De novo design of a picomolar nonbasic 5-HT1B receptor antagonist

Article abstract of DOI:10.1021/jm901200t

We describe herein the discovery of novel, de novo designed, 5-HT _1B receptor antagonists that lack a basic moiety and that provide improved hERG and in vitro phospholipidosis profiles. We used a known 5-HT _1B antagonist template as our starting point and focused on replacing the piperazine moiety. Pyrazole-based ideas were designed, and synthesized, among a small library of piperazine replacements. To our knowledge, these are the first, potent, nonbasic, functionally active antagonists of the 5-HT _1B receptor.

Full text of DOI:10.1021/jm901200t

1876 J. Med. Chem. 2010, 53, 1876–1880
DOI: 10.1021/jm901200t
De Novo Design of a Picomolar Nonbasic 5-HT1B Receptor Antagonist
David A. Nugiel,† Jennifer R. Krumrine,*,† Daniel C. Hill,† James R. Damewood, Jr.,† Peter R. Bernstein,†
Cynthia D. Sobotka-Briner,§ JianWei Liu,§ Anna Zacco,§ and M. Edward Pierson‡
†Department of CNS Chemistry, §Department of Neuroscience Biology, and ‡Neuroscience Therapeutic Area, AstraZeneca Pharmaceuticals,
1800 Concord Pike, Wilmington, Delaware 19850
Received August 11, 2009
We describe herein the discovery of novel, de novo designed, 5-HT1B receptor antagonists that lack a
basic moiety and that provide improved hERG and in vitro phospholipidosis profiles. We used a known
5-HT1B antagonist template as our starting point and focused on replacing the piperazine moiety.
Pyrazole-based ideas were designed and synthesized among a small library of piperazine replacements.
To our knowledge, these are the first potent, nonbasic, functionally active antagonists of the 5-HT1B
receptor.
Introduction
demonstrate subnanomolar potency, are selective for the
5-HT1B receptor, and are active in a pharmacodynamic model
of 5-HT1B receptor function. To our knowledge, these are the
first potent, nonbasic, functionally active antagonists of the
5-HT1B receptor. In addition, the hERG and phospholipido-
sis safety characteristics of this series are improved over the
piperazine series, which was our starting point for design.
Substituted quinolones8 and chromans9 are known 5-HT1B
antagonist templates. We used a chroman template as the
starting point for our design efforts (1a, Figure 1),10 as this
core had demonstrated superior drug metabolism and phar-
macokinetics (DMPK) properties compared to quinolones.11
With a focus on replacing the basic piperazine moiety, a
variety of ideas were generated and ranked using the de novo
design tool NovoFLAP.12 NovoFLAP is a unique, ligand
based, computer aided design approach that generates med-
icinally relevant ideas starting from compounds known to be
active at a biological target of interest. A pyrazole-based idea
(1b, Figure 1) was found to be particularly interesting, espe-
cially in light of a known inhibitor-bound crystal structure of
transforming growth factor β (TGFβ) kinase.13 This crystal
structure exemplified a hydrogen bond between an aspartic
acid residue and the pyrazole ring of the inhibitor. Motivated
by these findings, several pyrazole-based ideas were included
in a small library of piperazine replacements we designed and
synthesized. We pursued these initial pyrazoles using the
quinolone core because it was more amenable to library
synthesis. The library was generated using the chemistry
shown in Scheme 1.
5-HT1 is the largest subgroup of the 5-hydroxytryptamine
(5-HTa) family of receptors.1 There are five receptor subtypes
(5-HT1A, 5-HT1B, 5-HT1D, 5-HT1E, and 5-HT1F), all of which
are G-protein-coupled seven transmembrane receptors.2 The
5-HT1B receptor has been targeted for its potential in the
treatment of depression, anxiety, and other serotonergic
neurotransmission related psychiatric disorders. It has been
suggested that antagonists of the terminal 5-HT1 autoreceptors
(5-HT1B/1D) may effect immediate 5-HT release, thereby
increasing 5-HT transmission at the synapses. It is believed
that this may provide faster clinical antidepressant activity
than currently available therapies.3 Recent studies on 5-HT1B
selective antagonists are beginning to substantiate this hypothesis.4
It is generally accepted that known 5-HT receptor agonists
and antagonists bind protonated amino groups to the highly
conserved aspartic acid on transmembrane helix 3 (TM3).5
These compounds have a characteristic basic amine site (pKa >
7.5) embedded in an overall lipophilic (log D >2) molecule. It
is well-known that aryl-containing lipophilic bases6 have a
propensity to bind to the human ether-a-go-go related gene
(hERG) channel. In addition, such amphiphilic cations have
been implicated in phospholipidosis.7
Results
We describe here a series of novel, de novo designed 5-HT1B
receptor antagonists that lack a basic moiety. These compounds
*To whom correspondence may be addressed. Phone: (þ)1 302 885
4772. Fax: (þ)1 302 886 5382. E-mail: Jennifer.Krumrine@AstraZeneca.
com.
