Discovery and development of dimeric podocarpic acid leads as potent agonists of liver X receptor with HDL cholesterol raising activity in mice and hamsters

Article abstract of DOI:10.1016/j.bmcl.2005.03.100

Liver X receptors are nuclear receptors that regulate metabolism of cholesterol. They are activated by oxysterols resulting in increased transcription of the ABCA1 gene, promoting cholesterol efflux and HDL formation. We have identified podocarpic acid an

Full text of DOI:10.1016/j.bmcl.2005.03.100

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2824–2828
Discovery and development of dimeric podocarpic acid leads
as potent agonists of liver X receptor with HDL cholesterol
raising activity in mice and hamsters
Sheo B. Singh,* John G. Ondeyka, Weiguo Liu, Steve Chen, Tom S. Chen, Xiaohua Li,
Aileen Bouffard, James Dropinski, A. Brian Jones, Sherrie McCormick, Nancy Hayes,
Jianhua Wang, Neelam Sharma, Karen MacNaul, Melba Hernandez, Yu-Sheng Chao,
Joanne Baffic, My-Hanh Lam, Charlotte Burton, Carl P. Sparrow and John G. Menke
Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 10 January 2005; revised 7 March 2005; accepted 25 March 2005
Available online 27 April 2005
Abstract—Liver X receptors are nuclear receptors that regulate metabolism of cholesterol. They are activated by oxysterols resulting
in increased transcription of the ABCA1 gene, promoting cholesterol efflux and HDL formation. We have identified podocarpic acid
anhydride as a 1 nM agonist of LXRa and b receptors. Functionally this agonist was over 8–10-fold better activator of LXR recep-
tors compared to one of the natural ligands, 22-(R)-hydroxy cholesterol, in HEK-293 cells. An imide analog increased the level of
HDL by 26%, decreased LDL by 10.6%, and increased triglyceride by 51% in hamsters. Discovery, synthesis, SAR and details of the
activities of dimers have been described.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
including ABCA1. It has been shown that LXR agonists
cause increased expression of ABCA1 and raise the
HDL levels in mice. ABCA1 mediates the efflux of cho-
lesterol out of the cells and onto the ApoA1 protein of
HDL particles. Therefore, LXR agonists are expected
to provide an opportunity for the development of drugs
to increase reverse cholesterol transport and thus de-
crease the burden of atherosclerosis.8
The efficient regulation of cholesterol biosynthesis,
metabolism, acquisition, and transport is an essential
function of mammalian cells. Elevated plasma choles-
terol levels are a major risk factor correlated with coro-
nary heart disease and stroke. Liver X receptors (LXR)
are members of a superfamily of nuclear hormone recep-
tors represented by two subtypes, LXRa and LXRb.1–3
The a-subtype is predominantly present in liver whereas
the b-subtype is ubiquitously expressed. Oxysterols have
been identified as ligands for both subtypes4–6 and
27-hydroxy-cholesterol has been identified as a physio-
logically relevant endogenous ligand in human macro-
phages.7 These receptors have been shown to play a
role in cholesterol homeostasis.8 LXRs form heterodi-
mer with the retinoid X receptor (RXR) to regulate
the expression, directly or indirectly, of a number of
genes involved in cholesterol and fatty acid metabolism,
We screened our sample collection employing the ligand
binding domain of LXRa and b receptor in a scintilla-
tion proximity binding assays (LXRSPA)9 with the
radioactive synthetic ligand [3H2]-F3-methyl AA (1).
This allowed the identification of a screening hit identi-
fied as podocarpic acid (2) whose structure was con-
firmed by spectroscopic methods and by the
comparison with an authentic sample. However, signifi-
cant concern arose when several new and differently de-
rived samples of podocarpic acid turned out to be
inactive indicating that the confirmed activity of the ori-
ginal podocarpic acid sample may be due to a minor
contaminant. Therefore, about 20 mg of the originally
active sample of podocarpic acid was subjected to a bio-
assay-guided fractionation using LXRSPA assay. While
most of the traditional chromatographic techniques
Keywords: LXR; Atherosclerosis; Cholesterol lowering; Natural product
derivatives.
