Novel secoergoline derivatives inhibit both GABA and glutamate uptake in rat brain homogenates: Synthesis, in vitro pharmacology, and modeling

Article abstract of DOI:10.1021/jm040809c

Three of twelve secoergoline derivatives (Z ethyl 4-[(ethoxycarbonylmethyl) methylamino]-2-methyl-3-phenylpent-2-enoate, 8; ethyl 1,6-dimethyl-3-oxo-5- phenyl-1,2,3,6-tetrahydropyridine-2-carboxylate, 9; Z methyl 4-[(methoxycarbonylmethyl)methylamino)-2-m

Full text of DOI:10.1021/jm040809c

5620
J. Med. Chem. 2004, 47, 5620-5629
Articles
Novel Secoergoline Derivatives Inhibit Both GABA and Glutamate Uptake in
Rat Brain Homogenates: Synthesis, in Vitro Pharmacology, and Modeling
La´szlo´ He´ja,† Ilona Kova´cs,† EÄ va Sza´rics,† Ma´ria Incze,‡ Eszter Temesva´rine´-Major,‡ Ga´bor Do¨rnyei,‡
Ma´ria Peredy-Kajta´r,§ Eszter Ga´cs-Baitz,§ Csaba Sza´ntay,‡ and Julianna Kardos*,†
Department of Neurochemistry, Department of Natural Organic Compounds, and NMR Laboratory, Chemical Research Center,
Hungarian Academy of Sciences, H-1025 Pusztaszeri u´t 59-67, Organic Chemistry Institute, Budapest University of Technology
and Economics, H-1111 Gelle´rt te´r 4, Budapest, Hungary
Received March 12, 2004
Three of twelve secoergoline derivatives (Z ethyl 4-[(ethoxycarbonylmethyl)methylamino]-2-
methyl-3-phenylpent-2-enoate, 8; ethyl 1,6-dimethyl-3-oxo-5-phenyl-1,2,3,6-tetrahydropyridine-
2-carboxylate, 9; Z methyl 4-[(methoxycarbonylmethyl)methylamino)-2-methyl-3-phenylpent-
2-enoate, 11), containing bioisosteric sequences of GABA and Glu, inhibited both GABA and
Glu transport through cerebrocortical membranes specifically. Compounds 8, 9, and 11 appeared
to be equipotent inhibitors of GABA and Glu transport with IC50 values between 270 and 1100
µM, whereas derivatives 1-7, 10, and 12 were without effects. In the presence of GABA and
Glu transport-specific nontransportable inhibitors, inhibition of GABA and Glu transport by
8, 9, and 11 proceeded in two phases. The two phases of inhibition were characterized by IC50
values between 4 and 180 nM and 360-1020 µM and different selectivity sequences. These
findings may indicate the existence of some mechanism possibly mediated by a previously
unrecognized GABA-Glu transporter. Derivatives with the cis, but not the trans configuration
of bulky ester groups (8 vs 7 and 11 vs 12) showed significant inhibitory effect (IC50 values of
270 µM vs .1000 µM and 1100 µM vs .1000 µM on GABA transport, respectively). The cis-
trans selectivity can be explained by docking these secoergolines in a three-dimensional model
of the second and third transmembrane helices of GABA transporter type 1.
Introduction
GABA hypothesis of epilepsy.7 Accordingly, impairment
of GABAergic pathways is of deciding importance for
the pathogenesis of several types of epilepsy. Thus, the
GABA approach to antiepileptic therapy can be directed
at the development of novel GABA- and benzodiazepine-
mimetic derivatives targeting GABAA receptors as well
as GABA uptake inhibitors as potential drugs for
epilepsy.7-9 It is of note in this context that the residues
which form a salt bridge between the gamma and beta
subunits of GABAA receptors have been implicated in
epilepsy-inducing mutations.10 Despite advances, our
understanding of how neurotransmitter transporters
work is limited by a lack of three-dimensional structural
information.11
In the absence of high-resolution structural data,
computer-aided predictions of the three-dimensional
structure of GABA transporter type 1 (GAT1), the
prototypical member of a group of proteins with a GABA
transport role in the vertebrate central nervous system,
together with the localization of amino acid residues
crucial for GABA recognition12,13 can be explored. Based
on these studies and pharmacological data for a series
of known GABA transport inhibitors, and using the
transmembrane helix 3 and transmembrane helix 2
(TM3 and TM2) regions of GAT1, a model of the three-
dimensional structure of the ligand binding crevice in
the GAT1 transporter has been developed. We report
here the synthesis and pharmacological characterization
Epilepsy, one of the most common neurological dis-
order, which affects ∼1-2% of the population world-
wide,1 may be controlled by both the facilitation of
γ-aminobutyric acid (GABA)-mediated (GABAergic) in-
hibitory pathways2 and the inhibition of glutamic acid-
mediated (Gluergic) excitatory neurotransmission.3 Al-
though there is evidence that Gluergic excitatory
neurotransmission may contribute to the pathophysi-
ology of epilepsy, development of novel N-methyl-D-
aspartate (NMDA) receptor antagonists as antiepileptic
drugs has been terminated because of their low ef-
ficiency and wide-range side effects in clinical trials.4
Development of R-amino-3-hydroxy-5-methyl-4-isox-
azolepropionic acid (AMPA) receptor antagonists may
have a role for the control of symptoms. Indeed, sup-
pression of seizures in a low-[Mg2+] ion model of
experimental epilepsy was observed with the use of a
quinazolone-alkyl-carboxylic acid derivative,5 known to
block (S)(+)AMPA-induced transmembrane Ca2+ ion
influx.6
Among the rational strategies, the most widely used
way to develop new antiepileptic drugs is based on the
* To whom correspondence should be addressed. Tel.: +36 1 325
9101; fax: +36 1 325 7554; e-mail: jkardos@chemres.hu.
† Department of Neurochemistry.
‡ Department of Natural Organic Compounds.
§ NMR Laboratory.
10.1021/jm040809c CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/05/2004

Novel Secoergoline Derivatives
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5621
Scheme 1. Synthetic Routes to 1-4
Scheme 2. Synthetic Routes to 5-12
of new secoergoline derivatives. The results of docking
to the model have enabled us to rationalize the observed
efficiency and selectivity of 7 and 8.
piophenone according to Bennett et al.14 and House et
al.,15
was treated with diethyl R-formylsuccinate ac-
cording to a method described by Stoll et al.16 The
double bond of the geometric isomers of enamines ii
thus obtained was reduced, and a parallel reductive
methylation with formaldehyde and NaBH3CN fur-
nished diester iii. Stobbe-type ring closure of the latter
to piperidine derivative 1 was performed with super
base BuLi+tBuOK in tetrahydrofuran (THF). 1 is a
diastereomic mixture of iv and v, which could be
separated either by fractionated crystallization or by
column chromatography. Water elimination from 1 to
Results
Chemistry. All the chiral compounds are racemates;
however, they are represented by drawing only one of
the enantiomers in the schemes. The tested compounds
(indicated by Arabic numbers) were synthesized starting
from 1-phenyl-2-aminopropan-1-one hydrochloride (i) as
shown in Schemes 1 and 2. To prepare compounds 1-4
(Scheme 1), R-aminoketone (i), obtained from pro-

5622 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
He´ja et al.
Table 1. IC50 Values of Secoergoline Derivatives on Specific
[3H]GABA and [3H]DASP Transport through Cerebrocortical
Membranes at 30 °C. Number of Experiments: N ) 2 to N ) 6
configuration of bulky ester groups (8 vs 7 and 11 vs
12) showed significant inhibitory effect. The secoergoline
derivatives 8, 9, and 11 gave IC50 values of 270 ( 43,
680 ( 87 and 1100 ( 128 µM on GABA uptake,
respectively. Surprisingly, 8, 9, and 11 showed similar
inhibitory potential on Glu transport with IC50 values
of 420 ( 57, 660 ( 61, and 1100 ( 105 µM, respectively
(Table 1). Secoergoline derivatives found to be ineffective
on GABA uptake did not show any significant effect on
Glu transport, either (Table 1).
