Characterization of novel selective H1-antihistamines for clinical evaluation in the treatment of insomnia

Article abstract of DOI:10.1021/jm900933k

Analogues of the known H₁-antihistamine R-dimethindene were profiled as potential agents for the treatment of insomnia. Several highly selective compounds were efficacious in rodent sleep models. On the basis of overall profile, indene 1d and benzothiophene 2a had pharmacokinetic properties suitable for evaluation in night time dosing. Compound 2a did not show an in vivo cardiovascular effect from weak hERG channel inhibition.

Full text of DOI:10.1021/jm900933k

J. Med. Chem. 2009, 52, 5307–5310 5307
DOI: 10.1021/jm900933k
Histamine exerts a number of well documented peripheral
effects.10 In addition histamine is a neurotransmitter that has
acknowledged effects on arousal and wakefulness through
central action at H1 and H3 receptor subtypes.11 Indeed, first
generation H1-antihistamines including diphenhydramine,
doxylamine, and chlorpheniramine are sedating and have
been used for many years in over-the-counter sleep aids.
However, these agents also exhibit functional antagonism of
muscarinic receptors,12 a property thought to contribute to
undesirable side effects such as dry mouth, blurred vision,
constipation, tachycardia, urinary retention, and memory
deficits.13 Next-day impairment after bedtime use of these
antihistamines is common and has been attributed to long
plasma half-lives and protracted CNS exposure,14 suggesting
that an improved pharmacokinetic profile would be an
important determinant in the selection of novel H1-anti-
histamines for insomnia. Thus, selective H1-antihistamines
with appropriate duration of CNS exposure may provide an
alternative to current medications for the treatment of
insomnia, particularly for improved sleep throughout the
night and sleep efficiency.15 These agents would have low
potential for abuse and likely would be nonscheduled.
While many first generation H1-antihistamines are rela-
tively nonselective, some notable exceptions have been re-
ported including a derivative of the substructure 1 (Figure 1),
R-dimethindene (1a). 1a is the R enantiomer of the commer-
cially available racemate dimethindene and is highly selective
for the H1 receptor with sedating properties in daytime
studies.16 Single dose studies have indicated sedative effects
lasting up to 5 h,17 consistent with the reported half-life of
racemic dimethindene (∼5 h).18 More recently, doxepin, a
potent antihistamine with elimination half-life of 18 h,19 has
shown sleep maintenance effects lasting up to 8 h at low doses
with no next day residual effects.15 While 1a exhibited less
than ideal duration of action for the treatment of insomnia,
the compound represented an excellent starting point to
identify novel analogues for clinical evaluation that retained
receptor selectivity with potentially improved pharmacoki-
netic profile. Since metabolismof1aoccursprimarily via “first
pass” metabolic processes,20 pharmacokinetics of new related
compounds could be reasonably modeled from projections of
clearance. Subtle modifications around the structure of 1a
Characterization of Novel Selective
H1-Antihistamines for Clinical Evaluation in
the Treatment of Insomnia
Wilna J. Moree, Bin-Feng Li, Florence Jovic,
Timothy Coon, Jinghua Yu, Raymond S. Gross,
Fabio Tucci, Dragan Marinkovic, Said Zamani-Kord,
Siobhan Malany, Margaret J. Bradbury,
Lisa M. Hernandez, Zhihong O’Brien, Jianyun Wen,
Hua Wang, Samuel R. J. Hoare, Robert E. Petroski,
Aida Sacaan, Ajay Madan, Paul D. Crowe, and
Graham Beaton*
Neurocrine Biosciences, 12780 El Camino Real, San Diego,
California 92130
Received June 24, 2009
Abstract: Analogues of the known H1-antihistamine R-dimethin-
dene were profiled as potential agents for the treatment of insomnia.
Several highly selective compounds were efficacious in rodent sleep
models. On the basis of overall profile, indene 1d and benzothio-
phene 2a had pharmacokinetic properties suitable for evaluation in
night time dosing. Compound 2a did not show an in vivo cardio-
vascular effect from weak hERG channel inhibition.
