Structure dependence of long-chain [18F]fluorothia fatty acids as myocardial fatty acid oxidation probes

Article abstract of DOI:10.1021/jm301345v

In vivo imaging of regional fatty acid oxidation (FAO) rates would have considerable potential for evaluation of mammalian diseases. We have synthesized and evaluated ¹⁸F-labeled thia fatty acid analogues as metabolically trapped FAO probes to understand the effect of chain length, degree of unsaturation, and placement of the thia substituent on myocardial uptake and retention. 18-[¹⁸F]Fluoro-4-thia-(9Z)-octadec-9-enoic acid (3) showed excellent heart/background radioactivity concentration ratios along with highest retention in heart and liver. Pretreatment of rats with the CPT-1 inhibitor, POCA, caused >80% reduction in myocardial uptake of 16-[ ¹⁸F]fluoro-4-thiahexadecanoic acid (2) and 3, indicating high specificity for FAO. In contrast, 18-[¹⁸F]fluoro-4-thiaoctadecanoic acid (4) showed dramatically reduced myocardial uptake and blunted response to POCA. 18-[¹⁸F]Fluoro-6-thiaoctadecanoic acid (5) showed moderate myocardial uptake and no sensitivity of myocardial uptake to POCA. The results demonstrate relationships between structures of ¹⁸F-labeled thia fatty acid and uptake and their utility as FAO probes in various tissues.

Full text of DOI:10.1021/jm301345v

Article
pubs.acs.org/jmc
Structure Dependence of Long-Chain [18F]Fluorothia Fatty Acids as
Myocardial Fatty Acid Oxidation Probes
Mukesh K. Pandey,†,‡ Anthony P. Belanger,‡ Shuyan Wang,‡ and Timothy R. DeGrado*,†,‡
†Department of Radiology, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, United States
‡Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis Street, Boston, Massachusetts
02115, United States
S
* Supporting Information
ABSTRACT: In vivo imaging of regional fatty acid oxidation (FAO) rates would have considerable potential for evaluation of
mammalian diseases. We have synthesized and evaluated 18F-labeled thia fatty acid analogues as metabolically trapped FAO
probes to understand the effect of chain length, degree of unsaturation, and placement of the thia substituent on myocardial
uptake and retention. 18-[18F]Fluoro-4-thia-(9Z)-octadec-9-enoic acid (3) showed excellent heart/background radioactivity
concentration ratios along with highest retention in heart and liver. Pretreatment of rats with the CPT-1 inhibitor, POCA, caused
>80% reduction in myocardial uptake of 16-[18F]fluoro-4-thiahexadecanoic acid (2) and 3, indicating high specificity for FAO. In
contrast, 18-[18F]fluoro-4-thiaoctadecanoic acid (4) showed dramatically reduced myocardial uptake and blunted response to
POCA. 18-[18F]Fluoro-6-thiaoctadecanoic acid (5) showed moderate myocardial uptake and no sensitivity of myocardial uptake
to POCA. The results demonstrate relationships between structures of 18F-labeled thia fatty acid and uptake and their utility as
FAO probes in various tissues.
synthetase, transport into the mitochondrial matrix facilitated
by carnitine palmitoyltransferases I and II (CPT-I, CTP-II),
carnitine−acylcarnitine translocase (CAT), and the four
enzymes responsible for β-oxidation.1 CPT-I present at the
outer mitochondrial membrane often represents the major step
for flux control for FAO of fatty acids.1 Therefore, CPT-I
inhibitors such as POCA (sodium 2-[5-(4-chlorophenyl)-
pentyl]oxirane-2-carboxylate) have been extensively employed
to reduce FAO rates in experimental studies.2
The ideal FAO probe that is based on the metabolic trapping
concept should follow the flux of fatty acids through at least the
initial step(s) of β-oxidation, at which point the tracer becomes
metabolically trapped. The probe should not undergo
esterification into various complex lipids, as it renders the
accumulation of radioactivity less specific to FAO.1,3−5 Over the
decades, our research has been focused on the potential of
radiolabeled thia-substituted long-chain fatty acid analogues as
such metabolically trapped probes of FAO in heart and other
tissues.2,3,6−9 The first molecule developed was [18F]FTHA
INTRODUCTION
■
Fatty acid oxidation (FAO) plays a pivotal role in energy
production in many tissues and is a candidate biomarker for
various disease states in cardiology, endocrinology, neurology,
and oncology. Myocardial FAO is a highly regulated process,
governed by fatty acid supply, competing energy substrates,
energy demand, oxygen supply, and hormonal status among
other factors.1 The potential diagnostic utility of a noninvasive
PET (positron emission tomography) approach to estimate
regional FAO rates in the heart becomes apparent, as cardiac
dysfunction often displays metabolic changes that alter FAO,1
and could be applicable for evaluation of myocardial ischemia
or heart failure. According to the metabolic trapping approach
for radiotracer design, the ideal radiotracer for myocardial FAO
imaging will mimic natural long chain fatty acids in their
transport from the bloodstream into myocytes as well as their
intracellular trafficking to the mitochondrion for oxidation.1
FAO occurs in the mitochondrion matrix through the cyclical
oxidative process of β-oxidation. The overall rate of β-oxidation
is dependent on the flux control exerted by a multitude of
processes, including fatty acid transport across the cell
membrane (CD36/FATP), activation by fatty acyl CoA
Received: September 18, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Structure of thia-substituted fatty acids synthesized and evaluated as metabolically trapped myocardial FAO probes.
Scheme 1. Synthesis of 18-[18F]Fluoro-4-thiaoctadecanoic Acid (4)
Scheme 2. Synthesis of 18-[18F]Fluoro-6-thiaoctadecanoic Acid (5)
(1)2 (Figure 1), a 6-thia analogue that showed excellent
myocardial imaging characteristics . Surprisingly, 1 was found to
be metabolized in heart to protein-bound species, presumably
metabolic byproducts from β-oxidation. However, 1 lacked
specificity to measure FAO in hypoxic conditions because of
maintained myocardial retention of radiotracer.3,8 Furthermore,
metabolite analyses in liver and skeletal muscle tissues revealed
a predominant accumulation of 1 in lipid pools, further limiting
its application as a general mitochondrial FAO probe.2 The
second tracer developed was a 4-thia analogue of palmitic acid,
2 (Figure 1). The 4-thia analogue showed enhanced specificity
to myocardial FAO, correlating with FAO rates of natural fatty
B
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Uptake (% dose kg/g) of 16-[18F]Fluoro-4-thiahexadecanoic Acid (2) in Fasted Sprague−Dawley Rats
b
a
b
a
organ/tissue (N = 6)
control 30 min
poca treated 30 min
control 120 min
POCA treated 120 min
c
blood
heart
lung
0.070 0.059
0.0633 0.0077
0.0105 0.0028
0.1069 0.0649
0.0183 0.0072
0.4700 0.1888
0.1014 0.0512
0.0154 0.0056
0.3143 0.1334
0.0109 0.0059
0.0392 0.0094
0.1042 0.0387
0.0472 0.0098
0.4433 0.1201
0.1637 0.0659
0.0578 0.0114
0.6549 0.1112
0.0298 0.0069
d
e
0.589 0.257
0.099 0.140
0.899 0.314
0.209 0.144
0.067 0.049
0.235 0.168
0.031 0.013
0.0792 0.0174
c
0.0676 0.0126
0.6197 0.0949
0.2477 0.0310
0.04879 0.0142
0.1790 0.0758
liver
kidney
brain
bone
muscle
d
c
c
c
d
0.0333 0.0080
a
b
c
d
Pretreated with POCA (30 mg/kg ip) 2 h prior. Data from ref 10. p < 0.05 for 120 min control vs 120 min POCA. p < 0.05 for 30 min control
e
vs 120 min control. p < 0.01 for 30 min control vs 30 min POCA.
