Fluorous synthesis of 18F radiotracers with the [ 18F]fluoride ion: Nucleophilic fluorination as the detagging process

Article abstract of DOI:10.1002/anie.200803897

Tag team: The flouro-detagging of flourous sulfonates by the [ ¹⁸F]fluoride ion was found to be an advantageous strategy for the preparation of various ¹⁸F-labeled prosthetic groups and known radiotracers (see picture). Fluorous solid phase extraction (FSPE) was used to separate the excess fluorous precursor from the labeled material, which suggests that traditional purification protocols such as distillation or tedious separation can be avoided. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200803897

Communications
DOI: 10.1002/anie.200803897
Radiochemistry
18
Fluorous Synthesis of F Radiotracers with the [18F]Fluoride Ion:
Nucleophilic Fluorination as the Detagging Process**
Romain Bejot, Thomas Fowler, Laurence Carroll, Sophie Boldon, Jane E. Moore,
Jꢀrꢁme Declerck, and Vꢀronique Gouverneur*
Positron emission tomography (PET) is a molecular imaging
method of ever-increasing importance for diagnostic medi-
cine and clinical pharmaceutical research.[1] The availability
of structurally assorted radiotracers is crucial for break-
through progress and is currently limited by the speed and
efficiency of synthetic methods. This is particularly relevant
for radiolabeled molecular probes derived from short-lived
isotopes (for example, 11C or 18F). In radiosynthesis, the
nonradioactive precursor is often used in large excess (mmol–
mmol) relative to the amount of radiolabeling agent (pmol–
nmol). The radiolabeled compound must be separated from
the excess precursor prior to clinical use or for further
chemical transformations. This separation is critical when the
precursor can compete with the radiolabeled probe for the
in vivo target or cause unwanted biological effects. Purifica-
tion of radiolabeled compounds is conventionally performed
by using one of four techniques. Whilst HPLC is a very
powerful technique for separation, it is time-consuming and is
not an attractive option for short-lived radionuclides as decay
leads to a significant loss of radioactivity. Based on the same
principle as HPLC, chromatography by means of solid phase
extraction is fairly narrow in scope, as the technique requires
the precursor and the radiotracer to have significantly
different affinities for the stationary phase. Distillation is a
useful technique for volatile radiolabeled compounds but this
process is time consuming, difficult to implement, and may
suffer from poor reproducibility. In addition, heating of the
reaction mixtures may result in decomposition.[2,3] In recog-
nition of these limitations, the strategy of solid-phase labeling
has emerged as an attractive alternative technology. The
labeling reaction allows the radiotracer to be released from
the insoluble solid-supported substrate, which itself is
removed by simple filtration.[4] Although conceptually ele-
gant, the radiolabeling reaction itself typically requires
extensive optimization and could display an unfavorable
kinetic profile because of the heterogeneity of the reaction
mixture.
In contrast to solid-phase methods, fluorous chemistry is
best considered as solution phase synthesis with the benefit of
separation tagging. Light-fluorous technology is particularly
attractive since the unique separation properties of molecules
containing a highly perfluorinated domain allows for the
rapid purification of crude reaction mixtures by using
fluorous solid phase extraction (FSPE).[5] Radiolabeling
strategies that rely on the use of fluorous soluble supports
are extremely scarce, notable exceptions being the prepara-
tion of radiolabeled compounds 125I and 35S that have
relatively long half-lives.[6] Since 18F is one of the most
attractive radionuclides because of its routine accessibility by
cyclotron production, the high resolution of images obtained
and its intermediate half-life of 109.7 min, a key application of
fluorous synthesis is in the preparation of 18F-labeled
molecular probes. As part of our continuing investigations
into novel fluorination methodology, we developed a fluorous
strategy featuring an electrophilic fluorination as the detag-
ging process.[7] 18F Radiochemistry is currently dominated by
synthetic routes based on nucleophilic fluorination with the
[18F]fluoride ion, a reagent available in much higher specific
activity than electrophilic sources of [18F]fluorine. A platform
technology based on the detagging of a fluorous-tagged
precursor upon nucleophilic fluorination is therefore a
compelling application of fluorous chemistry (Figure 1).
