5-(2-[18F]fluoroethyl)-4-methylthiazole probe for positron emission tomography of the central nervous system

Full text of DOI:10.1007/s10593-014-1477-4

Chemistry of Heterocyclic Compounds, Vol. 50, No. 2, May, 2014 (Russian Original Vol. 50, No. 2, February, 2014)
5-(2-[18F]FLUOROETHYL)-4-METHYLTHIAZOLE PROBE
FOR POSITRON EMISSION TOMOGRAPHY OF THE
CENTRAL NERVOUS SYSTEM
N. M. Evdokimov1,2*, G. Flores2,4, P. M. Clark3, M. E. Phelps2,4,
1
O. N. Witte2,3,5,6, and M. E. Jung
*
Keywords: 5-(2-[18F]fluoroethyl)-4-methylthiazole, central nervous system, fluorination, imaging,
neuroprotection, positron emission tomography, radiolabeling.
Positron emission tomography (PET) is a non-invasive diagnostic method of high clinical utility [1].
Current clinical PET scanning of brain functions rely heavily on two tracers, 2-deoxy-2-[18F]fluoro-D-glucose
(18F-FDG) [2] and the more selective 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine (18F-FDOPA) [3]. Although
it is used to track fundamental sugar metabolism, the FDG probe sometimes lacks selectivity, especially for the
imaging of high-grade brain tumors after treatment [4]. Better resolution can be achieved with radiolabeled
amino acids [5]. Unfortunately, imaging of amino acid metabolism with FDOPA offers indirect conclusions
based on the low probe uptake in Parkinson's and Alzheimer's diseases [6]. The emerging interest in in vivo real
time imaging of memory functioning and age-caused neurodegeneration has been limited due to the small
number of existing selective PET tracers. Accordingly, development of new probes that target alternative brain
receptors could dramatically advance our understanding of the specific processes occurring within the brain and
would potentially offer an alternative avenue in clinical patient diagnosis.
_______
*To whom correspondence should be addressed, e-mail: nevdokim@chem.ucla.edu, jung@chem.ucla.edu.
1University of California, Los Angeles, Department of Chemistry and Biochemistry, 607 Charles E. Young
Drive East, Box 951569, Los Angeles CA 90095-1569, U.S.A.
2University of California, Los Angeles, Department of Molecular and Medical Pharmacology, 650 Charles
E. Young Dr. South; 23-120 Center for Health Sciences, Los Angeles CA 90095-1735, U.S.A.; e-mail:
owenwitte@mednet.ucla.edu, MPhelps@mednet.ucla.edu, GFlores@mednet.ucla.edu.
3University of Calfornia, Los Angeles, Department of Microbiology, Immunology, and Molecular Genetics, 609
Charles E. Young Drive East, Los Angeles CA 90095-1489, U.S.A.; e-mail: PClark@mednet.ucla.edu.
4University of California, Los Angeles, Crump Institute for Molecular Imaging, 570 Westwood Plaza,
Los Angeles CA 90095, U.S.A.
5Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research, University of California,
Los Angeles, Los Angeles CA 90095, U.S.A.
6Howard Hughes Medical Institute, David Geffen School of Medicine, University of California, Los Angeles,
Los Angeles CA 9009, U.S.A.
_________________________________________________________________________________________
Published in Khimiya Geterotsiklicheskikh Soedinenii, No. 2, pp. 330-334, February, 2014. Original
article submitted January 16, 2014.
0009-3122/14/5002-0303©2014 Springer Science+Business Media New York
303

