Synthesis of 3-[[4-(4-[18F] fluorophenyl)piperazin-1-yl]methyl]-1H-pyrrolo[2,3-b]pyridine

Article abstract of DOI:10.1002/jlcr.597

3-[[4-(4-[¹⁸F]fluorophenyl)piperazin-1-yl]methyl] -1H-pyrrolo[2,3-b]pyridine, a candidate to image dopamine D₄ receptors, was synthesised via electrophilic fluorination of a trimethylstannyl precursor with high specific radioactivity

Full text of DOI:10.1002/jlcr.597

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2002; 45: 687–696.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.597
Research Article
Synthesis of 3-[[4-(4-[18F]
fluorophenyl)piperazin-1-yl]methyl]-1H-
pyrrolo[2,3-b]pyridine
,1
1,2
1
.
Olli Eskola* , Jorgen Bergman , Pertti Lehikoinen , Merja
1,3
1,3
1,3
Haaparanta , Tove Gronroos , Sarita Forsback and Olof Solin1,2
1 Turku PET Centre, Radiopharmaceutical Chemistry Laboratory,
Porthaninkatu 3, FIN 20500 Turku, Finland
.
2 Accelerator Laboratory, Porthaninkatu 3, FIN 20500 Turku, Finland
3
.
Medicity/PET, Tykistokatu 6 A, FIN 20520, Turku, Finland
Summary
3-[[4-(4-[18F]fluorophenyl)piperazin-1-yl]methyl]-1H-pyrrolo[2,3-b]pyridine, a
candidate to image dopamine D4 receptors, was synthesised via electrophilic
fluorination of a trimethylstannyl precursor with high specific radioactivity
[18F]F2. The precursor was obtained by a facile four-step synthetic approach;
the trimethylstannyl leaving group was introduced by displacement of iodine
utilising palladium catalysis and hexamethyldistannane in an inert solvent. The
total radiosynthesis time was 50 min, including purification and formulation
for injection. Decay corrected radiochemical yield was 51% as calculated
from the amount of [18F]Fꢀ produced. Specific radioactivity at the end of
synthesis was 12.8–16.4 GBq/mmol. Radiochemical purity was 88–92%. Ex vivo
studies in rats showed homogeneous distribution of radioactivity within rat
brain. Copyright # 2002 John Wiley & Sons, Ltd.
Key Words: 18F; schizophrenia; dopamine D4 receptor; stannylation; electro-
philic fluorination
*Correspondence to: O. Eskola, Turku PET Centre, Radiopharmaceutical Chemistry Laboratory,
Porthaninkatu 3, FIN 20500 Turku, Finland. E-mail: olesko@utu.fi
Copyright # 2002 John Wiley & Sons, Ltd.
Received 17 September 2001
Revised 12 March 2002
Accepted 12 March 2002

688
O. ESKOLA ET AL.
Introduction
Schizophrenia is a complex psychotic disorder whose origin and
etiology as yet are not fully understood. An association between the
dopamine neuronal hyperactivity and schizophrenia has been studied
intensively over the past few years. An increased concentration of
dopamine D4 receptors has been demonstrated in the post-mortem
brains of schizophrenic patients,1 even though these findings are
controversial.2 It is evident that selective ligands with a high affinity
to dopamine D4 receptors have to be developed in order to measure the
dopamine D4 receptor density in brain and to highlight their
pharmacological role in schizophrenia.
Several ligands of different chemical classes have been reported to
bind selectively and with high affinity to the dopamine D4 receptor.
A few of these ligands contain arylpiperazine3–14 or azaindole
moieties.15–17 L-745,870 (Figure 1) is a structural combination of the
L-745,870,
L-750,667,
[
X = Cl
X = I
N
N
X
123I]L-750,667, X = 123
I
N
N
H
Figure 1. L-745, 870 and L-750, 667: Selective dopamine D4 receptor antag-
onists
1H-pyrrolo[2,3-b]pyridine (7-azaindole) unit and 4-chlorophenylpiper-
azine unit linked together with a methylene spacer.18 L-745,870 is a
selective (>2000-fold as compared to other dopamine receptor
subtypes) antagonist with high affinity to the D4 receptor
(Ki=0.43 nM). The 4-iodophenyl compound L-750,66719 (Figure 1)
has similar selectivity but slightly lower affinity (Ki=0.51 nM) than L-
745,870. L-750,667 has also been radioiodinated with 123I.20 As a
candidate to image dopamine D4 receptors, we have synthesised
molecule 7, 3-[[4-(4-[18F]fluorophenyl)piperazin-1-yl]methyl]-1H-pyrro-
lo[2,3-b]pyridine ([18F]F5P) (Figure 2), a fluorinated analogue of the
molecules L-745,870 and L-750,667. Since molecule 7 contains an 18F
isotope, a positron emitting nucleus, it can be used as a radiotracer with
positron emission tomography (PET).
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

