Design, synthesis, and in vitro testing of α-methylacyl-CoA racemase inhibitors

Article abstract of DOI:10.1021/jm0702377

The enzyme α-methylacyl-CoA racemase (AMACR) is overexpressed in prostate, colon, and other cancers and has been partially validated as a potential therapeutic target by siRNA knockdown of the AMACR gene. Analogs of the natural substrate branched chain α-methylacyl coenzyme A esters, possessing one or more β-fluorine atoms, have been synthesized using Wittig, conjugate addition, and asymmetric aldol reactions and found to be reversible competitive inhibitors. Each diastereomer of the previously reported inhibitor ibuprofenoyl-CoA was also tested. The compounds had K_i values of 0.9-20 μM and are the most potent inhibitors yet known. The presence of β-fluorine on the α-methyl group or the acyl chain results in a significant lowering of the K_i value compared with nonfluorinated analogs, and this is attributed to a lowering of the pK _a of the α-proton, facilitating enolization and binding. Several of the CoA ester inhibitors were formed by incubating the free carboxylic acid precursors with cell free extracts and CoA. α- Trifluoromethyltetradecanoic acid, the precursor to the most potent inhibitor, was shown to inhibit growth of cancer cell lines PC3, CWR22 Rv1, and Du145 in a dose-dependent manner and could be related to the expression level of AMACR.

Full text of DOI:10.1021/jm0702377

2700
J. Med. Chem. 2007, 50, 2700-2707
Design, Synthesis, and In Vitro Testing of r-Methylacyl-CoA Racemase Inhibitors
Andrew J. Carnell,*,† Ian Hale,† Simone Denis,‡ Ronald J. A. Wanders,‡ William B. Isaacs,§ Brice A. Wilson,§ and
Sacha Ferdinandusse‡
Department of Chemistry, UniVersity of LiVerpool, LiVerpool L69 7ZD, United Kingdom, Laboratory Genetic Metabolic Diseases, Academic
Medical Center, Amsterdam, the Netherlands, and Brady Urological Institute, Johns Hopkins Oncology Center, Baltimore, Maryland 21287
ReceiVed March 2, 2007
The enzyme R-methylacyl-CoA racemase (AMACR) is overexpressed in prostate, colon, and other cancers
and has been partially validated as a potential therapeutic target by siRNA knockdown of the AMACR
gene. Analogs of the natural substrate branched chain R-methylacyl coenzyme A esters, possessing one or
more â-fluorine atoms, have been synthesized using Wittig, conjugate addition, and asymmetric aldol reactions
and found to be reversible competitive inhibitors. Each diastereomer of the previously reported inhibitor
ibuprofenoyl-CoA was also tested. The compounds had Ki values of 0.9-20 µM and are the most potent
inhibitors yet known. The presence of â-fluorine on the R-methyl group or the acyl chain results in a significant
lowering of the Ki value compared with nonfluorinated analogs, and this is attributed to a lowering of the
pKa of the R-proton, facilitating enolization and binding. Several of the CoA ester inhibitors were formed
by incubating the free carboxylic acid precursors with cell free extracts and CoA. R-Trifluoromethyltet-
radecanoic acid, the precursor to the most potent inhibitor, was shown to inhibit growth of cancer cell lines
PC3, CWR22 Rv1, and Du145 in a dose-dependent manner and could be related to the expression level of
AMACR.
Introduction
chain fatty acids, are associated with increased prostate cancer
risk.6 Genome-wide scans for linkage in hereditary PCa families
implicate 5p13, the chromosomal region of AMACR, as the
location of the prostate cancer susceptibility gene, and it has
recently been reported that sequence variants of AMACR may
be responsible for increased prostate cancer risk.7 Two specific
mutations, S52P and L107P, in peroxisomal AMACR are
thought to be responsible for adult-onset sensory motor neur-
opathy. Patients with the disease show elevated plasma con-
centrations of pristanic acid and C27-bile acid intermediates.8
Recent data has shown that AMACR is functionally important
in the growth of PCa cells. Small interference RNA (siRNA)
knockdown of the AMACR gene reduced the expression of
AMACR and significantly impaired the proliferation of the
androgen responsive PCa cell line LAPC-4. Simultaneous
inhibition of both the AMACR pathway by siRNA and androgen
signaling by androgen withdrawal or antiandrogen suppressed
the growth of LAPC-4 cells to a greater extent than either
treatment alone. These results confirm that AMACR is an
androgen-independent growth modifier in prostate cancer and
suggest that the enzyme is a complementary drug target with
androgen ablation therapy.9 There are currently no effective
small molecule drug treatments for hormone refractory prostate
cancer. The prospects for drug treatment with AMACR inhibi-
tors are good because, although AMACR exists in many organs,
the major clinical manifestation of congenital racemase defi-
ciency is adult-onset sensory motor neuropathy, which is caused
by the slow, prolonged accumulation of branched-chain fatty
acids.8,10 During treatment with AMACR inhibitors, this could
be prevented by a low phytanic acid diet. AMACR is also
involved in bile acid synthesis, however, an alternative pathway
that does not rely on AMACR has recently been demonstrated
in AMACR-deficient mice.11
Prostate cancer (PCaa) is the most common solid tumor
malignancy in men in Western countries and was responsible
for 30 000 deaths during 2005 in the U.S.1 Androgen deprivation
by surgery or use of drugs is the most common treatment for
advanced PCa. However, despite a rapid initial response in the
majority of cases, PCa will ultimately progress to a hormone
refractory stage and metastatic disease. There is, therefore, a
great deal of interest in the identification of improved diagnostic
markers and therapeutic agents. Through mRNA expression
profiling, it has been shown that the R-methylacyl-CoA race-
mase gene (AMACR) is consistently upregulated in prostate
cancer (8-fold increase in mRNA), and overexpression of the
enzyme has been confirmed.2,3 AMACR is involved in the
â-oxidation of branched-chain fatty acids, and recent data show
that the entire peroxisomal pathway for branched-chain sub-
strates is up-regulated.4 AMACR overexpression starts early in
oncogenesis, with 70% of high-grade prostatic intraepithelial
neoplasia staining positive. In primary PCa, >95% of cases are
positive for AMACR, whereas <1% of adjacent normal
epithelial tissue stains positive. Once overexpressed in PCa, the
AMACR level remains elevated during progression to higher
grades and stages and even metastases. The enzyme is also
overexpressed in other cancers such as colorectal (92% of cases)
and breast (44% of cases).5 Epidemiological studies have shown
that red meat and dairy products, major sources of branched-
* To whom correspondence should be addressed. Tel.: +44 (0)151 794
3531. Fax: +44 (0)151 794 3588. E-mail: acarnell@liv.ac.uk.
† University of Liverpool.
‡ Academic Medical Center.