Michael addition of 2-bromo-5-methylaniline into diethy-
lacetylene dicarboxylate gave adduct 2 in 95% yield. This
precursor was cyclized to the quinolone 3 under thermal
conditions in 65% yield. Transamidation using sodium bis-
(trimethylsilyl)amide gave the key intermediate 4 in 90%
yield. Using a Suzuki-based boronic acid coupling strategy,14
a 40-member library of piperazine replacements was gener-
ated as exemplified by the synthesis of 5. The chemistry was
designed so thatthe piperazine replacementwasinserted atthe
last step, maximizing the library’s diversity potential.
a Abbreviations: 5-HT, 5-hydroxytryptamine; DMPK, drug metabo-
lism and pharmacokinetics; GTPγS, guanosine γ-thiophosphate;
TBTU, O-(benzotriazol-1-yl)-N,N,N0,N0-tetramethyluronium tetraflu-
oroborate; SPA, scintillation proximity assay; CHO, Chinese hamster
ovary; EDTA, ethylenediaminetetraacetic acid; TM3, transmembrane
helix 3; hERG, human ether-a-go-go related gene; TGFβ, transforming
growth factor β; P-gp, permeability glycoprotein; PDB, protein crystal
structure database; SEM, standard error of the mean; DCM, dichloro-
methane; EtOAc, ethyl acetate; THF, tetrahydrofuran; DME, 1,2-
dimethoxyethane; fcc, flash column chromatography.
r
pubs.acs.org/jmc
Published on Web 01/20/2010
2010 American Chemical Society

Brief Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1877
Table 1. 5-HT1B Receptor Binding Affinities and % Antagonist Effect
on [35S]GTPγS Binding in CHO Cells Expressing 5-HT1B Receptors,
and Selected Measured pKa Values
5-HT1B Ki (nM) % antagonist
(pKi ( SEM) effect ( SEM
pKa
base 1
pKa
base 2
compd
AZD1134 1.7 (8.76 ( 0.055) 150 ( 13
1a
5
7.3 ( 0.05 2.5 ( 0.05
7.66 ( 0.05
12 (7.91 ( 0.043) 120 ( 2.0
9.3 (8.03 ( 0.075) 33 ( 0.040
0.43 (9.37 ( 0.23) 90 ( 1.6
0.96 (9.02 ( 0.52) 62 ( 7.1
8
11
2.87 ( 0.01 2.05 ( 0.01
3.28 ( 0.05
Figure 1. Initial de novo design of a pyrazole replacement for the
piperazine ring in a 5-HT1B antagonist using NovoFLAP.
Scheme 2a
Scheme 1a
a Reagents: (a) diethylacetylene dicarboxylate, EtOH; (b) Dowtherm,
250 °C; (c) 4-morpholinoaniline, NaHMDS, THF; (d) boronic ester,
PdCl2(PPh3)2, K2CO3, 1:1 EtOH/water, 80 °C.
The library produced three hits having good binding affi-
nity for the 5-HT1B receptor, all of which contained pyrazole-
based piperazine replacements. The most potent was 5 (Table 1)
with a binding affinity of 9.3 nM (pKi = 8.03). Unfortunately,
functional activity evaluation of 5 revealed it to be a partial
agonist (33% antagonist effect in a GTPγS binding assay). As
we were interested in a compound having greater antagonist
properties, we continued to optimize this initial promising lead.
Our extensive experience with this class of 5-HT1B templates
allowed us to effectively incorporate the pyrazole-based piper-
azine replacement into a chroman-based template having all the
physiochemical and DMPK properties needed for an in vivo
tool compound. The chroman-based chemistry is detailed in
Scheme 2. Subjecting ethyl 8-bromo-6-fluoro-4-oxo-4H-chro-
mene-2-carboxylate15 6 to the same Suzuki reaction conditions
shown in Scheme 1 gave the desired chromenone acid 7 in 58%
yield. Compound 14 (Scheme 3) was coupled with chromenone 7
using TBTU in DMF to give the desired amide 8 in 85% yield.
Initial attempts to reduce chromenone7 to chroman 9 using
catalytic hydrogenation on a Parr apparatus only resulted in
partial reduction at elevated temperature (70 °C) and pressure
(80 psi). High pressure and temperature hydrogenation of
chromenone 7 at 90 bar and 70 °C using an H-Cube16
apparatus gave a nearly quantitative yield of hydroxyl chro-
man 9. The hydroxyl group was then smoothly reduced using
triethylsilane in trifluoroacetic acid at 80 °C to give the desired
chroman 10 in 55% yield. Standard amide coupling condi-
tions using TBTU in DMF gave the desired amide 11 in 65%
yield.
a Reagents: (a) PdCl2(PPh3)2, K2CO3, pyrazole, DME/EtOH/water,
90 °C; (b) H2, MeOH/HOAc/IPA, 90 bar, 70 °C; (c) Et3SiH, TFA, 80 °C;
(d) 14, TBTU, DMF.
Scheme 3a
a Reagents: (a) n-BuLi, dihydro-2H-pyran-4(3H)-one, THF, -78 °C;
(b) CuCl, Cu, 7 N NH3 in MeOH, 100 °C; (c) MeOH, TFA, 80 °C.
