3,4-Dihydro-2H-benzo[1,4]oxazine derivatives as 5-HT6 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2006.12.093

A series of novel 3,4-dihydro-2H-benzo[1,4]oxazine derivatives has been designed and synthesized as 5-HT₆ receptor antagonists. Many of the compounds displayed subnanomolar affinities for the 5-HT₆ receptor and good brain penetration in rats. The relationship of structure and lipophilicity to hERG inhibition of this series of compounds is discussed.

Full text of DOI:10.1016/j.bmcl.2006.12.093

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3504–3507
3,4-Dihydro-2H-benzo[1,4]oxazine derivatives
as 5-HT6 receptor antagonists
Shu-Hai Zhao,a,* Jacob Berger,a Robin D. Clark,a Steven G. Sethofer,a Nancy E. Krauss,a
Julie M. Brothers,b Renee S. Martin,b Dinah L. Misner,c
Dietmar Schwabd and Ludmila Alexandrovad
aRoche Pharmaceuticals, Medicinal Chemistry, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
bRoche Pharmaceuticals, Biochemical Pharmacology, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
cRoche Pharmaceuticals, Non-Clinical Drug Safety, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
dRoche Pharmaceuticals, Safety and Technical Sciences, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
Received 16 October 2006; revised 15 December 2006; accepted 29 December 2006
Available online 5 January 2007
Abstract—A series of novel 3,4-dihydro-2H-benzo[1,4]oxazine derivatives has been designed and synthesized as 5-HT6 receptor
antagonists. Many of the compounds displayed subnanomolar affinities for the 5-HT6 receptor and good brain penetration in rats.
The relationship of structure and lipophilicity to hERG inhibition of this series of compounds is discussed.
Ó 2007 Elsevier Ltd. All rights reserved.
The human 5-HT6 receptor is one of the more recently
identified members of the serotonergic receptor family.
The density of 5-HT6 receptors in rat brain is highest
in the olfactory tubercles, striatum, nucleus accumbens,
and hippocampus. Its distribution, together with its high
affinity for several tricyclic antipsychotics, suggests the
5-HT6 receptor may be involved in memory disorders,
psychosis, depression, and appetite control.1–3 More-
over, 5-HT6 receptor antagonists have been shown to in-
crease extracellular concentrations of neurotransmitters
that are important for cognition. Therefore, there has
been increasing interest in the 5-HT6 receptor as a ther-
apeutic target over the past few years, as evidenced by
numerous publications appearing in the literature.1
a 5-HT6 antagonist with an improved brain to plasma ra-
tio and lower hERG inhibitory activity compared to 2.
Herein, we report the design, synthesis, and biological
activity of a series of novel 3,4-dihydro-2H-benzo[1,4]ox-
azine derivatives 1 as 5-HT6 receptor antagonists.
In addition to 2, arylsulfonamides 3 and 4 were also
reported to be 5-HT6 receptor antagonists,5,6 albeit with
low brain penetration in the case of 3. In comparing 2, 3,
and 4, common features are evident, including a central
aromatic scaffold, a basic side chain, and an aryl sulfo-
nyl or arylsulfonamide group. These groups are ar-
ranged in a spatially similar fashion (Scheme 1).
Therefore, we envisioned that molecules of template 1,
which possess all these features, would also be high-af-
finity 5-HT6 receptor antagonists. Moreover, template
1 lacks the sulfonamide NH hydrogen bond donor in
3, which should help improve its brain penetration.
Our initial lead was RO-65-7674 (2),4 a potent and selec-
tive 5-HT6 antagonist with good DMPK properties in
preclinical studies. However, the whole brain homoge-
nate to plasma ratio of 2 in rats was found to be low
(B/P = 0.1). It is also a potent hERG channel inhibitor
with an IC20 of 0.2 lM. Hence, our goal was to discover
The synthesis of compounds 1 is exemplified by the reac-
tion sequences shown in Schemes 2 and 3. The latter is
more divergent, allowing more rapid synthesis of com-
pounds with different substituted arylsulfonyl groups.
Keywords: 5-HT6 antagonist; Benzoxazine; Brain penetration; hERG;
Lipophilicity.
Substituted 6-bromo-2-nitrophenols were either com-
mercially available or synthesized by bromination of
the corresponding 2-nitrophenols in acetic acid in good
*
Corresponding author. Tel.: +1 650 855 5107; fax: +1 650 855
6082; e-mail: shu-hai.zhao@roche.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.093

S.-H. Zhao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3504–3507
Boc
3505
H
N
Boc
N
H
N
N
Me
Br
HN
N
O
b
c, d
a
N
N
MeO
CH3
N
N
N
1
Me
N
O
O
N
N
H
R
N
NH
S
O
S
NH
S
Me
O
O
Cbz
S
O
NH2
R
N
H
R
N
O
O
S
Cl
Cbz
2
3
4
12
13
14
SB-271046
pKi =8.9
Brain/blood = 0.1
RO65-7674
pKi = 9.1
B/p = 0.1
RO63-0563
pKi = 7.9
Scheme 3. Synthesis of 3,4-dihydro-2H-benzo[1,4]oxazine derivatives.