SYNTHESIS OF [18F]F5P
689
N(CH3)2
i
N
N
H
N
N
N
N
I
H
iii
1
3
N
N
H
ii
HN
N
HN
N
I
5
2
4
iv
N
N
18F
N
N
Sn(CH3)3
v
N
N
H
N
N
H
7
6
Figure 2. Synthesis of the stannylated precursor 6 and the radioligand [18F]F5P
7: (i) dimethylammoniumchloride, paraformaldehyde, butanol, D; (ii) ICl, AcOH/
H2O (3/1), 608C; (iii) xylene, D; (iv) Sn2(CH3)6, (PPh3)4Pd(0), dioxane, D; and
(v) CCl3F, AcOH, [18F]F2.
Results and discussion
Precursor synthesis
Figure 2 illustrates the synthesis of precursor 6 containing a
trimethylstannyl functional group, a good leaving group that can be
displaced with [18F]F2. The precursor synthesis started with the
preparations of the building blocks 3 and 4. 3-N,N-dimethylamino-
methyl-1H-pyrrolo[2,3-b]pyridine 3 was prepared from commercially
available 7-azaindole 1 according to a literature method.21 Successive
crystallisations from diethylether afforded compound 3 (mp 1548C).
The other building block, 1-(4-iodophenyl)piperazine 4, was synthesised
by direct iodination of phenylpiperazine 2 with iodine monochloride, as
described by Hanson et al.22 With this method it was possible to
introduce the iodine in a fairly selective manner to the para-position of
the phenyl ring using a mild reaction medium and a short reaction time.
Furthermore, starting materials and side products were easily removed
by several recrystallisations from ethanol; no signals of ortho- or meta-
1
iodinated side products were detected from H NMR spectrum of the
purified compound 4 (mp 1218C).
Synthesis of the iodo-compound 5, the coupling reaction of the
building blocks 3 and 4, was performed according to the method
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

690
O. ESKOLA ET AL.
originally described by Kulagowski et al.23 The reaction was tested in
dry toluene and dry xylene at reflux. Of these two, dry xylene turned out
to be a more suitable solvent for the reaction, resulting in 56% yield in
comparison to the ꢁ20% yield in dry toluene. Compound 5 (mp 1798C,
accurate mass with HR-EI-MS: calc. for C18H19IN4 418.065449, found
418.065600) was purified by successive crystallisations from absolute
ethanol. The chemical purity of 5 exceeded 97%, as determined with
analytical HPLC (see experimental section). Recently, Staley et al.
reported a feasible synthesis of compound 5 by Mannich reaction
between 7-azaindole 1 and 4-iodophenylpiperazine 4.20 The yield of
compound 5 synthesised with their method was 92%.
Organostannane 6, the precursor for labelling, was synthesised from
the iodo-derivative 5 utilising palladium catalysis and hexa-alkyldis-
tannane, a method widely used for the preparation of functionally
substituted organotin compounds.24 After a 3 h reflux under an inert
argon atmosphere the reaction usually proceeded to completion, as
monitored by TLC, and compound 6 was purified by preparative
column chromatography. However, if unreacted iodo-derivative 5 was
present in the crude product, the separation of 5 and 6 with preparative
column chromatography was not satisfactory. In these cases, the
product was further purified with reversed-phase HPLC [mBp C18
column (7.8 ꢂ 300 mm), MeOH/H2O 8/2]. The traces of compound 5
were thus separated and the chemical purity of 6, checked with
analytical HPLC, exceeded 98%. Furthermore, no signals (m/z 418) of
the iodine derivative 5 were detected from the mass spectrum of
compound 6.
The reference compound 9 was prepared from 7-azagramine 3 and
commercially available 4-fluorophenylpiperazine. The synthesis resem-
bled that of the iodo-compound 5 resulting in a 51% yield. The chemical
purity of 9, determined with analytical HPLC, exceeded 99.5%.
Radiolabelling
Radioligand 7, [18F]F5P, was synthesised by electrophilic fluorodestan-
nylation of 6 (see Figure 2). The average synthesis time was 50 min. The
electrophilic labelling agent [18F]F2, produced from [18F]Fꢀ according
to literature methods,25,26 was bubbled at room temperature through a
CCl3F/CH3COOH mixture containing precursor 6. The major chemical
impurity that was formed using the method in question was found to be
the phenylpiperazine analogue 8 (Figure 3). Reversed-phase HPLC was
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