IC50 (µM)
compound
[3H]GABA transport
[3H]DASP transport
1
2
.1000
.1000
.1000
.1000
3
.1000
.1000
4
.1000
.1000
5
.1000
.1000
6
.1000
.1000
7
.1000
.1000
Measurements on the effects of 8, 9, and 11 on GABA
uptake were performed in the presence of 100 µM
1,2,5,6-tetrahydro-1-[2-[[(diphenylmethylene)amino]oxy]-
ethyl]-3-pyridine carboxylic acid hydrochloride (NNC-
711) (Table 2), a potent, specific, nontransportable
inhibitor of GAT1.24 In this case, the dose-response
curves showed two phases, indicating the presence of
two separate GABA transport processes (Figure 1). In
one of these phases 8, 9, and 11 gave IC50 values
comparable, but not equal to those obtained in the
absence of NNC-711 (380 ( 13, 360 ( 8, and 491 ( 14
µM, respectively), whereas IC50 values obtained for 8,
9, and 11 in the other phase were found to be 180 ( 7,
3.7 ( 0.6, and 44 ( 2 nM, respectively. In addition,
effects of 8, 9, and 11 on Glu transport were determined
in the presence of 100 µM DL-threo-â-benzyloxyaspartic
acid (DL-TBOA) (Table 2), a potent, nontransportable
inhibitor of Glu transporter types 1-5.25 As seen with
GABA transport by the presence of NNC-711, the dose-
response curves also showed two phases suggesting that
8, 9, and 11 act on two separate Glu transport processes
(Figure 1). In one of the phases 8, 9, and 11 gave IC50
values similar to those obtained in the absence of DL-
TBOA (385 ( 16, 591 ( 24, and 1019 ( 81 µM,
respectively), whereas the other phase was character-
ized by IC50 values of 34 ( 2, 9.0 ( 0.2, and 11 ( 0.9
nM, respectively.
Specificity of secoergoline derivatives as uptake in-
hibitors was addressed in measurements of GABAA,
GABAB, AMPA, NMDA, and (2S,3S,4R)-carboxy-4-(1-
methylethenyl)-3-pyrrolidineacetic acid (kainate) recep-
tor binding. Neither of secoergoline derivatives in 10 µM
(GABAA), 100 µM (GABAB), and 30 µM (AMPA, NMDA,
kainate) inhibited the binding of labeled specific ligands
to these receptors significantly (data not shown).
Molecular Modeling. The One-TM Model. To find
amino acid residues possibly involved in binding, the
nonflexible GABA transport inhibitor, guvacine, was
docked in the model of the TM3 region of GAT1.
H-Bonding interactions were found between functional
groups of residues, Tyr-140(OH), Asn-137(CdO), Ser-
133(OH), and guvacine (CdO, OH, NH, respectively).
In this way, all polar amino acids of GAT1 TM3
participated in H-bonding interactions except Tyr-139,
which is faced perpendicularly to Tyr-140. It is to note
in this respect, that Tyr-139 was found unnecessary for
ligand binding.13 To verify binding calculations obtained
by docking guvacine in the model of GAT1 TM3, known
inhibitors, such as GABA, (R)(-)-N-(4,4-di-(3-methyl-
thien-2-yl)but-3-enyl)nipecotic acid hydrochloride (ti-
agabine), and the (R)(-) and (S)(+) isomers of 3-piper-
idinecarboxylic acid ((R) and (S)nipecotic acid) were
docked in the model of GAT1 TM3, in addition to cis (8)
and trans (7) secoergoline derivatives (Table 3). H-
8
270 ( 43
680 ( 87
.1000
420 ( 57
660 ( 61
.1000
9
10
11
12
1100 ( 128
.1000
1100 ( 105
.1000
give 2 was accomplished by treatment with trifluoro-
acetic anhydride, both this and the previous steps
according to Moldvai et al.17 Aqueous KOH hydrolysis
of 1 at 60 °C (3 h) gave a ca. 2:1 mixture of half ester 3
and dicarboxylic acid 4. Separation of the two com-
pounds was performed by column chromatography. To
synthesize compounds 5-12 (Scheme 2), R-aminoketone
(i) was first alkylated with ethyl and methyl bromoace-
tate to produce 5 and vi, reductive methylation of which
gave tertiary amines 6 and 10, respectively. â-Hydroxy
esters vii and viii were prepared according to Rathke’s
method18 by reacting the keto function with lithio
acetates generated in situ with lithium bis(trimethyl-
silyl)amide at -78 °C. Water elimination from these
hydroxy compounds was performed with POCl3/pyridine
according to the method of Mehta et al.19 The two
geometric isomer pairs 7, 8 and 11, 12 were prepared
in another way as well, via phosphonate condensation
of ketones 6 and 10 according to the method described
by Wadworth et al.20 The isomers were separated by
column chromatography, and the geometry of the double
bond was elucidated by NOE measurements. Strong
interaction between the olefinic proton and 2′ and 6′
aromatic protons verifies the Z geometry in 8 and 11,
and that with all of the CH3CHNCH3 protons verifies
the E geometry in 7 and 12 (see Experimental Section).
Finally Dieckmann-condensation of isomer Z of ethyl
ester 8 in THF with lithium bis(trimethylsilyl)amide as
base at room temperature gave piperidone derivative
9. As seen by NMR measurements, piperidone deriva-
tive 9 equilibrates with the H-bond containing enol
tautomer in solution. For receptor binding and uptake
investigations, all the compounds were transformed into
water-soluble hydrochloride salts.
In Vitro Pharmacology. Effects of secoergoline
derivatives on GABA and Glu transport through cere-
brocortical membranes are summarized in Table 1. To
verify test conditions, the IC50 value of a specific GABA
uptake inhibitor, 1,2,5,6-tetrahydropiridine-3-carboxylic
acid (guvacine), has been determined. It was found to
be 6.25 ( 0.86 µM after 10 min preincubation with
guvacine. This value compares with data in the litera-
ture (8.7 ( 0.2 µM, no preincubation;21 4.92 ( 1.02 µM,
15 min preincubation22). Also in agreement with the
literature, the IC50 value of a specific Glu uptake
inhibitor, l-trans-pyrrolidine-2,4-dicarboxylic acid (t-
PDC) was found to be 5.64 ( 0.85 µM (5.1 ( 0.323).
Secoergoline derivatives with the cis, but not the trans,

Novel Secoergoline Derivatives
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5623
Table 2. Comparison of IC50 Values of Secoergoline Derivatives 8, 9, and 11 Obtained for Inhibition of [3H]GABA and [3H]DASP
Transport Processes, in the Presence and Absence (control) of NNC-711 and TBOA, Respectively, in Cerebrocortical Membrane
Suspensions at 30 °C. Number of Experiments: N ) 2 to N ) 6
IC50 for [3H]GABA transport
NNC-711a
IC50 for [3H]DASP transport
TBOAa
compound
control
phase 1
phase 2
control
phase 1
phase 2
8
9
11
270 ( 43 µM
680 ( 87 µM
1100 ( 128 µM
180 ( 7 nM
3.7 ( 0.6 nM
44 ( 2 nM
380 ( 13 µM
360 ( 8 µM
491 ( 14 µM
420 ( 57 µM
660 ( 61 µM
1100 ( 105 µM
34 ( 2 nM
385 ( 16 µM
591 ( 24 µM
1019 ( 81 µM
9.0 ( 0.2 nM
11.0 ( 0.9 nM
a 100 µM each.
fully docked in the one-TM model could also be docked
in the two-TM model (Table 4), suggesting that TM3
residues Tyr-140, Asn-137, Ser-133 and the TM2 residue
Thr-89 form a binding-crevice for GABA transport
inhibitors. Moreover, (S)nipecotic acid and secoergoline
derivative with the cis (8), but not with the trans (7),
configuration of ester groups was also found to bind in
the binding-crevice of the two-TM model (Table 4,
Figure 2C,D).
Discussion
Three of twelve secoergoline derivatives (8, 9, and 11)
inhibited both GABA and Glu transport processes (IC50
) 270-1100 µM), whereas neither these, nor the rest
of secoergoline derivatives, showed significant binding
to GABAA, GABAB, AMPA, NMDA, and kainate recep-
tors. Ineffectiveness of secoergolines 2 (a guvacine
derivative) and 1, 3, 4 (nipecotic acid derivatives) on
GABA transport suggests that carboxylic substitution
in the para position abolishes the inhibitory property
of the compounds. Moreover, the inhibitory effect emerges
by moving the carboxylic group from para (2) to ortho
position (9). Among the open secoergoline derivatives
(7, 8, 11, and 12), structures with the E configuration
(7 and 12) did not show remarkable effects.