Reagents and conditions: (a) 1-Boc-piperazine, Pd2(dba)3, BINAP,
toluene, NaO-t-Bu, 100 °C; (b) H2, Pd/C, EtOH; (c) ArSO2Cl, DCM,
pyridine; (d) HCl, EtOH, reflux.
H
N
NR2
Y
N
O
R
2, 3, 4
R
Ar
X
The in vitro binding activity of these compounds was
determined by displacement of [3H]LSD in HEK293
cells expressing the recombinant human 5-HT6 receptor
at 37 °C and pKi values are presented in Table 1.
R
N
S
O
O
S
Ar
O
O
1
Scheme 1. Design of novel benzoxazine derivatives as 5-HT6 ligands.
Halogen substitution on the 2 and/or 3 positions of the
phenylsulfonyl group seems beneficial for 5-HT6 affinity.
For example, 2,3-dichlorophenylsulfonyl (1a) gave the
highest affinity in this series with a pKi of 10.2, whereas
the parent unsubstituted compound 1b had relatively
modest affinity (pKi = 7.9). A simple 2-fluoro substitu-
tion on the phenylsulfonyl group (1c) increased the
pKi from 7.9 to 8.9. Polar substituents on the phen-
ylsulfonyl group such as methylsulfonyl or methylsulf-
onamide led to markedly lower activity (1g and 1h).
Naphthyl and some heteroaryl sulfonyl groups were well
tolerated (1i–k). The effect of substitution at positions 2
and 6 of the 3,4-dihydro-2H-benzo[1,4]-oxazine core
scaffold was also investigated. These substitution posi-
tions were chosen for synthetic considerations. Small
lipophilic substituents such as Me, F, Cl, and methoxy
on position 6 increased 5-HT6 affinity (1l–q). The bulky
6-tert-butyl group lowered affinity (1s). Analogs with
substitution on position 2 of the 3,4-dihydro-2H-ben-
zo[1,4]oxazine core were synthesized according to
Scheme 2 using corresponding substituted chloroacetyl
chlorides. The 2,2-dimethyl and 2,2-spirocyclobutyl
substituted compounds (Table 2, 15 and 16) had slightly
better affinity than the nonsubstituted analog (1c).
yields. The nitro compound 6 was reduced to aniline 7
by sodium dithionite in a mixture of ethanol and water
at ꢀ70 °C in 68% yield. Compound 7 was then reacted
with chloroacetyl chloride in a mixture of acetonitrile
and water in the presence of sodium bicarbonate at
room temperature for 30 min. followed by heating at
60 °C for 10 h to afford benzoxazinone 8.7 Reduction
by BH3 in THF followed by sulfonylation in dichloro-
methane in the presence of pyridine gave aryl bromide
10. The piperazine moiety was installed by a palladi-
um/Binap-catalyzed aryl amination reaction using
1-Boc-piperazine.8 The final piperazinyl benzoxazine
compound 1m was obtained after deprotection in etha-
nolic HCl and isolated as the HCl salt.
Alternatively, bromodihydrobenzoxazine 9 was protect-
ed with Cbz to give 12. Subsequent amination with
1-Boc-piperazine followed by hydrogenolysis to remove
Cbz gave intermediate 14, which was then reacted with
different arylsulfonyl chlorides to afford the desired
compounds1 after removal of the Boc group under acid-
ic conditions (Scheme 3).
O
O
OH
O
O
OH
OH
O
N+
N+
Br
NH2
Br
Br
NH
c
Br
NH
b
a
d
O
O
CH3
CH3
CH3
CH3
CH3
5
6
7
8
9
O
HN
O
BocN
O
F
O
S
O
g
F
O
S
O
F
O
S
O
Br
N
e
f
N
N
N
N
CH3
CH3
CH3
11
10
1m
Scheme 2. Synthesis of 3,4-dihydro-2H-benzo[1,4]oxazine derivative 1m. Reagents and conditions: (a) Br2, acetic acid, 95%; (b) Na2S2O4, EtOH/
water, heat, 68%; (c) ClCH2COCl, 95%; (d) BH3–DMS, THF, reflux, 89%; (e) 2-F-PhSO2Cl, pyridine, CH2Cl2, 75%; (f) 1-Boc-piperazine, Binap,
Pd2(dba)3, NaO-t-Bu, toluene, 52%; (g) HCl, EtOH, reflux, 90%.

