5006
P. C. Ting et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5004–5008
Table 3
O
In vitro H3 receptor binding affinity of right hand side analogues 8a–l
NH
O
N
O
N
NBOC
O
N
a
HN
HN
N
O
N
9
7
N
R
HN
d,b,c
b,c
O
Compd
R
Kia (nM)
O
N
N
N
O
N
O
6h
2.0
N
N
N
N
R
R
HN
N
R'
10a
10b
10c
174
40%b
440
11
10
N
Scheme 3. Reagents: (a) N-BOC-isonipecotic acid, DCC, HOBT, CH2Cl2, 95%; (b) TFA,
CH2Cl2, 100%; (c) RCHO, AcOH, NaB(OAc)3H, CH2Cl2 or RCOCl, Et3N, CH2Cl2; (d)
NaOH, nBu4N(HSO4), K2CO3, R0X.
O
N
Cl
N
N
10d
10e
222
route outlined in Scheme 3.15 Coupling of 4-(2-keto-1-benzimidaz-
olinyl)piperidine (7) with N-BOC-isonipecotic acid formed inter-
mediate 9. Acid catalyzed removal of the BOC protecting group
and further elaboration gave compounds 10a–l. Alternatively,
intermediate 9 could be alkylated before removal of the BOC pro-
tecting group, and additional elaboration provided compounds
11a–r.
The SAR around the pyridylmethyl moiety appeared to be quite
restricted (Table 3). Replacement of the 4-pyridyl ring in 6h with
the 3-pyridyl ring in 10a or the 2-pyridyl ring in 10b resulted in
a reduction of H3 affinity. The pyridine N-oxide 10c and the 3-chlo-
ropyridyl analog 10d were also less active. Furthermore, the meth-
ylene linker appeared to be important for biological activity.
Conversion of the basic amine to an amide in compound 10e or
extension to the ethylene linker in compound 10f significantly
decreased H3 receptor binding. Even addition of a single methyl
group as in compound 10g was not well-tolerated. Alternative het-
erocycles for the pyridine ring including the thiophene analog 10h,
the furan analog 10i, the pyrrole analog 10j, the quinoline analog
10k, and the indole analog 10l, all exhibited reduced H3 affinity.
In contrast, the urea nitrogen of the left side benzimidazolone
was very amenable to substitution (Table 4). H3 antagonists, in par-
ticular structures which contain an imidazole moiety, displayed
undesired cytochrome P450 inhibition, and therefore, we screened
active analogs for CYP liver enzyme 3A4 and 2D6 inhibition.14 The
11%b
O
10f
537
20
N
N
10g
Me
S
10h
10i
10j
104
204
42%b
O
N Me
N
10k
10l
681
Me
N
30%b
Inhibition of [3H]N -methylhistamine binding to guinea pig brain receptor. H3
binding Ki values are the average of at least two independent determinations.
Variation from the mean is generally within 50%.
a
a
lead compound 6h exhibited moderate 3A4 inhibition (9.5
and minimal 2D6 inhibition (>20 M). Target compounds which
met the criteria of guinea pig Ki for H3 <5 nM and IC50 for
3A4 >5 M were tested for H3 functional activity (pA2) in the gui-
nea pig ileum assay.14 In this assay, (R)-
-methylhistamine inhib-
ited electrical field-stimulated contractions in guinea pig ileum
segments. A H3 antagonist dose dependently inhibited the (R)-
lM)
b
% Inhibition at 1 lM (n = 2).
l
l
compounds were unfortunately potent inhibitors of the CYP3A4
enzyme. In addition, compounds 6h and 11i–l showed no liver
enzyme 2C9 inhibition (>20 lM).
a
a
-
Due to their acceptable guinea pig binding, guinea pig func-
tional H3 activity, and P450 profile (3A4, 2D6, 2C9), compounds
11i–k were validated in the guinea pig CNS-induced hypertension
methylhistamine activity. Overall, there was good correlation
between the guinea pig receptor binding assay and the guinea
pig functional assay. Introduction of an alkyl group such as the
methyl in 11a and the isobutyl in 11b retained reasonable H3
binding affinity and 3A4/2D6 profile. Ether chains as in compounds
11d–h were well-tolerated, but the more polar alcohol analog 11c
displayed good in vitro biological activity and improved (lower)
3A4 inhibition. This trend was also present with the more polar
substituted amino side chains in compounds 11i–l where
assay.16 In this in vivo model, (R)-
a-methylhistamine inhibited the
CNS-induced hypertension caused by electrical stimulation to the
medullary cardiopressor area. All three compounds 11i–k exhib-
ited in vivo efficacy by attenuating the effect of (R)-a-methylhista-
mine with an ED50 <0.3 mg/kg po. Compounds 11i–k displayed
AUC(0–6 h) of 240 nM h when dosed in the rat at 10 mg/kg po,
minimal hERG inhibition (<10%) at 10 lM, and reasonable clear-
Ki 6 0.5 nM and IC50 for 3A4 inhibition is P30
l
M. Compounds
ance levels in rat and human hepatocytes (<10
cells).
In conclusion, a new series of H3 receptor antagonists was dis-
covered. Optimization of the original lead compound 1 led to the
key intermediate 6h, and further studies of 6h identified the
biologically active compounds 11i–k. Initial studies focused on
lL/min/million
11j and 11l exhibited undesired 2D6 inhibition in comparison to
compounds 11i and 11k. Introduction of a benzyl group with either
an electron withdrawing chlorine atom as in analogs 11n–p or an
electron donating methoxy group as in analogs 11q and 11r also
displayed a high level of H3 receptor binding affinity, but these