(pKi = 9.02) but with a disappointing functional activity
profile of a partial agonist (62% antagonist effect).17 Fortu-
nately, chromenone 8 had the desired functional activity
profile. It had high affinity for the 5-HT1B receptor (0.43 nM,
pKi = 9.37) and behaved as a full antagonist in our GTPγS
binding assay (90% antagonist effect). In addition, the com-
pound had moderate clearance and was a moderate perme-
ability glycoprotein (P-gp) substrate.18 8was evaluated at 10 μM
in a broad selectivity screen of 100 targets (MDS Pharma)
and was found to be inactive except at the 5-HT1B receptor.19
Shown in Table 2 is the selectivity profile of 8 versus several
5-HT receptor subtypes that are closely related to the 5-HT1B
receptor and were among the 100 targets in the MDS Pharma
panel. These overall qualities were sufficient to examine the
Discussion
Table 1 shows the binding affinity and functional activity of
the chroman-based analogues. The fully reduced chroman 11
was a potent compound with a binding affinity of 0.96 nM

1878 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Nugiel et al.
Table 2. Selectivity Profile of 8 versus Several 5-HT Receptor Subtypesa
receptor
% effect at 10 μM
species
5-HT1A
5-HT2A
5-HT2B
5-HT4
-8
-5
4
human
human
human
guinea pig
30
a 8 was tested in several binding assays available through MDS
Pharma for 5-HT receptors closely related to the 5-HT1B receptor and
found to be selective. Additional information on assay protocols and
standards are available through MDS Pharma Services.19
compound’s behavior in vivo. Therefore, 8 was evaluated in
an agonist induced hypothermia model of central 5-HT1B
receptor function in the guinea pig.20 Details of the protocol
can be found in the Supporting Information. It has been
demonstrated previously that 5-HT1B receptors and not
5-HT1D receptors mediatethe hypothermicresponsein guinea
pigs.21 The results are shown in Figure 2. Here, 8 shows a dose
dependent reversal of agonist-induced hypothermia while
having no effect on body temperature when given alone,
indicating its temperature modifying effect was mediated by
a 5-HT1B antagonist-based mechanism.
Figure 2. Dose dependent effect of 8 in an agonist induced guinea
pig hypothermia model.
pH is not needed for in vitro and in vivo activity at the
5-HT1B receptor.
Experimental Section
In addition to the desirable in vitro and in vivo activities
22
described, pyrazoles generally had hERG IC50 improved
Biological Assays. 1. 5-HT1B Receptor Competition Binding
Studies. Frozen membrane preparations of stably transfected
Chinese hamster ovary (CHO) cells expressing 5-HT1B receptors
(Perkin-Elmer, 10 mg/mL) were rapidly thawed, briefly vor-
texed, and diluted in assay buffer (AB) containing 50 mM Tris-
HCl, 4 mM MgCl2, 4 mM CaCl2, 1 mM EDTA, pH 7.4.
Membranes were resuspended in AB at 125 μg/mL and dis-
pensed at 80 μL per well for a final concentration of 10 μg/well.
Assay plates are white, 96-well OptiPlates (PerkinElmer). Test
compounds were solubilized in DMSO at 10 mM. For concen-
tration response analyses, 11 serial dilutions (10 μM to 170 pM,
final concentration) of compound were prepared in DMSO
from 10 mM stock solutions. The stock radioligand was diluted
in AB/0.1% ascorbic acid to 10ꢀ the concentration needed for
the assay. AZD113424 (FW 521.6) was used to define nonspe-
cific binding at a final assay concentration of 1 μM. Final assay
volumes per well were 2 μL of compound/nonspecific, 80 μL
membranes, 20 μL of [3H]GR125743 (Amersham, TRK1046) at
0.5 nM final concentration. The assay plates were incubated for
1 h at room temperature with shaking. Yttrium silicate-WGA
SPA beads (Amersham RPNQ0011 or RPNQ0023) were resus-
pended at 15 mg/mL in AB, and an amount of 100 μL was added
for 1.5 mg/well final. Plates were incubated for an additional
30 min with high speed shaking to keep the beads in suspension.
Then the plates were centrifuged and counted on the Top Count
(PerkinElmer).
over the range typically observed among the piperazine
analogues, (0.7 to >33 μM). For example, the hERG IC50
of piperazine 1a was 19 μM (95% confidence interval,
15.6-24.7 μM) and 81% maximum observed effect at 100 μM.
In comparison, most pyrazoles had IC50 > 33 μM, the upper
limit of detection for our standard assay, and all had IC50
>
25 μM. For 8, the hERG IC50 was 95 μM (extrapolated value
due to precipitation at 100 μM, 95% confidence interval of
46.5-193 μM) with only a 26% maximum observed effect at
33 μM. Further, no pyrazole showed activity in our in vitro
phospholipidosis assay,23 an improvement over the pipera-
zines. For piperazine 1a, the maximum observed effect at
300 μM was 81%. For pyrazoles 5 and 8, the maximum
observed effect at 300 μM was 12% and 9%, respectively,
where 20% is the threshold of detection for phospholipogenic
activity. Compound 11 was not tested. In comparison, the
piperazine analogues of 5 and 11 had maximum observed
effects at 300 μM of 105% and 99%, respectively. The
piperazine analogue of 8 was not tested.