Although GABA and Glu transporters belong to
different family and there is no homology between
them,27 all of secoergoline derivatives investigated
showed similar effectiveness on both transport pro-
cesses. This unusual lack of selectivity among structur-
ally different transporters with different functions can
be explained by various reasons. As all the three
compounds contain not only a GABA, but also a Glu
bioisosteric sequence, they could act on different trans-
porters, however, in different orientation, which may
not allow similar activities on both systems. Another
explanation can be the existence of a novel transporter
type in cerebrocortical membrane that is able to trans-
port both GABA and Glu (GABA-Glu transporter). In
Figure 1. Comparison of inhibitory effects of 8 on [3H]GABA
and [3H]DASP transport through cerebrocortical membranes
at 30 °C in the presence of 100 µM NNC-711 and 100 µM
TBOA, respectively. Symbols: O, [3H]GABA uptake; 0, [3H]-
DASP uptake. N ) 3.
Bonding interactions summarized in Table 3 indicated
that GABA and tiagabine interacted with all the three
amino acids identified by guvacine. Enantioselectivity
of binding could also be observed on the basis of the
number of H-bonding interactions (Table 3). All the
three residues interacted with the more potent enanti-
omer, (R)nipecotic acid (IC50)5.9 ( 0.8 µM26). In con-
trast, the less active enantiomer, (S)nipecotic acid
(IC50)116 ( 8 µM26), was found to bind to Tyr-140 and
Asn-137 only (Table 3, Figure 2A,B). Using the one-TM
model, however, neither binding nor cis-trans selectiv-
ity of secoergoline derivatives observed in GABA trans-
port inhibition could be predicted (Table 3).
The Two-TM Model. To understand structural
requirements for transport inhibitors better, a two-TM
model has been developed by adding GAT1 TM2 region
to GAT1 TM3 and facing polar sides of helices against
each other. In this way, H-bonding interactions of TM2
Thr-89 with GABA transport inhibitors could also be
disclosed (Table 4). All compounds that were success-
Table 3. Molecular Mechanics Calculation of Putative H-Bonding Interactions between Transport Inhibitors and Amino Acid
Residues of GAT1 TM3 (Tyr-140, Asn-137, Ser-133): The One-TM Model. Distances (Å) Were Determined between Atoms Highlighted
in Bold. H-Bonding Interaction Was Accepted If Donor-Acceptor and Hydrogen-Acceptor Distances Were below 3.0 Å and 2.5 Å,
Respectively (in italic)
ligands (IC50 (µM))
GABA
guvacine
(R)nipecotic acid
(5.9 ( 0.8a)
(S)nipecotic acid
(116 ( 8a)
tiagabine
7c
8c
GAT1 TM3
(5 ( 1a)
(14 ( 3a)
(0.07 ( 0.01a)
(>1000b)
(270b)
Tyr-140(OH)-ligand(CdO)
Tyr-140(OH)-ligand(CdO)
Asn-137(C)O)-ligand(COOH)
Asn137(CdO)-ligand(COOH)
Ser-133(OH)-ligand(N)
2.60
1.79
2.50
1.62
2.75
2.12
2.65
1.87
2.79
1.91
2.77
1.85
2.63
1.97
2.69
1.90
2.86
1.94
2.78
2.45
2.73
2.21
4.96
4.14
2.50
1.71
2.52
2.13
2.88
2.12
2.67
1.73
-
3.17
2.43
-
-
-
5.05
4.52
2.61
2.18
Ser-133(OH)-ligand(N)
a Reference 26. b Present work. c Data are presented for racemic conformation with the stronger H-bonding interactions (values for the
other enantiomeric conformation were within (0.15 Å).

5624 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
He´ja et al.
Figure 2. Different models for the 3D pharmacophore of GABA transporter type GAT1. The one-TM model: (S)nipecotic acid
(A) and (R)nipecotic acid (B) were docked in GAT1 TM3. The two-TM model: secoergoline derivatives 7 (C) and 8 (D) were docked
in GAT1 TM2 coupled GAT1 TM3. Transmembrane R-helices are visualized as traces. Ligands (grey), and amino acid residues
involved in ligand binding (light blue) are shown in stick. Nitrogen and oxygen atoms are colored blue and red, respectively.
H-bonding interactions are represented by dotted lines.
Table 4. Molecular Mechanics Calculation of Putative H-Bonding Interactions between Transport Inhibitors and Amino Acid
Residues of GAT1 TM3 (Tyr-140, Asn-137, Ser-133) and GAT1 TM2 (Thr-89): The Two-TM Model. Distances (Å) Were Determined
between Atoms Highlighted in Bold. H-Bonding Interaction Was Accepted If Donor-Acceptor and Hydrogen-Acceptor Distances
Were below 3.0 Å and 2.5 Å, Respectively (in italic)
ligands (IC50 (µM))
GABA
guvacine
(R)nipecotic acid
(5.9 ( 0.8a)
(S)nipecotic acid
(116 ( 8a)
tiagabine
7c
8c
GAT1 TM3, GAT1 TM2
(5 ( 1a)
(14 ( 3a)
(0.07 ( 0.01a)
(>1000b)
(270b)
Tyr-140(OH)-ligand(CdO)
Tyr-140(OH)-ligand(CdO)
Asn-137(CdO)-ligand(COOH)
Asn-137(CdO)-ligand(COOH)
Ser-133(OH)-ligand(N)
2.73
1.83
2.64
1.80
2.88
2.15
2.88
1.97
2.59
1.68
2.56
1.68
2.90
2.26
2.66
1.76
2.78
1.88
2.60
1.95
2.74
2.00
2.56
2.14
2.66
1.77
2.66
1.83
2.97
2.10
2.95
2.49
2.76
1.87
2.72
1.78
2.92
2.14
2.68
1.79
2.61
1.67
-
2.68
1.75
-
-
-
4.56
3.61
3.60
2.99
2.84
2.20
2.68
1.93
Ser-133(OH)-ligand(N)
Thr-89(OH)-ligand(CdO)
Thr-89(OH)-ligand(CdO)
a Reference 26. b Present work. c Data are presented for racemic conformation with the stronger H-bonding interactions (values for the
other enantiomeric conformation were within ( 0.2 Å).
this respect, it is to note that transport rates of both
GABA- and Glu-containing synaptic vesicles were modu-
lated with the same endogenous protein.28 The existence
of this protein might substantiate a proposal that both
GABA and Glu transport processes can be regulated
with the same compound. Furthermore, underlying
mechanisms, that is inhibition on both GABA and Glu
transporters vs inhibition on GABA-Glu transporter can
operate in the same time. It follows that secoergoline
derivatives could manifest higher inhibitory potential
for an isolated GABA-Glu transporter. The pharmaco-
logical approach to the isolation of the transport process
mediated by a possible GABA-Glu transporter, by
blocking the major GABA and Glu transport processes
with high-affinity, nontransportable inhibitors, sepa-
rated high- and low-capacity GABA and Glu transport
processes. The high-capacity processes can correspond
to transports driven by known GABA and Glu trans-
porters. Since NNC-711 is a specific inhibitor of GAT1
(IC50 values were found to be 0.04, 171, 1700, and 622

Novel Secoergoline Derivatives
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5625
µM for the cloned GAT1, GAT2, GAT3, and BGT1,
respectively24), inhibitory effects of secoergoline deriva-
tives on both high- and low-capacity GABA transport
through cerebrocortical membranes were neither ex-
pected nor found to be the same in the presence and
absence of NNC-711. In contrast, inhibitory effects of
8, 9, and 11 on the high-capacity Glu transport process
were similar to those obtained in the absence of DL-
TBOA, a nearly equipotent inhibitor of Glu transporters
(IC50 values are 70, 6, 6, 4.4, and 3.2 µM for the cloned
Glu transporter types 1-5, respectively25), suggesting
that DL-TBOA affects less the transport properties of
secoergoline derivatives than NNC-711 does. The low-
capacity processes disclosed are unique as to secoer-
goline derivatives 8, 9, and 11 inhibit them similarly,
with substantially higher inhibitory potential and se-
lectivity. The low- and the high-capacity GABA and Glu
transport processes are also distinguishable on the basis
of different order of inhibitory potential of 8, 9, and 11.
Inhibitory sequences, in order of decreasing inhibition,
were 9 g 11 > 8 and 8 g 9 > 11, respectively. As the
inhibitory potential of compounds for the low-capacity
GABA and Glu transport phases is much higher than
that of any known GABA and Glu uptake inhibitors,
these low-capacity processes are unlikely to be mediated
by the known GABA- and Glu transporters. Indeed,
such a high affinity may only be functional when the
signaling amount of neurotransmitters had already been
taken up by the known high-capacity processes. This
way, we may conjecture the existence of a novel type of
transport process characterized by extreme high affinity
and low capacity with a role in some, rather general
clearance mechanism in or outside the synaptic cleft.