Current epidemiological data suggest that insomnia is one
of the most commoncentral nervoussystem (CNSa) disorders
with high prevalence in industrialized nations.1 Insomnia has
a significant economic impact on managed care2 with detri-
mental effects on nations’ workforces resulting from higher
work absenteeism and decreased job performance.3
The most common treatments for insomnia are pharmaco-
logic agents that act on the benzodiazepine binding site of
the γ-aminobutyric acid (GABAA) receptor and include
zolpidem, temazepam, and eszopiclone. These agents provide
positive effects in terms of sleep onset and sleep maintenance.
However, side effects including daytime sedation, cognitive
impairment, and motor effects raise concerns of increased
risk of motor vehicle accidents and injuries from falls.4,5
Moreover, recent safety labeling changes warn of complex
behaviors with amnesia following sedative-hypnotic use.6
These agents have also been associated with the potential
for abuse and dependence in at-risk populations and are
classified as schedule IV controlled substances in the U.S.7
Accordingly, there remains significant interest in developing
alternative sleep agents that demonstrate sleep onset and
maintenance and improved sleep quality throughout the night
without next day sedation. Novel hypnotic mechanisms8 that
have been recently evaluated or are still under investigation
include the atypical GABAA receptor agonist gaboxadol, the
recently approved melatonin receptor agonist ramelteon,
serotonin 5-HT2A receptor modulators, and orexin receptor
antagonists.9
*To whom correspondence should be addressed. Phone 858-337-
1801. Fax 858-486-9892. E-mail: beaton.graham@gmail.com.
a Abbreviations: CNS, central nervous system; GABAA, γ-aminobu-
tyric acid; QTc, corrected QT interval; EEG, electroencephalography;
NREM, nonrapid eye movement; REM, rapid eye movement.
Figure 1. R-Dimethindene 1a and related substructures.
r
2009 American Chemical Society
Published on Web 08/07/2009
pubs.acs.org/jmc

5308 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 17
Scheme 1. Synthesis of Indene Analoguesa
Moree et al.
Scheme 2. Synthesis of Benzothiophene and Benzofuran Ana-
loguesa
a (a) C4H3N2CH2R1, LDA, THF, 0 °C; (b) HCl (aq) 100 °C; (c) chiral
SFC; (d) n-BuLi, p-FC6H4CH2Br, THF, -78 °C.
were anticipated to display differences in pharmacokinetic
parameters such as clearance and volume of distribution while
maintaining the favorable H1 activity and selectivity profile.
On the basis of 1a, three substructures were targeted: 1,
benzothiophene (2), and benzofuran (3). Novelty was gener-
ated by selected R1 (methyl or hydrogen) and R2 (heteroaryl)
substitutions of the indene core or more generalized substitu-
tion in the benzothiophene or benzofuran structures. This
study had two objectives. The first was to identify potent,
brain penetrating H1-antihistamines with selectivity relative
to other key targets. Low muscarinic receptor affinity was
required to avoid anticholinergic side effects previously
noted. Selectivity relative to 5HT2A and H3 receptors was of
interest, since activity at these targets could affect the sleep
profile of candidate compounds. hERG potassium channel
inhibition was also an important consideration, as a number
of antihistamines are potent inhibitors of this channel.21 Some
of these, such as astemizole, are potent hERG channel
inhibitors and are thought to cause QT interval or corrected
QT interval (QTc) prolongation leading to potentially fatal
cardiac arrhythmias. P450 enzyme selectivity and metabolic
stability were other important considerations in primary
profile assessment. The second objective was to evaluate in
vivo properties of the more selective compounds for utility as
sleep agents for efficacy in a rat electroencephalography
(EEG) model and projected pharmacokinetic profile in
humans.
