Synthesis, and In Vitro and In Vivo Muscarinic Pharmacological Properties of a Series of 1,6-Dihydro-5(4H)-pyrimidinone Oximes

Article abstract of DOI:10.1016/S0968-0896(98)00074-1

A series of 1,6-dihydro-5-(4H)-pyrimidinone oxime derivatives I was synthesized (Scheme 1, Tables 1 and 2) and tested for muscarinic activity (Table 3) in receptor binding assays using <3H>-oxotremorine-M (Oxo-M) and <3H>-pirenzepine (Pz) as ligands.Poten

Full text of DOI:10.1016/S0968-0896(98)00074-1

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 1403±1420
Synthesis, and In Vitro and In Vivo Muscarinic
Pharmacological Properties of a Series of
1,6-Dihydro-5-(4H)-pyrimidinone Oximes
Ralf Plate,a,* Marc J. M. Plaum,a Peter Pintar,a Christan G. Jans,a
Thijs de Boer,b Fred A. Dijcks,b Ge Ruigtb and John S. Andrewsb
aDepartment of Medicinal Chemistry, Scienti®c Development Group, N. V. Organon, P. O. Box 20, 5340 BH Oss, The Netherlands
bDepartment of Neuropharmacology, Scienti®c Development Group, N. V. Organon, P. O. Box 20, 5340 BH Oss, The Netherlands
Received 12 December 1997; accepted 26 January 1998
AbstractÐA series of 1,6-dihydro-5-(4H)-pyrimidinone oxime derivatives I was synthesized (Scheme 1, Tables 1 and 2)
and tested for muscarinic activity (Table 3) in receptor binding assays using [3H]-oxotremorine-M (Oxo-M) and [3H]-
pirenzepine (Pz) as ligands. Potential muscarinic agonistic or antagonistic properties of the compounds were deter-
mined using binding studies that measured their potencies to inhibit the binding of Oxo-M and Pz. Preferential inhi-
bition of Oxo-M binding was used as an indicator for potential muscarinic agonistic properties; this potential was
con®rmed in functional studies on isolated organs. The series produced a wide range of active compounds with di€er-
ing degrees of selectivity in M1, M2, and M3 functional models. Several compounds that have mixed agonist/antagonist
pro®les were able to reduce cholinergic-related cognitive impairments in models of mnemonic function. Substitutions
(I, e.g. R2 or R3=Me) at the 1,6-dihydro-5-(4H)pyrimidine ring disrupted binding and ecacy, whereas systematic
variation of the oximes substituent R1 resulted in various degrees of potency and selectivity dependent on the nature of
the substitution. # 1998 Elsevier Science Ltd. All rights reserved.
Introduction
e€ects, although ecacy is still less than optimal.9
The ®nding that the degeneration of cholinergic neurons
in the forebrain of patients su€ering from Alzheimer's
disease correlated with the reduction in cognitive func-
tion led to the formulation of the cholinergic de®cit
hypothesis.1,2 Evidence suggesting a greater loss of pre-
synaptic rather than postsynaptic receptors3,4 has led to
the current treatment strategy concentrating on stimu-
lation of the postsynaptic receptors either by increasing
the availability of acetylcholine via cholinesterase inhi-
bitors such as tacrine,5 or by direct stimulation with
cholinergic agonists. To date, the cholinesterase inhibi-
tors have been shown to improve some aspects of cog-
nitive performance; however, ecacy has been poor and
side e€ects problematic.5±8 The recent introduction of
donepezil has led to a reduction in some but not all side
Similarly, studies involving cholinergic agonists have
been marred by the side e€ect pro®le of the agonists
(e.g. oxotremorine IV10 and RS 86 V11), although some
studies have indicated that careful titration of the dose
may lead to signi®cant improvements in function with-
out compromising side e€ects.12,13
There are at least ®ve known cholinergic receptors and
it is apparent that these receptors regulate a wide range
of peripheral and central functions apart from cogni-
tion.14 Therefore, it is not surprising that studies invol-
ving agonists acting at all subtypes produce many
undesirable e€ects. Thus, interest has focused on the
e€ects of selective muscarinic agonists as a means to
improve the therapeutic potential of cholinergic agents,
and especially for compounds with M1 and M1/M3 ago-
nist properties;15 (see ref. 16 for a discussion of some
potential problems with this approach). Recent animal
studies have shown positive results in tests of learning
Key words: Muscarinic agonist; muscarinic antagonist; 1,6-
dihydro-5-(4H)-pyrimidinone oximes; memory.