Moreover, it is hardly probable that two different targets
could be affected with the same compounds in the 3-200
nM concentration range; thus, we suppose that the novel
type of transport process described here may be medi-
ated by a single target, a GABA-Glu transporter.
Conclusively these findings suggest that molecules, such
as secoergoline derivatives 8, 9, and 11, which have
structural motifs allowing them to interact with both
GABA and Glu transporters, do highlight the possible
existence of a previously unrecognized GABA-Glu trans-
porter.
On the basis of mutagenesis studies, Kanner and co-
workers identified Tyr-140 residue in TM3 helix of
GAT1, which is likely involved in ligand binding.13
Results of GAT1 mutagenesis studies corroborate the
“alternate access” model of secondary transport viewed
as the alternating outward and inward reorientation of
the loaded and empty binding site.29 Thus, the minimal
functional unit for Na+ ion-coupled GABA translocation
can be viewed as the third transmembrane helix (TM3)
of GAT1. Thus, in the absence of the three-dimensional
structure of GABA transporters the TM3 region was
addressed first. On the basis of computer-aided model-
ing of interactions between guvacine and the GAT1 TM3
region, two additional amino acid residues have been
identified, which could possibly be involved in ligand
binding, Asn-137 and Ser-133. Binding interaction
between the amino acid residues, Tyr-140, Asn-137, and
Ser-133 and the GABA uptake inhibitors may provide
a minimal model for a 3D pharmacophore of the GABA
transporter. This one-TM model predicts enantioselec-
tivity of GABA transport inhibitors. However, even the
most active secoergoline derivative tested cannot be
docked into the binding core formed by all the three
amino acids involved in the one-TM model, suggesting
that more binding interactions are required to dock low-
potency compounds. Indeed, the proposed model of the
3D binding-crevice involving both GAT1 TM3 and TM2
helices is able to predict cis-trans selectivity of seco-
ergoline derivatives. Moreover, the two-TM model com-
bined with the one-TM model can be used to estimate
inhibitory potency as high- and low-potency compounds
can be docked to the two-TM model, but only compounds
with high affinity can fit in the one-TM model. Based
on knowledge of the role of GABA-mediated inhibitory
neurotransmission in the pathogenesis of several types
of epilepsy, the proposed 3D pharmacophore model may
provide new insights into the structural prerequisites
for the anticonvulsant activity of GABA transport
inhibitors.
Experimental Section
Chemistry. All compounds prepared and investigated are
racemates. Melting points are uncorrected. Mass spectra were
run on an AEI-MS-902 (70 eV; direct insertion) mass spec-
trometer. IR spectra were taken on a Nicolet 205 and Nicolet
7795 FT-IR spectrophotometers. NMR spectroscopy measure-
ments were carried out on a Varian Unity Inova (400 MHz)
instrument. Chemical shifts are given relative to TMS ) 0.00
ppm. Elemental analyses (C, H, N) were carried out by Vario
EL III (Elementar Analyzen Systeme GmbH) automatic mi-
croanalyzer, ionic halide content was measured by titration
with mercuric perchlorate. For thin-layer chromatography
(TLC) analyses MN Polygram SIL G/UV254 sheets were used.
Preparative separations were performed by column chroma-
tography on Merck Kieselgel 60 (0.063-0.200).
Glassware was flame dried before use. Solvents were
carefully dried and purified by appropriate methods. The
majority of the reactions were carried out under argon protec-
tion.
1-Methyl-2-oxo-2-phenylethylammonium chloride (i), diethyl
2-[(1-methyl-2-oxo-2-phenylethylamino)methylene]succinate (ii),
diethyl2-[methyl-(1-methyl-2-oxo-2-phenylethylamino)methyl]-
succinate (iii), diethyl 5â-hydroxy-1,6-dimethyl-5R-phenyl-
piperidine-3R,4â-dicarboxylate (1, iv, and v), and diethyl
1,6-dimethyl-5-phenyl-1,2,3,6-tetrahydropyridine-3,4-dicar-
boxylate (2) were prepared as described.17
Ethyl 1,6-Dimethyl-5â-hydroxy-5r-phenyl-3r-carbox-
ypiperidine-4â-carboxylate (3) and 1,6-Dimethyl-5â-hy-
droxy-5r-phenylpipereidine-3r,4â-dicarboxylic Acid (4).
To the solution of diastereomeric mixture of hydroxy-diester
1 (900 mg; 2.57 mmol) in ethanol (50 mL) was added a solution
of KOH (1.6 g; 28.8 mmol) in distilled water (30 mL). The
mixture was stirred at 60-70 °C for 3 h, cooled, and acidified
with 1:1 aq HCl (pH 3) and evaporated in vacuo. The residue
containing 3 and 4 was separated by column chromatography
(eluent: ethanol-25% NH4OH 5:1) to afford monocarboxylic
acid 3 (550 mg; 66%; Rf: 0.7) and dicarboxylic acid 4 (214 mg,
28%, Rf: 0.2). 3: pale tan crystals; mp 199-210 °C. IR (KBr):
1607 (CO2-); 1725 (CdO); 3140 cm-1 (OH). 1H NMR (diaster-
eomers; 400 MHz; DMSO-d6): 0.55+0.70 (3H; 2 × t; J ) 7.1
Hz; CH2CH3); 0.66+0.74 (3H; 2 × d; J ) 6.8 Hz; 6-CH3);
2.36+2.38 (3H; 2 × s; NCH3); 2.50+2.86+2.54+3.25 (2H; m;
H2); 2.64+2.76 (1H; 2 × q; J ) 7.1 Hz; H6); 2.98+3.41 (1H; 2
× d; J ) 11.6 Hz; H4); 3.20-3.30 (1H; m; H3); 3.52+3.66+3.76
(2H; 3 × q; J ) 7.1 Hz; OCH2); 4.66 (1H; brs; OH); 7.21 (1H;
m; H4′); 7.24 (2H; m; H3′+H5′); 7.41+7.60 (2 × 1H; m;
H2′+H6′). 13C NMR (100.6 MHz; DMSO-d6): 5.7+12.9 (6-CH3);
13.7+13.9 (CH2CH3); 41.0+41.4 (C3); 42.4+42.5 (NCH3);
44.6+54.1 (C4); 47.6+57.2 (C2); 59.3+59.7 (OCH2); 65.7+66.1
(C6); 75.9+76.2 (C5); 125.9+126.9+127.4+127.7+128.0
(C2′+C3′+C4′+C5′+C6′); 142.5+144.1 (C1′); 171.4+172.0 (CO2-

5626 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
He´ja et al.
Et); 174.0+174.2 (CO2H). Anal. (C17H23NO5) C, H, N. 4: pale
tan crystals; mp 229-231 °C. IR (KBr): 1600 (CO2-); 3120
cm-1 (OH). 1H NMR (diastereomers; 400 MHz; DMSO-d6):
0.93+1.00 (3H; 2 × d; J ) 6.8 Hz; 6-CH3); 2.78+2.84 (3H; 2 ×
s; NCH3); 2.96+3.22 (1H; 2 × d; J ) 11.4 Hz; H4); 3.28 (1H;
m; H3); 3.24+3.31+3.58 (2H; m; H2); 3.39+3.62 (1H; 2 × q; J
) 6.8 Hz; H6); 7.25 (1H; m; H4′); 7.34 (2H; m; H3′+H5′);
7.36+7.47 (2 × 1H; 2 × m; H2′+H6′). 13C NMR (100.6 MHz;
DMSO-d6): 6.1+10.8 (6-CH3); 40.9+41.5 (C3); 43.4+44.1
(NCH3); 45.7+55.5 (C4); 48.1+56.2 (C2); 65.5+66.2 (C6);
75.3+75.3 (C5); 125.5+126.6+128.2+128.8+129.0 (C2′+C3′+
C4′+C5′+C6′); 141.3+ 141.7 (C1′); 176.1+176.6+177.8+178.6
(2 × CO2H). Anal. (C15H19NO5) C, H, N.
8.03 (2H; d; H2′+H6′); 10.20 (1H; brs; +NH). Anal. (C14H20-
ClNO3) C, H, N, Cl.
