M. Gustafsson et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5918–5921
5919
disorders in chronic kidney disease, dominant hyperparathyroid-
OH
N
S
O
N
S
i
ism, familial hypocalciuric hypercalcemia, neonatal severe primary
hyperparathyroidism, and hypercalcemia caused by autoantibod-
ies that inhibit CaSR.
+
H
3
The calcium sensing receptor is a member of the GPCR super-
family and belongs to the group C family of GPCRs and is most clo-
sely related to the GABAB receptor and metabotropic glutamate
receptors.5–8 The receptor is expressed in calciotropic hormone-
secreting organs (parathyroid glands and C-cells of the thyroid
glands), kidney, bone, and intestinal cells.4 The CaSRs primarily
functions are to maintain systemic Ca2+ homeostasis mainly by
causing suppression of parathyroid hormone secretion by the para-
thyroid glands, and by influencing rates of renal tubular calcium
reabsorption and secretion of calcitonin by C-cells of the thyroid.9,10
A chemical library containing 250,000 small drug-like com-
pounds was screened for agonist activity using a cell-based func-
tional assay called Receptor Selection and Amplification
Technology (R-SAT).11 In the screen cells were maintained in stan-
dard tissue culture medium containing Dulbecco’s modified Eagle’s
medium (DMEM) and 0.8 mM MgCl2 and 1.6 mM CaCl2 to increase
the sensitivity for detecting hits. A number of active compounds rep-
resenting distinct structural classes were identified. Based on the
structurally interesting features compared with reported PAMS,
the benzothiazole class (e.g., 1) was chosen for further studies. In
contrast to 1, most of the reported CaSR PAMs today including,
NPS R-467, NPS R-568, calindol, and cinacalcet hydrochloride (2)
are phenylalkylamines, which are structurally derived from fendi-
line. In addition, compound 1 is not cationic at physiological pH
which further differentiates it from previously reported calcimimet-
ii, iii
N
S
16
Scheme 2. Reagents and conditions: (i) n-BuLi, THF, À78 °C, 46%; (ii) CH3SO2Cl/
Et3N, CH2Cl2, rt; (iii) CH3MgBr, THF, rt, 77% two steps.
O
OH
N
S
N
S
i
R1
8, 12, 14
Scheme 3. Reagents and conditions: (i) ArMgX, X = Cl, Br, THF, rt, 62–77%.
OH
N
S
N
S
i
7
13
ics, including the PAMs, the
L-amino acids, and obviously cations
such as Ca2+ and Mg2+
.
Scheme 4. Reagents and conditions: (i) PPA, THF/Ac2O, rt, 41%.
Compound 1 displayed partial agonism and nM activity at the
CaSR (pEC50 6.4 and 68% efficacy), but also had a high intrinsic
clearance in human liver microsomes of >200
l
L/min mg.
16 were synthesized. Lithiation of benzothiazole followed by reac-
tion with 2,4-dimethylbenzaldehyde gave alcohol 3 in 46% yield
(Scheme 2). Analogue 16 was made in two steps, one-pot, from
alcohol 3 as follows; compound 3 was reacted with CH3SO2Cl/
Et3N followed by nucleophilic substitution with CH3MgBr to yield
compound 16 in 77% yield.
Analogues 8, 12, and 14 were made by reacting acetylbenzo-
thiazole with the corresponding Grignard reagent (Scheme 3).
Treatment of compound 13 with poly phosphoric acid (PPA) gave
exo-methylene 7 in 41% yield (Scheme 4).
The in vitro activity of these analogs was measured using R-SAT
(Table 1). Experiments using media partially depleted of MgCl2 and
various concentrations of CaCl2 confirmed that our compounds
cooperate with divalent cations to activate the CaSR (Fig. 2).15
From the SAR it reads that the phenyl substituents have a large
impact on the activity of this class of compounds (Table 1). Replac-
ing the phenyl with the larger naphthyl gives a slightly less active
analogue compared with the hit 1. A more rigid structure was well
tolerated, and analogue 9 retained the activity (pEC50 6.5 0.2 and
Eff. 81 6). Whereas the o-methyl compound 11 had comparable
activity with the p-chloro compound 1, the electron donating o-
methoxy group (10) had a negative effect and the activity dropped
10 times to pEC50 5.4. With these substantial changes in activity in
mind we further elaborated with small substituents, for example,
Me, Cl, and F to give an initial SAR. Compounds 11 (o-CH3) and
12 (m-CH3-p-F), pEC50 at 6.6 and 6.8, respectively, had slightly
higher activity than the hit 1 (pEC50 6.4). However, a more substan-
tial increase in activity was seen when a disubstitution pattern
with methyl and chloro groups was introduced (Table 1, com-
pounds 5 and 13–15). The two most active racemic compounds
tested were 13 (pEC50 7.6 0.1 Eff. 91 4) and 14 (pEC50 7.5 0.2
Eff. 82 9). Another predominant feature of the structure–activity
relationship is that a small non-hydrogen R substituent is manda-
Based on the lead compound 1, a focused set of analogues were
synthesized to establish the SAR and to improve on the metabolic
stability. To be able to easily introduce a variety of substituents (R)
on the benzylic position, different synthetic strategies were devel-
oped (Schemes 1–4). Lithiation of benzothiazole at low tempera-
ture (À78 °C),12 followed by addition of a variety of ketones gave
products 1, 4–6, 9–11, 13, 15, 17–18 in medium to good yields
(Scheme 1).13,14 Of these only the synthesis of 9 was difficult,
and a modest 32% yield was achieved. The low yield was attributed
to the decreased reactivity of cyclic ketone compared to the ace-
tophenones in combination with the low stability of the lithiated
benzothiazole. The lithiated benzothiazole is reported to start
decomposing when increasing the temperature for at À40 °C being
totally cracked.12 To investigate the significance of the hydroxyl
and the methyl group on the benzylic carbon, analogues 3 and
OH
R
N
S
O
R
R1
R1
2, 4-6, 10, 11, 13, 15, 17-18
R=CH3, cPr, or Ph
or
N
S
i
+
O
HO
N
S
9
Scheme 1. Reagents and conditions: (i) n-BuLi, THF, À78 °C, 32–80%.
Gustafsson, Magnus