Neuronal excitotoxicity is one of the major mechanisms hypothesized to be involved in the development
of neurodegenerative conditions such as Alzheimer's disease [7]. It is characterized by the overstimulation of
excitatory glutamate receptors leading to an increased concentration of calcium ions, which trigger several cell-
damaging processes, resulting in dysfunction and the death of neurons [8]. Since they display neuroprotection
during Alzheimer's disease [9], GABA receptor agonists based on a benzodiazepine core are of particular
interest for imaging the adverse effects of neurodegeneration [10]. Selective agonists of GABA receptors have
been also shown to exhibit neuroprotective properties in ischemic stroke [11]. In particular, clomethiazole 1
(CMZ, Heminevrin) is a GABAA receptor agonist and a prescription sedative and hypnotic drug with
pronounced neuroprotective function [12]. However, it was shown to be ineffective in improving the outcome of
patients with major ischemic stroke [13].
We envisioned that substitution of the terminal chlorine atom with a radioactive fluorine atom would
afford an analog of CMZ, 5-(2-[18F]fluoroethyl)-4-methylthiazole (18F-FMT) ([18F]2), which would bind in a
similar fashion to a site on the GABA receptor. The non-radioactive 5-(2-fluoroethyl)-4-methylthiazole (FMT) (2),
which is readily available upon reaction of sulfurol (3) with DAST, was previously shown to have bacteriostatic
activity against E. coli [14]. It was anticipated that use of the 18F-labeled FMT probe would reveal regions of the
central nervous system (CNS) with an overexpression of GABA receptors and would likely image neuronal
inhibition in vivo.
18
The synthetic challenges in radiolabeling with the F isotope arise from its short half life – 110 min.
Therefore, the radiosynthesis of compound [18F]2 was planned so that the radiolabel would be introduced by a
fast SN2 reaction. Thus the reaction of the commercially available alcohol 3 with tosyl chloride produced
tosylate 4 [15], which in reaction with 18F-labelled potassium fluoride afforded the 18F-probe of compound
[18F]2 in one step with both high radio-corrected yield (RCY) and purity.
18
The F-FMT probe ([18F]2) thus obtained was subjected to quality control and used for PET scanning of
wild type mice. These mice were devoid of tumors or xenografts, and the results of these experiments show the
normal biodistribution of the probe. As early as 10 min after injection, strong radioactive accumulation in the liver
and nasal cavity was observed. Over the next two hours, the radioactivity continued to accumulate in the nasal cavity
and began to accumulate in the bladder and gall bladder (Fig. 1). This latter signal suggests hepatic and renal
clearance of the probe.
To determine whether the signal in the nasal cavity represents specific accumulation, we performed a
18
competition experiment between compound F-FMT ([18F]2) and non-radioactive FMT (2). Mice co-injected
with 6.25 mol of non-radioactive FMT and 18F-FMT showed significantly lower 18F-FMT accumulation in the
nasal cavity compared to mice co-injected with vehicle and 18F-FMT (Fig. 2). Alternatively, 18F-FMT
accumulation in the brain was unaffected by co-injection with non-radioactive FMT (Fig. 2). This suggests that
18F-FMT accumulates in the nasal cavity through a saturable accumulation mechanism.
304

In conclusion, attempting to selectively image the central nervous system, we have synthesized a new
18
PET probe based on a known thiazole heterocyclic drug clomethiazole. The probe F-FMT did not display a
strong and selective signal in the CNS, but we have observed that radioactivity accumulated in the nasal cavity.
Fig. 1. (a) Percent injected dose per gram (%ID/g) of
compound [18F]2 in several types of tissue.
(b) Representative images of radioactive label
distribution by a dynamic PET/CT scan of a mouse
treated with compound [18F]2. B – bladder, Br – brain,
GB – gall bladder, L – liver, NC – nasal cavity.
18
Fig. 2. F-FMT signal in the nasal cavity decreases in the presence of non-radioactive FMT. Quantification of
18
18F signal in the nasal cavity (b) and brain (c). Representative images of F signal in the nasal cavity of mice
injected with 18F-FMT and vehicle (a, left) or 18F-FMT and 6.25 µM non-radioactive FMT (a, right).
Although the probe was shown to be stable in bench stability experiments, a weak signal in the bones
suggested some formation of free fluoride in vivo. The experiment with non-radioactive FMT showed a diminished
accumulation of radioactivity within the nasal cavity, suggesting a competitive tissue-specific biodistribution of the
probe. The observed signal in the liver and bladder pointed to the hepatic and renal clearance of the probe or its
metabolites. Further work on this probe is under way.
Anhydrous solvents, K2CO3, and 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane (Kryptofix
222, K222) were purchased from Sigma Aldrich. Solid phase extraction cartridges, silica Sep-Pak WAT020520, and
Sep-Pak Plus Alumina N Cartridges WAT 020510 were purchased from Waters Associates. The silica gel cartridge
was preconditioned with 10 ml of anhydrous hexane, and the alumina cartridge was preconditioned with 10 ml of
water. The reference standard compound 2 was synthesized according to a published protocol [14].
Radiosynthesis of 5-(2-[18F]Fluoroethyl)-4-methylthiazole ([18F]2) and its Purification. [18F]KF
with no carrier added was produced in a RDS-111 cyclotron from [18O]H2O (98% isotopic purity, Rotem, Inc.)
on a silver target. An aliquot of aqueous [18F]KF (250 μl) was treated with a premixed solution of K2CO3 (1.5 mg)
in H2O (14 μl) and Kryptofix 222 (10 mg) in anhydrous MeCN (950 μl). The mixture was heated to 120°C
305