acids in hypoxic conditions.6,7 However, uptake of 2 in liver was
not sensitive to CPT-I inhibition, and there was poor retention
of the “metabolically trapped” radioactivity in rat myocardium
over a 2 h period.3 Subsequently, a 4-thia substituted oleic acid
analogue was developed as 310 (Figure 1). Compound 3
showed enhanced uptake and retention in rat myocardium
relative to both 1 and 2.10 Pharmacologic inhibition of CPT-1
with etomoxir evoked >80% decrease in myocardial uptake,
demonstrating excellent sensitivity to change in myocardial
FAO.10 The differences in the biodistribution properties of
compounds 1−3 raised the questions: What aspects of the
structure of 3 were responsible for its favorable myocardial
uptake properties, and could they be further enhanced? To
begin to answer these questions, two additional molecules were
designed and evaluated as stearic acid analogues: 18-[18F]-
fluoro-4-thiaoctadecanoic acid (18:0) (4) and 18-[18F]fluoro-6-
thiaoctadecanoic acid (18:0) (5) (Figure 1). The in vivo
characteristics of 4 and 5 were examined in comparison with
those of the previously developed 2 and 3, since these two
radiotracers fill the structural gap between the 2 and 3 to
investigate the effects of chain length (2 versus 4), thia position
(4 versus 5), and saturation (3 versus 4). Our present study
provides considerable information on the structural depend-
ence of thia-substituted fatty acid radiotracer uptake and
retention in myocardium and provides a stronger basis for
rational design of PET imaging probes for FAO.
of 9 was characterized by the appearance of a doublet of a
triplet (dt) at δ 4.46 ppm for two protons due to the formation
of the new carbon−fluorine bond (CH2F) and the absence of a
triplet peak at δ 3.41 ppm for two protons (CH2Br). The
formation of cold standard 9 was also confirmed and supported
by a doublet peak at δ 83.7−84.8 ppm due to fluorine−carbon
coupling in its carbon NMR. HRMS confirmed the product
formation with a molecular ion peak at 320.217 55 amu. The
radiofluorination of the bromo ester 8 using cyclotron-
produced K18F was carried out under microwave heating (75
°C) as previously described.12 Radiofluorination proceeded
efficiently, with >90% incorporation of [18F]fluoride into the
corresponding ester. The [18F]fluoro ester was quantitatively
hydrolyzed by addition of 0.2 N KOH and microwave heating
(75 °C) for 2 min. Overall radiochemical yields after
semipreparative HPLC isolation (tR = 9.4 min) were 30−33%
decay-corrected. Radiochemical purity of the product 4 was
>99% by radio-HPLC. The product was formulated in 0.5−
1.0% BSA (bovine serum albumin) in saline solution and
passed through a sterile filter before administration to rats.
The synthesis of the labeling precursor for 5 required a
modified approach because of the unavailability of methyl 6-
mercaptopentanoate. The methyl 6-mercaptopentanoate (10)
was obtained on reaction between methyl 5-bromopentanoate
and thiourea followed by treatment with ethanolic NaOH and
subsequent esterification with absolute methanol.13 Compound
10 was treated with 1,12-dibromododecane (11) in the
presence of sodium hydride in THF, forming methyl 18-
bromo-6-thiaoctadecanoate (12) as a precursor for radio-
fluorination. Compound 12 was characterized by the proton
NMR presence of a triplet at δ 3.42 ppm for two protons
(CH2Br) and another multiplet at δ 2.52−2.50 ppm for four
protons due to the formation of a thioether (CH2-S-CH2). The
product formation was further supported by carbon NMR and
HR(EI) mass spectrometry. The bromo ester 12 was also
treated with tetrabutylammonium fluoride in acetonitrile
followed by ester hydrolysis to obtain nonradioactive standard
13 as nonradioactive HPLC standard. Compound 13 was
confirmed by proton and carbon NMR and HRMS. The
radiofluorination, hydrolysis, and formulation were carried out
in a similar manner as with 4. The radiochemical yield of the
labeled product 5 after semipreparative HPLC isolation (tR =
8.7 min) was 30−33% decay-corrected. Radiochemical purity of
the product (5) by r-HPLC was >99%.
RESULTS
■
Synthesis of 18-[18F]Fluoro-4-thiaoctadecanoic Acid
(4) and 18-[18F]Fluoro-6-thiaoctadecanoic Acid (5).
Schemes 1 and 2 describe the synthesis of 4 and 5, respectively.
To synthesize the labeling precursor for 4, 1,14-tetradecadioic
acid was converted to 1,14-tetradecane diol (6) in two steps:
esterification using absolute methanol and catalytic amount of
sulfuric acid followed by LAH (lithium aluminum hydride)
reduction in THF (tetrahydrofuran).11 The obtained diol was
refluxed with 48% aqueous HBr to yield 1,14-dibromotetrade-
cane (7), which on subsequent reaction with methyl 3-
mercaptopropanoate in the presence of sodium hydride in THF
yielded methyl 18-bromo-4-thiaoctadecanoate (8) as precursor
for radiofluorination. The formation of compound 8 was
1
characterized by the presence of two triplets peaks in the H
NMR spectrum, one at δ 3.41 ppm for two protons (CH2Br)
and another at δ 2.60 ppm for two protons (CH2S). The
product formation was further confirmed by carbon NMR,
gradient HSQC, and HR(EI) mass spectrometry. The bromo
ester 8 was treated with tetrabutylammonium fluoride in
acetonitrile followed by ester hydrolysis to obtain non-
radioactive standard (9) as a HPLC standard. The formation
Biodistribution Studies in Rats. Tables 1−4 show the
biodistribution data for 18F-radioactivity in rats after intra-
venous administration of 2−5 at 30 and 120 min postinjection
(pi). To indicate the sensitivity of the biodisposition of the fatty
acid analogues to FAO impairment, data were obtained with
C
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. Uptake (% dose kg/g) of 18-[18F]Fluoro-4-thiaoctadecanoic Acid (4) in Fasted Sprague−Dawley Rats
a
a
organ/tissue (N = 6)
control 30 min
POCA treated 30 min
control 120 min
POCA treated 120 min
b
c
blood
heart
lung
0.0392 0.0063
0.3140 0.0738
0.0755 0.0168
0.6292 0.1114
0.1872 0.0248
0.0200 0.0037
0.2309 0.0588
0.0303 0.0061
0.0510 0.0026
0.0098 0.0006
0.2144 0.1017
0.0339 0.0122
0.3759 0.0655
0.0726 0.0166
0.0143 0.0015
0.2934 0.0826
0.0181 0.0070
0.0125 0.0033
0.1025 0.0260
0.0244 0.0092
0.1341 0.0686
0.0947 0.0350
0.0310 0.0111
0.4042 0.0612
b
c
c
c
c
0.1814 0.0276
0.0998 0.0245
b
d
liver
0.4064 0.0787
b
kidney
brain
bone
muscle
0.2350 0.0199
b
d
d
0.0385 0.0033
0.2289 0.0332
b
c
0.0443 0.0047
0.0197 0.0060
a
b
c
d
Pretreated with POCA (30 mg/kg i.p.) 2 h prior. p < 0.05 for 30 m control vs 30 m POCA. p < 0.05 for 30 m control vs 120 m control. p < 0.05
for 120 m control vs 120 m POCA.
Table 3. Uptake (% dose kg/g) of 18-[18F]Fluoro-6-thiaoctadecanoic Acid (5) in Fasted Sprague−Dawley Rats
a
a
organ/tissue (N = 6)
control 30 min
POCA treated 30 min
control 120 min
POCA treated 120 min
b
blood
heart
lung
0.0290 0.0120
0.4413 0.1749
0.0621 0.0180
1.1260 0.5112
0.2095 0.0734
0.0330 0.0223
0.1171 0.0493
0.0390 0.0268
0.0510 0.0190
0.0241 0.0137
0.2815 0.0895
0.0702 0.0118
0.7548 0.1403
0.1629 0.0587
0.0220 0.0041
0.0271 0.0038
0.2288 0.1051
0.0536 0.0090
0.3331 0.1210
0.1163 0.0421
0.8492 0.3636
0.3441 0.1484
0.0879 0.0479
0.1037 0.0297
0.0639 0.0242
d
liver
0.4190 0.0941
kidney
brain
bone
muscle
0.1554 0.0249
d
0.0519 0.0124
c
0.2467 0.0766
0.2104 0.0494
0.0599 0.0127
0.0263 0.0063
a
b
c
d
Pretreated with POCA (30 mg/kg ip) 2 h prior. p < 0.05 for 30 min control vs 30 min POCA. p < 0.05 for 30 min control vs 120 min control. p
< 0.05 for 120 min control vs 120 min POCA.
