Synthesis of [18F]fluoro-pivalic acid: An improved PET imaging probe for the fatty acid synthesis pathway in tumours

Article abstract of DOI:10.1039/c3md00169e

[¹⁸F]Fluoro-pivalic acid ([¹⁸F]FPIA), an analogue of [¹⁸F]fluoroacetate bearing a dimethyl moiety on C-2, has been radiosynthesised and evaluated in vivo. [¹⁸F]FPIA has high tumour uptake and, unlike [¹⁸

Full text of DOI:10.1039/c3md00169e

MedChemComm
Broadening the field of opportunity for medicinal chemists
www.rsc.org/medchemcomm
Volume 4 | Number 10 | October 2013 | Pages 1315–1414
ISSN 2040-2503
CONCISE ARTICLE
Federica Pisaneschi et al.
Synthesis of [18F]fluoro-pivalic acid: an
improved PET imaging probe for the
fatty acid synthesis pathway in tumours
2040-2503(2013)4:10;1-1

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
Synthesis of [18F]fluoro-pivalic acid: an improved PET
imaging probe for the fatty acid synthesis pathway in
tumours†
Cite this: Med. Chem. Commun., 2013,
4, 1350
*
Federica Pisaneschi,‡ Timothy H. Witney,‡ Lisa Iddon‡ and Eric O. Aboagye
Received 13th June 2013
Accepted 16th July 2013
[
18F]Fluoro-pivalic acid ([18F]FPIA), an analogue of [18F]fluoroacetate bearing a dimethyl moiety on C-2,
has been radiosynthesised and evaluated in vivo. [18F]FPIA has high tumour uptake and, unlike
18F]fluoroacetate, does not defluorinate.
DOI: 10.1039/c3md00169e
www.rsc.org/medchemcomm
[
Prostate cancer is the most prevalent of cancers in the male contributing to elevated acetate uptake and incorporation into
population, accounting for 40 841 incidents in the UK in 2009.1 the cell membrane.6,7 Furthermore, Yoshii and co-workers
In the USA prostate cancer is the second largest cause of cancer reported that tumour uptake of acetate reects cytosolic acetyl-
mortality amongst men. Positron emission tomography (PET) is CoA synthetase,8 an enzyme that is regulated in part by the fatty
routinely used in the clinic for cancer detection, with the acid content.
majority of all scans performed with the glucose analogue [18F]-
[11C]Acetate has shown promise for prostate cancer
2-uoro-2-deoxyglucose, [18F]FDG. A number of clinical studies, imaging,9 however the short half life of carbon-11 (20.4 min)
however, have shown poor sensitivity for prostate cancer requires an on-site cyclotron, limiting its wide-spread use. [18F]-
detection by [18F]FDG-PET, resulting from low basal glucose Fluoroacetate ([18F]FAC) was developed by Sykes et al.10 and
metabolism of some prostate tumours2 and the high renal investigated as an alternative to [11C]acetate for imaging of
clearance of [18F]FDG, which can oen mask tumour uptake.3 prostate cancer by Welch and co-workers.11 The advantage of
As a result, other PET tracers have been developed for prostate using uorine-18 is its longer half life of 109.5 min compared to
cancer imaging such as [11C]choline, [18F]uorocholine and carbon-11. The authors investigated the biodistribution of [11C]-
[11C]acetate.
