Design and synthesis of conformationally constrained N,N-disubstituted 1,4-diazepanes as potent orexin receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.01.138

Orexins are neuropeptides that regulate wakefulness and arousal. Small molecule antagonists of orexin receptors may provide a novel therapy for the treatment of insomnia and other sleep disorders. In this Letter we describe the design and synthesis of con

Full text of DOI:10.1016/j.bmcl.2010.01.138

Bioorganic & Medicinal Chemistry Letters 20 (2010) 2311–2315
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of conformationally constrained N,N-disubstituted
1,4-diazepanes as potent orexin receptor antagonists
a
b
a
a
Paul J. Coleman a, , John D. Schreier , Georgia B. McGaughey , Michael J. Bogusky , Christopher D. Cox ,
George D. Hartman a, Richard G. Ball c, C. Meacham Harrell d, Duane R. Reiss d, Thomayant Prueksaritanont e,
Christopher J. Winrow d, John J. Renger d
*
a Department of Medicinal Chemistry, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
b Department of Chemical Modeling and Informatics, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
c Department of Pharmaceutical Research and Development, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
d Department of Depression and Circadian Disorders, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
e Department of Drug Metabolism, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Orexins are neuropeptides that regulate wakefulness and arousal. Small molecule antagonists of orexin
receptors may provide a novel therapy for the treatment of insomnia and other sleep disorders. In this
Letter we describe the design and synthesis of conformationally constrained N,N-disubstituted 1,4-diaze-
panes as orexin receptor antagonists. The design of these constrained analogs was guided by an under-
standing of the preferred solution and solid state conformation of the diazepane central ring.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 7 January 2010
Revised 27 January 2010
Accepted 28 January 2010
Available online 8 February 2010
Keywords:
Orexin receptor
Sleep disorders
Diazepanes
Orexins (or hypocretins) are two neuropeptides that are se-
creted by a restricted group of neurons in the lateral hypothala-
mus. These neurons project to diverse regions of the central
nervous system including areas of the brain that modulate the
sleep-wake cycle.1 The orexin neuropeptides, Orexin A and Orexin
B, bind to two G protein-coupled receptors, OX1R and OX2R. There
is strong genetic and pharmacological evidence that orexins
promote and maintain wakefulness.2 Indeed, blockade of orexin
signaling by small molecule antagonists has been shown to pro-
mote sleep in preclinical species and in human clinical trials.3
In a previous publication from this laboratory, we described the
discovery of a novel series of diazepane dual orexin receptor
antagonists including 2-{4-[5-methyl-2-(2H-1,2,3-triazolyl-2-
yl)benzoyl]-1,4-diazepan-1-yl}quinazoline, 1.4 Compound 1 is a
brain-penetrant, potent dual orexin receptor antagonist that has
excellent OX2R/OX1R receptor affinity (Fig. 1). Indeed, when dosed
orally to rats with EEG-telemetry transmitters, compound 1 de-
creased wakefulness and increased slow-wave sleep at 100 mg/kg.
Despite the excellent intrinsic potency associated with 1, this com-
pound had poor in vitro metabolic stability and limited (<20%) oral
bioavailability in dogs and rats.
Spectroscopic, X-ray, and molecular modelling studies indicate
that N,N-disubstituted 1,4-diazepane antagonists such as 1 adopt
a low energy conformation that is characterized by an intramolec-
ular
p
-stacking interaction and a twist-boat ring conformation.5
This structure suggested to us that constraining the conformation-
ally mobile seven-membered diazepane ring could favorably im-
pact receptor potency and potentially address pharmacokinetic
limitations.6 Herein, we describe the synthesis and characteriza-
tion of a series of bridged diazepanes. These efforts identified a po-
N
N
N
O
N
N
N
N
Me
1 OX1R Ki = 1.2 nM
OX2R Ki = 0.6 nM
* Corresponding author. Tel.: +1 215 652 4618.
E-mail address: paul_coleman@merck.com (P.J. Coleman).