Table 4. Uptake (% dose kg/g) of 18-[18F]Fluoro-4-thia-(9Z)-octadec-9-enoic Acid (3) in Fasted Sprague−Dawley Rats
ab
,
ab
,
organ/tissue (N = 5)
control 30 min
POCA treated 30 min
control 120 min
POCA treated 120 min
d
blood
heart
lung
0.0441 0.0121
0.7028 0.3019
0.0496 0.0096
1.1847 0.1550
0.2035 0.0510
0.0206 0.0057
0.1531 0.0796
0.0201 0.007
0.0606 0.0096
0.0185 0.005
0.0141 0.0098
0.0760 0.0360
0.0179 0.0109
c
0.1155 0.0211
0.8164 0.2249
d
0.0458 0.0061
0.0228 0.005
c
d
e
liver
0.8162 0.1420
0.8488 0.2232
0.1244 0.0310
0.0215 0.0051
0.4610 0.4140
0.0171 0.0024
0.1666 0.0706
e
kidney
brain
bone
muscle
0.1932 0.0173
0.0611 0.0222
c
0.0368 0.0044
0.0206 0.009
0.2990 0.1645
0.0216 0.0137
0.1475 0.0355
c
0.0297 0.0039
a
b
c
d
Pretreated with POCA (30 mg/kg ip) 2 h prior. Reference 10. p < 0.05 for 30 min control vs 30 min POCA. p < 0.05 for 30 min control vs 120
e
min control. p < 0.05 for 120 min control vs 120 min POCA.
and without pretreatment of rats with the CPT-1 inhibitor
POCA (sodium 2-[5-(4-chlorophenyl)pentyl]oxirane-2-carbox-
ylate).2,14
76.9 10.8% of radioactivity bound to protein pellet in control
hearts compared to 24.5 5.4% in POCA hearts (Table 5).
Table 2 summarizes the biodistribution data for 4 (18:0). It
showed 1.7 fold higher (p < 0.05) heart uptake of tracer at 30
min control rats than 30 min POCA treated rat, indicating
considerable sensitivity of 4 toward FAO. More importantly, 4
showed longer retention in heart over time as indicated by 120
min control uptake studies. The Folch type extraction of heart
tissues treated with 4 showed 78.7 3.5% radioactivity bound
with protein pellet at 30 min control, supporting sensitivity of
the tracer toward FAO. On comparison with 2, 4 had 1.8-fold
lower myocardial uptake at 30 min control but longer retention
in the heart. The longer retention was also supported by
metabolic studies where it showed 71.5 5.5% of radioactivity
bound to protein pellet at 120 min control.
Table 3 summarizes the biodistribution data for 5 where the
sulfur atom replaced the sixth carbon in the stearate analogue.
CPT-I inhibition by POCA had no effect on myocardial uptake
at either time points, indicating poor sensitivity to indicate
mitochondrial FAO. Myocardial washout was slow over the
30−120 min period. The retained radioactivity was primarily in
the form of protein-bound metabolites in control hearts but in
The palmitate analogue 2 showed high myocardial uptake at
30 min (0.589 0.257% ID kg/g, Table 1). However, there
was >80% washout of radioactivity from control hearts from 30
to 120 min. The pretreatment of rats with POCA markedly
decreased (>90%) myocardial uptake at 30 min, indicating very
high sensitivity of 2 toward FAO. The effect of POCA was not
observed at 120 min because of clearance of 2 from control
hearts. The metabolic fate of 2 studied by Folch type extraction
of heart tissues (Table 5) showed 58.8
19.2% of activity
bound to the protein pellet at 30 min control compared to 32.8
6.1% of activity with POCA treated rats at 30 min.
Radioactivity in the protein pellet was interpreted as the
fraction of radiotracer that had undergone β-oxidation.2,3,5,10
Inhibition of β-oxidation by POCA results in reduced
radioactivity bound to protein and increased radioactivity in
organic soluble metabolites (nonmetabolized fatty acid or
complex lipids) or to a lesser extent aqueous metabolites.5
Therefore, these data further supported the sensitivity of 2 to
FAO. This effect was more pronounced at 120 min, showing
D
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
the organic phase of Folch extractions in POCA hearts (Table
5). Similar results were also obtained at the 120 min time point.
On comparison to 4, 5 had similar uptake and retention in
heart tissue at 30 and 120 min controls but uptake of 5 by the
heart was found to be less sensitive to POCA treatment.
Table 4 provides the biodistribution data for the oleate
analogue 3. Compound 3 showed the highest heart uptake of
the four tracers. Myocardial uptake of 3 in control animals was
statistically unchanged from 30 to 120 min, demonstrating
effective trapping in heart tissue. On treatment with POCA, the
heart uptake dramatically decreased at both 30 min (−84%)
and 120 min (−90%) time points, indicating very high
specificity and sensitivity of 3 for FAO (Figure 2). The
control animals. Bone uptake for the all the four tracers was
comparable at 120 min, suggesting their equivalent defluorina-
tion rates in rat.
Heart/organ ratios for 2−5 at 120 min after administration
of tracers are shown in Figure 4. Heart/blood ratio for 3 was
Figure 4. Comparison of heart/blood, heart/muscle, and heart/liver
ratios for compounds 2−5 in fasted rats at 120 min pi. Heart/blood
and heart/muscle are statistically higher for 3 than for 2, 4, and 5 (p <
0.05). Heart/liver values for 3 are also statistically higher than for 2
and 5 (p < 0.05). The data are derived from Tables 2−4.
Figure 2. Comparison of heart uptake of 2−5 in fasted rats.
Compound 3 uptake at 120 min control rats is statistically higher than
for 2, 4, and 5 (p < 0.05). Compound 3 showed the highest degree of
blockade of heart uptake at both 30 and 120 min time points. The data
are derived from Tables 2−4.
found to be 4.0-fold higher than 2, 1.8-fold higher than 4, and
3.0-fold higher than 5 (p < 0.005). The heart/muscle ratio for 3
was 3.0-, 3.5-, and 4.2-fold higher than 2, 4, and 5, respectively
(p < 0.005). Furthermore, heart/liver ratios for 3 were 3.3- and
2.5-fold higher than 2 and 5, respectively (p < 0.005).
Folch-type extraction of heart tissues after administration of 3
demonstrated the highest levels of protein binding for all
tracers studied (>88%), suggesting the highest metabolic
conversion to protein-bound metabolites. In control animals,
myocardial uptake of 3 was 1.2-, 2.2-, and 1.6-fold higher (p <
0.05) than that of 2, 4, and 5 at 30 min pi (Figure 2). At the
120 min time point, the heart uptake of 3 was 2.6- to 8.0-fold
higher (p < 0.05) than the other tracers (Figure 2).
With regard to the whole-body distribution of the tracers, 2
and 4 showed high uptake in liver, followed by heart, bone,
kidney, and other organs at 30 min in fasted control rats.
Compounds 3 and 5 showed highest uptake in liver, followed
by heart, kidney, bone, and other organs at 30 min control
fasted rats. Furthermore, all four tracers (2−5) showed the
highest uptake in liver at 120 min. However, this trend differed
slightly in the cases of 5 and 3, where the heart showed the
second highest uptake after liver, indicating longer myocardial
retention. Hepatic clearance was greater for 2, 4, and 5 in
comparison with 3 (Figure 3) between 30 and 120 min in
Folch-type extraction analysis of tissues showed that 3 is
metabolized predominantly to protein-bound species in control
hearts (Figure 5). At 120 min, the radioactivity of protein-
bound fraction in heart was significantly higher for 3 (89%)
relative to 2 (77%), 4 (71%), and 5 (67%) (p < 0.05). At 30
min, protein-bound, aqueous, and organic fractions in liver and
skeletal muscle were similar for 2, 4, and 5 (Figure 5).