*Corresponding author. Fax: 31 412 662 546.
0968-0896/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00074-1

1404
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
and memory with some of these compounds;17±19 how-
ever, the full results of clinical trials are still awaited for
the majority of these compounds. The putative M1
selective compound xanomeline has been reported to
have some bene®cial e€ects in patients, however e-
cacy, and side e€ects are still problematic.20
quinuclidine heterocyclic moiety by a 1,6-dihydro-(4H)-
pyrimidine22 based on molecular modelling considera-
tions. Analogous molecular modelling considerations
have been described by Messer et al.23 Thus, in this
paper we describe the synthesis of 1,6-dihydro-(4H)pyr-
imidine derivatives leading to a series of compounds I,
which are potent and ecacious muscarinic agonists
(Fig. 1, Table 1).
The aim of the experiments was to design and test a
series of compounds for selective e€ects at speci®c mus-
carinic receptor subtypes in vitro, and then to investi-
gate the central and peripheral e€ects in vivo, including
e€ects on memory. Previous studies had indicated that
compounds with agonist properties at M1, and to a les-
ser extent M3, receptors may have an acceptable
pharmacological pro®le; no e€ects or antagonist e€ects
at M2 receptors may also be desirable. The ultimate
object of these studies being to develop compounds that
preferentially a€ect the central nervous system and
show lesser peripheral side e€ects than the agonists
reported to date.
Chemistry
Synthesis of 1,6-dihydro-5-(4H)-pyrimidinone oxime
derivatives
and
1,4,5,6-tetrahydro-(N)-alkoxy-5-
pyrimidinamine derivatives22 Condensation of the O-
alkylhydroxylamine derivatives with diaminoacetone
gave the oxime derivatives 1 (Scheme 1). Alternatively,
we prepared these oximes 1 in a three step procedure
commencing with 1,3-dichloroacetone. O-Alkyl-
hydroxylamines were, if necessary, prepared according
to methods described in the literature.24,25 The inter-
mediate 1,3-dichloroacetone oxime derivatives 2 were
treated with potassium phthalimide and gave the 1,3-
diphthaloyl acetone oxime derivatives 3. Deprotection
of the amino groups using the sodium salt of hydro-
xylamine as reagent gave the diamino acetone oxime
derivatives 1.
A useful starting point for the design of the title series
was provided by the consideration of the muscarinic
activity of arecoline (II), however, the therapeutic bene-
®t of this compound is limited because of short duration
of action. Recently, we described the synthesis and test-
ing of a series of quinuclidine oxime derivatives III21
having an oxime function as a bioisosteric mimic of
arecoline's ester function. Compounds in this series
produced weak agonists selective for the M3 receptor.21
Ring closure of the diamino acetone oxime derivatives 1
using trimethyl orthoformate as the reagent gave the
desired 1,6-dihydro-5-(4H)-pyrimidinone oxime deriva-
tives 4±34, which were converted into their hydro-
chloride or maleic acid salts (overall yields and some
analytical data are given in Table 1). Treatment of 2
Our ongoing search for more potent and selective mus-
carinic agonists, starting from quinuclidine oxime deri-
vatives III,21 led to the bioisosteric replacement of the
Figure 1.

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1405
with trimethyl orthoacetate as reagent gave the corre-
sponding 2-methyl 1,6-dihydro-5-(4H)-pyrimidinone
oxime derivative 35.
desired 3,4,5,6-tetrahydro-N-hydroxy-5-pyrimidinamine
derivatives 38, 39, and 40, respectively (Scheme 2,
Table 1).
In an analogous fashion 1,3-diamino-l-methylacetone26
was treated with O-ethyl-hydroxylamine to give 36.