SYNTHESIS OF [18F]F5P
691
N
N
N
N
F
8
9
N
N
N
N
H
H
Figure 3. Structural formulae of compounds 8, the major chemical impurity
generated in labelling synthesis, and 9, the reference compound
used to separate this and other side products from compound 7. The
radiofluorination proceeded at relatively low radiochemical yield, in
average 0.7 ꢃ 0.1% (decay corrected to the end of bombardment) as
calculated from the amount of [18F]Fꢀ produced. This was due to the
formation of several radiofluorinated side products, more polar (shorter
retention time in the reversed-phase HPLC column) than compound 7.
Most likely these were a result of fragmentation, a phenomenon
common with highly reactive F2. Furthermore, the pyrrolic nitrogen of
precursor 6 was unprotected, which undoubtedly increased the amount
of radiolabelled side products. Since the absolute yield of 7, in average
183 ꢃ 32 MBq at the end of synthesis, was high enough to perform
the preclinical animal studies, no attempts were made to optimise the
labelling conditions.
HPLC analysis showed the presence of an unidentified 18F-labelled
contaminant in the formulated final solution of 7. This contaminant,
eluting as a bulky broad peak from the semi-preparative HPLC column
before compound 7, decreased the radiochemical purity of 7, which was
in average 90.3 ꢃ 1.7%. The chemical purity, however, exceeded 95%,
with respect to the phenylpiperazine analogue 8. The specific
radioactivity of 7 (at the end of synthesis) was in average
14.6 ꢃ 1.8 GBq/mmol.
Animal studies
In peripheral tissues in mice a high level of expression of the D4 mRNA
has been detected in the eye, adrenal gland and testis.27 In our
biodistribution studies in peripheral organs compound 7 showed high
uptake (%ID/g) in adrenal gland (7.2), pancreas (2.0), lung (2.0) and
liver (1.1) as well as in urine (5.2) at 5 min post-injection. The uptake in
liver tended to increase as a function of time.
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