However, 3 showed up to 71% (p < 0.01) and 59% protein-
bound fractions in liver and skeletal muscle, respectively (Table
5). The protein-bound fractions of 2, 4, and 5 were lower than
those for 3 in liver and skeletal muscle, indicating lower FAO-
dependent metabolic trapping relative to 3 in these organs.
It was also observed that POCA-treated rats had relatively
higher uptake in blood, brain, and muscle. For example, POCA-
treated rats with 2 and 5 showed 2.7-fold (p < 0.005) and 2.2-
fold (p < 0.005) higher muscle uptake at 120 min than control
Figure 3. Comparison of liver uptake of 2−5 in fasted rats. Compounds 3 and 4 showed the highest degree of blockade of liver uptake at 120 m time
point (p < 0.05). The data are derived from Tables 2−4.
E
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. Folch-type extraction analysis of metabolites at 120 min (pi) in control Sprague−Dawley rats. The data are derived from Table 5.
rats, respectively. In contrast, such enhancement with POCA
treatment was not observed for 3 and 4. POCA-treated rats
showed ∼3.0-fold higher fatty acid concentration and reduced
glucose concentration in plasma with respect to control rats
(Table 6), whereas triglyceride and lactate concentrations were
independent of CPT-I inhibition. As a CPT-I inhibitor, POCA
was developed to be a potential antidiabetic drug. It increases
plasma fatty acid levels by lowering the rate of fatty acid
clearance from the blood because of decreased β-oxidation in
liver, heart, and skeletal muscle. Conversely, POCA lowers
blood glucose. The decrease in FAO stimulates glucose
oxidation in liver, heart, and skeletal muscle to maintain energy
status. The higher glucose metabolic rates in these tissues
results in a decrease in plasma glucose levels.
MicroPET Studies in Rats. A preliminary microPET
imaging study was performed with 2−5 in fasted rats (Figure
6). Initial dynamic images showed rapid blood clearance of all
four tracers to low radioactivity concentrations within 2 min pi
(postinjection). Similarly, lung clearance was also very rapid
and gave low background to the heart by 2 min pi. The hepatic
uptake of all four thia fatty acid tracers (2−5) was extremely
high at 5 min pi, followed by slow clearance over time.
Myocardial uptakes of all four PET probes plateaued at about
10 min, but faster clearance was observed with 2, whereas
myocardial clearance was nominal for 4 and 5 but remained
stable for 3. Whole-body images acquired 55−115 min pi
demonstrated superior myocardial imaging abilities of 3 among
the thia fatty acid tracers (Figure 6). The biodistribution data
for all four tracers (Tables 1−4) were consistent with
microPET images (Figure 6). Bone uptake in ribs were seen
with all four radiotracers, indicating in vivo defluorination in the
rat.
documented in the literature.4,19−23 However, the fate of the
radiolabel involves multiple catabolic and anabolic processes
and requires complex kinetic models to derive an estimate for
FAO.4,19−24 All the radiotracers in the current study were
radiolabeled with 18F at the terminal (ω) position, a site known
to undergo metabolic defluorination in liver.25 Although we
have previously shown that the label can be stabilized against
defluorination if placed at the (ω-3) position,25 the synthesis
burden is greatly increased to produce the labeling precursors.
Furthermore, subsequent to our early work with (ω-3) labeled
tracers, we found that ω-labeling provided satisfactorily stability
against defluorination in pigs and humans.3 Defluorination of
ω-labeled F-18 fatty acids in rodents is significant but does not
render the present biodistribution data useless, since bone
uptake is quantitatively a minor contribution of the overall
biodistribution.
The present study extends our work on the structure uptake
relationship of thia-substituted fatty acid analogues as
metabolically trapped probes for PET imaging of FAO.2,3,10
Thia-substituted FA analogues are designed to undergo
metabolic trapping in myocardium after commitment of the
FA toward mitochondrial FAO.5 β-Oxidation of ω-[18F]fluoro
and 4- and 6-thia-fatty acid analogues give rise to protein-bound
metabolites in the myocardium, putatively explained by β-
oxidation dependent decomposition of the thia fatty acid
analogues to ω-[18F]fluoro long-chain thiols that covalently
react with mitochondrial proteins.2,3 The sulfur heteroatom
mimics the methylene group of normal fatty acids, although it
subtly increases the local polarity. The sulfur atom was initially
inserted at the sixth position in the first thia fatty acid
radiotracer 1, showing sensitivity toward CPT-I inhibition in
the myocardium in a mouse model.2 However, further studies
indicated FAO-independent mechanisms for retention of
radiotracer in hypoxic myocardium.2 Later, other positions of
thia substitutions were investigated and it was found that sulfur
substitution at the fourth position of the palmitate analogue 2
gave the best correlation with FAO rates in isolated rat hearts in
both normoxic and hypoxic conditions.3 Although 2 showed
higher specificity for FAO, it showed suboptimal retention in
rat myocardium.
DISCUSSION
■
Fatty acid oxidation (FAO) is the key energy producer in heart,
liver, and skeletal muscle. Furthermore, it has recently been
investigated that the brown adipose tissue plays a vital role in
thermogenesis during cold exposure via oxidative metabolism
of fat in human adults.15 A FAO-specific PET imaging
technique has a considerable potential to improve existing
medical care by determining disease state and progression and
also for evaluation of potential drug therapies targeted to alter
FAO rates.16−18 Several non-thia fatty acid analogues of
palmitic acid labeled with 18F and 11C were studied and
Therefore, we further explored the structure of the natural
FAO substrate oleic acid to form the 4-thia oleate analogue 3.10
The monounsaturated 3 (18:1) showed high myocardial uptake
and longer myocardial retention measured at 120 min in rat10
F
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 5. Distribution of 18F-Radioactivity (% of Total) in Aqueous, Organic, and Pellet Fractions after Folch-Type Extraction of
Tissues
a
a
30 min control
30 min POCA
120 min control
120 min POCA
Compound 2
b
b
b
heart (n = 6)
liver (n = 6)
muscle (n = 6)
aqueous
organic
pellet
2.63 1.00
15.28 3.96
51.90 3.86
32.81 6.12
20.23 4.88
41.68 6.53
38.07 4.09
25.13 2.90
42.23 6.23
32.63 5.99
4.17 1.31
7.53 2.25
b
38.56 19.03
58.80 19.21
13.52 8.30
48.25 8.35
38.21 13.23
23.24 8.37
33.56 13.11
43.18 16.90
18.88 10.79
76.93 10.77
5.32 0.82
67.98 6.82
b
24.47 5.35
b
aqueous
organic
pellet
11.52 1.13
b
31.39 9.80
63.27 9.25
16.17 4.64
34.00 12.3
49.81 16.94
50.00 8.31
b
38.47 8.58
aqueous
organic
pellet
19.16 4.09
48.14 7.73
32.69 5.13
compound 3
b
b
b
b
b
b
b
heart (n = 5)
liver (n = 5)
muscle (n = 5)
aqueous
organic
pellet
2.86 1.55
8.46 0.86
88.67 1.63
8.43 1.10
19.87 3.32
71.69 4.34
18.79 3.83
21.75 4.86
59.12 8.66
10.15 2.55
62.03 7.43
23.80 5.76
16.30 3.66
58.80 4.90
24.88 3.54
24.06 5.81
43.19 4.95
32.74 6.02
1.86 0.37
9.34 0.42
88.78 0.71
5.68 0.19
34.75 12.14
59.56 12.02
9.91 2.52
21.98 4.18
67.95 1.75
4.19 1.04
b
66.66 11.65
29.14 10.80
8.10 2.07
b
aqueous
organic
pellet
b
68.06 3.39
b
23.82 3.75
aqueous
organic
pellet
11.37 1.57
b
b
b
53.25 9.73
b
35.36 11.08
compound 4
7.73 0.62
b
heart (n = 6)
liver (n = 6)
muscle (n = 6)
aqueous
organic
pellet
3.97 1.01
17.29 2.61
78.72 3.49
15.36 1.68
51.66 5.80
32.97 6.71
28.99 3.73
38.77 5.36
32.22 5.24
2.39 0.95
3.39 3.17
b
b
b
b
67.58 3.94
24.67 3.92
31.69 5.95
51.76 7.03
16.53 1.87
25.80 5.01
47.15 2.67
27.04 3.35
26.14 5.50
71.45 5.49
8.18 4.73
74.91 6.52
b
21.68 3.63
aqueous
organic
pellet
10.23 3.80
67.69 13.75
22.07 13.63
9.12 2.18
55.69 8.48
36.13 11.75
11.58 3.65
48.90 7.54
39.51 4.831
b
b
aqueous
organic
pellet
58.25 8.75
32.62 8.20
compound 5
heart (n = 6)
liver (n = 6)
muscle (n = 6)
aqueous
organic
pellet
4.02 1.24
25.81 10.12
70.16 9.05
9.65 3.27
45.61 5.10
44.72 4.54
16.49 2.26
44.55 8.06
3.84 1.31
4.03 1.42
3.30 1.07
b
b
63.99 9.67
28.56 6.62
67.39 6.01
10.15 4.50
35.26 4.02
54.58 5.08
18.18 5.46
39.74 10.78
42.06 9.84
74.20 7.78
b
b
32.16 10.47
10.75 2.01
54.88 6.96
34.35 7.76
22.48 6.91
aqueous
organic
pellet
7.45 2.00
b
64.78 10.98
b
27.76 9.07
b
b
aqueous
organic
pellet
7.70 2.81
9.30 1.56
b
43.43 11.19
48.85 12.63
59.27 6.95
38.95 7.41
31.41 7.23
a
b
Pretreated with POCA (30 mg/kg ip) 2 h prior. p < 0.05 for control vs POCA.
