Discovery of HTL6641, a dual orexin receptor antagonist with differentiated pharmacodynamic properties

Article abstract of DOI:10.1039/c5md00027k

A novel series of potent, selective, and orally efficacious dual antagonists of the orexin receptors has been investigated, resulting in the identification of lead compound 27 (HTL6641). Comprehensive data for 27 are presented, including in vivo PK parame

Full text of DOI:10.1039/c5md00027k

MedChemComm
View Article Online
View Journal
CONCISE ARTICLE
Discovery of HTL6641, a dual orexin receptor
antagonist with differentiated pharmacodynamic
Cite this: DOI: 10.1039/c5md00027k
properties†
John A. Christopher, a Sarah J. Aves,a Jason Brown,a James C. Errey,a Suki S. Klair,b
*
Christopher J. Langmead,c Oliver J. Mace,a Richard Mould,a Jayesh C. Patel,a
Benjamin G. Tehan,a Andrei Zhukov,a Fiona H. Marshalla and Miles Congrevea
A novel series of potent, selective, and orally efficacious dual antagonists of the orexin receptors has been
investigated, resulting in the identification of lead compound 27 (HTL6641). Comprehensive data for 27 are
presented, including in vivo PK parameters, confirmation of receptor occupancy through ex vivo binding
and efficacy in a rat sleep model. A key feature of the series is a short dissociation half-life, measured by
surface plasmon resonance (SPR) using stabilized receptors, and confirmed by radioligand-binding experi-
ments. Based on a consideration of the requirements for a potential treatment for insomnia, compound 27
was identified as having the best balance of properties from the chemical series.
Received 20th January 2015,
Accepted 26th March 2015
DOI: 10.1039/c5md00027k
www.rsc.org/medchemcomm
The orexin neuropeptides orexin-A and orexin-B, derived from
a precursor expressed in the hypothalamus, were indepen-
dently revealed by two research groups in 1998.1,2 The orexins
bind to two G protein-coupled receptors (GPCRs), OX1 and
OX2, which are highly conserved across mammalian species,
as are the neuropeptides themselves.1,3 The orexin system is a
key regulator of behavioural arousal, sleep and wakefulness.
Transgenic mice and rats in which the orexin neurons have
been genetically ablated exhibit severe sleepiness, as do
orexin peptide knockout mice. In each case a phenotype simi-
lar to human narcolepsy patients is observed.4–6 In humans,
loss of the orexin neurons with intact receptor expression is
linked to narcolepsy, a chronic sleep disorder characterised
by excessive sleepiness during the day, fragmented sleep and
cataplexy.7 In canines, it has been established that disruption
of the OX2 receptor gene results in narcolepsy, providing a
clear genetic linkage between the orexin system and sleep
modulation.8
receptor antagonists, with both selective and dual profiles, to
investigate the potential for treatment of insomnia, other
sleep disorders and other diseases of the central nervous
system.9–12 Current therapies for insomnia treatment centre
on the use of benzodiazepines and related derivatives which
have sedative, hypnotic and anxiolytic actions resulting from
an enhancement of the effect of the neurotransmitter GABA
at the GABAA receptor. Due to a variety of side effects includ-
ing depression, dependence and sexual dysfunction, and the
potential for addiction, benzodiazepine therapies for insom-
nia are available only on prescription. A related treatment
paradigm uses newer drugs such as zolpidem which are posi-
tive allosteric modulators of GABAA receptors and share a
similar risk-benefit profile.13 The strategy of insomnia treat-
ment by modulation of GABAA receptors is most prominently
hampered by CNS-related side effects the morning after
administration, including drowsiness, cognitive hang-over
effects and a lack of coordination.14,15 Treatment of insomnia
by antagonising orexin receptors presents an opportunity to
more directly address the disorder than the existing GABA
based therapies, as orexin receptor antagonists, which act to
prevent wakefulness, should more selectively regulate the
sleep/wake cycle and lack the general CNS depressant effects
of current treatments. The majority of efforts in recent
years have centred on the identification and progression of
dual orexin receptor antagonists (DORAs), in line with obser-
vations that both receptors play a role in regulation of sleep
and wakefulness.16,17 The precise role of the OX1 receptor
remains controversial, as it has been suggested by one group
that antagonising the OX2 receptor alone should have greater
efficacy in the treatment of insomnia than a DORA,18
Over the past decade, there has been a significant drive
within the pharmaceutical industry to develop orexin
a Heptares Therapeutics Ltd.‡, BioPark, Welwyn Garden City, Hertfordshire, AL7
3AX, U.K.. E-mail: john.christopher@heptares.com
b Labstract Ltd., Stevenage Bioscience Catalyst, Gunnels Wood Road, Stevenage,
Hertfordshire, SG1 2FX, U.K.
c Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal
Parade, Parkville 3052, Victoria, Australia
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5md00027k
‡ ©Heptares Therapeutics 2015. The HEPTARES name is
a trademark of
Heptares Therapeutics Ltd
This journal is © The Royal Society of Chemistry 2015
Med. Chem. Commun.

View Article Online
Concise Article
MedChemComm
however both receptors have been found to play different
roles in sleep promotion.19,20 Selective antagonists of either
receptor were reviewed in 2013,21 and more recently OX2
selective molecules with efficacy in preclinical sleep or anti-
depressant models have been published by Merck and Eli
Lilly respectively.22,23
Several clinical studies have confirmed the therapeutic
utility and safety of DORAs in the treatment of insomnia.
Almorexant 1 (Fig. 1), developed by Actelion and subse-
quently licensed to GlaxoSmithKline, demonstrated efficacy
in a phase II trial, increasing sleep efficiency in primary
insomnia patients.24 Secondary end points, dose-dependent
decreases in latency to persistent sleep (LPS) and wake after
sleep onset (WASO), were also met. Although in 2009
almorexant was reported to have achieved its primary end
point in an initial phase III trial, development of the mole-
cule was discontinued in 2011 after a review of additional
studies conducted to further establish the clinical profile,
including tolerability.25 A second DORA from GSK, SB-649868
2, promoted sleep in male insomnia patients, but clinical
development was subsequently stopped due to a preclinical
safety observation in rats.26,27 Merck's suvorexant 3 is the
most advanced DORA,28 which was approved for use in
adults with insomnia in August 2014. Additionally, Merck
have progressed MK-6096 4 (filorexant) into phase II trials in
patients with primary insomnia,29 and data from phase I tri-
als of a further DORA from Actelion, ACT-462206 5, together
with preclinical data, have recently been published.30,31 The
clinical DORAs largely originate from high-throughput
screening and it has been challenging to optimize properties
such as lipophilicity, for example almorexant has c log P 5.9,32
which may be linked to the failure of several compounds due
to safety issues unrelated to the mechanism of action.
