1872
P. V. Fish et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1871–1875
O
Target compounds were also tested for their ability to inhibit bind-
ing of [3H]-meselurgine at the human 5-HT2C receptor utilizing SPA
technology and cellular membrane preparations generated from
recombinant Swiss 3T3 cells.12 Agonist activity at the 5-HT2B
receptor was measured in recombinant cell-based systems
expressing the human receptor.
CO2H
a, b
R3
R3
NH
MeO
OMe
6
7
c
The secondary sulfonamides 3a–f were found to be weak 5-
HT2C agonists with 5-HT2B activity, and several examples were
more active at the 5-HT2B receptor (3b) (Table 1). In contrast, ter-
tiary sulfonamides 3g–m, whether alkyl, benzyl or cyclic amines,
had improved 5-HT2C activity and some examples demonstrated
minimal 5-HT2B activity. Dihydrobenzofuran 3n was one of the
most potent 5-HT2C agonists from this set and had no significant
O
d, e
R3
R3
NH
NH
9
8
5-HT2B activity (<10% @ 10 lM). As part of our next iteration of
f
analogue synthesis, we elected to explore further this motif by
scaffold hopping from the dihydrobenzofuran to aryloxy amines
II-X (Fig. 1). These targets were designed to be more chemically en-
abled and promote a rapid exploration of the SAR of the amine ring
size and aryl ring substitution.
Azetidines 3o–q, pyrrolidines 3r,s and piperidines 3t–w all fur-
nished examples of potent 5-HT2C agonists with minimal 5-HT2B
activity. The aryloxy group was clearly making an important con-
tribution to both these SARs (e.g. 3r vs 3k and 3v vs 3l).
R3
O
O
g
R3
N
N
O
S
CF3
CF3
Cl
O
10
11
j, k, i
h, i
Introduction of a second or third substituent on the aryl ring of
the benzazepine 3x-aa proved detrimental within this very limited
set of R3 groups (e.g. 3v vs 3y vs 3aa). Further investigation of this
SAR, where R3 was from a broader set of substituents, would rep-
resent a suitable position for potential optimization of 3.
R3
R3
NH
NH
O
S
O
S
3
R1R2N
R1
7
O
O
An analysis of 5-HT2C activity of 3 aligned by R1R2N- clusters for
all the compounds (n = 169) was performed with scatter plots
which assisted in the identification of SAR trends and preferred
compounds (Fig. 2). For example, it proved difficult to divorce 5-
HT2C and 5-HT2B activity with aniline derivatives (3: R1 = Ar)
whereas compounds containing azetidines and pyrrolidines fur-
nished many potent 5-HT2C agonists with minimal activity at 5-
3
4
Scheme 1. Synthesis of sulfonamides 3 and sulfones 4. Reagents: (a) SOCl2; (b)
H2NCH2CH(OMe)2; (c) concd HCl; (d) H2, Pd/C, AcOH; (e) BH3ÁTHF; (f) (CF3CO)2O;
(g) ClSO3H; (h) R1R2NH; (i) K2CO3, MeOH; (j) KF, 18-crown-6 (cat), MeCN; (k)
R1MgX, THF.
acid-catalyzed cyclization of 7 to give 8 followed by sequential
reduction of the alkene and then amide.17 Protection of the azepine
N-atom followed by chlorosulfonylation furnished 11.
HT2B
.
Aryl sulfones 4 were potent 5-HT2C agonists which also demon-
strated significant 5-HT2B agonist activity and were not pursued.
Reversed sulfonamides 5 failed to show any significant 5-HT2C
activity which highlighted the importance of the sulfonamide of
3 in picking up specific favourable interactions and/or presenting
the amino substituents in a favourable trajectory.
Selected compounds were then screened in high throughput
in vitro ADME and safety screens as a wider assessment of their
properties: metabolic stability in human liver microsomes (HLM),
transit performance across MDCK-mdr1 cells to assess membrane
permeability,20 and binding to the potassium hERG channel as a
measure of ion channel activity (Table 2). In general, sulfonamides
3 had good metabolic stability consistent with low predicted clear-
ance, good membrane permeability with low affinity for P-gp ef-
flux transporters, and moderate affinity for the potassium hERG
channel.
Pharmacological evaluation for activity at the 5-HT2A receptor
was measured in a FLIPR assay employing Swiss 3T3 cells express-
ing the recombinant human 5-HT2A receptor, and in vitro tissue
preparations with femoral artery.21 Screening of 3v in the recombi-
nant assay gave a significant response with an EC50 32 nM and Emax
70%. However, our experience has taught us that this was a highly
expressed/coupled cell-line which over estimates 5-HT2A activity
and that the tissue preparation assay was a better predictor of
in vivo outcomes.22,23 Evaluation of 3v in the canine femoral artery
Reversed sulfonamides 5 were also prepared from benzazepine
9 (Scheme 2). Nitration of 9 with fuming HNO3 selectively intro-
duced a 7-NO2 group in good yield.18 Boc protection of the aze-
pine and hydrogenolysis of the 7-NO2 group afforded aniline
12.19 Reductive amination of 12 with aldehydes under standard
conditions gave sec-amines 13 and then reaction with sulfonyl
chlorides followed by deprotection of the azepine N-Boc group
gave 5.
The 5-HT2C agonist activity of target compounds (Table 1) was
initially evaluated by measuring the ability to induce a fluorescent
based calcium mobilization signal in a FLIPR assay employing re-
combinant CHO K1 cells expressing the human 5-HT2C receptor.12
a, b, c
N
Boc
NH
H2N
9
12
d
e, f
NH
N Boc
O
S
R2
N
HN
R1
gave a minimal response of <10% at 0.54 lM, and we concluded
that 3v has no significant 5-HT2A activity.
Compound 3v was then evaluated in additional ADME and
pharmacology screens as a representative example from this ser-
ies. Sulfonamide 3v had weak inhibition of CYP450 enzymes 1A2,
O
R1
5
13
Scheme 2. Synthesis of reversed sulfonamides 5. Reagents: (a) fuming HNO3; (b)
(Boc)2O; (c) H2, Pd/C; (d) aldehyde, MeOH then NaBH4; (e) R2SO2Cl; (f) CF3CO2H.
Fish, Paul V.