Finally, the 4-methyl-1,6-dihydro-5(4H)-pyrimidinone
oxime 37 was obtained by cyclisation of the 1,3-
diamino-1-methyl oxime derivative 36 using trimethyl-
orthoformate as reagent.
Results
The physical properties of the derivatives are shown in
Tables 1±3. These show the receptor binding properties
of the series. The ratio of antagonist/agonist binding has
been suggested15 to give an indication of agonist poten-
tial: a large ratio indicates an agonist (viz. Table 3 ratio
Pz/Oxo-M). This was explored further in various func-
tional in vitro models (Table 4). It appears that the
Reduction of the oxime function of 4, 5, and 6 using
trimethylamine borohydride complex (TMA BH3) in
methanolic hydrochloride as the reagent gave the
Table 1. 1,6-Dihydro-5-(4H)-pyrimidinone oximes
Compd
R1
R2 R3 mp (ꢀC)
Yield (%)
Molecular Formula
.
C4H7N3O HCl
4
H
methyl
ethyl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
168
170
164
147
145
172
199
178
158
165
136
129
176
151
140
144
205
176
167
146
188
186
146
189
168
187
176
176
168
147
137
135
153
54
48
58
45
49
58
49
40
57
57
16
32
8
.
5
C5H9N3O HC1
.
C6H11N3O HCl
6
.
7
propyl
butyl
C7H13N3O HCl
.
C9H15N3O HCl
8
.
9
sec-propyl
tert-butyl
C7H13N3O HCl
.
C8H15N3O HCl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
37
.
1-methyl-1-propyl
methylcyclopropyl
propenyl
C8H15N3O HCl
.
C8H13N3O HCl
.
C7H11N3O HCl
.
(E)-2-butenyl
(Z)-2-butenyl
C8H13N3O HCl
.
C8H13N3O HCl
.
1-methyl-2-propenyl
2-methyl-2-propenyl
3-methyl-2-butenyl
(E,E)-2,4-hexendienyl
2-propynyl
C8H13N3O HCl
.
C8H13N3O HCl
9
.
44
37
51
35
40
52
58
32
42
10
60
26
50
7
C9H15N3O HCl
.
C10H15N3O HCl
.
C7H9N3O HCl
.
2-butynyl
2-pentynyl
2-hexynyl
C8H11N3O HCl
.
C9H13N3O HCl
.
C10H15N3O HCl
.
C9H11N3O HCl
(E)-2-penten-4-ynyl
(E)-3-methyl-2-penten-4-ynyl
(Z)-3-methyl-2-penten-4-ynyl
1-methyl-2-propynyl
benzyl
.
C
C
10H13N3O HCl
.
10H13N3O HCl
.
C8H11N3O HCl
.
11H13N3O HCl
C
C11H12C1N3O HCl
.
(2-chlorophenyl)methyl
(2-¯uorophenyl)methyl
(2-tri¯uoromethylphenyl)methyl
3-phenyl-2-propynyl
3-(3-methoxyphcnyl)-2-propynyl
(E)-3-phenyl-2-propenyl
2-propenyl
.
11H12FN3O HCl
C
C12H12F3N3O HCl
.
.
55
44
39
32
45
C13H13N3O HCl
.
14H15N3O2 HCl
C
C13H15N3O C4H4O4
.
.
C8H13N3O HCl
.
C7H13N3O HCl
ethyl

1406
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
compounds with the highest ratios tend to have full
agonist properties in each of the subtype selective pre-
parations (Table 4). However, compounds with lower
ratios may also be agonists but this is then expressed as
partial agonistic activity or, alternatively, muscarinic-
subtype selective agonist e€ects in di€erent tissue pre-
parations.
Table 2. Tetrahydro-N-hydroxy-5-pyrimidinamines
Compd
R1 R2 R3 MP (ꢀC) Yield Molecular Formula
There is a signi®cant correlation between the Pz/Oxo-M
binding ratios and potencies as agonists in M1 (r=0.69,
p<0.05), M2 (r=0.823, p<0.01) and M3 (r=0.731,
p<0.01) mediated responses in isolated organ prepara-
tions (data from Tables 3 and 4). For the M1 and
M2 preparations this correlation was also evident for
.