Fig. 1 Clinically evaluated DORAs 1–5.
readily amenable to further analog synthesis. An additional
attraction of 6 was that the hit was largely dissimilar to
known DORAs.9–12 Initially, a number of analogs which var-
ied the benzyl substituent of 6, but kept the 3,4-dimethoxy-
phenyl portion constant, were purchased or synthesized
according to the general route in Fig. 3. Starting from
2-nitrobenzenesulfonyl chloride, sulfonamide coupling,
nitro reduction and cyclisation with triphosgene or 1,1′-
carbonyldiimidazole (CDI), yielded benzothiadiazin-3IJ4H)-one
1,1-dioxide intermediates which could then be readily
N-substituted with benzyl halides or benzyl alcohols. As pri-
mary assays for establishing SAR, we used OX1 and OX2
radioligand binding assays (see ESI†), focusing our attention
primarily on activity at the latter receptor. Data for key early
compounds in the series are summarized in Table 1.
To develop novel antagonists of the orexin receptors we
embarked upon a hit identification campaign using a combi-
nation of fragment, focused and virtual screening, selecting
several relatively small sets of molecules to screen by surface
plasmon resonance (SPR) and/or radioligand binding. A num-
ber of promising hits were identified from each approach
and several were followed up by further purchasing or synthe-
sis of close analogs. One particular screen of approximately
100 compounds contained molecules selected by medicinal
chemists after consideration of how a number of known
orexin antagonists were proposed to bind into a refined
homology model (see ESI†) of the OX2 receptor and then
targeting molecules considered to have similarity in their
pharmacophoric features. The effort uncovered several
related hits exemplified by 6 (Fig. 2) which had encouraging
binding affinity (OX2 pKi 6.6). Compound 6 and several other
classes of molecules were selected to test the hypothesis that
a central heterocyclic ring containing H-bond acceptors,
flanked by two further aromatic substituents which might
fold together into a hydrophobically collapsed conformation,
Removal of the methylenedioxyphenyl unit from 6, which
was deemed undesirable due to a potential bioactivation
risk,34 yielded an improvement in OX2 affinity for the
unsubstituted benzyl analog 7. Fluorine substitution at the
2-position was favourable (8), whereas 3- and 4-fluorination
would have the potential to bind to the orexin receptors.33
A
number of further commercially available analogs of 6 were
available to rapidly build SAR, and synthetic routes were
Fig. 2 Initial OX2 hit 6.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2015

View Article Online
MedChemComm
Concise Article
of alkyl and cycloalkyl substitutions in place of the benzyl
group were explored, as was replacement of the benzyl by
substituted (pyridinyl)methyl groups, but neither strategy
yielded sufficient affinity to be explored further (data not
shown). In general, analogs exhibited a moderate preference
for higher affinity at the OX2 receptor than OX1.
Having gained confidence that high affinity could be
achieved through optimization of the benzyl group, our atten-
tion turned to modification of the 3,4-dimethoxyphenyl unit
and the central heterocyclic scaffold, with the primary aim of
reducing lipophilicity and obtaining better developability
properties within the series. An additional benefit in moving
away from the dimethoxyphenyl unit was that the potential
for reactive metabolite risks associated with this moiety,22
or indeed with the presence of a masked aniline, would be
removed.
Synthesis in an analogous fashion to the route depicted in
Fig. 3 allowed replacement of the dimethoxyphenyl group by
a variety of dimethoxypyridines. In general this was tolerated
(Table 2), with pyridine variants having broadly comparable
affinity (compare 18 and 19 to 12; 20 to 13). As with the
phenyl variants (data not shown), two methoxy substituents
were required for high affinity, for example 21 is more than
30-fold less active at OX2 than the analogous 12, though as
in the phenyl series (data not shown) both 3,4- and
3,5-dimethoxy substitution patterns were well tolerated (18,
23 and 19, 22 respectively). Installation of the pyridine nitro-
gen at the 2-position was also well tolerated (24). Benzyl posi-
tion SAR was in general highly transferable, and by selection
of the most active substituents, OX2 binding could be readily
tuned to high levels, for example 22, 23 and 24 have affinities
in excess of pKi 9 (Ki < 1 nM).
Fig. 3 General synthetic route for synthesis of analogs of 6. Reagents
and conditions: (a) 3,4-dimethoxyaniline, 1,4-dioxane, 80 °C; (b) SnCl2,
EtOH, 100 °C or Pd/C, H2, EtOH/H2O, rt or Fe, AcOH, 60 °C; (c)
triphosgene, 1,4-dioxane, 100 °C or CDI, Et3N, DMF, 100 °C; (d)
substituted benzyl alcohol, Ph3P, diisopropyl azodicarboxylate (DIAD)
or 1,1′-azobisIJN,N-dimethylformamide) (TMAD), THF, rt or substituted
benzyl halide, K2CO3, DMF, 80 °C.
did not increase binding affinity significantly (data not
shown). Substitution with chloro at the 2-position was also
beneficial. 4-Substitution was envisaged to be advantageous
for metabolic stability, and this was tolerated in combination
with a 2-substituent (9, 10). 2,6-Disubstitution (11) yielded a
significant increase in affinity, which was largely maintained
in the 2,4,6-trisubstituted analogs 12 and 13, though not in
2-methoxy derivative 14. Useful SAR emerged at the 4-posi-
tion, methyl substitution (15) was tolerated in similar fashion
to fluoro but installation of methoxy yielded further signifi-
cant increases in OX2 affinity (16 vs. 12; 17 vs. 13). A range
In an exploration of close analogs of the highest affinity
compounds, we observed that one or both methoxy groups
could be replaced, for example by methyl (data not shown) or
methylamino (25). In general, when compared directly to the
methoxy analogs (data not shown), up to a 10-fold loss of
OX2 affinity was observed, and the majority of these changes
also resulted in a significant loss of in vitro metabolic stabil-
ity. Changes to the thiadiazin-3IJ4H)-one 1,1-dioxide portion
of the central scaffold were poorly tolerated, for example 28,
Table 2, whereas modification of the fused phenyl portion
was more encouraging. Most notably, installation of a nitro-
gen atom at the 5-position allowed the lipophilicity of the
series to be decreased whilst maintaining high levels of OX1
and OX2 binding affinity (25–27, Table 2). 5-Aza analogs of
this type could be readily accessed from 2-chloropyridine-3-
sulfonyl chloride using the synthetic route depicted in Fig. 4,
with sulfonamide coupling preceding SNAr reaction with a
substituted benzylamine, followed by ring closure with
triphosgene or CDI.
