Potent and selective biphenyl azole inhibitors of adipocyte fatty acid binding protein (aFABP)

Article abstract of DOI:10.1016/j.bmcl.2006.12.044

Herein we report the first disclosure of biphenyl azoles that are nanomolar binders of adipocyte fatty acid binding protein (aFABP or aP2) with up to thousand-fold selectivity against muscle fatty acid binding protein and epidermal fatty acid binding prot

Full text of DOI:10.1016/j.bmcl.2006.12.044

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3511–3515
Potent and selective biphenyl azole inhibitors of adipocyte fatty acid
binding protein (aFABP)
Richard Sulsky,a,* David R. Magnin,a Yanting Huang,a Ligaya Simpkins,a
Prakash Taunk,a Manorama Patel,a Yeheng Zhu,a Terry R. Stouch,b
Donna Bassolino-Klimas,b Rex Parker,c Thomas Harrity,c Robert Stoffel,c
David S. Taylor,c Thomas B. Lavoie,d Kevin Kish,d Bruce L. Jacobson,e
Steven Sheriff,d Leonard P. Adam,c,* William R. Ewinga and Jeffrey A. Robla
aDepartment of Metabolic Disease Chemistry, Bristol Myers-Squibb Pharmaceutical Research Institute, PO Box 5400,
Princeton, NJ 08543-5400, USA
bDepartment of Computationally Assisted Drug Discovery, Bristol Myers-Squibb Pharmaceutical Research Institute,
PO Box 5400, Princeton, NJ 08543-5400, USA
cDepartment of Metabolic Disease Biology, Bristol Myers-Squibb Pharmaceutical Research Institute, PO Box 5400,
Princeton, NJ 08543-5400, USA
dDepartment of Macromolecular Structure, Bristol Myers-Squibb Pharmaceutical Research Institute, PO Box 5400,
Princeton, NJ 08543-5400, USA
eSt. Cloud State University, Department of Biological Sciences, 262 Wick Science Building 720, 4th Avenue South,
St. Cloud, MN 56301-4498, USA
Received 14 November 2006; revised 12 December 2006; accepted 13 December 2006
Available online 21 December 2006
Abstract—Herein we report the first disclosure of biphenyl azoles that are nanomolar binders of adipocyte fatty acid binding protein
(aFABP or aP2) with up to thousand-fold selectivity against muscle fatty acid binding protein and epidermal fatty acid binding pro-
tein. In addition a new radio-ligand to determine binding against the three fatty acid binding proteins was also synthesized.
Ó 2007 Elsevier Ltd. All rights reserved.
Adipocyte fatty acid binding protein (aFABP, aP2) is a
14.6 kDa cytosolic protein located in adipocytes and
macrophages, and assists in the intracellular transport
of fatty acids.1a,b–d It is one of a class of fatty acid binding
proteins (FABPs) that are found predominately in the liv-
er, heart, intestine, and connective tissues. Hotamisligal
et al. have reported that aFABP-deficient mice, when
placed on a high fat diet (40% of caloric intake as fat),
were significantly protected from hyperinsulinemia and
insulin resistance compared to the wild type.2 Additional
genetic experiments have been reported in which aFABP
null mice have been crossed with ob/ob and in another in-
stance apoEꢀ/ꢀ mice. The aFABP-deficient ob/ob mice
were more insulin-sensitive when compared to ob/ob con-
trols as demonstrated by measuring circulating glucose
and insulin levels.3 While apoEꢀ/ꢀ animals developed se-
vere atherosclerosis on a high fat diet, the apoEꢀ/ꢀ
/
aFABPꢀ/ꢀ progeny have lesions that are much smaller
and less complex, suggesting a role for aP2 in atheroscle-
rosis.4 Based on these genetic knock-out models, we pur-
sued the development of inhibitors of aFABP for their
therapeutic potential in the treatment of diabetes, obesi-
ty, and atherosclerosis.