§ Johns Hopkins Oncology Center.
a Abbreviations: AMACR, R-methylacyl-CoA racemase; CoA, coenzyme
A; PCa, prostate cancer; THC, trihydroxycoprostanoyl; DAST, diethylami-
nosulfur trifluoride; CDI, carbonyl diimidazole; MEM, methoxyethoxym-
ethyl; MOPS, 3-(N-morpholino)propanesulfonic acid; LACS, long-chain
acyl-CoA synthetases; MTT, (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltet-
razolium bromide; PC3, CWR22Rv1, Du145, human prostate cancer cell
lines; PrEC, prostate epithelial cells; RIPA, radioimmunoprecipitation assay.
AMACR is a peroxisomal and mitochondrial enzyme cata-
lyzing the racemisation of branched chain R-methylacyl-CoA
esters such as pristanoyl-CoA 1 and the bile acid interme-
diate 3R,7R,12R-trihydroxycholestanoyl-CoA (THC-CoA) 2
10.1021/jm0702377 CCC: $37.00 © 2007 American Chemical Society
Published on Web 05/04/2007

R-Methylacyl-CoA Racemase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2701
oxazolidinone 11 and the tetradecanal afforded the syn aldol
product 12 as a single isomer. SN2 displacement of the hydroxyl
group with fluorine to give 13 was achieved with diethylami-
nosulfur trifluoride (DAST) at -78 °C. The fluorination proved
to be particularly low yielding (20%), a consequence of the
propensity for elimination of fluoride, which could be limited
at low temperature. Attempts to use TBAF displacement of the
mesylate derived from alcohol 12 resulted in almost complete
elimination. Removal of the oxazolidinone was achieved in good
yield without elimination of fluorine using lithium hydroper-
oxide to give 14. CDI-mediated coupling with CoA then
afforded the R-methyl-â-fluoro compound 5. The diastereomeric
inhibitor 6 was made similarly starting from the alternative
oxazolidinone auxiliary 15.
The THC analogue 7 was made using a similar approach
(Scheme 3). In this case, it was necessary to protect the hydroxyl
groups in the starting material, cholic acid 16, such that they
could be removed under conditions that would leave the
penultimate R-methyl-â-fluoro acid 19 intact. After a number
of attempts with protecting groups such as tetrahydropyranyl
and benzyl, where we were unable to achieve complete
protection, we found the methoxyethoxymethyl (MEM) group
to be the ideal choice. Thus after the four step conversion of
cholic acid to the aldehedye 17, Evans aldol and DAST
fluorination gave compound 18.
After removal of the auxiliary, the final deprotection to
remove the MEM groups was achieved using several drops of
concentrated hydrochloric acid (HCl) in tetrahydrofuran (THF)
to give the acid 19. Alternative conditions such as TiCl4 resulted
in total decomposition, ZnBr2 gave no reaction, and generation
of HCl by use of an acetyl chloride/MeOH mixture resulted in
re-esterification of the carboxylic acid. We were unable to
achieve the CoA ester coupling using the CDI method, therefore,
we used an enzymatic approach that involved incubation of the
acid 19 with purified rat liver microsomes in the presence of
ATP, MgCl2, and CoA.
Enzyme Inhibition Studies. Enzyme inhibition assays were
performed based on the ability of the compounds to inhibit the
AMACR-catalyzed racemisation of (25S)-THC-CoA 2. As the
AMACR source, an organellar pellet prepared by centrifugation
of homogenized rat liver was used. The incubations consisted
of 0.005 mg/mL of protein, 100 mM MOPS, pH 7.0, and 50
µM (25S)- or (25R)-THC-CoA, with increasing concentrations
of the inhibitor compounds (0-200 µM dissolved in 20 mM
MES, pH 6.0). Reactions were allowed to proceed for 20 min
at 37 °C and then terminated by the addition of HCl, followed
by resolution of (25S)- or (25R)-THC-CoA by reversed-phase
HPLC.16 Although our compounds had been designed as
mechanism-based irreversible inhibitors, our kinetic data showed
that the compounds (3-7) were actually competitive inhibitors,
with inhibition constants in the range 0.9-20 µM (Table 1).
Figure 1. Equilibration of the R-chiral center of pristanoyl-CoA 1 and
trihydroxycholestanoyl-CoA 2 by AMACR.
(Figure 1). Pristanoyl-CoA 1 is derived from the diet either
directly or by R-oxidation of dietary phytanic acid and exists
as (2R)- and (2S)-stereoisomers. THC-CoA is formed from
cholesterol as the (25R)-stereoisomer. The racemase intercon-
verts R- and S-stereoisomers of these R-methylacyl-CoA esters,
providing the (S)-isomer for the branched-chain acyl-CoA
oxidase in the first step of â-oxidation.
Little is known about the mechanism of action of this enzyme,
which has been purified and characterized by Schmitz et al.12,13
Based on the observation that the enzyme catalyzes the rapid
exchange of the substrate R-proton (2-3H-pristanoyl-CoA) with
water, it has been assumed that the enzyme functions by proton
abstraction to form an enolate, followed by reprotonation using
a one- or two-base mechanism. The nature of the basic groups
involved in the proton abstraction has not yet been established,
although a crystal structure of Mycobacterium tuberculosis
AMACR (43% sequence identity with the human enzyme) has
recently been determined.14 Four key residues, Arg91, His126,
Asp156, and Glu241, were found to be essential for catalysis,
with Asp156 being identified as the putative base.
Prior to the publication of the bacterial crystal structure, we
designed a series of substrate analogs 3-7 as mechanism-based
inhibitors containing one or more fluorine atoms in a â-position
to the CoA thioester (Figure 2). We envisaged that elimination
of fluoride from the proposed enolate by an E1cB mechanism
within the active site would generate an R,â-unsaturated thioester
that may then undergo attack by an active site base or
nucleophile to give the inactivated enzyme.
Results and Discussion
Chemistry. The synthesis of compound 3 commenced with
a Wittig reaction of dodecyl phosphonium bromide ylid with
methyl trifluoropyruvate (Scheme 1) to give a 1:1; E/Z mixture
of the 2-trifluoromethylester 9. Hydrogenation and ester hy-
drolysis followed by activation of the carboxylic acid with ethyl
chloroformate to form the mixed anhydride allowed coupling
with coenzyme A to give the thioester 3, which was purified
by reverse phase HPLC. The synthesis of difluoromethyl
thioester 4 was initiated by reacting dodecyl magnesium bromide
with trifluoromethyl acrylic acid according to Kitazume et al.15
The resulting difluoroalkene 10 proved unreactive to the reported
hydrogenation conditions using sonication and required suc-
cessive hydrogenations using Pd/C catalyst at 20 atm pressure.
Upon formation of the saturated difluoroacid, carbonyl diimi-
dazole (CDI)-mediated coupling with CoA generated the target
thioester 4 in 20% yield after purification by HPLC.
