains. By varying the chain
length of the ethylene glycol group in compounds 1 and 2, it
will be possible to adjust the lipophilicity and to maintain a
relatively small size. The calculated Log P values for the
compounds were 2.4770.59 (1) and 2.1170.69 (2), which are
lower compared with the 11C-labelled MTO lead.8,13 These
analogues may be interesting for further biological studies,
because of decreased lipophilicity, which is one of the key
determinants of pharmacokinetic properties.
R 1
R 2
O
O
O
O
[K/K2.2.2]+18F-
N
N
DMF/MeCN
N
N
(3) R1 = (CH2)2O(CH2)2Br
(4) R1 = (CH2)2O(CH2)2O(CH2)2Cl
Scheme 2. One-step labelling synthesis.
(1) R2 = (CH2)2O(CH2)218F
(2) R2 = (CH2)2O(CH2)2O(CH2)218F
Two precursors, compounds 3 and 4, were synthesized by
adding 2-bromoethyl ether or 1,2-bis(2-chloroethoxy)ethane to condition was found for each compound, the syntheses
the pre-treated 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylic were performed again and each reaction was HPLC-analysed
acid (Scheme 1). Compound 3 could not be synthesized using and the final d.c. radiochemical yield and SRA were determined
either conventional or microwave heating due to the formation of (Table 1).
byproducts. Instead, the best yield of compound 3, 53%, was
Five different temperatures (60, 100, 150, 170 and 2001C) were
obtained when the reaction temperature was allowed to slowly tested in the synthesis of analogue 1, and N,N-dimethylforma-
rise from 01C to room temperature over 20 h. In the synthesis of mide, acteonitrile and dimethyl sulfoxide were used as solvents.
The microwave syntheses of analogue
1 indicated that
precursor 4, both conventional and microwave heating were
evaluated. The disadvantage of conventional heating using an oil temperature was more important in determining the yield than
bath is that the vessel walls are heated up first, causing a the solvent was. Optimally, the reaction temperature should be
as high as the substrate and the product tolerate before
temperature gradient in the solution. However, very fast heating
can be obtained using microwave irradiation because the decomposing, or as high as the reaction solvent
sample is heated from inside, more uniformly at each point.14 allows. In this case 2001C was the temperature limit when
The synthesis of 4 using microwave heating gave a higher yield of N,N-dimethylformamide/acetonitrile, dimethyl sulfoxide/N,N-
the desired product and decreased the formation of the dimethylformamide or pure N,N-dimethylformamide was used.
byproducts compared with the synthesis using conventional The low yield of 1 might be a result from a reaction between
heating. The yield of compound 4 could be increased from 33 up dimethyl sulfoxide and the precursor 3. When acetonitrile was
to 56%.
used, the best temperature was 1501C; after that point, too
Compounds 1 and 2 were synthesized from their respective much pressure was built up in the vial. Two seconds of
halogen–ether chains by reacting compounds 3 and 4 with microwave heating was enough to obtain the highest d.c.
[K/K2.2.2]118Fꢀ (Scheme 2). Using conventional heating at
radiochemical yield in the reaction, regardless of the tempera-
1501C for 15 min produced 1 and 2 with specific radioactivities tures. A temperature increase to 2001C over 2 s gave the best
of 2074 and 175718 GBq/mmol, respectively, after a 50–60-min yields; at longer reaction times, the product started to
synthesis (Table 1). The radiochemical purity of both com- decompose. Compared with the conventional heating, micro-
pounds, as determined by analytical HPLC, was 495% and the wave heating increased the d.c. radiochemical yield of 1 from
decay-corrected (d.c.) radiochemical yields were 2678% for 1 2678 to 57712%.
and 2378% for 2.
The solvent mixture was an important factor in determining
In the literature, microwave heating has also been used to the yield of 2. However, by increasing the temperature from 150
enhance PET radiolabelling reactions, including one-step nu- to 2001C, the yield was doubled. To take advantage of the
cleophilic [18F]fluorination reactions, so that they become faster, microwave-heating effect, the reaction had to be carried out in a
solvent with high dielectric constant.17 Water was used first
cleaner and higher yielding.15,16
Microwave heating was performed on compounds 3 and 4 because it is the most common polar solvent, but in this case it
using different temperatures and solvents to find the optimized decomposed the precursor. The precursor was also destroyed
when the reaction was performed in N,N-dimethylformamide/
reaction condition (Tables 2 and 3). The reactions were heated in
the microwave oven for periods of 2–120 s and the samples water. N,N-dimethylformamide and dimethyl sulfoxide were also
were collected from the reaction mixtures at different time tested as solvents. A 3:1 mixture of N,N-dimethylformamide and
points and HPLC-analysed (Tables 2 and 3). When the optimized dimethyl sulfoxide gave the best yield. When more dimethyl
J. Label Compd. Radiopharm 2009, 52 278–285
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org