Methyl [Methyl-(1-methyl-2-oxo-2-phenylethyl)amino]-
acetate Hydrochloride (10). Alkylated amine HCl vi (1.6
g; 6.2 mmol) was methylated into the methyl ester analogue
of 6 (10) similarly to the previous procedure to afford the target
molecule, which was isolated as oily HCl salt (1.68 g; quantita-
tive yield) IR (KBr): 1680 (CdO ketone); 1740 (CdO ester);
3410 cm-1 (+NH2). 1H NMR (base; 400 MHz; CDCl3): 1.31 (3H;
d; J ) 6.8 Hz; CH3CHN); 2.45 (3H; s; NCH3); 3.50+3.58 (2 ×
1H; 2 × d; Jgem ) 16.9 Hz; NCH2CO); 3.70 (3H; s; OCH3); 4.53
(1H; q; J ) 6.8 Hz; CH3CHN); 7.44 (2H; t; H3′+H5′); 7.58 (1H;
t; H4′); 8.06 (2H; d H2′+H6′). Anal. (C13H18ClNO3) C, H, N,
Cl.
Ethyl (1-Methyl-2-oxo-2-phenylethylamino)acetate Hy-
drochloride (5). To the suspension of amine HCl (i) (9.25 g;
50 mmol) in THF (100 mL) was added triethylamine (13.9 mL;
10.1 g; 100 mmol), and the mixture was cooled in ice bath to
0 °C. Into the stirred mixture was dropped in small portions
ethyl bromoacetate (5.5 mL; 8.35 g; 50 mmol) over 30 min.
The mixture was then stirred at ambient temperature for
1-1.5 h, and progress of the alkylation could be followed by
TLC (hexane-EtOAc 1:1 after treating the sheet with NH3
vapor; Rf 5 > Rf i). Triethylammonium chloride precipitated
was filtered off and the solution evaporated in vacuo. The
residue was purified by column chromatography (eluent:
chloroform-acetonitrile 4:1) to afford the alkylated base as an
oil (7.05 g; 60%). The compound was isolated and characterized
as hydrochloride salt 5 (prepared in EtOAc) as off-white
crystals (7.0 g; 52%), mp 131-133 °C (ethyl acetate). IR
(KBr): 1680 (CdO ketone); 1755 (CdO ester); 3410 cm-1
(+NH2). 1H NMR (base; 400 MHz; CDCl3): 1.23 (3H; t;
CH2CH3); 1.38 (3H; d; J ) 6.9 Hz; CH3CHN); 2.20 (1H; brs;
NH); 3.35+3.45 (2 × 1H; 2 × d; Jgem ) 16.5 Hz; NCH2CO);
4.18 (2H; q; OCH2); 4.41 (1H; q; J ) 6.9 Hz; CH3CHN); 7.49
(2H; t; H3′+H5′); 7.61 (1H; t; H4′); 7.98 (2H; d; H2′+H6′), Anal.
(C13H18ClNO3) C, H, N, Cl.
Methyl (1-Methyl-2-oxo-2-phenylethylamino)acetate
Hydrochloride (vi). The methyl ester analogue of 5 (vi) was
prepared similarly to the previous procedure. Amine HCl i (3.7
g; 20 mmol) was alkylated with methyl bromoacetate to afford
the target molecule after chromatography as a pale yellow oil
(2.5 g; 57%). The compound was isolated and characterized
as hydrochloride salt vi (prepared in chloroform) as off-white
crystals (2.5 g; 48%), mp 156-158 °C (methanol-diethyl
ether). IR (KBr): 1680 (CdO ketone); 1740 (CdO ester); 3420
cm-1 (+NH2). 1H NMR (400 MHz; CDCl3+DMSO-d6): 1.64 (3H;
d; J ) 7.0 Hz; CH3CHN); 3.83 (3H; s; OCH3); 4.01+4.09 (2 ×
1H; 2 × d; Jgem ) 16.5 Hz; +NCH2CO); 5.39 (1H; q; J ) 7.0
Hz; CH3CHN+); 7.58 (2H; t; H3′+H5′); 7.71 (1H; t; H4′); 8.03
(2H; d; H2′+H6′); 10.10 (2H; brs; +NH2). Anal. (C12H16ClNO3)
C, H, N, Cl.
Ethyl [Methyl-(1-methyl-2-oxo-2-phenylethyl)amino]-
acetate Hydrochloride (6). The HCl salt 5 (6.8 g; 25 mmol)
was transformed into the corresponding base (oil) and its
solution in dry acetonitrile (400 mL) and acetic acid (95 mL)
was cooled to 0 °C. Into the stirred solution, sodium cy-
anoborohydride (6.1 g; 96 mmol) and 36% aq formaldehyde
solution (140 mL; 1.66 mol) was added. The mixture was
stirred at 0 °C, progress of the methylation could be followed
by TLC (hexanes-EtOAc 1:1 after treating the sheet with NH3
vapor; Rf 6>Rf 5). The reaction was complete after 30-40 min.
The mixture was diluted with water (500 mL), its pH was
adjusted to neutral with Na2CO3 solution and extracted with
chloroform (3 × 300 mL). The combined organic phase was
dried (Na2SO4) and evaporated. The residue was dissolved in
ethanol-diethyl ether 1:5 (50 mL), and acidified with 2N HCl/
dioxane (pH 2) to give white crystals of target compound 6
(6.35 g; 89%). Mp 155-157 °C (acetone). IR (KBr): 1680 (Cd
O ketone); 1738 (CdO ester); 3400 cm-1 (+NH2). 1H NMR (400
MHz; DMSO-d6): 1.25 (3H; t; CH2CH3); 1.50 (3H; d; J ) 6.8
Hz; CH3CHN) 2.89 (3H; s NCH3); 4.08+4.11 (2 × 1H; 2 × d;
Jgem ) 16.8 Hz; +NCH2CO); 4.22 (2H; q; OCH2); 5.35 (1H; q; J
) 6.8 Hz; CH3CHN); 7.59 (2H; t; H3′+H5′); 7.70 (1H; t; H4′);
Ethyl 4-[(Ethoxycarbonylmethyl)methylamino]-3-hy-
droxy-3-phenylpentanoate (vii). A well-dried three-neck
flask was immersed in an acetone/dry CO2 cooling bath (-78
°C), and under argon protection, 1 N lithium bis(trimethylsi-
lyl)amide/THF solution (6 mL; 6 mmol) and ethyl acetate (0.60
mL; 0.54 g; 6 mmol) were admixed. The solution was stirred
for 15 min at -78 °C. Then a solution of aminoketone in dry
THF (40 mL), obtained from the solution of the HCl salt (6)
(570 mg; 2 mmol) in water (10 mL) by having treated with
10% aq solution of NaHCO3 (pH 7), extracted with chloroform,
and evaporated, was admixed at -78 °C. Progress of the
addition could be followed by TLC (hexane-ethyl acetate 4:1;
Rf vii> Rf 6). The reaction was usually complete after 1 h. At
this temperature 20% aq HCl (0.4 mL; ∼2.4 mmol) was added
to the mixture, which was left to warm to room temperature.
The pH of the mixture was adjusted to 7 when needed. After
distilling off THF in vacuo, the residue was dissolved in a
mixture of chloroform (50 mL) and water (30 mL). The organic
phase was dried (Na2SO4) and evaporated and the residue
purified by column chromatography (eluent: hexane-ethyl
acetate 9:1). After evaporation to dryness the inseparable
diastereomeric mixture of hydroxy-diesters (vii) was obtained
as a pale yellow oily base (500 mg; 75%). As the compound
proved to be rather unstable, it was used in the next step
immediately. IR (KBr): 1730 (broad; CdO ester); 3460 cm-1
1
(OH). H NMR (400 MHz; CDCl3): 1.00 (3H; d; J ) 6.8 Hz;
CH3CHN); 1.05+1.26 (2 × 3H; 2 × t; 2 × CH2CH3); 2.20 (3H;
s; NCH3); 2.86+3.44 (2 × 1H; 2 × d; NCH2CO); 2.93 (1H; q; J
) 6.8 Hz; CH3CHN); 3.06+3.11 (2 × 1H; 2 × d; Jgem)16.7 Hz;
CCH2CO); 4.00+4.18 (2 × 2H; 2 × q; 2 × OCH2); 7.20-7.50
(5H; m; aromatic protons).
Methyl 4-[(Methoxycarbonylmethyl)methylamino]-3-
hydroxy-3-phenylpentanoate (viii). The dimethyl ester
analogue of vii (viii) was prepared from amine 10 similarly
to the previous procedure; instead of ethyl acetate, an equiva-
lent amount of methyl acetate was used in the reaction. In a
2 mmol charge after purification with column chromatography
(eluent: hexane-ethyl acetate 9:1) the inseparable diastere-
omeric mixture of hydroxy-diesters (viii) was obtained as a
pale yellow oily base (409 mg; 66%). As the compound proved
to be rather unstable, it was used in the next step immediately.