Attempted lipase-mediated resolution of the carboxylic acid
precursors to 3 and 4 did not give useful enantioselectivity, so
the CoA esters were made from the racemic acids and tested as
diastereoisomeric mixtures. For the synthesis of compounds 5
and 6, we employed an Evans asymmetric aldol reaction to have
control over absolute and relative stereochemistry of the
2-methyl and 3-fluoro substituents (Scheme 2). Thus, asym-
metric aldol reaction between the boron enolate derived from
Figure 3 shows a kinetic plot for inhibitor 6 clearly showing
the classical profile for reversible competitive inhibition. These
are encouraging values, because the only previous reported Ki
for an inhibitor of AMACR is 56 µM for ibuprofenoyl-CoA
20.12 We assume that this was measured using a diastereomeric
mixture made from racemic ibuprofen. We also assayed each
diastereomer of ibuprofenoyl-CoA separately and found (R)-
ibuprofenoyl-CoA 20 to be around 3.5-fold more effective as
inhibitor (Ki ) 5.4) than the (S)-isomer 20 (Ki ) 19.2).17
Our results show that incorporation of fluorine in the
â-position to the CoA ester gives compounds that can inhibit
AMACR, with the best inhibitor 3 containing the R-trifluorom-
ethyl group. Compound 4 containing the R-difluoromethyl group

2702 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Carnell et al.
Figure 2. Compounds designed as mechanism-based inhibitors of AMACR, with the proposed mechanism shown in the inset.
Scheme 1a
a Reagents and conditions: (i) PPh3, toluene, ∆, 48 h, 98%; (ii) n-BuLi, F3C(O)CO2Me, THF, -78 °C, 3 h, 54%; (iii) H2/Pd/C, MeOH, rt, 24 h, 94%;
(iv) 5 N NaOH, THF, MeOH, rt, 24 h, 97%; (v) NEt3, ClCO2Et, DCM, rt, 2 h; (vi) CoALi3, KHCO3, t-BuOH, rt, 24 h; (vii) Mg turnings, THF, rt, 30 min;
(viii) trifluoromethyl acrylic acid, THF, 5 h, -40 °C, 35%; (ix) H2/Pd/C, MeOH, rt, 20 atm., 67%; (x) CDI, THF; (xi) CoALi3, THF, H2O, rt, overnight,
20%.
Scheme 2a
a Reagents and conditions: (i) n-Bu2BOTf, DIPEA, Me(CH2)12CHO, THF, -78 °C, 3 h, 41%; (ii) DAST, DCM, -78 °C, 3 h, 22%; (iii) H2O2, LiOH‚H2O,
THF, H2O, 0 °C, overnight, 90%; (iv) CDI, THF; (v) CoALi3, THF, H2O, rt, overnight, 20%.
was much less effective for reasons that are as yet unclear.
Compounds 5-7 containing one fluorine atom on the â-position
of the chain have similar low Ki values. The reversal of absolute
configuration of the methyl and fluorine groups in compounds
5 and 6 appeared to make little difference to the Ki. The similar
Ki of 5 and 6 and the fact that both diastereomers of pristanoyl-
CoA 1 are substrates for the racemase led us to believe that
stereochemistry at the R-position in this series may not be an
important factor. However, the marked difference in Ki values
for (R)- and (S)-ibuprofenoyl-CoA 20 is interesting and probably

R-Methylacyl-CoA Racemase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2703
Scheme 3a
a Reagents and conditions: (i) amberlite ion ex. res., MeOH, rt, overnight,
97%; (ii) MEM-Cl, DIPEA, DCM, rt, 24 h, 41%; (iii) LiAlH4, THF, rt, 4
h, 90%; (iv) PCC, DCM, rt, 3 h, 61%; (v) n-Bu2BOTf, DIPEA, CAi-Pr,
THF, -78 °C, 3 h, 23%; (vi) DAST, DCM, -78 °C, 3 h, 47%; (vii) H2O2,
LiOH‚H2O, THF, H2O, 0 °C, overnight, 95%; (viii) conc aq HCl, THF, 3
h, rt, 98%; (ix) rat liver microsomes, ATP, MgCl2, CoASH.
Figure 3. Kinetic plot for inhibitor 6 showing competitive inhibition.
the natural substrate by facilitating enolate formation in the
active site, although the Ki for the difluoromethyl compound 4
is anomalous. The polar fluorine-carbon bonds may also increase
the binding affinity although, if this were the case, one might
expect to see a larger difference in inhibition between inhibitors
3 and 5-7, where C-F bonds would be orientated differently
upon binding.
Table 1. Competitive Inhibitors of AMACRa
Our inhibitors were conceived prior to the publication of the
M. tuberculosis crystal structure when we were unaware of the
possibility of Asp 152 in human AMACR (Asp 156 in the M.
tuberculosis structure) acting as the active site base. Because
the conjugate addition of an aspartate carboxylate to an unsatu-
rated ester is likely to be reversible, one would not expect mech-
anism-based irreversible inhibition to be observed. In theory it
is possible that our inhibitors eliminate fluorine and that the
unsaturated esters are the real inhibitors, but this is unlikely
because 2-methylmyrist-2-enoyl CoA is not an inhibitor.12
The use of CoA in drug compounds is likely to prevent them
from being orally bioavailable. However, the CoA thioesters
may be biosynthesized in vivo by a long-chain acyl-CoA
thioester synthetases (LACS)18 or other CoA ester ligases,19 from
the carboxylic acid as a pro-drug. For example, (R)-ibuprofen
is thioesterified by a stereoselective synthetase in vivo prior to
racemisation.19 Incubation of the free carboxylic acids corre-
sponding to inhibitors 3, 5, and 7, with rat liver homogenate or
purified microsomes in the presence of ATP, MgCl2, and CoA
indeed showed that all three acids were thioesterified by an
endogenous synthetase/ligase. It can be speculated that an acid,
which is very similar to the natural substrate precursor, could
become thioesterified by a synthetase and would be transported
into the peroxisome and/or mitochondrion where AMACR is
localized. This in vivo arming of the inhibitor might afford
selective toxicity, because the free acid or ester acid pro-drug
would not be active as an inhibitor.
a Ki values were determined using direct linear plot analysis. AMACR
activity was determined by monitoring the interconversion of (25R/S)-THC-
CoA.16 Assays consisted of 5 µg/mL protein, 100 mM MOPS pH 7, 50
µM (25S) or (25R)-THC-CoA with increasing concentrations of inhibitor.
b Ki for racemic ibuprofenoyl-CoA and Ki values for compounds 21 and 22
calculated from reported kinetic data.12
Inhibitory Activity of (()-2-Trifluoromethyltetradecanoic
Acid on Prostate Tumor Cell Growth. Several prostate cancer
cell lines (PC3, CWR22Rv1, Du145) as well as cultured prostate
epithelial cells (PrEC) were incubated with 50 µM and 100 µM
(()-2-trifluoromethyltetradecanoic acid, the carboxylic acid
precursor of compound 3. The cells were then tested for viability
at 24 h intervals for up to 10 days using the MTT cell
proliferation assay. The results presented in Figure 4a clearly
show that this compound has a cytotoxic effect. With both
concentrations of the drug tested, the number of viable cells
was significantly diminished after 4 days for all cell lines.