692
O. ESKOLA ET AL.
In rat brain the dopamine D4 receptors are mainly localised in
entorhinal cortex, hippocampus and hypothalamus.28 As measured with
digital autoradiography, the radioactivity uptake in brain at 5 min was
about 1%ID/g in cortex, hippocampus, thalamus and striatum as well
as in various other brain regions, including cerebellum. At 2 h post-
injection the radioactivity was uniformly distributed in whole brain.
Based on these results it was concluded the 7 is not a suitable
compound for visualising the D4 receptor in vivo in rat. Even though the
compound showed relatively high uptake in brain, the radioactivity
distribution in brain was uniform and not markedly concentrated into
known dopamine D4-rich areas. Recently, Staley et al. reported the
synthesis of [123I]L-750,667, the radioiodinated analogue of compound
7.20 Their conclusion was that [123I]L-750,667 was not suited as a
radiotracer for the in vivo imaging of the D4 receptor but, however, the
ligand is selective for D4 receptor in vitro. The specific radioactivity of
tracer 7 was 12.8–16.4 GBq/mmol. However, the D4 receptor density in
rodent brain is low, in the order of 30 fmol/mg protein.28 Thus the
possibility that the amount of tracer mass (ꢁ1 mg/kg per rat) injected in
this study saturated the receptor binding sites cannot be excluded.
The metabolism of the tracer was not studied, although the increasing
uptake in liver as a function of time does indicate metabolic degradation
of the tracer. Unchanged uptake in bone as a function of time indicated
that the molecule 7, or its metabolites were not defluorinated to any
larger extent.
The number of animals in this study was kept small. This was
warranted as early results demonstrated that the tracer was clearly less
than optimal as a dopamine D4 tracer.
Experimental
General methods
All the chemicals used in this study were bought, unless otherwise noted,
from commercial suppliers. NMR spectra were recorded on JEOL
JNM-GX400 spectrometer with chemical shifts reported in d values
(parts per million) relative to tetramethylsilane (d=0). Melting points
(mp) were detected with Mettler FP80 apparatus equipped with a
Mettler FP81 MBC cell and they are uncorrected. High-resolution mass
spectrometry was done with VG ZabSpec spectrometer operated in
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

SYNTHESIS OF [18F]F5P
693
electron ionisation (EI) mode. Merck silica gel 60 (230–400 mesh
ASTM) was used for preparative column chromatography separations.
Thin layer chromatography was done with Merck TLC aluminium
sheets (silica gel 60 F254). A Merck Hitachi L-6200 pump was used for
semi-preparative HPLC. The semi-preparative column was Waters
mBondapak C18 HPLC column (7.8 ꢂ 300 mm, 10 mm) connected in
series with a Merck Hitachi L-4000A UV absorption detector
(l=280 nm) and a 2 ꢂ 200 sodium iodide crystal for radioactivity
detection. The chromatograms were registered with a Goerz Metra
Watt SE 120 two-channel recorder. The column was eluted isocratically
with 0.1 M ammonium formate solution for 2 min with a flow rate of
4 ml/min and then isocratically with a mobile phase of 0.1 M
ammonium formate/MeOH (45:55) maintaining the same flow rate.
A Merck Hitachi L-6200 pump was used for analytical HPLC. A
Waters mBondapak C18 HPLC column (3.9 ꢂ 300 mm, 10 mm) was
eluted isocratically (flow rate 1.5 ml/min) using 0.1 M ammonium
formate/MeOH (40:60) as a mobile phase. The column was connected in
series with a Shimazu (model SPD-M6A) UV/VIS photodiode array
detector (l=210–250 nm), 2 ꢂ 200 NaI crystal radioactivity detector and
Waters 740 integrator. This analytical HPLC set-up (l=254 nm) was
also used in determination of the chemical purities of the compounds 5
(Rt 12.3 min), 6 (Rt 33 min) and 9 (Rt 5.6 min). Radioactivity was
measured with a radioisotope calibrator (CRC-15R, Capintec Inc.,
Pittsburgh, USA).
Radiochemical yields were calculated based on the amount of 18F
produced at the end of bombardment. Determinations of chemical and
radiochemical purity and specific radioactivity were done by compar-
isons of HPLC retention times and peak intensities with the reference
compound 9 (Figure 3) of known concentration.
Chemistry
Compounds 3,21 422 and 523 were prepared as described in the literature.
3-[[4-(4-trimethylstannylphenyl)piperazin-1-yl]methyl]-1H-pyrrolo[2,3-b]
pyridine 6. The iodo-derivative 5 (1.0 g, 2.39 mmol) was refluxed with
hexamethylditin (1.3 g, 4.0 mmol) and tetrakistriphenylphosphinepalla-
dium(0) in dry dioxane (40 ml) under argon atmosphere for 3 h. The
mixture was allowed to cool down and the solvent was removed by a
rotary evaporator. The crude product was purified by column
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