[*] Dr. R. Bejot, T. Fowler, L. Carroll, S. Boldon, Prof. V. Gouverneur
Chemistry Research Laboratory
Mansfield Road, OX1 3TA Oxford (UK)
Fax: (+44)1865-28-502
E-mail: veronique.gouverneur@chem.ox.ac.uk
Figure 1. Fluorous tag displaceable upon nucleophilic fluorination.
Dr. J. E. Moore
AstraZeneca, Alderley Park
Macclesfield, SK10 4GT Cheshire (UK)
A fluorous variant of the commonly used nucleophilic
fluorination of alkyl sulfonates was chosen as a proof-of-
principle study,[8] and a light fluorous tag that could be
substituted with the [18F]fluoride ion was identified and
prepared. Though some perfluorinated sulfonates (triflate
and nonaflate) are widely used to carry out aliphatic
nucleophilic substitution, none of them have the character-
Dr. J. Declerck
Siemens Molecular Imaging
23-38 Hythe Bridge Street, OX1 2EP Oxford (UK)
[**] This work was generously supported by the DIUS and Siemens
Molecular Imaging (Project No TP/16636)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803897.
586
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 586 –589

Angewandte
Chemie
istic feature of a fluorous tag. A modification of the sulfonate
with a polyfluorinated alkyl chain (typically with 9–17
fluorine atoms) should provide a leaving group with such
properties. The first attempts performed with the commer-
cially available perfluorooctanesulfonyl halides 1, typically
used to obtain sulfonamides,[9] were unsuccessful. Although
the synthesis of alkyl perfluorooctane sulfonates has been
reported,[10] an ammonium sulfonate salt was the only product
formed in this case, because of the high reactivity of the alkyl
perfluorosulfonate that was presumably formed.[11,12] The
alternative tags 2a–b with a spacer of two carbon atoms
between the fluorous chain and the sulfonyl chloride group
were therefore prepared as the distancing of the sulfonyl
chloride functionality from the fluorous domain should
modulate reactivity (Scheme 1).
Consequently, transformations that are referred to as “click
reactions”, for example, the CuI-catalyzed Huisgen 1,3-cyclo-
addition or the nucleophilic opening of an epoxide,[14] have
proven highly valuable for the coupling of 18F-radiolabeled
prosthetic groups with sensitive substrates under mild or
aqueous conditions.[3,15] To test the practical utility of the
fluoro-detagging methodology, various fluorous precursors of
18F-radiolabeled prosthetic groups were synthesized and
subjected to fluorination (Table 1). 18F Fluorination was
Table 1: Fluorous synthesis of 18F prosthetic groups.
Entry
Precursors[a]
18F-Radiolabeled
products[b]
RCY
FSPE
ratio
[%][c]
MeCN/
H2O
>76[d]
(m=2)
1
7:3
7 (n=8)
9 (n=8)
11 (n=8)
8
2
3
>79[d]
>89[d]
7:3
7:3
10
12
Scheme 1. First- and second-generation tagging reagents.
The second-generation fluorous tags 2a–b were synthe-
sized following the two-step procedure outlined in
Scheme 2.[13] Compound 2a was further reacted with 4-
50
(m=3)
4
5
7:3
13 (n=8)
15 (n=6)
14
16
82
(m=3)
9:1[e]
[a] n=8 Rf =C8F17(CH2)2-SO2-; n=6 Rf =C6F13(CH2)2-SO2-. [b] Precursor
(10 mg), MeCN (0.3 mL), 15 min, 1208C. [c] crude RCY: decay-corrected
radiochemical yield or mean RCY based on m experiments. [d] Volatile
product; RCY was determined after subsequent coupling. [e] FSPE was
also successful with MeCN/DMSO 9:1.