Table 6. Plasma Concentration of Metabolic Substrates in Sprague−Dawley Rats
fatty acid (mM)
glucose (mg/dL)
triglyceride (mg/dL)
lactate (mmol)
control (n = 12)
POCA (n = 6)
p
0.242 0.161
0.905 0.085
4.84 × 10−09
111.660 14.768
66.748 15.945
0.000226
46.484 36.701
47.971 18.730
0.9107
1.797 0.577
1.850 0.523
0.8470
(Tables 4 and 5). To increase our understanding of the
different biochemical behavior of 2 and 3, two additional
radiotracers 4 and 5 were synthesized and allowed for the study
of the effects of chain length (2 (16:0) versus 4 (18:0)),
monounsaturation (4 (18:0) versus 3 (18:1)), and position of
the thia substituent (4-thia 4 versus 6-thia 5). It was observed
that 2 (16:0), 4 (18:0), and 5 (18:0) have almost similar
myocardial uptake at 30 min. However, the stearate analogues
(18:0) showed better retention in the myocardium than 2 at
120 min. Similar to the stearate (18:0) analogues, the
monounsaturated (18:1) 3 also showed excellent myocardial
retention. On comparison of the myocardial kinetics of the
unsaturated oleate analogue 3 to the saturated stearate analogue
4, the presence of monounsaturation resulted in more than
double the myocardial uptake at 30 min with excellent
retention of both radiotracers at 120 min. Furthermore, 3
showed the highest protein-bound fractions of 18F-radioactivity
in heart homogenates and markedly higher sensitivity to POCA
induced inhibition of myocardial uptake and formation of
protein-bound metabolites in the heart. Thus, our findings
G
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 6. Representative microPET images of 2−5 at 55−115 min pi. Images are presented as maximum intensity projections (MIPs) of rats in
coronal orientation with snout up. Accumulation of the radiotracers is visualized in heart, liver, and ribs (indicating defluorination).
suggest that monounsaturation enhances FAO-dependent
transport and intracellular trafficking of 3 to the mitochondrion
for oxidative metabolic trapping. The inhibition of myocardial
uptake and formation of protein-bound metabolites of FTO
(and to a lesser extent the other FA analogues) with CPT-I
inhibition (POCA effect) is a strong indication of mitochon-
drial trafficking. The significance of this fact is that FTO may
serve as a noninvasive biomarker of the flux of fatty acids to the
mitochondrion for β-oxidation. We had confirmed in an earlier
work with FTP that the tracer had localized to the
mitochondrial fraction by ultracentrifugation.3 Extending the
chain length of the palmitate analogue (2) by two carbons to
form 4 results in similar uptake and metabolic trapping
properties but lengthened myocardial retention, suggesting
longer retention of the putative radiolabeled long-chain thiol
metabolite for 4.
Thia substition at the fourth carbon appears to optimize
specificity of metabolic trapping to FAO. FAO-dependent
metabolism of the 4-thia analogues to protein-bound
metabolites in heart was strongly indicated from the finding
that pretreatment of rats with the CPT-I inhibitor POCA
resulted in dramatic reductions of protein-bound fractions from
70−89% to 20−25%. To further clarify the nature of the
protein binding binding of metabolites, a gel electrophoresis
experiment was performed with heart and liver homogenates
using 2 and 3. 18F-Radioactivity was found to be distributed
among at least five different molecular weight proteins across
the gels and not specific to one particular protein (data not
shown). This finding is consistent with the previously reported
mechanism of metabolism of thia fatty acids to long-chain thiol
fragments which continues to bear the 18F-radiolabel and
nonspecifically bind to mitochondrial protein by formation of
disulfide bonds through various cysteine residues.5 Interest-
ingly, the 6-thia fatty acid analogue 5 showed no significant
change in myocardial uptake in response to CPT-1 inhibition
(Table 3), although protein-bound fractions of 18F radioactivity
in heart homogenates were decreased in a manner similar to the
4-thia fatty acid analogues. It appears that 5 was taken up by
POCA treated hearts by an amount near control levels despite
the inhibition of FAO at the CPT-1 step. This behavior differs
from that previously seen for 1 in POCA-treated mouse hearts
where myocardial uptake was reduced by 87% (p < 5 × 10−4) at
60 min2. It is not clear whether this discrepancy is caused by the
difference in fatty acid length (C-18 versus C-17), placement of
the 18F (ω versus ω-3), or species differences (rat versus
mouse). Profound differences in metabolism have been
observed for fatty acids differing by as little as one or two
methylene groups.2,3,5,10
CONCLUSIONS
■
Several long-chain, thia-substituted fatty acid analogues were
synthesized and preliminarily evaluated in rats as PET probes
for myocardial fatty acid oxidation (FAO). The radiosyntheses
of 4 and 5 were developed in analogy to those previously
developed for 2 and 3. Biodistribution and PET imaging studies
of the tracers were performed to establish a plausible structure−
uptake relationship. The enhanced myocardial uptake and
metabolic trapping of 3 are primarily due to the presence of
monounsaturation, whereas improved retention in myocardium
can be attributed to the longer chain length. Myocardial uptake
and metabolism of 4-thia stearate analogues appear to be more
specific to mitochondrial CPT-1 mediated trafficking to the
mitochondrion than the 6-thia analogue 5. Preliminary images
of the accumulation of compound 3 in the rat myocardium
were clearly superior to those of 2, 4, and 5. In summary, 3
remains the most specific and sensitive FAO probe of the 4-thia
fatty acid analogues evaluated and warrants further preclinical
evaluation.
EXPERIMENTAL SECTION
■
Materials and Methods. All chemicals and solvents were
purchased from Sigma-Aldrich and used as received. Anhydrous
solvents were also purchased from Sigma-Aldrich in “Sure/Seal”
bottles. TLC analysis of reaction mixtures and products was performed
on Merck silica gel 60 F254 TLC plates. Liquid chromatography was
carried out on Merck 60 silica gel (32−63 μm). H and 13C NMR
1
results were recorded on a Varian 600 MHz spectrometer and
referenced to CDCl3 (1H NMR CDCl3 at 7.26 ppm and 13C NMR
CDCl3 at 77.0 ppm). Mass spectral data were obtained from the Mass
Spectral Lab of School of Chemical Sciences, University of Illinois,
Urbana, IL. Analytical HPLC was performed on a Phenomenex Luna
C-18 column (5 μm, 4.6 mm × 250 mm) with a flow rate of 0.8 mL/
min. Semipreparative HPLC was performed on the final 18F labeled
product using a Phenomenex Luna C-18 column (5 μm, 10 mm × 250
mm) with a flow rate of 5.0 mL/min. POCA (sodium 2-[5-(4-
chlorophenyl)pentyl]oxirane-2-carboxylate) was obtained as a gen-
erous gift from Dr. S. John Gatley, Department of Pharmaceutical
Sciences, Northeastern University, Boston, MA. The “5P30 DK
36836” specialized assay core facility of Joslin DERC, Boston, MA, was
used to estimate the plasma concentration of metabolic substrates like
glucose, triglyceride, free fatty acid, and lactate. All these substrate
were measured by standard enzymatic laboratory assay.