acetate and [18F]FAC in Sprague-Dawley rats. They found fairly
[11C]Acetate was initially developed as a radiotracer to eval- rapid clearance of [11C]acetate from most of the organs except
uate oxidative metabolism in the myocardium. Following entry the pancreas at 1 h, perhaps due to the oxidative metabolism of
into the cell, either by passive diffusion or membrane transport [11C]acetate, releasing [11]CO2, whereas [18F]FAC clearance was
via the monocarboxylate transporters,4 [11C]acetate is converted slower from most organs. The main drawback of [18F]FAC is its
to [11C]acetyl-CoA by acetyl-CoA synthetase before its rapid substantial bone uptake, characteristic of radiotracer deuori-
metabolism via the citric acid cycle to [11C]CO2.5 As well as being nation.11 There are a number of putative routes for deuorina-
a substrate for the citric acid cycle, acetyl-CoA also contributes tion (Scheme 1). Tecle and Casida,12 for example, found that
to de novo fatty acid synthesis, with tumour-associated [11C]-
acetate accumulation shown to result from cell membrane
incorporation following ux through the fatty acid synthesis
pathway. Experiments carried out by Yoshimoto and co-workers
showed that [1-14C]acetate was incorporated into the lipid
soluble fraction, mostly as phosphatidylcholine and neutral
lipids. The accumulation of [1-14C]acetate was positively corre-
lated with the growth of the tumour cells. Fatty acid synthase
(FAS) has been shown to be overexpressed in tumours,
Comprehensive Cancer Imaging Centre, Department of Surgery and Cancer, Imperial
Centre for Translational and Experimental Medicine, Hammersmith Hospital, Du
Cane Road, London W12 0NN, UK. E-mail: f.pisaneschi@imperial.ac.uk; Tel: +44 (0)
20 3313 3204
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c3md00169e
‡ These authors contributed equally to this work.
Scheme 1 Mechanisms of [18F]FAC defluorination.
1350 | Med. Chem. Commun., 2013, 4, 1350–1353
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Concise Article
MedChemComm
incubation of [13C]uoroacetate with rat and mouse liver cytosol
led to deuorination, with uoride ion displacement resulting
from nucleophilic attack by glutathione (GSH). [18F]FAC also
undergoes deuorination via its metabolism in the citric acid
cycle. Once [18F]FAC enters the citric acid cycle, it is converted
into 2-uorocitrate 3. The same authors,12 together with other
groups13,14 showed that (ꢀ)-erythro-2-uorocitrate 3 is further
converted to uoro-cis-aconitate by aconitase, undergoing
addition of hydroxide and subsequent loss of uoride to form
4-hydroxy-trans-aconitate 4. Compound 4 binds very tightly to
the enzyme and it is responsible for the toxicity of uoroacetate
at pharmacological levels. Of note, evidence of deuorination
has been found in rodents and pigs, although the rate of
deuorination was reduced in non-human primates.15,16
Given the inadequate performance of [18F]FAC as a tracer
that can be used to image biology of acetate metabolism pre-
clinically in prostate cancer and beyond, there is a requirement
to nd a stable imaging agent which does not undergo
deuorination.
Results and discussion
Herein we describe the synthesis of a novel analogue for
imaging of acetate metabolism. The criteria for a new analogue
was that it should be stable to deuorination in rodents, but
still show good uptake in tumour cell lines with increased
fatty acid metabolism. An acetate analogue, [18F]uoro-pivalic
acid ([18F]FPIA, 5, Fig. 1), also called 3-[18F]uoro-2,2-dimethyl
propionic acid, was selected, as it was hypothesised that the
gem-dimethyl group would prevent this compound from Fig. 1 Citrate synthase mechanism: first step. The previously published mecha-
nism for entry of acetate into the citric acid cycle17 has been adapted for [11C]
entering the citric acid cycle, not being a substrate for citrate
synthase (Fig. 1), thought to contribute to deuorination and
acetate, [18F]FAC and [18F]FPIA. The lack of a proton a to the carbonyl prevents
Asp375 from attacking [18F]FPIA, disabling the compound from entering the citric
slow background clearance from other organs. It was also
acid cycle.
proposed that
[
18F]FPIA would be slightly less reactive
towards glutathione attack due to the inductive effect of the
gem-dimethyl group and the increased distance of the
carboxylic acid.
[18F]FPIA is, however, hypothesised to have the capacity to
map the initial step of the fatty acid synthesis pathway because
it has the potential to be converted into an acetyl-CoA derivative.
