Structure and activity studies of glycine receptor ligands. Part 7. Structural remarks on arylidene-imidazoline-4-one glycinates and glycinamides

Article abstract of DOI:10.1016/S0022-2860(01)00584-1

To improve the physicochemical properties (lipophilicity) of the potential ligands of glycine binding site of NMDA receptor, esters and amides of glycine derivatives of arylidene-imidazolidine-4-ones were obtained. The analysis of their properties (X-ray crystallography, NMR spectral data and theoretical calculations) was performed with respect to the existence of the probable tautomeric forms. Their possible interaction with hypothetical active points of the receptor (taking into the account the model of glycine binding site of NMDA receptor) was discussed on the basis of simulation with IsoStar. program.

Full text of DOI:10.1016/S0022-2860(01)00584-1

Journal of Molecular Structure 597 02001) 73±81
www.elsevier.com/locate/molstruc
Structure and activity studies of glycine receptor ligands.
Part 7. Structural remarks on arylidene±imidazoline-4-one
glycinates and glycinamides
a,
b
Janina Karolak-Wojciechowska , Katarzyna Kiec-Kononowicz , Agnieszka Mrozek
a
*
Â
a
Ç
Institute of General and Ecological Chemistry, Technical University, Zwirki 36, PL 90-924 Lodz, Poland
bDepartment of Chemical Technology of Drugs, Faculty of Pharmacy, Jagiellonian University Medical College, Medyczna 9,
PL 30-688 Krakow, Poland
Received 29 December 2000; revised 9 March2001; accepted 9 March2001
Abstract
To improve the physicochemical properties 0lipophilicity) of the potential ligands of glycine binding site of NMDA receptor,
esters and amides of glycine derivatives of arylidene±imidazolidine-4-ones were obtained. The analysis of their properties
0X-ray crystallography, NMR spectral data and theoretical calculations) was performed with respect to the existence of the
probable tautomeric forms. Their possible interaction with hypothetical active points of the receptor 0taking into the account the
model of glycine binding site of NMDA receptor) was discussed on the basis of simulation with IsoStar. program. q 2001
Elsevier Science B.V. All rights reserved.
Keywords: Glycine receptor ligands; X-ray structure; Isostar
1. Introduction
improve the activity of this series of compound,
increasing the log P value by chemical modi®cation
was postulated. At the same time, we were working on
the three-dimensional model of ligand±receptor inter-
action, based on crystallographic data and molecular
modelling results [4]. Moreover, it was presumed
that an active form of imidazolone-glycines and
-glycinates [4±6] could be probably of the b-tautomer
form 0Scheme 2).
The carboxyl group seemed to be the interesting
region for the modi®cation. Consequently, ester II-2
was obtained and studied, including X-ray structure
analysis [4]. Recently, additional ester II-1 and series
of amide derivatives III and IV were prepared 0Table
1) and selected for pharmacological studies. Subse-
quently, structure and activity studies incorporating
newly synthesised derivatives, particularly amides
III and IV, were planned. Due to crystallisation
As a part of our researchprogram aimed at ®nding
new ligands of the glycine Ð NMDA binding site, the
glycine derivatives of arylidene±imidazoline-4-ones
were prepared [1,2]. Compounds belonging to type I
0Scheme1) have shown noticeable af®nity to that
receptor [1].
The preliminary analysis of structure-activity
relationship has shown that the active arylidene±
imidazolone glycines 0I) possess log P , 1. It is
known that compounds acting at CNS should possess
log P values of ca. 2 [3]. So, withthe objective to
* Corresponding author. Tel.: 148-42-6313122; fax: 148-42-
6313103.
E-mail address: jkarolak@ck-sg.p.lodz.pl 0J. Karolak-Wojcie-
chowska).
0022-2860/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-2860001)00584-1

74
J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
Scheme 1.
problems presented below the discussion is based on
the molecular modelling and quantum chemistry
calculations.
obtained as the result of I cyclisation upon in¯u-
ence of acetic anhydride [8]. The bicyclic deriva-
tive was cleaved withtwo fold excess of ammonia
[8] and different amines: 0un)substituted-benzyl
amines, 2-phenylethylamine, and 2-0N-morpholine)
ethylamine. The reactions were carried out in
heterogeneous conditions, in toluene suspensions.
Upon action of substituted ethylamines, the
deacetylated derivatives of group III were
obtained. In the reactions with benzylamines,
mainly the acetyl derivatives of group IV were
obtained. In the case of o-chlorobenzylamine both
types of compounds III-3 and IV-5 were isolated.
