Isochroman-6-carboxamides as highly selective 5-HT(1D) agonists: Potential new treatment for migraine without cardiovascular side effects [2]

Full text of DOI:10.1021/jm980137o

2180
J . Med. Chem. 1998, 41, 2180-2183
Ta ble 1. CNS Binding Affinities (Ki, nM) for Sumatriptan (1),
8, and Isochroman-6-carboxamidesa
Isoch r om a n -6-ca r boxa m id es a s High ly
Selective 5-HT1D Agon ists: P oten tia l New
Tr ea tm en t for Migr a in e w ith ou t
Ca r d iova scu la r Sid e Effects
compd
5-HT1D 5-HT1B 5-HT1A 5-HT2A
D2
D4
1
7
8
2.0
8.6
6.7
3.8
238
0.9
6.4
5862
7242
3023
26021
5775
150300
341
2463
899
376
140
76
248
220
168
586
>218 Ib
348 1054
1431 2.4
(S)-(-)-7
(R)-(+)-7
(S)-(-)-10
967
325
386
I
I
Michael D. Ennis,*,† Nabil B. Ghazal,†
Robert L. Hoffman,† Martin W. Smith,‡
Siusaidh K. Schlachter,‡ Charles F. Lawson,‡
Wha Bin Im,§ J effrey F. Pregenzer,§
Kjell A. Svensson,§ Richard A. Lewis,§
Edward D. Hall,§ Dorothy M. Sutter,§
Lorri T. Harris,| and Robert B. McCall|
>3356
1092
>3356
241 >3704
(R)-(+)-10 298
158
I
a
Values for 5-HT1D and 5-HT1B were determined using mem-
branes prepared from cloned gorilla receptors expressed in cul-
tured HEK 293 cells (see ref 13). Cloned human receptors were
used for the 5-HT1A, 5-HT2A, D2, and D4 determinations (see
b
Supporting Information). I ) inactive, defined as less than 50%
inhibition of test ligand binding at 10-6 M.
Pharmacia & Upjohn, Inc., Kalamazoo, Michigan 49001
Received March 4, 1998
to give 6. Metal-halogen exchange of 6 with t-BuLi and
subsequent quenching of the resulting aryl anion with
N-trimethylsilylisocyanate afforded the 6-carboxamide
7. When evaluated in a battery of CNS receptor binding
assays (Table 1), 7 was found to possess poor affinity
at the D4 site (Ki 1054 nM), in sharp contrast to the
unsubstituted isochroman analogue 8 (Ki 2.4 nM), a
selective dopamine D4 antagonist.12 However, amide 7
displayed unexpected affinity for the 5-HT1D receptor
(Ki 8.6 nM). Furthermore, 7 showed little affinity for
the 5-HT1B site, exhibiting at least a 100-fold preference
Migraine headache is a debilitating, chronic disease
affecting the lives of millions of people worldwide.1,2
Migraine is most prevalent in working age populations
(35-40 years old), and its economic impact in the United
States alone has been estimated in the billions of
dollars.3,4 Currently, the primary treatment for mi-
graine is Sumatriptan (1), a nonselective serotonergic
agent which binds with high affinity to both the 5-HT1D
and 5-HT1B receptors (Ki of 3.4 and 7.7 nM respec-
tively).5,6 Recent theories regarding the etiology of
for 5-HT1D
.
Thus, it appeared that isochroman-6-
carboxamides such as 7 may represent a new, non-
tryptamine-based family of selective 5-HT1D agents.
The serendipitous finding of 5-HT1D selectivity within
the isochroman-6-carboxamide family prompted us to
prepare the enantiomers of 7. For this purpose, we
employed an enzymatic-based resolution procedure
similar to that previously described for the related D4
isochroman compounds.12 As illustrated in Scheme 2,
treatment of racemic bromo ester 3 with Amano P-30
lipase results in a highly selective hydrolysis of the (S)-
ester antipode to provide a readily separable mixture
of the carboxylic acid (S)-(-)-4 and the recovered ester
(R)-(+)-3 in yields of 86% and 91% and optical purities
of 99% and 96% ee, respectively.14 From these resolved
products were prepared the corresponding enantiomers
(S)-(-)-7 and (R)-(+)-7. As can be seen from Table 1,
binding affinity at the 5-HT1D receptor was exquisitely
sensitive to the stereochemical configuration of the
isochroman C-1 position, with the more potent isomer
being the (S)-(-)-7 compound. Interestingly, the other
serotonin and dopamine receptors evaluated showed
little dependence upon stereochemistry in their binding
affinities for these compounds.