Structurally, aFABP is a 10-stranded, 136-amino acid,
2-sheeted anti-parallel b-barrel with two short helixes
forming a ‘cap’. The protein contains essentially one
large cavity with a volume of approximately 950 cubic
angstroms. From crystallographic data of ligands bound
to the protein, the interior amino acids that are known
to specifically bind to endogenous fatty acids include
Keywords: Fatty acid binding protein; aP2; aFABP; Biphenyl azoles;
Radioligand assay.
*
Corresponding author. Tel.: +609 818 4743; e-mail: sulskyr@bms.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.044

3512
R. Sulsky et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3511–3515
Arg126 and Tyr128 (Fig. 2), which usually ligate an
acidic moiety directly, and Arg106, which often ligates
via a water bridge. A number of hydrophobic residues
also line the pocket. An unusual feature of this pocket
is that, dependent on substrate, it typically binds nine
or ten water molecules tightly. The bound water mole-
cules form part of the surface against which ligands
contact.5
substrates of aFABP using a homology model of the hu-
man adipocyte protein generated from the mouse adipo-
cyte protein crystal structure5 (PDB code 1LIE, 91%
identity, 95% similarity) and the LOOK program,8 led
to the identification of 1a, a 58 nM inhibitor of aFABP
that is essentially equipotent against the mFABP (Fig. 1
and Table 1). Similarity searches using 1a as a template
resulted in the identification of 1b (Fig. 1) as a potent
inhibitor of aFABP with 170-fold selectivity over
mFABP (Table 1). The profile of compound 1b is inter-
esting in that it binds with significantly greater affinity to
aFABP and has better isoform selectivity as compared
to known endogenous substrates such as palmitic or ole-
ic acid (Table 1) and the initial screening hit 1a.
The ubiquity of endogenous fatty acids in living systems
and the high intracellular concentration of aFABP (of
the order of micromolar)1a,b–d implied that we needed
an inhibitor of significantly greater intrinsic potency
than endogenous fatty acids. In addition, muscle FABP
(mFABP), which is expressed in both heart and skeletal
muscle, has been correlated to fatty acid oxidation
capacity in these tissues, and is necessary for their prop-
er function.6 Thus selectivity versus mFABP is thought
to be critical for an aFABP inhibitor as a therapeutic
agent. Epidermal FABP (eFABP) was also used as a
measure of selectivity in our assay protocol. eFABP is
widely distributed in many tissue and organs including
adipose, macrophages, stomach, heart, brain, liver,
spleen, muscle, lung, intestine, bone marrow, testis, ret-
ina, and mammary and endothelial cells.1d Because of
the high intracellular concentration of aFABP, its high
affinity for endogenous ligands, and need for selectivity
versus other FABPs, our program objectives were to dis-
cover potent aFABP inhibitors with selectivity over
mFABP and eFABP.
The weaker binding affinity of 1b compared to 1a for
mFABP suggested to us that the biphenylazole chemo-
type could provide a useful platform for obtaining selec-
tivity for aFABP over other FABP isoforms. Ligand 1b
contains a highly substituted oxazole ring and a distal
carboxylic acid bridged together by a biphenyl core.
This manuscript highlights the in vitro structure–activity
relationships (SAR) of several alternate heterocyclic ring
replacements for the oxazole in 1b and the resulting
FABP selectivity profiles. Compounds were assayed
against aFABP, eFAPB, and mFABP to determine
inhibition constants and selectivity. The 1,8-ANS
(8-anilino-1-naphthalene-sulfonic acid) displacement
assay developed by Kurian et al. was used to determine
binding affinity (Table 1).9b
Directed screening (via DOCK software)7 of our com-
pound library based on known endogenous fatty-acid
Our initial approach to explore the SAR of 1b was to
replace the oxazole with an imidazole. The imidazole
allows for additional functionalization via N-alkylation
to provide a new set of ligands. The facile method for
the synthesis of substituted imidazoles is shown in
Scheme 1. Imidazoles 4a and 4b were less active vs
aFABP when compared to 1b, but maintained selectivity
against m- and eFABP. When N-substituents such as
methyl (4c and 4d) or ethyl (4e and 4f) were added to
the heterocyclic core, improved binding to aFABP was
achieved, with Ki values in the range of 2 nM (the lower
O
N
N
O
HO
O
O
HO
O
O
1a
1b
Figure 1. Virtual screening leads for aFABP binding.