3506
S.-H. Zhao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3504–3507
Table 1. Biological activities of novel 3,4-dihydro-2H-benzo[1,4]oxazine derivatives: 5-HT6 affinity, hERG inhibition and brain to plasma ratio in
rats
O
R
HN
Ar
N
N
S
O
O
Compound
R
Ar
5-HT6 pKia
hERG at 1 lMb
clogPc
B/Pd
1a
1b
H
2,3-Dichloro-Ph
Ph
2-F-Ph
10.2
7.9
8.9
8.5
8.7
8.0
6.7
6.5
9.1
9.3
9.2
8.6
9.4
8.7
9.2
9.4
9.8
9.4
8.1
8.9
H
1c (R1485)
H
9.9 3.3%
28.2 5.4%
2.90
3.34
1.8
1.6
1.4
1d
1e
1f
H
2-Cl-Ph
3-Cl-Ph
3-F-Ph
H
H
1g
1h
1i
H
3-MeSO2NH-Ph
2-MeSO2-Ph
Naphthalen-1-yl
H
Cl
1j
Cl
Cl
Benzo[1,2,5]oxadiazol-4-yl
Quinolin-8-yl
Ph
1k
1l
Me
Me
Me
F
23.2 9.9%
26.5 1.8%
14.2 0.8%
26.0 3.5%
45.2 12%
3.25
3.45
2.79
3.11
3.54
1.0
1m
1n
1o
1p
1q
1r
1s
1t
2-F-Ph
2-CN-Ph
2-F-Ph
2-F-Ph
2-F-Ph
Ph
Cl
2.7
0.8
MeO
MeO
t-Bu
F
34.0 3.8%
2.78
2-F-Ph
4-Amino-Ph
a Standard deviation of the pKi estimates was 0.05–0.3.
b Inhibition of hERG outward current at 1 lM concentration; mean SEM; baseline = 0%, n = 2 cells.11
c Ko/w–clogP values (calculated values of octanol–water partition coefficient) calculated using KowWin (V1.57, Syracuse Research, North Syracuse,
NY, USA).
d Whole brain homogenate to plasma ratio at 1 h post 2 mg/kg iv dosing in rats, drug concentrations were determined by HPLC–MS/MS.
Several benzoxazine derivatives with basic side chains
other than piperazine were synthesized (Table 2). Simi-
larly to the synthesis of the piperazine derivatives, the
homopiperazine, 2,6-dimethylpiperazine, and dimeth-
ylaminoethylamine derivatives (18, 19 and 22) were syn-
thesized by the palladium-catalyzed amination reaction
of the corresponding aryl bromide with the required
amine. The aniline compound (17) was synthesized using
benzophenone imine as an ammonia surrogate.9 The
piperidine derivatives (20, 21) were synthesized via the
Suzuki coupling reaction of the corresponding aryl
bromide with 4-(4,4,5,5-tetramethyl-[1,3,2]dioxaboro-
lan-2-yl)-3,6-dihydro-2H-pyridine- 1-carboxylic acid
tert-butyl ester,10 followed by hydrogenation and
removal of the Boc group. The homopiperazine and
the piperidine derivatives (19, 20 and 21) had 5-HT6
affinity comparable to the corresponding piperazine
analog. Other side chains led to lower 5-HT6 affinity.
Among the basic side chains investigated, piperazine
derivatives offered the combination of good 5-HT6 affin-
ity and ease of chemical synthesis.
such as Me, F, and Cl on the aromatic rings seem to
increase hERG blockade, with Cl substitution having
a stronger effect (1d, 1m, 1o and 1p). The more polar
cyano substituted analog (1n) had weaker hERG inhibi-
tion than the fluoro analog (1m). In fact, a correlation
seems to exist between calculated logP values and the
measured hERG inhibition, with the methoxy com-
pound 1r being the outlier. Within this chemical series,
the general trend is that the more lipophilic the mole-
cule, the stronger its hERG inhibitory activity.
Several compounds were profiled in rat in vivo following
2 mg/kg single intravenous dose (iv). The brain to plas-
ma ratios at 1 h post iv dose were calculated from drug
levels in whole brain homogenate versus plasma. The
drug levels were determined by using HPLC–MS/MS.
The B/P ratios for the compounds tested ranged from
0.8 to 2.7, a significant improvement over our initial lead
RO65-7674 (2) or SB-271046 (3).5 This improvement is
probably due to the fact that template 1 has fewer het-
eroatoms than 2, and is also devoid of the sulfonamide
NH hydrogen bond donor in 3.