C4H9N3O 2HCl
38
39
40
H
H H
221
214
178
18
46
76
.
methyl H H
ethyl H H
C5H11N3O 2HCl
.
C6H13N3O 2HCl
Table 3. Receptor binding anity and anity ratios for
assays using agonist, oxotremorine-M (Oxo-M), and antagonist
pirenzepine (Pz) ligands (for further details see Experimental)
Compd
Oxo-M KI
(mM)
Pz KI
(mM)
Ratio
Pz/Oxo-M
4
5.855
0.0698
0.0122
0.162
0.0631
0.115
0.5205
0.895
0.1807
0.0465
0.0396
0.0389
0.199
0.33
126
21.52
1,315.51
1,397.38
7.04
5
91.8667
17.0667
1.14
6
7
8
7.94
125.83
102.6
12.47
6.17
9
10
11.795
6.493
5.52
11
12
9.85
54.52
443.01
234.85
378.12
63.32
36.06
201.49
1.88
13
14
20.6
9.3
15
16
14.69
12.6
11.9
17
18
0.0202
0.992
0.0226
0.0096
0.0097
0.173
0.0631
0.0074
0.0019
0.112
0.107
0.159
0.0579
0.388
0.8905
1.285
5.16
4.07
1.86
19
20
15.95
13.6
8.38
705.75
1,413.72
863.92
41.50
125.83
643.44
894.86
149.11
55.61
10.38
32.30
4.30
21
22
23
24
7.18
7.94
Scheme 1. Synthesis of dihydropyrimidine oxime derivatives.
(A) O=C(CH2NH2)2, MeOH, 20±96 h re¯ux; (B)
O=(CH2Cl2)2, MeOH, 2 h, 40 ꢀC; (C) PhtNK, DMF, 4 h,
100 ꢀC; (D) NaONH2, MeOH, 2 h, rt; (E) HC(OCH3)3, MeOH,
20 h, re¯ux; (F) CH3C(OCH3)3, MeOH, 20 h re¯ux; (G)
H2NH2C(C=O)CH(CH3)NH2MeOH, 20 h re¯ux.
25
26
4.77
1.742
16.7
5.95
27
28
29
30
1.65
1.87
31
32
1.67
5.2845
4.135
15.215
4.4795
71.425
11.6
5.93
3.22
33
34
2.95
2.11
35
37
2.125
3.085
11.1
23.15
1.05
38
39
6.4
500
101
1.58
0.25
38.2
78.13
298.82
200
40
0.338
0.0079
0.00056
0.0148
Arecoline II
Oxotremorine IV
Carbachol VI
Scheme 2. Synthesis of tetrahydro-N-alkoxy-5-pyrimidinamine
derivatives. (H) Trimethylamine±borohydride complex,
HCl/MeOH, 2 h, rt.
446
2581

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1407
measures of ecacy (M1, r=0.902, p<0.01; M2,
r=0.833, p<0.01) as well as for potency, for M3, this is
true only for potency.
peripheral and central (cognitive or EEG) cholinergic-
related activity.
Although compounds that are strong agonists at the
M1, M2, and M3 receptors are clearly identi®ed in the
models indicative of acetylcholine-related side e€ects (e.g.
