Radiosynthesis and in vivo tumor uptake of 2-deoxy-2-[18F] fluoro-myo-inositol

Article abstract of DOI:10.1016/j.bmcl.2012.08.022

Inositols play an important role in membrane lipid metabolism and mitogenic signaling of most cancer cells. There is paucity of data on the distribution of radiolabelled inositols. Based on work previously carried out on 1-deoxy-1-[¹⁸F]fluoro-scyllo-inositol ([¹⁸F]2), we began a program of work to label myo-inositol (2-deoxy-2-[¹⁸F]fluoro-myo- inositol, [¹⁸F]1), the most abundant inositol in cells. Fluorination of a triflate precursor 4 afforded the desired [¹⁸F]1 following deprotection with a radiochemical yield of 8% n.d.c. [¹⁸F]1 showed higher uptake in vivo in a human breast cancer xenograft model, MDA-MB-231, compared to [¹⁸F]2. Thus, we have developed a new inositol radiotracer that could have utility for studying inositol uptake in tumors.

Full text of DOI:10.1016/j.bmcl.2012.08.022

Bioorganic & Medicinal Chemistry Letters 22 (2012) 6148–6150
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Radiosynthesis and in vivo tumor uptake of 2-deoxy-2-[18F]fluoro-myo-inositol
Laurence Carroll a, Meg Perumal a, Neil Vasdev b, Edward Robins c, Eric O. Aboagye a,
⇑
a Comprehensive Cancer Imaging Centre, Department of Surgery and Cancer, Hammersmith Campus, Imperial College, London W12 0NN, UK
b Department of Radiology, Harvard Medical School and Department of Nuclear Medicine and Molecular Imaging, Massachusetts General Hospital, 55 Fruit St., Boston,MA 02114, USA
c GE Healthcare Immanet, Cyclotron Building, Du Cane Road, Hammersmith Campus, London W12 0NN, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Inositols play an important role in membrane lipid metabolism and mitogenic signaling of most cancer
cells. There is paucity of data on the distribution of radiolabelled inositols. Based on work previously car-
ried out on 1-deoxy-1-[18F]fluoro-scyllo-inositol ([18F]2), we began a program of work to label myo-ino-
sitol (2-deoxy-2-[18F]fluoro-myo-inositol, [18F]1), the most abundant inositol in cells. Fluorination of a
triflate precursor 4 afforded the desired [18F]1 following deprotection with a radiochemical yield of 8%
n.d.c. [18F]1 showed higher uptake in vivo in a human breast cancer xenograft model, MDA-MB-231, com-
pared to [18F]2. Thus, we have developed a new inositol radiotracer that could have utility for studying
inositol uptake in tumors.
Received 13 June 2012
Revised 2 August 2012
Accepted 3 August 2012
Available online 11 August 2012
Keywords:
Fluorine-18
Inositol
Ó 2012 Elsevier Ltd. All rights reserved.
Cancer
PET Imaging
myo-Inositol, the most abundant inositol in cells, is taken up by
diverse cell types including normal and malignant cells, and incor-
porated into phospholipids. In humans, cellular uptake is mediated
by high affinity human sodium-inositol co-transporter and hydro-
gen-dependent symporter.1,2 myo-Inositol is involved in insulin
signaling, and signaling via the epidermal growth factor receptor
(EGF) is known to enhance myo-inositol uptake in EGF responsive
tumor cells.3,4 There is limited knowledge, however, of the uptake
of labelled inositols including myo-inositol in malignant cells in
vivo.
18F chemistry has been widely developed over the past thirty
years, with a number of advances in both labelling methodology
and purification of compounds at the end of synthesis, to widen
the available pool of radiotracers and synthons for new targets.