Synthesis of Methyl 18-Bromo-4-thiaoctadecanoate (8).
Synthesis of methyl 18-bromo-4-thiaoctadecanoate (8) was achieved
by coupling methyl 3-mercaptopropionate with 1,14-dibromotetrade-
cane (7) using sodium hydride. To a stirred solution of methyl 3-
mercaptopropanoate (500 mg, 4.16 mmol) in dry THF (100 mL)
under nitrogen at 0 °C was added NaH (100 mg, 4.16 mmol). The
resulting solution was stirred for 15 min at 0 °C. A solution of 7 (1.32
g, 3.70 mmol) in THF (10 mL) was then gradually added, and the
H
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
resulting mixture was maintained at 0 °C for 30 min. The mixture was
stirred for 3 h at room temperature and quenched with water. After
extraction with the dichloromethane, the organic layer was dried over
Na2SO4 and solvent was evaporated under vacuum. The residue was
subjected to column chromatography (5:95 ethyl acetate/hexane) to
over anhydrous Na2SO4, and filtered. The filtrate was evaporated
under vacuum to dryness. The obtained fluoro ester residue was
further subjected to hydrolysis using 0.5 mL of aqueous solution of
KOH (10 mg, 0.18 mmol) and stirred at room temperature for 2 h.
The solvent was evaporated under vacuum. The obtained residue was
dissolved in 25 mL of dichloromethane, washed with 1 N HCl, dried
over Na2SO4, and filtered. The filtrate was evaporated under vacuum
and purified by column chromatography using 4:96 methanol/
chloroform as solvent and silica gel as an adsorbent to obtain
1
yield 8 (1.30 g, 79% yield, mp 37 1 °C) as a white solid. H NMR
(23 °C, 599.77 MHz, CDCl3) δ ppm: 3.70 (s, 3H), 3.41 (t, 2H, J1,2
=
12.0 Hz), 2.78 (t, 2H, J1,2 = 12.0 Hz), 2.60 (t, 2H, J1,2 = 12.0 Hz), 2.52
(t, 2H, J1,2 = 12.0 Hz), 1.88−1.82 (m, 2H), 1.58−1.53 (m, 2H), 1.44−
1.40 (m, 2H), 1.38−1.34 (m, 2H), 1.32−1.22 (brs, 16H). 13C NMR
(23 °C, 150.81 MHz, CDCl3) δ ppm: 172.6 (COOCH3), 51.9
(OCH3), 34.7, 34.1, 32.9, 32.2, 29.6−29.4 (CH2x6), 29.3, 29.2, 28.9,
28.7, 28.1, 26.9. HRMS (ES) calcd for C18H35O2SBr(M+) 394.154 11,
found 394.152 27 and (M + 2) at 396.0 (due to isotopic abundance).
Synthesis of 18-Fluoro-4-thiaoctadecanoic Acid (9). An
anhydrous acetonitrile solution (3 mL) of bromo ester 8 (80 mg,
0.20 mmol) and 1 M THF solution of TBAF (3 mL, 3.0 mmol) was
stirred at room temperature for 72 h under nitrogen (additional 0.5
mL (0.5 mmol) of 1 M THF solution was further added at 24 and 48
h). Solvent was evaporated under vacuum. The resulting solid was
dissolved in 100 mL of dichloromethane, washed with water, dried
over anhydrous Na2SO4, and filtered. The filtrate was evaporated
under vacuum to dryness. The obtained fluoro ester residue was
further subjected to hydrolysis using 0.5 mL of aqueous solution of
KOH (10 mg, 0.18 mmol) and stirred at room temperature for 2 h.
The solvent was evaporated under vacuum. The obtained residue was
dissolved in 25 mL of dichloromethane, washed with 1 N HCl, dried
over Na2SO4, and filtered. The filtrate was evaporated under vacuum
and purified by column chromatography using 4:96 methanol/
chloroform as solvent and silica gel as an adsorbent to obtain
compound 13 as a white solid (38 mg, 61.7% over all yield, mp 51
1
°C). 1H NMR (23 °C, 599.77 MHz, CDCl3) δ ppm: 4.46 (dt, 2H, J =
47.3 Hz, 6.2 Hz), 2.56−2.51 (m, 4H,), 2.41 (t, 2H, J1,2 = 7.3 Hz),
1.79−1.66 (m, 6H), 1.61−1.58 (m, 2H), 1.40−1.38 (m, 4H), 1.33−
o
1.29 (brs, 12H). 13C NMR (23 C, 150.81 MHz, CDCl3) δ ppm:
179.2, 84.8, and 83.7 (d, J = 163.5 Hz, F−CH2 due to carbon−fluorine
coupling), 33.7, 33.2, 32.1, 31.6, 30.4−30.3 (d, J = 19.1 Hz long-range
fluorine carbon coupling), 29.6−28.9 (CH2 × 8), 25.1, 23.9. HRMS
(ES) calcd for C17H33O2SF (M+), 320.2151, found 320.2194.
General Procedure for 18F-Labeling. The labeling precursors for
all the four tracers 2,3,6 3,10 4, and 5 were the methyl esters of their
corresponding bromides (Schemes 1 and 2). Nucleophilic [18F]-
fluorination and subsequent hydrolysis of the ester were performed as
previously described.19 Briefly, cyclotron-produced [18F]fluoride was
dried under nitrogen at 95 °C in a 2 mL glass vial containing Kryptofix
2.2.2 (10 mg), acetonitrile (0.8 mL), and K2CO3 (4 mg) solution in
water (0.15 mL). The residue was further dried by azeotropic
distillation with acetonitrile (2 × 0.5 mL). A solution of the precursor
(approximately 2−3 mg) in acetonitrile (0.5 mL) was added, and the
vial was sealed (screw tight cap) and subjected to microwave
irradiation at 75 °C for 5 min. Subsequent hydrolysis of the resulting
[18F]fluoro ester was performed in the same vessel by the addition of
0.15 mL of 0.2 N KOH followed by microwave irradiation at 75 °C for
2 min. The mixture was then cooled, acidified with acetic acid (0.2 N),
filtered, and injected onto the semipreparative HPLC column
(Phenomenex Luna C-18, 5 μm, 10 mm × 250 mm). The mobile
phase was acetonitrile/water/TFA (90:10:0.005 v/v), and flow rate
was 5 mL/min (tR = 9.4 min for 4 and tR = 8.7 min for 5). An in-line
ultraviolet detector (210 nm) was used to monitor the elution of
unlabeled materials. The [18F]fluoro fatty acid fraction was diluted in
50 mL water and trapped on a C-18 Sep-Pak cartridge (Accel C-18
Plus, Waters, Milford, MA), followed by washing of the Sep-Pak with
10 mL of water. The product was eluted from the Sep-Pak using 0.1
mL of ethanol (multiple fractions). The fraction having the highest
activity was used for formulation in 0.5−1% BSA in isotonic NaCl
solution and filtered through a 0.22 μm filter (Millex-GS, Millipore,
Bedford, MA, U.S.). Radiochemical purity (>99%) was analyzed by
radio-HPLC in the system described above.
compound 9 as a white solid (40 mg, 61.6% overall yield, mp 57
1
°C). 1H NMR (23 °C, 599.77 MHz, CDCl3) δ ppm: 4.46 (dt, 2H, J1 =
47.3 Hz, J2 = 6.2 Hz), 2.82 (t, 2H, J1,2 = 7.3 Hz), 2.70 (t, 2H, J1,2 = 7.3
Hz), 2.57 (t, 2H, J1,2 = 7.3 Hz), 1.74−1.68 (m, 2H), 1.63−1.58 (m,
2H), 1.42−1.38 (m, 4H), 1.32−1.22 (brs, 16H). 13C NMR (23 °C,
150.81 MHz, CDCl3) δ ppm: 176.2, 84.8, and 83.7 (d, J = 163.8 Hz,
F−CH2 due to carbon−fluorine coupling), 34.3, 32.2, 30.4−30.3 (d, J
= 19.6 Hz long-range fluorine carbon coupling), 29.6−29.2 (CH2 × 7),
29.2, 28.8, 26.6, 25.2, 25.1. HRMS (ES) calcd for C17H33O2SF(M+)
320.218 53, found 320.217 55.