The optically pure primary amide (S)-(-)-7 served as
a useful starting point for the preparation of a number
of secondary and tertiary amide derivatives. Treatment
of (S)-(-)-7 with di-tert-butyl dicarbonate generates the
bis-BOC derivative (S)-(-)-9 (Scheme 3), which can be
reacted with a variety of primary and secondary
amines.15 Use of methylamine in this reaction provides
the N-methyl analogue (S)-(-)-10 (PNU-109291), which
proved to be among the most potent and selective
5-HT1D compounds prepared to date, displaying a
5-HT1D Ki of 0.9 nM and an B/D ratio of over 5000. As
migraine suggest that selective agonists at only the
5-HT1D site could be effective at ameliorating migraine
symptoms.7,8 Indeed, it has been proposed that the
5-HT1B properties of Sumatriptan are responsible for
its vasoconstrictive properties and associated cardio-
vascular side effects.9,10 Nearly all of the compounds
currently in development for treating migraine are
tryptamine derivatives which possess little or no selec-
tivity between the 5-HT1D and 5-HT1B sites.11 In this
Communication, we describe the first class of highly
selective, non-indole 5-HT1D agonists. One of these
compounds, (S)-(-)-10, is significantly more potent than
Sumatriptan in animal models of migraine yet devoid
of vasoconstrictive properties.
While exploring the structure-affinity relationships
in a series of isochroman-arylpiperazines targeting the
dopamine D4 receptor,12 we prepared the 6-carboxamide
analogue 7 shown in Scheme 1. Reaction of 4-bro-
mophenethyl alcohol 2 with ethyl 3,3-diethoxypropi-
onate in the presence of TiCl4 provided the isochroman
3. Hydrolysis of 3 to the carboxylic acid 4 and subse-
quent DECP coupling with 4-methoxyphenylpiperazine
provided the amide 5, which was reduced with borane
† Structural, Analytical and Medicinal Chemistry.
‡ Discovery Technologies
§ CNS Diseases Research
| Cardiovascular Pharmacology
S0022-2623(98)00137-X CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/03/1998

Communications to the Editor
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13 2181
Sch em e 1
Sch em e 2
Sch em e 3
with the primary amide 7, the (S)-(-)-10 isomer pos-
sessed far superior affinity for the 5-HT1D site relative
to the (R)-(+)-10 enantiomer. In addition to its affinity
at the 5-HT1D site, (S)-(-)-10 also possesses weak
affinity for the 5-HT2A (168 nM) and dopamine D2 (241
nM) sites. Finally, (S)-(-)-10 was found inactive (<50%
inhibition of test ligand at 10-6 M) in a broad panel of
CNS receptors, including the 5-HT2C, 5-HT3, 5-HT4, and
5-HT uptake sites.
The discovery of the isochroman-6-carboxamide (S)-
(-)-10 as the first truly selective 5-HT1D ligand has
provided a valuable tool for the elucidation of the
pharmacological role played by the 5-HT1D receptor. We
have been able to demonstrate in vivo activity with (S)-
(-)-10 which we believe reflects the 5-HT1D agonist
properties of this unique compound and suggests ef-
ficacy for the treatment of migraine headache.16 In
addition, we have determined that (S)-(-)-10 does not

2182 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13
Communications to the Editor
Ta ble 2. Guinea Pig Hypothermia
hypothermiaa
dose
(µmol/kg, sc)
∆ °C at
30 min
∆ °C at
60 min
compd
vehicle
sumatriptan (1)
(S)-(-)-7
(S)-(-)-10
(S)-(-)-10
(S)-(-)-10
0.06 ( 0.16
0.08 ( 0.19
25
50
25
50
100
-0.14 ( 0.13
-1.42 ( 0.14b -1.14 ( 0.18b
-0.80 ( 0.11b -0.54 ( 0.33
-1.84 ( 0.26b -2.34 ( 0.53c
-1.38 ( 0.09b -1.52 ( 0.07b
a
The hypothermic response is presented as the difference in
°C from control (vehicle) and is expressed as mean ( SEM (n )
b
4-5). P < 0.01 vs vehicle. c P < 0.05 vs vehicle.
possess the vasoconstrictive activity which can be
demonstrated with Sumatriptan.