Table 1 OX1 and OX2 SAR of 7–17 IJR1–R5 = H unless specified)
R1–R5
OX1 pKia
OX2 pKia
7
8
9
—
5.7
6.5
7.2
<5.6
8.1
7.3
7.6
7.2
8.2
8.7
9.4
7.1
8.2
8.0
7.6
9.1
8.7
9.0
8.2
9.1
9.3
10.0
R1 = F
R1 = F, R3 = Cl
10
11
12
13
14
15
16
17
R1 = Cl, R3 = F
R1 = Cl, R5 = F
R1, R3, R5 = F
R1 = Cl, R3, R5 = F
R1 = OMe, R3, R5 = F
R1 = Cl, R3 = Me, R5 = F
R1 = F, R3 = OMe, R5 = F
R1 = Cl, R3 = OMe, R5 = F
An important consideration that we were aware of during
the course of the project is that the duration of action in vivo
of orexin antagonists will be dependent upon both the phar-
macokinetic and receptor–ligand kinetic parameters of the
molecule. Pharmacokinetic parameters are an important
a
Data represent geometric means of two or more measurements.
This journal is © The Royal Society of Chemistry 2015
Med. Chem. Commun.

View Article Online
Concise Article
MedChemComm
Table 2 OX1 and OX2 SAR and rat microsomal stability of 18–28
R1
R3
W
X
Y
Z
OX1 pKia
OX2 pKia
RLM t1/2 (min)
c log Pb
18
19
20
21
22
23
24
25
26
27
28
F
F
Cl
F
F
F
F
F
F
F
F
F
F
OMe
OMe
OMe
OMe
F
CH
CH
CH
CH
CH
CH
N
CH
N
CH
—
N
COMe
N
COMe
COMe
N
COMe
N
COMe
N
COMe
N
COMe
COMe
COMe
CH
COMe
COMe
CH
CNHMe
CH
COMe
—
7.9
7.8
7.4
6.4
9.0
8.6
nd
8.9
7.6
7.7
<5.2
8.3
8.9
8.8
7.2
9.6
9.4
9.7
9.2
8.8
8.6
<5.0
28
16
23
23
16
9
8
7
16
23
nd
2.9
3.1
2.5
2.7
3.0
2.8
2.8
0.3
1.4
1.4
4.0
N
COMe
COMe
COMe
COMe
COMe
—
COMe
N
—
F
—
F
—
a
Data represent geometric means of two or more measurements, nd = not determined. b Ref. 32.
consideration during medicinal chemistry optimization, how-
ever whilst these measurements are routine it is not always
the case that pharmacodynamic effects are measured at the
level of the receptor.35,36 At Heptares we use biochemical
and/or biophysical approaches to routinely probe structure–
kinetic relationships (SKR) to enable us to select molecules
with appropriate attributes for progression. A notable feature
of the advanced DORAs suvorexant and almorexant is that
they exhibit slow kinetics at the orexin receptors.37,38 In the
treatment of sleep disorders, a potential issue for compounds
with slow receptor kinetics is that they may exhibit a
prolonged duration of action which could potentially pro-
mote next-day residual effects. Therefore, as part of our
progression strategy we sought to identify molecules with
appropriate parameters, both in terms of moderate pharma-
cokinetic half-lives and a fast receptor off-rate.
characteristics of the wild-type receptors.40 The enhanced sta-
bility of the purified proteins allowed capture on SPR sensor
chips (see ESI†) and examination of the kinetics of receptor–
ligand binding interactions of standard molecules and multi-
ple compounds in our series. Dilution series of each com-
pound were injected, and blank-subtracted data were fitted to
a 1 : 1 interaction model to obtain kinetic and affinity con-
stants. Our series, exemplified by 18 and 27, displayed fast
OX2 receptor kinetics, with 18 (see ESI†) having an on rate
(ka) of 2.8 × 104 M−1 s−1, an off rate (kd) of 2.7 × 10−3 s−1 and a
dissociation half-life (t1/2 = (ln 2/kd)/60) of 4.3 min. The close
To address these twin challenges, we first examined exam-
ples of our series in biophysical and biochemical assays to
determine receptor binding kinetics. Wild-type GPCRs are
rarely stable enough to be successfully captured on to biosen-
sor chips for evaluation by SPR,39 a limitation which can be
successfully overcome by increasing thermostability through
protein engineering.40,41 Introduction of a small number of
mutations, which do not affect the binding site characteris-
tics of the receptor but dramatically increase thermostability,
yield modified GPCRs known as StaR® proteins. The stabi-
lized receptors can be successfully immobilized on biosensor
chips and used for fragment screening,42 Biophysical Map-
ping™,43 and for evaluation of receptor–ligand kinetic
parameters,44 as we have previously described. OX1 and OX2
StaR proteins were generated which had significantly
increased thermostability and retained the ligand binding
Fig. 4 Synthesis of 25–27. Reagents and conditions: (a) heteroaryl
amine, pyridine, DCM, 0 °C or heteroaryl amine, 1,4-dioxane, 0 °C; (b)
substituted benzylamine, MeCN, microwave heating, 180 °C, 1–2 h; (c)
triphosgene, 1,4-dioxane, 100 °C or CDI, Et3N, DMF, 100 °C.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2015

View Article Online
MedChemComm
Concise Article
analog 27 (Fig. 5) was profiled at both the OX1 (ka 2.0 × 104
M−1 s−1, kd 1.6 × 10−2 s−1, t1/2 = 0.8 min) and OX2 receptors (ka
3.3 × 104 M−1 s−1, kd 2.8 × 10−3 s−1, t1/2 = 4.2 min). Suvorexant
by comparison displayed much slower receptor kinetics with
dissociation half-lives of 24 and 210 min for the OX1 and OX2
receptors respectively. Evaluation of OX2 receptor kinetics
using Motulsky–Mahan radioligand-binding studies was also
undertaken,38,45 and yielded comparable data: 18 t1/2 = 3.0
min; 27 t1/2 = 7.6 min; suvorexant t1/2 = 79 min; the latter fig-
ure is additionally in line with that reported in the literature
by workers from Merck.37 In OX2 radioligand-binding studies
almorexant has a dissociation half-life of 242 min.38 The sig-
nificantly faster receptor kinetics observed with members of
our series including 18 and 27 should facilitate rapid re-
equilibration with changing orexin levels in vivo, to help
reduce the potential risk for next day somnolence effects.37
During the medicinal chemistry progression of the series
described herein, in addition to the studies presented in this
manuscript, multiple co-crystal structures of compounds
from the benzothiadiazine series, another novel series from