The purity of the compounds obtained was checked
by means of the thin layer chromatography. Their
structures were con®rmed on basis of elemental and
spectral analyses 01H-NMR).
2. Materials and methods
2.1. Chemical part
2.1.1. General procedure
As starting materials for the designed compounds,
chlorosubstituted arylidene derivatives of 2-thio-
hydantoin were used [1,7], which reacted with methyl
iodide giving the methylthioderivatives 0Scheme 3).
The methylthioderivatives were converted into
aminoacids I upon reaction withglycine in re¯uxing
acetic acid.
0a) Esters II were prepared according to previous
work [8] as a result of acid catalysed esteri®cation
of I.
0b) For the amides III, IV synthesis, as starting
material, the bicyclic acetyl imidazo[2,1-b]imida-
zoline-3,5-dione was used. This compound was
2.1.2. Experimental
Uncorrected mp's were determined on Wagner
1
Munz apparatus. H-NMR spectra were recorded on
Gemini 200 or Bruker VM250 in d6DMSO, chemical
shifts are reported in [ppm] values relative to internal
Scheme 2.

J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
75
Table 1
Chemical details and biological evaluation results for esters and amide derivatives of the types II±IV from Scheme 1
X
R2
R1
Inhibition 0%)
ASP
log P
2.09
Ref.
II-1
II-2
m-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
p-Cl
H
OC2H5
OC2H5
40
1.7
III
I
H
2.03
2 0.31
2.64
[8]
III-1
III-2
III-3
IV-1
IV-2
IV-3
IV-4
IV-5
IV-6
IV-7
H
NH±CH2CH2±Morph14
NH±CH2CH2±Ph6
NH±CH2CH2±o-Cl±Ph-
NH2
III
III
H
H
-
2.67
CH3CO
CH3CO
CH3CO
CH3CO
CH3CO
CH3CO
CH3CO
6.4
III
I
2 0.17
1.64
2.36
2.29
2.22
1.79
1.53
[8]
[8]
NH-CH2-Ph9.2
NH±CH2±p-Cl±Ph12
NH-CH2-m-Cl-Ph-
NH-CH2-o-Cl-Ph-
NH±CH2±p-F±Ph11
NH±CH2±p-OCH3±Ph4
I
III
I
III
III
Scheme 3.

76
J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
reference of TMS. Infrared spectra were measured
withFT IR 410 spectrometer 0Jasco) in KBr pellets.
The elemental analyses were performed for C, H, N
and results were within ^0.4% of theoretical values.
TLC was conducted on Al sheets, 0.2 mm layer of
silica gel 060F254, Merck). The developing systems
used were 0A) CHCl3/izopropanol/NH3 aq. 9:11:2;
0B) CHCl3/acetone: 1:1.
0382.84); mp 237±2408C 0DMF); yield 68%; RfꢀA† ˆ
1
0:70; H-NMR 0200 MHz) s ˆ 2:75 0t, J ˆ 7:70 Hz;
2H, CH2Ph); 3.33 0def t, J ˆ 6:81 Hz; 2H,
NHCH2CH2); 3.99 0s, 2H, HNCH2); 7.11±7.28 0m,
5H, H-20, H-30, H-40, H-50, H-60); 7.38 0d, J ˆ
8:33 Hz; 2H, H-3, H-5); 7.56 0br.s, 1H, NH); 8.09 0d,
J ˆ 8:00 Hz; 2H, H-20, H-60); 8.22 0def t, J ˆ 5:29 Hz;
1H, CONH); 10.35 [s, 1H, 09%) 1-NH; 10.71 0s, 1H,
091%) 3-NH); IR 0KBr) n: 3282 0NH), 1704 ꢀC4yO†;
1674 0CyO), 1652 0ArCHy), 1607, 1507, 1282, 1088,
864, 820, 682 cm21.
2.1.2.1. Ethyl-N-acetyl-N-[5-:Z)-:3-chlorobenzylidene)-
4-oxo-2-imidazolidinyl]glycinate II-1. The suspension
of N-[5-0Z)-03-chlorobenzylidene)-4-oxo-2-imidazoli-
dinyl] glycine [1] 02.8 g, 0.01 mol) in 100 ml of
etahnol was re¯uxed for 5 hwith1.5 ml of conc.
