J. M. Keith et al. / Bioorg. Med. Chem. Lett. 17 (2007) 702–706
703
from our laboratories.10,11 Herein we describe the combi-
nation of our potent H3 antagonist 412 with tetrahydro-
isoquinoline derived serotonin transporter (SERT)
inhibitors such as 513 to give potent dual H3 antago-
nist/SERT inhibitors 6.14 In this paper, we describe our
efforts to optimize the binding affinity at H3 and the
physical properties of molecules 6, while retaining SERT
activity, through modification of the H3 pharmacophore.
MeOH in the presence of NaCNBH3, bromocresol
green, and MeOH until a color change, generally from
green/brown to yellow, persisted to give 13. Conversion
of the alcohol to the mesylate was accomplished under
standard conditions and the mesylate used immediately
in the subsequent step. Displacement of the mesylate
was generally carried out in a parallel fashion using a
variety of primary and secondary amines to provide
the desired products 6a–6cc. To improve yields, hydro-
chloric acid salts of the amines were neutralized with
NaOt-Bu in n-BuOH prior to use. Yields ranged from
0.4% to 58%.
Synthesis of tetrahydroisoquinolines 6a–6cc was accom-
plished according to the procedure outlined in Scheme
1.15 Thus, 3-hydroxybenzaldehyde (7) was treated with
3-bromo-propan-1-ol to give 3-(3-hydroxy-propoxy)-
benzaldehyde 8. Subsequent reductive amination with
methylamine afforded intermediate 9 in quantitative
yield. Alkylation of the resulting secondary amine gave
10, which underwent an acid catalyzed ring-closure reac-
tion with methanesulfonic acid (MSA). The ring-closure
afforded the expected 1,2-dihydro-isoquinoline 11 along
with the ring oxidized isoquinolinium salt 12. The latter
product was generally found to be the predominant
component following workup and the product mixture
was carried forward without further purification. The
ring-closed products were treated with 1.25 M HCl in
Rat and human SERT and human histamine H3 binding
data16 for the tetrahydroisoquinolines 6 are shown in
Tables 1 and 2.17 Modifications to the pendant amine
portion of the molecule had little effect on the SERT
affinity of the compounds. All were generally potent
serotonin transporter ligands with good correlation be-
tween the human and rat transporters. Affinity for the
human histamine H3 receptor varied greatly. Morpho-
line and substituted piperidines usually gave potent
compounds with the exception of 6j. Analogs bearing
a tertiary amine were consistently more potent than
the corresponding secondary amines (e.g., 6n and 6o).
Substituted piperazines generally had high affinity for
the H3 receptor when the substituent was small (e.g.,
6r and 6s) although a cyclopropyl group had a detrimen-
tal effect on potency (6t). In all but one case, aryl piper-
azines displayed weak affinity for the H3 receptor. The
exception was the 4-pyridyl-piperazine derivative 6x,
which had high affinity. All piperazine amides examined
displayed low affinity for the H3 receptor, regardless of
the substitution (6aa–6cc).
OH
(a)
(b)
H
H
O
HO
OH
O
O
7
8
OMe
O
OH
H
(c)
(d)
Me
N
N
O
Me
Compounds 6b and 6c were examined in the 5-hydroxy-
tryptophan (5-HTP) induced head twitch model in
mice.18 This animal model gives a qualitative measure
of the blockade of the SERT, with a behavioral output.
The mice were dosed at 10 mg/kg, ip (n = 4 for each) and
observations made at 1 and 24 h. At 1 h both com-
pounds gave a very robust response (>500% of control)
and in each case, the response was diminished at 24 h,
suggesting the compounds penetrate the brain quickly
but are subsequently eliminated. To verify this interpre-
tation, blood–brain barrier (BBB) experiments were per-
formed on 6b and 6c to compare pharmacological
responses to brain and plasma concentrations. Com-
pounds were dosed 10 mg/kg, ip and brain concentra-
tions determined at 1 and 24 h (n = 3). Compounds 6b
and 6c showed high brain concentrations at 1 h (12.5
and 6.1 lM, respectively) and significantly lower brain
concentrations at 24 h (<1 and 2.5 lM, respectively).
These observed brain concentrations correlate well with
the 5-HTP induced head twitch data.
O
9
10
OMe
OMe
OH
OH
(e)
+
N
N
Me
O
O
Me
11
12
OMe
OMe
R1
R2
OH
N
(f, g)
N
N
O
Me
O
Me
13
6
Scheme 1. Synthesis of tetrahydroisoquinolines. Reagents and condi-
tions: (a) K2CO3, 3-bromo-propan-1-ol, acetonitrile, reflux, 48 h, 90%;
(b) 40% aqueous MeNH2, MeOH, 0 °C, then NaBH4, 0 °C, 0.5 h, then
23 °C 18 h, 100%; (c) Hunig’s base, 2-bromo-40-methoxy-acetophe-
none, THF, 23 °C, 45 min; (d) MSA, 60 °C, 18 h; (e) NaCNBH3,
MeOH, bromocresol green, 23 °C, 5 min, then 1.25 M MeOHÆHCl,
0.5 h, 28% three steps; (f) MeSO2Cl, Et3N, DCM, 0–23 °C, 20 min; (g)
R1R2NH, Na2CO3, KI, n-BuOH, 50–80 °C, 18 h, 0.4–58%, two steps.
Compounds 6b and 6c were also examined in an hH3
functional assay and were found to be potent antago-
nists (pA2 = 8.6 and 9.6, respectively).19
In conclusion, we have found a wide range of amine sub-
stituents tolerated by the H3 receptor and the SERT.
Keith, John M.