Figure 1. Structure and X-ray of orexin receptor antagonist 1.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.138

2312
P. J. Coleman et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2311–2315
1. LiAlH4, THF
2. BOC-anhydride
O
HN
N
90%
BOC
2
3
1. O3 then DMS
2. BnNH2, NaCNBH3
51%
Pd(OH)2, MeOH
2 (g), 65 C
Bn
H
NH
N
N
N
1000 PSI, 99%
BOC
BOC
5
4
N
N
N
EDC, HOBT,
DMF, 48%
CO2H
Me
6
N
N
1. HCl, dioxane
2. 2-chloroquinazoline,
K2CO3, DMF, 125 C
O
N
N
N
O
N
N
N
N
N
N
BOC
N
3. Chiral HPLC separation
37%
Me
Me
7
8
Scheme 1. Synthesis of 3,6-diazabicyclo[3.2.1]octanes as orexin receptor antagonists.
Table 1
Diazabicyclo[3.2.1]octane orexin receptor antagonists
R2
N
N
N
N
O
R1
N
Me
Compd
R1
R2
OX1R (Ki, nM)
OX2R (Ki, nM)
FLIPR
OX1R (IC50, nM)
OX2R (IC50, nM)
1
1a
H
F
1.2
1.0
0.6
0.2
29
36
27
27
N
N
N
N
N
N
N
N
N
N
Het
Het
Het
Het
Het
8a (rac)
8b (end A)
8c (end B)
H
H
2.3
110
0.7
3.6
155
0.9
42
2500
24
65
6200
44
H
H
F
9
12
11
77
110
14 (rac)
15 (rac)
870
1700
260
3500
>5700
3900
>10,000
3700
F

P. J. Coleman et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2311–2315
2313
tent, constrained analog of 1 with improved physicochemical prop-
erties and enhanced brain penetration.
man orexin receptors (expressed as Ki), and a FLIPR (Fluorometric
Imaging Plate Reader) assay in which calcium flux is measured as
a functional determinant of orexin antagonism (expressed as
IC50).11 These compounds all incorporated a 2-triazolyl-5-methylb-
enzamide discovered from previous optimization studies. The most
potent compound in this group was compound 8a. This racemic
mixture could be resolved by chiral HPLC and provided two enan-
tiomers 8b and 8c. The two stereoisomers differed in receptor
affinity by greater than 50-fold favoring single enantiomer 8c
which was equipotent to 1 with respect to OX2R binding and
slightly more potent on OX1R. In contrast, bridged compounds 9,
14, and 15 lost significant receptor potency (Fig. 1).
In order to constrain the flexible diazepane central ring in these
dual orexin receptor antagonists, we designed and synthesized
bridged diazepanes where we systematically varied the linkage
of a one carbon bridge on the seven-membered diazepane ring.
Preliminary modeling studies suggested that several of these con-
strained diazepane rings would enforce the low energy conforma-
tion previously observed with compound 1.7
A
series of 3,6-diazabicyclo[3.2.1]octanes were prepared
according to Scheme 1. In order to enable comparison between
various constrained cores, the 2-(2H-1,2,3-triazol-2-yl)-5-meth-
ylbenzamide was maintained as a conserved feature. Further, the
bridged diazepane was coupled with either an unsubstituted qui-
nazoline or with a 6-fluoroquinazoline.8 A solution of 2-azabicy-
clo[2.2.1]hept-5-en-3-one 2 was reduced with lithium aluminum
hydride and the amine product protected with BOC-anhydride.
Ozonolysis of alkene 3 followed by double, in situ reductive
alkylation between the unpurified dialdehyde and benzylamine
afforded the bridged diazepane 4 in moderate yield.9 Hydrogenol-
ysis of the benzyl group provided 5. Amide coupling with 2-(2H-
1,2,3-triazol-2-yl)-5-methylbenzoic acid 6, followed by acid
deprotection, nucleophilic aromatic substitution using 2-chloro-
quinazoline (or 6-fluoro-2-chloroquinazoline) afforded the recep-
tor antagonist 8 in Table 1. Rac-8a could be resolved by HPLC on
a chiral stationary support to afford enantiomers 8b and 8c.10
The sequence of transformations could be reversed and intermedi-
ate 5 treated with 2-chloroquinazoline followed by BOC-deprotec-
tion and amide coupling to give compound 9 (Table 1).