694
O. ESKOLA ET AL.
chromatography (CH2Cl2:EtOAc:Et3N 40:60:10; Rf 0.12). This afforded
1
the stannyl derivative 6 (620 mg, 57%) as a white solid (mp 1808C). H
NMR (CDCl3) 10.8 (1 H; s), 8.33 (1 H; d; J=6 Hz), 8.11 (1 H; d;
J=9 Hz), 7.36 (2 H; d; J=9 Hz), 7.33 (1 H; s), 7.10 (1 H; dd; J=4.5 and
12.5 Hz), 6.90 (2 H; d; J=9 Hz), 3.77 (2 H; s), 3.21 (4 H; m; piperazinyl
CH2), 2.66 (4 H; m; piperazinyl CH2), 0.28 (9 H; s; Sn(CH3)3). Accurate
mass with HR-EI-MS: calculated for C21H28N4Sn 454.133004, found
454.133800.
3-[[4-(4-[18F]fluorophenyl)piperazin-1-yl]methyl]-1H-pyrrolo[2,3-b]
pyridine 7. The stannylated precursor 6 was radiolabelled with
electrophilic [18F]F2 gas according to methods discussed in detail in
the literature.25,26 Three hundred micrograms (0.66 mmol) of 6 was
dissolved in a solution containing freon-11 and dry acetic acid. [18F]F2
gas (specific radioactivity ꢁ55 GBq/mmol at EOB) was bubbled through
the reaction mixture. Freon-11 was evaporated and the residue was
dissolved in 0.1 M HCO2NH4 solution, which was injected on the semi-
preparative HPLC column. The fraction containing compound 7,
eluting at ꢄ18 min, was collected and the radioactivity was measured.
This fraction was then evaporated to dryness and the dry residue was
dissolved in 0.9% NaCl solution (pH 4.7). This solution was used for
the in vivo studies with rats.
3-[[4-(4-fluorophenyl)piperazin-1-yl]methyl]-1H-pyrrolo[2,3-b]pyridine
9. A solution of 3 (0.5 g, 2.8 mmol) and 1-(4-fluorophenyl)piperazine
(0.51 g, 2.8 mmol) was refluxed in dry xylene in argon atmosphere for
7 h. The reaction mixture was allowed to cool down and the solids were
separated by filtration and washed with ethanol. Three recrystallisations
from ethanol afforded compound 9 (440 mg, 51%) as a light white
1
powder (mp 2118C). H NMR (DMSO) 8.18 (1 H; d; J=6 Hz), 8.03
(1 H; d; J=9 Hz), 7.38 (1 H; s), 6.98–7.05 (3 H; m), 6.90 (2 H; m), 3.66
(2 H; s), 3.03 (4 H; m; piperazinyl CH2), 2.49 (4 H; m; piperazinyl CH2).
Accurate mass with HR-EI-MS: calculated for C18H19FN4 310.159375,
found 310.159600.
Animal studies
The biodistribution of compound 7 was determined in male Sprague-
Dawley (Hsd:SD) rats (Harlan Sprague–Dawley, IN, USA). About
22 MBq (about 0.5 mg) of compound 7 was injected intravenously
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

