Discovery of 4-[3-(trans-3-Dimethylaminocyclobutyl)-1H-indol-5-ylmethyl](4S)-oxazolidin- 2-one (4991W93), a 5HT1b/1d receptor partial agonist and a potent inhibitor of electrically induced plasma extravasation

Article abstract of DOI:10.1021/jm000956k

Utilizing a pharmacophoric model of binding of 3-(2-aminoethyl)indoles to 5HT_1B/1D receptors, we identified the 3-aminocyclobutyl group as a potential ethylamine isostere. A novel multidimensional chemometric approach was used to predict the intrinsic activity (degree of agonism) at the receptor. A qualitative model for pharmacokinetic properties was then used to guide the synthesis toward molecules likely to have oral bioavailability in humans. A novel synthetic route to 3-(3-dimethylaminocyclobutyl)indoles was developed. Analogues showed generally lower intrinsic activity at 5HT_1B/1D receptors than their ethylamine counterparts. 4-[3-(trans-3-Dimethylaminocyclobutyl)-1H-indol-5-ylmethyl]-(4S)-oxazolidin- 2-one (4991W93, 1) was identified as a partial agonist against 5HT_1B/1D receptors, with low intrinsic activity. This molecule also has significant activity against 5HT_1F receptors but is selective over other 5HT receptors. In addition this compound was found to be an exceptionally potent inhibitor of electrically induced plasma extravasation. Compound 1 may have utility in the treatment and prophylaxis of migraine.

Full text of DOI:10.1021/jm000956k

J . Med. Chem. 2001, 44, 681-693
681
Discover y of 4-[3-(tr a n s-3-Dim eth yla m in ocyclobu tyl)-1H-in d ol-5-ylm eth yl]-
(4S)-oxa zolid in -2-on e (4991W93), a 5HT1B/1D Recep tor P a r tia l Agon ist a n d a
P oten t In h ibitor of Electr ica lly In d u ced P la sm a Extr a va sa tion
Karamjit Singh J andu,† Vikki Barrett,† Michael Brockwell,| David Cambridge,† Duncan R. Farrant,†
Christopher Foster,† Heather Giles,† Robert C. Glen,‡ Alan P. Hill,† Heather Hobbs,† Andrew Honey,†
Graeme R. Martin,§ J ohn Salmon, Donna Smith,† Patrick Woollard,† and David L. Selwood*,
University Chemical Laboratory, University of Cambridge, Cambridge, U.K., GlaxoWellcome, Medicines Research Centre,
Gunnels Wood Road, Stevenage SG1 2NY, U.K., GlaxoWellcome, Chemical Development, Dartford, Kent, U.K.,
Roche Bioscience, Palo Alto, California, and The Wolfson Institute for Biomedical Research, University College London,
The Cruciform Building, Gower Street, London WC1E 6AU, U.K.
Received March 28, 2000
Utilizing a pharmacophoric model of binding of 3-(2-aminoethyl)indoles to 5HT1B/1D receptors,
we identified the 3-aminocyclobutyl group as a potential ethylamine isostere. A novel
multidimensional chemometric approach was used to predict the intrinsic activity (degree of
agonism) at the receptor. A qualitative model for pharmacokinetic properties was then used to
guide the synthesis toward molecules likely to have oral bioavailability in humans. A novel
synthetic route to 3-(3-dimethylaminocyclobutyl)indoles was developed. Analogues showed
generally lower intrinsic activity at 5HT1B/1D receptors than their ethylamine counterparts.
4-[3-(trans-3-Dimethylaminocyclobutyl)-1H-indol-5-ylmethyl]-(4S)-oxazolidin-2-one (4991W93,
1) was identified as a partial agonist against 5HT1B/1D receptors, with low intrinsic activity.
This molecule also has significant activity against 5HT1F receptors but is selective over other
5HT receptors. In addition this compound was found to be an exceptionally potent inhibitor of
electrically induced plasma extravasation. Compound 1 may have utility in the treatment and
prophylaxis of migraine.
In tr od u ction
We report the design, synthesis, and pharmacological
evaluation of 4-[3-(trans-3-dimethylaminocyclobutyl)-
1H-indol-5-ylmethyl]-(4S)-oxazolidin-2-one (4991W93,
1), a novel 3-(3-dimethylaminocyclobutyl)indole 5HT1B/1D
receptor partial agonist with potential for the treatment
and prophylaxis of migraine. This compound demon-
strates potent inhibition of electrically evoked plasma
extravasation. Migraine is a debilitating condition af-
flicting some 5-10% of the population with the majority
being women of young to middle age. In recent years
several 5HT1B/1D receptor1 agonists have been developed
by pharmaceutical companies as antimigraine agents.
The first of these was sumatriptan 2, which has been
followed to the market by a series of other “triptans”.2
These molecules such as naratriptan3 3 and zolmitrip-
tan4 4, are in general more lipophilic than sumatriptan
and may have improved clinical effectiveness as a result
of increased blood-brain barrier penetration and acti-
vation of 5HT1B/1D receptor or other receptors within the
CNS.5 All of these molecules were developed as se-
lective vasoconstrictors of the cranial vasculature, though
several workers, notably Moskowitz6 and Goadsby,7
have proposed additional direct neuronal actions for
* To whom correspondence should be addressed. Tel: +44 (0)20 7679
these molecules. In Moskowitz’s model the 5HT1B/1D
receptor agonists act on prejunctional 5HT hetero-
receptors to inhibit the release of vasoactive peptides,
such as substance P (SP) and calcitonin gene-related
peptide (CGRP). Recent studies on the localization of
6716. Fax: +44 (0)20 7679 6799. E-mail: d.selwood@ucl.ac.uk.
‡ University of Cambridge.
† GlaxoWellcome, Medicines Research Centre.
| GlaxoWellcome, Chemical Development.
§ Roche Bioscience.
University College London.
10.1021/jm000956k CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/24/2001

682 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
J andu et al.
Sch em e 1. General Synthetic Methods for
5HT1D and 5HT1B receptors have indicated that only
5HT1D receptors are located on human trigeminal
sensory neurons whereas only 5HT1B receptors were
detected on dural arteries.8 A refinement to the model
suggests that stimulation of C fibers causes release of
SP and leads to extravasation while stimulation of Aδ
fibers causes release of CGRP and leads to vasodilation.9
A further complication arose with the report of the po-
tent activity of CP-122,288 5: this novel pyrrolidinyl-
methylindole exhibits potent inhibition of electrically
stimulated plasma extravasation which is not consistent
with its 5HT1B/1D receptor activity. The existence of an
undiscovered 5HT receptor has been invoked to explain
the potency of inhibition.10 CP-122,288 itself is reported
to be ineffective in the treatment of migraine.11 Despite
a decade of research the precise involvement of the
various receptors in the mechanism of action of the
5HT1B/1D agonists is still unclear,12 though more infor-
mation is expected from the clinical testing of selective
5HT1D receptor agonists. In the case of the 5HT1F
receptor the selective agonist LY334370 has been pro-
posed to have a central action blocking noiciception.13
It may be that both 5HT1B and 5HT1D receptor activities
are necessary for clinical effectiveness. Clearly inhibi-
tion of extravasation alone does not account for the
action of sumatriptan as NK1 antagonists are ineffective
in the treatment of acute migraine.14
Cyclobutylindolesa
In our program to develop novel analogues with
potential antimigraine properties, we sought a novel
ethylamine isostere with reduced susceptibility to
monoamine oxidase (MAO), acceptable oral bioavail-
ability, and significantly lower intrinsic activity, i.e.,
very low partial agonism at 5HT1B/1D receptors. We also
needed to maintain activity at the receptors thought to
be important to the antimigraine action (5HT1B and
5HT1D). Selectivity over 5HT2A receptors is required or
the molecule should be functionally silent at this recep-
tor.
a
Reagents and conditions: (a) 200 °C; (b) KOH/H2O; (c) DPPA/
BnOH; (d) (Ph3P)3RhCl/CO/H2; (e) 1% aq H2SO4/EtOH; (f) Pd(OH)2/
C/HCOOH; (g) NaCNBH3/AcOH/CH2O. For structures of R, see
Table 1.
Ch em istr y
A combination of molecular modeling and pharmaco-
kinetic considerations identified the 3-cyclobutylamine
indoles as strong candidates for selective 5HT1B/1D
receptor ligands likely to have low intrinsic activity (vide
infra). Scheme 1 shows the general synthetic route used
for most of the analogues. In all cases the Fischer indole
synthesis was utilized to construct the indole core. The
use of the trans-cyclobutyl moiety as an ethylamine
isostere was dependent on achieving a viable synthetic
route. Initially a synthetic scheme utilizing a cycloalkyl-
ation of diethyl malonate with 1,3-dibromo-2-benzyloxy-
propane was attempted (data not shown).15 Although
this route did provide small amounts of the desired
analogues, it did not form the basis for a medicinal
chemistry program. The key novel chemistry concerned
the preparation of the cyclobutyl aldehyde precursor for
the Fischer indole. Using the known cyclobutane olefin16
6 we were able to conduct a hydroformylation reaction
using (Ph3P)3RhCl as catalyst in the presence of 1:1 CO/
H2 to give a 3:1 mixture of straight chain and branched
aldehydes which were separated by flash chromatogra-
phy. The predominant trans product 7 could be crystal-
lized from the mixture of straight chain aldehydes in
20% yield. A literature report17 indicates that increased
PPh3 and H2 concentrations increase the proportion of
normal/branched aldehyde. However, this is at the cost
of greatly increased amounts of reduced olefin side
products. The simple 1:1 ratio of reactants provided the
best compromise of yield versus normal/branched ratio.
With quantities of the desired aldehyde available we
could for the first time explore the structure-activity
relationships (SARs) for 5HT ligands containing this
novel ethylamine isostere. 5-Methylheterocyclic hydra-
zines 8 were prepared by methods reported previously18,
23 for 2-(indol-3-yl)ethylamine 5HT receptor ligands. The
CBz-protected amines 9 were cleaved using Pd(OH)2/C
in formic acid at reflux and the primary amines directly
dimethylated using NaCNBH3 and formaldehyde to give
the 5-substituted analogues 10. Scheme 2 details the
synthesis of a chain-extended and also a monomethyl-
ated analogue. The homologation of the aldehyde pro-
ceeded smoothly with Ph3PCH2OMe/NaH to give the
enol ether 11, which was utilized directly in the Fischer
indole synthesis and dimethylation strategy to give 12.
The formation of 13 involved conversion of the aldehyde
to the diethyl acetal using triethyl orthoformate/pTsOH.