H2SO4. After cooling the precipitate was ®ltered out
and recrystallised from ethanol. C14H14N3O3 0307.73);
mp 231±2348C 0EtOH); Yield 44%; RfꢀA† ˆ 0:80; 1H-
NMR 0200 MHz) s ˆ 1:22 0t, J ˆ 7:09 Hz; 3H, CH3);
4.13 0s, 2H, HNCH2); 4.16 0q, J ˆ 7:10 Hz; CH2CH3);
6.34 0s, 1H, CHy); 7.33 0m, 2H, H-4, H-5); 7.82 0br.d,
J ˆ 4:64 Hz; 1H, H-6); 8.21 0br.s, 1H, H-2); IR 0KBr)
n: 3330, 3173 0NH), 1736 ꢀC4yO†; 1694 0COO), 1671
0ArCHy), 1599, 1507, 1219, 1120, 900, 783, 686 cm21.
2.1.2.4.
N-[5-:Z)-:4-chlorobenzylidene)-4-oxo-2-
imidazolidinyl]glycin-2-chlorobenzamide III-3. To
the DMF ®ltrate after IV-5 separation water was
added. The obtained precipitate was recrystallised
from DMF/H2O to give III-3. C19H16N4O2Cl2
0403.26); mp 228±2308C 0DMF/H2O); yield 41%;
1
RfꢀB† ˆ 0:41; H-NMR 0250 MHz) s ˆ 4:15 0d, J ˆ
4:75 Hz; 2H, NHCH2); 4.48 0d, J ˆ 5:75 Hz; 2H,
NHCH2CO); 6.39 0s, 1H, CHy); 6.99±7.07 0m, 1H,
H-500); 7.19±7.31 0m, 1H, H-400); 7.43 0d, J ˆ
8:75 Hz; 2H, H-30, H-50); 7.38±7.62 0m, 2H, H-300,
H-600); 7.76 0br.s, 1H, NHCH2); 8.17 0d, J ˆ 8:5 Hz;
2H, H-20, H-60); 8.69 0t, J ˆ 5:75 Hz; 1H,
NHCH2CO); 10.35 0s, 1H, 08%) 1-NH); 10.88 0s,
1H, 092%) 3-NH); IR 0KBr) n: 3301, 3085 0NH),
1706 ꢀC4yO†, 1662 0CyO), 1618 0ArCHy), 1515,
2.1.2.2.
N-[5-:Z)-:4-chlorobenzylidene)-4-oxo-2-
imidazolidinyl]glycin [2-:N-morpholinoethyl)]amide
III-1. To the stirred suspension of 1-acetyl-6-0Z)-04-
chlorobenzylidene) 2,3,5,6-tetrahydroimidazo-[2,1-
b]imidazole-3,5-dione [8] 00.303 g, 1 mmol in 10 ml
of toluene N-02-aminoethyl)-morpholine 00.260 g,
2 mmol) was added. The mixture was re¯uxed for
5 h. The precipitate was ®ltered off and
recrystallised from ethanol. C20H24N5O4Cl 0433.99);
mp 254±2568C 0EtOH); yield 43%; 1H-NMR
0250 MHz) s ˆ 2:33 0m, 6H, CH2N0CH2)2); 3.22
0m, 2H, NHCH2CH2); 3.52 0m, 4H, O0CH2)2); 3.94
0d, J ˆ 5:0 Hz; 2H, NHCH2CO); 6.28 0s, 1H, CHy);
7.35 0d, J ˆ 7:5 Hz; 2H, H-30, H-50); 7.54 0t, J ˆ
5:0 Hz; 1H, NHCH2); 8.05 0d, J ˆ 7:5 Hz; 2H, H-20,
H-60); 8.09 0t, 1H, CONH); 10.24 0br.s, 1H, 08%) 1-
NH); 10.70 0br.s, 1H, 092%) 3-NH); IR 0KBr) n: 3398,
3293 0NH), 2941, 2890, 2856, 2810 0CH2), 1726
ꢀCH4yO†; 1690 0CyO), 1661 0ArCHy), 1596, 1503,
1278, 1247, 1117, 1089, 818, 748, 683 cm21
.