salivation, mydriasis) these compounds are not as uni-
formly active in the cognitive models (Table 5, Figs 2 and
3). Generally speaking, full agonists (high ecacy and
potency at the M1, M2, and M3 receptors) are generally
capable of reversing cholinergic-induced de®cits in the
The series contain compounds with a wide range of
properties: full, partial and selective agonists are all
represented. M3 activity is the most common property,
and all chemical modi®cations tend to reduce M1 and
M2 agonist properties before removing M3 activity in
vitro. Similar e€ects are evident in vivo. Tables 5 and 6
illustrate the activity in vivo in various models of
Table 4. E€ects of series members on isolated central and peripheral organs exhibiting receptor selective e€ects (for further details see
Experimental)
M1 Hippocampal slice
pD2
M2 Rat left atrium
a
M3 Guinea Pig ileum
a
Compd
a
pD2
pA2
pD2
pA2
4
nt
4.8
5
0.84
0.89
0
5.61 (0.07)
6.15 (0.07)
0.93
0.95
5.47 (0.1)
6.73 (0.03)
1.25
1.27
1.0
6
5.5
7
<4
<4
4.8
4.86 (0.04)
5.12 (0.05)
5.2 (**
)
8
0
4.61 (0.02)
5.5 (0.01)
0.73
1.22
9
10
0.49
0
5.1(0.003)
0.71 4.74 (0.07)
4.63
<4
<4.5
4.5
11
12
5.03 (0.05)
5.03 (0.05)
0.65
<4
5.0
0
4.66 (0.07)
5.45 (0.02)
5.6 (0.07)
5.51 (0.04)
5.24 (0.04)
4.8
1.04
1.4
13
14
0.86
0.7
5.33 (0.07)
4.84 (0.05)
5.27 (0.01)
<4.2
4.6
4.5
0.63
0.77
1.27
1.29
1.2
15
16
0.54
0.24
0.08
0.5
4.93 (0.04)
4.39
17
15
5.0
<4.3
0.7
1.32
5.07 (0.02)
0.52
6.04 (0.08)
19
20
4.76
<4
4.8
5.1
0.85
0.82
0.96
0
6.43 (0.04)
5.95 (0.01)
6.02 (0.16)
0.89
0.91
5.98 (0.03)
5.89 (0.02)
5.74 (0.03)
1.41
0.95
1.51
21
22
4.9
<4
4.5
0.95
23
24
4.9
4.5
4.14
0.75
0.6
0.9
0.48
0
5.0
0.8
0.8
0.87
0.7
5.1
1.1
25
26
<4.3
5.3
4.5
5.55 (0.17)
6.03 (0.01)
5.19 (0.002)
5.58 (0.04)
5.83 (0.02)
6.75 (0.02)
5.31 (0.05)
1.05
1.15
1.44
27
28
4.8 (0.05)
4.96 (0.1)
5.46 (0.11)
5.36 (0.09)
<4
<3.5
<4
nt
4.2
5.64 (0.15)
29
30
0
0
<4.5
31
32
<4
<3.5
<3.5
nt
0
0
0
33
34
4.58 (0.13)
4.54 (0.13)
5.06 (0.13)
4.64
0.2
35
37
<4
nt
0
38
39
nt
40
4.6
5.4
0.64
0.72
0.56
1
5.35 (0.03)
0.82
5.08 (0.02)
1.57
Arecoline II
Oxotremorine IV
Carbachol VI
6.9
7
(1)
6.5
7.4
6.6
(1)
(1)
6.1
5.7
(1)
(1)
6.86
(1.2)
Agonist values pD2, antagonist values pA2; a intrinsic activity; ®gures represent mean (SEM) where appropriate; ntÐnot tested in any
preparation.

1408
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
swim maze, but are less e€ective in the DMTP model. In
contrast partial agonists, or compounds with a mixed
pro®le (e.g. M1, M3 agonist, M2 partial antagonist 24)
tend to have similar positive e€ects in both models
(Table 5, Figs 2 and 3). There are also di€erential e€ects
of full agonists and compounds with a mixed agonist/
antagonist pro®le on sleep±waking behavior as mea-
sured by the EEG (Table 6). For example, the full ago-
nist 6 increases quiet waking and decreases REM sleep
in a manner similar to that observed with oxotremorine;
24 (M1 and M3 agonist, M2 antagonist) has similar
depressant e€ects on REM sleep but in contrast to the
full agonists increases active waking rather than quiet
waking. The e€ects on active waking may play a role
in the more positive e€ects of the mixed agonist/
antagonists on cognitive performance (Table 5, Figs 2
and 3).
Table 5. E€ects of series members on peripherally and
centrally mediated behaviours
Compd
Mydriasisa Salivationb DMTPc SWIMd
The ability of compounds 6 (top panel) and 24 (bottom
panel) to reverse mnemonic impairments in the two-
island-swim-maze task induced by central acetylcholine
depletion following icv administration of hemi-
cholinium-3 (HC3). Both the full agonist (6) and the
mixed M1/M3 agonist, M2 antagonist (24) can reverse
the HC3-induced de®cit. (See also Table 5, Fig. 3).