Discussion and Conclusions
With the de novo design tool NovoFLAP, pyrazole repla-
cements were identified for the ubiquitous piperazine moiety
found in 5-HT1B antagonists, which is believed to form a key
interaction with an aspartic acid. Analysis of the protein
crystal structure database (PDB) provided supporting evi-
dence for a pyrazole ring having the capacity to hydrogen-
bond with an aspartic acid residue in the required manner. A
40-membered scouting library was designed in which three
pyrazole-containing analogues with excellent 5-HT1B binding
affinity were found. Further refinement led to compounds
with subnanomolar affinity for the 5-HT1B receptor and a
range of in vitro functional activities from partial agonist to
antagonist. Additionally, these same compounds exhibited
acceptable DMPK properties, excellent profiles with respect
to in vitro hERG and phospholipidosis assays, and potent
5-HT1B antagonist activity in our guinea pig hypothermia
pharmacodynamic model.
Data were analyzed by calculating IC50 and Ki using the
Kenakin correction for ligand depletion (see equations below),
2
1=2
B ¼ ½ðKD þ LT þ RTÞ -fðKD þLT þ RTÞ -4RTLTg ꢁ=2
Ki ¼ ð0:5BÞðIC50ÞðKDÞ=½ðLTRTÞ þ 0:5BðRT - LT þ 0:5B -KDÞꢁ
where LT and RT are the total concentrations of the ligand and
receptor, respectively (in nM).
2. GTPγ35S Functional Antagonist Assay. Membranes and
compound dilutions used for the GTPγS assay were the same as
those in the 5-HT1B receptor binding assay. In this scintillation
proximity assay (SPA), membranes (15 mg of protein/well) and
WGA PVT beads (50 mg/well) (Amersham RPNQ0001) were
preincubated in bulk in assay buffer containing 20 mM HEPES,
100 mM NaCl, 10 mM MgCl2, 0.1% BSA, pH 7.4, for 30 min.
After addition of GTPγ35S (200 pM final assay concentration)
and GDP (10 mM final) to the membrane/bead mixture, 150 mL
of the mixture was added to 96-well OptiPlates (PerkinElmer)
containing 2 mL of compounds. After 5 min of preincubation,
an amount of 50 mL of buffer or 108 nM 5-HT (27 nM final
We have thus demonstrated, for the first time, that
an amine that will be positively charged at physiological

Brief Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1879
assay concentration, EC80) was added. The plates were incu-
bated for 1 h with shaking and then counted in a TopCount
(PerkinElmer).
Percent effect with respect to basal (buffer unstimulated) and
stimulated (EC80 of 5-HT) response was determined. IC50 was
defined as 50% inhibition of the stimulated response.
isolated by preparative HPLC to give the desired product as a
solid (83 mg, 5%). Mp 153 °C; LC/MS 2.30 min, m/z 430 (M þ
H); 1H NMR (300 MHz, CDCl3) δ 2.39 (s, 3H), 2.58 (m, 4H),
3.72 (m, 4H), 6.70 (m, 3H), 6.88 (m, 1H), 7.07 (m, 1H), 7.74 (m,
2H), 8.15 (m, 2H), 8.89 (bs, 1H), 9.11 (bs, 1H), 9.84 (bs, 1H).
HRMS: calcd M þ H = 430.4906, found M þ H = 430.4922.
7: Potassium 6-Fluoro-4-oxo-8-(1,3,5-trimethyl-1H-pyrazol-
4-yl)-4H-chromene-2-carboxylate. A slurry of ethyl 8-bromo-6-
fluoro-4-oxo-4H-chromene-2-carboxylate (1.24 g, 3.93 mmol),
1,3,5-trimethyl-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-
1H-pyrazole (1.23 g, 5.21 mmol), bis(triphenylphosphine)-
palladium(II) chloride (276 mg, 0.39 mmol), potassium carbonate
(1.63 g, 11.8 mmol), DME (50 mL), water (21 mL), and ethanol
(14 mL) was vacuum degassed (three cycles) and then heated to
90 °C for 16 h. Reaction was cooled to room temperature
and filtered through Celite. Solids were washed with DME (3 ꢀ
20 mL), and filtrate was evaporated. DME was added (30 mL)
and evaporated (3ꢀ) to azeotrope water. This solid was suspended
in 50 mL of DME, filtered, washed with more DME, and dried
under high vacuum for 3 h. The pink crude solid potassium salt
was typically carried on in further reactions without purification
(0.66 g, 58%). LC/MS 0.82 min, m/z 317 (M þ H); 1H NMR (300
MHz, DMSO-d6) δ ppm 2.13 (s, 3 H), 2.23 (s, 3 H), 3.74 (s, 3 H),
6.68 (s, 1 H), 7.51 (dd, J = 9.0, 2.8 Hz, 1 H), 7.57-7.77 (m, 1 H).