M. Erlandsson et al.
Table 1. Radiochemical yield and specific activity for compounds 1 and 2
Specific activityb
Compound
Radiochemical yielda
(%)
Radiochemical yielda,c
(%)
Specific activityb,c
(GBq/mmol)
(GBq/mmol)
Conventional heating
Conventional heating
Microwave heating
Microwave heatingc
1
2
2678 (n = 5)
2378 (n = 4)
2074 (n = 2)
175718 (n = 2)
57712 (n = 4)
51718 (n = 6)
4678 (n = 2)
1977 (n = 4)
aD.c., based upon [18F]fluoride at start and radioactivity of HPLC-purified product. n = number of experiments.
bRatio of radioactivity to amount of substance, d.c. to time at start of synthesis.
cCompound 3 reacted in acetonitrile/N,N-dimethylformamide at 2001C for 2 s; compound 4 reacted in dimethyl sulfoxide/N,N-
dimethylformamide 3:1 at 2001C for 2 s.
Table 2. Results from the microwave synthesis to find the optimized reaction condition for analogue 1
Product yield at different
Heating time (s)
reaction conditionsa (%)
2
10
15
20
30
60
120
MeCN/DMF 2001C
MeCN/DMF 1701C
MeCN/DMF 1501C
MeCN/DMF 1001C
MeCN/DMF 601C
DMSO/DMF 2001C
MeCN 1501C
6675
6474
5576
2673
18710
1574
6371
53721
6071
5674
5576
2878
1076
1371
6375
48715
5874
5873
5573
3076
1075
1371
6271
48711
5775
5873
5673
3175
1074
1872
6877
46713
5572
5975
5574
3176
1176
1872
6474
47712
5574
5874
6377
3177
1174
1475
6371
3972
5474
5774
5771
3075
1072
1574
6773
44711
DMF 2001C
aMeCN, acetonitrile; DMF, N,N-dimethylformamide. Average values from two experiments. The yields are d.c. based upon
[18F]fluoride at start and radioactivity of HPLC-purified product.
Table 3. Results from the microwave synthesis to find the optimized reaction condition for analogue 2
Product yield at different
Heating time (s)
reaction conditionsa (%)
2
10
15
20
30
60
120
MeCN/DMF 2001C
MeCN/DMF 1501C
DMSO/DMF 3:1 2001C
DMSO/DMF 6:1 2001C
DMF 2001C
1571
972
2274
1174
3375
1972
2171
2074
17711
2273
1175
3275
2272
2072
2276
20713
2072
1072
3073
2576
2072
2576
19711
2172
1073
3172
2172
2172
2476
19711
2272
1175
3274
1872
2371
2477
19712
2174
1173
3276
2173
2371
2676
17712
3575
2172
1971
1571
18712
MeCN 2001C
MeCN/DMF 1:1 2001C
aMeCN, acetonitrile; DMF, N,N-dimethylformamide. Average values from two experiments. The yields are d.c. based upon
[18F]fluoride at start and radioactivity of HPLC-purified product.
sulfoxide was used, the yield decreased due to higher water for 2 s increased the d.c. radiochemical yield for analogue 2 to
content. When the reaction using precursor 4 was performed in 51718%, compared with 2378% with conventional heating.
dimethyl sulfoxide, the yield was higher compared with the
For analogue 1, it was possible to increase the specific activity
reaction with precursor 3. The reason for this may be that the with microwave heating to 4678 GBq/mmol, compared with
reaction rate was slower with the chloro derivative then with the 2074 GBq/mmol with conventional heating. However, the result
bromo derivative. A practical problem with dimethyl sulfoxide is was the opposite for analogue 2; the specific activity decreased
that the workup procedure is complicated by the need to from 175718 GBq/mmol with conventional heating to
remove the solvent at the end of the reaction. However, when 1977 GBq/mmol with microwave heating (Table 1).