CWR22Rv1 was the least sensitive to the cytotoxic effect of
2-trifluoromethyltetradecanoic acid, with a viability of 95.3%
and 74% after 3 days of incubation with 50 µM and 100 µM of
the drug, respectively. Interestingly, this cell line had the lowest
reflects the greater steric demands of the aromatic ring im-
mediately adjacent to the R-carbon.
Schmitz et al. reported an increase in Km for racemisation
of THC-CoA in the presence of 2-methyloctanoyl CoA 21
(270 µM) and 2-methylmyristoyl CoA 22 (175 µM).12 With the
substrate and inhibitor concentrations reported, Ki values of 45
µM for 2-methyloctanoyl 21 and 137 µM for 2-methylmyristoyl
CoA 22 can be calculated. The presence of fluorine in our
inhibitors clearly has a significant effect, because the analogous
fluorinated compounds 3, 5, and 6 have Ki values between 0.9
and 2.3 µM. The effect exerted by fluorine is not yet clear, but
we hypothesize that the fluorine-induced reduction in pKa of
the R-proton may make these compounds better substrates than

2704 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Carnell et al.
Figure 4. (a) Cell viability for PC3, CWR22 Rv1, Du 145, and PrEC following treatment with 100 µM (()-2-trifluoromethyltetradecanoic acid.
A two-tailed student’s T-test was performed for each triplicate day compared to its vehicle control; any time point for which the p-value is less that
0.05 is marked with an asterisk (*). (b) Western Blot analysis of AMACR expression. Immunoblot analysis of AMACR in whole cell lysates of
PC3, CWR22 Rv1, Du145, and PrEC. An immunoblot for â-actin expression was performed as a loading control.
AMACR expression as seen by Western Blot (Figure 4b), which
suggests that the cytotoxic effect of the compound could be
AMACR-dependent.
2-methyloctanoyl-CoA, 2-methylmyristoyl-CoA, or ibupro-
fenoyl- CoA. The presence of fluorine is postulated to lower
the pKa of the R-proton facilitating enolization and binding in
the active site. (R)- and (S)-ibuprofenoyl-CoA were tested
separately and, unlike the fluorinated inhibitors or the natural
substrates, which have no â-branching, showed significant
differential activity with the isomer derived from (R)-ibuprofen,
exhibiting a 3.5-fold lower Ki. (()-2-Trifluoromethyltetrade-
canoic acid, the precursor of the most potent AMACR inhibitor
3, showed cytotoxic activity on cancer cell lines PC3, CWR22
Rv1, and Du145, with CWR22 Rv1, expressing low levels of
AMACR, being least affected. Normal prostate epithelial PrEC
cells expressing an unexpectedly high level of AMACR were
also affected. Although the 100 µM concentration is relatively
high, the carboxylic acid needs to be taken up by the cells and
activated to its CoA ester prior to targeting the enzyme. The
concentration of inhibitor required for substantial cell growth
inhibition and time taken to exert its affect are, therefore, not
surprising, but these data provide proof of concept for this first
series of AMACR inhibitors and encourage us to examine
related compounds as potential drug leads in the future.
Surprisingly, nonmalignant PrEC cells were most sensitive
to the toxic effect of this compound. Previous studies consis-
tently demonstrate very low expression of AMACR in normal
prostate epithelium.20 However, little data exists on the expres-
sion of this protein in cultured prostate epithelial cells.21,22 Our
results show that PrEC expresses AMACR at a readily detectable
level. Whether the sensitivity we observed for these cells reflects
a combination of effects, both compound-mediated as well as
the normal senescence program that these cells inevitably
progress through, is not clear at present. These data suggest
that follow-up studies in nonprostate and nontumorgenic cell
lines both with high and with low AMACR expression are
necessary to assess the role of this activity in the cytotoxic
effects observed.
Conclusion
A range of AMACR substrate analogs containing an R-dif-
luoromethyl, an R-trifluoromethyl, or a fluorine in the â-position
on the acyl chain have been synthesized. Synthetic routes were
developed that avoided elimination of fluorine at key stages to
allow sufficient quantities to be synthesized for the biological
assays. Three R-methyl-â-fluoroacyl inhibitors were made as
single isomers to test the effect of chirality. These compounds,
originally designed as irreversible mechanism-based inhibitors,
are promising reversible competitive inhibitors of AMACR and
are significantly more potent than the previously reported
Experimental Section
General. All solvents were distilled and kept over 3 Å molecular
sieves prior to use. Analytical grade solvents were used for reactions
involving aqueous media. Dichloromethane (DCM) was distilled
from calcium hydride before use. Sureseal anhydrous N,N-dimeth-
ylformamide (DMF) was purchased from Aldrich or distilled from
calcium hydride before use. Methanol (MeOH) was distilled from

R-Methylacyl-CoA Racemase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2705
a magnesium and iodine still for 2-3 h under argon atmosphere
before use. THF was dried from sodium and benzophenone until
purple coloration persisted and then distilled under atmospheric
pressure. Chiral auxiliaries-(S)-4-isopropyl-3-propionyloxazolidin-
2-one and (4R,5S)-4-methyl-5-phenyl-3-propionyloxazolidin-2-one
were purchased from Fluka. Coenzyme A trithium salt (CoA-Li3)
was purchased from Roche. (R)- and (S)-Ibuprofen were a kind
gift from Knoll Pharmaceuticals, Amsterdam (a branch of BASF
Pharma). The CoA esters were synthesized according to Rasmussen
et al.23 All other reagents were purchased from Sigma-Aldrich
unless otherwise stated. Analytical thin layer chromatography was
performed on 2 × 5 cm aluminum sheets, precoated with Merck
Kieselgel 60, and detection by UV 254 nm light. The developed
chromatograms were visualized with a UV lamp. Plates were then
treated with potassium permanganate dip and developed with a heat
gun. Flash column chromatography was performed using silica
(Merck 9385 Kieselgel 60). Crude products were applied to the
column either dissolved in a small amount of the eluent or absorbed
onto silica, and then the column was eluted with additional solvent
of increasing polarity. Fractions of 5-100 mL were collected and
analyzed by TLC. 1H, 13C, COSY, and HMQC NMR spectra were
recorded on a Bruker AMX-400 machine. Coupling constants (J
values) are reported in Hz. Chemical shifts are given in ppm,
downfield from an internal standard of tetramethylsilane (TMS).