*
Corresponding author. Tel.: +1 732 594 3222; fax: +1 732 594
6880; e-mail: sheo_singh@merck.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.100

S. B. Singh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2824–2828
2825
such as silica gel and RPHPLC led to the loss of the
activity, gel permeation chromatography on Sephadex
LH20 in methanol led to the separation of the activity
away from podocarpic acid. The activity (0.5% by
weight) eluted significantly earlier than podocarpic acid,
suggesting a compound with higher molecular weight
and identified as podocarpic acid anhydride (3) (ESIMS
m/z 548 [M+NH4], IR mmax 1795 cmÀ1). This was con-
verted to a diacetate (4) by a reaction with acetic anhy-
dride. Its identity was confirmed by comparison (1H
NMR, ESIMS, IR) with an authentic synthetic sample
of podocarpic acid anhydride diacetate (4) prepared by
heating of podocarpic acid with acetic anhydride fol-
lowed by purification of both samples by Sephadex
LH20.
to DMSO control. The activity of APD was 1000-fold
more potent and afforded 8–10-fold higher maximal
stimulation than the natural ligand 22-(R)-hydroxy cho-
lesterol.10 Additionally APD induced the ABCA1
mRNA levels and increased efflux of cholesterol and
phospholipids from a number of cell types.10 The activ-
ity of APD was not dependent on covalent modification
of LXR. The steric factors lead to hydrolytic stability of
this anhydride. While APD 4 showed equal potent activ-
ity for a and b LXR it exhibited no activity against other
nuclear receptors (FXR, PPAR a, c, d, estrogen a and b,
glucocorticoid or thyroid hormone) at 10 lM, thus dis-
playing strong selectivity for LXR although no selecti-
vity between LXR subtypes. In order to further study
this lead and elucidate the SAR, we synthesized a series
of dimeric podocarpic acids varying the point and length
of dimerization linkers by employing chemical and bio-
transformation based modifications. This led to the syn-
thesis of compounds 5–14. Syntheses, structure activity
relationship and details of biological activities of dimeric
podocarpic acids are described.
2. Chemistry
Podocarpic acid (2) was originally reported in 1873 from
plant resins and later from several species of Podocarpus
in 1938.11 It was obtained from Sigma–Aldrich for the
synthesis of dimeric compounds. Podocarpic acid anhy-
dride (3) was synthesized by first protecting the carboxyl
group either as a methyl ester (15) or MOM ester (16)
followed by protection of phenol as benzyl ether (17
or 18). The esters were then hydrolyzed to produce the
benzyl protected acid (19), which was coupled to give
benzyl anhydride 20 that was hydrogenolyzed to furnish
3 in a good overall yield (Scheme 1).
Reaction of the benzyl ether (19) with thionyl chloride
afforded acid chloride (21), which was reacted with
ammonia to yield carboxy amide (22). Deprotonation
of 22 followed by acylation with acid chloride (21)
We recently reported biological characterization of acet-
yl podocarpic acid anhydride (APD, 4).10 APD exhib-
ited potent binding activities against LXRa and b
receptors (EC50 = 2 nM each), transactivation activity
(EC50 = 10 nM), showed a 60- and 54-fold maximal
induction of a and b receptors, respectively compared
Scheme 1. Reagents: (i) TMSCHN2, MeOH, PhH; (ii) DIEA, DMF,
MOMCl; (iii) Cs2CO3, BnBr, DMF; (iv) NaH, BnBr, DMF; (v)
KtBuO, DMSO, H3O+, (vi) H2SO4/THF/MeOH/H2O; (vii)
C2H5N@C@N(CH2)3 NMe2 Æ HCl, ClCH2CH2Cl; (viii) H2, Pd/C,
45 psi.

2826
S. B. Singh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2824–2828
Scheme 5. Reagents: (i) Bop, DMF; (ii) 0.5 mol equiv NH2NH2,
DMF, 70 °C; (iii) 0.5 mol equiv NH2CH2CH2NH2, DMF, 70 °C.
arately heated in DMF at 70 °C with 0.5 equiv each of
hydrazine and ethylene diamine to give dimers 10 and
11, respectively (Scheme 5).