acetonitrile and N,N-dimethylformamide were used as solvents,
Microwave heating should produce an efficient internal heat
the yield decreased to lower levels than were obtained with transfer and result in minimized wall effects compared with
conventional heating. To find out if solvatization was a problem, conventional heating, in which the temperature of the reaction
the less polar solvent dioxane was added to N,N-dimethylfor- vessel is higher than that of the reaction mixture.18 The activity left
mamide and N,N-dimethylformamide/dimethyl sulfoxide, but in on the walls of the reaction vial after washing was 1376% (n= 10)
these cases no products were obtained; only unreacted starting in the microwave vial and 974% (n= 6) in the vial used for
material and 18F were recovered. Microwave heating at 2001C conventional heating. Hence, our experiments indicate that the
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 278–285

M. Erlandsson et al.
fluoride was attracted to the glass walls of the reaction vial adrenals, not visible in the normal and pathologic human tissue)
regardless of whether microwave or conventional heating was used. but was approximately 20–30% for 1 (pig adrenals). The two
radioligands showed qualitatively similar binding to the
Biology
different human normal and pathologic tissues, although some
quantitative differences could be observed.
Organ distribution
Analysis of radiolabelled metabolites in plasma
The distribution (relative to blood) of 1 and 2 in various male rat
organs is shown in Figure 1. The main distribution was to lungs,
adrenals, pancreas, kidney and liver, whereas the accumulation
was considerably lower in other organs. The accumulation of 1
was approximately tenfold higher in the lung than that of 2. The
accumulation in the adrenals was time-dependent, and was
markedly lower after 60 min with both compounds, whereas it
increased over time in the bone. Low radioactivity was found in
the brain (cerebrum or cerebellum).
Metabolite studies were performed on 1 and 2 in rats. Blood
samples were collected 5 and 30 min after injection, and the
amounts of unchanged tracer and radioactive metabolites they
contained were measured. Duplicate experiments were per-
formed for each tracer and time point (Table 4). For tracer 1, the
metabolite analysis revealed that 2076 and 271% of the
material were intact 5 and 30 min after injection, respectively.
The recovery was 79734%. One hydrophilic metabolite was
observed. For tracer 2, the analysis revealed that 2775 and
674% were intact 5 and 30 min after injection, respectively. The
recovery was 86716%. One hydrophilic metabolite was
Frozen section autoradiography
Both ligands bound specifically to different tissues from pig and observed.
human normal and pathologic adrenals (Figure 2). The
The biological evaluation of the metomidate analogues 1 and
nonspecific binding was low for 2 (2.9% of total in the pig 2 was in accordance with previous results.3,4,9 Hence, strong
100
75
1
50
25
15
10
5
Mean 5 min
Mean 30 min
Mean 60 min
0
20
15
10
5
2
Mean 5 min
Mean 10 min
Mean 30 min
0
Figure 1. Organ distribution of 1 and 2 in male rats (SUV, relative blood). The rats were injected intravenously with 10 MBq/per animal (n = 4 per time point) of one of the
two tracers.
J. Label Compd. Radiopharm 2009, 52 278–285
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org

M. Erlandsson et al.
Figure 2. Autoradiography showing total binding of 1 and 2 to various human (A–J) and pig (K) adrenal tissues. A, B: normal adrenals; C, D: adrenal cortex cancer; E, F:
aldosterone-producing adenoma; G, H: cortisol-producing adenoma; I, J: adrenal cortex hyperplasia; K: pig adrenal; L: standards on paper. The left tissue shows the tracer
binding and the tissue to the right is under blocking conditions. The slides were incubated for 50 min with 0.01–0.1 MBq/mL of one of the two tracers. Ten micromolar
etomidate was added to block specific binding.
Table 4. Results from metabolite studies for compounds 1 and 2 performed in rats
Intact tracera after 5 min (%)
Tracer
Intact tracera after 30 min (%)
Recoveryb (%)
1
2
2076
2775
271
674
79734
86716
n = number of experiments.
aIntact radiotracer = (radiotracer fraction/all fractions) ꢁ 100. n = 2.
bRecovery = (all fractions/analysed sample) ꢁ 100. n = 4.
accumulation was seen in normal pig and human adrenal cortex formate, pH 3.5 (C), and acetonitrile (D). For analytical LC, an ACE
in autoradiography and in rat adrenals in the organ distribution.
The activity is slightly increased in bone, which indicates
defluorination. The autoradiographic studies also show that
both compounds are likely candidates for the detection of
different types of adrenal carcinomas. The metabolism, probably
due to cleavage of the ester, is rapid for both compounds, which
might limit their usefulness as tracers for the in vivo visualization
of adrenal tumours. Both compounds also accumulate in the
lung, pancreas and liver, further indicating rapid metabolism in
line with earlier metabolism data.19
5 C18-HL 250 mm ꢁ 4.6 mm column was used at a flow rate of
1.5 mL/min. For semi-preparative LC, an ACE C18-HL
5
250 mm ꢁ 10 mm column was used at a flow rate of 5 mL/min.