4
>1
0.03
0.003
>0.3
>1
0.3
>10
1
5
>10
>1
2.2
1.5
6
0.1
7
10
10
>25
8
9
10
1
15
15
>1
>1
>1
>10
ꢁ10
(10)
1
11
12
Discussion
13
14
0.03
>5
>1
>5
2.2
0.1
0.03
0.3
0.3
0.1
3
Cholinergic properties in binding studies
15
16
3
3
The present series of muscarinic cholinergic ligands was
evaluated for potential muscarinic agonistic or antag-
onistic activity by measuring Oxo-M and Pz binding
and by assessing the Pz/Oxo-M potency ratio (Table 3).
In general modi®cations of the substituent RI had a
greater e€ect on agonist (Oxo-M) than on antagonist
(Pz) binding.
17
18
>10
3
>10
19
20
>1
>10
1
0.03
0.03
0.03
>2
>1
(10)
21
22
0.3
3
>10
3
23
24
>0.3
0.3
0.3
1
>5
0
1.0
25
26
<0.1
0.1
0.3
The role of chain length and branching of substituent R1
27
28
3
ꢁ10
Extending the side chain substituent R1 in a linear
manner from hydrogen
>22
4 (R1=H) via methyl 5
29
30
>10
>10
>10
>10
>10
>1
>1
(R1=methyl) to ethyl 6 (R1=ethyl), increased the ago-
nist binding and decreased the antagonist binding
resulting in a large increase in the Pz/Oxo-M potency
ratio (compounds 4±6; Tables 1±3). Further extension of
the chain length to 7 (R1=propyl) and 8 (R1=butyl)
however led to a decrease in the agonist binding as well
as in the Pz/Oxo-M potency ratio, indicating a bell-
shaped relationship with 6 (R1=ethyl) as most potent
agonist. Addition of a cyclopropyl group to the methyl
side chain of 5, viz. 12 (R1=methylcyclopropyl) lowered
the agonist binding as well as the Pz/Oxo-M potency
ratio. Moreover, addition of methyl substituents to 6
(R1=ethyl) led to 9 (R1=sec-propyl), 10 (R1=tert-
butyl) and 11 (R1=1-methyl-l-propyl), compounds with
decreased Oxo-M binding and Pz/Oxo-M binding ratios.
32
33
35
37
>1
0.3
0.3
0.3
38
40
Arecoline I
Carbachol VI*
Oxotremorine IV
0.01
0.0001
0.0003
>0.5
1
0.05
>0.05
0.03
aMydriasis (MYD): reversal of clonidine-induced mydriasis.
bSalivation: induction of salivation in anaesthetised mice.
cDMTP: reversal of hemicholinium-3-induced memory de®cit
in the delayed matching to position task in rats.
dSWIM: reversal of hemicholinium-3-induced performance
de®cit in the two choice island task of spatial memory.
All values represent the minimal e€ective dose (MED) in
mg/kg, or in the case of mydriasis mg per rat; >x represents
the highest dose tested at which no signi®cant e€ect was
observed; see experimental section for further details.
The role of unsaturation
The e€ect of unsaturation in the side chain of 7 (R1=n-
propyl) was studied. An increase in potency in the
*Not tested further due to strong convulsant properties in vivo.

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1409
Table 6. E€ects of series members on rat sleep±waking behaviour as measured by EEG
Compd
Dose (mg/kg) Active waking Quiet waking Quiet sleep Deep sleep Pre-REM sleep REM sleep
Oxotremorine IV
0.01
0.1
2.2
3.2
2
34
4
73
75
46
13
16
49
30
14
21
37
1
2
51
52
31
30
46
100
100
48
6
24
25
96
44
1
10
100
22
74
Compounds showing full agonist properties at M1, M2, and M3 muscarinic subtypes show similar e€ects to oxotremorine (e.g. 6);
compounds with a mixed M1/M3 agonist and M2 partial antagonist pro®le (e.g. 24) show a di€erent pattern of changes in sleep EEG.
