J. Med. Chem. 2007, 50, 1445-1448
1445
Analogues of Acifran: Agonists of the High and
Low Affinity Niacin Receptors, GPR109a and
GPR109b
Jae-Kyu Jung,† Benjamin R. Johnson,† Tracy Duong,†
Marc Decaire,† Jane Uy,† Tawfik Gharbaoui,†
P. Douglas Boatman,† Carleton R. Sage,† Ruoping Chen,‡
Jeremy G. Richman,‡ Daniel T. Connolly,‡ and
Graeme Semple*,†
Figure 1. Structures of known antilipolytic compounds shown to be
agonists of GPR109a.
agonists of GPR109a (Table 1). In our hands, these compounds
are able to decrease forskolin-induced elevation of cAMP levels
in cells stably transfected with the receptor. Acifran is less active
at this receptor than niacin but more potent than the clinically
used agent acipimox.4 We found the agonist potency of acifran
in a whole cell cAMP assay was around 1 µM for GPR109a
and it showed only a 3-4-fold selectivity for this receptor over
GPR109b (Table 1), whereas niacin and acipimox showed no
effect up to 30 µM in the GPR109b assay. Recent data has
shown that â-hydroxybutyrate can function as ligand for
GPR109a, with an EC50 in the mM range. Despite the low
affinity, under various stressful conditions in vivo, the plasma
levels of â-hydroxybutyrate can significantly exceed this
concentration.7 Our data indicate, therefore, that acifran should
be considered to be a full agonist of GPR109a, as it has both
comparable agonist efficacy to this purported natural ligand and
it is able to fully reverse the cAMP elevating effect of forskolin.
There is no known natural ligand for GPR109b, but again acifran
was able to fully reverse the cAMP elevating effect of forskalin
in the GPR109b assay. While acifran was clinically evaluated
and was shown to lower free fatty acids in humans, in the
absence of a molecular target, only four analogs were prepared
and reported.8 A recent paper described further investigations
around the acifran series, but significant improvements in
potency were not achieved. Only the 4-fluorophenyl derivative
of acifran was equipotent with the parent molecule, and
derivatives with substitution on the furan ring or four-membered
ring analogues were all inactive or considerably less potent than
acifran. Chiral versions of acifran were also prepared via a
Sharpless asymmetric dihydroxylation, and all of the activity
of racemic acifran could be assigned to the (S)-enantiomer, as
the (R)-enantiomer was completely inactive.9
Departments of Medicinal Chemistry and DiscoVery Biology,
Arena Pharmaceuticals, 6166 Nancy Ridge DriVe,
San Diego, California 92121
ReceiVed January 8, 2007
Abstract: Recently identified GPCRs, GPR109a and GPR109b, the
high and low affinity receptors for niacin, may represent good targets
for the development of HDL elevating drugs for the treatment of
atherosclerosis. Acifran, an agonist of both receptors, has been tested
in human subjects, yet until recently very few analogs had been reported.
We describe a series of acifran analogs prepared using newly developed
synthetic pathways and evaluated as agonists for GPR109a and
GPR109b, resulting in identification of compounds with improved
activity at these receptors.
Niacin (nicotinic acid) is a water-soluble vitamin that, at high
doses in humans, favorably affects essentially all serum lipid
and lipoprotein parameters.1 Niacin’s ability to increase high-
density lipoprotein (HDL) to a greater extent than other currently
used drugs is of particular interest, and it has been used for
many years in the treatment of atherosclerotic disease.2 Recent
mechanistic investigations have shown that niacin may exert
its beneficial effects on lipids through activation of a G-protein
coupled receptor (GPCR) located on adipocytes.3 The resultant
decrease in intracellular cAMP leads to inhibition of lipolysis
via negative modulation of hormone-sensitive lipase activity,
thereby decreasing plasma free fatty acid levels and, ultimately,
it has been postulated, raising HDL. Two Gi-coupled orphan
GPCRs that share 95% identity and which are both expressed
in human adipocytes have been identified as possible molecular
targets for niacin.4 GPR109a (HM74a) is the human orthologue
of the previously described rodent receptor (PUMA-G),5 whereas
GPR109b (HM74) appears to have arisen from a very late gene
duplication. It differs from GPR109a and PUMA-G mainly in
the C-terminal tail portion and has no rodent equivalent.6 Niacin
has been shown to activate GPR109a, with an EC50 of 250 nM
The above observations, coupled with the availability of a
high-throughput functional screening assay, prompted us to
investigate the use of acifran as a starting point for the discovery
of new receptor agonists and ultimately new agents to elevate
HDL.
3
in a GTPγS assay and displaces H-niacin from GPR109a,
expressing CHO cell membranes with an IC50 of 81 nM.4 It is
a much weaker ligand for GPR109b, with an EC50 in the
millimolar range. Unfortunately, the use of niacin as a thera-
peutic is limited by a number of associated side-effects,
including an uncomfortable cutaneous flushing response and
free fatty acid rebound. The development of novel agonists of
the niacin receptor that retain the beneficial effect on athero-
sclerosis but with fewer side effects would clearly be of value.
With the discovery of these putative molecular targets for
niacin, we and others quickly identified the known lipid
modulating agents, acifran and acipimox (Figure 1) as functional
We quickly found the scope of the reported procedure for
the preparation of acifran (Scheme 1) to be somewhat limited
and, hence, inadequate for fully exploring the SAR. These
limitations were identified as the instability of the R-tert-hydroxy
ketone intermediate 3 under the acid hydrolysis conditions
employed for the conversion of the alkyne 2 to 3 and the lack
of reproducibility of the conversion of ketone 3 to acifran analog
5 via the six-membered intermediate 4. In particular, when the
substituent R1 was heterocyclic in nature, no formation of the
desired analogues 5 was detected. The search for new ap-
proaches to acifran analogs led us to investigate a route making
use of dithiane chemistry so that R-tert-hydroxy ketone inter-
mediate 3 could be produced under neutral conditions (Scheme
1). The dithiane anion derived from 6 was added to various
ketones, resulting in tert-alcohols 7 in good yields. The resulting
* Author to whom correspondence should be addressed. Tel.: (858)-
† Department of Medicinal Chemistry.
‡ Department of Discovery Biology.
10.1021/jm070022x CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/15/2007