Scheme 2. Reagents: (i) SOCl2, ClCH2CH2Cl, reflux; (ii) NH3,
dioxane; (iii) NaHMDS, THF; (iv) H2, Pd/C, 45 psi.
3. Biotransformation
afforded protected imide (23), which upon hydrogenoly-
sis furnished imide 5 (Scheme 2). Reduction of the
methyl ester 17 to give hydroxymethyl analog 24, which
was acylated with acid chloride 21 and hydrogenolyzed
to give 6 (Scheme 3).
The biphenyl ether (12) and biphenyl dimers (13–14)
were synthesized by biotransformation of podocarpic
acid. Incubation of 2 with MA4111 (Streptomyces
griseus) or MA6230 (S. griseus) afforded a mixture of
biphenyl ether 12 and biphenyl dimer 13. These were
easily purified by RPHPLC. The biphenyl a-mannopyr-
anoside (14) was produced by incubation of 2 with
MA6559 (Actinoplanes sp.).
The benzoate ester 7 was synthesized in 71% overall
yield by acylation of podocarpic acid benzyl ester with
acid chloride 21 followed by hydrogenolysis. The syn-
thesis of b-diketo dimer 8 began with the oxidation of
the hydroxymethyl derivative 24 to aldehyde 25, which
was methylated with methyl lithium and oxidized to give
methyl ketone 26. Deprotonation followed by acylation
with 21 afforded benzyl protected dimer 27, which was
hydrogenolyzed to afford diketo dimer 8 (Scheme 4).
4. LXRa and b activities and SAR
All compounds were first evaluated in LXRSPA binding
assay using [3H2]-F3-methyl AA (1), a compound origi-
nally identified in PPAR program as an agonist in a cell
based transactivation assay using a protocol described
earlier.9 The anhydrides 3 and 4 were potent ligands to
both receptors and exhibited binding IC50 (concentra-
tion of compound required to exhibit 50% displacement
of radioligand) values of 1–2 nM (Table 1). The imide 5
was equally potent binder to both receptors each dis-
playing IC50 value of 1 nM. The nature of the dimeriza-
tion was very important for the binding activities. The
reduction of one of the ketone groups of the anhydride
leading to the ester linked dimer (6), which maintained
the length of the three-atom linker caused a 7–12-fold
reduction of the binding activities. This loss of the bind-
ing activities was more profound (45- and 29-fold for a
and b) of the alternatively arranged dimeric benzoate
ester (7) where the separation of the two monomers was
only by two atoms. The replacement of the anhydride
oxygen with a methylene group that produced a more
stable b-diketone (8) also resulted in the loss of the bind-
ing activities by 33- and 163-fold for a and b receptors,
respectively. This indicates the importance of the con-
nectivity of the two carbonyls by a heteroatom (O or
N). Dimerizations by extending the length of linkers
by extra nitrogen (e.g., 10) or by an ethylene amine
(e.g., 11) resulted in the complete loss of the binding
activities. Likewise, other modes of dimerizations
involving the aromatic ring resulting in phosphate (9),
biphenyl ether (12) and biphenyls (13 and 14) also led
to complete loss of binding activity.
The bis-podocarpic acid phosphate (9) was prepared in
ꢀ20% yield by reaction with PCl5 in CH2Cl2 followed
by aqueous quench and RPHPLC purification. Reaction
of podocarpic acid with Bop reagent produced activated
benzotriazole ester 28 in very high yield, which was sep-
Scheme 3. Reagents: (i) DIBAL, toluene; (ii) pyridine, 21; (iii) H2, Pd/
C, 45 psi.
The agonist activity of these compounds were measured
using an in vitro cofactor association assay in which the
Scheme 4. Reagents: (i) TPAP, NMO; (ii) (a) MeLi, THF, (b) TPAP,
NMO; (iii) (a) NaHMDS, THF, (b) 21; (iv) H2, Pd/C, 45 psi.