Synthesis of the 2,3-diazabicyclo[3.2.1]octanes is shown in
Scheme 2. Piperidone 10 underwent reductive alkylation with
benzylamine and intramolecular cyclization of 11 under basic con-
ditions gave bicyclic lactam 11. Removal of the CBZ-group, lactam
reduction, BOC-protection, and debenzylation afforded 12. Nucleo-
philic aromatic substitution with 6-fluoro-2-chloroquinazoline
provided 13. BOC-deprotection and standard amide coupling pro-
vided 14. The reaction sequence could be easily reordered from
intermediate 12 (amide formation and heteroarylation) to provide
15 (Table 1).
Detailed 1H NMR studies were performed on compound 8c to
determine its solution conformation. In solution, 8c exists as a
mixture of four components in equilibrium at À35 °C related to re-
stricted rotation of the amide and 2-aminoquinazoline rotamers.
The two major rotamers possess several strong NOE correlations
between the phenyl triazole ring and the quinazoline protons
(Fig. 2). For the two major rotamers, the 1H NMR methyl reso-
nances from the phenyltriazole group (1.37 ppm major; 1.17 ppm
minor) and the aromatic singlets (6.11 ppm, major; 5.97 ppm, min-
or) are shifted upfield which is characteristic of an aromatic ring
current shift. Reference compound C was prepared to determine
the inherent chemical shifts of the C6 phenyl and aromatic methyl
protons in a molecule in which a
available.
p-stacking interaction is not
Attempts to generate diffraction quality crystals of 8c were
unsuccessful; however we were able to crystallize a Cl-analog of
8c, which replaces the methyl group in 8c with a chlorine atom
to give 16 (Fig. 3). This molecule retains the binding affinity of 8c
and the structure of this molecule was solved by X-ray crystallog-
raphy.12 The solid state structure of 16 was consistent with the 1H
NMR structure of 8c and superimposes well on the X-ray generated
for compound 1 (Fig. 3). Based on the solved X-ray, the stereo-
chemistry of 29 was assigned as R, R.
We were unsuccessful in obtaining X-ray structures of the less
active, constrained diazepanes 9, 14, and 15. Interestingly, detailed
1H NMR studies were done on 15, and no significant NOE’s were
observed between the phenyltriazole and quinazoline resonances
indicating the ring systems were not proximal. Further, the key
chemical shifts of the methyl resonances from the phenyl triazole
(major rotamer 2.45 ppm; minor rotamer 2.41 ppm) and the
aromatic singlets (major rotamer 7.27 ppm; minor rotamer
Results: Rigidifying the conformationally mobile diazepane in 1
by installing a methylene bridge provides the compounds shown in
Table 1. To measure potency, we utilized a binding assay that
measures displacement of a high-affinity radioligand bound to hu-
O
1. Pd/C, H2 (g)
2. LiAlH4, THF/Et2O
1. BnNH2, NaCNBH3,
MeOH
O
CBZ
BOC
N
N
N
CBZ
CO2Me
3. BOC2O
4. Pd(OH)2, H2(g)
62% (4 steps)
2. NaOH (aq), THF/MeOH
37% (2 steps)
N
HN
Bn
10
11
12
N
Cl
N
F
Et3N, DMF, 62%
BOC
N
N
N
N
N
1. HCl, dioxane
O
N
F
N
N
N
F
2. EDC, HOBT, DMF
N
Acid 6, 55% (3 steps)
13
N
Me
14
Scheme 2. Synthesis of 2,3-diazabicyclo[3.2.1]octanes as orexin receptor antagonists.

2314
P. J. Coleman et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2311–2315
1.37
H3C
1.17
2.43
H3C
H3C
8.75
7.12
7.69
6.99
7.67
9.17
7.70
7.77
NOE's
7.42
7.87
NOE's
N
4
N
7.29
6.11
6
4
7.26
5.97
6
7.23
6
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
7.21
7.22
O
O
N
O
C
8c-rotamer B
8c-rotamer A
Figure 2. 1H NMR chemical shifts for compound 8c and reference compound C.