8: 6-Fluoro-N-(6-(4-methoxytetrahydro-2H-pyran-4-yl)pyri-
din-3-yl)-4-oxo-8-(1,3,5-trimethyl-1H-pyrazol-4-yl)-4H-chromene-
2-carboxamide. To a solution of potassium 6-fluoro-4-oxo-
8-(1,3,5-trimethyl-1H-pyrazol-4-yl)-4H-chromene-2-carboxylate
(240 mg, 0.68 mmol) in DMF (3 mL) was added 6-(4-methox-
ytetrahydro-2H-pyran-4-yl)pyridin-3-amine (111 mg, 0.53 mmol),
and Hunig’s base (200 μL, 1.15 mmol). When all had dissolved,
2-(1H-benzo[d][1,2,3]triazol-1-yl)-1,1,3,3-tetramethylisouronium
tetrafluoroborate (340 mg, 1.06 mmol) was added all in one
portion and the mixture was stirred for 1 h. DMF was evapo-
rated. Residue was mixed with EtOAc (30 mL) and extracted
with 20% K2CO3 (2 ꢀ 10 mL), then brine. Organic layer was
dried over Na2SO4, filtered, and evaporated to give an oil which
was purified by fcc on silica (12 g) DCM to 10% MeOH in DCM.
This gave a light-yellow solid (230 mg, 85%). Mp 132-134 °C
LC/MS 2.22 min, m/z 507 (M þ H); 1H NMR (500 MHz, CDCl3)
δ ppm 1.57 (br. s., 3 H), 1.95 (d, J = 13.1 Hz, 2 H), 2.06 - 2.42 (m,
6 H), 2.80 (s, 3 H), 2.96-3.16 (m, 3 H), 3.68-4.07 (m, 4 H), 7.34
(dd, J = 7.9, 3.1 Hz, 1 H), 7.54 (d, J = 8.5 Hz, 1 H), 7.87 (dd, J =
7.8, 3.2 Hz, 1 H), 8.10-8.34 (m, 2 H), 8.59 (d, J = 2.4 Hz, 1 H).
HRMS: calcd M þ H = 507.2038 found M þ H = 507.2061.
9: 6-Fluoro-4-hydroxy-8-(1,3,5-trimethyl-1H-pyrazol-4-yl)chro-
man-2-carboxylic Acid. A solution of potassium 6-fluoro-4-
oxo-8-(1,3,5-trimethyl-1H-pyrazol-4-yl)-4H-chromene-2-carboxy-
late (780 mg, 2.47 mmol) in MeOH (100 mL)/acetic acid
(40 mL)/IPA (60 mL) was pumped through the H-Cube using
10% Pd/C catalyst (medium CatCart part no. THS 01121, 1 mL/
min, 90 barr of H2, 70 °C, hydrogen controlled). Eluate was eva-
porated under reduced pressure and residue was purified by fcc on
silica (12 g) DCM to 20% MeOH in DCM to give a white solid
(750 mg, 95%). LC/MS 0.75 min, m/z 321 (M þ H); 1H NMR (500
MHz, MeOD) δ ppm 2.10 (s, 3 H), 2.15 (s, 3 H), 2.45-2.53 (m,
1 H), 3.71 (s, 3 H), 4.54-4.62 (m, 1 H), 4.92 (dd, J = 8.7, 6.3 Hz,
1 H), 6.74 (dd, J= 8.9, 3.1 Hz, 1 H), 7.12 (dd, J= 9.0, 3.2 Hz, 1 H).
10: 6-Fluoro-8-(1,3,5-trimethyl-1H-pyrazol-4-yl)chroman-2-
carboxylic Acid. To a flask containing 6-fluoro-4-hydroxy-
8-(1,3,5-trimethyl-1H-pyrazol-4-yl)chroman-2-carboxylic acid
(750 mg, 2.34 mmol) was added triethylsilane (10 mL, 62.61 mmol)
followed by trifluoroacetic acid (5 mL, 64.90 mmol). The
mixture was heated to 90 °C for 3 h. Excess reagents were
evaporated under reduced pressure. Residue was dissolved in
20 mL of 3 N HCl and extracted with CHCl3 (4 ꢀ 20 mL).
Organic layers were dried over Na2SO4, filtered, and evapo-
rated. Crude product was purified by fcc on silica (4 g) DCM to
10% MeOH/DCM to give 6-fluoro-8-(1,3,5-trimethyl-1H-pyr-
azol-4-yl)chroman-2-carboxylic acid (392 mg, 55.0%). LC/MS
Chemistry. All commercially available reagents and solvents
were used without further purification unless otherwise stated.
Automated flash chromatography was performed on an ISCO
CombiFlash using Biotage Flash cartridges with peak detection
at 254 nm. 1H NMR spectra were recorded on a Bruker Avance
instrument operating at 300 or 500 MHz with tetramethylsilane
or residual protonated solvent used as a reference. Low resolu-
tion mass spectra were recorded using a Waters Micromass ZQ
quadrupole mass spectrometer linked to a Hewlett-Packard
HP1100 LC system with UV detector at 254 nm. Sample
injection is done by a Gilson 215 autosampler. The mobile
phase consisted of a gradient using water with 0.1% formic acid
and acetonitrile with 0.1% formic acid. The spectrometer has an
electrospray source operating in positive and negative ion mode.
Additional detection is achieved using a Sedex 55 ELS detector.
All compounds were found to have a purity of >95% by this
method. High-resolution mass spectra were recorded on an
Agilent Technologies 6210 time-of-flight LC/MS spectrometer.