2.1.2.5. N-acetyl-N-[5-:Z)-:4-chlorobenzylidene)-4-
oxo-2-imidazolidynyl]glycin-4-chlorobenzamide IV-
3. IV-3 was obtained as IV-5. C21H18N4O3Cl2
0445.29); mp 257±2588C 0DMF); yield 56%; 1H-
NMR 0200 MHz) s ˆ 2:40 0s, 3H, CH3CO); 4.32 0d,
J ˆ 5:74 Hz; 2H, NHCH2); 4.68 0s, 2H, NCH2); 6.83
0s, 1H, CHy); 7.15 0d, J ˆ 8:39 Hz; 2H, H-30, H-500);
7.25 0d, J ˆ 8:44 Hz; 2H, H-200, H-600); 7.35 0d, J ˆ
8:45; 2H, H-30, H-50); 8.12 0d, J ˆ 8:49 Hz; 2H, H-20,
H-60); 8.87 0t, J ˆ 5:50 Hz; 1H, NHCH2); 11.30 0s,
1H, 1-NH); IR 0KBr) n: 3332, 3287 0NH), 1721
ꢀC4yO†; 1688 0CyO), 1644 0ArCHy), 1554, 1464,
1213, 1090, 992, 818, 670 cm21
.
1259, 1117, 1084, 1037, 679 cm21
.
2.1.2.6. N-acetyl-N-[5-:Z)-:4-chlorobenzylidene)-4-
oxo-2-imidazolidinyl]glycin-3-chlorobenzamide IV-
4. IV-4 was obtained as IV-5. C21H18N4O3Cl2
0445.29); mp 236±2388C 0DMF); yield 28%; RfꢀB† ˆ
2.1.2.3. N-[5-:Z)-:4-chlorobenzylidene)-4-oxo-2-
imidazolidinyl]glycin[2-:phenylethyl)]amide III-2. III-
2 was obtained as described for III-1. C20H19N4O2Cl

J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
77
H-600); 6.81 0s, 1H, CHy); 7.15 0dd, J ˆ 7:5; 2:5 Hz;
2H, H-300, H-500); 7.32 0dd, J ˆ 8:5; 2:0 Hz; 2H, H-30,
H-50); 8.08 0dd, J ˆ 8:5; 2:0 Hz; 2H, H-20, H-60); 8.75
0t, J ˆ 7:50 Hz; 1H, NHCH2); 11.28 0s, 1H, 1-NH); IR
0KBr) n: 3308, 3286 0NH), 1726 ꢀC4yO†; 1690 0CyO),
1648 0ArCHy), 1550, 1250, 1212, 1084, 992, 825,
811 cm21.
1
0:65; H-NMR 0250 MHz) s ˆ 2:40 0s, 3H, CH3CO);
4.34 0d, J ˆ 5:80 Hz; 2H, NHCH2); 4.68 0s, 2H,
NCH2); 6.83 0s, 1H, CHy); 7.13±7.33 0m, 3H, H-400,
H-500, H-600); 7.20 0s, 1H, H-200); 7.35 0d, J ˆ 7:50 Hz;
2H, H-30, H-50); 8.12 0d, J ˆ 7:50 Hz; 2H, H-20, H-
60); 8.89 0t, J ˆ 6:2 Hz; 1H, NHCH2); 11.31 0s, 1H, 1-
NH).
2.2. Pharmacological part
2.1.2.7. N-acetyl-N-[5-:Z)-:4-chlorobenzylidene)-4-oxo-
2-imidazolidinyl]glycin-2-chlorobenzamide IV-5. The
suspension of 1-acetyl-6-0Z)-04-chlorobenzylidene)-
2,3,5,6-tetrahydroimidazo-[2,1-b]-imidazole-3,5-dione
[8] 01.51 g, 5 mmol) and 2-chlorobenzylamine 01.41 g,
10 mmol) in 50 ml of toluene was re¯uxed for 5 hr. The
precipitate was ®ltred off and recrystallised from DMF to
give IV-5. C21H18N4O3Cl2 0445.29); mp 248±2508C
2.2.1. Receptor binding determinations
Radioligand binding studies were performed
according to the method described by Grimwood et
al. [9]. Af®nities of tested compounds II-1, II-2; III-1,
III-2; IV-1±IV-3, IV-6, IV-7 used at concentrations
of 100 mM, for the glycine site of the NMDA receptor
were determined by displacement of the glycine site
antagonist [3H]-L-689,560 binding to rat cortex
0hippocampus membranes). The percent of inhibition
of the ligand was estimated and the results are
presented in Table 1.