our laboratories, and literature compounds such as
suvorexant and SB-334867,46 were solved in complex with
OX1 and/or OX2 StaR proteins. These crystal structures
revealed, in fine detail, the binding modes of the various
series and how they relate to one another, as well as giving
some insight into the binding kinetics of the receptors. In
particular the X-ray structures provided a detailed under-
standing of the differences between the two receptors and
have been used extensively to inform further iterations of our
medicinal chemistry project leading to the design of uniquely
selective OX1 antagonists which have potential in the treat-
ment of addiction disorders.47 These data are beyond the
scope of this communication and will be the subject of future
publications in specialized journals. During the preparation
of this manuscript, the crystal structure of suvorexant bound
to an OX2 fusion protein construct was disclosed.48
In parallel with profiling key compounds through stan-
dard in vitro assays such as the hERG ion channel and P450
inhibition we routinely monitored in vitro stability in rat-liver
microsomal preparations (RLM t1/2, Table 2). For the reasons
detailed above, we needed to strike a delicate balance
between having sufficient metabolic stability to not compro-
mise oral bioavailability through high first-pass metabolism,
yet having a relatively short pharmacokinetic half-life in rat
as an initial pre-clinical species. Profiling of early molecules
in PK experiments in male Sprague Dawley rats by both iv
and po routes allowed us to identify a half-life range in the
RLM assay of 15–35 minutes as an appropriate criteria for
progression; data for key advanced compounds are summa-
rized in Table 3. Compound 18 and the regioisomeric 19 have
low in vivo clearance (18 and 17 mL min−1 kg−1 respectively)
with 18 having a shorter half-life by virtue of a lower volume
of distribution than 19. In line with its higher measured
plasma protein binding (PPB) in rat, an additional in vitro
parameter which we closely monitored during series progres-
sion, 19 additionally compares poorly to its regioisomer in
terms of the unbound fraction in the brain (approximated as
the ratio of drug concentration in the cerebrospinal fluid
(CSF) to that in brain at the same time point);49 in common
with the majority of compounds profiled, both molecules
had very good brain penetration. Chloro derivative 20, with
both higher clearance and volume of distribution than 18,
had a similar half-life and a higher free fraction in brain. All
three molecules, together with the aza-examples 26 and 27,
had acceptable oral bioavailability (29–66%). Selected com-
pounds were additionally progressed to PK studies in male
beagles; 18 had low clearance, moderate volume of distribu-
tion, long half-life and excellent bioavailability. Compound
27, the aza-variant of 18, also had a good profile in beagle
with moderate clearance and volume of distribution, good
bioavailability and a shorter half-life than 18. The mean resi-
dence time (MRT) of 27 was 1.9 h and 4.0 h in rat and beagle,
respectively. Overall, these PK parameters suggest a short to
moderate half-life in human.
After consideration of the overall profiles of 27 and related
close analogs, 27 was selected for extensive further profiling
as summarized below and in Table 4. The dual-antagonist
profile of 27 was confirmed in OX1 and OX2 antagonist func-
tional cell assays measuring receptor-stimulated ERK1/2
phosphorylation. A clean in vitro profile was apparent, with
no issues identified with inhibition of P450 isoforms or the
hERG ion channel. Evaluation of bidirectional apparent per-
meability (Papp) across a Caco-2 cell monolayer indicated that
the molecule had high passive permeability with low efflux.
Binding to human, dog and mouse plasma proteins was
high, in common with the measured value in rat (99.5%),
and the latter is consistent with the CSF concentrations
observed in rat PK experiments. Evaluation of suvorexant
as a benchmark in these assays returned comparable data
(99.4, 98.3, 99.3% for human, dog and mouse respectively).
Fig. 5 (a) OX1 and (b) OX2 surface plasmon resonance sensorgrams of 27.
This journal is © The Royal Society of Chemistry 2015
Med. Chem. Commun.

View Article Online
Concise Article
MedChemComm
Table 3 RLM and PK parameters of 18–20, 26 and 27 in rat and beagle
iv
po
RLM t1/2
(min)
Cl
Vss
Brain :
plasma (0.5 h)
CSF : brain
(0.5 h)
t1/2
(po) (h)
AUCinf
Tmax
(h)
Fpo
(%)
Species
(mL min−1 kg−1
)
(L kg−1
)
(ng h−1 mL−1
)
18a
28
Rat
Beagle
Rat
Rat
Rat
18
5
2.2
3.6
3.9
3.5
2.3
2.1
3.0
0.77
—
0.01
—
1.6
10.6
3.3
1.8
1.3
1.8
7.1
558
6957
705
300
401
1.5
0.4
0.4
0.8
0.5
0.5
0.4
29
100
36
29
34
66
70
19a
20a
26a
27b
15
23
16
23
17
34
29
18
13
1.51
0.79
0.84
0.49
—
BQLc
0.02
0.01
0.01
—
Rat
Beagle
1217
2442
a
b
Dosed at 1 & 2 mg kg−1 iv/po respectively in rat (2 & 2 mg kg−1 in beagle), using 10% DMA, 10% solutol HS15, 80% saline as vehicle. Dosed
at 1 & 2 mg kg−1 iv/po respectively in the rat PK studies (1 & 1 mg kg−1 in beagle), using 10% DMA, 10% solutol HS15, 80% (10% aqueous
IJ2-hydroxypropyl)-β-cyclodextrin) as vehicle. c BQL = below quantifiable limit of 1.00 ng mL−1
.
Screening in an in vitro cytotoxicity assay in HepG2 cells,
determining cell viability by ATP measurements, revealed no
issues, and the molecule was inactive in a 5-strain AMES test
with and without S9 fractions. Stability in human,
cynomolgus monkey, dog and rat hepatocytes was high, and
subsequently the major metabolites arising from incubation
with hepatocytes from these four species were characterized.