under dry nitrogen and dried repeatedly with 4×0.5 ml of anhydrous MeCN. In the next step, 16 mg of
2-(4-methylthiazol-5-yl)ethyl 4-methylbenzenesulfonate (4) dissolved in anhydrous MeCN (500 μl) was added
18
to the dried F-KF/K222 complex, and the reaction mixture was heated to 76°C for 15 min in a sealed vessel.
The cooled reaction mixture was loaded onto a silica gel cartridge and the fluorinated product was eluted with
anhydrous ethyl acetate (3 ml) into a second reaction vessel. The solvent was then evaporated at 80°C under a
flow of dry nitrogen. The resulting dried residue was dissolved in water (500 μl) and passed through an alumina
cartridge and eluted with sterile injectable water (4 ml) that was sterilized by filtration through a 0.22 μm filter.
The final product was formulated in 10% EtOH in H2O.
Analytical HPLC was performed with a WellChrom K-505 HPLC pump (Knauer, Berlin, Germany),
analytical phase Luna column (5 μm, 10×250 mm, Phenomenex), UV detector (254 nm, WellChrom Spectro-
Photometer K-505, Knauer), and gamma radiation detector and counter (B-FC-3300 and B-FC-1000, Bioscan,
Inc.). The mobile phase was 50% MeCN in H2O (flow rate 1 ml/min, retention time 3.5 min). All chromatograms
were collected using a GinaStar analog-to-digital converter (Raytest U.S.A., Inc.) and GinaStar software (Raytest
U.S.A., Inc.) running on a PC. The radiochemical purity of the probe was shown to be 99%.
The PET experiment. Approximately 100 Ci of 18F-FMT was injected intravenously into the tail vein
of anesthetized 8-12 week female C57Bl/6 mice. Each mouse was subject to a 2 h dynamic PET scan on a
microPET imager (Inveon, Siemens Medical Solutions U.S.A., Inc.). For the competition experiments,
18
approximately 20 Ci of F-FMT was co-injected intravenously into the tail vein with non-radioactive FMT
(6.25 mol) or vehicle (1.25% DMSO in 1× phosphate buffered saline). After 1 h, each mouse was subjected to
a 10 min static PET scan on a microPET imager (G4, Sofie Biosciences). Following the PET scans, each mouse
was subjected to a 10 min computed tomography (CT) scan on a microCT imager (MicroCAT, Imtek, Inc.). The
PET images were reconstructed and co-registered with the CT image [16]. PET/CT images were analyzed and
quantified using the AMIDE software [17]. Images are representative of n = 3-5 independent experiments.
N. M. E. was supported by The Phelps Family Foundation. P. M. C. was supported by the California
Institute of Regenerative Medicine (Training Grant TG2-01169) and by the UCLA Scholars in Molecular
Imaging Program (NCI R25T CA098010).
REFERENCES
1.
2.
3.
M. E. Phelps, Proc. Natl. Acad. Sci. U. S. A., 97, 9226 (2000).
D. H. S. Silverman, J. Nucl. Med., 45, 594 (2004).
W. Chen, D. H. S. Silverman, S. Delaloye, J. Czernin, N. Kamdar, W. Pope, N. Satyamurthy,
C. Schiepers, and T. Cloughesy, J. Nucl. Med., 47, 904 (2006).
M. V. Padma, S. Said, M. Jacobs, D. R. Hwang, K. Dunigan, M. Satter, B. Christian, J. Ruppert,
T. Bernstein, G. Kraus, and J. C. Mantil, J. Neurooncol., 64, 227 (2003).
F. Bénard, J. Romsa, and R. Hustinx, Semin. Nucl. Med., 33, 148 (2003).
D. J. Brooks, J. Nucl. Med., 51, 596 (2010).
R. Cacabelos, M. Takeda, and B. Winblad, Int. J. Geriatr. Psychiatry, 14, 3 (1999).
M. P. Mattson, Neuromol. Med., 3, 65 (2003).
P. R. Louzada, A. C. Paula Lima, D. L. Mendonça-Silva, F. Noël, F. G. De Mello, and S. T. Ferreira,
FASEB J., 18, 511 (2004).
4.
5.
6.
7.
8.
9.
10.
H. Boecker, A. Weindl, D. J. Brooks, A. O. Ceballos-Baumann, C. Liedtke, M. Miederer, T. Sprenger,
K. J. Wagner, and I. Miederer, J. Nucl. Med., 51, 1030 (2010).
11.
12.
A. R. Green, A. H. Hainsworth, and D. M. Jackson, Neuropharmacol., 39, 1483 (2000).
J. W. B. Marshall, A. J. Cross, D. M. Jackson, A. R. Green, H. F. Baker, and R. M. Ridley, Brain Res.
Bull., 52, 21 (2000).
306

13.
P. Lyden, A. Shuaib, K. Ng, K. Levin, R. P. Atkinson, A. Rajput, L. Wechsler, T. Ashwood,
L. Claesson, T. Odergren, and E. Salazar-Grueso, Stroke, 33, 122 (2002).
14.
15.
16.
17.
G. Lowe and B. V. L. Potter, J. Chem. Soc., Perkin Trans. 1, 2026 (1980).
H. Zhao, F. W. Foss, and R. Breslow, J. Am. Chem. Soc., 130, 12590 (2008).
P. L. Chow, D. B. Stout, E. Komisopoulou, and A. F. Chatziioannou, Phys. Med. Biol., 51, 379 (2006).
A. M. Loering and S. S. Gambhir, Mol. Imaging, 2, 131 (2003).
307