Synthesis of Methyl 18-Bromo-6-thiaoctadecanoate (12).
Synthesis of methy18-fluoro-6-thiaoctadecanoate (12) was achieved by
coupling of methyl 6-mercaptopentanoate with 1,12-dibromododecane
(11) using sodium hydride. To a stirred solution of methyl 6-
mercaptopentanoate (500 mg, 3.37 mmol) in anhydrous THF (100
mL) under nitrogen at 0 °C was added NaH (81 mg, 3.37 mmol). The
resulting solution was stirred for an additional 15 min at 0 °C. A
solution of 11 (990 mg, 3.09 mmol) in THF (10 mL) was slowly
added, and the resulting mixture was maintained at 0 °C for 30 min.
The mixture was then stirred for 3 h at room temperature and
quenched with water. The dichloromethane extract was dried over
Na2SO4, filtered, and the filtrate was evaporated under vacuum. The
residue was subjected to column chromatography (5:95 ethyl acetate/
hexane) to yield compound 12 (1.09 g, 2.76 mmol, 82% yield) as a
Biodistribution Studies in Rats. Female Sprague−Dawley rats
were fasted overnight. The rats were anesthetized with isoflurane
(3.5% induction, 1.5% maintenance). The 18F-labeled radiotracer
(0.7−1.4 MBq) was injected into the tail vein. The biodistributions
were performed at either 30 or 120 min before procurement of heart,
liver, lung, blood, kidney, bone (femur), brain, and skeletal muscle.
The tissues were counted for 18F-radioactivity and weighed. All 18F-
radioactivity measurements were corrected for radioactivity decay.
Radiotracer uptake was calculated as
1
colorless oil. H NMR (23 °C, 599.77 MHz, CDCl3) δ ppm: 3.68 (s,
3H), 3.42 (t, 2H, J1,2 = 12.0 Hz), 2.52−2.50 (m, 4H), 2.34 (t, 2H, J1,2
=
12.0 Hz), 1.86−1.82 (m, 2H), 1.76−1.71 (m, 2H), 1.61−1.64 (m,
2H), 1.59−1.54 (m, 4H), 1.44−1.39 (m, 2H), 1.38−1.34 (m, 2H),
1.32−1.22 (brs, 10H). 13C NMR (23 °C, 150.81 MHz, CDCl3) δ
ppm: 174.1 (COOCH3), 51.8 (OCH3), 34.0, 33.6, 32.8, 32.1, 31.6,
29.6−28.7 (CH2 × 9), 28.1, 24.1. HRMS (ES) calcd for C18H35O2SBr-
(M+) 394.154 11, found 394.152 27 and (M + 2) at 395.9 (due to
isotopic abundance)
Synthesis of 18-Fluoro-6-thiaoctadecanoic Acid (13). An
anhydrous acetonitrile solution (3 mL) of bromo ester 12 (76 mg,
0.19 mmol) and 1 M THF solution of TBAF (3 mL, 3.0 mmol) was
stirred at room temperature for 72 h under nitrogen (additional 0.5
mL (0.5 mmol) of 1 M THF solution was further added at 24 and 48
h). Solvent was evaporated under vacuum. The resulting solid was
dissolved in 100 mL of dichloromethane, washed with water, dried
CPM(tissue) × body wt (kg)
tissue wt (g) × CPM(dose)
uptake (% dose kg/g) =
× 100
(1)
In one group of fasted rats, the CPT-I inhibitor POCA (30 mg/kg)
was administered intraperitoneally 120 min prior to 3 injection. Crude
analysis of the nature of the metabolites in the heart, liver, and muscle
was performed by a Folch-type extraction procedure. Approximately
0.5 g of tissue was excised and thoroughly homogenized and sonicated
(20 s) in 7 mL of chloroform/methanol (2:1) at 0 °C. Urea (40%,
1.75 mL) and 5% sulfuric acid (1.75 mL) were added, and the mixture
was sonicated for an additional 20 s. After centrifugation for 10 min at
1800g, aqueous, organic, and protein interphase (pellet) fractions were
I
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
separated and counted for 18F-radioactivity. Furthermore, the three
fractions are obtained by this extraction method, an organic phase
(chloroform/methanol), aqueous phase, and protein-pellet. The
percentage of 18F-radioactivity distributed among the three phases
indicated the following: (1) organic phase contained nonmetabolized
fatty acid analogue and complex lipid metabolites such as triglycerides
and phospholipids; (2) aqueous phase contained F-18 fluoride or
other aqueous soluble metabolites; and (3) the remaining protein-
pellet contained protein-bound 18F-labeled β-oxidation metabolites.
MicroPET/CT Imaging Studies in Rats. Female Sprague−
Dawley rats were fasted overnight. The rats were anesthetized with
isoflurane (3.5% induction, 1.5% maintenance). PET/CT scanning
was performed using an eXplore Vista PET/CT (General Electric
Health Care, Milwaukee, WI). The animal was positioned in the prone
position with the snout positioned into the scanner. Before the
radiotracer imaging study, a whole-body CT scan was performed. This
scan was used for anatomical visualization and for attenuation
correction of the emission data. The radiotracers were studied in
different rats to allow the biodistribution to be done following the PET
scan. The 18F-labeled radiotracer (9−11 MBq) was injected into a tail
vein, and the emission data were collected in dynamic acquisition for
50 min over the upper thorax, followed by a static whole-body scan
from 55 to 115 min postinjection. For 3−5, two to five rats were
imaged with each tracer. In this preliminary study, representative
images are presented for each 18F-labeled fatty acid analogue. The
images were displayed as maximum intensity projections (MIPs) with
coronal orientation. The pixel intensities were normalized to the
highest pixel value within each image. The images were interpreted
strictly qualitatively to show the imaging feasibility of the radiotracers
as myocardial imaging agents and confirmation with the biodistribu-
tion data.
tetrahydrofuran; BSA, bovine serum albumin; TBAF, tetrabu-
tylammonium fluoride; pi, postinjection; amu, atomic mass unit
REFERENCES
■
(1) Lopaschuk, G. D.; Ussher, J. R.; Folmes, C. D.L.; Jaswal, J. S.;
Stanley, W. C. Myocardial fatty acid metabolism in health and disease.
Physiol. Rev. 2010, 90, 207−258.
(2) DeGrado, T. R.; Coenen, H. H.; Stocklin, G. 14(R,S)-
[18F]Fluoro-6-thia-heptadecanoic acid (FTHA): evaluation in mouse
of a new probe of myocardial utilization of long chain fatty acids. J.
Nucl. Med. 1991, 32, 1888−1896.
(3) DeGrado, T. R.; Wang, S.; Holden, J. E.; Nickles, R. J.; Taylor,
M.; Stone, C. K. Synthesis and preliminary evaluation of 18F-labeled 4-
thia palmitate as a PET tracer of myocardial fatty acid oxidation. Nucl.
Med. Biol. 2000, 27, 221−231.
(4) Bergmann, S. R.; Weinheimer, C. J.; Markham, J.; Herrero, P.
Quantitation of myocardial fatty acid metabolism using PET. J. Nucl.
Med. 1996, 37, 1723−1730.
(5) Pandey, M. K.; Bansal, A.; DeGrado, T. R. Fluorine-18 labeled
thia fatty acids for PET imaging of fatty acid oxidation in heart and
cancer. Heart Metab. 2011, 51, 15−19.