IR (KBr): 1730 (broad; CdO ester); 3450 cm-1 (OH). 1H NMR
(400 MHz; CDCl3): 1.01 (3H; d; J ) 7.0 Hz; CH3CHN); 2.21
(3H; s; NCH3); 2.89+3.43 (2 × 1H; 2 × d; Jgem) 16 Hz; NCH2-
CO); 2.94 (1H; q; J ) 7.0 Hz; CH3CHN); 3.11+3.13 (2 × 1H; 2
× d; Jgem) 16.7 Hz; CCH2CO); 3.55+3.67 (2 × 3H; 2 × s; 2 ×
OCH3); 3.65 (1H; brs; OH); 7.23 (1H; m; H4′); 7.28 (2H; m;
H3′+H5′); 7.46 (2H; m; H2′+H6′).
E and Z Ethyl 4-[(Ethoxycarbonylmethyl)methylami-
no]-2-methyl-3-phenylpent-2-enoate (7 and 8). Method A
(via phosphonate condensation of ketone 6). In a two-necked
flask 50% NaH dispersion in mineral oil (1.8 g; 36 mmol) was
washed with dry THF (2 × 25 mL) until oil-free. Then it was
suspended in dry THF (20 mL) and cooled to 0-5 °C, and a
solution of triethyl phosphonoacetate (6.72 g; 30 mmol) in dry
THF (10 mL) was added. After stirring the mixture at 0-5 °C
for 30 min, the temperature was allowed to warm to room
temperature, and a suspension of ketone HCl 6 (1.14 g; 4
mmol) in dry THF (20 mL) was admixed. The progress of the
condensation followed by TLC (hexane-ethyl acetate 4:1; Rf

Novel Secoergoline Derivatives
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5627
7 and 8 > Rf 6). The reaction usually was complete (over 90%
conversion) after 3 days. Then water (100 mL) was added, the
THF was distilled off in reduced pressure, and the water phase
was extracted with chloroform (3 × 50 mL). The combined
organic phase was dried and evaporated in vacuo, and the
residue was separated by column chromatography (eluent:
hexane-ethyl acetate 9:1). The first compound eluted from the
column proved to be the Z geometric isomer 8 (310 mg; 25%).
IR (KBr): 1690 (CdO conjug ester); 1735 cm-1 (CdO ester).
1H NMR (400 MHz; CDCl3): 1.20 (3H; d; J ) 7.0 Hz; CH3-
CHN); 1.27+1.32 (2 × 3H; 2 × t; 2 × CH2CH3); 2.43 (3H; s;
NCH3); 3.27+3.52 (2 × 1H; 2 × d; Jgem)16.2 Hz; NCH2CO);
4.16+4.21 (2 × 2H; 2 × q; 2 × OCH2); 4.78 (1H; q; J ) 7.0 Hz;
CH3CHN); 5.98 (1H; s; dCH); 7.31-7.35 (3 × 1H; m; H3′+
H4′+H5′); 7.67 (2H; m; H2′+H6′). NOE: 5.98 (dCH) f 7.67
(H2′+H6′). Anal. (C18H25NO4) C, H, N. The second compound
eluted from the column proved to be the E geometric isomer 7
(330 mg; 26%). IR (KBr): 1700 (CdO conjug ester); 1735 cm-1
The reaction was usually complete after 12 h. At this temper-
ature the mixture was neutralized with 20% aq HCl (pH7).
After distilling off the THF in vacuo, the residue was dissolved
in a mixture of chloroform (20 mL) and water (10 mL). The
organic phase was dried (Na2SO4) and evaporated, and the
residue was purified by column chromatography (eluent:
chloroform-acetonitrile 9:1). After evaporation the pyridone
9 was isolated as an oily base (101 mg; 42%), which exists in
equilibrium with chelated enol form. IR (KBr): 1640 (CdC;
CdO conjug H-bridged ester); 1680 (CdO ketone); 1730 (Cd
O ester); 3300 cm-1 (chel OH). 1H NMR (400 MHz; CDCl3):
1.20 (3H; d; 2-CH3); 1.39 (3H; t; CH2CH3); 2.40 (0.5 × 1H; s;
H6 keto form); 2.55 (3H; s; 1-CH3); 3.94 (1H; q; H2); 4.31 (0.5
× 2H; q; OCH2 keto form); 4.46 (0.5 × 2H; q; OCH2 enol form);
6.39 (1H; s; H4); 7.30-7.50 (5H; m; aromatic protons); 11.10
(0.5 × 1H; s; OH enol form). Anal. (C16H19NO3) C, H, N. In
binding and uptake measurements the HCl salt was investi-
gated.
1
(CdO ester). H NMR (base; 400 MHz; CDCl3): 1.05+1.27 (2
In Vitro Pharmacology. Three to six week old male
Wistar rats purchased from Toxicoop (Budapest, Hungary)
were kept and used in accordance with the European Council
Directive of 24 November 1986 (86/609/EEC) and with the
Hungarian Animal Act, 1998, and associated guidelines.
Compounds used in the experiments: tris(hydroxymethyl)-
aminomethane (TRIS) (Reanal: Budapest, Hungary); N-[2-
hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid] (HEPES),
aprotinin, leupeptin, antipain, pepstatin A, phenyl methane-
sulfonyl fluoride (PMSF), butylated hydroxytoluene (BHT),
t-PDC, guvacine, (1S,9R)(-)bicuculline methiodide (BMI),
(()2(2-chlorophenyl)-2-(methylamino)cyclohexanone ((()ket-
amine), and (L)glutamic acid (Sigma-Aldrich, Budapest, Hun-
gary); CaCl2, NaCl, KCl, MgCl2, NaOH, and glucose (Fluka:
Budapest, Hungary); (S)(-)-5-fluorowillardiine, [3H](S)(-)5-
fluorowillardiine (36 Ci/mmol), (RS)4-amino-3-(4-chlorophen-
yl)butanoic acid ((RS)baclofen), [3H](+)MK 801 (32,75 Ci/mmol)
(Tocris: Bristol, U.K.); [3H]kainic acid (58 Ci/mmol) (NEN:
Boston, MA); [3H](D)asparatic acid ([3H]DASP) (24 Ci/mmol),
[3H]GABA (61 Ci/mmol) (Amersham: Little Chalfont, U.K.);
HiSafe 3 scintillation mixture (LKB-Wallac: Bromma, Swe-
den) and spectroscopic grade dimethyl sulfoxide (DMSO,
Merck: Budapest, Hungary). Buffers had the following com-
positions: A, 145 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM
MgCl2, 10 mM glucose, 0.1 mM aminooxyacetate, 2 mM
NaHCO3, and 20 mM HEPES adjusted to pH 7.5 with NaOH;
B, 50 mM TRIS and 2.5 mM CaCl2 adjusted to pH 7.4 with
HCl;
× 3H; 2 × t; 2 × CH2CH3); 1.15 (3H; d; J ) 6.8 Hz; CH3CHN);
2.47 (3H; s; NCH3); 3.38+3.49 (2 × 1H; 2 × d; Jgem)16.6 Hz;
NCH2CO); 3.60 (1H; q; J ) 6.8 Hz; CH3CHN); 3.98+4.16 (2 ×
2H; 2 × q; 2 × OCH2); 6.19 (1H; d; J ) 1.5 Hz; dCH); 7.17
(2H; m; H2′+H3′); 7.33 (3 × 1H; m; H3′+ H4′+H5′). NOE: 6.19
(dCH) f 1.15+3.60 (CH3CHN); 2.47 (N-CH3). Anal. (C18H25-
NO4) C, H, N. In binding and uptake measurements the HCl
salts of both isomers were investigated. Method B (via water
elimination from vii). To the solution of the oily hydroxy-
diester vii (675 mg; 2 mmol) in pyridine (20 mL) was added
POCl3 (0.6 mL; 6.5 mmol), and the mixture was refluxed on a
120 °C oil bath. Progress of the water elimination could be
followed by TLC (chloroform-acetonitrile 9:1; Rf 7 and 8 > Rf
vii). The conversion was usually complete after 2-3 h. The
mixture was cooled and evaporated in vacuo, and the residue
was dissolved in a mixture of chloroform (50 mL) and water
(20 mL), while the pH of the water phase was adjusted to a
value of 7-7.5. The organic phase was dried (Na2SO4) and
evaporated, and the residue was purified by column chroma-
tography (eluent: chloroform-acetonitrile 15:1 or hexane-
ethyl acetate 9:1). In this way mainly Z geometric isomer 8 is
obtained (290 mg; 45%) along with some isomer E (7) as minor
product (45 mg; 7%). Both isomers obtained in this way proved
to be identical (TLC, NMR) with those prepared according to
Method A.