The synthesis of seven analogues is outlined in Schemes 1 and
2. Indene analogues 1b-d (Scheme 1) were either synthesized in
a fashion analogous to that previously reported for 1a,22
through base-mediated coupling of a suitable alkyl pyrazine
to indanone 4 and acidic work up (analogues 1c, 1d) or
alkylation of the indene 5 (1b). Benzothiophene or benzofuran
versions of 1a were synthesized according to one of three
approaches outlined in Scheme 2. Direct analogues of 1a were
prepared from acetyl intermediates 7a or 7b. Compound 7a was
obtained from a Heck coupling with 6a, while 7b was prepared
by lithiation of intermediate 6b followed by quenching with
acetic anhydride. Final products 2a and 3a were obtained
following addition of 2-lithiopyridine to 7a or 7b and stepwise
dehydration and reduction. Arene substituted benzothio-
phene 2b was prepared by Friedel-Crafts acylation of the
intermediate 8. Reduction of the ketone 9 afforded 2b.
Pyrazine 2c was prepared from lithiation of intermediate
6a and quenching with pyrazine carboxaldehyde. Deoxy-
genation of the alcohol 10 yielded 2c. Chiral materials were
obtained by fractional crystallization or preparative chiral
SFC or HPLC of racemate. The chiral products were
characterized by chiral SFC and/or optical rotation. Crys-
tallographic data were obtained for one product to confirm
absolute stereochemistry.
a (a) n-BuOCHdCH2, Pd(OAc)2, P(o-Tol)3, Et3N, CH3CN, 100 °C;
(b) 20% HCl, THF; (c) n-BuLi, TMEDA, Ac2O, toluene, -78 °C;
(d) C5H4NLi, TMEDA, toluene or DCM, -78 °C; (e) H2SO4, TFA;
(f) H2, Pd/C, MeOH; (g) fractional crystallization; (h) chiral HPLC;
(i) p-FC6H4COCl, 1M AlCl3 in C6H5NO2, C2H4Cl2, 0 °C; (j) LiAlH4,
THF, 0 °C; (k) NaBH4, TFA; (l) n-BuLi, TMEDA, C4H3N2CHO,
toluene, -78 °C; (m) HI, HOAc.
in any of the examples. Human microsome stability data
displayed sufficient variation in predicted clearance rates,
suggesting that it would be possible to identify compounds
with varying pharmacokinetic profiles in the analogues pre-
pared. With the exception of 2c, all of the compounds tested
were potent binders to the H1 receptor with greater than 100-
fold selectivity relative to M1. Examples 1a-d in the indene
series and benzothiophene 2a were greater than 1000-fold
selective for this receptor. Data for 1a-d indicated that
secondary and tertiary carbon analogues had similar H1
potency. In the benzothiophene class, arene substitution in
2b provided a slightly less potent compound. Interestingly,
pyrazine analogue 2c was a markedly less potent H1-antihis-
tamine, indicating subtle differences between the indene and
benzothiophene classes. Introduction of the p-fluorophenyl
substitution to 1b and 2b resulted in significant CYP2D6
inhibition and potent affinity for 5HT2A. In contrast, 1a, 1c,
1d, 2a, and 3a were greater than 1000-fold selective for
CYP2D6, highlighting the value of the heterocyclic group
for improved drug likeness. Benzofuran 3a was potent in H1
binding but an order of magnitude less selective for M1
compared to 1a and benzothiophene 2a. Compound 3a also
displayed significant affinity for M3 (Ki=76 nM) that was not
observed for the other analogues. Indene 1c was less selective
versus 5HT2A compared to its chiral analogues 1a and 1d.
None of the compounds showed significant affinity for H3
receptor. hERG channel data for 1c, 1d, 2a, and 3a showed
weak inhibition comparable to 1a. Overall these data indi-
cated suitable receptor selectivity profiles for 1c, 1d, and 2a.
These compounds were further assessed for effects on sleep-
wake parameters in rats.
Table 2 summarizes sleep parameters for 1a, 1c, 1d, and 2a.