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 683
Sch em e 2. Synthesis of Homocyclobutyl- and
Sch em e 4. Synthesis of Methylene-Linked 5-Amide and
Monomethylindole Analoguesa
5-Heterocyclic Cyclobutylindolesa
a
Reagents and conditions: (a) Ph3PCH2OMe/NaH; (b) 1% aq
H2SO4/EtOH; (c) Pd(OH)2/C/HCOOH; (d) NaCNBH3/AcOH/CH2O;
(e) (EtO)3CH/pTsOH; (f) NaH/MeI/DMF.
Sch em e 3. Synthesis of 5-Amide and 5-Heterocyclic
Cyclobutylindolesa
a
Reagents and conditions: (a) liq NH3/TBTU; (b) (MeO)2CMe-
NMe2/PhMe/120 °C; (c) H2NOH/70% aq AcOH/p-dioxane; (d)
Pd(OH)2/C/HCOOH; (e) NaCNBH3/AcOH/CH2O; (f) H2NNH2/70%
aq AcOH/120 °C.
were synthesized using the methodology of Lin.19 Reac-
tion of the primary amide 17 with N,N-dimethylacet-
amide dimethyl acetal gave the intermediate acyl-
amidine 18 which was reacted with hydroxylamine or
hydrazine to give the oxadiazole 19 and triazole 20,
respectively. In a similar fashion the ethylene-linked
oxadiazole 38 was prepared from the primary amide 14.
Finally the intermediate acid 16 (Scheme 3) was also
coupled with benzylamine to provide the benzylamide
21 or with ammonia to give the primary amide 22.
a
Reagents and conditions: (a) NaOH; (b) BnNH2/TBTU; (c)
(MeO)2CMeNMe2/PhMe/120 °C; (d) H2NOH/70% aq AcOH/p-
dioxane.
Design F ea tu r es
This product was alkylated using NaH/MeI then treated
as above to give the monomethylated analogue 13.
The factors considered in seeking to employ a novel
ethylamine isostere were (i) resistance to monoamine
oxidase attack and cleavage to the inactive indole acetic
acid metabolite, (ii) good oral absorption, and (iii) high
affinity and good selectivity over 5HT2A and other
monoamine receptors together with reduced intrinsic
activity at the 5HT1B/1D receptors represented by func-
tional activity in the rabbit saphenous vein. This last
feature was intended to counter the ability of triptan
molecules to constrict the coronary vasculature. This
pharmacological action contraindicates the use of such
agents in patients with coronary heart disease.
Schemes 3 and 4 show the strategy adopted for the
synthesis of various amides and heteroaryls at the
5-position. In these cases it was found to be more
convenient to modify the pre-formed indole rather than
synthesize the individual hydrazines. Thus, the primary
amide 14, formed from the 4-carboxamidophenylhydra-
zine, was hydrolyzed to the acid and coupled with
benzylamine using TBTU to give 15 (Scheme 3). The
5-acetic acid 16 (Scheme 4), prepared from the corre-
sponding hydrazine, was similarly converted to the
amide 17 using 0-(1H-benzotriazol-1-yl)-N,N,N1,N1-tet-
ramethyluronium tetrafluoroborate (TBTU) coupling. In
the case of the oxadiazole and triazole analogues these
(i) The initial metabolite of the related compound
zolmitriptan is the 3-(2-methylaminoethyl) derivative.

684 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
J andu et al.
F igu r e 1. Relationship between CMR and calculated log
DpH7.4 for a range of triptan and tryptamine analogue values.
Average plasma concentrations (at 0.5 and 2 h) after oral
administration (10 mg/kg) to rats were determined and
compounds with >25% of the level for sumatriptan are
indicated as triangles. Other triptans known to be bioavailable
in humans are indicated by single letters: A ) almotriptan,
E ) eletriptan, F ) frovitriptan, R ) rizatriptan, N )
naratriptan, S ) sumatriptan, and Z ) zolmitriptan. Gray
shading indicates a zone of physicochemical properties where
little oral absorption was observed.
F igu r e 2. Pharmacophore model used to overlay novel
analogues. Distance ranges are in Ångstroms, selectivity
volume in yellow.
route from our model. Clearly the proposed molecules
fit into our model (with the possible exception of
eletriptan) and have reasonable predicted absorption
properties. CNS penetration is expected to be low with
such hydrophilic molecules again with the exception of
eletriptan. Some increased CNS penetration has been
reported for zolmitriptan, however, and is proposed to
be beneficial for its antimigraine action.7
(iii) We utilized our model of the 5HT receptor23 and
a novel multidimensional chemometric approach for
predicting agonism as described below.
This conversion has been shown to be CYP1A2-medi-
ated.20 Cleavage to the monomethyl derivatives results
in an active metabolite. That the 3-(3-dimethylamino-
cyclobutyl) moiety cannot be cleaved to the indole acetic
acid is reasonable to expect, although the compound
could still be demethylated.
Com p u t er Mod elin g. We utilized a theoretical
pharmacophore model (Figure 2) for the 5HT1D receptor
vasoconstrictor receptor23 and extended the model to
enable a quantitative assessment of intrinsic activity
to be made. The receptor model utilizes distance con-
straints between the protonated amine, an aromatic
center (usually the indole), a hydrophobic volume, and
hydrogen-bonding sites to predict whether novel ligands
will have significant affinity. Given the spatial distribu-
tion of the common pharmacophore points, it is possible
to perform conformational analysis and least-squares
fitting of novel compounds to the desired pharmaco-
phoric points and hence suggest a binding mode for new
compounds. The relative ease with which these phar-
macophore points are accommodated within the volume
of existing ligands gives a qualitative measure of their
predicted affinity. The principal binding points were
deduced to be the protonated amine and the hydrogen-
bonding acceptor (e.g., the ketone group in the oxazoli-
dinone ring). In some cases, due to the conformational
and geometric constraints present in some molecular
structures, this mode of least-squares fitting results in
a displacement of the aromatic center of the indole (if
present) from the aromatic site occupied by, for example,
5HT. It was observed that in such cases the affinity of
a compound could be high while its intrinsic activity was
greatly reduced.18,24 This suggested to us a hypothesis
that the region of the molecule that could be necessary
for agonism in this series of compounds was contained
within the double-bond region of the indole fragment
(which, in antagonist structures that were fitted to the
pharmacophore model, was displaced from its usual
(ii) To establish that our planned analogues would
have a reasonable chance of oral absorption we utilized
the model previously employed by our group to predict
oral absorption for a range of 5HT1B/1D agonists.21 In
this model oral absorption for these relatively hydro-
philic molecules is thought to occur via a paracellular
route, i.e., crossing of the intestinal membrane through
aqueous pores. We observed that a relationship existed
between the molecular size of the molecule, as measured
by the calculated molar refractivity (CMR) and absorp-
tion. Larger hydrophilic molecules tended to have poor
absorption. Our findings were derived from a study,
which estimated the plasma levels in rats following oral
dosing in rats. The plot of calculated log D versus CMR
is shown in Figure 1 with compound 1 highlighted.
Average plasma concentrations (at 0.5 and 2 h) after
oral administration (10 mg/kg) to rats were determined,
and compounds with >25% of the level for sumatriptan
(0.2 µg/L) are indicated. By way of validation other
triptans known to be bioavailable in humans are
indicated by single letters: A ) almotriptan, E )
eletriptan, F ) frovitriptan, R ) rizatriptan, N )
naratriptan, S ) sumatriptan, and Z )zolmitriptan.
Gray shading indicates a zone of physicochemical
properties where little oral absorption was observed.
Eletriptan appears to be very different in terms of its
physicochemical properties, being significantly more
lipophilic with a higher CMR. The calculated log D is
higher than the reported value22 (∼1.17 versus -0.5),
the reason for this is not readily apparent. This molecule
would be expected to be absorbed by a transcellular

