Design and synthesis of phthalimide-based fluorescent liver X receptor antagonists

Article abstract of DOI:10.1248/cpb.58.1525

Based on our structure-activity relationship study of liver X receptor (LXR) ligands, we designed and synthesized fluorescent LXR antagonists containing an unsubstituted or substituted amino group on a phthalimide unit.

Full text of DOI:10.1248/cpb.58.1525

November 2010
Note
Chem. Pharm. Bull. 58(11) 1525—1528 (2010)
1525
Design and Synthesis of Phthalimide-Based Fluorescent Liver X Receptor
Antagonists
a
b
b
,a
*
Kenji KISHIDA, Atsushi AOYAMA, Yuichi HASHIMOTO, and Hiroyuki MIYACHI
a Division of Pharmaceutical Sciences, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama
University; 1–1–1 Tsushima-Naka, Kita-ku, Okayama 700–8530, Japan: and b Institute of Molecular and Cellular
Biosciences, University of Tokyo; Yayoi, Bunkyo-ku, Tokyo 113–0032, Japan.
Received May 29, 2010; accepted August 6, 2010
Based on our structure–activity relationship study of liver X receptor (LXR) ligands, we designed and syn-
thesized fluorescent LXR antagonists containing an unsubstituted or substituted amino group on a phthalimide
unit.
Key words liver X receptor; liver X receptor antagonist; fluorescent ligand
Liver X receptors (LXR) are members of the nuclear re- tive depends on the three-dimensional geometry of the two
ceptor superfamily, and are involved in the regulation of cho- phenyl ring moieties. Based on this concept, we performed
lesterol, lipid, and glucose metabolism.1,2) Two subtypes, further structural modification of T0901317, and found that
LXRa and LXRb, which have characteristic distribution pat- other types of conformationally restricted analogs of
terns, are known so far. LXRa is expressed abundantly in dibenz[b,f ][1,4]oxazepin-11-one, 11,12-dihydrodibenz[b,f ]-
liver, adipose tissue and macrophages, while LXRb is ubiq- azocin-6-one, and (Z)-dibenz[b,f ]azocin-6-one derivatives,
uitously distributed in organs and tissues. The endogenous exhibited LXR-agonistic activity.10)
ligands of LXRs are reported to be cholesterol metabolites
As a part of our continuing structural development of sub-
(oxysterols), such as 22(R)-hydroxycholesterol (1) and type-selective LXR ligands, we required a rapid and simple
24(S),25-epoxycholesterol (2) (Fig. 1).3) In macrophages, method to determine LXR-binding activity without using
liver, and intestine, activation of LXRs induces expression of isotope-labeled material(s). Here, we report the design and
genes involved in cholesterol homeostasis.4) It may be possi- the synthesis of fluorescent LXR antagonists suitable for use
ble to prevent or even reverse atherosclerosis by modulating in a fluorescent ligand binding assay system for the identifi-
the expression or activity of LXRs, which therefore represent cation of endogenous and exogenous LXR ligands.
attractive potential therapeutic targets. Much research has
We have already reported structurally simple LXR antago-
been done to discover LXR ligands suitable for the treatment nists with a phthalimide skeleton, such as 5 and 6 (Fig.
of metabolic disorders, as exemplified by the representative 1).6—8) The phthalimide structure is not fluorescent itself, but
compound T-090317 (3) (Fig. 1).5)
the introduction of an amino group at the benzene ring of ph-
We have studied the design and synthesis of ligands of var- thalimide was reported to be fluorogenic, probably due to in-
ious nuclear receptors, such as peroxisome proliferator-acti- tramolecular charge transfer.11,12) Based on these insights, we
vated receptor (PPAR) and farnesoid X receptor (FXR), and anticipated that if such amino group-bearing phthalimide de-
recently we expanded our focus to include LXR.6—8) We have rivatives exhibit LXR activity, we would be able to obtain
also demonstrated that the 5-substituted phenanthridin-6-one fluorescent LXR ligands.