2: (E)-2-(2-Bromo-5-methylphenylamino)but-2-enedioic Acid
Diethyl Ester. A solution of 2-bromo-5-methylphenylamine
(1.4 g, 10.0 mmol) in ethanol (10 mL) was treated with diethy-
lacetylene dicarboxylate (1.7 mL, 10.0 mmol) and heated to
60 °C for 2 h. The mixture was cooled and the solvent evapo-
rated at reduced pressure. Chromatography (silica, 0-30%
EtOAc/hexane) gave the product as an oil (3.4 g, 96%). LC/
MS 2.96 min, m/z 356 (M þ H); 1H NMR (300 MHz, CDCl3) δ
1.08 (t, 3H), 1.12 (t, 3H), 2.02 (s, 3H), 4.07 (q, 2H), 4.16 (q, 2H),
5.49 (s, 1H), 6.37 (m, 1H), 7.22 (m, 2H), 8.53 (bs, 1H).
3: 8-Bromo-5-methyl-4-oxo-1,4-dihydroquinoline-2-carboxylic Acid
Ethyl Ester. A solution of (E)-2-(2-bromo-5-methylphenylami-
no)but-2-enedioic acid diethyl ester (3.4 g, 9.6 mmol) in Dow-
therm (5 mL) was added dropwise to Dowtherm (30 mL) at 250 °C
over 2 min. During the addition the internal temperature was kept
above 240 °C. After 5 min the mixture was cooled to room
temperature. The mixture was subjected directly to chromato
graphy (silica, 0-70% EtOAc/hexane) to give the desired pro-
duct as a solid (2.5 g, 84%). Mp 135 °C; LC/MS 2.40 min, m/z 310
(M þ H); 1H NMR (300 MHz, CDCl3) δ 1.09 (t, 3H), 2.43 (s, 3H),
4.05 (q, 2H), 6.54 (s, 1H), 6.89 (m, 1H), 7.58 (m, 1H), 9.20 (bs, 1H).
4: 8-Bromo-5-methyl-4-oxo-1,4-dihydroquinoline-2-carboxylic Acid
(4-Morpholin-4-ylphenyl)amide. A solution of 8-bromo-5-meth-
yl-4-oxo-1,4-dihydroquinoline-2-carboxylic acid ethyl ester (4.6
g, 14.9 mmol) and 4-morpholin-4-ylphenylamine (2.65 g, 14.9
mmol) in THF (40 mL) was cooled to 0 °C. Sodium bis-
(trimethylsilyl)amide (45 mL, 45 mmol) was added dropwise
over 2 min. The mixture was stirred at 0 °C for 1 h, then poured
into water and extracted with DCM. The organic layer was
separated and dried (MgSO4) and the solvent removed at
reduced pressure. The product was recrystallized from acetoni-
trile to give a yellow solid (5.0 g, 76%). Mp 144 °C; LC/MS 2.54
min, m/z 442 (M þ H); 1H NMR (300 MHz, CDCl3) δ 2.41 (s,
3H), 2.58 (m, 4H), 3.72 (m, 4H), 6.70 (m, 2H), 6.91 (m, 2H), 7.61
(m, 1H), 7.74 (m, 2H), 9.11 (bs, 1H), 9.84 (bs, 1H).
5: 5-Methyl-4-oxo-8-(1H-pyrazol-4-yl)-1,4-dihydroquinoline-
2-carboxylic Acid (4-Morpholin-4-yl-phenyl)amide. A suspen-
sion of 8-bromo-5-methyl-4-oxo-1,4-dihydroquinoline-2-car-
boxylic acid (4-morpholin-4-ylphenyl)amide (1.5 g, 4.59 mmol),
4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-1H-pyrazole (1.6 g,
6.88 mmol), and potassium carbonate (1.9 g, 13.76 mmol) in a
7:3:2 THF/EtOH/water mixture (70 mL) was treated with bis-
(triphenylphosphine)palladium(II) dichloride (0.322 g, 0.46 mmol)
and heated to 100 °C for 3 h. The mixture was cooled, dilu-
ted with water, extracted with DCM and the organic layer
separated, dried (MgSO4), and concentrated. The product was

1880 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Nugiel et al.
2.00 min, m/z 305 (M þ H); 1H NMR (500 MHz, CDCl3) δ ppm
2.00-2.47 (m, 8 H), 2.74-3.08 (m, 2 H), 3.74-4.10 (m, 3 H),
4.65 (dd, J = 7.5, 3.5 Hz, 1 H), 5.55 (br s, 1 H), 6.66 (dd, J = 8.5,
3.1 Hz, 1 H), 6.82 (dd, J = 8.4, 2.9 Hz, 1 H).
(8) Horchler, C. L.; McCauley, J. P., Jr.; Hall, J. E.; Snyder, D. H.;
Moore, W. C.; Hudzik, T. J.; Chapdelaine, M. J. Synthesis of novel
quinolone and quinoline-2-carboxylic acid (4-morpholin-4-yl-phe-
nyl)amides: a late-stage diversification approach to potent 5-HT1B
antagonists. Bioorg. Med. Chem. 2007, 15, 939–950.