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 685
Ta ble 1. Calculated Molecular Properties for Compound 1
discrim.
property
mean
SEM
vector
sum
ESDL_1 -0.2369 -0.3114 0.04293
ESDL_2 -0.2812 -0.2502 0.01654
ESDL_3 -0.2215 -0.2652 0.01345
-0.23064
0.25885
0.050932
-0.16556
0.28664
0.041854
-0.03031
0.024389
-0.01277
-0.07944
result ) 0.06
ESDL_9 -0.2401 -0.2549 0.005488 -0.03237
PE_X
PE_Y
PE_Z
9.072
5.145
5.49
8.447
6.088
4.821
0.8981
0.9872
0.4312
0.13866
0.08989
-0.11227
0.0200
F igu r e 3. Ellipsoid axis and indole atom-numbering system.
position). To attempt to quantify these effects, an
analysis of the molecular properties of molecules having
a range of intrinsic activities (including silent com-
pounds) was undertaken (15 agonists, 14 antagonists,
and 10 test compounds were used to generate and test
the model).25 From these it was found that the molecular
shape of the 3-substituent, determined by the ellipsoid
axes (Figure 3), within which this substituent could be
contained, and the electrophilic superdelocalizability26
(computed from the AMI wave function, a measure of
the electron density available for interactions) of the
atoms in the 1-, 2-, 3-, and 9-positions of the indole ring
(Figure 3) could be incorporated into a QSAR model (a
hyperplane calculation constructed from these seven
parameters separated agonists and antagonists). An
extensive quantitative investigation using these param-
eters to separate agonists, partial agonists, and antago-
nists27 generated a quantitative model which would
predict the intrinsic acitivity of compounds closely
related to the training set. To validate the model 10
randomly selected agonists and antagonists were omit-
ted from the model generation and the model tested by
incorporating their computed parameters into the gen-
erated equation. All 10 were predicted correctly. Using
this model, we generated many new test structures and
quantitatively predicted their intrinsic activity. The
incorporation of a cyclobutane fragment replacing the
ethyl fragment of the side chain (as found in zolmitrip-
tan) resulted in a compound predicted (from the hyper-
plane calculation) to have very low intrinsic activity.
Subsequent investigations narrowed the region which
appears to be responsible for intrinsic activity down to
the double bond (π-electron density) region of the indole
and the steric parameters associated with the 3-sub-
stituent.
If suitable modifications are made to the remainder
of the molecule such that it can still accommodate the
principal pharmacophoric binding points in order to
maintain affinity, then an antagonist may be obtained
which has high affinity for 5HT1B/1D. In addition, oc-
cupation of the selectivity volume promotes selectivity
for the 5HT1B/1D receptor over the 5HT2A receptor.
Compounds 31 and 35 were predicted from the
pharmacophore model to have difficulty in accommodat-
ing the pharmacophoric binding points necessary for
high affinity. Compounds 23 and 1 were predicted to
better fit the pharmacophoric model and hence have
higher affinity. In addition, examination of their proper-
ties as antagonists using the 7-parameter model sug-
gested that 1 should be of very low intrinsic activity or
an antagonist. The values for the ellipsoid axes in 1 are
8.072, 6.145, and 5.490 Å and the values for the
electrophilic superdelocalizabilities are -0.2369, -0.2812,
-0.2215, and -0.2401 electrons. These values are
autoscaled using the mean and standard deviations
according to Fukui,26 which are shown in Table 1. The
constant
subsequent values are then inserted into the equa-
tion for the discrimination vector:26 P ) -0.23064-
(ESDL_1) + 0.25885(ESDL_2) + 0.50932(ESDL_3) -
0.03237(ESDL_9) + 0.13866(PE_1) + 0.089895(PE_2)
- 0.11227(PE_3) + 0.020071, which then yields the
prediction of agonist or antagonist and distance from
the zero hyperplane. It was seen that compounds near
to the zero hyperplane were very partial agonists, as 1
(0.04) was predicted to be.
Resu lts
Bin d in g a n d Va scu la r Activity. Compounds were
evaluated for their binding in CHO-K1 cells28 stably
transfected with human recombinant 5HT1B and 5HT1D
receptors. Functional activity was assessed in rings of
rabbit saphenous vein29 and their intrinsic activity
expressed relative to 5HT (1.0). The results are pre-
sented in Table 2.
A highly focused group of analogues were synthesized
restricting the substitution at the 5-position to generally
small heterocyclic substituents. Many of the compounds
were seen to be antagonists in the rabbit saphenous vein
(marked *). These included all the ethylene-linked
compounds such as 25-28. Not all the analogues
displayed low intrinsic activity; methylation of the (S)-
oxazolidinone as in 30 resulted in a molecule displaying
an intrinsic activity, i.e., maximum effect, of 0.48
(relative to 5HT). The (R)-oxazolidinones 31 and 34 also
displayed high intrinsic activity as did the methylene
sulfonamide 29. The heteroaryl, amide, and phenoxy
analogues were all antagonists in the rabbit saphenous
vein with the exception of the primary amide 14 which
had an intrinsic activity value of 0.52. Thus some
structural features which tend to full agonism in ethyl-
amine indole 5HT ligands, e.g., 5-carboxamidotryptamine,
can overcome the tendency to lower intrinsic activity
in the 3-(3-aminocyclobutyl)indoles. Only the trans
analogues displayed good activity with cis examples 24
and 35 showing reduced affinity. Extension of the 3-(3-
aminocyclobutyl) side chain with a methylene spacer,
i.e., in 12, broadly retained affinity but the compound
was an antagonist. In general the binding data on the
human 5HT1B and 5HT1D receptors paralleled the data
gathered in the rabbit saphenous vein. Few of the
molecules demonstrated any selectivity between the
receptors, though 1 did show 7-fold selectivity for 5HT1D
.
Alone among this group of molecules 1 demonstrated
very low intrinsic activity in the rabbit saphenous vein,
together with reasonable binding affinity. Evaluation
against other 5HT receptors is shown in Table 3; clearly
the molecule has significant activity against the related
calf caudate 5HT1D and against human 5HT1F stably
expressed in CHO-K1 cells but less potent activity

686 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Ta ble 2. 5HT1D, 5HT1B, and Rabbit Saphenous Vein Binding Data†
J andu et al.

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 687
Ta ble 2 (Continued)

688 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Ta ble 2 (Continued)
J andu et al.
a
b
† For the general structure n ) 0, except 12 where n ) 1. Human receptors expressed in CHO cell lines. Figures in parentheses for
RbSv are intrinsic activity, where 1.0 is the value for 5HT. c Data from ref 20. Ox )
d
e Values are means ( SEM, n ) 3 or more. Other values are means of duplicates.
Ta ble 3. Binding Data for 1 against 5HT Receptors
The ability of the compound to affect blood flow in the
guinea pig ear was also assessed by laser doppler and
gives a demonstration of the separation of the vaso-
constrictor effects of 1 versus the PPE inhibition. The
data for zolmitriptan is included in the figure for
comparison. The separation is at least 4 orders of
magnitude. Table 4 illustrates data for some other 5HT
agonists such as sumatriptan and CP-122,288. Supris-
ingly 1 is even more potent than CP-122,288 in this
assay, as in our hands CP-122,288 had an ED50 of 100
ng/kg. Sumatriptan, in common with most other trip-
tans, had an ED50 of approximately 10 µg/kg.
P h a r m a cok in etics. The qualitative model for oral
absorption had predicted that 1 would show similar
properties to other bioavailable triptans. Indeed when
the compound was evaluated in rats, it was shown to
have an oral bioavailability comparable to that of
zolmitriptan; see Figure 1. The detailed pharmaco-
5HT receptor
h5HT1B h5HT1D h5HT1A h5HT1F 5HT1D 5HT2A 5HT2C
6.95 7.80 5.45 6.44 7.05 >10 µM >10 µM
a
b
c
pIC50
a
b
Binding to calf caudate receptors. Binding in rabbit aorta,
no functional activity was noted. c Binding to guinea pig cortex in
the presence of [3H]mesulergine.
against other 5HT receptors such as 5HT2A (rabbit
aorta) and 5HT2C (guinea pig cortex). The compound
was functionally silent at the 5HT2A receptor. Com-
pound 1 was therefore selected for evaluation in rel-
evant animal models of migraine.
Activity a ga in st Electr ica lly Evok ed P la sm a
P r otein Extr a va sa tion (P P E) in Gu in ea P ig Du r a .
Figure 4 illustrates the activity of 1 against PPE in
guinea pigs. The compound was exceptionally potent
with an ED50 of approximately 10 ng/kg when given iv.

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 689
an antimigraine effect in humans. Other studies have
shown that 1 does not affect CGRP release in the cat at
non-5HT1B/1D receptor levels32 and also that only 5HT1B/1D
receptor doses inhibit trigeminocervical neurons.33 The
mechanism for the potent inhibition of PPE in rodents
remains unclear. It does seem however, as in the case
of CP-122,288, that the potent effects may not be
mediated by 5HT1B/1D receptors. In a multicenter,
randomized double-blind, placebo-controlled study, clini-
cal studies34 with 1 have shown that low doses of the
compound (0.125 and 2.5 mg/kg/iv) carefully designed
to be below the threshold for 5HT1B/1D receptor effects
were not effective in reducing the traditional 2-h acute
migraine headache. In these studies 117 patients were
randomized (placebo ) 42; 0.125 mg of 1 ) 40; 2.5 mg
of 1 ) 35). Traditional 2-h response rates (moderate or
severe to mild or none) were 45%, 50%, and 46% in the
placebo, 0.125 mg of 1, and 2.5 mg of 1 treatment
groups, respectively. For patients treating “early” (0-2
h), response rates were 53%, 42%, and 71%. Fewer
patients in the groups receiving 1 had severe headache
pain at 2 h (33%, 23%, and 14%). Recurrent headache
occurred in 53%, 58%, and 13% of the patients with a
median time to recurrence of 6.7, 7.8, and 11.9 h. The
compound was well-tolerated. There were no treatment
effects on blood pressure, ECGs, or clinical laboratory
assessments. Therefore although 1 was not effective in
the traditional 2-h HA response, the data regarding
early treatment and recurrence suggest that it may
modulate certain aspects of the migraine process at
doses which reflect the PPE activity. The profile of 1
would seem to make it well-suited among the triptan
family to be of use in the prophylaxis of migraine. The
arrival of generic sumatriptan to the market is likely
to have a big impact on the progression of second-
generation triptan molecules.
F igu r e 4. Effect of 1 on [125I]albumin extravasation into dura
following electrical stimulation of trigeminal ganglion (EC50
and vascular (ear) conductance in the guinea pig (mean (
SEM).
)
Ta ble 4. Inhibition of [125I]Albumin Extravasation into Dura
following Electrical Stimulation of Trigeminal Ganglion
compd
ED50, guinea pig
RbSva
1
∼10 ng/kg
∼100 ng/kg
∼12 µg/kg
∼10 µg/kg
5.71 (0.1)
6.33 (0.85)
6.8 (0.77)
6.6 (0.83)
5 (CP-122,288)
zolmitriptan
sumatriptan
a
pA50, figures in parentheses are intrinsic activity relative to
5HT.
Ta ble 5. Pharmacokinetic Parameters for 1 in Rats
â-phase
half-life
(h)
vol of
distrib
(L/kg)
plasma
clearance
(mL/min/kg)
oral
bioavail
(%)
compd
1
1.4
2.6
40
35
kinetic parameters are shown in Table 5. The half-life
of 1.5 h and oral bioavailability of 35% show that our
qualitative model was predictive, at least for the
tryptamine series. In humans 1 showed oral bioavail-
ability of 41%30 compared to the structurally related
zolmitriptan at 40-49% and sumatriptan at 14%.31
Although 1 was metabolized to the desmethyl metabo-
lite (compound 13 in Table 2) at 20-30% of the level of
1, no cleavage to an indole acetic acid metabolite such
as that observed for zolmitriptan was reported. Thus
the cyclobutylamino moiety was found to be resistant
to obvious metabolic degradation.
Exp er im en ta l Section
Ch em istr y. Melting points are uncorrected. Flash chroma-
tography was carried out by the method of Still35 except that
the columns were slurry packed. Merck silica art. no. 9385
was used. NMR spectra were run at 200 MHz unless otherwise
indicated. Chemical shifts are in ppm relative to TMS.
Coupling constants are in Hz.
tr a n s-N-(Ben zyloxyca r bon yl)cyclobu ta n a m in e-3-a cet-
a ld eh yd e (7). trans-N-(Benzyloxycarbonyl)-3-methylene-
cyclobutanamine (24.27 g, 112 mmol) (prepared as described
in EP-A-0366059) and tris(triphenylphosphine)rhodium chlo-
ride (400 mg 0.43 mmol) were heated to 70 °C in toluene (250
mL) under 100 atm of CO:H2, 1:1 mixture, for 18 h. The solvent
was evaporated under reduced pressure and the residue
chromatographed on silica eluting with 25% EtOAc in cyclo-
hexane. First product eluted as a mixture of cis and trans
branched chain aldehydes (7.01 g, 28.4 mmol). Second product
eluted as a mixture of cis and trans straight chain aldehydes
(18.65 g, 75.5 mmol). The trans straight chain isomer was
crystallized from ether (100 mL) as white needles (5.4 g, 22.0
mmol, 19.5%): mp 66-67 °C; MS (FAB) 248 (M + 1)+; 1H NMR
(CDCl3) δ 9.71 (t, 1H, J ) 1.3 Hz), 7.44-7.28 (m, 5H), 5.08 (s,
2H), 5.01 (broad d, J ) 6.2, 1H), 4.25 (m, 1H), 2.65 (m, 3H),
2.13 (m, 4H). Anal. (C14H17NO3) C, H, N.
Gen er a l Meth od A: F isch er In d ole Syn th esis a n d CBz
Dep r otection . Exa m p le: 4-[3-(tr a n s-3-Am in ocyclobu tyl)-
1H-in d ol-5-ylm eth yl]-(4S)-oxa zolid in -2-on e (23). The ap-
propriate hydrazine (6.3 g, 30 mmol) and trans-N-(benzyloxy-
carbonyl)cyclobutanamine-3-acetaldehyde (6.3 g, 25.5 mmol)
were heated to 80 °C for 7 h in 1% aqueous sulfuric acid (100
mL) and EtOH (150 mL). The reaction mixture was evaporated
Discu ssion
The most pleasing aspect of this project was the
success of the design process. This built on the knowl-
edge base in this area from previous studies.23 In
particular the hyperplane model for agonism suggested
that a novel ethylamine isostere might exhibit low
partial agonism against the vasoconstrictor 5HT1 recep-
tors. Consideration of the physicochemical parameters
necessary for good oral absorption predicted that small
substituents at the 5-position might maintain the oral
bioavailability of sumatriptan. Detailed investigation of
potential synthetic routes eventually provided a work-
able synthesis to the desired 3-(3-dimethylamino-
cyclobutyl)indoles.
The finding of potent PPE inhibitory activity similar
to that of CP-122,288 was intriguing, though PPE
activity of itself has not been shown to be predictive of