The guinea pig hypothermia assay has been described
as a simple in vivo assay for 5-HT1D agonists.17 The
hypothermic response is believed to reflect agonist
action within the CNS, since compounds with poor brain
penetration (e.g., Sumatriptan) are essentially inactive
after systemic administration. Given in Table 2 are the
results from this assay for Sumatriptan, (S)-(-)-7, and
(S)-(-)-10. Both of the isochroman amides exhibited
good hypothermic effects, indicating that these com-
pounds are agonists that readily cross the blood-brain
barrier. The excellent binding selectivity of both (S)-
(-)-7 and (S)-(-)-10 strongly suggests that this response
is being mediated by 5-HT1D receptors.18
F igu r e 1. Inhibition of guinea pig dural plasma extravasa-
tion: comparison of Sumatriptan and (S)-(-)-10. Values )
mean ratio of 125I albumin extravasation in the stimulated side
of the dura to that in the unstimulated side (( standard error)
for n ) 4-8 animals/group. P < 0.05 versus control.
The data presented in Tables 1 and 2 clearly indicate
that the isochroman-6-carboxamide (S)-(-)-10 is a selec-
tive and potent 5-HT1D agonist with good CNS penetra-
tion. Recent reports regarding the involvement of
5-HT1D receptors in meningeal pain prompted us to
examine the potential antimigraine activity of (S)-(-)-
10. Moskowitz has described a neurogenic inflamma-
tion model for migraine in which unilateral stimulation
of the trigeminal ganglia results in the colateral release
of permeability-enhancing autacoids (such as substance
P and CGRP) from the trigeminal nerve terminals
innervating meningeal blood vessels.19 The release of
these autacoids causes extravasation of plasma protein
from these dural vessels. Pretreatment of guinea pigs
with radiolabeled plasma protein (125I bovine serum
albumin) allows for quantification of the plasma ex-
travasation in the stimulated and unstimulated halves
of the dura and provides a means to measure the
efficacy with which compounds inhibit this neurogenic
inflammation. Moskowitz has shown that nonselective
5-HT1D/1B agonists such as Sumatriptan and dihydro-
ergotamine are effective at blocking plasma extravasa-
tion in this model.20,21 As shown in Figure 1, we have
demonstrated that (S)-(-)-10, in addition to being
5-HT1D selective, is also significantly more potent than
Sumatriptan in blocking neurogenic inflammation in the
guinea pig model of migraine. Nearly complete inhibi-
tion of plasma extravasation can be achieved in this
model using 1.0 µg/kg iv of (S)-(-)-10, whereas a dose
of 300 µg/kg iv of Sumatriptan is required for compa-
rable activity.22 The difference in potency between (S)-
(-)-10 and Sumatriptan in the neurogenic inflammation
assay is not predicted by their nearly equivalent 5-HT1D
binding affinities (0.9 and 2.0 nM, respectively). One
possible explanation for this difference recognizes the
superior brain penetration of (S)-(-)-10, allowing for
F igu r e 2. Comparison of Sumatriptan and (S)-(-)-10 on
carotid artery resistance in the cat. Values ) mean increase
( standard error.
augmentation of the antiinflammatory effect by access-
ing centrally located 5-HT1D receptors.
It has been proposed that the 5-HT1B affinity of
Sumatriptan is responsible for its vasoconstrictive
properties and that a selective 5-HT1D agonist might be
an effective treatment for migraine without cardiovas-
cular side effects.9,10 We have investigated the effects
of Sumatriptan and (S)-(-)-10 in a cat model of carotid
resistance (Figure 2).23 Sumatriptan (0.01-1.0 mg/kg,
iv, n ) 5) produced a dose-related increase in carotid
resistance in the absence of a change in mean arterial
blood pressure. The maximum increase in resistance
was 286% of pretreatment. In contrast, (S)-(-)-10

Communications to the Editor
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 13 2183
(8) Moskowitz, M. A.; Macfarlane, R. Neurovascular and molecular
mechanisms in migraine headaches. Cerebrovasc. Brain Metab.