SYNTHESIS OF [18F]F5P
695
through the tail vein. After specific times (5, 15, 30 and 120 min) the rats
were sacrificed in a CO2 chamber. Blood, urine and organs to be
examined were rapidly removed, measured for radioactivity and
weighed. The results are expressed as % injected dose per gram tissue
(%ID/g). The brains were frozen and sliced with a cryomicrotome.
Coronal sections (20 mm) were cut and thaw-mounted on microscopic
slides. Brain sections were placed in an exposure cassette with imaging
plate (BAS-TR2025, Fuji Photo Film Co., Japan) and exposed for 4 h.
The spatial distribution of radioactivity in the brain sections was
recorded with a phosphoimaging device (Fujifilm BAS-5000, Fuji Photo
Film Co Ltd, Japan). Brain sections were then stained with
haematoxylin/eosin and the brain regions were identified from stained
sections using a rat brain atlas.29
Conclusion
[18F]F5P 7 was synthesised from trimethylstannyl precursor 6 by
electrophilic destannylation with high specific radioactivity [18F]F2.
Radiochemical yields of 7 were low and the specific radioactivity was
moderately high. Ex vivo biodistribution studies showed homogeneous
distribution of radioactivity within rat brain. Thus, we conclude that 7 is
probably not a suitable radiotracer to image dopamine D4 receptors.
References
1. Seeman P, Guan H-C, Van Tol H. Nature 1993; 365: 441–445.
2. Reynolds GP, Mason SL. Eur J Pharmacol 1995; 281: R5–R6.
3. Perrone R, Berardi F, Colabufo NA, Leopoldo M, Tortorella V. J Med
Chem 2000; 43: 270–277.
.
4. Haubmann C, Hubner H, Gmeiner P. Bioorg Med Chem Lett 1999; 9:
1969–1972.
5. Sanner MA, Chappie TA, Dunaiskis AR, et al. Bioorg Med Chem Lett
1998; 8: 725–730.
.
6. Arlt M, Bottcher H, Rietmuller A, et al. Bioorg Med Chem Lett 1998; 8:
.
2033–2038.
7. He XS, Woodruff K, Brodbeck R. Bioorg Med Chem Lett 1997; 7:
2399–2402.
8. Thurkauf A, Yuan J, Chen X, et al. J Med Chem 1997; 40: 1–3.
9. Glase SA, Akunne HC, Georgic LM, et al. J Med Chem 1997; 40:
1771–1772.
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696

696
O. ESKOLA ET AL.
10. Perrone R, Berardi F, Colabufo NA, Leopoldo M, Tortorella V. Bioorg
Med Chem Lett 1997; 7: 1327–1330.
11. TenBrink RE, Bergh CL, Duncan JN, et al. J Med Chem 1996; 39:
2435–2437.
12. Boyfield I, Coldwell MC, Hadley MS, et al. Bioorg Med Chem Lett 1996; 6:
1227–1232.
13. Thurkauf A, Yuan J, Chen X, et al. J Med Chem 1995; 38: 4950–4952.
14. Perrone R, Berardi F, Colabufo NA, Leopoldo M, Tortorella V. J Med
Chem 1998; 41: 4903–4909.
.
.
15. Lober S, Hubner H, Gmeiner P. Bioorg Med Chem Lett 1999; 9: 97–102.
16. Curtis NR, Kulakowski JJ, Leeson PD, et al. Bioorg Med Chem Lett 1999;
9: 585–588.
.
17. Prante O, Lober S, Hubner H, Gmeiner P, Kuwert T. J Labelled Compd
.
Radiopharm 2001; 44: 849–858.
18. Patel S, Freedman S, Chapman KL, et al. J Pharmacol Exp Ther 1997; 283:
636–647.
19. Patel S, Patel S, Marwood, R, et al. Mol Pharmacol 1996; 50: 1658–1664.
20. Staley JK, Tamagnan G, Baldwin RM, et al. Nucl Med Biol 2000; 27:
547–556.
21. Williamson WRN. J Chem Soc Part II: 1962; 2833.
22. Hanson RN, Hariharan S, Astik R. J Heterocycl Chem 1985; 22: 47–48.
23. Kulagowski JJ, Broughton HB, Curtis NR, et al. J Med Chem 1996; 39:
1941–1942.
24. Ali H, van Lier JE. Synthesis 1996; 4: 423–445.
25. Bergman J, Lehikoinen P, Solin O. J Labelled Compd Radiopharm 1997;
40: 38–39.
26. Bergman J, Solin O. Nucl Med Biol 1997; 24: 677–683.
27. Suzuki T, Kobayashi K, Nagatsu T, Neurosci Lett 1995; 199: 69–72.
28. Primus RJ, Thurkauf A, Xu J, et al. J Pharmacol Exp Ther 1997; 282:
1020–1027.
29. Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. Academic
Press, Inc.: San Diego, CA, USA, 1986.
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 687–696