9–11
As aforementioned, Vasdev et. al. synthesised [18F]2 in a 25%
n.d.c. radiochemical yield starting from triflate precursor 3 by
nucleophilic substitution followed by sequential deprotections of
the acid- and base-labile protecting groups. To allow comparison
of [18F]1–[18F]2, we repeated the synthesis of [18F]2. The overall
procedure was carried in an almost identical fashion to that from
the literature, with only minor modifications in the purification
and formulation steps. Whereas [18F]2 had previously been puri-
fied by solid-phase cartridge purification, we were able to use re-
verse-phase semi-preparative HPLC to afford the desired
radiotracer [18F]2 in ꢀ1–2 mL 10% ethanol in saline solution,
compared to 5–6 mL which had been previously reported, with
no detriment to the radiochemical yield (RCY). This enables a
greater range of in vivo work, using a much lower starting activity
within 90 min from the end of radionucleide production. Radio-
chemical purity was found to be greater than 95%. Confirmation
Recently, Vasdev et al. radiolabelled scyllo-inositol, the 2-epimer
of myo-inositol, with fluorine-18, and showed that the radiotracer,
1-deoxy-1-[18F]fluoro-scyllo-inositol ([18F]2), was taken up into a
human breast tumor xenograft model.5,6 Here, we have synthesized
2-deoxy-2-[18F]fluoro-myo-inositol ([18F]1) (Fig. 1) and evaluated
its tumor uptake in the same xenograft model. We compare
2-deoxy-2-[18F]fluoro-myo-inositol uptake to that of 1-deoxy-1-
[
18F]fluoro-scyllo-inositol.
Precursor 3 was subjected to potassium acetate at 60 °C, leading
to an inversion of the axial triflate present in 3, in relatively low
isolated yield of 27% (Scheme 1).7,8 A large amount of the remain-
ing mass balance of the reaction appeared to be various elimina-
tion products, caused by the use of
a base at elevated
temperature. Conversion of the acetyl protected 5 to the desired
equatorial triflate 4 was carried over two steps in a moderate over-
all yield of 44%, similar to that previously reported in the literature
(see Supplementary data for further details).
⇑
Corresponding author.
E-mail address: eric.aboagye@imperial.ac.uk (E.O. Aboagye).
Figure 1. Structures of fluorine labeled inositols.
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.08.022

L. Carroll et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6148–6150
6149
MeO
O
MeO
OTf
O
O
OMe
AcO
OMe
KOAc, DMA
60oC, 1 hr
O
O
O
O
OBz
Yield = 27 %
5
OBz
MeO
MeO
O
OMe
OMe
3
AcCl, MeOH
RT, 16 hrs
MeO
O
OMe
HO
O
O
OBz
MeO
O
Yield = 55 %
OMe
6
Tf2O, Pyr.
DCM
-78oC, 2.5 hrs
MeO
O
OMe
O
TfO
O
OBz
Yield = 79 %
MeO
O
OMe
4
Scheme 1. Synthesis of compound 4.
Figure 3. PET imaging of MDA-MB-231 tumor bearing mice with [18F]1 or [18F]2. A.
Coronal, and B. transverse (30–60 min; fused 0.5 mm slice) PET-CT slices through a
mouse injected with [18F]1. C. Time versus radioactivity curves showing time
dependent uptake of [18F]1 and [18F]2 in vivo over 60 min. Inset: First 10 min D.
Scheme 2. Radiosynthesis of [18F]1.
Comparison of
[ [
18F]1 and 18F]2 at 60 min after injection of the respective
radiotracers. n = 6 mice per group.
of the product identity was carried out by radio-TLC, where the Rf
of [18F]2 was found to correlate with a 19F standard of 2 upon
potassium permanganate staining. Unfortunately, the radiochemi-
cal yield of [18F]2 was significantly lower than that previously
encountered within the most recent publication of the desired
compound (8% vs 25% n.d.c.). However, this is most likely due to
this reported synthesis being carried out on an automated plat-
form, and when the radiosynthesis of [18F]2 was previously carried
out using analogous equipment to our set-up, an identical radio-
deprotections, giving [18F]1 in an 8% n.d.c. radiochemical yield.