1
0DMF); yield 30%; H-NMR 0250 MHz) s ˆ 2:47 0s,
3H, CH3CO); 4.46 0d, J ˆ 5:50 Hz; 2H, CH2NH); 4.80
0s, 2H, CH2CO); 6.91 0s, 1H, CHy); 7.03 0dt., J ˆ
7:5; 1:25 Hz; 1H, H-500); 7.27 0dt, J ˆ 7:75; 1:75 Hz;
1H, H-400); 7.38±7.48 0m, 2H, H-300, H ˆ 600); 7.46 0d,
J ˆ 8:50 Hz; 2H, H-30, H-50), 8.23 0d, J ˆ 8:75 Hz;
2H, H-20, H-60); 8.95 0t, J ˆ 5:0 Hz; 1H, NHCH2);
11.38 0s, 1H, 1-NH); IR 0KBr) n: 3317, 3073 0NH),
2981, 2930 0CH2), 1728 ꢀC4yO†; 1692 0CyO), 1644
0ArCHy), 1548, 1213, 1085, 992, 824, 752 cm21.
2.2.2. Anticonvulsant assays
All animal anticonvulsant and neurotoxicity assays
were conducted by the Antiepileptic Drug Develop-
ment Program, Epilepsy Branch, Neurological Dis-
orders Program, National Institutes of Neurological
and Communicative Disorders and Stroke 0NINCDS)
according to the testing procedures described earlier
[16,17]. Phase I of anticonvulsant screening project
0ASP), involved three tests: maximal electroshock
seizure 0MES), subcutaneous pentylenetetrazol
0ScMet), and neurological toxicity 0Tox). All
compounds apart from III-3 were injected intraperi-
toneally into mice as suspensions in 0.5% methyl-
cellulose. Small groups of animals were used 01±8)
and dose levels were 30, 100 and 300 mg/kg. These
data are presented in Table 1. The classi®cations were
as follows: I: anticonvulsant activity at 100 mg/kg or
less; II: anticonvulsant activity at doses larger than
100 mg/kg; III: compound inactive at 300 mg/kg.
2.1.2.8. N-acetyl-N-[5-:Z)-:4-chlorobenzylidene)-4-
oxo-2-imidazolidinyl]glycin-4-¯uorobenzamide IV-6.
IV-6was obtained as IV-5. C21H18N4O3ClF 0428.84);
mp 255±2588C 0DMF); yield 61%; RfꢀA† ˆ 0:85; 1H-
NMR 0200 MHz) s ˆ 2:40 0s, 3H, CH3CO); 4.32 0d,
J ˆ 5:76 Hz; 2H, NHCH2); 4.68 0s, 2H, NCH2); 6.84
0s, 1H, CHy), 6.95 0dt, J ˆ 9:00; 2:33 Hz; 2H, H-200,
H-600); 7.27 0dt, J ˆ 9:00; 2:23 Hz; 2H, H-300, H-500);
7.35 0d, J ˆ 8:62 Hz; 2H, H-30, H-50); 8.12 0d, J ˆ
8:52 Hz; 2H, H-20, H-60); 8.87 0t, J ˆ 5:96 Hz; 1H,
NHCH2); 11.32 0s, 1H, 1-NH); IR 0KBr) n: 3321
0NH), 1729 ꢀC4yO†; 1694 0CyO), 1664 0ArCHy),
1548, 1213, 994, 822, 670 cm21
.
2.1.2.9.
N-acetyl-N-[5-:Z)-:4-chlorobenzylidene)-4-
2.3. Computational procedure
oxo-2-imidazolidinyl]glycin-4-methoxybenzamide IV-
7. IV-7 was obtained as IV-5. C22H21N4O4Cl 0440.88);
All molecules were modelled based on the crystal-
lographic structure of II-2 [4] and roughly optimised
using pcmodel.6 program [10]. MO calculations were
carried out withAM1 method [10] using MOPAC
program 0version 6.0) [11] in water environment
1
mp 248±2508C 0DMF); yield 68%; RfꢀA† ˆ 0:85; H-
NMR 0250 MHz) s ˆ 2:37 0s, 3H, CH3CO); 3.64 0s,
3H, OCH3); 4.24 0d, J ˆ 7:50 Hz; 2H, NHCH2); 4.64
0s, 2H, NCH2); 6.69 0dd, J ˆ 7:5; 2:5 Hz; 2H, H-200,

78
J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
esters 0II-1 and II-2). At the same time, the activities
of both m-Cl-substituted species are almost 40 units
0%) higher than that of p-Cl ones. Remaining
compounds 0III and IV amides) are practically
inactive, independently of log P value 0averaged
inhibition ,8%).
During the in vivo anticonvulsant tests 0Phase I
ADD Program), all compounds in doses up to
300 mg/kg were devoid of neurotoxicity. In the
group of esters 0II), p-chloro substituted II-2 was
evaluated as active and classi®ed to the class I ASP.