performed in acetonitrile using [18F]À/K+-Kryptofix 222 and
FSPE implemented to separate the radiolabeled product from
the fluorous precursor, which was typically used in large
excess. After dilution with a fluorophobic solvent, the crude
reaction mixture was eluted through a FSPE cartridge and
washed with a fluorophobic eluent. Pleasingly, less than 5%
of radiolabeled material remained trapped on the cartridge,
while most of the unreacted fluoride was not eluted with
organic solvent. The fluorous starting material was recovered
in 20–40% yield by eluting the FSPE cartridge with aceto-
nitrile; another fraction of the precursor usually precipitated
in the reaction vessel while diluting with the fluorophobic
solvent. None or only a trace amount of fluorous precursors
could be detected in the eluted fraction by HPLC. The click
reagent [18F]-2-fluoroethylazide 8, an 18F building block useful
for “click coupling” was prepared upon nucleophilic fluoro-
detagging of 7 and was subsequently purified by FSPE. This
protocol was superior to previous radiosyntheses that rely on
purification by distillation (Table 1, entry 1).[3,16,17] [18F]-(2-
Fluoroethoxy)prop-1-yne 10, another compound belonging to
the class of “click partners” was obtained with good radio-
chemical yield (RCY) after FSPE without the need to have
recourse to distillation (Table 1, entry 2).[15,18] Both the 18F-
labeled dipole 8 and dipolarophile 10 successfully underwent
Scheme 2. Synthesis of the fluorous tagging reagents 2a–b.
phenyl-butan-1-ol, which resulted in the production of 4-
phenyl-butyl-1H,1H,2H,2H-perfluorodecane-1-sulfonate 5 in
78% yield. The nucleophilic fluorination of 5 was carried out
with tetrabutylammonium fluoride (TBAF), a reagent used as
a surrogate for “hot” [18F]fluoride ions (Scheme 3). Although
only traces of the desired fluorinated product were seen, this
encouraging result prompted us to validate the radiosynthesis
of both 18F-labeled prosthetic groups and molecular probes
commonly used for PET.
Although it is preferable to carry out the radiolabeling as
the last step of the synthetic sequence, this is not always
possible since the conditions required for the fluorination
may not be compatible with the radiotracer of interest.
Scheme 3. Validation of nucleophilic fluoro-detagging.
Angew. Chem. Int. Ed. 2009, 48, 586 –589
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
587

Communications
the click reaction in a subsequent step (see the Supporting
ical maximum value of 6.3 ꢀ 104 GBqmmolÀ1 is never reached
because of contamination with the stable isotope originating
from the radionuclide production, solvents, chemicals, and
impurities. It is thus essential to verify to what extent the
polyfluorinated tag may induce unwanted contamination with
“cold” 19F. The specific activity was evaluated for the
fluorination of both 4-phenylbutyl tosylate and its fluorous
analogue 5. Starting from both the fluorous and the non-
Information). [18F]-Epifluorohydrin 12, a prosthetic group
known to couple successfully with various nucleophiles, was
synthesized in higher RCY than previously reported (70–
75%) and successfully purified by FSPE (Table 1, entry 3).[19]
Similarly, the detagging of 13 afforded [18F]-N-(2-fluoroe-
thyl)phthalimide 14 in good RCY (Table 1, entry 4). FSPE
purification of 14 followed by deprotection with hydrazine
hydrate led to unmasked [18F]-2-fluoroethylamine.[2,20] The
doubly tagged diol 15, derived from ethylene glycol, was
found to be amenable to monofluorination and afforded the
monotagged substrate 16 in 82% RCY. For this reaction, it
was essential to use the lighter C6F13 fluorous tag 2b.
Importantly, the product 16 was separated from the heavy
fluorous tagged precursor (C12F26) 15 by FSPE, which suggests
that it is no longer necessary to use HPLC for the purification
of the non-fluorous 18F-fluoroalkylating agent [18F]-2-fluo-
roethyltosylate (Table 1, entry 5).[21]
fluorous precursors, [18F]-4-phenylbutyl fluoride
6 was
obtained in 91% and 90% RCY, with specific activities of
1–10 GBqmmolÀ1 (m = 3) and greater than 100 GBqmmolÀ1
(m = 2), starting from approximately 200 MBq [18F]À ions. It
therefore appears that, under the radiolabeling conditions,
the fluorous tag induced a decrease of the specific activity
possibly from leaching of “cold” fluoride ion from the tag,[28]
a
limitation which has to be taken into consideration if the
tracer is toxic or able to saturate the receptor sites.