All the compounds tested exhibited a similar and significant
increase in nonrapid eye movement (NREM) sleep. None of
the compounds tested showed appreciable effects on latency
Initial profile data of 1a and its analogues are shown in
Table 1. No significant inhibition of CYP3A4 was observed

Letter
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 17 5309
Table 1. In Vitro Profile of R-Dimethindene Analoguesa
H1,
Ki ( SEMb (nM)
M1,
Ki ( SEMb (μM)
5HT2A
,
CYP2D6,
IC50 (μM)
hERG,c
human systemic
clearanced ((mL/min)/kg)
compound
Ki ( SEMb (μM)
IC50 (μM)
1a
1b
1c
1d
2a
2b
2c
3a
0.4 ( 0.04
0.60 ( 0.03
4.0 ( 0.3
1.3 ( 0.2
4.0 ( 0.5
11 ( 5
2.4 ( 0.2
14% at 10 μM
19 ( 2
20% at 10 μM
5.3 ( 0.3
2.0 ( 0.0
ND
33% at 10 μM
0.14 ( 0.15
0.60 ( 0.04
32% at 10 μM
31% at 10 μM
0.09 ( 0.04
ND
32.5
0.3
0.9
7
ND
5.5
13
12
5
6.7
50.5
28.2
0.2
1.4
1.4
10
16
ND
10
ND
ND
1.2
75 ( 6
0.9 ( 0.2
ND
3.3
0.32 ( 0.03
62% at 10 μM
a ND: not determined. b Minimum of 3 experiments. c For comparison, cisapride and the H1-antihistamines astemizole and diphenhydramine were
tested in these cells (IC50 = 36 nM, 10 nM, and 2 μM, respectively). d Predicted from human liver microsomes.
Table 2. Efficacy Profile of R-Dimethindene Analoguesa
human microdose study where measured half-lives for
compounds 2a and 1d were 6.8 and 12 h, respectively.24
Measured clearance values for 2a and 1d were low at 5.2 and
3.3 (mL/min)/kg, respectively.
rat
intrinsic
clearance,c
latency
dose, NREM to 1 min concn (1 h),
compound (mg/kg) in 4 hb NREMb
plasma
(ng/mL) ((mL/min)/kg)
The absence of specific reports concerning QT effects in the
use of dimethindene and the weak hERG inhibition signal
observed for these compounds prompted further assessment
of the risk for QTc prolongation. A cardiovascular safety
study of 2a in conscious monkeys produced a slight decrease
in heart rate (16-19%) compared to vehicle control at
all doses up to 200 mg/kg, beginning 1 h postdose and
returning to near control levels approximately 10 h postdose.
Reflective of this decrease, there were slight increases in the
electrocardiograph RR and PR intervals, QTc and duration
of the QRS complex.25 When QTc interval was corrected for
heart rate, there were no notable differences between the test
article treated groups and the control group (range of QTc
changeis-1 to 5 ms) atexposures atleast 100-fold higherthan
those projected for clinically efficacious doses of 2a. This
margin was significantly higher than previous guidelines for
the assessment of hERG inhibitors21 and underscored the
importance of an in vivo assessment of 2a to qualify the in
vitro findings. From the overall data profile 2a was advanced
to clinical trials for further assessment.
In summary, a number of novel, selective H1-antihistamines
were identified and examined for utility in the treatment of
insomnia. A key driver in compound selection focused on
pharmacokinetic properties which led to the identification of
analogues 1d and 2a that were efficacious in a rat EEG model
with suitable clearance profiles. While they are weakinhibitors
of the hERG channel, in vivo assessment of 2a in monkeys
showed no significant effects on QTc, indicating a large safety
margin for this parameter. As such, 2a appears to be a suitable
candidate for further assessment as a sleep agent.
1a
1c
60
60
60
30
30
132 ( 7 134 ( 7
159 ( 6 79 ( 7
145 ( 5 109 ( 7
142 ( 6 124 ( 12 4.0 ( 0.1
220 ( 6 18 ( 2 ND
ND
16 ( 13
86 ( 45
2600
1500
1500
2300
ND
1d
2a
zolpidem
a ND: not determined. b Calculated as a percentage of vehicle. c Pre-
dicted from rat liver microsomes.
to NREM or rapid eye movement (REM) sleep parameters.
Only 2a displayed a significant increase in REM sleep (168%
of vehicle). Plasma concentrations from surrogate animals for
1c, 1d, and 2a were low despite the large doses employed.