Synthia, an automated synthesis system, was used for LC
injection and fraction collection.20 Data collection and LC
control were performed using
a Beckman System Gold
chromatography software package (USA). Radioactivity was
measured in an ion chamber (Veenstra Instrumenten bv, VDC-
˚
¨
202, Holland). A portable dose-rate meter (Langenas elektriska
AB, Sweden) was used to estimate radioactivity during synthesis.
Unlabelled reference substrates were synthesized and used for
comparison in the LC analysis of the 18F-labelled compounds.
Syntheses of precursors and references were conducted under a
nitrogen atmosphere using dried glassware and magnetic
stirring. The 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylates
were synthesized from the corresponding enantiopure phenyl-
ethylamine as described in the literature.21 All other chemicals
Experimental
General chemistry
Liquid chromatographic (LC) analysis was performed with a
Hitachi LaChromElite, pump L2130, injector L-220 and an L2400
UV detector in series with a b1-flow detector. The following
mobile phases were used: acetonitrile:water (50:7) (A), 25 mM and anhydrous solvents were bought from Sigma-Aldrich or
aqueous potassium phosphate (B), 0.05 M aqueous ammonium
Acros Organics. Analytical thin layer chromatography (TLC) was
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 278–285

M. Erlandsson et al.
performed using Merck silica gel 60 F254, and spots were 30, 60 and 120 s and repeated twice. Samples were taken from the
reaction mixture at the different time points and analysed by
preparative LC to determine the d.c. radiochemical yield. The
recovery for the analogues, calculated from the following formula:
recovery = (injected activity on the HPLC column/collected
activity) ꢁ 100, was 102714% (n= 21). When the optimized
condition was found the experiment was performed again and
the crude product was purified using the method described for
conventional heating. The d.c. radiochemical yield and the specific
radioactivity were finally determined.
visualized using UV light or I2. Preparative TLC was carried out
1
on 1-mm plates pre-coated with silica gel 60 F254. H and 13C
NMR spectra were recorded in CDCl3 (7.26 ppm 1H, 77.0 ppm
1
13C) or (CD3)2SO (2.5 ppm H, 39.5 ppm 13C) on a Varian Unity
400 spectrometer (400 MHz for 1H, 100.6 MHz for 13C and
376.3 MHz for 19F nuclei) or on a Varian Inova spectrometer
1
(500 MHz for H and 125 MHz for 13C nuclei). 19F NMR spectra
were recorded in CDCl3 with CFCl3 as the internal standard. Mass
spectra was recorded on a Quattro Premier instrument from
Waters, which has a triple quadrupole with electrospray
ionization and was operated in positive mode, or on a Finnigan
AQA using the mobile phases acetonitrile (0.1% formic acid) and
water (0.1% formic acid). A Gilson 322 pump and a Pheno-
menexs Gemeni 5-mm C18 110A 150 mm ꢁ 3.00 mm column
were also used. Microwave experiments were performed with a
SmithCreatorTM oven with monomodal radiation or on an
Initiator 2.0 (Biotage AB, Uppsala, Sweden).
[18F]fluoride was produced at Uppsala Imanet, GE Healthcare,
using the Scanditronix MC17 cyclotron and the 18O(p, n)18F
nuclear reaction through proton irradiation of 18O-enriched
water (95%, Rotem). The product solution of 18Fꢀ in water was
transferred from the cyclotron target by an HPLC pump and
trapped on a QMA filter (ABX, advanced biochemical com-
pounds, pre-conditioned Sep-PAKs, light QMA cartridge with
CO23ꢀ as counter-ions, Radeberg). The column was purged with
helium for 1 min. The [18F]fluoride adsorbed on the resin was
eluted into a reaction vial with 2 mL of the following solution:
12 mL of a 96:4 (by volume) acetonitrile:water mixture contain-
ing Kryptofix 2.2.2 and potassium carbonate (molar ratio 2:1).22
The solution was then evaporated and co-evaporated with
anhydrous acetonitrile (2 ꢁ 1 mL) to dryness in a nitrogen stream
at 1101C to give the dried [K/K2.2.2]118Fꢀ complex.
Precursor and reference synthesis
2-(2-Bromoethoxy)ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-car-
boxylate (3)
A solution of 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylic
acid (0.036 g, 0.16 mmol) in dichloromethane was activated by
adding tetrabutylammonium hydroxide (1 M in methanol,
0.2 mL). The reaction mixture was evaporated and co-evapo-
rated with dichloromethane, then further dried under vacuum.