S. B. Singh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2824–2828
2827
Table 1. LXRa and b SPA binding and cofactor association (HTRF) assays of dimeric podocarpic acids (2–14)
Compounds
LXR SPA binding
IC50 (nM)
Cofactor association HTRF assay,
EC50 (nM)
LXRa
LXRb
LXRa
LXRb
1 (T = H)
2
35.4
>50,000
24.9
>50,000
35.4
>50,000
16.1
>50,000
<3
3
2
1
1
4
2
2
2
2
5
1
14
1
12
1
NT
1
6
NT
NT
7
90
67
29
163
NT
320
8
22-(R)-OH-cholesterol
27% @ 83 nM
>15 lM
70% @100 lM
60% @100 lM
>15 lM
Compounds 9–14 were not active in the SPA binding and HTRF assays at 10 lM against either of the two receptors.
association of recombinant steroid receptor coactivator
1 (SRC1) coactivator protein with recombinant LXRa
and b ligand binding domains were measured using a
homogeneous time resolved fluorescence (HTRF)
assay.9 The strong binders such as anhydrides (3 and
4) and imide (5) strongly stimulated the SRC1 and
LBD association and exhibited EC50 (effective concen-
tration of compound requiring 50% stimulation of
cofactor association) values of 1–2 nM indicating that
they were potent agonists. The poor ligand binders were
weak agonists (e.g., b-diketone 8) (Table 1).
maximal induction was unaffected (88- and 76-fold)
(Table 2). The C-linked dimer 8 was significantly less
active in this transactivation assay (Table 2).
The ability of these compounds to efflux cholesterol was
measured in Caco-2 cells.10 The activity of dimer 3 and
imide 5 was identical to 4.10 All exhibited EC50 values of
1 nM with 7% of maximal efflux with plateau at 100–
1000 nM.
Of the dimers, the more stable and potent imide 5 was
selected for in vivo studies. The PK analysis of this com-
pound in Male Sprague-Dawley Rats at 1 mg/kg intra-
venous and 2 mg/kg oral administration produced
following results: (iv) AUC = 0.47 lM h kg/mg; Clp =
66.3 mL/min/kg; Vdss = 5.81 L/kg; t1/2 = 2.29 h; (po)
AUCn = 0.103 lM h kg/mg; Cmax = 0.037 lM; F = 22%.
To evaluate its effect on the lipid levels this compound
was administered at 10 mg/kg twice daily for eight days
to hamsters (n = 10/group) and male C57B1/6J (n = 10/
group) mice. In hamsters, total plasma cholesterol levels
increased by 28% (p < 0.05) with concomitant increase
of HDL-cholesterol levels by 22% (p < 0.05) and an in-
crease in the triglyceride levels of 51% (p < 0.05). These
results were corroborated by FPLC analysis which
showed increase in VLDL levels by 72% (p = 0.09) and
HDL by 30% (p < 0.001) and decrease of LDL levels
by 11% (p = 0.19). Similar results were also observed
in mice where total cholesterol levels increased by 26%
Cell-based transactivation assay using a chimeric LXR
constructs were used to measure the LXRa and/or b ago-
nist or antagonist functional activity of podocarpic acid
dimers in HEK-293 cells. This assay uses fusion proteins
with the yeast Gal4 DNA binding domain connected to
the hinge region and the LBD domain of either LXR
receptors and has been previously described.9,10 As re-
ported earlier, the APD (4) showed potent activity in this
assay and exhibited significantly (6–8-fold) greater
expression of the reporter gene than 22-(R)-hydroxy cho-
lesterol (See Table 2). The anhydride dimer 3 was more
potent and exhibited EC50 value of 1 nM against both
receptors and showed 50- and 80-fold maximal induction
of a and b LXR receptors, respectively. The more stable
imide dimer 5 was equally active (EC50 = <3 nM) but
had lower maximal induction (24- and 30-fold). Ester 6
was moderately less active (EC50 = ꢀ100 nM) but the
Table 2. LXRa and b cell-based transactivation assays of dimeric podocarpic acids (2–8)
Compounds
Transactivation
EC50 (nM)
Transactivation max fold
induction
LXRa
LXRb
LXRa
LXRb
1 (T = H)
2
1790
NT
1080
NT
20
NT
50
60
24
88
33
4
35
NT
85
54
30
76
51
8
3
1
10
1
10
4
5
<3
ꢀ100
<3
ꢀ100
6
8
22-(R)-OH-cholesterol
231% @10 lM
**
124% @10 lM
**
Compounds 7, 9–14 were not tested in transactivation assays; **A transactivation EC50 was not calculated for this compound since a 100% plateau
was not obtained at the doses tested.