Z and E Methyl 4-[(Methoxycarbonylmethyl)methyl-
amino)-2-methyl-3-phenylpent-2-enoate (11 and 12). The
two geometric isomers 11 and 12 were prepared in analogous
way described in the previous example by both methods and
were characterized as oily hydrochloride salts. 11 HCl salt:
oil; yield after chromatography: 30%. IR (KBr): 1690 (CdO
conjug ester); 1735 cm-1 (CdO ester). 1H NMR (400 MHz;
CDCl3): 1.76 (3H; d; J ) 6.8 Hz; CH3CHN+); 3.21 (3H; s; +N-
CH3); 3.83+3.86 (2 × 3H; 2 × s; 2 × OCH3); 4.25+4.46 (2 ×
1H; 2 × d; Jgem)16.5 Hz; +NCH2CO); 5.93 (1H; q; J ) 6.8 Hz;
CH3CHN+); 6.27 (1H; s; dCH); 7.20-7.55 (3H; m;
H3′+H4′+H5′); 7.68 (2H; m; H2′+H6′). NOE: 6.27 (dCH) f
7.68 (H2′+H6′). Anal. (C16H22ClNO4) C, H, N, Cl. 12 HCl salt:
oil; yield after chromatography: 30%. IR (KBr): 1690 (CdO
conjug ester); 1730 cm-1 (CdO ester). 1H NMR (400 MHz;
CDCl3):1.76 (3H; d; J ) 6.9 Hz; CH3CHN); 3.05 (3H; s; +N-
CH3); 3.60+3.76 (2 × 3H; 2 × s; 2 × OCH3); 4.02+4.06 (2 ×
1H; 2 × d; Jgem)17.0 Hz; +NCH2CO); 4.62 (1H; q; J ) 6.9 Hz;
CH3CHN+); 7.06 (1H; s; dCH); 7.21 (2H; m; H2′+H6′); 7.30-
7.45 (3H; m; H3′+H4′+H5′). NOE: 7.06 (dCH) f 1.76+4.62
(CH3CHN); 3.05 (+N-CH3). Anal. (C16H22ClNO4) C, H, N, Cl.
GABA and Glu Transport.21 Native plasma membrane
vesicle fraction was obtained as described before with the
exception that the final pellet was suspended in 15 mL of
buffer A.
Receptor Binding Assays. Affinity of test compounds to
GABAA,30 NMDA,31 AMPA,32 and kainate31 receptors were
determined as described before. IC50 values for standards were
found to be 0.5 ( 0.1 µM (BMI on GABAA receptor), 1.6 ( 0.2
µM ((S)AMPA on AMPA receptor), 0.04 ( 0.01 µM (kainate
on kainate receptor), and 0.51 ( 0.14 µM (ketamine on NMDA
receptor). Affinity for GABAB receptor was determined accord-
ing to Hill and Bowery33 with modifications. Briefly, rats were
decapitated and the brains were rapidly removed and homog-
enized in 15 vol of ice-cold 0.32 M sucrose solution. The
homogenate was centrifuged at 1000g for 10 min, the pellet
was discarded, and the supernatant fluid was centrifuged at
20000g for 20 min. The pellet was resuspended in 15 vol of
distilled water, lyzed on ice for 30 min, and centrifuged at
8000g for 20 min. The supernatant and the “buffy coat layer”
loosely attached to the pellet was collected and centrifuged at
40000g for 20 min. The pellet was resuspended and pelleted
at 40000g for 20 min in 15 vol of distilled water. The pellet
was frozen at -20 °C for at least 18 h. On the following day
the pellet was thawed, resuspended in 10 vol of buffer B, and
after incubation at 25 °C for 45 min, it was centrifuged at
7000g for 10 min. The pellet was resuspended in 10 vol of
buffer B and centrifuged at 7000g for 10 min. This washing
procedure was repeated one more time, and the final pellet
Ethyl 1,6-Dimethyl-3-oxo-5-phenyl-1,2,3,6-tetrahydro-
pyridine-2-carboxylate (9). To the stirred suspension of
hydrochloride salt of pure Z geometric isomer 8 (310 mg; 0.87
mmol) in dry THF (10 mL) was added 1 N lithium bis-
(trimethylsilyl)amide/THF solution (2 mL; 2 mmol) at ambient
temperature. Progress of the Dieckmann condensation could
be followed by TLC (chloroform-acetonitrile 9:1; Rf 9 > Rf 8).
After 10 h, a further amount of 1 N lithium bis(trimethylsilyl)-
amide/THF (3 mL; 3 mmol) was added to the stirred mixture.

5628 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23
He´ja et al.
(6) Sza´rics, EÄ .; Nyikos, L.; Baraba´s, P.; Kova´cs, I.; Skuban, N.;
Temesva´rine´-Major, E.; Egyed, O.; Nagy, P. I.; Ko¨ko¨si, J.;
Taka´cs-Nova´k, K.; Kardos, J. Quinazolone-alkyl-carboxylic acid
derivatives inhibit transmembrane Ca2+ ion flux to (+)[S]-R-
amino-3-hydroxy-5-methylisoxazole-4-propionic acid. Mol. Phar-
macol. 2001, 59, 920-928.
(7) Lo¨scher, W. Current status and future directions in the phar-
macotherapy of epilepsy. Trends Pharmacol. Sci. 2002, 23, 113-
118.
(8) Dalby, N. O. GABA-level increasing and anticonvulsant effects
of three different GABA uptake inhibitors. Neuropharmacol.
2000, 39, 2399-2407.
(9) Frølund, B.; Ebert, B.; Kristiansen, U.; Liljefors, T.; Krogsgaard-
Larsen, P. GABAA receptor ligands and their therapeutic
potential. Curr. Topics Med. Chem. 2002, 2, 817-832.
(10) Cromer, B. A.; Morton, C. J.; Parker, M. W. Anxiety over GABAA
receptor structure relieved by AChBP. Trends Pharmacol. Sci.
2002, 27, 280-287.
was resuspended in 10 vol of buffer B. Aliquots (1000 µL) of
crude synaptic membrane homogenate in buffer B were
incubated with a 1000 µL solution of test compounds at 20 °C
for 40 min in the presence of a 200 µL solution of [3H]GABA
and isoguvacine in final concentration of 15 nM and 40 µM,
respectively. The nonspecific binding was determined with 100
µM (RS)baclofen. After the incubation, samples of 1000 µL
were centrifuged in Eppendorf tubes at 11000g for 3 min at
20 °C. The supernatant of the samples was aspirated under
vacuum, and the pellets were rinsed two times with ice-cold
buffer B. The pellet was solubilized with vortex in 150 µL of
10% SDS solution, and the radioactivity of samples was
counted in HiSafe 3 scintillation mixture. IC50 of the standard,
(RS)baclofen, was found to be 0.4 ( 0.09 µM.
Molecular Modeling. All calculations were performed
using Sybyl 6.6 package (SYBYL 6.7.1 Tripos Inc., 1699 South
Hanley Rd., St. Louis, Missouri, 63144) with Tripos force field
and Gasteiger-Hu¨ckel charges. The Tripos force field was used
throughout the calculations for its simplicity, robustness, and
considerably faster execution times as compared to other, more
sophisticated force fields (e.g. AMBER or MMFF94) and
because of the inherent uncertainty of side chain positions in
a model to be constructed on the basis of the amino acid
sequence alone. The TM regions (TM3 and TM2) were built
by Sybyl Build Biopolymer command based on amino acid
sequence of TM3 (TM2) followed by minimization with sim-
plexing and applying Powell algorithm. The relative orienta-
tion of TM3 and TM2 was calculated using a knowledge-based
scale for transmembrane helix structure prediction.34 Geom-
etry of each ligand molecule was obtained upon molecular
mechanics energy minimization of the positively charged form
using simplexing and applying Powell and BFGS algorithm
alternately. Docking procedures were performed as follows:6
ligand molecules were positioned in the hypothesized manner
(CdO group of COOH faced to Tyr-140 of TM3 and Thr-89 of
TM2, OH group of COOH faced to Asn-137 and N faced to Ser-
133). After minor manual adjustments for finding a favorable
initial position for the ligand, the energy of the system was
minimized throughout the docking procedure during which
neither site nor ligand geometry was fixed. H-Bonding interac-
tion was accepted if donor-acceptor and hydrogen-acceptor
distances were below 3.0 Å and 2.5 Å, respectively.35
(11) Wipf, D.; Ludewig, U.; Tegeder, M.; Rentsch, D.; Koch, W.;
Frommer, W. B. Conservation of amino acid transporters in
fungi, plants and animals. Trends Biol. Sci. 2002, 27, 139-147.