The dried complex was then dissolved in anhydrous acetonitrile
(0.5 mL) and cooled to 01C. 2-Bromoethyl ether (40 mL,
0.32 mmol) was added and the mixture was brought to room
temperature and stirred for 20 h. Dichloromethane was added,
and the resulting mixture was extracted with water. The organic
phase was concentrated to a small volume. The crude product
was purified by preparative TLC (dichloromethane:methanol,
1
9.5:0.5) to give the title compound 3 as an oil (32 mg, 53%). H
NMR (CDCl3): d 7.84–7.81 (m, 2H), 7.35–7.26 (m, 3H), 7.19–7.17
(m, 2H), 6.34 (q, J = 7.4 Hz, 1H), 4.37 (t, J = 4.9 Hz, 2H), 3.81–3.75
(m, 4H), 3.43 (t, J = 6.1 Hz, 2H), 1.86 (d, J = 7.4 Hz, 3H). 13C NMR
(CDCl3): d 159.9, 140.9, 137.8, 129.0, 128.7, 128.2, 127.0, 126.3,
71.2, 69.0, 63.55, 55.85, 30.3, 22.3. LC-MS (ESI1), m/z 367 (50.7%),
369 (49.3%) [M1H]1.
General method for one-step 18F-labelling with conventional
heating
2-(2-Fluoroethoxy)ethyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-car-
boxylate (1)
The dried [K/K2.2.2]118Fꢀ complex was dissolved in anhydrous
N,N-dimethylformamide (0.2 mL) followed by the addition of the
precursor (3.370.3mg), dissolved in anhydrous acetonitrile (50 mL)
and N,N-dimethylformamide (0.2 mL). The mixture was heated in a
closed vessel at 1501C for 15 min. The crude product was dissolved
in water (2 mL) and injected onto the semi-preparative LC (1 and
2; tR = 8.88 and 8.59min, respectively) using the mobile phase C:D
(55:45). The organic solvent was evaporated from the collected
product fractions and the residue was dissolved in 2mL of
propylene glycol and phosphate buffer (pH = 7.5, 1:16). The
product was then collected into a sterile vial. The radiochemical
purities determined by analytical LC were 498% (1 and 2;
tR = 5.46 and 4.61 min, respectively, mobile phase A:B (50:50)).
Tetrabutylammonium fluoride (1 M in tetrahydrofuran, 0.33 mL)
was added to a solution of 3 (0.1 g, 0.29 mmol) in anhydrous
N,N-dimethylformamide (0.5 mL). The reaction mixture was
heated at 1301C for 3 h. Dichloromethane was added and the
resulting mixture was extracted with water. The organic phase
was dried with magnesium sulfate and concentrated to a small
volume. The crude product was purified by preparative TLC
(dichloromethane:methanol, 9.9:0.1) to give the title compound
1 as an oil (4.6 mg, 5%). 1H NMR (CDCl3, 500 MHz): d 7.83 (s, 1H),
7.74 (s, 1H), 7.35–7.17 (m, 5H), 6.33 (q, J = 7.0 Hz, 1H), 4.60–4.58
(m, 1H), 4.50–4.36 (m, 2H), 4.22–4.19 (m, 1H), 4.06–4.04 (m, 1H),
3.94–3.92 (m, 1H), 3.79–3.70 (m, 2H), 1.85 (d, J = 7.0 Hz, 3H). 13C
NMR (CDCl3, 300 MHz): d 167.8, 151.6, 144.5, 129.1, 129.0, 128.2,
126.4, 125.8, 87.5, 69.5, 66.1, 62.7, 55.7, 22.3. 19F NMR d ꢀ223.3
(m, F). LC-MS (ESI1), m/z 307 [M1H]1.
General method for one-step 18F-labelling with microwave
heating
To a microwave vessel (0.2–0.5 mL) containing the precursor
(2.570.4 mg of 3 or 4) dissolved in an appropriate solvent, the
2-[2-(2-Chloroethoxy)ethoxy]ethyl 1-[(1R)-1-phenylethyl]-1H-imida-
zole-5-carboxylate (4)
dissolved [K/K2.2.2]118 ꢀ
F complex was added. A magnetic stir bar
was used in all syntheses. The temperature and pressure were
monitored during the course of the reaction. The reaction vial was
cooled with pressurized air after irradiation. To find out the best
reaction condition, the experiments were performed for 2, 10, 15, 20,
A solution of 1-[(1R)-1-phenylethyl]-1H-imidazole-5-carboxylic
acid
activated by adding tetrabutylammonium hydroxide (1 M in
methanol, 0.35 mL). The reaction mixture was evaporated and
(0.062 g,
0.29 mmol)
in
dichloromethane
was
J. Label Compd. Radiopharm 2009, 52 278–285
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org