We have demonstrated that 18F-radiolabeled material can
be prepared by nucleophilic fluorination using fluorous-
tagged precursors and purified by FSPE regardless of the
affinity of the untagged substrate for the stationary phase.
FSPE-purified labeled compounds can then be used in
subsequent reactions or more easily purified by HPLC
before administration. FSPE is an easily implemented
separation technique, which can be run by automated systems.
One key feature of fluorous radiochemistry is the possibility
to perform reactions in homogeneous phases. This allows for
more favorable reaction kinetics than solid-phase synthesis.[29]
Furthermore, the light fluorous approach requires minimal
optimization, as the reaction conditions are comparable to
those reported in the literature using conventional sulfonates.
Further extensions are underway to tune the reactivity of the
precursors by using various linkers between the fluorous
chain and the sulfonyl moiety. Indeed, the optimum choice of
the leaving group is usually strongly dependent on the
substrate (e.g., for FDG production, a triflate group is used
as the leaving group in preference to a tosylate group). The
above technology offers an appealing versatility for the
synthesis of radiotracers used in PET, but can be applied to
the preparation of biomarkers required for alternative imag-
ing modalities, for example, single photon emission computed
tomography (SPECT).
Our attention then turned towards the synthesis of
18F radiotracers frequently used for PET (Scheme 4).
[18F]FMISO, a biomarker for tissue hypoxia, was successfully
prepared in moderate radiochemical yield from epifluorohy-
drin 12 (Scheme 4, reaction (1)). [18F]FECh, a potential tracer
for metabolic cancer imaging,[22] was obtained in 84% RCY
from 16. For this transformation, a mixture of acetonitrile/
DMSO was used to elute the FSPE cartridge in order to avoid
the presence of water that can compete with the amine for the
substitution of the sulfonate group (Scheme 4, reaction (2)).
The fluorous protocol was also applied to the synthesis of
[18F]-cis-4-fluoro-l-proline 17, a biomarker used for the
imaging of brain tumors.[23] A variant on the reported
radiosynthesis method,[24,25] was developed using the fluorous
precursor 18. [18F]-cis-N-(tert-Butoxycarbonyl)-4-fluoropro-
line methyl ester was prepared in a moderate radiochemical
yield of 42% (m = 3) and the nucleophilic fluorination
occurred with clean inversion.[24] FSPE purification and
quantitative deprotection with triflic acid afforded 17
(Scheme 4, reaction (3)).[26]
An important factor that requires consideration in PET is
the specific activity, defined as the level of radioactivity per
amount of tracer.[27] For 18F-labeled radiotracers, the theoret-
Experimental Section
Typical procedure: FSPE cartridges were prepared in-house using a
Sep-Pak Light tube (Waters, Milford, MA) and fluorous silica
(180 mg; Fluorous Technology, Pittsburgh, PA). 18F Fluorination
was typically performed in acetonitrile (0.3–0.5 mL) using dry
[18F]À/K+-Kryptofix 222 and precursor (10 mg). After heating for
15 min at 100–1208C, the crude reaction mixture was cooled to room
temperature and diluted with a fluorophobic solvent such as water or
DMSO to obtain a fluorophilic/fluorophobic ratio of typically 7:3,
prior to elution through a FSPE cartridge. The cartridge was
subsequently washed with the fluorophilic/fluorophobic eluent
(0.5 mL) to collect the 18F-radiolabeled compound.
Received: August 7, 2008
Revised: October 10, 2008
Published online: December 3, 2008
Scheme 4. Fluorous synthesis of known 18F radiotracers.
588
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 586 –589

Angewandte
Chemie
[14] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. 2001, 113,
Keywords: fluorine · positron emission tomography ·
radiopharmaceuticals · synthetic methods
.
2056; Angew. Chem. Int. Ed. 2001, 40, 2005.
[15] J. Marik, J. L. Sutcliffe, Tetrahedron Lett. 2006, 47, 6681.