derivative (4), which is a conformationally restricted hetero-
Chemistry 11a—d were prepared from 2- (or 4)-ni-
cyclic analogue of T0901317, lacks LXR-agonistic activity, trobenzyltriphenylphosphonium bromide (7), which was it-
but shows LXR-antagonistic activity.9) This prompted us to self prepared by the reaction of 2- (or 4)-nitrobenzyl bromide
speculate that the nature of the LXR-modulating activity with triphenylphosphine. Wittig reaction of 7 with benzalde-
elicited by the conformationally restricted T0901317 deriva- hyde derivative in the presence of NaOMe as a base, and sub-
Fig. 1. Structures of Known LXR Ligands
∗ To whom correspondence should be addressed. e-mail: miyachi@pharm.okayama-u.ac.jp
© 2010 Pharmaceutical Society of Japan

1526
Vol. 58, No. 11
sequent reduction with 10% Pd–C, afforded the aniline deriv- nometer and a microplate reader. DNA cotransfection experi-
atives (9a, b). 9a, b were treated with nitrophthalic anhydride ments included 50 ng of reporter plasmid, 20 ng of pCMX®-
in acetic acid to afford 10a—d. Hydrogenation of 10a—d b-galactosidase, 15 ng of each LXR receptor expression plas-
with 10% Pd–C afforded the desired amino derivatives mid, and pGEM® carrier DNA to make a total of 150 ng of
(11a—d). The amino group of 11a—d was methylated with DNA per well in a 96-well plate. Luciferase data were nor-
trioxane in the presence of NaBH(OAc)3 as a reducing agent malized to an internal b-galactosidase control and reported
to afford N,N-dimethyl derivatives (12a—d) and N- values are the means of triplicate assays.
monomethyl derivatives (13a—d).13)
Fluorescence Measurements Fluorescence spectra were
Cell Culture and Cotransfection Assay Human embry- measured at 25 °C on a HITACHI F2500 spectrofluoro-Pho-
onic kidney HEK293 cells were cultured in Dulbecco’s modi- tometer with a 1-cm path-length quartz cell.
fied Eagle’s medium (DMEM) containing 5% fetal bovine
serum and antibiotic–antimycotic mixture (Nacalai) at 37 °C Results and Discussion
in a humidified atmosphere of 5% CO2 in air. Transfections
As can be seen from Table 1, the position and shape of the
were performed by calcium phosphate coprecipitation. Eight amino substituents located on the benzene ring of the synthe-
hours after transfection, LXR full agonist, T0901317 with or sized compounds is important for the inhibitory activity to-
without the test compound were added. Cells were harvested wards both LXR subtypes. In the case of 5 derivatives, intro-
approximately 16—20 h after the treatment, and luciferase duction of an unsubstituted amino group at the 4- and 5-posi-
and b-galactosidase activities were assayed using a lumi- tions did not result in apparent LXRa- or LXRb-inhibitory
Reagents and conditions: (a) benzaldehyde, MeONa, dehydrated THF, rt., 3 d, 80—90%; (b) H2, 10% Pd–C, AcOEt, rt, 1 h, 74—89%; (c) nitrophthalic anhydride,
AcOH, reflux, 3 h, quant.; (d) H2, 10% Pd–C, AcOEt, rt, 1 h, 64%-quant.; (e) trioxane, NaBH(OAc)3, AcOH, rt., 3—7 d, 12a—d 6—61%, 13a—d 16—61%.
Chart 1. Synthesis of the Present Series of Amino-Substituted Phthalimides
Table 1. LXR-Antagonistic Activities of the Amino-Substituted Phthalimides
IC50 (mM)
IC50 (mM)
Compd.
NR1R2
Compd.