In all species mono-demethylation of 27 was the major
metabolite, with subsequent glucuronide conjugation also
observed. Low levels of mono-oxidation, with and without
sulfation or glucuronidation, were also noted. The two regio-
isomeric de-methylated metabolites of 27 were synthesized
and profiled in binding assays, and in each case a significant
drop in affinity (approximately 80-fold) compared to the par-
ent was observed, suggesting a low risk of active metabolite
formation in vivo. No metabolites that were unique to human
were observed, and there was no evidence of addition of glu-
tathione to the parent molecule or metabolites, which would
be indicative of a risk of formation of a reactive metabolite.
Cross-screening of 27 in a small number of in-house GPCR
radioligand binding assays (M1, M2, M3, M4, CGRP, GLP1,
mGlu5) provided initial confidence that the molecule had
good selectivity for the OX receptors. Wider profiling in an
external panel of 14 GPCR, kinase, ion-channel and nuclear
receptor targets indicated that 27 had at least 1000 fold selec-
tivity against the panel members. Compound 18, a close ana-
log of 27, had been earlier screened against an extended
panel of 68 targets and also had excellent selectivity, provid-
ing overall confidence in the selectivity of the chemotype.
With these data in hand, 27 was advanced to a rat ex vivo
autoradiography experiment (RenaSci Ltd, Nottingham, UK)
to evaluate OX2 receptor occupancy in the CNS from an oral
dose. Male Sprague Dawley rats were dosed orally with vehi-
cle alone (10% DMA, 10% solutol HS15, 80% (10% aqueous
IJ2-hydroxypropyl)-β-cyclodextrin)) or 27 (1 mg kg−1, po) and
sacrificed 1 h post-dose. Frontal cortex sections were pre-
pared and incubated with [3H]EMPA,50 with levels of bound
radioactivity in the sections determined using a beta imager.
Receptor occupancy was determined from mean specific
binding with the vehicle treated control taken as 100%. In
this study 27 demonstrated a mean receptor occupancy of
57% from an oral dose of 1 mg kg−1 (Fig. 6); by comparison,
suvorexant from a 1 mg kg−1 oral dose achieved 31% receptor
occupancy. Plasma exposures at this time point were 72 ng
mL−1 and 22 ng mL−1 for 27 and suvorexant respectively. The
study confirmed that despite the fast receptor kinetics of 27,
high occupancy of the OX2 receptors could be achieved from
an oral dose.
The effect of 27 treatment on rat sleep architecture in vivo
was investigated using a telemetered CT18 sleep study (Aptuit
(Verona) Srl., Italy), a model which has previously been used
to support orexin lead optimisation programmes prior to pos-
itive clinical proof of concept in insomnia patients.51 Adult
male CD rats were implanted with telemetric probes to
record electroencephalogram (EEG) and neck electromyo-
gram (EMG) readings. The study employed a balanced cross-
over design in which all animals were alternatively treated
with drug and vehicle. Rats were dosed with vehicle (80%
Table 4 Profile of 27
OX1 pKi 7.7a
Permeabilityc 52 × 10−6 cm s−1 (efflux ratio 0.7)
PPB%d 99.3, 98.1, 99.2
OX2 pKi 8.6a
OX1 pKb 6.6a
OX2 pKb 7.3a
OX1 kinetics: t1/2 0.8 min (SPR)
OX2 kinetics: t1/2 4.2 min (SPR), 7.6 min (biochemical)
P450 inhibition pIC50 < 5b
hERG pIC50 < 5
Hepatocyte t1/2: >110 min (rat, dog), >250 min (cyno, human)
No reactive metabolites
Ames negativee
a
b
c
d
See ESI for assay details. 1A2, 2C8, 2C9, 2C19, 2D6 and 3A4 isoforms. Caco-2 monolayer, Papp A–B. Binding to human, canine and
murine plasma proteins respectively. e 5 strains, with and without S9 fractions.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2015

View Article Online
MedChemComm
Concise Article
respectively, over 3 hours. The time spent asleep was also
increased by both doses of 27, which promoted sleep primar-
ily by increasing REM, with smaller effects on NREM. Over 3
hours, the time spent in REM sleep was increased by 70.6
and 69.6% (p < 0.05; Fig. 7F) by 3 and 10 mg kg−1 27. The
time spent in NREM sleep was significantly increased by 28.8
and 30.7% (p < 0.01; Fig. 7E) at 3 and 10 mg kg−1 27, respec-
tively. Consistent with a smaller effect on NREM, there was
no significant effect on the latency to the first NREM episode
following treatment with 27 at the doses tested (Fig. 7G).
However, the latency to the first REM episode was reduced by
treatment with 10 mg kg−1 27, reducing the time taken by
35.0% from 70.9 7.7 to 46.1 6.6 min (p < 0.05; Fig. 7H)
compared to vehicle. In a similar rat sleep model, suvorexant
demonstrated significant changes in sleep architecture from
a 30 mg kg−1 dose with greater effects on REM than NREM,
providing overall confidence that 27 demonstrates encourag-
ing efficacy in vivo.52
Fig. 6 [3H]EMPA ex vivo autoradiography in rat frontal cortex 1 hour
following oral administration of 27 and suvorexant at 1 mg kg−1
.
PEG-400 and 20% Cremophor® EL) or with 27 (3 and 10 mg
kg−1) po, administered at circadian time 18h (CT18, six hours
after lights off). The effect of 27 on sleep patterns was evalu-
ated starting the recording 6 hours in to the dark phase. EEG
and neck EMG readings were recorded and analysed for 5
hours after treatment to determine time spent awake, NREM
and REM sleep (Fig. 7A–C). Sleep and wake stages in rats
were altered by 27 compared to vehicle, with the hypnotic
effect of 27 starting to decline after 3 hours. Peak effects on
sleep parameters occurred at approximately 2 h, in contrast
to a Tmax of approximately 0.5 h in the rat oral PK study, and
a clear dose response in several of the sleep parameters
examined was not apparent. Given the limitations of this ini-
tial study further in vivo efficacy experiments are warranted
to better understand the properties of 27. During episodes of
insomnia, it is important to decrease the time spent awake
during the night, a clinical end point measured during a
sleep trial, as well as increasing the duration of sleep. At
3 and 10 mg kg−1, 27 promoted sleep and reduced the
time spent awake by 17.4 and 18.5% (p < 0.05; Fig. 7D),
Conclusions
In summary, the series described herein, exemplified by com-
pounds such as 18 and 27, represents a novel, high affinity
dual orexin receptor antagonist chemotype, with a good over-
all in vitro and in vivo developability profile. Based on a con-
sideration of the requirements for a potential treatment for
insomnia, compound 27 (HTL6641), was identified as having
the best balance of properties in the chemical series. Key fac-
tors which warrant the further investigation of 27 are its
appropriate pre-clinical PK profile, high CNS receptor occu-
pancy, fast off-rate kinetics and relevant in vivo efficacy in a
rat sleep study.