[16] a) L. Carroll, R. Bejot, R. Huting, S. Bayly, J. Dilworth, A. Gee,
J. Declerck, V. Gouverneur, J. Nucl. Med. 2008, 49, 98P; b) M.
Glaser, M. Solbakken, J. Arukwe, H. Karlsen, A. Cuthbertson, S.
Luthra, E. ꢁrstad, J. Nucl. Med. 2008, 49, 96P.
[1] a) Handbook of Radiopharmaceuticals (Eds.: M. J. Welch, C. S.
Redvanly), Wiley, New York, 2003; b) S. M. Ametamey, M.
Honer, P. A. Schubiger, Chem. Rev. 2008, 108, 1501; c) D. W.
Townsend, J. Nucl. Med. 2008, 49, 938.
[17] The RCY and distillation efficiency of [18F]-2-fluoroethylazide
were reported as 54% and 63%, respectively.[3]
[18] U. Sirion, H. J. Kim, J. H. Lee, J. W. Seo, B. S. Lee, S. J. Lee, S. J.
Oh, D. Y. Chi, Tetrahedron Lett. 2007, 48, 3953.
[19] a) J. R. Grierson, J. M. Link, C. A. Mathis, J. S. Rasey, K. A.
Krohn, J. Nucl. Med. 1989, 30, 343; b) E. L. Kꢄmꢄrꢄinen, T.
Kyllꢅnen, O. Nihtila, H. Bjꢅrk, O. Solin, J. Labelled Compd.
Radiopharm. 2004, 47, 37.
[20] a) C. Gilissen, G. Bormans, T. de Groot, A. Verbruggen, J.
Labelled Compd. Radiopharm. 1998, 41, 491; b) S. Lehel, G.
Horvꢂth, I. Boros, P. Mikecz, T. Mꢂriꢂn, A. J. Szentmiklꢆsi, L.
Trꢆn, J. Labelled Compd. Radiopharm. 2000, 43, 807; c) J. S.
Rasey, P. D. Hofstrand, L. K. Chin, T. J. Tewson, J. Nucl. Med.
1999, 40, 1072.
[21] a) D. Block, H. H. Coenen, G. Stocklin, J. Labelled Compd.
Radiopharm. 1988, 25, 201; b) A. Bauman, M. Piel, R. Schirr-
macher, F. Rꢅsch, Tetrahedron Lett. 2003, 44, 9165.
[22] a) M. Piel, A. Bauman, R. P. Baum, S. Hꢅhnemann, I. Klette, R.
Wortmann, F. Rꢅsch, Bioorg. Med. Chem. 2007, 15, 3171; b) T.
Hara, N. Kosaka, H. Kishi, J. Nucl. Med. 2002, 43, 187; c) T. R.
DeGrado, S. W. Baldwin, S. Wang, M. D. Orr, R. P. Liao, H. S.
Friedman, R. Reiman, D. T. Price, R. E. Coleman, J. Nucl. Med.
2001, 42, 1805.
[23] a) K. J. Langen, K. Hamacher, D. Pauleit, F. W. Floeth, G.
Stoffels, D. Bauer, G. Reifenberger, K. Zilles, H. H. Coenen,
Anat. Embryol. 2005, 210, 455; b) K.-J. Langen, K. Hamacher, D.
Bauer, S. Broer, D. Pauleit, H. Herzog, F. Floeth, K. Zilles, H. H.
Coenen, J. Cereb. Blood Flow Metab. 2005, 25, 607.
[24] K. Hamacher, J. Labelled Compd. Radiopharm. 1999, 42, 1135.
[25] a) S. M. Mazza, J. Labelled Compd. Radiopharm. 2000, 43, 1047;
b) H.-J. Wester, M. Herz, R. Senekowitsch-Schmidtke, M.
Schwaiger, G. Stocklin, K. Hamacher, Nucl. Med. Biol. 1999,
26, 259.
[2] T. J. Tewson, Nucl. Med. Biol. 1997, 24, 755.
[3] M. Glaser, E. ꢁrstad, Bioconjugate Chem. 2007, 18, 989.