NR1R2
LXR a
LXR b
LXR a
LXR b
11a
13a
12a
11b
13b
12b
4-NH2
Ͼ30
2.9
1.6
Ͼ30
6.9
Ͼ30
10
21
Ͼ30
20
17
11c
13c
12c
11d
13d
12d
4-NH2
Ͼ30
Ͼ30
Ͼ30
1.8
Ͼ30
Ͼ30
Ͼ30
5.3
4-NHCH3
4-N(CH3)2
5-NH2
5-NHCH3
5-N(CH3)2
4-NHCH3
4-N(CH3)2
5-NH2
5-NHCH3
5-N(CH3)2
0.75
1.0
4.2
5.0
4.2
5
9.8
Ͼ30
6
0.2
Ͼ30

November 2010
1527
activity at the concentration of 30 mM. Methylation of the hibited non-selective LXR antagonistic activity. The presence
amino group caused the appearance of LXR-antagonistic ac- of a methyl group did not affect the LXR-antagonistic activ-
tivity. All the methylated derivatives (12a, 12b, 13a, 13b) ex- ity, i.e., all the 5-amino-substituted derivatives 11d, 12d, and
hibited more potent LXR antagonistic activity than the lead 13d exhibited equipotent non-selective LXR-antagonistic ac-
compound 5. They exhibited partial subtype preference for tivity. These structure–activity relationship studies indicated
LXRa, as did 5. However, in the 6 series, a somewhat differ- that the binding modes of amino-5 derivatives and amino-6
ent structure–activity relationship was obtained. All the 4- derivatives are somewhat different, even though these com-
substituted derivatives (11c, 12c, 13c) lacked LXR-antago- pounds have similar structural motifs.
nistic activity at the concentration of 30 mM, whereas the lead
Next, the absorption and fluorescence spectra of represen-
6 exhibited LXRa-selective antagonistic activity. On the tative LXR-positive compounds (12a, 12b, 11d, 12d) were
other hand, the 5-amino derivative (11d) exhibited apparent examined in dimethyl sulfoxide (DMSO), which is a co-sol-
LXR-antagonistic activity towards both subtypes; the LXRa- vent frequently used in bioassay and for preparing stock so-
antagonistic activity of 11d was decreased to some extent, lutions. The excitation and fluorescence spectra of 5 mM con-
while the LXRb-antagonistic activity was increased consid- centration of each compound were depicted in Fig. 2. In the
erably, as compared with those of 6. Consequently, 11d ex- case of 6 derivatives, the 5-unsubstituted amino derivative
Fig. 2. (A) Excitation and (B) Fluorescence Spectra of Amino-Substituted Phthalimides
Fig. 3. Effect of PBS on the Excitation and Fluorescence Spectra of (A) 11d and (B) 12d in DMSO

1528
Vol. 58, No. 11
Acknowledgement This research was supported in part by a Grant from
the “Okayama Medical Foundation Okayama University Medical School.”
(11d) exhibited excitation and emission spectra, with max-
ima at 385 nm and 495 nm, respectively. The 5-N,N-dimethyl-
amino derivative 12d exhibited weak excitation and emission
spectra, as compared to 11d, with maxima at 401 nm and
516 nm, respectively. The excitation and emission intensities
of 11d are about 6 greater than those of 12d. In the 5 series,
the 5-N,N-dimethylamino derivative 12b exhibited weak ex-
citation and emission as compared to 12d, with maxima at
404 nm and 515 nm, respectively. It is of interest that the 4-
N,N-dimethylamino derivative 12a did not exhibit apparent
excitation or emission (data not shown). The rank order of
the fluorescence intensity decreased in the order of 5-
NH2Ͼ5-N(Me)2ϾϾ4-N(Me)2. This prompted us to speculate
that methylation of the 5-amino group of phthalimide de-
creased both excitation and emission, probably due to loss of
the hydrogen-bonding nature of the amino group and/or
change of the electron donor–acceptor balance of the phthal-
imide ring.11) The shift of the N,N-dimethylamino group
from the 5-position to the 4-position of the phthalimide ring
decreased both excitation and emission, but the exact reason
unknown. Based on the spectral intensity, we selected the
amino-6 series for further study.
References and Notes
1) Alberti K. G., Zimmet P. Z., Diabetic Medicine, 15, 539—553 (1998).
2) Mukherjee R., Jow L., Noonan D., McDonnell D. P., J. Steroid
Biochem. Mol. Biol., 51, 157—166 (1994).