M. Erlandsson et al.
co-evaporated with dichloromethane, then further dried under bladder, testis, muscle, bone (femur), brain (cerebrum)
vacuum. The dried complex was dissolved in anhydrous and cerebellum were collected and weighed, and their
acetonitrile (0.30 mL) and cooled to 01C. 1,2-Bis(2-chloroethoxy)- radioactivity was determined and corrected for radioactive
ethane (0.45 mL, 2.85 mmol) was added and the reaction decay in an automated g-counter. Organ values were
mixture was brought to room temperature and stirred for 1 h. calculated as standardized uptake value (SUV), calculated as
The reaction was then microwave heated at 601C for 15 h. follows:
Dichloromethane was added and the resulting mixture was
extracted with water and concentrated to a small volume. The
ACTðBq=gÞ
DOSEðBqÞ=BWðgÞ
crude product was purified by preparative TLC (dichlorometha-
ne:methanol, 9.5:0.5) to give the title compound 4 as an oil
(58 mg, 56%). 1H NMR (CDCl3): d 7.85 (s, 1H), 7.83 (s, 1H),
7.36–7.18 (m, 5H), 6.36 (q, J = 7.1 Hz, 1H), 4.37 (q, J = 5.4 Hz, 2H),
3.77–3.60 (m, 10H), 1.87 (d, J = 7.1 Hz, 3H).13C NMR (CDCl3): d
160.0, 146.2, 140.6, 139.5, 137.0, 129.0, 128.3, 126.6, 126.4, 71.4,
70.7, 69.1, 63.8, 55.9, 42.8, 22.2. LC-MS (ESI1), m/z 367 (75.8%),
369 (24.2%) [M1H]1.
SUV ¼
where ACT is the measured concentration of radioactivity
corrected for physical decay, DOSE is the administered amount
of radioactivity (Bq) and BW is the body weight of the
subject (g).
Frozen section autoradiography
2-[2-(2-Fluoroethoxy)ethoxy]ethyl 1-[(1R)-1-phenylethyl]-1H-imida-
zole-5-carboxylate (2)
Sections (25 mm) were prepared in a cryomicrotome and put on
s
SuperFrost glass slides (SuperFrost Plus, MenzelGlaser, Ger-
¨
Tetrabutylammonium fluoride in tetrahydrofuran (45 mL) was
added to a solution of 4 (0.012 g, 0.03 mmol) in anhydrous N,N-
dimethylformamide (0.4 mL). The reaction mixture was heated in
the microwave at 2001C for 1.4 h. Dichloromethane was added
and the resulting mixture was extracted with water. The organic
phase was concentrated to a small volume. The crude product
was purified by semi-preparative LC (C:D, 50:50) to give the title
many). The slides were kept in a freezer (ꢀ201C) for 1 day until
used. At the start of the experiment, the slides were pre-
incubated for 10 min in Tris–HCl (50 mM, pH 7.4) buffer. The
slides were then transferred to containers that held 1 or 2
(0.01–0.1 MBq/mL) in Tris–HCl buffer. In a duplicate set of
containers, 10 mM etomidate was added to block specific
binding. After incubation for 50 min, the slides were washed
3 ꢁ 3 min in cold Tris–HCl buffer. The slides were dried in a
heated (371C) oven and exposed to phosphor imaging plates
(Molecular Dynamics) for 2 h. Twenty microlitres of the incubate
was pipetted onto a filter paper (standard activity area), which
was exposed in parallel to the experimental sections. The plates
were then scanned in a Phosphor Imager Model 400S using
100-mm pixel width (Molecular Dynamics).
1
compound 2 as an oil (8.2 mg, 70%). H NMR (CDCl3): d 7.76 (s,
1H), 7.71 (s, 1H), 7.35–7.17 (m, 5H), 6.36 (q, J = 7.1 Hz, 1H),
4.26–4.16 (m, 2H), 1.86 (d, J = 7.1 Hz, 4H), 1.68–1.65 (m, 3H)
1.42–1.38 (m, 3H), 0.96–0.93 (m, 3H). 13C NMR (CDCl3): d 155.8,
138.7, 130.0, 129.0, 128.1, 126.4, 125.0, 104.9, 84.1, 70.9, 69.3,
63.7, 62.2, 59.7, 55.5, 20.0. 19F NMR d ꢀ223.2 (m, F). LC-MS (ESI1),
m/z 351 [M1H]1.