Rev. 1993, 5, 159-177.
(0.01-1.0 mg/kg, iv, n ) 5) failed to alter either carotid
resistance or mean arterial blood pressure. These data
confirm numerous previous studies which demonstrate
vasoconstrictor properties of Sumatriptan. The fact that
(S)-(-)-10 has no carotid constrictor effect supports the
notion that the vasoconstrictor properties of Sumatrip-
tan are mediated by its 5-HT1B agonist properties.
Furthermore, these data suggest that (S)-(-)-10 may be
free of the cardiovascular complications observed with
Sumatriptan.
We have discovered a class of isochroman-6-carboxa-
mides which possesses excellent affinity and selectivity
for the 5-HT1D receptor. One such compound, (S)-(-)-
10 (PNU-109291), is active as an agonist in the guinea
pig hypothermia model and potently blocks neurogenic
dural inflammation in the guinea pig model of migraine.
Furthermore, we have demonstrated a lack of vasocon-
strictive effects of (S)-(-)-10 in a rabbit model of blood
flow. When tested in the rat, (S)-(-)-10 displayed
excellent oral bioavailability (70 ( 20%, n ) 4).24 The
debate whether the antimigraine efficacy of the current
class of triptan drugs is due to neurogenic or vascular
mechanisms remains unresolved. The discovery of truly
selective 5-HT1D agonists such as (S)-(-)-10 should
provide a valuable tool for probing therapeutic ap-
proaches to the treatment of migraine. We believe that
the isochroman-6-carboxamides described herein rep-
resent a significant structural breakthrough in the
search for safe, new therapies for migraine headache.
(9) Hamel, E. 5-HT1D Receptors: pharmacology and therapeutic
potential. Serotonin 1996, 1, 19-29.
(10) Hamel, E.; Gregoire, L.; Lau, B. 5-HT1 receptors mediating
contraction in bovine cerebral arteries: a model for human
cerebrovascular ‘5-HT1Dâ’ receptors. Eur. J . Pharmacol. 1993,
242, 75-82.
(11) Branchek, T.; Audia, J . E. Recent Advances in Migraine Therapy.
Annu. Rep. Med. Chem. 1997, 32, 1-10.
(12) TenBrink, R. E.; Bergh, C. L.; Duncan, J . N.; Harris, D. W.; Huff,
R. M.; Lahti, R. A.; Lawson, C. F.; Lutzke, B. S.; Martin, I. J .;
Rees, S. A.; Schlachter, S. K.; Sih, J . C.; Smith, M. W. (S)-(-)-
4-[4-[2-(Isochroman-1-yl)ethyl]-piperazin-1-yl]benzenesulfon-
amide, a Selective Dopamine D4 Antagonist. J . Med. Chem.
1996, 39, 2435-2437.
(13) Pregenzer, J . F.; Alberts, G. L.; Bock, J . H., Slightom, J . L.; Im,
W. B. Characterization of ligand binding properties of the 5-HT1D
receptors cloned from chimpanzee, gorilla, and rhesus monkey
in comparison with those from the human and guinea pig
receptors. Neurosci. Lett. 1997, 235, 117-120.
(14) The acid (S)-(-)-4 was assayed for optical purity by reduction
to the corresponding alcohol and analysis via HPLC (Chiracel
OJ column; see Supporting Information for details). The ester
(R)-(+)-3 was hydrolyzed to the corresponding acid (R)-(+)-4 and
assayed in an identical fashion. A single recrystallization of this
enriched chiral acid brought the optical purity to 99%.
(15) Davidsen, S. K.; May, P. D.; Summers, J . B. Di-tert-butyl
N-acylimido dicarbonates as isolable acylating agents: Mild
conversion of primary carboxamides to substituted amides. J .
Org. Chem. 1991, 56, 5482-85.
(16) We have also been able to demonstrate agonist properties for
(S)-(-)-10 in vitro using the GTPγ35S binding assay with cloned
gorilla receptors. Unfortunately, similar evaluation in the more
relevant guinea pig 5-HT1D receptors fails because these recep-
tors appear to be poorly coupled to intracellular G-proteins when
expressed in HEK 293 cells (Wha Bin Im, manuscript in
preparation).