R2
R1
N
N
N
N
N
N
N
O
Compound 8c (R1 =H; R2 =Me);
OX1R Ki = 0.7 nM
OX2R Ki = 0.9 nM
Compound 16 (R1 =F; R2 =Cl);
OX1R Ki = 0.4 nM
OX2R Ki = 0.6 nM
X-ray structure of 16
X-ray overlap of 16 and 1
Figure 3. X-ray of orexin receptor antagonist 16.
that CSF concentrations of an orexin antagonist would be critical
component for predicting receptor-available free compound in
the brain.14 Compounds were administered by intravenous infu-
sion at 2 mg/kg and then the animals were sacrificed and brain,
plasma, and CSF levels were measured. Both 1 and 8c were
brain-penetrant with brain-plasma ratios of 0.7–0.8 (Table 2). Nei-
ther were substrates of the P-glycoprotein transporter.15 CSF con-
centrations of 8c were threefold higher than 1 and consistent
with the higher unbound concentration measured in plasma.
The pharmacokinetics of bridged compound 8c were evaluated
after oral administration to rats and dogs. As previously noted, 1
had poor pharmacokinetics in the rat with low oral bioavailability
(F = 2%). Orexin receptor antagonist 1 also had poor pharmacokinet-
ics in the dog with moderate clearance and short T1/2 (Table 3).
Unfortunately, bridged diazepane antagonist 8c did not exhibit im-
proved pharmacokinetics in the rat or dog. Bioavailability and plas-
ma exposure remained low in both species with moderate to high
plasma clearance.
Figure 4. Minimized Structure of 15.
7.25 ppm) do not indicate ring current shifts. Modeling studies on
15 (Fig. 4) provided a minimized structure where aryl-aryl overlap
was minimal (Fig. 4).
Compound 8c emerged as the most promising compound iden-
tified from these studies and was further characterized in terms of
the physicochemical properties, pharmacokinetic parameters, and
brain penetration. Despite introduction of the conformational con-
straint, none of the bridged compounds were significantly more
potent than the lead molecule 1.
Introduction of a ring constraint altered the physicochemical
properties of 8c relative to 1. The experimentally measured Log P
of 8c was lower than 1 (Table 2). Brain, plasma, and CSF concentra-
tions of compounds were assessed for compounds 1 and 8c in
anesthetized Sprague–Dawley rats.13 Although total brain concen-
tration is an important indicator of CNS penetration, we postulated
In summary, we have designed and synthesized constrained
derivatives of compound 1 that were inspired by an understanding
of the solution and solid state conformations of 1. The 3,6-diazabi-
cyclo[2.2.1]octane core in 8c enforces the energetically favorable
p–p interaction seen in 1 and maintains binding affinity for OX1R
and OX2R. Solution conformational analysis of 8c and X-ray crys-
tallography of analog 16 demonstrate that these conformational-
ly-restricted analogs have good overlap with the lowest energy
Table 3
Dog and rat pharmacokinetics for bridged compound 8c versus 1
Compd Species %F Cl (mL/min/
T1/2
(h)
AUC0–24
M h)
Cmax
M)
Table 2
kg)
(
l
(
l
Rat brain penetration and physical properties of key compounds
1
Dog
Rat
16 12
1.3
0.3
0.44
0.19
0.23
0.30
Compd
Rat PK at 2 mpk, 30 min; iv infusion
PB%
Log P
2
53
Plasma (nM)
Brain (nM)
CSF (nM)
Human (rat)
8c
Dog
Rat
8
2
24
45
0.9
0.7
0.49
0.23
0.51
0.15
1
8c
664
835
462
690
19
60
97.6 (97.1)
93.6 (92.7)
2.91
2.60
Rats dosed orally at 10 mpk, po in VitE-TPGS and dosed iv at 2 mpk in DMSO.