General biology
Metabolite study
Male Sprague–Dawley rats were used. All animals were handled
according to the guidelines set forth by the Swedish Animal
Welfare Agency, and the experiments were approved by the
local Ethics Committee for Animal Research, permit no: C234/5.
Tissues from the adrenals of human subjects were taken during
surgery at Uppsala University Hospital in accordance with
permission obtained from the local ethical committee. Adrenals
from pigs were obtained from the local slaughterhouse. All
tissues were frozen immediately after removal and stored at
ꢀ701C until cryosectioning.
Data analysis: Calculations were performed using Microsoft
Excel (Microsoft, USA) and Prism v.5.01 (GraphPad Software,
USA). The images were prepared using ImageQuant 5.1
(Molecular Dynamics, USA) and Microsoft Excel and PowerPoint
(Microsoft).
The metabolite analysis in plasma was performed in rats at the
5- and 30-min time points. At each time point, one male rat was
injected via the tail vein with 50 MBq radioligand. The animals
were sacrificed after 5 or 30 min in a high concentration of CO2.
The plasma was obtained by a 4-min centrifugation of the
venous blood at 4000 rpm at 41C. Plasma proteins were
precipitated by adding acetonitrile (600 mL) that contained
unlabelled reference to the plasma (600 mL). The resulting
mixture was centrifuged at 4000 rpm at 41C for 2 min. The
supernatant was filtered through a 0.2-mm pore size membrane
(Corning Incorporated, Corning, NY, USA) under centrifugation
at 4000 rpm at 41C for 2 min. The sample was analysed by HPLC
to separate the radioligand from its radioactive metabolites. This
was performed by injecting the sample (99 mL) on the same
column used for semi-preparative LC. The tracer and its
metabolite were eluted with C:D (1: 50:50 vol/vol and 2:
45:55 vol/vol) at a flow rate of 5 mL/min. The eluent was
collected in two fractions. Both tracers eluted in the second
fraction (1: tR = ꢀ2.20 and 5.29 min and 2: tR = 1.90 and 5.08 min),
whereas their metabolites eluted in the first fraction. Observa-
tion of the UV peak from the unlabelled standard indicated that
1 and 2 eluted in the second fraction. Even if no detectable
radiopeak co-eluted with the unlabelled standard, the fraction
was collected and the radioactivity of the fraction was measured
in a g-counter.
Organ distribution
In order to evaluate uptake in normal tissues, biodistribution
studies were performed. The rats were injected intravenously
with 10 MBq/animal (1: 281711, range 260–295 g (n = 12; n = 4
per time point); 2: 318712, range 300–340 g (n = 12; n = 4 per
time point)) of the two tracers, respectively. The animals were
sacrificed in a high concentration of CO2 after 5, 10 and 30 min,
and selected organs were dissected out. Blood, heart, lung,
liver, pancreas, spleen, adrenal, kidney, intestine, urinary
www.jlcr.org
Copyright r 2009 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2009, 52 278–285

M. Erlandsson et al.
˚
[4] M. Bergstrom, C. Juhlin, T. A. Bonasera, A. Sundin, B. Langstrom,
¨
J. Nucl. Med. 2000, 41, 275–282.
¨
Conclusion
One-step 18F-labelling method for the synthesis of 1 and 2 is
presented. These compounds, which are ethoxy-modified at the
ester part of the methyl 1-[(1R)-1-phenylethyl]-1H-imidazole-5-
carboxylate, are less lipophilic than the alkyl-substituted
metomidate tracer, and the biological evaluation indicates that
both compounds are likely candidates for the detection of
different types of adrenal carcinomas.
The d.c. radiochemical yields of 1 and 2 were 2678 and
2378%, respectively, after a 50–60-min synthesis with conven-
tional heating. The reaction became faster and higher yielding
when microwave heating was employed. Specifically, the d.c.
radiochemical yield for analogues 1 and 2 was increased to
57712 and 51718%, respectively.
˚
¨
¨
[5] T. S. Khan, A. Sundin, C. Juhlin, B. Langstrom, M. Bergstrom,
B. Eriksson, Eur. J. Nucl. Med. 2003, 30, 403–410.
[6] G. Zettinig, M. Mitterhauser, W. Wadsak, A. Becherer, C. Pirich,
H. Vierhapper, B. Niederle, R. Dudczak, K. Kletter, Eur. J. Nucl. Med.
2004, 31, 1224–1230.