Table 1. Ki (nM) Binding constantsa of Oxazoles (1) and Imidazoles (4) with aFABP, eFABP, and mFABP in the 1,8-ANS assay compared to
endogenous fatty acids
Compound
X
R
aFABP Ki (nM)
mFABP Ki (nM)
eFABP Ki (nM)
Palmitic acid
Oleic acid
—
—
—
—
336 ( 164)
185 ( 35)
58 ( 4)
6 ( 1)
NTb
NT
802 ( 336)
248 ( 12)
>2000
1a
1b
4a
4b
4c
4d
4e
4f
—
—
—
—
32 ( 4)
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>2000
>2000
O
H
9.3 ( 0.6)
18 ( 6)
<2c
NH
O
H
Me
>2000
830 ( 100)
>2000
NH
O
Me
Et
2.7 ( 0.5)
<2c
830 ( 70)
>2000
>2000
NH
NH
NH
O
Et
3.4 ( 0.3)
130 ( 6)
>2000
<2c
4g
4h
4i
Isobutyl
CH2COOH
FCH2CH2
>2000
>2000
a Values are means of three experiments, standard deviation is given in parentheses.
b NT, not tested.
c Resolution of the assay does not allow for Ki quantification below 2 nM.

R. Sulsky et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3511–3515
3513
X
Et
Br
Ph
O
N
N
Ph
Ph
N
X
O
N
O
Ph
NH
O
Ph
a, b, c
COOH
9
N
Ph
H
2
32-74%
O
Br
3
11a
X = NH or O
c,d,e
a,b
X
Et
Ph
Ph
f,g
d, e
54-82%
O
N
N
N
HO
Ph
N
N
Br
Ph
N
R
Ph
O
Ph
4
OH
Et
h,i
11b
OH
Scheme 1. Reagents and conditions: (a) Pd(PPh3)4, 2-formylphenylbo-
ronic acid, Na2Co3, EtOH/toluene; (b) NH4OAc, Benzyl, HOAc,
reflux; (c) H2SO4, EtOH, rt; (d) alkyl halide, NaHCO3, DMF; (e)
NaOH, H2O, then H3O+.
10
Et
N
N
Ph
Ph
OH
11c
O
detection limit for the ANS assay).9 Yet when the larger
iso-butyl (4g) group was appended to the imidazole, a
>50-fold loss in binding to aFABP was observed. Polar
substituents (e.g., 4h, R = CH2COOH) markedly de-
creased activity. The most potent compounds (4c, 4e,
and 4i) achieved high selectivity ratios over eFABP
and mFABP (Table 1).
Scheme 3. Reagents and conditions: (a) NaH, Propionic anhydride/
NMP, 72%; (b) NaH/DMF, 82%; (c) Pd(PPh3)4, 3-methoxyphenylbo-
ronic acid, 90%; (d) BBr3/CH2Cl2, 90%; (e) i—BrCH2CO2Et, NaH/
DMF, 88%; ii—NaOH, H2O, H3O+, 88%; (f) Pd(PPh3)4, 3-form-
ylphenylboronic acid, 35%; (g) i—TMSCN, Et3N; ii—HCl (concd),
100 °C, 42%; (h) Pd(PPh3)4, 3-carbomethoxyphenylboronic acid, 30%;
(i) LiOH, 90%.