Both types of compounds suppress REM sleep, but di€er considerably in their e€ects on active and quiet waking. Values represent
percent change from placebo control. For further details see Experimental.
Oxo-M binding and subsequent Pz/Oxo-M potency
ratio was found for the propenyl derivative 13 and the 2-
propynyl derivative 20 as compared with the saturated
propyl 7. Addition of substituents, at the propenyl's 3-
position of 13, further increased anity in agonist binding
in 14 (R1=(E)-2-butenyl), 15 (R1=(Z)-2-butenyl), 18
(R1=3-methyl-2-butenyl), 25 (R1=(E)-3-methyl-2-pen-
ten-4-ynyl), and with 26 (R1=(Z)-3-methyl-2-penten-4-
ynyl) having the highest anity in the agonist binding in
this series of muscarinics. In contrast, the anity of the
1-methyl and 2-methyl derivatives 16 (R1=1-methyl-2-
propenyl) and 17 (R1=2-methyl-2-propenyl), in agonist
binding, as well as the anity of the 3-alkenyl derivative
19 (R1=(E,E)-2,4-hexendienyl), were found to be lower
than the parent propenyl derivative 13, whereas the
agonist binding of 24 (R1=(E)-2-penten-4-ynyl) was
found again to be comparable to that of 13. Modi®ca-
tion of the side chain of 20 (R1=propynyl) to 21
(R1=butynyl) or 22 (R1=pentynyl) led to an increase in
agonist binding anity and hence increased the Pz/Oxo-
M potency ratio. Further extension of the side chain
with one methylene group more (23 R1=hexynyl)
resulted in a signi®cantly lowered agonist binding a-
nity, indicating a bell-shaped relationship. In agreement
with this observation, addition of a methyl group to the
propenyl side chain 13, gave 27 (R1=1-methyl-2-pro-
pynyl) a compound with decreased agonist anity. As
compared with 13 (R1=propenyl) the benzyl derivatives
28 (R1=benzyl) 29 (R1=(2-chlorophenyl)methyl) 30
(R1=(2-¯uorophenyl)methyl) 31 (R1=(2-tri¯uoro-
methylphenyl)methyl) did not show an increase in agonist
binding anity or in the PZ/Oxo-M ratio.
results in the loss of potential agonistic properties as
re¯ected in the Pz/Oxo-M binding anity ratio.
Cholinergic properties in isolated organs
Many of the 1,6-dihydro-5-(4H)-pyrimidinone oximes
described are agonists on all three muscarinic cholin-
ergic receptor subtypes M1, M2, and M3 tested (viz. 5, 6,
13, 15, 20, 21, 22, 26, Tables 3 and 4) of these com-
pounds 5, 6, 13, 20, 21, 22, and 26 are carbachol-like in
having high ecacy at all three subtypes, the rest are
oxotremorine-like in having a much lower ecacy at
M1, receptors. Three compounds are M2 and M3 selec-
tive agonists (viz. 14, 18, 25), two compounds are M3
selective agonists (viz. 7, 8) and several compounds have
combined M1, and M3 agonist with M2 partial antago-
nist properties (viz. 9, 24, 27). All the agonist pro®les
identi®ed using functional assays in isolated organs are
characterized by a Pz/Oxo-M binding ratio >100.
However, this ratio showed no apparent relationship to
the observed subtype selectivity.
Within the series several compounds approached the
desired pharmacological pro®le (9, 16, 24, 26, 27;
Tables 3 and 4). This pro®le included mixed M1/M3
agonistic and M2 partial antagonist properties, activity
in several cholinergic-related models of CNS activity,
and a reduced side-e€ect liability. Changing 6 (R1=ethyl)
to 7 (R1=propyl) resulted in a loss of both M1 (hippo-
campal slice) and M2 (left atrium) agonistic properties.
However, 9 (R1=isopropyl) produced the desired pro®le
with a retention of M1 and M3 agonist properties and a
reduction in M2 agonist e€ects; a similar pro®le is obtained
with 24 (R1=(E)-2-penten-4-ynyl; Tables 4 and 5).