(6) DeGrado, T. R.; Kitapci, M. T.; Wang, S.; Ying, J.; Lopaschuk, G.
D. Validation of 18F-fluoro-4-thia-palmitate as a PET probe for
myocardial fatty acid oxidation: effects of hypoxia and composition of
exogenous fatty acids. J. Nucl. Med. 2006, 47, 173−181.
(7) Wallhaus, T. R.; Taylor, M.; DeGrado, T. R.; Russell, D. C.;
Stanko, P.; Nickles, R. J.; Stone, C. K. Myocardial free fatty acid and
glucose use after carvedilol treatment in patients with congestive heart
failure. Circulation 2001, 103, 2441−2446.
(8) Renstrom, B.; Rommelfanger, S.; Stone, C. K.; DeGrado, T. R.;
Carlson, K. J.; Scarbrough, E.; Nickles, R. J.; Liedtke, A. J.; Holden, J.
E. Comparison of fatty acid tracers FTHA and BMIPP during
myocardial ischemia and hypoxia. J. Nucl. Med. 1998, 39, 1684−1689.
(9) Stone, C. K.; Pooley, R. A.; DeGrado, T. R.; Renstrom, B.;
Nickles, R. J.; Nellis, S. H.; Liedtke, A. J.; Holden, J. E. Myocardial
uptake of the fatty acid analog 14-fluorine-18-fluoro-6-thia-heptadeca-
noic acid in comparison to beta-oxidation rates by tritiated palmitate. J.
Nucl. Med. 1998, 39, 1690−1696.
Statistical Analyses. All results were expressed as the mean SD.
Comparison of the mean values was made using a two-tailed Student’s
t test. The acceptable level of significance was set at p < 0.05. Microsoft
Excel was used to perform the statistical analysis.
ASSOCIATED CONTENT
■
S
* Supporting Information
1H NMR spectra of nonradioactive standards and radio-HPLC
chromatograms of compounds used in the present study. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(10) DeGrado, T. R.; Bhattacharyya, F.; Pandey, M. K.; Belanger, A.
P.; Wang, S. Synthesis and preliminary evaluation of 18-[18F]fluoro-4-
thia-oleic acid (FTO) as a PET probe of fatty acid oxidation. J. Nucl.
Med. 2010, 51, 1310−1317.
(11) Jain, S. C.; Kumar, R.; Goswami, R.; Pandey, M. K.; Khurana, S.;
Rohatgi, L.; Gyanda, K. Synthesis of novel non-isoprenoid phenolic
acids and 3-alkylpyridines. Pure Appl. Chem. 2005, 177, 185−193.
(12) Belanger, A. P.; Pandey, M. K.; DeGrado, T. R. Microwave-
assisted radiosynthesis of [18F] fluorinated fatty acid analogs. Nucl Med
Biol 2011, 38, 435−441.
(13) Jung, C. M.; Kraus, W.; Leibnitz, P.; Pietzsch, H. J.; Kropp, J.;
Spies, H. Synthesis and first crystal structure of rhenium complexes
derived from ω-functionalized fatty acids as model compounds of
technetium tracers for myocardium metabolism imaging. Eur. J. Inorg.
Chem. 2002, 1219−1225.
AUTHOR INFORMATION
Corresponding Author
*Phone: (507) 538-4319. Fax: (507) 266-4461. E-mail:
degrado.timothy@mayo.edu.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Institutes of Health
(Grants R01 HL-63371 (T.R.D.) and R01 CA108620
(T.R.D.)) and the Department of Radiology, Brigham and
Women’s Hospital, MA.
(14) DeGrado, T. R.; Holden, J. E.; Ng, C. K.; Raffel, D. M.; Gatley,
S. J. Comparision of 16-iodohexadecanoic acid (IHDA) and 15-p-
iodophenlpentadecanoic acid (IPPA) metabolism and kinetics in the
isolated rat heart. Eur. J. Nucl. Med. 1988, 14, 600−606.
(15) Ouellet, V.; Labbe,
́
S. M.; Blondin, D. P.; Phoenix, S.; Brigitte
ABBREVIATIONS USED
Guerin, B.; Haman, F.; Turcotte, E. E.; Richard, D.; Carpentier, A. C.
́
■
Brown adipose tissue oxidative metabolism contributes to energy
expenditure during acute cold exposure in humans. J. Clin. Invest. 2012,
122 (2), 545−552.
FAO, fatty acid oxidation; CPT-I, carnitine palmitoyltransferase
I; POCA, sodium 2-[5-(4-chlorophenyl)pentyl]oxirane-2-car-
boxylate; PET, positron emission tomography; CD36, cluster of
differentiation 36, which is an integral membrane protein;
FATP, fatty acid transport protein; CAT, carnitine−acylcarni-
tine translocase; 1, 14(R,S)-[18F]fluoro-6-thiaheptadecanoic
acid; (16:0), 16 carbon chain with no double bond; (18:0),
18 carbon chain with no double bond; (18:1), 18 carbon chain
with one double bond; LAH, lithium aluminum hydride; THF,
(16) Folmes, C. D.; Clanachan, A. S.; Lopaschuk, G. D. Fatty acid
oxidation inhibitors in the management of chronic complications of
atherosclerosis. Curr. Atheroscler. Rep. 2005, 7, 63−70.
(17) Di Napoli, P.; Taccardi, A. A.; Barsotti, A. Long term
cardioprotective action of trimetazidine and potential effect on the
inflammatory process in patients with ischaemic dilated cardiomyop-
athy. Heart 2005, 91, 161−165.
J
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(18) Wambolt, R. B.; Lopaschuk, G. D.; Brownsey, R. W.; Allard, M.
F. Dichloroacetate improves postischemic function of hypertrophied
rat hearts. J. Am. Coll. Cardiol. 2000, 36, 1378−1385.
(19) Knust, E. J.; Kupfernagel, C.; Stocklin, G. Long-chain F-18 fatty
acids for the study of regional metabolism in heart and liver; odd-even
effects of metabolism in mice. J. Nucl. Med. 1979, 20, 1170−1175.
(20) Fox, K. A.; Abendschein, D. R.; Ambos, H. D.; Sobel, B. E.;
Bergmann, S. R. Efflux of metabolized and nonmetabolized fatty acid
from canine myocardium. Implications for quantifying myocardial
metabolism tomographically. Circ. Res. 1985, 57, 232−243.
(21) Livni, E.; Elmaleh, D. R.; Barlai-Kovach, M. M.; Goodman, M.
M.; Knapp, F. F., Jr.; Strauss, H. W. Radioiodinated beta-methyl
phenyl fatty acids as potential tracers for myocardial imaging and
metabolism. Eur Heart J. 1985, 6 (Suppl. B), 85−89.
(22) Livni, E.; Elmaleh, D. R.; Levy, S.; Brownell, G. L.; Strauss, W.
H. Beta-methyl[1-11C]heptadecanoic acid: a new myocardial metabolic
tracer for positron emission tomography. J. Nucl. Med. 1982, 23, 169−
175.
(23) DeGrado, T. R.; Holden, J. E.; Ng, C. K.; Raffel, D. M.; Gatley,
S. J. β-Methyl-15-p-iodophenylpentadecanoic acid metabolism and
kinetics in the isolated rat heart. Eur. J. Nucl. Med. 1989, 15, 78−80.
(24) Goodman, M. M.; Knapp, F. F., Jr.; Callahan, A. P.; Ferren, L. A.
Synthesis and biological evaluation of 17-[131I]iodo-9-telluraheptade-
canoic acid, a potential imaging agent. J. Med. Chem. 1982, 25, 613−
618.
(25) Moka, D. C.; DeGrado, T. R. Non-β-oxidizable ω-[18F]fluoro
long chain fatty acid analogs show cytochrome P-450-mediated
defluorination: implications for the design of PET tracers of
myocardial fatty acid utilization. Int. J. Radiat. Appl. Instrum., Part B
1992, 19, 389−397.
K
dx.doi.org/10.1021/jm301345v | J. Med. Chem. XXXX, XXX, XXX−XXX