690 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
J andu et al.
in vacuo and brine added. Extraction with EtOAc gave the
crude CBz product 10.5 g (83%). The product was refluxed in
10% formic acid-MeOH with Pd(OH)2 on carbon (1 g) for 7 h.
The solvent was removed in vacuo and brine added. The
solution was then washed with EtOAc and then made basic
(pH 10-12) with dilute ammonium hydroxide solution. Ex-
traction with THF gave the crude product 2.0 g (28%) which
was purified by flash chromatography (2:14:84 NH3:EtOH:
CHCl3) to give the product 23. The compound was obtained
as a foam: 1H NMR (DMSO-d6) δ 10.71 (bs, 1H), 7.72 (bs, 1H),
7.27 (s, 1H), 7.25 (d, J ) 8.2, 1H),7.16 (dd, J ) 1.2, 0.5, 1H),
6.93 (dd, J ) 8.2, 1.2, 1H), 4.22 (m, 1H), 4.04 (m, 1H), 4.00
(m, 1H), 3.8 (partly obscured m, 2H), 2.88 (dd, J ) 13.5, 4.6,
1H), 2.77 (dd, J ) 13.5, 6.9, 1H) 2.40 (bm, 2H), 2.24 (bm, 2H);
MS (FAB) m/z 286 (M + 1). Anal. (C16H19N3O2‚1.26H2O‚
0.23C2H6O) C, H, N.
1.32 (s, 6H); MS (EI) m/z 368, 297, 194, 156, 143; HRMS for
C21H28N4O2 calcd 368.22123, found 368.22119.
2-{2-[3-(tr a n s-3-Dim eth yla m in ocyclobu tyl)-1H-in d ol-5-
1
yl]eth yl}isoin d ole-1,3-d ion e (28): mp 52-53 °C; H NMR
(CDCl3) δ 7.87 (bs, 1H), 7.82 (m, 2H), 7.68 (m, 2H), 7.34 (dt, J
) 1.5, 0.5, 1H), 7.29 (dd, J ) 8.3, 0.5, 1H), 7.13 (dd, J ) 8.3,
1.5, 1H), 7.01 (dd, J ) 2.1, 1.3, 1H), 3.96 (ct, J ) 7.7, 5.8, 2H),
3.56 (ttt, J ) 9.1, 4.6, 1.2, 1H), 3.08 (ct, J ) 7.9, 5.9, 2H), 2.86
(pd, J ) 7.2, 0.9, 1H), 2.39 (m, 2H), 2.22 (m, 2H), 2.17 (s, 6H);
MS (FAB) m/z 388 (M + 1). Anal. (C24H25N3O2‚0.4H2O‚
0.3CHCl3) C, H, N.
N-Met h yl-[3-(tr a n s-3-d im et h yla m in ocyclob u t yl)-1H -
in d ol-5-yl]m eth a n esu lfon a m id e (29): using the hydrazine
prepared as reported;36 mp 220-222 °C; 1H NMR (DMSO-d6)
δ 10.82 (bs, 1H), 7.44 (m, 1H), 7.31 (d, J ) 8.1, 1H), 7.22 (dd,
J ) 1.9, 0.9, 1H), 7.08 (dd, J ) 8.2, 1.5, 1H), 6.73 (q, J ) 4.9,
1H), 4.32 (s, 2H), 3.51 (ttt, J ) 9.0, 4.5, 1.0, 1H), 2.81 (pd, J )
6.8, 0.6, 1H), 2.53 (d, J ) 4.9, 1H), 2.30 (m, 2H), 2.21 (m, 2H),
2.06 (s, 6H); MS (FAB) m/z 322 (M + 1). Anal. (C16H22N3O2S‚
0.089CHCl3) C, H, N.
4-[3-(tr a n s-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-
ylm eth yl]-3-m eth yl-(4S)-oxa zolid in -2-on e (30): mp 46-
47 °C; 1H NMR (CDCl3) δ 8.06 (bs, 1H), 7.40 (m, 1H), 7.32
(dd, J ) 7.2, 0.7, 1H), 7.08 (dd, J ) 2.3, 1.2, 1 H), 6.97 (dd, J
) 7.2, 1.5, 1H), 4.17 (t, J ) 8.1, 1H), 4.05 (m, 1H), 3.94 (m,
1H), 3.64 (ttt, J ) 9.2, 4.1, 1.0, 1H), 3.24 (dd, J ) 13.6, 4.6,
1H), 2.92 (s, 3H), 2.91 (m, 1H), 2.77 (dd, J ) 13.6, 8.6, 1H),
2.47 (m, 2H), 2.30 (m, 2H), 2.19 (s, 6H); MS (FAB) m/z 328 (M
+ 1). Anal. (C19H25N3O2‚0.13CHCl3) C, H, N.
Compounds prepared using general method A:
tr a n s-3-{2-[3-(3-Am in ocyclobu tyl)-1H-in d ol-5-yl]eth yl}-
5,5-d im eth ylim id a zolid in e-2,4-d ion e (27): mp 85-87 °C;
1H NMR (DMSO-d6) δ 10.63 (bs, 1H), 8.11 (bs, 1H), 7.22 (dd,
J ) 8.2, 0.7, 1H), 7.14 (m, 1H), 7.12 (dd, J ) 2.2, 1.0, 1H),
6.87 (dd, J ) 8.4, 1.6, 1H), 3.57 (ct, J ) 7.5, 2H), 3.5 (m, 1H),
2.90 (m, 1H), 2.88 (bt, J ) 7.5, 2H), 2.31 (m, 2H), 2.10 (m,
2H), 1.16 (s, 6H); MS (EI) m/z 297, 181, 169. Anal. (C19H24N4O2‚
0.09CHCl3) C, H, N.
5-P h en oxy-3-(tr a n s-3-a m in ocyclobu tyl)-1H-in d ole (35):
mp 164-166 °C; 1H NMR (DMSO-d6) δ 10.90 (bs, 1H), 7.38
(dd, J ) 8.6, 0.4, 1H), 7.31 (m, 1H), 7.30 (m, 1H), 7.09 (d, J )
2.3, 1H), 7.02 (tt, J ) 7.3, 1.0, 1H), 6.90 (m, 1H), 6.83 (dd, J )
8.6, 2.3, 1H), 3.74 (m, 2H), 2.42 (m, 4H); MS (EI) m/z 278, 235,
158. Anal. (C18H18N2O‚C6H6O4) C, H, N.
4-[3-(tr a n s-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-
1
ylm eth yl]-(4R)-oxa zolid in -2-on e (31): mp 163-165 °C; H
NMR (DMSO-d6) δ 10.69 (bs, 1H), 7.72 (bs, 1H), 7.27(m, 1H),
7.25 (d, J ) 8.2, 1H), 7.17 (dd, J ) 1.6, 0.6, 1H), 6.93 (dd, J )
8.3, 1.2, 1H), 4.22 (m, 1H), 4.03 (m, 1H), 3.99 (m, 1H), 3.51
(ttt, J ) 9.2, 4.4, 1.0, 1H), 2.88 (m, 1H), 2.78 (m, 1H), 2.75 (m,
1H), 2.31 (m, 2H), 2.19 (m, 2H), 2.06 (s, 6H); MS (FAB) m/z
314 (M + 1). Anal. (C18H23N3O2‚0.66H2O‚0.33C4H8O2) C, H,
N.
Gen er a l Meth od B: Red u ctive Am in a tion . Exa m p le:
tr a n s-4-[3-(3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-yl-
m eth yl]-(4S)-1,3-oxa zolid in -2-on e (1). Formaldehyde (0.18
mL, 2.22 mmol) in MeOH (5 mL) was added to the product
method A (250 mg, 0.88 mmol), AcOH (0.26 mL, 4.55 mmol)