(12) Kleinberger-Doron, N.; Kanner, B. I. Identification of Tryptophan
Residues Critical for the Function and Targeting of the γ-Ami-
nobutyric Acid Transporter (Subtype A). J. Biol. Chem. 1994,
269, 3063-3067.
(13) Bismuth, Y.; Kavanaugh, M. P.; Kanner, B. I. Tyrosine 140 of
the γ-aminobutyric acid transporter GAT-1 plays a critical role
in neurotransmitter recognition. J. Biol. Chem. 1997, 272,
16096-16102.
(14) Bennett, G. B.; Mason, R. B.; Alden, L. J.; Roach, J. B., Jr.
Synthesis and antiinflammatory activity of trisubstituted py-
rimidines and triazines. J. Med. Chem. 1978, 21, 623-628.
(15) House, H. O.; Berkowitz, W. F. The Neber rearrangement of
substituted desoxybenzoin oxime tosylates. J. Org. Chem. 1963,
28, 307-311.
(16) Stoll, A.; Rutschmann, J.; Petrzilka, Th. U¨ ber Derivate des
1;3;4;5-Tetrahydro-benz[c; d]indols. Helv. Chim. Acta 1950, 33,
2257-2261.
(17) Moldvai, I.; Temesva´ri-Major, E.; Incze, M.; Platthy, T.; Ga´cs-
Baitz, E.; Sza´ntay, Cs. Chemistry of indoles carrying a basic
function. Part VII. A new aspect of Stobbe reaction. Heterocycles
2003, 60, 309-319.
(18) Rathke, M. W. The preparation of lithio ethyl acetate. A simple
procedure for the conversion of aldehydes and ketones to
â-hydroxy esters. J. Am. Chem. Soc. 1970, 92, 3222-3223.
(19) Mehta, G.; Murthy, A. N.; Sivakumar-Reddy, D.; Veere-Reddy,
A. A general approach to linearly fused triquinane natural
products. Total synthesis of (()-hirsutene, (()-corioline and (()-
capnellene. J. Am. Chem. Soc. 1986, 108, 3443-3452.
(20) Wadworth, W. S., Jr.; Emmons, W. D. The utility of phosphonate
carbanions in olefin synthesis. J. Am. Chem. Soc. 1961, 83,
1733-1738.
(21) Kardos, J.; Kova´cs, I.; Blandl, T.; Cash, D. J. Modulation of
GABA flux across rat brain membranes resolved by a rapid
quenched incubation technique. Neurosci. Lett. 1994, 182, 73-
76.
(22) Ali, F. E.; Bondinell, W. E.; Dandridge, P. A.; Frazee, J. S.;
Garvey, E.; Girard, G. R.; Kaiser, C.; Ku, T. W.; Lafferty, J. J.;
Moonsammy, G. I.; Oh, H. J.; Rush, J. A.; Setler, P. E.; Stringer,
O. D.; Venlavsky, J. W.; Volpe, B. W.; Yunger, L. M.; Zirkle, C.
L. Orally active and potent inhibitors of γ-aminobutyric acid
uptake. J. Med. Chem. 1985, 28, 653-660.
Data Analysis. Duplicate measurements of binding and
transport were repeated in two to six experiments (N ) 2 to
N ) 6). Data are expressed as means ( SD and were analyzed
using one-way analysis of variances (ANOVAs, Origin ver. 6.1).
A value of P < 0.05 was considered significant. Concentration-
inhibition curves were fitted with the use of Origin ver. 6.1.
Acknowledgment. This work was supported by
grants 1/047 NKFP MediChem (Hungary), OTKA
T031753, Center of Excellence on Biomolecular Chem-
istry QLK2-CT-2002-90436 (EU), and Bolyai fellowship
to EÄ va Sza´rics, Ph.D. Professor Istva´n Simon (Institute
of Enzymology, Biological Research Center, Hungarian
Academy of Sciences) is gratefully acknowledged for his
advice on two-TM model building.
(23) Wang, G. J.; Chung, H. J.; Schnuer, J.; Pratt, K.; Zable, A. C.;
Kavanaugh, M. P.; Rosenberg, P. A. High affinity glutamate
transport in rat cortical neurons in culture. Mol. Pharmacol.
1998, 53, 88-96.
(24) Borden, L. A.; Murali Dhar, T. G.; Smith, K. E.; Weinshank, R.
L.; Branchek, T. A.; Gluchowski, C. Tiagabine, SK&F 89976-A,
CI-966, and NNC-711 are selective for the cloned GABA trans-
porter GAT-1. Eur. J. Pharmacol. 1994, 269, 219-224.
(25) Shimamoto, K.; Lebrun, B.; Yasuda-Kamatani, Y.; Sakaitani, M.;
Shigeri, Y.; Yumoto, N.; Nakajima, T. DL-threo-beta-benzyloxy-
aspartate, a potent blocker of excitatory amino acid transporters.
Mol Pharmacol. 1998, 53, 195-201.
References
(26) Dhar, T. G. M.; Borden, L. A.; Tyagarajan, S.; Smith, K. E.;
Branchek, T. A.; Weinshank, R. L.; Gluchowski, C. Design,
synthesis and evaluation of substituted triarylnipecotic acid
derivatives as GABA uptake inhibitors: identification of a ligand
with moderate affinity and selectivity for the cloned human
GABA transporter GAT-3. J. Med. Chem. 1994, 37, 2334-2342.
(27) Amara, S. G. Neurotransmitter transporters: new insights into
structure; function and pharmacology. Rev. Bras. Biol. 1996, 56,
5-19.
(28) O¨ zkan, E. D.; Lee, F. S.; Ueda, T. A protein factor that inhibits
ATP-dependent glutamate and γ-aminobutyric acid accumula-
tion into synaptic vesicles: Purification and initial characteriza-
tion. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4137-4142.
(1) Browne, T. R.; Holmes, G. L. Epilepsy. New Eng. J. Med. 2001,
344, 1145-1151.
(2) Treiman, D. M. GABAergic mechanisms in epilepsy. Epilepsia
2001, 42, 8-12.
(3) Kwan, P.; Sills, G. J.; Brodie, M. J. The mechanisms of action of
commonly used antiepileptic drugs. Pharmacol. Ther. 2001, 90,
21-34.
(4) Lo¨scher, W. New visions in the pharmacology of anticonvulsion.
Eur. J. Pharmacol. 1998, 342, 1-13.
(5) Laszto´czi, B.; Kova´cs, R.; Nyikos, L.; Kardos, J. A glutamate
receptor subtype antagonist inhibits seizures in rat hippocampal
slices. Neuroreport 2002, 13, 351-356.

Novel Secoergoline Derivatives
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 23 5629
(29) Veenhoff, L. M.; Heuberger, E. H. M. L.; Poolman, B. Quaternary
structure and function of transport proteins. Trends Biol. Sci.
2002, 27, 242-249.
(30) Kardos, J.; Blasko´, G.; Kerekes, P.; Kova´cs, I.; Simonyi, M.
Inhibition of [3H]GABA binding to rat brain synaptic membranes
by bicuculline related alkaloids. Biochem. Pharmacol. 1984, 33,
3537-3545.
(31) Kova´cs, I.; Laszto´czi, B.; Sza´rics, EÄ .; He´ja, L.; Sa´gi, G.; Kardos,
J. Characterisation of an uridine-specific binding site in rat
cerebrocortical homogenates. Neurochem. Int. 2003, 43, 101-
112.
(32) Kova´cs, I.; Simon, AÄ .; Sza´rics, EÄ .; Baraba´s, P.; He´ja, L.; Nyikos,
L.; Kardos, J. Cyclothiazide binding to functionally active AMPA
receptor reveals genuine allosteric interaction with agonist
binding sites. Neurochem. Int. 2004, 44, 271-280.
(33) Hill, D. R.; Bowery, N. G. 3H-baclofen and 3H-GABA bind to
bicuculline-insensitive GABAB sites in rat brain. Nature 1981,
290, 149-152.
(34) Pilpel, Y.; Ben-Tal, N.; Lancet, D. kPROT: A knowledge-based
scale for the propensity of residue orientation in transmembrane
segments. Application to membrane protein structure prediction.
J. Mol. Biol. 1999, 294, 921-935.
(35) McDonald, I. K.; Thornton, J. M. Satisfying hydrogen bonding
potential in proteins. J. Mol. Biol. 1994, 238, 777-793.
JM040809C