Compounds with high 5-HT6 affinity (pKi > 8.5) and
acceptable initial DMPK properties, such as good sta-
bility in microsomes and low cytochrome P450 inhibi-
tion, were selected for hERG inhibition measurement
by conventional patch clamp electrophysiology assays.11
Inhibition of hERG current by the tested compounds at
1 lM concentration ranged from 10% to 45% depending
on the substitution (Table 1). Lipophilic substituents
Compound 1c (R1485) showed good B/P ratio (1.8) and
low hERG inhibition (10% at 1 lM) while maintaining
high affinity for the 5HT6 receptor (pKi = 8.9). Its
hERG IC20value at 37 °C was determined using four
concentrations in a patch clamp physiology assay and
shown to be 2.2 lM, representing a significant improve-
ment over our initial lead molecule 2 (IC20 = 0.2 lM). In

S.-H. Zhao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3504–3507
Table 2. 5-HT6 affinity of 3,4-dihydro-2H- benzo[1,4]oxazines
3507
nists. The 5-HT6 affinity of these compounds is depen-
dent on the substitution on the aromatic rings. A
general correlation between lipophilicity and hERG
inhibition was observed in this chemical series. R1485
was selected for further development because of its over-
all superiority, and was eventually advanced to human
clinical trials.
Compound
Structure
5HT6 pKi
HN
N
O
O
F
F
O
N
15
9.5
S
O
HN
N
O
16
17
18
19
9.2
6.2
7.4
8.4
N
S
O
Acknowledgments
O
F
O
The authors thank Keith A. Walker, Ralph N. Harris
III, Hans Maag, and Zhongjiang Jia for helpful
discussions.
H2N
N
S
O
O
HN
N
O
O
N
S
References and notes
O
Cl
H
N
1. Davies, S. L.; Silvestre, J. S.; Guitart, X. Drugs Future
2005, 30, 479.
O
N
N
S
2. Mitchell, E. S.; Neumaier, J. F. Pharmacol. Ther. 2005,
108, 320.
O
Cl
O
3. Bourson, A.; Borroni, E.; Austin, R. H.; Monsma, F. J.;
Sleigh, A. J. J. Pharmacol. Exp. Ther. 1995, 274, 173.
4. (a) Boes, M.; Riemer, C.; Stadler. H. Eur. Pat. Appl.
EP941994, 1999; (b) Stadler, H.; Boes, M.; Borroni, E.;
Bourson, A.; Martin, J.; Poli, S.; Riemer, C.; Sleight, A. J.
37th IUPAC Congress, Berlin, August 14–19, 1999;
Abstract MM-7.
5. (a) Bromidge, S. M.; Brown, A. M.; Clarke, S. E.;
Dodgson, K.; Gager, T.; Grassam, H. L.; Jeffrey, P. M.;
Joiner, G. F.; King, F. D.; Middlemiss, D. N.; Moss, S. F.;
Newman, H.; Riley, G.; Routledge, C.; Wyman, P.
J. Med. Chem. 1999, 42, 202; (b) Bromidge, S. M.; Clarke,
S. E.; Gager, T.; Griffith, K.; Jeffrey, P.; Jennings, A. J.;
Joiner, G. F.; King, F. D.; Lovell, P. J.; Moss, S. F.;
Newman, H.; Riley, G.; Rogers, D.; Routledge, C.;
Serafinowska, H.; Smith, D. R. Bioorg. Med. Chem. Lett.
2001, 11, 55.
HN
HN
O
N
20
21
S
8.4
8.3
O
F
O
O
N
S
F
O
F
O
H
O
N
N
22
6.9
S
F
O
N
6. Sleight, A. J.; Boess, F. G.; Bos, M.; Levet-Trafit, B.;
Riemer, C.; Bourson, A. Br. J. Pharmacol. 1998, 124, 556.
7. Combs, D. W.; Rampulla, M. S.; Bell, S. C.; Klaubert, D.
H.; Tobia, A. J.; Falotico, R.; Haertlein, B.; Lakas-Weiss,
C.; Moore, J. B. J. Med. Chem. 1990, 33, 380.
8. Yang, B. H.; Buchwald, S. L. J. Organomet. Chem. 1999,
576, 125.
9. Wolf, J. P.; Ahman, J.; Sadighi, J. P.; Singer, R. A.;
Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6367.
10. Eastwood, P. R. Tetrahedron Lett. 2000, 41, 3705.
11. Ly, J. Q.; Shyy, G.; Misner, G. D. L. J. Assoc. Lab. Auto.
2005, 10, 225.
the functional assay, 1c was devoid of agonist activity
and behaved as an antagonist at 5-HT6 receptor. Profil-
ing against standard receptor, ion channel, and uptake
system panels including other serotonergic receptor sub-
types indicated that R1485 was >100-fold selective over
>50 other targets.
In summary, a series of novel 3,4-dihydro-2H-ben-
zo[1,4]oxazine derivatives was designed and synthesized,
and found to comprise potent 5-HT6 receptor antago-