This would be analogous to the use of [18F]uorodeoxyglucose
to map glucose metabolic ux by tumour cells. The synthesis of
[18F]uoroacetate is generally carried out using a precursor
protected as a methyl or ethyl ester. Ponde et al. puried the
ester intermediate using a HLB cartridge method. We initially
rationalised the synthesis of a FPIA precursor to involve a non-
base sensitive carboxylic acid protecting group such as
p-methoxybenzyl ester that would be stable during the aqueous
dilution of the reaction before preparative HPLC. The
commercially available 3-bromo-2,2-dimethylpropionic acid 6
was used in a Mitsunobu reaction with 4-methoxybenzyl
Scheme 2 Reagents and conditions: (a) PPh3, DIAD, THF, 0 ꢁC. (b) HBTU, DBU,
CH2Cl2, rt. (c) TsCl, DMAP, pyridine, rt (70%).
We therefore decided to exploit the methyl ester protecting
alcohol. The reaction gave no desired product. A reaction was group strategy as shown by Ponde et al.11 To synthesise a suit-
then attempted using HBTU to form the activated ester, in the able methyl ester protected precursor, methyl 2,2-dimethyl-3-
presence of DBU. This showed evidence of the desired product hydroxypropionate (7) was utilised. The hydroxyl group was
but was difficult to separate from the 4-methoxybenzyl alcohol converted to the tosylate leaving group in good yield (70%)
and other by-products. We suspected that the bromine atom (Scheme 2). The precursor 8 was then taken forward for radio-
competed in the reaction (Scheme 2).
chemistry. Initial experiments to incorporate uorine-18 were
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1350–1353 | 1351

View Article Online
MedChemComm
Concise Article
carried out using K2CO3 and kryptox to form the K18F/kryp-
Once isolated from the HPLC eluent, the product was
tand complex, followed by addition of the precursor in aceto- hydrolysed using NaOH (1 M) in 5 min at 60 ꢁC which was then
nitrile at 80 ꢁC. The reaction was monitored by analytical HPLC. neutralised with HCl (1 M) and phosphate buffered saline was
These conditions gave a mixture of radiolabelled products aer added to achieve a pH of 7.4 and 10% EtOH/PBS or less. The
5 minutes and only 6% product; further heating led to decom- compound identity was proved by co-elution with the cold
position of the precursor and only 12% product [18F]9. The reference compound FPIA (220 nM). In aggregate, the entire
reaction was then undertaken using DMF as solvent at 105 C, synthesis and formulation takes 1.5 h and delivers [18F]FPIA
ꢁ
and this showed a robust reaction, giving the product [18F]9 (up ready for injection in the end of synthesis (EOS) yield of 11.3.ꢂ
to 75% of radiolabel incorporation) and [18F]toluenesulfonyl 4.1% (n ¼ 4) and >99% radiochemical purity. Accurate and
uoride (25% yield) as a by-product aer 15 minutes. At this reliable measurement of specic activity was not possible by
temperature the precursor was degraded to give methyl-3- HPLC due to low UV absorbance of FPIA (see the ESI†). The
hydroxy-2,2-dimethylpropionate. The intermediate was then synthesis was optimised manually with ꢃ750 MBq starting
isolated using reverse phase preparative HPLC and 30% activity. Transfer to an automated platform is currently under
ethanol/water as the eluent. The yield of the isolated compound investigation.
was remarkably lower than expected from the ratio of conver-
[18F]FPIA pharmacokinetics were tested in vivo in non-
sion. It was suspected that on dilution of the reaction mixture tumour bearing mice by PET imaging (Fig. 2). Substantial
with water, hydrolysis of the methyl ester had occurred giving uptake was observed in the cortex of the kidney, with clearance
the nal desired product [18F]FPIA. The product could not be primarily via urinary excretion. Tracer localisation was also
isolated as it co-eluted at the solvent front with the DMF and any observed in the heart, liver and intestines. Notably, there was a
un-reacted uoride-18. Due to hydrolysis, the isolated yield of lack of bone-associated radioactivity, suggesting the absence of
the methyl ester intermediate [18F]9 was lower than anticipated (or low) deuorination. A similar radiotracer biodistribution
(13.9 ꢂ 9.1% decay corrected, n ¼ 7). As a less basic environ- has been previously observed with [11C]acetate,15 except for the
ment should disfavour the hydrolysis of the methyl ester, route of excretion: elimination via [11C]CO2 for [11C]acetate and
KHCO3 was employed instead of K2CO3 to form the kryptand via the renal route (kidney and urine) for [18F]FPIA.