P. J. Coleman et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2311–2315
2315
9. (a) Bunnelle, W. H.; Daanen, J. F.; Ryther, K. B.; Schrimpf, M. R.; Dart, M. J.;
Gelain, A.; Meyer, M. D.; Frost, J. M.; Anderson, D. J.; Buckley, M.; Curzon, P.;
Cao, Y.-J.; Puttfarcken, P.; Searle, X.; Ji, J.; Putman, C. B.; Surowy, C.; Toma, L.;
Barlocco, D. J. Med. Chem. 2007, 50, 3627; (b) Fray, A. H.; Augeri, D. J.; Kleinman,
E. F. J. Org. Chem. 1988, 53, 896.
conformation of diazepane 1. These findings further substantiate
our hypothesis that the interaction in the diazepane series
p–p
leads to a highly favorable binding conformation with the orexin
receptors. While compound 8c does not have improved pharmaco-
kinetics in rats and dogs, the bridged core provides reduced plasma
protein binding and enhanced brain penetration. Further investiga-
tion of the in vitro and in vivo properties of diazepane-based scaf-
folds as orexin receptor antagonists will be reported in due course.
10. Racemic 8a was resolved by chiral HPLC on a ChiralPak AD column (2 Â 25 cm)
eluting with 40% hexanes/60% EtOAc. The first eluting isomer (8b) has a
retention time = 14.3 min and the second eluting enantiomer (8c) has
retention time = 17.9 min.
a
11. For detailed descriptions of these assays see: Supplementary data in Ref. 4.
12. Compound 16, C23H19Cl
a = 10.1978(7), b = 10.4373(6), c = 19.6892(11) Å, V = 2095.7(2) Å3, Z = 4,
Dx = 1.470 gcmÀ3, monochromatized radiation k(Cu) = 1.5418 Å, = 1.97 mmÀ1
F(0 0 0) = 960, T = 100 K. Data were collected on Oxford Diffraction CCD
diffractometer to a h limit of 66.4838° which yielded 6974 reflections. There are
2548 unique reflections with 2287 observed at the 2 level. The structure was
F N7O, Mr = 463.900, orthorhombic, P212121,
References and notes
l
,
a
1. For reviews: (a) Boss, C.; Brisbare-Roch, C.; Jenck, F. J. Med. Chem. 2009, 52, 891;
(b) Boss, C.; Brisbare-Roch, C.; Jenck, F.; Aissaoui, H.; Koberstein, R.; Sifferlen, T.;
Weller, T. Chimia 2008, 62, 974; (c) Roecker, A. J.; Coleman, P. J. Curr. Top. Med.
Chem. 2008, 8, 977.
2. (a) De Lecea, L.; Kilduff, T. S.; Peyron, C.; Gao, X.-B.; Foye, P. E.; Danielson, P. E.;
Fukuhara, C.; Battenberg, E. L. F.; Gautvik, V. T.; Bartlett, F. S., II; Frankel, W. N.;
Van Den Pol, A. N.; Bloom, F. E.; Gautvik, K. M.; Sutcliffe, J. G. Proc. Natl. Acad. Sci.
U.S.A. 1998, 95, 322; (b) Sakurai, T.; Amemiya, A.; Ishii, M.; Matsuzaki, I.;
Chemelli, R.; Tanaka, H.; Williams, S. C.; Richardson, J. A.; Kozlowski, G. P.;
Wilson, S.; Arch, J. R. S.; Buckingham, R. E.; Haynes, A. C.; Carr, S. A.; Annan, R.
S.; McNulty, D. E.; Liu, W.; Terrett, J. A.; Elshourbagy, N. A.; Bergsma, D. J.;
Yanagisawa, M. Cell 1998, 92, 573.
3. Brisbare-Roch, C.; Dingemanse, J.; Koberstein, R.; Hoever, P.; Aissaoui, H.;
Flores, S.; Mueller, C.; Nayler, O.; van Gerven, J.; de Haas, S. L.; Hess, P.; Qiu, C.;
Buchmann, S.; Scherz, M.; Weller, T.; Fischli, W.; Clozel, M.; Jenck, F. Nat. Med.