and sodium cyanoborohydride (70 mg, 1.17 mmol) in MeOH
(15 mL) and stirred at room temperature under a nitrogen
atmosphere overnight. Water was added and the mixture
washed with EtOAc. The aqueous phase was then adjusted to
pH 10 with K2CO3 and saturated with NaCl. Extraction with
EtOAc gave a sticky gum which was purified by flash chro-
matography (1:9:90 NH3:MeOH:CHCl3) to give the product 1
as an off white powder 137 mg (44%): mp 159-160 °C; 1H
NMR (DMSO-d6) δ 10.69 (bs, 1H), 7.73 (bs, 1H), 7.27 (partly
obscured m, 1H), 7.25 (d, J ) 8.2, 1H), 7.17 (dd, J ) 2.1, 0.8,
1H), 6.93 (dd, J ) 8.3, 1.4, 1H), 4.22 (m, 1H), 4.04 (m, 1H),
3.99 (m, 1H), 3.51 (m, 1H), 2.88 (m, 1H), 2.78 (m, 1H), 2.76
(m, 1H), 2.29 (m, 2H), 2.19 (m, 2H), 2.06 (s, 1H); MS (FAB)
m/z 314 (M + 1). Anal. (C18H23N3O2) C, H, N.
3-{2-[3-(tr a n s-3-Am in ocyclobu tyl)-1H-in dol-5-yl]m eth yl}-
5,5-d im eth ylim id a zolid in e-2,4-d ion e (32): mp 70-72 °C;
1H NMR (CDCl3) δ 7.94 (bs, 1H), 7.55 (dt, J ) 1.5, 0.8, 1H),
7.29 (dd, J ) 8.3, 0.8, 1H), 7.24 (dd, J ) 8.3, 1.5, 1H), 7.05
(dd, J ) 2.4, 1.2, 1H), 5.2 (bs, 1H), 4.75 (s, 1H), 3.63 (ttt, J )
9.4, 4.5, 1.2, 1H), 2.90 (pd, J ) 7.4, 1.0, 1H), 2.30 (m, 2H),
2.44 (m, 2H), 2.18 (s, 6H), 1.41 (s, 6H); MS (FAB) m/z 355 (M
+ 1). Anal. (C20H26N4O2‚0.1H2O‚0.2CHCl3) C, H, N.
3-[3-(tr a n s-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-
ylm eth yl]im id a zolid in e-2,4-d ion e (33): mp 106-109 °C
1
dec; H NMR (DMSO-d6) δ 10.76 (bs, 1H), 8.02 (bs, 1H), 7.37
(dt, J ) 1.6, 0.6, 1H), 7.27 (dd, J ) 8.3, 0.6, 1H), 7.20 (dd, J )
2.2, 1.1, 1H), 7.02 (dd, J ) 8.2, 1.6, 1H), 4.56 (s, 2H) 3.93 (s,
2H), 3.49 (ttt, J ) 9.5, 4.4, 1.1, 1H), 2.80 (pd, J ) 7.1, 1.0,
1H), 2.29 (m, 2H), 2.17 (m, 2H), 2.06 (s, 6H); MS (EI) m/z 326,
295, 226, 211; HRMS for C18H22N2O2 calcd 326.17428, found
326.17371.
Compounds prepared using general methods A and B:
5-Ca r boxa m id o-3-(tr a n s-3-d im eth yla m in ocyclobu tyl)-
1
1H-in d ole (14): mp 93-95 °C; H NMR (DMSO-d6) δ 11.00
(bs, 1H), 8.07 (dt, J ) 1.6, 0.6, 1H), 7.80 (bs, 1H), 7.65 (dd, J
) 8.5, 1.6, 1H), 7.33 (dd, J ) 8.6, 0.6, 1H), 7.28 (dd, J ) 2.2,
1.1, 1H), 7.02 (bs, 1H), 3.57 (ttt, J ) 9.0, 4.4, 1.1, 1H), 2.81
(pd, J ) 7.0, 0.7, 1H), 2.35 (m, 2H), 2.21 (m, 2H), 2.07 (s, 6H);
MS (FAB) m/z 258 (M + 1). Anal. (C15H19N3O‚0.34H2O‚
0.1AcOEt) C, H, N.
4-[3-(tr a n s-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-
ylm eth yl]-3-m eth yl-(4R)-oxa zolid in -2-on e (34): obtained
as a foam; 1H NMR (DMSO-d6) δ 10.72 (bs, 1H), 7.29 (m, 1H),
7.27 (dd, J ) 8.2, 0.5, 1H), 7.18 (dd, J ) 2.3, 1.0, 1H), 6.94
(dd, J ) 8.2, 1.6, 1H), 4.14 (m, 1H), 4.00 (m, 1H), 3.96 (m, 1H),
3.51 (ttt, J ) 9.0, 1.1, 4.4, 1H), 3.17 (dd, J ) 13.5, 3.7, 1H),
3.05 (m, 1H), 2.80 (s, 3H), 2.77 (m, 1H), 2.30 (m, 2H), 2.18 (m,
2H), 2.06 (s, 6H); MS (EI) m/z 327, 312, 283, 256, 100. Anal.
(C19H25N3O2‚0.75H2O) C, H, N.
5-P h en oxy-3-(tr a n s-3-d im et h yla m in ocyclob u t yl)-1H -
in d ole (36): prepared from 35 above; mp 189-190 °C; 1H NMR
(DMSO-d6) δ 11.06 (d, J ) 1.6, 1H), 7.41 (m, 2H), 7.31 (m,
1H), 7.13 (d, J ) 2.2, 1H), 7.02 (tt, J ) 7.3, 1.1, 1H), 6.89 (m,
1H), 6.85 (dd, J ) 8.6, 2.2, 1H), 3.86 (pd, J ) 1.0, 8.0, 1H),
3.59 (m, 1H), 2.69 (s, 6H), 2.67 (m, 2H), 2.40 (m, 2H); MS (EI)
m/z 306, 291,262, 235, 158. Anal. (C20H22N2O‚HBr‚0.1C3H8O‚
0.1H2O) C, H, N.
tr a n s-3-{2-[3-(3-Dim eth yla m in ocyclobu tyl)-1H-in d ol-5-
yl]eth yl}im id a zolid in e-2,4-d ion e (25): mp 197-199 °C; 1H
NMR (CDCl3) δ 7.96 (bs, 1H), 7.37 (dt, J ) 1.5, 0.6, 1H), 7.29
(dd, J ) 8.3, 0.6, 1H), 7.10 (dd, J ) 8.5, 1.6, 1H), 7.04 (dd, J )
2.2, 1.2, 1H), 5.51 (bs, 1H), 3.92 (d, J ) 17.7, 1H), 3.85 (d, J )
17.7, 1H), 3.79 (ct, J ) 8.0, 6.0, 2H), 3.62 (ttt, J ) 9.2, 4.0, 1.2,
1H), 3.02 (ct, J ) 8.2, 6.1, 2H), 2.89 (pd, J ) 7.2, 1.1, 1H) 2.44
(m, 2H), 2.30 (m, 2H), 2.18 (s, 6H); MS (FAB) m/z 341 (M +
1). Anal. (C19H24N4O2‚0.22H2O) C, H, N.
5,5-Dim e t h yl-3-{2-[3-(t r a n s-3-d im e t h yla m in ocyclo-
bu tyl)-1H-in d ol-5-yl]eth yl}im id a zolid in e-2,4-d ion e (26):
mp 159-160 °C; 1H NMR (CDCl3) δ 7.89 (bs, 1H), 7.35 (s, 1H),
7.32 (d, J ) 8.3, 1H), 7.09 (dd, J ) 8.3, 1.7, 1H), 7.03 (dd, J )
1.7, 1.3, 1H), 5.13 (bs, 1H), 3.78 (t, J ) 7.6, 2H), 3.62 (m, 3H),
2.90 (q, J ) 7.6, 1H), 2.46 (m, 2H), 2.31 (m, 2H), 2.20 (s, 6H),
Gen er a l Meth od C: F or Syn th esis of th e cis-Am in o-
cyclobu tyl An a logu es. The mixture of cis- and trans-CBz