complex. The optimal reaction conditions were found to be
Dynamic [18F]FPIA-PET imaging was next performed in
heating at 110 ꢁC for 10 min. No other radiochemical peaks BALB/c mice bearing EMT6 murine breast adenocarcinoma
apart from that for the desired product were observed. This gave xenogras (Fig. 3). In this proof-of-concept study, in vivo
up to 80% of incorporation of the uoride as shown by analyt- imaging of implanted EMT6 xenogras showed rapid tumour
ical HPLC (40% MeOH/water) and improved the yield of the labelling and retention, with good target : background signal
isolated product to 39.3 ꢂ 12.6% (n ¼ 5) decay corrected measured 30–60 min post-injection (Fig. 3A and B).
(Scheme 3).
Intracellular EtOH is rapidly catabolised to acetate, which,
Good results were also obtained when [18F]TBAF was used as at high levels, was predicted to compete with the formation
the source of uoride. The uoride was dried in the presence of of [18F]FPIA-CoA and therefore tumour trapping. Tumour-
tetrabutylammonium hydrogen carbonate (TBAHCO3) to give associated radioactivity was increased following EtOH removal
[18F]TBAF. The use of the less basic TBAHCO3 led to reduced
hydrolysis before preparative HPLC. Analysis of the reaction
mixture showed complete incorporation of uoride to give the
desired product. Aer addition of water to the reaction mixture,
the preparative HPLC showed 44% product [18F]9 and 52% of
the hydrolysed nal product, together with unreacted uorine-
18. The methyl ester protected intermediate [18F]9 was isolated
in 37% decay corrected yield.
Scheme 3 Synthesis of [18F]FPIA. Reagents and conditions: (a) K2CO3, K222, [18F]
Fig. 2 PET imaging with [18F]FPIA. Representative sagittal CT, PET and PET-CT
images (30–60 min summed-activity) for [18F]FPIA. Key organs are identified.
Abbreviations: B, brain; H, heart; L, liver; K, kidney; I, intestines.
F
ꢀ or KHCO3, K222, [18F]Fꢀ or [18F]TBAF, 105–110 ꢁC, 10 min. (b) NaOH (1 M), 60
ꢁC, 5 min, then HCl (1 M) and PBS.
1352 | Med. Chem. Commun., 2013, 4, 1350–1353
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Concise Article
MedChemComm
and one HPLC purication step has been successfully per-
formed in the EOS yield of 11.3 ꢂ 4.1% (n ¼ 4). A key design goal
to avoid deuorination was achieved. Preliminary in vivo PET
imaging showed no bone uptake and high tumour uptake in
EMT6 tumour-bearing mice. These encouraging results have
prompted further biological evaluation of the radiotracer to
assess its potential for the imaging of elevated de novo fatty acid
synthesis.
Notes and references
1 http://info.cancerresearchuk.org/cancerstats/types/prostate/
incidence/.
2 A. H. Hou, K. F. Sullivan and E. D. Crawford, Curr. Opin.
Urol., 2009, 19, 283–289.
3 R. Wahl, J. Harney, G. Hutchins and H. Grossman, J. Urol.,
1991, 146, 1470–1474.
4 R. A. Waniewski and D. L. Martin, J. Neurosci., 1998, 18,
5225–5233.
5 R. R. Wolfe and F. Jahoor, Am. J. Clin. Nutr., 1990, 51, 248–
252.
6 F. P. Kuhajda, Nutrition, 2000, 16, 202–208.
7 S. J. Kridel, W. T. Lowther and C. W. Pemble, Expert Opin.
Invest. Drugs, 2007, 16, 1817–1829.