IR spectra were recorded as nujol mulls in the range 4000-600
cm-1 using a Perkin-Elmer 883 spectrophotometer and a Perkin-
Elmer FT-IR Paragon 1000 spectrometer on sodium chloride plates
washed with ethanol. Mass spectrometry analyses (ES+, ES-, EI+)
were performed on a Trio1000 spectrometer and run by the mass
spectrometry service within the University of Liverpool. Where
chemical ionizations (CI) were used, ammonia was used as the
carrier gas. Gas chromatography was carried out on a Shimadzu
GC-14B module using an EC-1 column (30 m × 0.32 mm, 1 µm
film thickness). Chiral separations used a FS-Lipodex-E column
(25 m × 0.25 mm). Analytical reverse phase HPLC was performed
on a Waters 2695 separations module equipped with a Waters 996
photodiode array detector. Separations were carried out using an
Alltech Altima C-18 5µ, 250 mm × 4.6 mm column packed with
5 µm spherisorb ODS silica, with a gradient using 90% MeOH/
H2O and 50% MeOH/25 mM K2HPO4/KH2PO4. Analyses were also
carried out using a Perkin-Elmer Biocompatible Binary Pump 250
and a Pharmacia UV detector LKB-UV-MII. Separations were
carried out using a SupelcoSIL LC-18-DB column 250 mm × 4.6
mm, the gradient used was 40 to 70% A/B; (A ) 70% MeCN/16.9
mM sodium phosphate, pH 6.9; B ) 10% MeCN/16.9 mM sodium
phosphate, pH 6.9).
Methyl 2-Trifluoromethyltetradec-2-enoate (9). Triphenylphos-
phine (5.41 g, 20.66 mmol, 1 equiv) was added to a solution of
1-bromododecane (5 mL, 20.66 mmol, 1 equiv) in toluene (50 mL).
The solution was heated under reflux for 48 h. The solvent was
then removed under reduced pressure to yield dodecyltriph-
enylphosphonium bromide (8.74 g, 98%) as a dark yellow oil, which
was used without further purification. n-Butyllithium (1.6 M, 3.72
mL, 5.95 mmol, 1.1 equiv) was added dropwise to a cooled solution
of dodecyltriphenylphosphonium bromide (2.33 g, 5.41 mmol, 1
equiv) in THF (50 mL) at -70 °C. The mixture was allowed to
warm to room temperature over 2 h and stirred for an additional 1
h. The dark red mixture was again cooled to about -70 °C, and
methyl trifluoropyruvate (843 mg, 5.41 mmol, 1 equiv) was added
dropwise. The resulting yellow solution was monitored by TLC
until conversion was complete. The mixture was then warmed to
room temperature and filtered through a short column of silica gel
(30 g) to remove triphenylphosphine oxide and salts. The crude
product was concentrated in vacuo and chromatographed on silica
gel (eluent: Et2O/hexane (1:20) to yield both geometric isomers
(900 mg, 54%) of methyl 2-trifluoromethyltetradec-2-enoate 9 in
an approximate 1:1 ratio; Rf ) 0.68 (E-isomer), 0.62 (Z-isomer),
Et2O/hexane (1:20).
activated carbon. The vessel was then evacuated under reduced
pressure and flushed with hydrogen three times. The reaction was
allowed to stir overnight. The mixture was then flushed with
nitrogen and opened to the atmosphere. The mixture was filtered
through a pad of Celite and washed with MeOH. The solvent was
then concentrated in vacuo to yield methyl 2-trifluoromethyltet-
radecanoate as a pale yellow oil (837 mg, 94%). Saponification
was carried out by dropwise addition of 5 N sodium hydroxide (1
mL) to a solution of methyl 2-trifluoromethyltetradecanoate (70
mg, 0.23 mmol) in THF (3 mL)/MeOH (2 mL). The resulting
yellow solution was allowed to stir at room temperature overnight.
After acidification with 1 M HCl (5 mL), the crude product was
redissolved in DCM (10 mL) and extracted with water (2 × 5 mL).
The organic layer was dried (MgSO4), concentrated in vacuo, and
chromatographed on silica gel (eluent: Et2O/hexane (2:5)) to yield
2-trifluoromethyltetradecanoic acid as a pale yellow oil (61.4 mg,
91%): 1H NMR (400 MHz, CDCl3) 0.88 (3H, t, J ) 7.0 Hz, H-14),
1.22-1.36 (20H, m, H-4 to H-13), 1.72-1.78 (1H, m, H-3), 1.83-
1.92 (1H, m, H-3), 3.06-3.14 (1H, m, H-2); 13C NMR (100 MHz,
CDCl3) 14.4 (C-14), 23.0-32.3 (C-3 to C-13), 50.8 (C-2), 168.5
(C-1); νmax /cm-1 1756 (CdO); νmax/cm-1 3300-2950 (O-H acid),
1729 (CdO); HRMS calcd for C15H31F3O2N (M + NH4+),
314.2307; found, 314.2315. To a stirred solution of carbonyldi-
imidazole (1.84 mg, 11.35 µmol, 1.2 equiv, desiccated over silica
crystals overnight) in THF (0.2 mL) was added a solution of
2-trifluoromethyltetradecanoic acid (2.81 mg, 9.45 µmol, 1 equiv)
in THF (0.2 mL). After 2 h of stirring at room temperature, a
solution of coenzyme A trilithium salt (CoA-Li3; 8.2 mg, 10.39
µmol, 1.1 equiv) in water (0.2 mL) was added. The resulting
solution was allowed to stir at room temperature overnight. Two
drops of 0.1 M HCl was then added, and the solution was diluted
with MeOH (0.5 mL). The product was then purified by semi-
preparative reverse phase HPLC (elution gradient MeOH/H2O/25
mM phosphate buffer) to give the title compound 3 as a white
amorphous solid (1.89 µmol, 20% yield, calculated from UV
absorption of a specific concentration): UV/vis absorption maxima
at 260 nm; HRMS calculated for C36H60F3N7O17P3S (M + H-),
1044.2982; found, 1044.2931 (for HPLC chromatographs, see
Supporting Information, S10).
2-Difluoromethylenepentadecanoic Acid (10). 1-Bromodode-
cane (3.25 mL, 13.45 mmol, 2.5 equiv) and magnesium turnings
(325 mg, 13.45 mmol, 1 equiv) were added to a solution of trifluoro-
methylacrylic acid (750 mg, 5.38 mmol, 1 equiv) and trimethylsilyl
chloride (0.68 mL, 5.38 mmol, 1 equiv) in THF (20 mL) at -40
°C. After stirring for 5 h at that temperature, the mixture was
quenched with saturated ammonium chloride (20 mL) below -20
°C. The resulting solution was extracted with ether (100 mL). The
combined extracts were then washed with water (100 mL) and dried
over anhydrous MgSO4. The solvent was removed and the crude
product was chromatographed on silica gel (eluent: EtOAc/hexane
(1:4)) to yield 2-difluoromethylenepentadecanoic acid 10 as an
amorphous yellow solid (550 mg, 35%): Rf ) 0.46 EtOAc/hexane
(3:7).