Only three amides, namely: benzyl, p-chlorobenzyl,
and o-chlorobenzyl amides were rated among the
class I ASP. The other compounds were found to be
inactive.
3.1. Imidazolone±glycines tautomerisation
Fig. 1. Scattergram for log D vis log P.
Three resonance structures for imidazolone±
glycine derivatives are possible 0Scheme 2). It has
been shown [4,5] that the concentration of b-tautomer
is essential for the inhibition to the glycine receptor.
In order to understand the relationship between struc-
ture and af®nity to the receptor based on the model
presented previously [4], this concentration should be
0dielectric constant 78.4). The values of log P were
calculated by means of pallas 0version 1.2) program
[12].
3. Results and discussion
1
estimated tentatively. This can be achieved via H-
The differences in chemical nature of all the
investigated compounds 0Scheme 1) become clear
after looking at Fig. 1. So, all arylidene±imidazolone
glycine derivatives studied until now are classi®ed in
two groups. The ®rst group incorporates acidic
glycines I 0set A) withvariously substituted aromatic
ring, the second one Ð glycinates II, and amides III
and IV 0set B). Synthesis and pharmacological test
results of set B were a direct outcome of our search
for modi®ed arylidene±imidazolone glycines 0set A)
withincreased log P values [3]. However, indepen-
dent of log P value, the compounds from set B are
explicitly less interesting than parent glycines I. In
the in vitro binding assays, all modi®ed compounds
of series B possess lower inhibition to the receptor,
only the ester II-1 0from family of m-Cl arylidene
derivatives) showed noticeable af®nity to the receptor
040% of inhibition), yet, it was still lower than that of
the parent molecule I-1 090% of inhibition)[1]. Hence,
just two acid±ester pairs are insuf®cient for general
conclusion, the activities of acids 0Table 1, I-1 and
I-2) are almost exactly 50 units 0%) higher than that of
NMR spectra analysis and/or based on theoretical
calculations. It is also of special interest to compare
the results from both analyses.
Glycinamides of type III may exist in all three
tautomeric forms 0Scheme 2). However, the spectral
analyses revealed that these derivatives exist 0in
DMSO) in solution mainly in c-form. In the 1H-
NMR spectra, the double set of signals was related
to protons attached to the imidazolone ring nitrogens
at positions 1-NH and 3-NH. The singlets at 10.24±
10.35 ppm were related to protons attached to the
nitrogen at position 1 of the imidazolone ring 0b-
tautomer), and the singlets at 10.70±10.88 ppm Ð
to more acidic protons attached to the nitrogen at
position three of the imidazolone ring 0c-tautomer).
The signals of protons from the exocyclic NH groups
of b-tautomers were very weak and hidden under the
aromatic proton signals. The analysis of the signals'
integration of 1-N and 3-N protons enabled us to
conclude that in the described amides the c-tautomers
were observed in predominance to the b-ones 0ca. 9:1
ratio) 0Table 2), contrary to the previously described

J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
79
Table 2
Imidazolone-glycines tautomerization
HfÐheat of formation and pi for tautomers a, b and c according to AM1 method
1H-NMR spectra analysis
Comp.no
a
b
c
II-2
Dynamic state
Dynamic state
1
Dynamic state
9
III-1
IV-1
, 1
±
Dominant
Not existing
Hf 0kJ/mol)
271.38
8.97
pi 0%)
Hf0kJ/mol)
268.77
11.45
pi 0%)
23
Hf0kJ/mol)
267.63
12.73
pi 0%)
II-2
62
62
-
15
14
65
III-1
IV-1
24
35
-
256.05
257.66
diphenyl glycinamides [2]. Glycinamides of type IV
may exist only in the two tautomeric forms: b and c
0Scheme 2). Only single signals for N±H protons were
observed at the range 11.28±11.38 ppm. These
signals were observed in even more deshielded region
as more acidic protons of c tautomers of amides III
are present 0the signal range observed was 10.70±
10.88 ppm). We presume that the only tautomers,
which exist in DMSO solution, are b tautomers,
where the 1-N±H protons are being under anisotropic
in¯uence of carbonyl group attached to the exocyclic
nitrogen causing down®eld shift of these signals.
Suchpredominance of one of the tautomeric forms
was not observed for the esters of type II. Th e N±H
protons were not detected, indicating the dynamic
behaviour of the structures in the solution for structure
II-1 and presumably only the c tautomer of p-chloro
substituted II-2 existed in the solution 0H±N proton
observed at 10.95 ppm).