11: 6-Fluoro-N-(6-(4-methoxytetrahydro-2H-pyran-4-yl)pyr-
idin-3-yl)-8-(1,3,5-trimethyl-1H-pyrazol-4-yl)chroman-2-carbo-
xamide. To a solution of 6-fluoro-8-(1,3,5-trimethyl-1H-pyra-
zol-4-yl)chroman-2-carboxylic acid (33 mg, 0.11 mmol) and
6-(4-methoxytetrahydro-2H-pyran-4-yl)pyridin-3-amine (22.6 mg,
0.11 mmol) in DMF (3 mL) containing Hunig’s base (70 μL,
0.40 mmol was added 2-(1H-benzo[d][1,2,3]triazol-1-yl)-1,1,3,3-
tetramethylisouronium tetrafluoroborate (52.2 mg, 0.16 mmol).
The mixture was stirred for 1 h, and then additional 2-(1H-
benzo[d][1,2,3]triazol-1-yl)-1,1,3,3-tetramethylisouronium tetr-
afluoroborate (52.2 mg, 0.16 mmol) was added. The mixture was
stirrred for an additional 1 h. DMF was evaporated, residue was
mixed with 20 mL of EtOAc and extracted with 20% K2CO3
(3 ꢀ 10 mL), then brine (10 mL). Organic layer was dried over
Na2SO4, filtered, and evaporated. Product was purified by fcc
on silica (4 g) DCM to 10% MeOH/DCM to give 6-fluoro-
N-(6-(4-methoxytetrahydro-2H-pyran-4-yl)pyridin-3-yl)-8-(1,3,5-
trimethyl-1H-pyrazol-4-yl)chroman-2-carboxamide (35.0 mg,
65.3%). LC/MS 2.29 min, m/z 495 (M þ H); 1H NMR
(300 MHz, CDCl3) δ ppm 0.62-1.48 (m, 2 H), 1.66 (br s,
1 H), 1.83-2.47 (m, 10 H), 2.47-2.79 (m, 1 H), 2.78-3.25
(m, 4 H), 3.56-4.10 (m, 6 H), 4.59 (d, J = 10.3 Hz, 1 H),
6.59-6.99 (m, 2 H), 7.45 (d, J = 8.6 Hz, 1 H), 7.89-8.24 (m,
2 H), 8.30-8.79 (m, 1 H). HRMS: calcd M þ H = 495.2402,
found M þ H = 495.2400.
(9) Bernstein, P.; Hill, D.; Nugiel, D.; Pierson, E.; Shenvi, A.; Jacobs,
R. Chroman Compounds as 5-HT1B Receptor Antagonists, Their
Preparation, Pharmaceutical Compositions, and Use in Therapy.
PCT Int. Appl. WO 2007053095, 2007.
(10) Information on this chroman was presented previously: Bernstein,
P. R. Challenge of CNS Drug Discovery. Presented at the 235th
National Meeting of the American Chemical Society, New Orleans,
LA, Apr 6-10, 2008; MEDI-143.
(11) Shenvi, A. Manuscript detailing the SAR of chroman and chro-
menone 5-HT1B ligands is in preparation.
(12) Damewood, J. R.; Lerman, C. L. Flexible Ligand Alignment
Protocols and Their Use in de Novo Design. Presented at the
234th National Meeting of the American Chemical Society, Bos-
ton, MA, Aug 19-23, 2007; CINF-34. NovoFLAP is a proprietary
computer-aided de novo tool that uses an evolutionary algorithm
(EA-Inventor is available from Tripos International, 1699 South Hanley
Rd, St. Louis, MO 63144) and a scoring function based on shape and
pharmacophoric features.
(13) Sawyer, J. S.; Beight, D. W.; Britt, K. S.; Anderson, B. D.; Campbell,
R. M.; Goodson, T.; Herron, D. K.; Li, H. Y.; McMillen,
W. T.; Mort, N.; Parsons, S.; Smith, E. C.; Wagner, J. R.; Yan,
L.; Zhang, F.; Yingling, J. M. Synthesis and activity of new aryl-
and heteroaryl-substituted 5,6-dihydro-4H-pyrrolo[1,2-b]pyra-
zole inhibitors of the transforming growth factor-beta type I
receptor kinase domain. Bioorg. Med. Chem. Lett. 2004, 14,
3581–3584.
(14) Suzuki, A. Synthetic studies via the cross-coupling reaction of
organoboron derivatives with organic halides. Pure Appl. Chem.
1991, 63, 419–422.
(15) Robinson, G. E.; Cunningham, O. R.; Dekhane, M.; McManus,
J. C.; O’Kearney-McMullan, A.; Mirajkar, A. M.; Mishra, V.;
Norton, A. K.; Venugopalan, B.; Williams, E. G. Successful
development and scale-up of a palladium-catalyzed amination
process in the manufacture of ZM549865. Org. Process Res. Dev.
2004, 8 (6), 925–930.
Acknowledgment. The authors thank Jim Hall for provid-
ing NMR support, Timothy Blake for MS support, Christo-
pher Holmquist for the initial library synthesis, Paul Warwick
for synthesis of intermediates, Paul Ciaccio for evaluating the
phospholipogenic potential, and Herbert Barthlow and Mat-
thew Bridgland-Taylor for evaluating the hERG potential.
(16) For references and instrument applications, see www.thalesnano.
com.
(17) This compound was resolved by chiral chromatography. Both
enantiomers are partial agonists.
Supporting Information Available: Preparation details of 14,
described in Scheme 3, preparative reverse phase chromato-
graphy, guinea pig hypothermia assay protocol. This material is
available free of charge via the Internet at http://pubs.acs.org.