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 691
aldehydes produced in the preparation of 7 was utilized as in
methods A and B above. The cis dimethylated products were
separated by HPLC using SiO2, 2:8:90 880 NH3:MeOH:CHCl3.
was purified by flash chromatography (1:10:89 0.88 NH3:
MeOH:CHCl3) (175 mg, 32%). The crude material was treated
by the methods described above to give 15 as a foam: 1H NMR
(DMSO-d6) δ 10.96 (bs, 1H), 8.78 (t, J ) 6.0, 1H), 8.09 (m, 1H),
7.67 (dd, J ) 8.5, 1.7, 1H), 7.35 (m, 1H), 7.32 (m, 2H), 7.28
(dd, J ) 2.2, 1.0, 1H), 7.22 (tt, J ) 6.2, 3.2,), 4.5 (d, J ) 6.0,
2H), 3.58 (ttt, J ) 9.3, 4.5, 1.0, 1H), 2.84 (pd, J ) 7.3, 1.0,
1H), 2.34 (m, 2H), 2.20 (m, 2H), 2.08 (s, 6H); MS (FAB) m/z
348 (M + 1). Anal. (C22H25N3O‚0.71H2O) C, H, N.
3-(tr a n s-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-yl-
a ceta m id e (22). To 4-hydrazinophenylacetic HCl in 1% aque-
ous sulfuric acid (75 mL) was added trans-N-(benzyloxycar-
bonyl)cyclobutanamine-3-acetaldehyde (5 g, 20 mmol). The
mixture was heated to 90 °C for 7 h. The semisolid formed
was filtered and washed with 1% H2SO4 and water. The solid
was then taken up into EtOAc and washed with water. The
organic phase was dried (MgSO4). The product was obtained
as a sticky solid: yield 6.5 g (74%).
To a solution of the crude 3-[trans-N-(benzyloxycarbonyl)-
cyclobutanamine]-1H-indol-5-ylacetic acid (0.5 g, 1.38 mmol)
in DMF (5 mL) was added TBTU (0.44 g, 1.38 mmol) and Et3N
(0.21 mL, 1.52 mmol). Anhydrous ammonia was then bubbled
through the solution for 1 h at room temperature with cooling
in ice as necessary. The mixture was stirred overnight at room
temperature. The solvent was evaporated under vacuum. The
residue was chromatographed on silica using 1:10:89 880 NH3:
MeOH:CHCl3. Yield: 0.28 g (56%). This was treated by the
general method B outlined above to give 22: mp 57-59 °C;
1H NMR (DMSO-d6) δ 10.76 (bs, 1H), 7.30 (dt, J ) 1.5, 0.9,
1H), 7.29 (bs, 1H), 7.24 (dd, J ) 8.1, 0.5, 1H), 7.17 (dd, J )
2.5, 1.2, 1H), 6.98 (dd, J ) 8.3, 1.6, 1H), 6.74 (bs, 1H), 3.50
(ttt, J ) 8.8, 4.4, 1.1, 1H), 3.39 (s, 2H), 2.81 (pd, J ) 7.0, 0.8,
1H), 2.30 (m, 2H), 2.19 (m, 2H), 2.07 (s, 6H); MS (FAB) m/z
272 (M + 1); HRMS for C16H21N3O calcd 271.16846, found
271.16713.
N -Be n zyl-3-(t r a n s-3-d im e t h yla m in ocyclob u t yl)-1H -
in d ol-5-yla ceta m id e (21). To a solution of 3-[trans-N-
(benzyloxycarbonyl)-3-aminocyclobutyl]indol-5-ylacetic acid (1
g, 2.76 mmol) in DMF (5 mL) was added TBTU (0.97 g 3.04
mmol), Et3N (0.42 mL 3.04 mmol) and benzylamine (0.33 mL
3.04 mmol). The mixture was stirred overnight at room
temperature. The solvent was evaporated under vacuum and
the residue purified by flash chromatography (1:5:94 NH3:
MeOH:CHCl3) to give a pale yellow oil 0.81 g (80%). The crude
product of this step was then treated by the general methods
A and B outlined above to provide 21 as a light yellow oil: 1H
NMR (CDCl3) δ 8.08 (bs, 1H), 7.40 (dt, J ) 1.5, 0.6, 1H), 7.32
(dd, J ) 8.3, 0.5, 1H), 7.25 (m, 1H), 7.22 (m, 1H), 7.15 (m, 1H),
7.06 (m, 2H), 5.79 (bt, J ) 5.6, 1H), 4.39 (d, J ) 5.8, 2H), 4.32
(s, 2H), 3.63 (ttt, J ) 9.2, 4.0, 1.1, 1H), 2.92 (pd, J ) 7.3, 1.1,
1H), 2.44 (m, 2H), 2.28 (m, 2H), 2.19 (s, 6H); MS (FAB) m/z
362 (M + 1); HRMS for C23H27N3O calcd 361.21541, found
361.2154.
Prepared using method C were:
cis-4-[3-(3-Am in ocyclobu tyl)-1H-in dol-5-ylm eth yl]-(4S)-
oxa zolid in -2-on e a ceta te (24): mp 180-181 °C; 1H NMR
(DMSO-d6) δ 10.70 (bs, 1H), 7.79 (bs, 1H), 7.41 (s, 1H), 7.24
(d, J ) 8.2, 1H), 7.08 (dd, J ) 1.5, 0.7, 1H), 6.92 (dd, J ) 8.2,
1.5, 1H), 4.23 (tm, 1H), 4.04 (m, 3H), 3.19 (bm, 1H), 2.89 (dd,
J ) 13.5, 4.5, 1H), 2.78 (dd, J ) 6.8, 13.5, 1H), 2.61 (bm, 2H),
1.97 (bm, 2H); MS (FAB) m/z 286 (M + 1); HRMS m/z calcd
for C16H19N3O2 285.1488, found 285.1477.
4-[3-(cis-3-Dim et h yla m in ocyclob u t yl)-1H -in d ol-5-yl-
m eth yl]-(4S)-oxa zolid in -2-on e (35): obtained as a hygro-
scopic solid; 1H NMR (CDCl3) δ 8.27 (bs, 1H), 7.44 (dt, J )
1.7, 0.7, 1H), 7.29 (dd, J ) 8.3, 0.6, 1H), 7.03 (dd, J ) 2.4, 0.7,
1H), 6.95 (dd, J ) 8.3, 1.6, 1H), 5.28 (bs, 1H), 4.47 (m, 1H),
4.18 (m, 1H), 4.14 (om, 1H), 3.31 (tdd, J ) 10.2, 7.5, 1.0, 1H),
3.00 (dd, J ) 13.5, 5.7, 1H), 2.90 (dd, J ) 13.5, 7.8, 1H), 2.75
(m, 1H), 2.62 (m, 2H), 2.25 (s, 6H), 2.13 (m, 2H); MS (FAB)
m/z 314 (M + 1); HRMS C18H23N3O2 calcd 313.17903, found
313.17683.
4-[3-(tr a n s-3-Dim eth yla m in ocyclobu t-1-ylm eth yl)-1H-
in d ol-5-ylm eth yl]-(4S)-oxa zolid in -2-on e (12): mp 64-65
1
°C; H NMR (DMSO-d6) δ 10.66 (bs, 1H), 7.73 (bs 1H), 7.34
(dt, J ) 1.6, 0.7, 1H), 7.24 (dd, J ) 8.3, 0.5, 1H), 7.05 (d, J )
2.3, 1H), 6.92 (dd, J ) 8.3, 1.5, 1H), 4.22 (m, 1H), 4.04 (m,
1H), 4.00 (m, 1H), 2.89 (dd, J ) 13.7, 4.5, 1H), 2.80 (d, J )
8.0, 2H), 2.77 (dd, J ) 13.7, 6.8, 1H), 2.75 (m, 1H), 2.42 (m,
1H), 2.00 (s, 6H), 1.91 (m, 2H), 1.78 (m, 2H); MS (EI) m/z 325,
312, 299, 241, 212. Anal. (C19H25N3O2‚0.42H2)O C, H, N.
4-[3-(tr a n s-3-Met h yla m in ocyclob u t yl)-1H -in d ol-5-yl-
m eth yl]-(4S)-oxa zolid in -2-on e (13). The aldehyde 7 (1 g, 4
mmol), triethyl orthoformate (1.35 mL, 8 mmol) and pTsOH
(100 mg) were refluxed in EtOH for 3 h. The reaction mixture
was evaporated in vacuo and purified by flash chromatography
(10:90 EtOAc/cyclohexanone) to give a colorless oil (1.2 g, 94%),
which was used crude for the next stage. The trans-N-
(benzyloxycarbonyl)cyclobutanamine-3-acetaldehyde diethyl
acetal (1.2 g, 3.7 mmol) in dry DMF (10 mL) was added
dropwise to a cold suspension of sodium hydride (60% in oil)
(165 mg, 4.1 mmol) in dry DMF (10 mL). After the addition
was complete the mixture was allowed to stirred at 10 °C for
30 min, MeI (0.23 mL, 3.7 mmol) in dry DMF (5 mL) was then
added dropwise. The mixture was then allowed to warm to
room temperature and stirred for 2 h. The reaction was then
poured onto ice and extracted with ether. Flash chromatog-
raphy (10:90 EtOAc:cyclohexane) gave the product as a color-
less oil (1 g, 83%). The product of this step was then treated
by the general methods A and B above to provide 13 as a
foam: 1H NMR (CDCl3) δ 8.11 (bs, 1H), 7.32 (dt, J ) 1.5, 0.7,
1H), 7.31 (dd, J ) 8.3, 0.6, 1H), 7.07 (dd, J ) 2.0, 1.0, 1H),
6.98 (dd, J ) 8.3, 1.6, 1H), 5.16 (bs, 1H), 4.47 (t, J ) 8.0, 1H),
4.19 (dd, J ) 8.0, 5.6, 1H), 4.12 (td, J ) 8.0, 5.6, 1H), 3.73 (ttt,
J ) 9.0, 4.0, 1.2, 1H), 3.48 (s, 1H), 3.42 (m, 1H), 2.98 (m, 1H),
2.91 (m, 1H), 2.43 (m, 2H), 2.40 (s, 3H), 2.28 (m, 2H); MS (EI)
m/z 325, 312, 299, 241, 212. Anal. (C19H25N3O2‚0.42H2O) C,
H, N.
5-N-Ben zylca r b oxa m id o-3-(tr a n s-3-d im et h yla m in o-
cyclob u t yl)-1H -in d ole (15). 5-Carboxamido-3-(trans-3-di-
methylaminocyclobutyl)-1H-indole (14; 0.4 g, 1.6 mmol), pre-
pared using the methods described above, was refluxed in 10
M NaOH solution (15 mL) and MeOH (10 mL) for 7 h. The
resulting solution was cooled in ice and neutralized with dil
HCl. This was then evaporated to dryness in vacuo and MeOH
added. The NaCl was filtered off and the solution evaporated
to give the crude product. The crude 3-(trans-3-dimethylamino-
cyclobutyl)-1H-indole-5-carboxylic acid (0.4 g, 1.6 mmol), O-(1H-
benzotriazol-1-yl)-N,N,N1,N1-tetramethyluronium tetrafluoro-
borate (TBTU) (0.57 g, 1.8 mmol), benzylamine (0.18 mL, 1.6
mmol) and Et3N (0.25, 1.8 mmol) were stirred at room
temperature in dry DMF (15 mL) for 5 h. The reaction was
quenched with water and extracted with EtOAc. This was
dried (MgSO4) and evaporated to give the crude product which
3-Meth yl-5-[3-(tr a n s-3-d im eth yla m in ocyclobu tyl)-1H-
in d ol-5-ylm eth yl]-1,2,4-oxa d ia zole (38). To a solution con-
taining 3-(trans-N-benzyloxycarbonyl-3-aminocyclobutyl)indol-
5-ylacetamide (4.68 g 13 mmol) in toluene (10 mL) was added
N,N-dimethylacetamide dimethyl acetal (40 mL). The mixture
was then heated at 120 °C for 2 h under Dean & Stark condi-
tions. The solution became very dark. The solvent was evap-
orated under vacuum to give N-[(dimethylamino)ethanylidene]-
3-[trans-3-(benzyloxycarbonylamino)cyclobutyl]-1H-indol-5-
ylacetamide (5.7 g crude), which was used directly for subse-
quent operations.
To a mixture containing hydroxylamine HCl (10 mg 1.6
mmol), 5 N NaOH (0.2 mL), p-dioxane (3 mL) and 70% AcOH
(10 mL) was added the amidine (500 mg 1.6 mmol) and the
mixture heated to 120 °C for 1.5 h. The solvent was evaporated
under vacuum and the residue was purified by flash chroma-
tography (40:60 EtOAc:cyclohexane) to give the product, 150
mg (31%), as a foam. The product of this step was then treated
by the general methods A and B above to provide 38 as a
foam: 1H NMR (CDCl3) δ 7.97 (bs, 1H), 7.42 (dt, J ) 1.6, 0.7,
1H), 7.28 (dd, J ) 8.3, 0.7, 1H), 7.11 (dd, J ) 8.3, 1.7, 1H),