Again, purification was carried out using reverse-phase semi-pre-
parative HPLC, and the radiochemical purity was >95% as deter-
mined by both radio-HPLC and radio-TLC, which afforded a
product with high radioactivity concentration for in vivo studies
in the mouse (Fig. 2).
The human breast cancer cell line, MDA-MB-231, previously
used with [18F]2 was selected for in vivo evaluation.4 All in vivo
studies were done in accordance with relevant national guide-
lines.12 The absolute uptake (%ID/mL of tissue) and tracer kinetics
of [18F]1 and [18F]2 were determined by PET. PET images of tumor
5
chemical yield of 8% n.d.c. was achieved.
Radiolabelling to give [18F]1 was achieved in an analagous man-
ner to that for synthesising [18F]2 (Scheme 2). Precursor 4 was sub-
jected to similar conditions to that of precursor 3, with an initial
nucleophilic fluorination followed by sequential acid and base
Figure 2. Semi-preparative Radio-HPLC of [18F]1.

6150
L. Carroll et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6148–6150
bearing mice (cumulative images of 30–60 min dynamic data;
Fig. 3A and 3B) showed distribution of [18F]1 in most tissues with
the highest activity occurring in bladder and heart (blood pool).
Radiotracer uptake increased rapidly over ꢀ10 min in tumors
and continued to increase more slowly over the 60 min period
for [18F]1, but plateaued after ꢀ10 min for [18F]2 (Fig. 3C). Hence,
the uptake of [18F]1 was higher than [18F]2 at 60 min post-radio-
tracer injection (Fig. 3D: p =0.03 (⁄)). Gamma counting of tissues
after the imaging study showed significantly higher tumor/mus-
cle ratio for [18F]1 but unremarkable tumor/blood ratio (Fig. 3E).
We report the radiosynthesis of [18F]1 and comparison of its
biological properties with [18F]2 previously reported by Vasdev
et al.5,6 Radiolabelling of [18F]1 could be achieved using a similar
method previously reported for [18F]2. Radiochemical yield, radio-
chemical purity and radioactivity concentration of both radiotra-
cers were adequate for in vivo injection. The initial tumor uptake
of [18F]1 and [18F]2 (0—ꢀ10 min (Fig. 3C)) were similar. This possi-
bly means that both radiotracers enter the cells in an analogous
fashion. We have developed a small molecule inositol radiotracer
that allows tumor visualization in vivo. Previous studies by McL-
arty and co-workers showed similar levels of uptake between
in tumor/blood ratio suggests that the higher tumor radioactivity
levels for [18F]1 may be due in part to slower clearance of [18F]1
from blood compared to [18F]2. Hepatic metabolism is unlikely to
be the reason for the higher tumor levels as liver/blood ratios were
unremarkable (60 min gamma counting for [18F]1 and [18F]2 were,
‘respectively’, 1.7 0.2 and 1.5 0.1; P>0.05). Future studies would
include plasma and tissue metabolite analysis to rule out the pos-
sibility that the higher tumor levels detected by imaging are due in
part to the contribution of labeled metabolites.
[
18F]1 was synthesized in adequate yields. There was higher
levels of [18F]1-derived radioactivity in MDA-MB-231 tumors com-
pared to [18F]2 making [18F]1 an interesting candidate for further
evaluation as an inositol radiotracer.
Acknowledgments
Funding provided by CR-UK&EPSRC Cancer Imaging Centre at
Imperial College London, in association with the MRC and Depart-
ment of Health (England) Grant C2536/A10337 and UK Medical
Research Council core funding grant U.1200.02.005.00001.01. We
would also like to thank Dr. Mark Nitz for providing a sample of
compound 3.
[
18F]2 and [18F]2-2-fluoro-2-deoxy-D-glucose in MDA-MB-231 tu-
mors, while uptake of the [18F]2 in inflammation was lower.6 Given
the higher tissue abundance of myo- compared to scyllo-inositol, it
was not altogether surprising to see higher tumor uptake of [18F]1
compared to [18F]2. The quantitative higher tumor localization of
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
08.022.