2828
S. B. Singh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2824–2828
4. Janowski, B. A.; Willy, P. J.; Devi, T. R.; Falck, J. R.;
Mangelsdorf, D. J. Nature 1996, 383, 728–731.
5. Lehmann, J. M.; Kliewer, S. A.; Moore, L. B.; Smith-
Oliver, T. A.; Oliver, B. B.; Su, J.-L.; Sundseth, S. S.;
Winegar, D. A.; Blanchard, D. E.; Spencer, T. A.; Wilson,
T. M. J. Biol. Chem. 1997, 272, 3137–3140.
(p = 0.0238), HDL-cholesterol levels by 19% (p = 0.01)
and triglyceride levels by 36% (p = 0.8). All data were
measured compared to vehicle controls. These observa-
tions are consistent with the results reported for
T0901317.12
6. Janowski, B. A.; Grogan, M. J.; Jones, S. A.; Wisely, G.
B.; Kliewer, S. A.; Corey, E. J.; Mangelsdorf, D. J. PNAS
1999, 96, 266–271.
7. Fu, X.; Menke, J. G.; Chen, Y.; Zhou, G.; MacNaul, K.
L.; Wright, S. D.; Sparrow, C. P.; Lund, E. G. J. Biol.
Chem. 2001, 276, 38378–38387.
In summary, in this paper we have described the discov-
ery of podocarpic acid anhydride, a derivative of plant
natural product, as a potent and balanced LXRa and
b agonist lead. Initial lead optimization led to the imide
analog which demonstrated potent agonistic activities
against both receptors and showed increased levels of
cholesterol efflux in cells. This compound also raised
HDL in vivo. However, it also increased levels of triglyc-
eride which is undesirable. The dissection of the effects
of HDL and triglyceride would require a and/or b
LXR selective agonists.
8. Repa, J. J.; Mangelsdorf, D. J. Nat. Med. 2002, 8, 1243–
1248.
9. Menke, J. G.; Macnaul, K. L.; Hayes, N. S.; Baffic, J.;
Chao, Y.-S.; Elbrecht, A.; Kelly, L. J.; Lam, M.-H.;
Schmidt, A.; Sahoo, S.; Wang, J.; Wright, S. D.; Xin, P.;
Zhou, G.; Moller, D. E.; Sparrow, C. P. Endocrinology
2002, 143, 2548–2558.
10. Sparrow, C. P.; Baffic, J.; Lam, M.-H.; Lund, E. G.;
Adams, A. D.; Fu, X.; Hayes, N.; Jones, A. B.; Macnaul,
K. L.; Ondeyka, J.; Singh, S.; Wang, J.; Zhou, G.; Moller,
D. E.; Wright, S. D.; Menke, J. G. J. Biol. Chem. 2002,
277, 10021–10027.
References and notes
1. Willy, P. J.; Umesono, K.; Ong, E. S.; Evans, R. M.;
Heyman, R. A.; Mangelsdorf, D. J. Genes Dev. 1995, 9,
1033–1045.
11. Sherwood, R.; Short, W. F. J. Chem. Soc 1938, 1006–
1013.
2. Miyata, K. S.; McCaw, S. E.; Patel, H. V.; Rachubinski,
R. A.; Capone, J. P. J. Biol. Chem. 1996, 271, 9189–9192.
3. Shinar, D. M.; Endo, N.; Rutledge, S. J.; Vogel, R.;
Rodan, G. A.; Schmidt, A. Gene 1994, 147, 273–276.
12. Schultz, J. R.; Tu, H.; Luk, A.; Repa, J. J.; Medina, J. C.;
Li, L.; Schwendner, S.; Wang, S.; Thoolen, M.; Mangels-
dorf, D. J.; Lustig, K. D.; Shan, B. Genes Dev. 2000, 14,
2831–2838.