Experimental details
All studies involving live animals were performed in compli-
ance with national laws and institutional guidelines, follow-
ing approval by the relevant committees of the study centres.
Synthetic methods
The purity of the final compounds was determined by LC/MS
analysis to be ≥95%. Supplier information for purchased
compounds, LCMS QC data for all compounds, experimental
details and 1H NMR data for all synthesized compounds in
Tables 1 and 2 are described in the ESI.† Suvorexant was syn-
thesized in accordance with literature conditions.53 J values
are given in Hz. Synthesis of 27 is described below.
2-Chloro-N-IJ5,6-dimethoxypyridin-3-yl)pyridine-3-sulfonamide.
A mixture of 2-chloropyridine-3-sulfonyl chloride (530 mg,
2.50 mmol), 5,6-dimethoxypyridin-3-amine (424 mg, 2.75 mmol)
and pyridine (0.60 mL, 7.42 mmol) in DCM (10 mL) was
stirred at 0 °C for 3 h before concentration in vacuo. Puri-
fication by gradient column chromatography, eluting with
12–60% EtOAc in iso-hexane yielded the title compound
(542 mg, 1.64 mmol, 66%) as a pale orange solid. m/z
328.2, 330.2 (M − H)−; δH (400 MHz; d6-DMSO) 3.70 (3 H, s),
Fig. 7 Telemetered rat circadian time (CT) 18 sleep study of 27.
This journal is © The Royal Society of Chemistry 2015
Med. Chem. Commun.

View Article Online
Concise Article
MedChemComm
3.77 (3 H, s), 7.02 (1 H, d, J 2.3), 7.39 (1 H, d, J 2.3), 7.61
(1 H, dd, J 7.9, 4.8), 8.38 (1 H, dd, J 7.9, 1.8), 8.64 (1 H, dd,
J 4.8, 1.8), 10.71 (1 H, s).
2 L. de Lecea, T. S. Kilduff, C. Peyron, X.-B. Gao, P. E. Foye,
P. E. Danielson, C. Fukuhara, E. L. F. Battenberg, V. T.
Gautvik, F. S. Bartlett II, W. N. Frankel, A. N. van Den Pol,
F. E. Bloom, K. M. Gautvik and J. G. Sutcliffe, Proc. Natl.
Acad. Sci. U. S. A., 1998, 95, 322.
3 J. Gatfield, C. Brisbare-Roch, F. Jenck and C. Boss,
ChemMedChem, 2010, 5, 1197.
4 R. M. Chemelli, J. T. Willie, C. M. Sinton, J. K. Elmquist, T.
Scammell, C. Lee, J. A. Richardson, S. C. Williams, Y. Xiong,
Y. Kisanuki, T. E. Fitch, M. Nakazato, R. E. Hammer, C. B.
Saper and M. Yanagisawa, Cell, 1999, 98, 437.
5 C. T. Beuckmann, C. M. Sinton, S. C. Williams, J. A.
Richardson, R. E. Hammer, T. Sakurai and M. Yanagisawa,
J. Neurosci., 2004, 24, 4469.
6 J. Hara, C. T. Beuckmann, T. Nambu, J. T. Willie, R. M.
Chemelli, C. M. Sinton, F. Sugiyama, K.-i. Yagami, K. Goto,
M. Yanagisawa and T. Sakurai, Neuron, 2001, 30, 345.
7 N. Tsujino and T. Sakurai, Pharmacol. Rev., 2009, 61, 162.
8 L. Lin, J. Faraco, R. Li, H. Kadotani, W. Rogers, X. Lin, X.
Qiu, P. J. de Jong, S. Nishino and E. Mignot, Cell, 1999, 98,
365.
N-IJ5,6-Dimethoxypyridin-3-yl)-2-ij(2,4,6-trifluorobenzyl)-
amino]pyridine-3-sulfonamide. A mixture of 2-chloro-N-IJ5,6-
dimethoxypyridin-3-yl)pyridine-3-sulfonamide (165 mg, 0.50
mmol) and 2,4,6-trifluorobenzylamine (0.18 mL, 1.47 mmol)
in MeCN (3 mL) was heated in a microwave reactor at 180 °C
for 1 h. After concentration in vacuo purification by gradient
column chromatography, eluting with 7–60% EtOAc in iso-
hexane yielded the title compound (218 mg, 0.48 mmol, 96%)
as a yellow oil. m/z 453.2 (M − H)−; δH (400 MHz; d6-DMSO)
3.62 (3 H, s), 3.73 (3 H, s), 4.60 (2 H, d, J 5.5), 6.49–6.72 (2 H,
m), 6.83 (1 H, d, J 2.3), 7.14 (2 H, t, J 8.7), 7.21 (1 H, d, J 2.3),
7.78 (1 H, dd, J 7.8, 1.8), 8.22 (1 H, dd, J 4.8, 1.6), 10.27 (1 H, s).
2-IJ5,6-Dimethoxypyridin-3-yl)-4-IJ2,4,6-trifluorobenzyl)-2H-
pyridoij2,3-e]ij1,2,4]thiadiazin-3IJ4H)-one 1,1-dioxide (27,
HTL6641).
A
mixture of N-IJ5,6-dimethoxypyridin-3-yl)-
(218
2-ij(2,4,6-trifluorobenzyl)amino]pyridine-3-sulfonamide
mg, 0.48 mmol), 1,1′-carbonyldiimidazole (311 mg, 1.92
mmol) and triethylamine (0.13 mL, 0.93 mmol) in DMF (1.5
mL) in a sealed tube was heated at 100 °C for 2 h. After con-
centration in vacuo purification by gradient column chroma-
tography, eluting with 10–80% EtOAc in iso-hexane yielded
the title compound (159 mg, 0.33 mmol, 69%) as a white
solid. m/z 481.1 (M + H)+; δH (400 MHz; d6-DMSO) 3.77 (3 H,
s), 3.93 (3 H, s), 5.59 (2 H, s), 7.19 (2 H, t, J 8.8), 7.35 (1 H, d,
J 2.0), 7.53 (1 H, dd, J 7.8, 4.8), 7.75 (1 H, d, J 2.0), 8.51 (1 H,
dd, J 7.8, 1.8), 8.85 (1 H, dd, J 4.9, 1.6); δF (376 MHz;
d6-DMSO, C6H5CF3) −111.0 (1 F), −113.6 (2 F); δC (101 MHz;
9 P. J. Coleman and J. J. Renger, Expert Opin. Ther. Pat.,
2010, 20, 307.