While relatively stable in human microsomes, all the com-
pounds were rapidly metabolized by the rat counterpart
(Table 2) with observed clearance values at the limit of assay
detection. Surrogate PK studies of 2a at a dose of
30 mg/kg showed brain levels of 29 ng/g at 1 h and 23 ng/g at
4 h. Thesedataindicated that this analogue was brain penetra-
ting and achieved suitable exposure over the course of the
EEG experiment despite relatively low plasma concentra-
tions. Overall, these studies were consistent with reported
effects of H1-antihistamines on sleep-wake parameters
in rats.23
From the available data, compounds 1c, 1d, and 2a were
selective H1-antihistamines with rodent efficacy comparable to
1a. It was hypothesized that analogues of 1a with similar or
incrementally longer half-lives would represent suitable candi-
dates for further evaluation of dose, exposure, and sedative
action. Human microsome stability data indicated that 1c
displayed higher clearance compared to 1a. In contrast, 1d
and 2a exhibited predicted systemic clearance values similar to
1a [7 (mL/min)/kg]. Allometric scaling of these two compounds
projected human clearances of approximately 9 (mL/min)/kg,24
a value close to that reported for racemic dimethindene
[8.0 (mL/min)/kg],18 indicating that these were candidates of
interest. Both compounds were subsequently assessed in a
Acknowledgment. We thank John Saunders, Paul Conlon,
Haig Bozigian, and Wendell Wierenga for support of this
work and Tao Hu, Mila Lagman, Chris DeVore, and John
Harman for analytical chemistry support.
Supporting Information Available: Experimental procedures,
analytical and spectral characterization data, EEG data and

5310 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 17
Moree et al.
crystallographic information in CIF format. This material is
available free of charge via the Internet at http://pubs.acs.org.
(15) Roth, T.; Rogowski, R.; Hull, S.; Schwartz, H.; Koshorek, G.;
Corser, B.; Seiden, D.; Lankford, A. Efficacy and safety of doxepin
1mg, 3mg, and 6mg in adults with primary insomnia. Sleep 2007,
30, 1555–1561.
(16) Nicholson, A. N.; Pascoe, P. A.; Turner, C.; Ganellin, C. R.;
Greengrass, P. M.; Casy, A. F.; Mercer, A. D. Sedation and
histamine H1-receptor antagonism: studies in man with the en-
antiomers of chlorpheniramine and dimethindene. Br. J. Pharma-
col. 1991, 104, 270–276.
References
(1) Ohayon, M. M.; Lemoine, P. Sleep and insomnia markers in the
general population. Encephale 2004, 30, 135–140.
(2) Fullerton, D. S. P. The economic impact of insomnia in managed
care: a clearer picture emerges. Am. J. Managed Care 2006,
12 (8, Suppl), S246–S252.
(3) Zammit, G. K.; Weiner, J.; Damato, N.; Sillup, G. P.; McMillan, C.
A. Quality of life in people with insomnia. Sleep 1999, 22, S379–S385.
(4) Glass, J.; Lanctot, K. L.; Herrmann, N.; Sproule, B. A.; Busto, U.
E. Sedative hypnotics in older people with insomnia: A meta-
analysis of risks and benefits. Br. Med. J. 2005, 331, 1169–1175.
(5) Vermeeren, A. Residual effects of hypnotics. CNS Drugs 2004, 18,
297–328.
(6) Tsai, M.-J.; Tsai, Y.-H.; Huang, Y.-B. Compulsive activity and
anterograde amnesia after zolpidem use. Clin. Toxicol. 2007, 45,
179–181.
(7) Hajak, G.; Muller, W. E.; Wittchen, H. U.; Pittrow, D.; Kirch, W.
Abuse and dependence potential for the non-benzodiazepine hyp-
notics zolpidem and zopiclone: a review of case reports and
epidemiological data. Addiction 2003, 98, 1371–1378.
(8) Navab, P.; Guilleminault, C. Emerging pharmacotherapeutic
agents for insomnia: a hypnotic panacea? Expert Opin. Pharmac-
other. 2006, 7, 1731–1738.
(9) Roecker, A. J.; Coleman, P. J. Orexin receptor antagonists: med-
icinal chemistry and therapeutic potential. Curr. Top. Med. Chem.