692 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
J andu et al.
7.03 (dd, J ) 2.2, 1.1, 1H), 3.71 (s, 2H), 3.68 (s, 3H), 3.63 (ttt,
J ) 9.3, 4.0, 1.2, 1H), 2.89 (pd, 1.2, 7.3, 1H), 2.44 (m, 2H),
2.29 (m, 2H), 2.17 (s, 6H); MS (FAB) m/z 310 (M + 1).
mm laterally and 2.0 mm posteriorly to bregma for electrode
placement. Each animal then received a single bolus dose of 1
(0.01, 0.03 or 0.1 µg/kg) or vehicle (0.9% w/v saline) via the
right jugular vein and, after 5 min, 50 µCi/kg [125I]bovine
serum albumin (BSA) (50 µCi/mL saline) was administered
via the same cannula. Subsequently, paired nonconcentric
bipolar electrodes (Rhodes NE-200, Clark Electromedical
Instruments, U.K.) were lowered into the trigeminal ganglia,
under stereotactic control, to a depth of 10.5 mm from the dura
mater. The left or right side was arbitrarily designated for
stimulation. Five minutes after the administration of [125I]BSA
the test side was electrically stimulated consecutively for a
further 5 min using paired rectangular pulses of opposite
polarity with 5 pulses/s of 5-ms duration at an intensity of 1.2
mA (Stimulator, model S88; Stimulus Isolation Unit, model
SIU5A; Constant Current Unit, model CCU1A; Grass Instru-
ments). Ipsilateral contraction of the muscles of mastication
during stimulation was assumed to indicate correct placement
of the electrodes. Immediately following the stimulation period
the thorax was opened, the descending aorta clamped and the
right atrium incised for drainage. Saline was perfused via the
left ventricle at a pressure of 100 mm Hg for 3 min in order to
flush all blood from the intravascular compartment. The skull
was removed and the dura mater both overlying and underly-
ing the brain was carefully dissected from both the stimulated
and nonstimulated sides. The large dural sinuses, the region
of dura through which the electrodes penetrated and the dura
covering the trigeminal ganglia were excluded from analysis.
In some experiments a mantle of frontoparietal cortex from
each side of the brain was also dissected. After rinsing in
saline, the samples were dried on tissue paper, weighed and
counted for [125I] for 20 min in a Wallac Gamma Counter
(Wizard 1470, Wallac). Extravasation of [125I]albumin into the
dura was calculated as cpm/mg wet weight for both the
stimulated and the unstimulated sides from each animal and
the data (mean ( SEM) are expressed as the ratio between
each side.
3-Meth yl-5-[3-(tr a n s-3-d im eth yla m in ocyclobu tyl)-1H-
in d ol-5-ylm eth yl]-1,2,4-tr ia zole (20). To a solution of the
amidine, prepared as above (3 g, 6.99 mmol) in 70% aqueous
AcOH (100 mL), was added hydrazine hydrate (10.9 mL, 8.39
mmol). The mixture was stirred at 90 °C for 5 h. The mixture
was concentrated under reduced pressure and diluted with
water (100 mL). The aqueous phase was extracted with EtOAc
(2 × 50 mL). The organic phase was dried (MgSO4) and
evaporated. The residue was chromatographed using 1:10:89
NH3:MeOH:CHCl3 to give an oil, 200 mg. The product of this
step was then treated by general methods A and B to provide
the product 20 as a foam: 1H NMR (DMSO-d6) δ 13.12 (bs,
1H), 10.59 (bs, 1H), 7.31 (m, 1H), 7.23 (dd, J ) 7.9, 05, 1H),
7.15 (dd, J ) 2.0, 0.7, 1H), 6.96 (dd, J ) 8.2, 1.6, 1H), 3.97 (m,
2H), 3.49 (ttt, J ) 9.0, 4.6, 1.0, 1H), 2.82 (pd, J ) 1.0, 7.2,
1H), 2.30 (m, 2H), 2.17 (m, 2H), 2.07 (s, 6H); MS (FAB) 310
(M + 1). Anal. (C18H23N5‚1.0H2O) C, H, N.
3-Meth yl-5-[3-(tr a n s-3-d im eth yla m in ocyclobu tyl)-1H-
in d ol-5-yl]-1,2,4-oxa d ia zole (19). The primary amide 14, was
treated as for 38 above to give 19 as a foam. The product of
this step was then treated by the general methods A and B
above to provide 19 as a foam: 1H NMR (DMSO-d6) δ 11.26
(bs, 1H), 8.20 (dd, J ) 1.5, 0.5, 1H), 7.79 (dd, J ) 8.4, 1.6, 1H),
7.53 (dd, J ) 8.5, 0.5, 1H), 7.39 (dd, J ) 2.2, 1.2, 1H), 3.61
(ttt, J ) 9.3, 4.3, 0.9, 1H), 2.87 (pd, J ) 0.9, 8.0, 1H), 2.39 (s,
3H), 2.38 (m, 2H), 2.22 (m, 2H), 2.10 (s, 6H); MS (EI) m/z 296,
252, 225, 170. Anal. (C17H20N4O‚0.57H2O‚0.04C4H8O2) C, H,
N. 1H NMR (CDCl3) δ 7.97 (bs, 1H), 7.42 (dt, J ) 1.6, 0.7, 1H),
7.28 (dd, J ) 8.3, 0.7, 1H), 7.11 (dd, J ) 8.3, 1.7, 1H), 7.03
(dd, J ) 2.2, 1.1, 1H), 3.71 (s, 2H), 3.68 (s, 3H), 3.63 (ttt, J )
9.3, 4.0, 1.2, 1H), 2.89 (pd, 1.2, 7.3, 1H), 2.44 (m, 2H), 2.29 (m,
2H), 2.17 (s, 6H); MS (FAB) m/z 310 (M + 1).
Biologica l Assa ys. Compound binding in CHO-K1 cells
stably transfected with human recombinant 5HT1B, 5HT1D
,
At least one animal was treated with vehicle (0.9% w/v
saline) on each day of experimentation to control for variability
in albumin leakage noted in the perfused tissues from week
to week.
5HT1A and 5HT1F receptors and also the 5HT2C receptor in
guinea pig cortex was performed as described by Martin.28
Rabbit saphenous vein functional assays and the 5HT2A assays
in rabbit aorta were as detailed by Martin.29
3. Ea r Blood F low /Con d u cta n ce Mea su r em en t. In a
separate group of guinea pigs the effect of 1 on vascular
conductance in the ear was determined. Animals were anes-
thetized and surgically prepared as described above. Blood flow
was continuously measured in the right ear using a single laser
Doppler flow probe attached to an MBF3 laser Doppler blood
flow monitor (Moor Instruments, U.K.). The flow probe and
flow monitor were set to operate at a frequency of 10 Hz and
a time constant of 1 s. Data were output to an IBM PC and
displayed and analyzed using Moorsoft 4.1 software. Blood flow
was recorded as red blood cell flux (the product of mean
velocity and concentration of moving red blood cells) deter-
mined as a percentage of a standardized signal obtained in a
particulate control medium. Effects on vascular conductance
were calculated by dividing blood flow by blood pressure
changes in each animal studied. Both measurements were
made at the time that maximum changes in flow were
observed.
P r otocol for Bin d in g Assa y Usin g Ca lf Ca u d a te Mem -
br a n es. Membranes were prepared from homogenates of calf
caudate nucleus. Competition binding studies were performed
with 3.5 nM [3H]5HT (ca. 25 Ci/mL) in the presence of 15 µM
mesulergine and 15 µM 8-OH-DPAT, using 5 mg wet weight/
mL of membranes in a total volume of 1 mL/tube. Binding data
was fitted to a four-parameter logistic function to obtain
estimates of pIC50. These were converted to pKi values using
the Cheng-Prussof equation.
P r otocols for in Vivo Stu d ies. 1. Su r gica l P r ep a r a tion .
Male Dunkin Hartley guinea pigs (200-250 g; David Hall
Supplies) were housed under diurnal lighting conditions and
allowed food and water ad libitum. Each animal was anes-
thetized using pentabarbitone (50 mg/kg ip) and the rectal
temperature was maintained at 37 °C with a homeothermic
blanket control unit (Harvard model 50-7061, Harvard Ap-
paratus Ltd.). The trachea was cannulated and the animals
were artificially ventilated with room air throughout the
experiment using a mechanical respirator (Harvard model 50-
1718, Harvard Apparatus Ltd.). A tidal volume of 10 mL/kg
and a frequency of 80 breaths/min was used. A 2FG polythene
cannula (Portex Ltd.) was inserted into the right external
jugular vein. In the animals used for flow studies a 3FG
polythene cannula (Portex Ltd.) was inserted into the right
carotid artery to monitor blood pressure via a Statham P23
pressure transducer attached to a model 7D Polygraph (Grass
Instruments).
Su p p or tin g In for m a tion Ava ila ble: Complete NMR
assignments for the target compounds; table of combustion
analyses; and data used to construct the pharmacokinetic
absorption model, i.e., the tryptamine structures, calculated
log D values, and measured plasma levels. This informa-
tion is available free of charge via the Internet at http://
pubs.acs.org.
Refer en ces
2. Electr ica l Tr igem in a l Ga n glion Stim u la tion a n d
P la sm a P r otein Extr a va sa tion Mea su r em en t. Animals
were placed in a stereotaxic frame (model 900, Kopf Instru-
ments, Tujunga, CA), with the incisor bar set at -4 mm from
the horizontal, and the skull exposed by a midline incision.
Symmetrical burr holes (1-2 mm diameter) were drilled 3.2
(1) Earlier literature referrs to 5HT1DR (5HT1D) and 5HT1Dâ (5HT1B).
(2) Yevich, J . P.; Yocca, F. D. Second generation antimigraine
5HT1B/1D agonists: structure activity relationship and pre-
clinical pharmacological considerations. Curr. Med. Chem. 1997,
4, 295-312.
(3) Salonen, R. Naratriptan. Int. J . Clin. Pract. 1999, 53, 552-556.