With the understanding that the central oxazole in 1b
could be replaced with an imidazole and maintain good
binding affinity, our attention turned to examining addi-
tional heterocycles, with a view to varying both the elec-
tron density of the core and subtly influencing the
spatial relationship of the peripheral substituents. We
prepared electron-rich heterocycles from the common
intermediate diketone 6.10 Furans were generated by
dehydration of 6 with boron trifluoride, pyrroles by
weak acid catalysis in the presence of amines, and thi-
ophenes by reaction of 6 with Lawesson’s reagent
(Scheme 2).11
oles maintained selectivity for aFABP over mFABP
and eFABP.
Additional analogs of 11a were prepared (Scheme 3) to
probe activity relative to the position of the carboxylic
acid group. As shown in Table 2, the a-hydroxy acid
11b was found to be 10-fold less active for aFABP when
compared to 11a. The benzoic acid 11c showed an even
greater loss in aFABP binding. This SAR suggests that
the position of the carboxylic acid group for the chemo-
type is critical for aFABP binding. Interestingly, these
modifications had negligible impact on the binding affin-
ity to m- and eFABPs.
The comparatively electron-poor pyrazole 11a was pre-
pared in 5 steps from hydrazone 9 (Scheme 3).12 The
biological results for compounds of chemotypes 8 and
11a (Table 2) suggest a similar SAR to the imidazoles.
For these chemotypes, the ethylated compounds 8c
and 11a were more active than the corresponding unsub-
stituted analogs (8a and 8b). Both oxazoles and pyraz-
Subsequent X-ray crystallographic analysis of 11a
bound to aFABP (Fig. 2) suggested that the ethyl at-
tached to the heterocyclic ring comes in close proximity
to Ser53. The residue Ser53 appears to play a role in the
intrinsic binding affinity of the azoles to aFABP. Substi-
tution with increasingly large groups (lone pair, H, Me,
Et) may enhance binding to aFABP through improved
van der Waals interactions with Ser53. Groups signifi-
cantly larger (e.g., 4g, isobutyl) or more polar (e.g., 4h
and 4i) than ethyl lead to reduced binding affinities.
Ph
Ph
O
R
Ph
O
Br
O
Br
b, c or d
a
R
5
6
Several of our compounds shown in Tables 1 and 2 test-
ed the limits of the fluorescence-based assay that we had
relied upon for much of our work. Resolution of Ki val-
ues in this 1,8-ANS assay is limited to the range of 2–
2000 nM.9a,b This arises from the relatively low fluores-
cent sensitivity of ANS, necessitating a high protein level
in the assay. As the lower limit of Ki measurement is
directly dependent on the protein concentration, this
in itself limits the assay sensitivity. In addition, the
1,8-ANS assay is subject to compound interference by
light absorption, because ANS fluoresces at 460 nm.
Because of these complications, a radioligand binding
assay was developed. The non-tritiated derivative of
Ph
O
Y
Br
Ph
Ph
e, f
Y = N, O or S
O
Y
HO
Ph
R
R
7
8
Scheme 2. Reagents and conditions: (a) PhCHO, N-3, 4-dimethyl-5-(2-
hydroxyethyl)thiazolium iodide, Et3N, 90%; (b) BF3–Et2O, Y@O, 58–
78%; (c) NH4OAc, or RNH2, HOAc Y@N–R0, 47–68%. (d) Lawes-
son’s reagent, Y@S, 85–100%; (e) i—Pd(PPh3)4, 4-methoxyphenylbo-
ronic acid, 63–84%; ii—BBr3, CH2Cl2, ꢀ78 °C 66–81%; (f) i—NaH,
ethyl bromoacetate, THF, 44–69%; ii—NaOH, H2O, then 1 N HCl,
95%.