2-Difluoromethylpentadecanoyl-CoA (4). A solution of 10
(160 mg, 0.55 mmol) in distilled MeOH (5 mL) containing 1 spatula
of palladium over activated carbon was hydrogenated at a pressure
of 20 atm for 16 h. HCl (1 M, 2 mL) was added, and the mix-
ture was filtered through a pad of Celite. The crude product was
concentrated in vacuo and chromatographed on silica gel (eluent:
EtOAc/hexane (1:4)) to yield 2-difluoromethylpentadecanoic
acid as a pale yellow oil (108 mg, 67%): Rf ) 0.52 EtOAc/hexane
1
(3:7); δ H (400 MHz, CDCl3) 0.88 (3H, t, J ) 7.0 Hz, H-15),
1.22-1.38 (20H, m, H5-14), 1.42-1.51 (2H, m, H-4), 1.71-1.78
(2H, m, H-3), 2.82-2.89 (1H, m, H-2), 5.96 (1H, td, JH-F ) 85.7
Hz, J2,3 5.2 Hz, CHF2); δ 13C (100 MHz, CDCl3) 14.4 (C-15), 23.0-
30.1 (C-4 to C-12), 32.2 (C-3), 50.2 (C-2), 116.1 (CF2, t, J )
240 Hz), 168.3 (C1); HRMS calcd for C16H29F2O2 (M + H+),
310.4022; found, 310.4008. 2-Difluoromethylpentadecanoic
acid (2.81 mg, 9.58 µmol) was reacted according to the procedure
previously described above for 2-trifluoromethyltetradecanoic
acid to afford the title compound 4 as a crystalline solid (1.89
2-Trifluoromethyltetradecanoyl-CoA (3). To a solution of
methyl 2-trifluoromethyltetradec-2-enoate 9 (890 mg, 2.88 mmol)
in distilled MeOH (10 mL) was added 1 spatula of palladium over

2706 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Carnell et al.
µmol, 20%): UV/vis- absorption maxima at 260 nm; HRMS calcd
for C38H63F3N7O18P3S, 1068.3034 (M + H- + MeOH); found,
1068.3028 (for HPLC chromatographs see Supporting Information
S10).
aluminum hydride (3.5 M, 4 equiv, 6.88 mL, 24.1 mmol) was added
dropwise to a solution of methyl (3R,7R,12R)-tris(methoxy-
ethoxymethyl) cholate (1.49 g, 3.53 mmol) in THF (35 mL). The
resulting suspension was allowed to stir at room temperature
for 4 h. The reaction was quenched by the addition of water
(5 mL), followed by 1 M sodium hydroxide solution (2 mL) and
water (5 mL). The resulting precipitate was filtered, and the fil-
trate was concentrated in vacuo. The crude was then chroma-
tographed on silica gel (eluent: EtOAc/hexane (1:8) to yield
(3R,7R,12R)-tris(methoxyethoxymethyl)-5â-cholestan-24-ol as a
clear oil (606 mg, 45%): Rf ) 0.26 DCM/MeOH (9:1). (3R,7R,-
12R)-Tris(methoxyethoxymethyl)-5â-cholestan-24-ol (606 mg,
0.92 mmol, 1 equiv) was added to a solution of pyridinium chloro-
chromate (397 mg, 1.84 mmol, 2 equiv) in DCM (9 mL). The re-
sulting solution was allowed to stir at room temperature for 3 h.
The reaction was quenched with ethanol (2 mL) and filtered through
a silica pad. The crude was concentrated in vacuo and chro-
matographed on silica gel (eluent: EtOAc/hexane (1:4)) to yield
(3R,7R,12R)-tris(methoxyethoxymethyl)-5â-cholestan-24-al 17 as
a clear oil (367 mg, 61%): Rf ) 0.36 DCM/MeOH (9:1).
N-(24S,25R,4′S)-(3R,7R,12R)-Tris(methoxyethoxymethyl)-24-
fluoro-5â-cholestan-26-oyl-4′-isopropyl-oxazolidin-2′-one (18).
(3R,7R,12R)-Tris(methoxyethoxymethyl)-5â-cholestan-24-al 17 (367
mg, 0.56 mmol) was reacted according to the procedure described
above (for 12) to yield N-(24R,25S,4′S)-(3R,7R,12R)-tris(methoxy-
ethoxymethyl)-24-hydroxy-5â-cholestan-26-oyl-4′-isopropyl-oxazo-
lidin-2′-one as a pale yellow oil (109 mg, 23%): Rf ) 0.59 DCM/
MeOH(9:1).N-(24R,25S,4′S)-(3R,7R,12R)-tris(methoxyethoxymethyl)-
24-hydroxy-5â-cholestan-26-oyl-4′-isopropyloxazolidin-2′-one (109
mg, 0.13 mmol) was reacted according to the procedure described
above (for 13) to yield N-(24S,25R,4′S)-(3R,7R,12R)-tris(methoxy-
ethoxymethyl)-24-fluoro-5â-cholestan-26-oyl-4′-isopropyl-oxazo-
lidin-2′-one 18 as a pale yellow oil (51 mg, 47%): Rf ) 0.3 DCM/
MeOH (9:1).
(24S,25R)-(3R,7R,12R)-Trihydroxy-24-fluoro-5â-cholestan-26-
oic Acid (19). N-(24S,25R,4′S)-(3R,7R,12R)-Tris(methoxyethoxym-
ethyl)-24-fluoro-5â-cholestan-26-oyl-4′-isopropyloxazolidin-2′-
one 18 (44.5 mg, 61 mmol) was reacted according to the procedure
described above (for 14) to yield (24S,25R)-(3R,7R,12R)-tris-
(methoxyethoxymethyl)-24-fluoro-5â-cholestan-26-oic acid as a
clear oil (41 mg, 92%): Rf ) 0.21 DCM/MeOH (9:1). Concentrated
HCl (five drops) was added to a solution of (24S,25R)-(3R,7R,-
12R)-tris(methoxyethoxymethyl)-24-fluoro-5â-cholestan-26-oic acid
(12 mg, 16.4mmol) in THF (5 mL). The resulting solution was
allowed to stir at room temperature for 2 h. The solvent was
concentrated in vacuo, and the crude residue was triturated with
DCM (2 mL) to yield (24S,25R)-(3R,7R,12R)-trihydroxy-24-fluoro-
5â-cholestan-26-oic acid (19) as a clear oil (5 mg, 65%): Rf )
0.07 DCM/MeOH (9:1).
N-(2S,3R,4′S)-3-Hydroxy-2-methylhexadecanoyl)-4′-isopropy-
loxazolidin-2′-one (12). n-Dibutylboron triflate (1.1 M in DCM,
2.02 mL, 2.05 mmol, 1.4 equiv) and diisopropylethylamine (0.41
mL, 2.35 mmol, 1.6 equiv) were added successively to a solution
of (S)-4-isopropyl-3-propionyloxazolidin-2-one 11 (0.35 mL, 2.05
mmol, 1.4 equiv) in DCM (5 mL) at -78 °C. After stirring for 30
min, a solution of tetradecanal (313 mg, 1.46 mmol, 1 equiv) in
DCM (2 mL) was added dropwise. The resulting pale yellow
solution was then allowed to stir at -78 °C for an additional 30
min and then 2 h at room temperature. The solution was quenched
with 0.1 M phosphate buffer (pH 7.2, 5 mL). The organic layer
was extracted with 1 M HCl (5 mL), NaHCO3 (5 mL), and brine
(5 mL). The organic layer was dried (MgSO4), concentrated in
vacuo and chromatographed on silica gel (eluent: Et2O/hexane (1:
1)) to yield N-(2S,3R,4′S)-(3-hydrox-2-methylhexadecanoyl)-4′-
isopropyloxazolidin-2′-one 12 as a clear oil (236 mg, 41%): Rf )