To calculate the percentage 0pi) of molecules in a
given energetic state, non-degenerate Boltzmann
distribution was used 0second part of Table 2). The
heat of formation Hf for eachtautomer was calculated
using the semiempirical AM1 method [11] in a simu-
lated water environment 0dielectric constant equals
78.4 units). As all studied amides 0III and IV) belong
to p-chloro-arylidene derivatives, also the calculations
for ester II-2 were performed for compatibility. From
the series of amides, the molecules III-2 and IV-1
0Table 1) were selected for further modelling. The
calculations for derivatives from set B 0II and III)
showed that the a tautomer is the most probable
form in the aqueous environment 0,62%). According
to that calculations, there are about 23% of b tautomer
and only 15% of c. In amides of type IV 0only the two
tautomeric forms are taken into account), the c
tautomer is the most stable and the ratio b/c is about
1:2. Bothtautomeric forms of N-acylated amide IV-1
are stabilised by intramolecular N±H´´´O hydrogen
bonds 0Fig. 2).
The theoretical calculation results for glycine
derivatives of all types 0see Table 2) do not exactly
1
agree with the conclusions based on the H-NMR
spectra analysis. These results 0at physiological
temperature 378C) could be a direct outcome of
1
different character of environment: H-NMR spectra
were measured in DMSO aprotonic solvent, while the
calculations were performed in H2O being a highly
protonic one. However, in the equilibrium state in
the solution all tautomeric forms may be present
and, thus, none of them could be neglected. Therefore,
this permits to conclude that the tautomers, essential
for binding to the receptor, are coming into existence
dynamically, during binding to the receptor. So, it is
important to consider the other additional conditions,
which divert signi®cantly the inhibition to the
receptor of the compounds from sets A and B.
3.2. Structure-activity relationships
Molecules from sets A and B differ from the view-
point of responsibility for coulombic interactions
between receptor and ligand. Therefore, considering
only b tautomeric forms, we were able to limit the
analyses of partial charges 0from AM1 approximation
in aqueous solution) just to the carboxylic group

80
J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
Fig. 2. Two tautomeric forms of N-acylated amide IV-1. Doted lines mark intramolecular H-bonds.
residue: two oxygens for I and II, and/or oxygen and
nitrogen for III and IV, respectively 0Table 3).
con®rmed areas of non-bonded contacts for COO2,
COOR, and CONH2 groups withH±X ꢀXyO; N†
[13±15]. This representative scattegrams 0Fig. 3 a±
c) were created using IsoStar. program [12±15].
Areas of contact are explicitly different.
For ionic I-2, charges of about 20.7e are located on
bothoxygens. However, for II-2, nature of oxygens is
not identical having charges equal 20.46e and
20.30e, respectively. For III-2 and IV-1, charges
are signi®cantly lower than for I-1 and they equal,
on average, 20.53e on oxygen and 20.30e on
amide nitrogen 0Table 3). The proton donoring object
0NH 2 from CONH±R group) Ð in the coulombic
attraction area of all amides Ð additionally hampered
the ligand±receptor interactions. Moreover for IV,
intramolecular H-bond 0N±H´´´O) hinders the proton
donoring group capable of forming N±H´´´A bond
with the receptor 0Fig. 2). Presumably, the af®nity to
the receptor of imidazoline±glycine derivatives
depends mainly on the negative charge over the
carboxylic group residue. Every substitution,
lowering this charge and blocking ionisation, restrains
the derivative contact with another chemical objects.
These differences, intuitively clear and con®rmed
by the value of charges 0Table 3), are visualised by
scattergams, which summarise crystalographically
4. Conclusions and recommendations for future
work
The obtained III and IV amides 0Table 1) are much
less active in a series of p-Cl±arylidene±imidazoline±
glycine derivatives than other compounds of the same
origin. The p-Cl±arylidene±imidazoline±glycine I-2
was chosen as the leading structure, because it has
shown the af®nity to the glycine binding site in
NMDA receptor and was active in MES test 0ASP
class II); up to dosage of 300 mg/kg, and contrary to
m-chloro derivative I-1 it was devoid of neurotoxic
effects. Nevertheless, their m-Cl analogue II-1 would
be also less active than parent acid. This means that
there is the necessity to verify the direction in
searching for the new improved ligands with glycine
binding site in NMDA receptor.