3) Lehmann J. M., Kliewer S. A., Moore L. B., Smith-Oliver T. A., Blan-
chard D. E., Spencer T. A., Willson T. M., J. Biol. Chem., 272, 3137—
3140 (1997).
4) Repa J. J., Turley S. D., Lobaccaro J. M. A., Medina J., Li L., Lustig
K., Shan B., Heyman R. A., Dietschy J. M., Mangelsdorf D. J.,
Science, 289, 1524—1529 (2000).
5) Schults J. R., Tu H., Luk A., Repa J. J., Medina J. C., Li L., Schwend-
ner S., Wang S., Thoolen M., Mangelsdorf D. J., Lustig K. D., Shan B.,
Genes Develop., 14, 2831—2838 (2000).
6) Noguchi-Yachide T., Aoyama A., Makishima M., Miyachi H.,
Hashimoto Y., Bioorg. Med. Chem. Lett., 17, 3957—3961 (2007).
7) Noguchi-Yachide T., Miyachi H., Aoyama A., Makishima M., Aoyama
H., Hashimoto Y., Chem. Pharm. Bull., 55, 1750—1754 (2007).
8) Dodo K., Aoyama A., Noguchi T., Makishima M., Miyachi H.,
Hashimoto Y., Bioorg. Med. Chem., 16, 4272—4285 (2008).
9) Aoyama A., Aoyama H., Dodo K., Makishima M., Hashimoto Y.,
Miyachi H., Heterocycles, 76, 137—141 (2008).
10) Aoyama A., Aoyama H., Makishima M., Hashimoto Y., Miyachi H.,
Heterocycles, 78, 2209—2216 (2009).
11) Soujanya T., Fessenden R. W., Samanta A., J. Phys. Chem., 100,
3507—3512 (1996).
For binding assay application, we have to use aqueous so-
lution, because the active conformation of LXR protein is 12) Soujanya T., Krishna T. S. R., Samanta A., J. Photochem. Photobiol.
A: Chem., 66, 185—192 (1992).
unstable in organic media. Therefore, we investigated the ef-
1
13) 11a; H-NMR (500 MHz, CDCl3) d: 7.49 (d, Jϭ8.4 Hz, 1H), 7.06—
fect of the addition of phosphate-buffered saline (PBS) (pH
7.40 (m, 10H), 6.92 (d, Jϭ8.4 Hz, 1H), 5.30 (s, 2H), 2.84 (m, 4H);
7.4). In the case of 11d, the excitation and emission intensity
decreased dramatically, especially in the case of its fluores-
cence, and exhibited red shift of fluorescence, depending on
the solvent polarity. That is, the excitation and emission max-
ima (nm) shifted from 385/495 (DMSO) to 393/527
(DMSO : PBSϭ1 : 1 v/v). The excitation and emission max-
ima (nm) of 11d in DMSO : PBSϭ1 : 7 v/v was undetectable.
This bathochromic shift was consistent with the reported re-
sult for an aza-crowned phthalimide.14) To the contrary, a
hypochromic shift of fluorescence was observed with in-
creasing solvent polarity in the case of 12d. The excitation
and emission maxima (nm) shifted from 401/516 (DMSO) to
403/469 (DMSO : PBSϭ1 : 1 v/v) and 403/470 (DMSO :
PBSϭ1 : 7 v/v). The excitation and emission intensities of
12d were also decreased in polar solvents, but the degree of
the decrease was less than in the case of 11d. The mecha-
nisms underlying these results remain unclear.
In conclusion, fluorescent LXR antagonists based on pre-
viously reported N-phenylphthalimide were developed. The
fluorescence intensity and emission maxima of the present
series of probes were dependent on environmental polarity.
As the binding site of LXR is highly hydrophobic, an in-
crease in fluorescence intensity is expected to occur upon
specific binding of the present series of compounds to
LXR,15) so 11d and 12d should be good candidate probes for
an LXR-binding assay system. Optimization of the assay for-
mat is in progress.