0.15 Et2O/hexane (1:1).
N-(2R,3S,4′S)-(3-fluoro-2-methylhexadecanoyl)-4′-isopropyl-
oxazolidin-2′-one (13). N-(2S,3R,4′S)-3-Hydroxy-2-methylhexa-
decanoyl)-4′-isopropyloxazolidin-2′-one 12 (30 mg, 75.5 mmol) in
DCM (2 mL) was added dropwise to a 0.5 M solution of DAST
(0.15 mL, 1 equiv) in DCM (1.2 mL) at -78 °C. After 2 h, the
solution was allowed to warm to room temperature and an equal
volume of water was carefully added. The organic layer was washed
with NaHCO3 (1 mL) and water (1 mL), dried (MgSO4), concen-
trated in vacuo, and chromatographed on silica gel (eluent: EtOAc/
hexane (1:15)) to yield N-(2R,3S,4′S)-(3-fluoro-2-methylhexade-
canoyl)-4′-isopropyl-oxazolidin-2′-one 13 as a pale yellow oil (6.7
mg, 22%): Rf ) 0.67 Et2O/hexane (1:1).
(2R,3S)-3-fluoro-2-methylhexadecanoic Acid (14). Hydrogen
peroxide (30% v/v, 1.2 µL, 6 equiv) was added dropwise to a stirred
solution of 13 (6.7 mg, 16.8 µmol) in THF (0.5 mL) at 0 °C.
Lithium hydroxide monohydrate (2 equiv) was then added, and the
mixture was allowed to stir at room temperature overnight. Saturated
sodium sulfite (0.5 mL) was added, followed by 1 M HCl (0.5
mL). The crude product was then redissolved in DCM (5 mL) and
extracted with water (2 × 4 mL). The organic layer was dried
(MgSO4), concentrated in vacuo, and chromatographed on silica
gel (eluent: EtOAc/hexane (1:3)) to yield (2R,3S)-3-fluoro-2-
methylhexadecanoic acid 14 as a clear oil (4.35 mg, 90%): Rf )
0.33 Et2O/hexane (1:1).
(2R,3S)-3-Fluoro-2-methylhexadecanoyl-CoA
(5). (2R,
3S)-3-Fluoro-2-methylhexadecanoic acid 14 (3.1 mg, 7.76 µmol)
was reacted according to the procedure described above for 2-tri-
fluoromethyltetradecanoic acid to yield the title compound 5 (1.89
µmol, 26% yield, calculated from UV absorption of a specific
concentration): UV/vis absorption maxima at 260 nm; HRMS calcd
for C38H63FN7O17P3S (M + H-), 1036.3423; found, 1036.3433 (for
HPLC chromatographs, see Supporting Information S10).
(24S, 25R)-(3R,7R,12R)-Trihydroxy-24-fluoro-5â-cholestan-
26-oyl-CoA (7). The CoA ester (24S,25R)-(3R,7R,12R)-trihy-
droxy-24-fluoro-5â-cholestan-26-oyl-CoA, 7, was synthesized
using purified rat liver microsomes. The microsomes were pre-
pared by differential centrifugation of a rat liver homogenate as
previously described.24 One millimolar of compound 19 dissolved
in 10 mM Tris pH 8.0 and 1 mg/mL â-cyclodextrin was incubated
in 100 mM Tris pH 8.0, 10 mM ATP, 10 mM MgCl2, 1 mM CoA,
and 0.5 mM DTT with 0.3 mg/mL microsomal protein for 6 h at
37 °C. The synthesized CoA ester was purified by HPLC8 (for
HPLC chromatographs, see Supporting Information S10).
(3R,7R,12R)-Tris(methoxyethoxymethyl)-5â-cholestan-24-al
(17). Amberlite ion-exchange resin (20 g, prewashed with acetone,
1 M HCl, water) was added to a solution of cholic acid 16 (7.48 g,
18.33 mmol) in MeOH. The resulting mixture was allowed to stir
at room temperature overnight. The crude product was filtered,
redissolved in hot ethyl acetate (100 mL), refiltered, and concen-
trated in vacuo to yield to methyl cholate as a white crystalline
solid (7.52 g, 97%): Rf ) 0.11 DCM/MeOH (9:1). Methoxy-
ethoxymethyl chloride (3.98 mL, 34.9 mmol, 7 equiv) was added
dropwise to a solution of methyl cholate (2.1 g, 4.99 mmol, 1 equiv)
and diisopropylethylamine (6.1 mL, 34.9 mmol, 7 equiv) in DCM
(30 mL). The resulting solution was allowed to stir at room
temperature overnight. The reaction was quenched with saturated
ammonium chloride (20 mL) and extracted with brine (2 × 30 mL).
The organic layer was dried (MgSO4), concentrated in vacuo, and
chromatographed on silica gel (elution: DCM/MeOH (9:1)) to yield
methyl (3R,7R,12R)-tris(methoxyethoxymethyl) cholate as a pale
yellow oil (1.41 g, 41%): Rf ) 0.5 DCM/MeOH (9:1). Lithium
Acknowledgment. We thank the EPSRC for a studentship
(I.H.). This work was supported by The Netherlands Organiza-
tion for Scientific Research (NWO, Grant 916.46.109.)
Supporting Information Available: Experimental and analyti-
cal data for 6 and spectroscopic data for compounds 3-5 and 7
and all synthetic intermediates, HPLC traces for CoA esters 3-7,
the AMACR inhibition and cytotoxicity assays, and Western Blot
analyses. This material is available free of charge via the Internet
at http://pubs.acs.org.

R-Methylacyl-CoA Racemase Inhibitors
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2707
(14) Savolainen, K.; Bhaumik, P.; Schmitz, W.; Kotti, T. J.; Conzelmann,
E.; Wierenga, R. K.; Hiltunen, J. K. R-Methylacyl-CoA racemase
from Mycobacterium tuberculosis: Mutational and structural char-
acterization of the active site and the fold. J. Biol. Chem. 2005, 280,
12611-12620.
(15) Kitazume, T.; Ohnogi, T.; Miyauchi, H.; Yamazaki, T.; Watanabe,
S. Ultrasound-promoted synthesis of R-difluoromethylated carboxylic
acids. J. Org. Chem. 1989, 54, 5630-5632.
(16) Ferdinandusse, S.; Denis, S.; Clayton, P. T.; Graham, A.; Rees, J.
E.; Allen, J. T.; McLean, B. N.; Brown, A. Y.; Vreken, P.; Waterham,
H. R.; Wanders, R. J. A. Mutations in the gene encoding peroxisomal
R-methylacyl-CoA racemase cause adult-onset sensory motor neu-
ropathy. Nat. Genet. 2000, 24, 188-191.