˚
[7] H. Minn, A. Salonen, J. Friberg, A. Roivainen, T. Viljanen, J. Langsjo,
J. Salmi, M. Valimaki, K. Nagren, P. Nuutila, J. Nucl. Med. 2004, 45,
¨
˚
¨
¨
972–979.
¨
˚
¨
[8] F. Karimi, M. Erlandsson, O. Lindhe, B. Langstrom, J. Labelled
Compd. Radiopharm. 2008, 51, 273–276. DOI: 10.1002/jlcr.1517.
[9] W. Wadsak, M. Mitterhauser, J. Labelled Compd. Radiopharm. 2003,
46, 379–388. DOI: 10.1002/jlcr.680.
[10] W. Wadsak, M. Mitterhauser, G. Rendl, M. Schuetz, L. K. Mien,
D. E. Ettlinger, R. Dudczak, K. Kletter, G. Karanikas, Eur. J. Nucl. Med.
Mol. Imaging 2006, 33, 669–672.
[11] N. Elander, J. R. Jones, S. Y. Lu, S. Stone-Elander, Chem. Soc. Rev.
2000, 29, 239–249.
[12] I. M. Zolle, M. L. Berger, F. Hammerschmidt, S. Hahner, A. Schirbel,
B. Peric-Simov, J. Med. Chem. 2008, 51, 2244–2253.
[13] www.acdlabs.com.
According to the literature,23 microwave reactors have been
integrated into fully automated radiofluorination modules. The
one-step microwave heated 18F-labelling synthesis proposed in
this article could also be fully automated by integrating a
microwave reactor into, for example, the commercial synthesizer
¨
[14] J. Tierney, P. Lindstrom, Microwave Assisted Organic Synthesis,
Blackwell Publishing Ltd, Oxford, 2005.
[15] J. R. Jones, S. Y. Lu, Microwave-enhanced radiochemistry, in
Microwaves in Organic Synthesis (Ed.: A. Loupy), 1st ed., Wiley-VCH
Verlag GmbH & Co., Weinheim, 2002, pp. 456–459.
[16] S. Stone-Elander, N. Elander, J. Labelled Compd. Radiopharm. 2002,
45, 715–746.
TRACERlab FX
.
FN
Acknowledgement
[17] N. Leadbeater, H. Torenius, J. Org. Chem. 2002, 67, 3145–3148.
[18] www.personalchemistry.com.
[19] D. E. Ettlinger, W. Wadsak, L. K. Mien, M. Machek, L. Wabnegger,
G. Rendl, G. Karanikas, H. Viernstein, K. Kletter, R. Dudczak,
M. Mitterhauser, Eur. J. Nucl. Med. Mol. Imaging 2006, 33,
928–931.
This work was conducted in collaboration with Uppsala Imanet,
¨
GE Healthcare. We are grateful to Elisabeth Bergstrom-Petter-
man for her assistance in the biological study.
[20] P. Bjurling, R. Reineck, G. Westerberg, A. D. Gee, J. Sutcliffe,
References
˚
B. Langstrom, Proceedings of the VIth Workshop on Targetry and
Target Chemistry, Vancouver, Canada, 1995, pp. 282–284.
¨
´
´
´
[1] I. Varga, K. Racz, R. Kiss, L. Futo, M. Toth, O. Sergev, E. Glaz, Steroids [21] E. F. Godefroi, P. A. Janssen, C. A. Van Der Eycken, A. Van Heertum,
¨ ¨
1993, 58, 64–68. C. Niemegeers, J. Med. Chem. 1965, 6, 220–223.
[2] H. Vanden Bossche, G. Willemsens, W. Cools, D. Bellens, Biochem. [22] N. A. Gomzina, D. A. Vasil’ev, R. N. Krasikova, Radiochemistry 2002,
Pharmacol. 1984, 33, 3861–3868. 44, 366–409.
¨
¨
´
[3] M. Bergstrom, T. A. Bonasera, L. Lu, E. Bergstrom, C. Backlin,
[23] N. Lazarova, F. G. Simeon, J. L. Musachio, S. Lu, V. W. Pike,
J. Labelled Compd. Radiopharm. 2007, 50, 463–465.
˚
C. Juhlin, B. Langstrom, J. Nucl. Med. 1998, 39, 982–989.
¨
J. Label Compd. Radiopharm 2009, 52 278–285
Copyright r 2009 John Wiley & Sons, Ltd.
www.jlcr.org