[
18F]1 measured by small animal PET imaging (Fig. 3D) makes this
radiotracer interesting as a candidate radiotracer for imaging ino-
sitol uptake.
[3H]Myo-inositol has been previously used to study inositol flux
into cancer cells.13 These early studies demonstrated that labeled
inositols are rapidly transported into cells where they exist within
the acid-soluble pool or within phospholipids. Scyllo-inositol inhib-
its the cellular uptake of [3H]myo-inositol suggesting similar mech-
anisms of uptake.13 The ability to measure the flux of inositols into
cellular macromolecules will allow important biological processes
in cancer to be studied as myo-Inositol is involved in insulin signal-
ing, and signaling via the epidermal growth factor receptor
(EGF).3,4 Although the imaging studies (typically 60 min for most
radiotracers) demonstrated tumor uptake, we also saw high uptake
in some normal tissues including the kidneys and heart. Given the
continued increase in uptake over time, future studies of [18F]1
should explore longer timepoints to assess whether tumor to back-
ground signal will improve over time as well as comparison with
other radiotracers that measure glucose metabolism (2-deoxy-2-
References and notes
1. Calker, D. v.; Belmaker, R. H. Bipolar Disord. 2000, 2, 102.
2. Uldry, M.; Ibberson, M.; Horisberger, J. D.; Chatton, J.-Y.; Riederer, B. M.;
Thorens, B. EMBO J. 2001, 20, 4467.
3. Potter, B. V. L.; Lampe, D. Angew. Chem., Int. Ed. Engl. 1933, 1995, 34.
4. Sawyer, S. T.; Cohen, S. Biochemistry 1981, 20, 6280.
5. Vasdev, N.; Chio, J.; Oosten, E. M. V.; Nitz, M.; McLaurin, J.; Vines, D. C.; Houle,
S.; Reilly, R. M.; Wilson, A. A. Chem. Commun. 2009, 5527.
6. McLarty, K.; Moran, M. D.; Scollard, D. A.; Chan, C.; Sabha, N.; Mukherjee, J.;
Guha, A.; McLaurin, J.; Nitz, M.; Houle, S.; Wilson, A. A.; Reilly, R. M.; Vasdev, N.
Nucl. Med. Biol. 2011, 38, 953.
7. Sun, Y.; Zhang, G.; Hawkes, A.; Shaw, J. E.; McLaurin, J.; Nitz, M. Bioorg. Med.
Chem. Lett. 2008, 16, 7177.
8. Lowe, G.; McPhee, F. J. Chem. Soc., Perkin Trans. 1 1991, 1249.
9. Ametamey, S. M.; Honer, M.; Schubiger, P. A. Chem. Rev. 2008, 108, 1501.
10. Bejot, R.; Fowler, T.; Carroll, L.; Boldon, S.; Moore, J. E.; Declerck, J.; Gouverneur,
V. Angew. Chem., Int. Ed. 2009, 48, 586.
11. Lee, E.; Kamlet, A. S.; Powers, D. C.; Neumann, C. N.; Boursalian, G. B.; Furuya,
T.; Choi, D. C.; Hooker, J. M.; Ritter, T. Science 2011, 334, 639.
12. Workman, P.; Aboagye, E. O.; Balkwill, F.; Balmain, A.; Bruder, G.; Chaplin, D. J.;
Double, J. A.; Everitt, J.; Farningham, D. A. H.; Glennie, M. J.; Kelland, L. R.;
Robinson, V.; Stratford, I. J.; Tozer, G. M.; Watson, S.; Wedge, S. R.; Eccles, S. A.
Br. J. Cancer. 2010, 102, 1555.
[
18F]fluoro-deoxyglucose) and choline metabolism ([18F]fluorocho-
line) to establish comparative sensitivity for tumor detection. The
higher tumor/muscle ratio for [18F]1 compared to [18F]2 is was in
keeping with the imaging data. The lack of a significant difference
13. Moyer, J. D.; Malinowski, N.; Napier, E. A.; Strong, J. Biochem. J. 1988, 254, 95.