FAB-MS m/z: 343 (MϩH)ϩ. 11b; 1H-NMR (500 MHz, CDCl3) d: 7.74
(d, Jϭ8.4 Hz, 1H), 7.05—7.40 (m, 10H), 6.89 (dd, Jϭ8.4, 2.1 Hz, 1H),
4.40 (s, 2H), 2.80 (m, 4H); FAB-MS m/z: 343 (MϩH)ϩ. 11c; 1H-NMR
(500 MHz, CDCl3) d: 6.80—7.50 (m, 11H), 5.31 (s, 2H), 3.80 (s, 3H),
2.92 (m, 4H); FAB-MS m/z: 373 (MϩH)ϩ. 11d; H-NMR (500 MHz,
1
CDCl3) d: 6.80—7.50 (m, 11H), 4.40 (s, 2H), 3.80 (s, 3H), 2.90 (m,
4H); FAB-MS m/z: 373 (MϩH)ϩ. 12a; 1H-NMR (500 MHz, CDCl3) d:
7.57 (d, Jϭ5.1 Hz, 1H), 7.00—7.40 (m, 11H), 3.13 (s, 6H), 2.83 (m,
4H); FAB-MS m/z: 371 (MϩH)ϩ. 12b; 1H-NMR (500 MHz, CDCl3) d:
7.71 (d, Jϭ8.4 Hz, 1H), 7.05—7.40 (m, 10H), 6.80 (dd, Jϭ8.4, 2.4 Hz,
1H), 4.60 (s, 1H), 2.98 (d, Jϭ4.8 Hz, 3H), 2.83 (m, 4H); FAB-MS m/z:
371 (MϩH)ϩ. 12c; 1H-NMR (500 MHz, CDCl3) d: 6.84—7.54 (m,
11H), 3.80 (s, 3H), 3.13 (s, 6H), 2.90 (m, 4H); FAB-MS m/z: 401
(MϩH)ϩ. 12d; 1H-NMR (500 MHz, CDCl3) d: 7.74 (d, Jϭ8.4 Hz,
1H), 6.83—7.34 (m, 10H), 3.80 (s, 3H), 3.15 (s, 6H), 2.90 (m, 4H);
FAB-MS m/z: 401 (MϩH)ϩ. 13a; 1H-NMR (500 MHz, CDCl3) d: 7.56
(d, Jϭ8.4 Hz, 1H), 7.00—7.40 (m, 11H), 3.53 (q, Jϭ7.5 Hz, 1H), 3.05
(s, 3H), 2.81 (m, 4H); FAB-MS m/z: 357 (MϩH)ϩ. 13b; 1H-NMR
(500 MHz, CDCl3) d: 7.71 (d, Jϭ8.4 Hz, 1H), 7.05—7.40 (m, 10H),
6.80 (dd, Jϭ8.4, 2.4 Hz, 1H), 4.60 (s, 1H), 2.98 (d, Jϭ4.8 Hz, 3H),
2.83 (m, 4H); FAB-MS m/z: 357 (MϩH)ϩ. 13c; H-NMR (500 MHz,
1
CDCl3) d: 6.83—7.53 (m, 11H), 3.80 (s, 3H), 3.35 (q, Jϭ7.5 Hz, 1H),
3.09 (s, 3H), 2.90 (m, 4H); FAB-MS m/z: 387 (MϩH)ϩ. 13d; 1H-NMR
(500 MHz, CDCl3) d: 7.73 (d, Jϭ8.4 Hz, 1H), 6.82—7.35 (m, 10H),
3.80 (s, 3H), 3.55 (q, Jϭ7.5 Hz, 1H), 3.09 (s, 3H), 2.90 (m, 4H); FAB-
MS m/z: 387 (MϩH)ϩ.
14) Okamoto H., Kohno M., Satake K., Kimura M., Bull. Chem. Soc. Jpn.,
78, 2180—2187 (2005).
15) Soujanya T., Krishna T. S. R., Samanta A., J. Phys. Chem., 96, 8544—
8548 (1992).