3514
R. Sulsky et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3511–3515
Table 2. Fatty acid protein binding constants of Furans, Pyrroles, Thiophenes, and Pyrazoles in the 1,8-ANS assay
O
Ph
O
Y
HO
Ph
R
8
Compound
Y
R
aFABP Ki (nM)
mFABP Ki (nM)
eFABP Ki (nM)
8a
8b
O
NH
H
H
Et
36 ( 3)
12 ( 1)
3.5 ( 0.7)
<2b
>1000
>1000
>2000
>2000
8c
O
220 ( 20)
250 ( 15)
360 ( 50)
110 ( 5)
290 ( 60)
350 ( 3)
550 ( 70)
240 ( 10)
11a
11b
11c
(Scheme 3)
(Scheme 3)
(Scheme 3)
28 ( 5)
650 ( 60)
aValues are means of three experiments, standard deviation is given in parentheses.
b Resolution of the assay does not allow for Ki quantification below 2 nM.
Cl
Cl
I
I
Cl
a, b
Cl
Cl
I
O
O
87%
Cl
OHC
13
12
COOH
O
O
c, d
48%
Cl
Cl
T
T
14
Cl
Cl
Scheme 4. Reagents: (a) NBS, AIBN; (b) K2CO3, ethanol, 2,3-
dihydroxybenzaldehyde; (c) H2NSO3H, NaClO2, THF–water; (d) T2,
ethanol, KOAc, 10% Pd/C.
Figure 2. X-ray crystal structure of 11a bound to the active site of
aFABP.13a The X-ray model of 11a bound to the binding site of
aFABP with the initial 2Fo-Fc electron density map contoured at 1r.
Hydrogen bonds are shown as small spheres. For clarity only a limited
number of surrounding residues are included in the figure. In
particular, Ser53, which is a Thr in m- and eFABP, is proximal
that Ki values quantified in the fluorescent assay are
potentially in error by a factor of 3 to 4 and the signif-
icant differences in assay format,9a,b variation in calcu-
lated Ki values between the two assays was expected.
In addition, X-ray crystallographic analysis shows that
ANS and analogs of 14 bind in distinctly different posi-
tions. These differences in binding orientation suggest
another possible reason to expect differences in Ki.
˚
(3.8 A) to the ethyl substituent of the pyrrozole ring. Figure prepared
with PyMol.13b
benzoate 14 was identified as a potent aFABP inhibitor
from other SAR studies.14 We selected compound 14 as
the radioligand since the non-labeled compound demon-
strates potent binding in the 1,8-ANS assay for several
FABPs including aFABP (Ki = 2 nM), eFABP
(Ki = 2 nM), and mFABP (Ki = 4 nM), and is therefore
of potential use as a probe for binding to each of these
proteins. Compound 14 was generated by the method
outlined in Scheme 4 in four steps starting from the
commercially available iodide 12. The aFABP Kd of
14 was determined by Scatchard analysis and found to
be 21 nM ( 2).15
Table 3. Binding constants of selected compounds in the radioligand
assay and comparison to 1,8-ANS displacement assay
Compound
1,8-ANS assay Ki (nM)
3H 14 assay Ki (nM)b
1b
4b
4c
6 ( 1)
18 ( 6)
<2a
95 (1)
300 (1)
13 (1)
48 (1)
3.3 (2)
13 (2)
8.7 (1)
7.4 (1)
32 (1)
1000
4d
4e
2.7 ( 0.5)
<2a
4f
3.4 ( 0.3)
<2a
<2a
4i
11
8c
3.5 ( 0.7)
Kd = 150
Selected compounds were assayed against aFABP and
the results were compared to the original 1,8-ANS assay
data (Table 3). Within the set of compounds tested, Ki
values were greater when determined with the radioli-
gand binding assay than with the 1,8-ANS assay. Given
ANS
a Resolution of the assay does not allow for Ki quantification below
2 nM.
b Values in parentheses refer to number of experiments.