We intend to obtain and test imidazoline±glycine
derivatives withvarious aromatic substituents
attached over the region of liphophilic pocket. So
far, just one diphenyl analogue was found to be
inactive [5]. At the same time, we are going to investi-
gate other imidazoline±aminoacids. It should be
emphasised that the research direction proposed here
is the consequence of the thorough analysis of the
Table 3
The partial charges [e] in tautomer b at oxygens from COO residue
for I and II or at oxygen and nitrogen from CON residue from III
and IV
Partial charge at
I-2
II-2
III-1
IV-1
O1
20.70
20.69
20.46
20.30
20.53
20.30
20.52
20.30
O2 0or N)

J. Karolak-Wojciechowska et al. / Journal of Molecular Structure 597 :2001) 73±81
81
Fig. 3. Surfaces of crystalographically registered interactions with H±X ꢀXyO; N† for: 0a) COO2; 0b) COOR; 0c) CONH2.
0Eds.), A textbook of Drug Design and Development, 2nd
model of interaction between arylideno-imidazoline±
glicynes and the glycine binding site of NMDA
receptor created by us [4].
ed., 1996, p. 35.
Â
[4] J. Karolak-Wojciechowska, A. Mrozek, K. Kiec-Kononowicz,
J. Mol.Struct. 516 02000) 113.
[5] J. Karolak-Wojciechowska, A. Mrozek, W. KsiaÎzÇek, K. Kiec-
Kononowicz, J. Handzlik, J. Mol.Struct. 447 01998) 89.
Acknowledgements
Â
Â
[6] A. Drabczynska, J. Karolak-Wojciechowska, K. Kiec-
Kononowicz, Pol. J. Chem. 73 01999) 783.
[7] F.A. Shalaby, H.A. Daboun, S.S.M. Boghdadi, Z. Naturforsch
29b 01974) 99.
The support of Polish State Committee for Scien-
ti®c Research, project 4 P05F 007 17, is greatly
acknowledged.The authors wish to acknowledge
Professor P.Leeson and Dr S.Grimwood 0Merck
Sharp and Dohme Research Laboratories, Harlow,
UK) for the ®nancial support and providing the tests
of inhibition of [3H]-L-689.560 binding to rat brain
membranes. We are also grateful to Professor J.P.
Stables for providing pharmacological data through
the Antiepileptic Drug Development Program,
National Institutes of Health.
Â
[8] K. Kiec-Kononowicz, J. Karolak-Wojciechowska, Arch.
Pharm. 0Weinheim) 328 01995) 119.
[9] S. Grimwood, A.M. Moseley, M.W. Carling, P.D. Leeson,
A.C. Foster, Mol. Pharmacol. 41 01992) 923.
[10] PCMODEL.6 Molecular Modeling Software for the IBM PC,
Serena Software, Bloomington, 1996.
[11] MOPAC Programme, Version 6, QCPE No. 455; Department
of Chemistry, Indiana University, Bloomington, IN 1990.
[12] PALLAS for Windows 1.2, CompuDrug Chemistry Ltd, 1995.
[13] F.H. Allen, J. Davies, J.E. Galloy, J.J. Johnson, O. Kennard,
C.F. Macrae, E.M. Mitchell, J.M. Smith, D.G. Watson, J.
Chem. Inf. Comput. Sci 31 01991) 187.
[14] F.C. Bernstein, T.F. Koetzle, G.J.B. Williams, E.F. Meyer, M.
Dbrice, J.R. Rodgers, O. Kennard, T. Shimanouchi, M.
Tasumi, J. Mol. Biol. 112 01977) 535.
References
Â
[1] K. Kiec-Kononowicz, J. Karolak-Wojciechowska, J. Handzlik,
[15] http://www.ccdc.cam.ac.uk/prods/isostar.html.
[16] R.L. Krall, J.K. Penry, B.G. White, H.J. Kupferberg, E.A.
Swinyard, Epilepsia 19 01978) 400.
Acta Pol. Pharm.-Drug Res. 55 01998) 381.
[2] K. Kiec-Kononowicz, J. Karolak-Wojciechowska, Acta Pol.
Pharm.-Drug Res. 55 01998) 389.
[17] R.J. Porter, J.J. Cereghino, G.D. Gladding, B.J. Hessie, H.J.
Kupferberg, B. Scoville, B.G. White, Cleve Clin. Q 51 01984)
293.
[3] U. Hacksell, Structural and Physicochemical Factors in Drug
Action, in: P. Krogsgaard-Larsen, T. Liljefors, U. Madsen