(18) Moderate human cLint (60 (μL/min)/mg) and an efflux ratio of 5.9
(B:A/A:B ratio in MDR1 transfected cells is described in the
following: Hochman, J. H.; Pudvah, N.; Qiu, J.; Yamazaki, M.;
Tang, C.; Lin, J.; Prueksaritanont, T. Interactions of human
P-glycoprotein with simvastatin, simvastatin acid, and atorvastatin.
Pharm. Res. 2004, 21, 1686–1691.
(19) Unpublished results. A variety of receptors that might have inter-
fered with the in vivo experiments were specifically targeted. For
binding assay information and values for standard reference
compounds, see http://discovery.mdsps.com/Catalog/.
(20) Hatcher, J. P.; Slade, R. C.; Hagan, J. J. 5-HT1D receptors mediate
SKF 99101H-induced hypothermia in the guinea pig. J. Psycho-
pharmacol. 1995, 9, 234–241.
(21) Hagan, J. J.; Slade, P. D.; Gaster, L.; Jeffrey, P.; Hatcher, J. P.;
Middlemiss, D. N. Stimulation of 5-HT1B receptors causes hy-
pothermia in the guinea pig. Eur. J. Pharmacol. 1997, 331, 169–174.
(22) Bridgland-Taylor, M. H.; Hargreaves, A. C.; Easter, A.; Orme, A.;
Henthorn, D. C.; Ding, M.; Davis, A. M.; Small, B. G.; Heapy,
C. G.; Abi-Gerges, N.; Persson, F.; Jacobson, I.; Sullivan, M.;
Albertson, N.; Hammond, T. G.; Sullivan, E.; Valentin, J.-P.;
Pollard, C. E. Optimisation and validation of a medium-through-
put electrophysiology-based hERG assay using IonWorks HT.
J. Pharmacol. Toxicol. 2006, 54, 189–199.
(23) Morelli, J. K.; Buehrle, M.; Pognan, F.; Barone, L. R.; Fieles, W.;
Ciaccio, P. J. Validation of an in vitro screen for phospholipidosis
using a high-content biology platform. Cell Biol. Toxicol. 2006,
22, 15–27.
References
(1) Hoyer, D.; Hannon, J. P.; Martin, G. R. Molecular, pharmacolo-
gical and functional diversity of 5-HT receptors. Pharmacol.,
Biochem. Behav. 2002, 71, 533–554.
(2) Hartig, P. R.; Hoyer, D.; Humphrey, P. P.; Martin, G. R. Align-
ment of receptor nomenclature with the human genome: classifica-
tion of 5-HT1B and 5-HT1D receptor subtypes. Trends Pharmacol.
Sci. 1996, 17, 103–105.
(3) Slassi, A. Recent advances in 5-HT1B/1D receptor antagonists and
agonists and their potential therapeutic applications. Curr. Top.
Med. Chem. 2002, 2, 559–574.
(4) Helal, C. J. Potent and Selective 5-HT1B/D Antagonists/Inverse
Agonists with Reduced hERG Affinity for the Treatment of
Depression. Presented at the 235th National Meeting of the
American Chemical Society, New Orleans, LA, Apr 6-10, 2008;
MEDI-154.
(5) Trumpp-Kallmeyer, S.; Hoklack, J.; Bruinvels, A.; Hibert, M.
Three-dimensional models of neurotransmitter G-binding protein
coupled receptors. Mol. Pharmacol. 1991, 40, 8–15.
(6) Hancox, J. C.; McPate, M. J.; Harchi, A. E.; Zhang, Y. The hERG
potassium channel and hERG screening for drug-induced torsades
de pointes. Pharmacol. Ther. 2008, 119, 118–132. Redfern, W. S.;
Carlsson, L.; Davis, A. S.; Lynch, W. G.; MacKenzie, I.; Palethorpe, S.;
Siegl, P. K.; Strang, I.; Sullivan, A. T.; Wallis, R.; Camm, A. J.;
Hammond, T. G. Relationships between preclinical cardiac electrophy-
siology, clinical QT interval prolongation and torsade de pointes for a
broad range of drugs: evidence for a provisional safety margin in drug
development. Cardiovasc. Res. 2003, 58, 32–45.
(24) In vivo pharmacology: Hudzik, T.; Smolka, J; Litwin, L.; Porrey,
T.; Pierson, E. In vivo pharmacology of AZD1134, a novel 5-HT1B
antagonist. Eur. Neuropsychopharmacol. 2003, 13 (Suppl. 4), S181–
S182. Process Chemistry (as M549865): Robinson, G. E.; Cunningham,
O. R.; Dekhane, M.; McManus, J. C.; O'Kearney-McMullan, A.;
Mirajkar, A. M.; Vikas Mishra; Norton, A. K.; Venugopalan, B.;
Williams, E. G. Successful development and scale-up of a palladium-
catalysed amination process in the manufacture of ZM549865. Org.
Process Res. Dev. 2004, 8, 925–930. Development status: http://
integrity.prous.com, entry no. 323892 .
(7) Reasor, M. J.; Hatings, K. L.; Ulrich, R. G. Drug-induced phos-
pholipidosis: issues and future directions. Expert Opin. Drug Saf.
2006, 5, 567–583.