(17) The difference in magnitude between our Ki values for ibuprofenoyl-
CoA and the previously reported 56 µM can be attributed to the
different assay method used and the AMACR source. Schmitz used
pristanoyl CoA as the substrate and reported kinetic constants
Km ) 60 µM and Vmax ) 31 µmol‚min-1‚(mg protein)-1 compared
to our values Km ) 24 µM and Vmax ) 141 nmol‚min-1‚(mg
protein).12
(18) Wanders, R. J. A.; Vreken, P.; Ferdinandusse, S.; Jansen, G. A.;
Waterham, H. R.; Van Roermund, C. W. T.; Van Grunsven, E. G.
Peroxisomal fatty acid R- and â-oxidation in humans: Enzymology,
peroxisomal metabolite transporters, and peroxisomal diseases.
Biochem. Soc. Trans. 2001, 29, 250-267.
(19) Knights, K. M.; Drogemuller, C. J. Xenobiotic-CoA ligases:
Kinetic and molecular characterization. Curr. Drug Metab. 2000, 1,
49-66.
(20) Jiang, Z.; Woda, B. A.; Wu, C. L.; Yang, X. J. Discovery and
clinical application of a novel prostate cancer marker: R-Methylacyl
CoA racemase (P504S). Am. J. Clin. Pathol. 2004, 122,
275-289.
(21) Dalrymple, S.; Antony, L.; Xu, Y.; Uzgare, A. R.; Arnold, J. T.;
Savaugeot, J.; Sokoll, L. J.; De Marzo, A. M.; Isaacs, J. T. Role of
notch-1 and E-cadherin in the differential response to calcium in
culturing normal versus malignant prostate cells. Cancer Res. 2005,
65, 9269-9279.
References
(1) Jemal, A.; Murray, T.; Ward, E.; Samuels, A.; Tiwari, R. C.; Ghafoor,
A.; Feuer, E. J.; Thun, M. J. Cancer Statistics, 2005. CA-Cancer J.
Clin. 2005, 55, 10-30.
(2) Luo, J.; Zha, S.; Wesley, G. R.; Dunn, T. A.; Hicks, J. L.; Bennet,
C. J.; Ewing, C. M.; Platz, E. A.; Ferdinandusse, S.; Wanders, R. J.;
Trent, J. M.; Issacs, W. B. R-Methylacyl-CoA racemase: A new
molecular marker for prostate cancer. Cancer Res. 2002, 62, 2220-
2226.
(3) Rubin, M. A.; Zhou, M.; Dhanasekaran, S. M.; Varambally, S.;
Barrette, T. R.; Sanda, M. G.; Pienta, K. J.; Ghosh, D.; Chinnaiyan,
A. M. R-Methylacyl coenzyme A racemase as a tissue biomarker
for prostate cancer. JAMA, J. Am. Med. Assoc. 2002, 287, 1662-
1670.
(4) Zha, S.; Ferdinandusse, S.; Hicks, J. L.; Denis, S.; Dunn, T. A.;
Wanders, R. J.; Luo, J.; De Marzo, A. M.; Isaacs, W. B. Peroxisomal
branched chain fatty acid â-oxidation pathway is upregulated in
prostate cancer. Prostate 2005, 63, 316-323.
(5) Zhou, M.; Chinnaiyan, A. M.; Kleer, C. G.; Lucas, P. C.; Rubin, M.
A. R-Methylacyl-CoA racemasesA novel tumor marker over-
expressed in several human cancers and their precursor lesions. Am.
J. Surg. Pathol. 2002, 26, 926-931.
(6) Kolonel, L. N. Fat, meat, and prostate cancer. Epidemiol. ReV. 2001,
23, 72-81.
(7) Zheng, S. L.; Chang, B.-L.; Faith, D. A.; Johnson, J. R.; Issacs, S.
D.; Hawkins, G. A.; Turner, A.; Wiley, K. E.; Bleecker, E. R.; Walsh,
P. C.; Meyers, D. A.; Isaacs, W. B. Sequence variants of R-methy-
lacyl-CoA racemase are associated with prostate cancer risk. Cancer
Res. 2002, 62, 6485-6488.
(8) Ferdinandusse, S.; Denis, S.; Clayton, P. T.; Graham, A.; Rees, J.
E.; Allen, J. T.; McLean, B. N.; Vreken, P.; Waterham, H. R.;
Wanders, R. J. A. Mutations in the gene encoding peroxisomal
R-methylacyl-CoA racemase cause adult-onset sensory motor neu-
ropathy. Nat. Genet. 2000, 24, 188-191.
(9) Zha, S.; Ferdinandusse, S.; Denis, S.; Wanders, R. J.; Ewing, C. M.;
Luo, J.; De Marzo, A. M.; Isaacs, W. B. R-Methylacyl-CoA racemase
as an androgen-independent growth modifier in prostate cancer.
Cancer Res. 2003, 63, 7365-7376.
(10) Clayton, P. T. Clinical consequences of defects in peroxisomal
â-oxidation. Biochem. Soc. Trans. 2001, 29, 298-305.
(11) Savolainen, K.; Kotti, T. J.; Schmitz, W.; Savolainen, T. I.; Sormunen,
R. T.; Ilves, M.; Vainio, S. J.; Conzelmann, E.; Hiltunen, J. K. A
mouse model for R-methylacyl-CoA racemase deficiency: Adjust-
ment of bile acid synthesis and intolerance to dietary methyl-branched
lipids. Hum. Mol. Gen. 2004, 13, 955-965.
(12) Schmitz, W.; Fingerhut, R.; Conzelmann, E. Purification and proper-
ties of an R-methylacyl-CoA racemase from rat liver. Eur. J. Biochem.
1994, 222, 313-323.
(13) Schmitz, W.; Albers, C.; Fingerhut, R.; Conzelmann, E. Purification
and characterization of an R-methylacyl-CoA racemase from human
liver. Eur. J. Biochem. 1995, 231, 815-822.
(22) Mobley, J. A.; Leav, I.; Zielie, P.; Wotkowitz, C.; Evans, J.; Lam,
Y. W.; L’Esperance, B. S.; Jiang, Z.; Ho, S. Branched fatty acids in
dairy and beef products markedly enhance R-methylacyl-CoA
racemase expression in prostate cancer cells in vitro. Cancer
Epidemiol., Biomarkers PreV. 2003, 12, 775-783.
(23) Rasmussen, J. T.; Borchers, T.; Knudsen, J. Comparison of the
binding affinities of acyl-CoA-binding protein and fatty-acid-binding
protein for long-chain acyl-CoA esters. Biochem. J. 1990, 265, 849-
855.
(24) Wanders, R. J.; Romeyn, G. J.; Schutgens, R. B.; Tager, J. M.
L-Pipecolate oxidase: A distinct peroxisomal enzyme in man.
Biochem. Biophys. Res. Commun. 1989, 164, 550-555.
JM0702377