A 5HT1B/ 1D Receptor Partial Agonist
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 693
(4) Proietti-Cecchini, A.; Afra, J .; Schoenen, J . Intensity dependence
of the cortical auditory evoked potentials as a surrogate marker
of central nervous system serotonin transmission in man:
demonstration of a central effect for the 5HT1B/1D agonis
zolmitriptan (311C90, Zomig). Cephalalgia 1997, 17, 849-854.
(5) Hoskin, K. L.; Goadsby, P. J . Comparison of more and less
(21) Hill, A. P.; Hyde, R. M.; Robertson, A. D.; Woollard, P. M.; Glen,
R. C.; Martin, G. R. Oral delivery of 5-HT1D receptor agonists:
towards the discovery of 311C90, a novel anti-migraine agent.
Poster at 10th Migraine Trust International Symposium, Lon-
don, U.K., Sept 5-8, 1994.
(22) Fox, A. W. Comparitive tolerablility of oral 5-HT1B/1D agonists.
Headache 2000, 40, 521-527.
lipophilic serotonin (5HT1B/1D) agonists in
a model of tri-
geminovascular nociception in cat. Exp. Neurol. 1998, 150, 45-
51.
(23) Glen, R. C.; Martin, G. R.; Hill, A. P.; Hyde, R. M.; Woollard, P.
M.; Salmon, J . A.; Buckingham, J .; Robertson, A. D. Computer-
aided design and synthesis of 5-substituted tryptamines and
their pharmacology at the 5-HT1D receptor: discovery of
compounds with potential anti-migraine properties. J . Med.
Chem. 1995, 38, 3566-3580.
(6) Moskowitz, M. A. Neurogenic versus vascular mechanisms of
sumatriptan and ergot alkaloids in migraine. Trends Pharmacol.
Sci. 1992, 13, 307-311.
(7) Goadsby, P. J .; Hoskin, K. L. Serotonin inhibits trigeminal
nucleus activity evoked by craniovascular stimulation through
a 5HT1B/1D receptor: a central action in migraine. Neurology
1998, 43, 711-718.
(8) Longmore, J .; Shaw, D.; Smith, D.; Hopkins, R,; McAllister, G.;
Pickard, J . D.; Sirinathsinghji, D. J . S.; Butler, A. J .; Hill, R. G.
Differential distribution of 5HT1D and 5HT1B immunoreactivity
within the human trigemino-cerebrovascular system: implica-
tions for the discovery of new antimigraine drugs. Cephalalgia
1997, 17, 833-842.
(24) Moloney, G. P.; Martin, G. R.; Mathews, N.; Hobbs, H.; Dods-
worth, S.; Sang, Y. S.; Knight, C.; Maxwell, M.; Glen, R. C.
Synthesis and pharmacological profile of
a series of 2,5-
substituted-N,N-dimethyltryptamine derivatives as novel an-
tagonists for the vascular 5-HT1B-like receptor. J . Chem. Soc.,
Perkin Trans. 1, 1999, 2713.
(25) Arthur 81, version 3; B&B Associates, P.O. Box 85505, Seattle,
WA 98105.
(9) Shepherd, S. L.; Williamson, D. J .; Beer, M. S.; Hill, R. G.;
Hargreaves, R. J . Differential effects on neurogenic dural plasma
extravasation and vasodilatiation in anaethetized rats. Neuro-
pharmacology 1997, 36, 525-523.
(26) Fukui, K.; Yonezawa, T.; Nagata, C. Special Report. Bull. Chem.
Soc. J pn. 1954, 27, 423.
(27) Glen, R. C.; Moloney, G. P. Analysis of 5HT receptors and the
design of therapeutic analogues. In Rational Molecular Design
in Drug Research; Alfred Benzon Symposium 42; Liljefors, T.,
J orgensen, F. S., Krogsgaard-Larsen, P., Eds.; Munksgaard:
Copenhagen, 1998; pp 33-46.
(28) Martin, G. R.; Robertson, A. D.; Maclennan, S. J .; Prentice, D.
J .; Barrett, V. J .; Buckingham, J .; Honey, A. C.; Giles, H.;
Moncada, S. Receptor specificity and trigemino-vascular inhibi-
tory actions of a novel 5-HT1B/1D receptor partial agonist,
311C90 (zolmitriptan). Br. J . Pharmacol. 1997, 121, 157-164.
(29) Martin, G. R.; MacLennan, S. J . Analysis of the 5-HT receptor
in rabbit saphenous vein exemplified the problems of using
exclusion criteria for receptor classification. Naunyn-Schmiede-
berg’s Arch Pharmacol. 1990, 342, 111-119.
(30) McColm, J . A.; Peck, R. W.; Koch, K. M.; J ulious, S.; Collyer, S.
M.; Hulbert, R. A study to investigate the tolerability, pharma-
cokinetics and effects on blood pressure, heart rate and sub-
stance P-induced wheal and flare response of single, oral doses
of 4991W93 in healthy volunteers. Cephalalgia 1999, 19, 406-
406.
(31) Deleu, D.; Hanssens, Y. Current and emerging second generation
triptans in acute migraine therapy: a comparative review. J .
Clin. Pharmacol. 2000, 40 687-700.
(32) Knight, Y. E.; Connor, H. E.; Edvinsson, L.; Goadsby, P. J . Only
5HT1B/1D agonist doses of 4991W93 inhibit CGRP release in the
cat. Cephalalgia 1999, 19, 401.
(33) Storer, R. J .; Connor, H. E.; Goadsby, P. J . Microinotophoretic
application of 4991W93 inhibits trigeminocervical neurons.
Cephalalgia 1999, 19, 314.
(34) Earl, N. L.; McDonald, S. A.; Lowy, M. T. Efficacy and tolerability
of the neurogenic inflammation inhibitor, 4991W93, in the acute
treatment of migraine. Cephalalgia 1999, 19, 357.
(35) Still, W. C.; Kahn, M.; Mitra, A. Rapid chromatographic tech-
nique for preparative separations with moderate resolution J .
Org. Chem. 1978, 43, 2923.
(36) Remuzon, P.; Dussy, C.; J acquet, J . P.; Soumeillant and Bouzard,
D. 4-Hydroxy-2-methyl-3(2H)-isothiazolone-1,1-dioxide as a pro-
tecting group of 4-(N-methylaminosulphonyl)methylphenyl-
hydrazone during Fischer indole synthesis. Tetrahedron Lett.
1995, 36, 6227.
(10) Goadsby, P. J . Serotonin receptors and the acute attack of
migraine. Clin. Neurosci. 1998, 5, 18-23.
(11) Roon, K.; Diener, H. C.; Ellis, P.; Hettiarachchi, J .; Poole, P.;
Christiansen, I.; Ferrari, M. D.; Olesen, J . CP-122,288 blocks
neurogenic inflammation, but is not effective in aborting mi-
graine attacks: results of two controlled clinical trials. Ceph-
alalgia 1997, 17, 245.
(12) For an excellent review of the recent development of migraine
therapy, see Goadsby, P. J . Advances in the pharmacotherapy
of migraine. Drugs R&D 1999, 2, 361-374.
(13) Shepheard, S.; Edvinsson, L.; Cumberbatch, M.; Williamson, D.;
Mason, G.; Webb, J .; Boyce, S.; Hill, R.; Hargreaves, R. Possible
antimigraine mechanisms of action of the 5HT1F receptor agonist
LY334370. Cephalalgia 1999, 19, 851-858.
(14) Connor, H. E.; Bertin, L.; Gillies, S.; Beattie, D. T.; Ward, P.
Clinical evaluation of a novel, potent, CNS penetrating NK1
receptor antagonist in the acute treatment of migraine. Ceph-
alalgia 1998, 18, p392, 9.2.
(15) Castro, P. J . L.; Guiblin, A. R.; Matassa, V. G.; Reeve, A. J .;
Sternfeld, F.; Street, L. J . Preparation of imidazole, triazole, and
tetrazole derivatives as 5-HT1-like receptor agonists. PCT Patent
Application WO 9402477, 1994.
(16) Norbeck, D. W.; Plattner, J . J .; Rosen, T. J .; Pariza, R. J .; Sowin,
T. J .; Garmaise, D. L.; H., S. M. Preparation of nucleoside
cyclobutane analogues as antiviral and antitumor agents. Eu-
ropean Patent Application EP-A-0366059, 1990.
(17) Pittman, C. U.; Honnick, W. D. Rhodium-catalysed hydro-
formylation of allyl alcohol. A potential route to 1,4-butanediol.
J . Org. Chem. 1980, 45, 2132-2139.
(18) Moloney, G. P.; Martin, G. R.; Mathews, N.; Milne, A.; Hobbs,
H.; Dodsworth, S.; Sang, P. Y.; Knight, C.; Williams, M.;
Maxwell, M.; Glen, R. C. Synthesis and serotonergic activity of
substituted 2,N-benzylcarboxamido-5-(2-ethyl-1-dioxoimidazo-
lidinyl)-N,N-dimethyltryptamine derivatives: novel antagonists
for the vascular 5-HT(1B)-like receptor. J . Med. Chem. 1999,
42, 2504-2526.
(19) Lin, Y.-I.; Lang, S. A., J r.; Lovell, M. F.; Perkinson, N. A. New
synthesis of 1,2,4-triazoles and 1,2,4-oxadiazoles. J . Org. Chem.
1979, 44, 4160-4164.
(20) Wild, M. J .; McKillop, D.; Butters, C. J . Determination of the
human cytochrome P450 isoforms involved in the metabolism
of zolmitriptan. Xenobiotica 1999, 29, 847-857.
J M000956K