10 J. A. Christopher, Pharm. Pat. Anal., 2012, 1, 329.
11 C. Boss, Expert Opin. Ther. Pat., 2014, 24, 1367.
12 J. A. Christopher, Pharm. Pat. Anal., 2014, 3, 625.
13 J. Wagner, M. L. Wagner and W. A. Hening, Ann.
Pharmacother., 1998, 32, 680.
14 R. Hoque and A. L. Chesson, J. Clin. Sleep Med., 2009, 5, 471.
15 T. Roehrs and T. Roth, Neurotherapeutics, 2012, 9, 728.
16 J. T. Willie, R. M. Chemelli, C. M. Sinton, S. Tokita, S. C.
Williams, Y. Y. Kisanuki, J. N. Marcus, C. Lee, J. K. Elmquist,
K. A. Kohlmeier, C. S. Leonard, J. A. Richardson, R. E.
Hammer and M. Yanagisawa, Neuron, 2003, 38, 715.
17 S. R. Morairty, F. G. Revel, P. Malherbe, J.-L. Moreau, D.
Valladao, D, J. G. Wettstein, T. S. Kilduff and E. Borroni,
PLoS One, 2012, 7, e39131.
2
CD3CN) 36.8 (1 C), 54.5 (1 C), 56.7 (1 C), 101.3 (2 C, ddd, JCF
2
4
2
4
26.1, JCF 26.1, JCF 2.3), 110.4 (1 C, td, JCF 18.4, JCF 4.6),
119.9, 120.8, 122.1, 122.9, 133.7 (1 C), 138.8 (1 C), 145.6 (1 C),
147.9 (1 C), 151.0 (1 C), 154.4 (1 C), 156.3 (1 C), 162.9 (2 C,
1
3
3
1
ddd, JCF 249.6, JCF 15.3, JCF 11.1), 163.3 (1 C, dt, JCF 247.7,
3JCF 16.1).
18 C. Dugovic, J. E. Shelton, L. E. Aluisio, I. C. Fraser, X. Jiang,
S. W. Sutton, P. Bonaventure, S. Yun, X. Li, B. Lord, C. A.
Dvorak, N. I. Carruthers and T. W. Lovenberg, J. Pharmacol.
Exp. Ther., 2009, 330, 141.
19 C. Dugovic, J. E. Shelton, S. Yun, P. Bonaventure, B. T.
Shireman and T. W. Lovenberg, Front. Neurosci., 2014, 8, 28.
20 K. Etori, Y. C. Saito, N. Tsujino and T. Sakurai, Front.
Neurosci., 2014, 8, 8.
Acknowledgements
Giles A. Brown is acknowledged for assistance with
radioligand and suvorexant synthesis. Members of the
Heptares pharmacology group are acknowledged for the gen-
eration of selectivity data for HTL6641. Liz Jagger and col-
leagues at RenaSci Ltd are acknowledged for generation of
ex vivo autoradiography data, and Laura Piccoli and col-
leagues at Aptuit (Verona) Srl. are acknowledged for genera-
tion of data from the CT18 sleep study.
21 T. P. Lebold, P. Bonaventure and B. T. Shireman, Bioorg.
Med. Chem. Lett., 2013, 23, 4761.
22 A. J. Roecker, S. P. Mercer, J. D. Schreier, C. D. Cox, M. E.
Fraley, J. T. Steen, W. Lemaire, J. G. Bruno, C. M. Harrell,
S. L. Garson, A. L. Gotter, S. V. Fox, J. Stevens, P. L.
Tannenbaum, T. Prueksaritanont, T. D. Cabalu, D. Cui, J.
Stellabott, G. D. Hartman, S. D. Young, C. J. Winrow, J. J.
Renger and P. J. Coleman, ChemMedChem, 2014, 9, 311.
23 T. E. Fitch, M. J. Benvenga, C. D. Jesudason, C. Zink, A. B.
Vandergriff, M. M. Menezes, D. A. Schober and L. M. Rorick-
Kehn, Front. Neurosci., 2014, 8, 5.
Notes and references
1 T. Sakurai, A. Amemiya, M. Ishii, I. Matsuzaki, R. M.
Chemelli, H. Tanaka, S. C. Williams, J. A. Richardson, G. P.
Kozlowski, S. Wilson, J. R. S. Arch, R. E. Buckingham, A. C.
Haynes, S. A. Carr, R. S. Annan, D. E. McNulty, W.-S. Liu,
J. A. Terrett, N. A. Elshourbagy, D. J. Bergsma and M.
Yanagisawa, Cell, 1998, 92, 573.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2015

View Article Online
MedChemComm
Concise Article
24 P. Hoever, G. Dorffner, H. Beneš, T. Penzel, H. Danker-
Hopfe, M. J. Barbanoj, G. Pillar, B. Saletu, O. Polo, D. Kunz,
J. Zeitlhofer, S. Berg, M. Partinen, C. L. Bassetti, B. Högl,
I. O. Ebrahim, E. Holsboer-Trachsler, H. Bengtsson, Y.
Peker, U.-M. Hemmeter, E. Chiossi, G. Hajak and J.
Dingemanse, Clin. Pharmacol. Ther., 2012, 91, 975.
25 GlaxoSmithKline press release. http://www.gsk.com/en-gb/
media/press-releases/2011/gsk-and-actelion-discontinue-
clinical-development-of-almorexant (accessed January 15,
2015).
L. Yao, S. D. Kuduk, J. Uslaner, P. J. Coleman and J. J.
Renger, BMC Neurosci., 2013, 14, 90.
38 R. Mould, J. Brown, F. H. Marshall and C. J. Langmead, Br.
J. Pharmacol., 2014, 171, 351.
39 S. G. Patching, Biochim. Biophys. Acta, Gen. Subj., 2014, 1838, 43.
40 N. Robertson, A. Jazayeri, J. Errey, A. Baig, E. Hurrell, A.
Zhukov, C. J. Langmead, M. Weir and F. H. Marshall,
Neuropharmacology, 2011, 60, 36.
41 M. Congreve, R. L. Rich, D. G. Myszka, F. Figaroa, G. Siegal
and F. H. Marshall, in Methods in Enzymology, ed. L. C. Kuo,
Academic Press, Burlington, MA, U.S., 2011, ch. 5, vol. 493,
pp. 115–136.