8 Y. Yoshii, A. Waki, T. Furukawa, Y. Kiyono, T. Mori,
H. Yoshii, T. Kudo, H. Okazawa, M. J. Welch and
Y. Fujibayashi, Nucl. Med. Biol., 2009, 36, 771–777.
9 C. Schiepers, S.-C. Huang and M. Dahlbom, J. Nucl. Med.,
2013, 54, 326.
Fig. 3 Dynamic [18F]FPIA PET imaging in EMT6 xenografts. Representative
coronal (A) and axial (B) PET-CT images (30–60 min summed-activity) for [18F]FPIA.
White arrowheads indicate the tumour, identified from the CT image. (C) EMT6
tumour time versus radioactivity curve (TAC) obtained from 60 minute dynamic
PET imaging with and without EtOH in the final prep. Mean ꢂ SD (n ¼ 4–6 mice
per group). (D and E) Semi-quantitative imaging variables extracted from the TAC:
tumour radioactivity 60 min post-injection (D) and area under the TAC, 0–60 min
(E). Mean ꢂ SD (n ¼ 4–6 mice per group).
10 T. R. Sykes, T. J. Ruth and M. J. Adam, Int. J. Radiat. Appl.
Instrum., Part B, 1986, 13, 497–500.
11 D. E. Ponde, C. S. Dence, N. Oyama, J. Kim, Y.-C. Tai,
R. Laforest, B. A. Siegel and M. J. Welch, J. Nucl. Med.,
2007, 48, 420–428.
12 B. Tecle and J. E. Casida, Chem. Res. Toxicol., 1989, 2, 429–
435.
13 H. Lauble, M. C. Kennedy, M. H. Emptage, H. Beinert and
C. D. Stout, Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 13699–
13703.
of the nal prep by heating at 45 ꢁC under vacuum and a stream
of nitrogen, shown by the time versus radioactivity curves (TAC;
Fig. 3C). A signicant 1.2-fold increase in tumour radioactivity
was measured following EtOH removal at 60 min post-injection,
rising from 6.49 ꢂ 0.51% ID mLꢀ1 to 7.92 ꢂ 0.88% ID mLꢀ1
(P ¼ 0.011; Fig. 3D). Furthermore, a 1.2-fold increase in [18F]-
FPIA was also measured in EMT6 xenogras over the entire 60
minute time course following EtOH removal, dened by the
area under the TAC, increasing from 378 ꢂ 30% ID mLꢀ1 minꢀ1
to 472 ꢂ 40% ID mLꢀ1 minꢀ1 (P ¼ 0.0029; Fig. 3E). Although the
metabolic fate of [18F]FPIA is yet to be experimentally deter-
mined, [18F]FPIA shows great promise as a cancer imaging
agent. Future work will determine its mode of action and
provide further biological characterisation.
14 J. J. Villafranca and E. Platus, Biochem. Biophys. Res.
Commun., 1973, 55, 1197–1207.
¨
˚
¨
15 O. Lindhe, A. Sun, J. Ulin, O. Rahman, B. Langstrom and
¨
J. Sorensen, Eur. J. Nucl. Med. Mol. Imaging, 2009, 36, 1453–
1459.
16 R. Nishii, W. Tong, R. Wendt, III, S. Soghomonyan,
U. Mukhopadhyay, J. Balatoni, O. Mawlawi, L. Bidaut,
P. Tinkey, A. Borne, M. Alauddin, C. Gonzalez-Lepera,
B. Yang and J. Gelovani, Mol. Imaging Biol., 2012, 14, 213–
224.
Conclusions
A new uorine-18 labelled tracer, [18F]FPIA, analogue of the
known [18F]FAC, has been designed to potentially image acetate
metabolism and the related fatty acid pathway, overexpressed in 17 G. Wiegand and S. J. Remington, Annu. Rev. Biophys. Biophys.
certain types of cancer. The radiosynthesis involving two steps
Chem., 1986, 15, 97–117.
This journal is ª The Royal Society of Chemistry 2013
Med. Chem. Commun., 2013, 4, 1350–1353 | 1353