R. Sulsky et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3511–3515
3515
tion with varying [ANS]. Ki values, therefore, were
calculated according to the following (variant of Equation
3a in Kurian et al.): Ki = IC50KANS/(A ꢀ 0.5P ꢀ
KANS) ꢀ C. C is constant for any given assay condition
and is specified as: C = 0.5PKANS(1 ꢀ KANS/(A ꢀ 0.5P))/
(A ꢀ 0.5P ꢀ KANS) A and P refer to total [ANS] and
[FABP] used in the assay. The assay was run in 384-well
format using 10 lM ANS and 0.5 lM FABP. In addition,
assay [ANS] was verified in each plate by the addition of
saturating protein to determine maximal fluorescence, in
comparison to an ANS standard. In this manner, Ki errors
were minimized because (i) [ANS] was significantly greater
than KANS, (ii) Ki values were minimally affected by
variations in assay protein concentration, and (iii) IC50
values better reflected compound effects than inhibition
with a single concentration. The Kd(ANS) (KANS) used in
all calculations for the 1,8-ANS assay was 150 nM.
However, experimentally determined Kd values were
found to be highly dependent on the concentration of
aFABP used in the assay. Calculated Kd values increased
to approximately 500 nM as the [aFABP] decreased to
50 nM. The use of concentrations of aFABP below this
value was not technically feasible. Because of this, Ki
values reported for the 1,8-ANS assay (Tables 1–3) are
likely to be lower, up to a factor of 3 or 4, than what may
be their true affinity for the lipid binding site; (c) van
Dongen, M. V. P.; Uppenberg, J.; Svensson, S.; Lun-
Because the Ki values in the ANS fluorescent assay are
highly variable in the low nanomolar range, the radioli-
gand binding assay offers a significant advantage in sen-
sitivity for tight binding ligands.
In conclusion, we have identified a novel structural class
of compounds that bind to aFABP with significantly
greater affinity and FABP isoform selectivity than
known endogenous fatty acid substrates. In particular
we have identified several compounds which are potent
(nM) and selective inhibitors of aFABP. These ligands
can serve as useful probes for further investigation of
the potential utility of aFABP inhibitors for the treat-
ment of diabetes, obesity, and atherosclerosis. Other ser-
ies of aFABP inhibitors have been reported in the
literature by Biovitrum.16a,b In comparison, the inhibi-
tors described in this manuscript are of greater potency.
We have also described a new and useful radioligand
binding assay for the determination of binding constants
for aFABP (and potentially m- and eFABP) with a
greater dynamic range than the standard fluorescence
assay and capable of distinguishing the SAR of our most
active compounds.
˚
dba¨ck, T.; Akerud, T.; Wikstro¨m, M.; Schultz, J. J. Am.
Chem. Soc. 2002, 124, 11874.
References and notes
10. Stetter, H.; Simons, L. Chem. Ber. 1985, 118, 3172–3187.
11. Robl, J. A.; Sulsky, R. B.; Magnin, D. R. U.S. Patent
6,548,529, 2003; . Chem. Abstr 2000, 133, 296346.
12. Bondiou, J. C.; Legrand, J. J. Eur. Pat. Appl. 70231, 1983;
Chem. Abstr. 1983, 99, 38454.
13. (a) PDB Deposition ID code is 2NNQ; (b) DeLano, W. L.
The PyMol Molecular Graphics System 2002. DeLano
Scientific, San Carlos, CA, US, http:/www.pymol.org.
14. (a) McDonnell, P. A.; Constantine, K. L.; Goldfarb, V.;
Johnson, S. R.; Sulsky, R.; Magnin, D. R.; Robl, J. A.;
Caulfield, T. J.; Parker, R. A.; Taylor, D. S.; Adam, L. P.;
Metzler, W. J.; Mueller, L.; Farmer, B. T. J. Med. Chem.
2006, 49, 5013; (b) Magnin, D. R.; Sulsky, R. B.; Robl, J.
A.; Caulfield, T. J.; Parker, R. A. Dual Inhibitors of
Adipocyte Fatty Acid Binding Protein and Keratinocyte
Fatty Acid Binding Protein. W.O. Patent 03,043,624,
2003.