26 P. B. Bettica, L. Squassante, S. Zamuner, G. Nucci, H.
Danker-Hopfe and E. Ratti, Sleep, 2012, 35, 1097.
27 ClinicalTrials
record.
http://clinicaltrials.gov/ct2/show/
42 J. A. Christopher, J. Brown, J. A. S. Doré, J. C. Errey, M.
Koglin, F. H. Marshall, D. G. Myszka, R. L. Rich, C. G. Tate,
B. Tehan, T. Warne and M. Congreve, J. Med. Chem.,
2013, 56, 3446.
43 A. Zhukov, S. P. Andrews, J. C. Errey, N. Robertson, B.
Tehan, J. S. Mason, F. H. Marshall, M. Weir and M.
Congreve, J. Med. Chem., 2011, 54, 4312.
44 M. Congreve, S. P. Andrews, A. S. Doré, K. Hollenstein, E.
Hurrell, C. J. Langmead, J. S. Mason, I. W. Ng, B. Tehan, A.
Zhukov, M. Weir and F. H. Marshall, J. Med. Chem.,
2012, 55, 1898.
45 H. J. Motulsky and L. C. Mahan, Mol. Pharmacol., 1984, 25, 1.
46 R. A. Porter, W. N. Chan, S. Coulton, A. Johns, M. S. Hadley,
K. Widdowson, J. C. Jerman, S. J. Brough, M. Coldwell, D.
Smart, F. Jewitt, P. Jeffrey and N. Austin, Bioorg. Med. Chem.
Lett., 2001, 11, 1907.
NCT00534872?term=SB-649868&rank=1 (accessed January 15,
2015).
28 W. J. Herring, K. M. Connor, N. Ivgy-May, E. Snyder, K. Liu,
D. B. Snavely, A. D. Krystal, J. K. Walsh, R. M. Benca, R.
Rosenberg, R. B. Sangal, K. Budd, J. Hutzelmann, H.
Leibensperger, S. Froman, C. Lines, T. Roth and D.
Michelson, Biol. Psychiatry., 2014, DOI: 10.1016/j.
biopsych.2014.10.003.
29 P. J. Coleman, J. D. Schreier, C. D. Cox, M. J. Breslin, D. B.
Whitman, M. J. Bogusky, G. B. McGaughey, R. A. Bednar, W.
Lemaire, S. M. Doran, S. V. Fox, S. L. Garson, A. L. Gotter,
C. M. Harrell, D. R. Reiss, T. D. Cabalu, D. Cui, T.
Prueksaritanont, J. Stevens, P. L. Tannenbaum, R. G. Ball, J.
Stellabott, S. D. Young, G. D. Hartman, C. J. Winrow and J. J.
Renger, ChemMedChem, 2012, 7, 415.
30 M. Hoch, H. van Gorsel, J. van Gerven and J. Dingemanse,
J. Clin. Pharmacol., 2014, 54, 979.
31 C. Boss, C. Roch-Brisbare, M. A. Steiner, A. Treiber, H.
Dietrich, F. Jenck, M. von Raumer, T. Sifferlen, C. Brotschi,
B. Heidmann, J. T. Williams, H. Aissaoui, R. Siegrist and
J. Gatfield, ChemMedChem, 2014, 9, 2486.
47 C. Baimel, S. E. Bartlett, L.-C. Chiou, A. J. Lawrence, J. W.
Muschamp, O. Patkar, L.-W. Tung and S. L. Borgland, Br. J.
Pharmacol., 2015, 172, 334.
48 J. Yin, J. C. Mobarec, P. Kolb and D. M. Rosenbaum, Nature,
2015, 519, 247.
49 Z. Rankovic, J. Med. Chem., 2015, 58, 2584.
32 ACD/Log P calculator, version 12.02, within ACD/ChemSketch,
release 12.00, was used to calculate c log P values. http://
www.acdlabs.com.
33 C. D. Cox, G. B. McGaughey, M. J. Bogusky, D. B. Whitman,
R. G. Ball, C. J. Winrow, J. J. Renger and P. J. Coleman,
Bioorg. Med. Chem. Lett., 2009, 19, 2997.
50 P. Malherbe, E. Borroni, L. Gobbi, H. Knust, M. Nettekoven,
E. Pinard, O. Roche, M. Rogers-Evans, J. G. Wettstein and
J.-L. Moreau, Br. J. Pharmacol., 2009, 156, 1326–1341.
51 P. A. Gerrard, R. Porter, V. Holland, M. Massagrande, A.
Poffe, L. Piccoli, P. Bettica, M. Corsi, F. Hagan, J. Hagan and
E. Ratti, Sleep, 2009, 32(Suppl. 0125), A42.
34 A. S. Kalgutkar, I. Gardner, R. S. Obach, C. L. Shaffer, E.
Callegari, K. R. Henne, A. E. Mutlib, D. K. Dalvie, J. S. Lee, Y.
Nakai, J. P. O'Donnel, J. Boer and S. P. Harriman, Curr. Drug
Metab., 2005, 6, 161.
35 R. Zhang and F. Monsma, Curr. Opin. Drug Discovery Dev.,
2009, 12, 488.
36 R. A. Copeland, Expert Opin. Drug Discovery, 2010, 5, 305.
37 A. L. Gotter, C. J. Winrow, J. Brunner, S. L. Garson, S. V. Fox,
J. Binns, C. M. Harrell, D. Cui, K. L. Yee, M. Stiteler, J.
Stevens, A. Savitz, P. L. Tannenbaum, S. J. Tye, T. McDonald,
52 C. D. Cox, M. J. Breslin, D. B. Whitman, J. D. Schreier, G. B.
McGaughey, M. J. Bogusky, A. J. Roecker, S. P. Mercer, R. A.
Bednar, W. Lemaire, J. G. Bruno, D. R. Reiss, C. M. Harrell,
K. L. Murphy, S. L. Garson, S. M. Doran, T. Prueksaritanont,
W. B. Anderson, C. Tang, S. Roller, T. D. Cabalu, D. Cui,
G. D. Hartman, S. D. Young, K. S. Koblan, C. J. Winrow, J. J.
Renger and P. J. Coleman, J. Med. Chem., 2010, 53, 5320.
53 C. A. Baxter, E. Cleator, K. M. J. Brands, J. S. Edwards, R. A.
Reamer, F. J. Sheen, G. W. Stewart, N. A. Strotman and D. J.
Wallace, Org. Process Res. Dev., 2011, 15, 367.
This journal is © The Royal Society of Chemistry 2015
Med. Chem. Commun.