[4] a) K. Kowai, H. Ohta, M. A. Channing, A. Kubodera, W. C.
Eckelman, Appl. Radiat. Isot. 1996, 47, 37; b) D. H. Hunter, X.
Zhu, J. Labelled Compd. Radiopharm. 1999, 42, 653; c) R.
Dunn-Dufault, A. Pollak, J. Fitzgerald, J. R. Thornback, J. R.
Ballinger, Nucl. Med. Biol. 2000, 27, 803; d) A. Pollak, D. G.
Roe, C. M. Pollock, L. F. L. Lu, J. R. Thornback, J. Am. Chem.
Soc. 1999, 121, 11593; e) S. Mundwiler, L. Candreia, P. Hafliger,
K. Ortner, R. Alberto, Bioconjugate Chem. 2004, 15, 195; f) L. J.
Brown, D. R. Bouvet, S. Champion, A. M. Gibson, Y. Hu, A.
Jackson, I. Khan, N. Ma, N. Millot, H. Wadsworth, R. C. D.
Brown, Angew. Chem. 2007, 119, 959; Angew. Chem. Int. Ed.
2007, 46, 941.
[5] a) Handbook of Fluorous Chemistry (Eds.: J. A. Gladysz, D. P.
Curran, I. T. Horvꢂth), Wiley-VCH, Weinheim, 2004; b) D. P.
Curran, Synlett 2001, 1488; c) W. Zhang, D. P. Curran, Tetrahe-
dron 2006, 62, 11837.
[6] a) A. Donovan, J. Forbes, P. Dorff, P. Schaffer, J. Babich, J. F.
Valliant, J. Am. Chem. Soc. 2006, 128, 3536; b) A. S. Zhang, C. S.
Elmore, M. A. Egan, D. G. Melillo, D. C. Dean, J. Labelled
Compd. Radiopharm. 2005, 48, 203.
[7] S. Boldon, J. E. Moore, V. Gouverneur, Chem. Commun. 2008,
3622.
[8] a) F. Dollꢃ, D. Roeda, B. Kuhnast, M.-C. Lasne in Fluorine and
Health (Eds.: A. Tressaud, G. Haufe), Elsevier, Dordrecht, 2008;
b) L. Cai, S. Lu, V. W. Pike, Eur. J. Org. Chem. 2008, 2853.
[9] a) X. Hao, A. Yoshida, J. Nishikido, J. Fluorine Chem. 2006, 127,
193; b) L. Zu, J. Wang, H. Li, W. Wang, Org. Lett. 2006, 8, 3077.
[10] P. M. Savu, US Patent 08/341448, 1994.
[11]
[26] 50 mL triflic acid (2m) was sufficient to deprotect 18 after FSPE,
compared to the reported amount of 0.5 mL.
[27] a) M.-C. Lasne, C. Perrio, J. Rouden, L. Barrꢃ, D. Roeda, F.
Dolle, C. Crouzel, in Contrast Agents II, Vol. 222 (Ed.: W.
Krause), Springer, Berlin, 2002, pp. 201; b) W. C. Eckelman, M.
Bonardi, W. A. Volkert, Nucl. Med. Biol. 2008, 35, 523.
[28] P. S. Johnstrꢅm, S. Stone-Elander, Appl. Radiat. Isot. 1996, 47,
401.
[12] a) Y. Li, Z.-Q. Chen, J. Tian, Y.-B. Zhou, Z.-X. Chen, J. Fluorine
Chem. 2004, 125, 1077; b) Y. Li, Z.-Q. Chen, J. Tian, Y.-B. Zhou,
Z.-X. Chen, Z.-J. Liu, J. Fluorine Chem. 2005, 126, 888.
[13] a) F. Szonyi, A. Cambon, J. Fluorine Chem. 1989, 42, 59; b) L.
Foulletier, Deutsches Patent P1667117, 1969; c) H. Stach,
Deutsches Patent P2059257, 1970.
[29] C. Chen, W. Zhang, Mol. Diversity 2005, 9, 353.
Angew. Chem. Int. Ed. 2009, 48, 586 –589
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
589