2008, 8, 977–987.
ꢀ
^
(17) Cicurel, L.; Pechadre, J. C.; Grandjean, E.; Duchene-Marullaz, P.
Comparison of the central and peripheral antihistaminic effects of
dimethindene maleate (fenistil) and its enantiomers. Agents Actions
1992, C440–C443.
(18) Arnera, V.; Wermeille, M.; Wellman, M.; Llull, M.; Althaus,
M. A.; Balant, L. P. Pharmacokinetics of dimetindene after in-
travenous and oral administration to healthy volunteers. Arzneim.-
Forsch./Drug Res. 1990, 40 (II), 1346–1348. Clearance calculations
estimated on an average human weight of 60 kg.
(19) Virtanen, R.; Scheinin, M.; Iisalo, E. Single dose pharmacokinetics
of doxepin in healthy volunteers. Acta Pharmacol. Toxicol. 1980,
47, 371–376.
(20) Prein, D.; Blaschke, G. Studies of the metabolism of the antihist-
aminic drug dimethindene by high-performance liquid chromato-
graphy and capillary electrophoresis including enantioselective
aspects. J. Chromatogr., B 1997, 688, 309–318.
(21) Redfern, W. S.; Carlsson, L.; Davis, A. S.; Lynch, W. G.; MacK-
enzie, I.; Palethorpe, S.; Siegl, P. K. S.; Strang, I.; Sullivan, A. T.;
Wallis, R.; Camm, A. J.; Hammond, T. G. Relationships between
preclinical cardiac electrophysiology, clinical QT interval prolon-
gation 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.
(10) Simons, F. E. R. Advances in H1-antihistamines. N. Engl. J. Med.
2004, 351, 2203–2217.
(22) Bohme, T. M.; Keim, C.; Kreutzmann, K.; Linder, M.; Dingermann,
¨
(11) Barbier, A. J.; Bradbury, M. J. Histaminergic control of sleep wake
cycles: recent therapeutic advances for sleep and wake disorders.
CNS Neurol. Disord.: Drug Targets 2007, 6, 31–43.
(12) Kubo, N.; Shirakawa, O.; Kuno, T.; Tanaka, C. Antimuscarinic
effects of antihistamines: quantitative evaluation by receptor-bind-
ing assay. Jpn. J. Pharmacol. 1987, 43, 277–282.
(13) Meolie, A. L.; Rosen, C.; Kristo, D.; Kohrman, M.; Gooneratne,
N.; Aguillard, R. N.; Fayle, R.; Troell, R.; Townsend, D.; Claman,
D.; Hoban, T.; Mahowald, M. Oral nonprescription treatment for
insomnia: an evaluation of products with limited evidence. J. Clin.
Sleep Med. 2005, 1, 173–187.
(14) Kay, G. G.; Plotkin, K. E.; Quig, M. B.; Starbuck, V. N.; Yasuda,
S. Sedating effects of AM/PM antihistamine dosing with evening
chlorpheniramine and morning terfenadine. Am. J. Managed Care
1997, 3, 1843–1848.
T.; Dannhardt, G.; Mutschler, E.; Lambrecht, G. Structure-
activity relationships of dimethindene derivatives as new M2-
selective muscarinic receptor antagonists. J. Med. Chem. 2003,
46, 856–867.
(23) Saitou, K.; Kaneko, Y.; Suginoto, Y.; Chen, Z.; Kamei, C. Slow
wave sleep-inducing effects of first generation H1-antagonists. Biol.
Pharm. Bull. 1999, 22, 1079–1082.
(24) Madan, A.; O’Brien, Z.; Wen, J.; O’Brien, C.; Farber, R. H.;
Beaton, G.; Crowe, P. D.; Oosterhuis, B.; Garner, R. C.; Lappin,
G.; Bozigian, H. P. A pharmacokinetic evaluation of five H1
antagonists after oral and intravenous microdose to human sub-
jects. Br. J. Clin. Pharmacol. 2009, 67, 288–298.
(25) Becker, D. E. Fundamentals of Electrocardiography Interpreta-
tion. Anesth. Prog. 2006, 53, 53–64.