2007, 13, 150.
4. Whitman, D. B.; Cox, C. D.; Breslin, M. J.; Brashear, K. M.; Schreier, J. D.;
Bogusky, M. J.; Bednar, R. A.; Lemaire, W.; Bruno, J. G.; Hartman, G. D.; Reiss, D.
R.; Harrell, C. M.; Kraus, R. L.; Li, Y.; Garson, S. L.; Doran, S. M.; Prueksaritanont,
T.; Li, C.; Winrow, C. J.; Koblan, K. S.; Renger, J. J.; Coleman, P. J. ChemMedChem
2009, 4, 1069.
r
solved by direct methods (SHELXS-97, Sheldrick, G.M. Acta Crystallogr., 1990, A46,
467–473) and refined using full-matrix least-squares on F2 (SHELXL-97, Sheldrick,
G.M. SHELXL-97. Program for the Refinement of Crystal Structures. Univ. of Göttingen,
Germany). The finalmodelwas refinedusing298 parametersand all2548data. All
non-hydrogen atoms were refined with anisotropic thermal displacements. The
final agreement statistics are: R = 0.068 (based on 2287 reflections with I > 2
r(I)),
wR = 0.177, S = 1.10 with (D/r)max = 0.01. The maximum peak height in a final
difference Fourier map is 0.774 eÅÀ3 and this peak is without chemical
significance. CCDC 758078 contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
13. Seven to ten week old Sprague–Dawley rats were anesthetized with isoflurane
(5% with oxygen at 2 L/min). Once anesthetized, the rat was shaved from the
ears to the shoulders and then placed on a circulating water warming blanket
set at 37 °C. Anesthesia was maintained using 2% isoflurane by means of a nose
cone. The tail vein was cannulated with
a 25 G winged infusion needle
connected via tubing to a syringe containing the test compound. The syringe
was placed on a Harvard Apparatus infusion pump set to deliver 2 mg/kg over
30 min. At the end of the infusion, CSF was collected from the cisterna magna
by needle puncture with slow aspiration using a 25 G butterfly connected by
5. Cox, C. D.; McGaughey, G. B.; Bogusky, M. J.; Whitman, D. B.; Ball, R. G.;
Winrow, C. J.; Renger, J. J.; Coleman, P. J. Bioorg. Med. Chem. Lett. 2009, 19, 2997.
6. Reduction in numbers of rotatable bonds has been associated with improved
oral bioavailability. See: Veber, D. F.; Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 2615.
tubing to a 1 cc syringe. The CSF sample was dispensed into
a vial and
immediately frozen on dry ice. A 1 ml blood sample was then obtained via
cardiac puncture, centrifuged, plasma collected and frozen on dry ice. The brain
was rapidly harvested and frozen on dry ice. CSF, plasma, and brain samples
were then analyzed to determine compound levels.
7. Conformational searching using the mixed torsion/low-mode sampling
algorithm as implemented in Schrödinger, v9.0 (Schrödinger, LLC, v9.0) was
14. Martin, I. Drug Discovery Today 2004, 9, 161.
15. The MDR1 B–A/A–B ratio for 8c = 0.9 with an apparent permeability of
37 Â 10À6 cm/s in LLC-PK1 cells. For a description of this method see (a)
Hochman, J. H.; Yamazaki, M.; Ohe, T.; Lin, J. H. Curr. Drug Metab. 2002, 3, 257;
(b) Lin, J. H.; Yamazaki, M. Drug Metab. Rev. 2003, 35, 417.
performed using the OPLS force field in
a constant dielectric of 1. All
conformers within 5 kcal/mol of the global minimum were saved and
visually inspected for evidence of pi-stacking. See: Kolossváry, I.; Guida, W.
C. J. Am. Chem. Soc. 1996, 118, 5011.
8. The 6-fluoroquinazoline imparted additional binding affinity for OX2R (e.g.
Table 1; compound 1 vs 1a). For preparation of 6-fluoro-2-chloroquinazoline
see: Eur. Pat. Appl. 248554 A2, 1987.