1. (a) Banaszak, L.; Winter, N.; Xu, Z.; Bernlohr, D. A.;
Cowan, S.; Jones, T. A. Adv. Protein Chem. 1994, 45, 89;
(b) Bernlohr, D. A.; Doering, T. L.; Kelly, T. J., Jr.; Lane,
M. D. Biochem. Biophys. Res. Commun. 1985, 132, 850; (c)
Zimmerman, A. W.; Veerkamp, J. H. Cell. Mol. Life Sci.
2002, 59, 1096; (d) Hertzel, A. V.; Bernlohr, D. A. Trends
Endrocrinol. Metab. 2000, 11, 175.
2. Scheja, L.; Makowski, L.; Uysal, K. T.; Wiesbrock, S. M.;
Shimshek, D. R.; Meyers, D. S.; Morgan, M.; Parker, R.
A.; Hotamisligil, G. S. Diabetes 1999, 48, 1987.
3. Uysal, K. T.; Scheja, L.; Wiesbrock, S. M.; Bonner-Weir,
S.; Hotamisligil, G. Endocrinology 2000, 141, 3388.
4. (a) Layne, M. D.; Patel, A.; Chen, Y. H.; Rebel, V. I.;
Carvajal, I. M.; Pellacani, A.; Ith, B.; Zhao, D.; Schreiber,
B. M.; Yet, S. F.; Lee, M. E.; Storch, J.; Perrella, M.
FASEB J. 2001, 15, 2733; (b) Makowski, L.; Boord, J. B.;
Maeda, K.; Babaev, V. R.; Uysal, K. T.; Morgan, M. A.;
Parker, R. A.; Suttles, J.; Fazio, S.; Hotamisligil, G. S.;
Linton, M. F. Nature 2001, 7, 699.
5. LaLonde, J. M.; Bernlohr, D. A.; Banaszak, L. J.
Biochemistry 1994, 33, 4885.
6. Glatz, J. F. C.; van Breda, E.; Van Der Vusse, G. J. Adv.
Exp. Med. Biol. 1998, 441, 207.
7. Ewing, J. A. T.; Kuntz, I. D. J. Comput. Chem. 1997, 18,
1175.
8. (a) Levitt, M. J. Mol. Biol. 1992, 226, 507; (b) Levitt, M.
J. Mol. Biol. 1992, 170, 723.
15. 3H-14 (39 Ci/mmol) was used in a binding assay based on
the methods described in Vork et al. (Mol. Cell. Biochem.
1990, 98, 111). Briefly, free radioligand was separated
from bound radioligand using hydroxyalkoxypropyl dex-
tran beads (Sigma) that do not bind protein. The Kd(3H-
14) (K3H-14) was determined by varying the concentration
of radioligand in the assay, and Ki values for ligands were
calculated based on the correction of Cheng and Prusoff
(Biochem. Pharmacol. 1973, 22, 3099): Ki = IC50/(1 +
[3H-14]/K3H-14).
16. (a) Ringom, R.; Axen, E.; Uppenberg, J.; Lundbaeck,
T.; Rondahl, L.; Barf, T. Bioorg. Med. Chem. Lett.
2004, 14, 4449; (b) Lehmann, F.; Haile, S.; Axen, E.;
Medina, C.; Uppenberg, J.; Svensson, S.; Lundbaeck,
T.; Rondahl, L.; Barf, T. Bioorg. Med. Chem. Lett.
2004, 14, 4445, These compounds, benzylamino-6-(tri-
fluoromethyl)pyrimidine-4(1H)-ones were reported to
have IC50 > 0.5 lM.
9. (a) Kurian, E.; Kirk, W. R.; Prendergast, F. G. Biochem-
istry 1996, 35, 3865; (b) Ki values in the 1,8-ANS assay
were calculated based on a variation of the methodology
presented in Kurian et al.9a The main methodological
difference involved the quantification of Ki values from an
experimentally determined IC50, instead of quantification
based on the inhibition at a single compound concentra-