(17) Skingle, M.; Higgins, G. A.; Feniuk, W. Stimulation of central
5-HT1D receptors causes hypothermia in the guinea pig. J .
Psychopharmacol. 1994, 8, 14-21.
(18) As observed with gorilla clones, (S)-(-)-10 shows excellent
selectivity for guinea pig 5-HT1D receptors over 5-HT1B (350-
400-fold).
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the preparation of all compounds and detailed
descriptions of the biological assays (22 pages). Ordering
information is given on any current masthead page.
(19) Moskowitz, M. A.; Cutrer, F. M.; Lee, W. S.; Waeber, C.;
Limmroth, V.; Yu, X. Drug Discovery in Migraine Circa 1995.
In Experimental Headache Models; Olesen, J ., Moskowitz, M.
A., Eds.; Lippincott-Raven Publishers: Philadelphia, 1995; pp
63-73.
(20) Markowitz, S.; Saito, K.; Moskowitz, M. A. Neurogenically
mediated leakage of plasma protein occurs from blood vessels
in dura mater but not brain. J . Neurosci. 1991, 7, 4129-4136.
(21) Buzzi, M. G.; Moskowitz, M. A. The antimigraine drug sumatrip-
tan (GR43175), selectively blocks neurogenic plasma extrava-
sation from blood vessels in dura mater. Br. J . Pharmacol. 1990,
99, 202-206.
(22) Similar results for (S)-(-)-10 have been obtained indepen-
dently: Cutrer, F. M.; Yu, X.-J .; Ayata, G.; Moskowitz, M. A.;
Waeber, C. Effects of PNU-109291, a selective 5-HT1D receptor
agonist, on electrically induced dural plasma extravasation and
capsaicin-evoked c-Fos immunoreactivity within trigeminal
nucleus caudalis. Neuropharmacology, submitted.
(23) Carotid artery resistance was determined by dividing carotid
blood flow (as measured by an implanted transit-time ultrasonic
probe) by mean arterial blood pressure. Complete details are
provided in the Supporting Information.
Refer en ces
(1) Olesen, J . Migraine and other headaches: the vascular mecha-
nisms; Raven Press: New York, 1991.
(2) Mushet, G. R.; Miller, D.; Clements, B.; Pait, G.; Gutterman, D.
L. Impact of Sumatriptan on Workplace Productivity, Nonwork
Activities, and Health-Related Quality of Life Among Hospital
Employees With Migraine. Headache 1996, 36, 137-143.
(3) de Lissovoy, G.; Lazarus, S. S. The economic cost of migraine.
Neurology 1994, 44 (suppl. 4), S56-S62.
(4) Osterhaus, J . T.; Gutterman, D. L.; Plachetka, J . R. Healthcare
resource and lost labour costs of migraine headache in the US.
Pharmacoeconomics 1992, 2, 67-76.
(5) Weinshank, R. L.; Zgombic, J . M.; Macchi, M. J .; Branchek, T.
A.; Hartig, P. R. Human serotonin 1D receptor is encoded by a
subfamily of two distinct genes: 5-HT1DR and 5-HT1Dâ. Proc. Natl.
Acad. Sci. U.S.A. 1992, 89, 3630-3634.
(6) The 5-HT1D and 5-HT1B receptors were formerly referred to as
5-HT1DR and 5-HT1Dâ, respectively. See: Hartig, P. R.; Hoyer,
D.; Humphrey, P. P. A.; Martini, G. R. Alignment of receptor
nomenclature with the human genome: classification of 5-HT1B
and 5-HT1D receptor subtypes. Trends Pharmacol. Sci. 1996, 17,
103-105.
(24) We thank Ms. Grace Wilson, Endocrine Pharmacology and
Metabolism, Pharmacia & Upjohn, for this determination.
(7) Rebeck, G. W.; Maynard, K. I.; Hyman, B. T.; Moskowitz, M. A.
Selective 5-HT1DR serotonin receptor gene expression in trigemi-
nal ganglia: Implications for antimigraine drug development.
Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 3666-3669.
J M980137O