The separate series of tetrahydro-N-hydroxy-5-pyr-
imidinamine derivatives 38 (R1=H), 39 (R1=methyl)
and 40 (R1=ethyl) that were derived from the oximes
by selective reduction of the oxime double bond showed
signi®cantly lower agonist binding anity compared to
the parent oxime derivatives 4, 5, and 6, respectively.
Additional methyl substituents to the 1,6-dihydro-(4H)-
pyrimidine ring (viz. 13 versus 35, R3=Me and 6 versus
37, R2=Me) decreases agonist binding anity and
Cholinergic properties in behavioural models
The e€ects of the compounds in vivo support but do not
exactly mirror the in vitro e€ects. In general, com-
pounds with the highest anity and overall potency in
vitro have the greatest and most potent e€ects in the
mydriasis and salivation tests (e.g. 5, 6, 20; Tables 4

1410
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
and 5). However, e€ects in the mydriasis test appear to
be more closely correlated with in vitro activity than
e€ects on salivation. Thus, within some subseries (e.g.
4±8) the in vivo e€ects accurately re¯ect the in vitro
properties, whereas some particular compounds do not
produce the muscarinic pro®le in vivo which would be
predicted from the in vitro data (e.g. 28 and 38). This
may re¯ect interference from activity at other receptors
known to mediate drug e€ects in these models, for
example alpha adrenergic receptors in the eye.
Figure 2. Left: accuracy of responding, as measured by % correct responding, after di€erent delays: * placebo; ^ HC3 replacement;
~ low dose; & mid dose; ! high dose. Right, top: accuracy of responding collapsed across delays per dose per session. Middle:
trials completed per dose per session. Bottom: speed of responding per dose per session. All values mean and standard deviation;
* signi®cant di€erence from HC3. For further details see Experimental.

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1411
Although designed to measure the same area of mem-
ory, the nature of the two tasks di€ers considerably. The
swim maze can be seen as a stressful, forced response
task: rats have to swim and thus make a response. The
DMTP task is positively as opposed to aversely moti-
vated, the only consequence of not responding is less
food in the session. Detrimental side e€ects are thus more
readily able to interfere with performance in the DMTP
than in the swim maze, where lack of responding may be
life-threatening. The most likely cause for this di€erence
in ecacy in the two tests is probably related to the M2-
activity. The results of the salivation experiments shown
in Table 5 are from anaesthetized animals. Anaesthetized
rats appear to be more sensitive to M3-mediated periph-
eral e€ects than conscious animals: at the doses used in
the mnemonic models signi®cant salivation is seen only
with the most potent compounds. General full agonists
(e.g. 6) cause dramatic changes in the cardiovascular
system in conscious rats at doses slightly above those
used in the DMTP and similar to the doses that are active
in the swim maze (unpublished observations). It may be
that the stress of being immersed in the water negates
some of the adverse e€ects of full agonists on M2-medi-
ated cardiac responses, a protective e€ect not evident in
the DMTP. Recent studies with cholinergic agonists
suggest that sympathetic activation can alleviate some
peripheral e€ects of muscarinic stimulation.28
General aspects
In general, there exists a good correlation between the
Pz/Oxo-M binding ratio and agonistic properties.
Compounds showing the highest ratios have the highest
intrinsic activity at all receptors (e.g. carbachol (VI), 6,
21, 22). Indeed, the binding ratios of 6 and 21 are only
twofold lower than carbachol and three to seven times
higher than that of arecoline or oxotremorine which are
known partial agonists at some receptors. These com-
pounds show no antagonistic properties in any model.
For M1- and M2-mediated e€ects, the correlation exists
for both potency, which might be expected, but also for
ecacy that may actually be more useful. However, in
contrast to previous reports that agonist/antagonist
ratios may be used to predict selectivity,28 in this study
the ratio could not be used to predict properties other
than agonist-like or not agonist-like. For example,
although for compounds 22 and 26 the ratios and the
pharmacological pro®le are similar (both full agonist in
all preparations, Table 4), this is not true for com-
pounds 8 and 24, which also have similar binding ratios.
In the latter case, 8 is a weak partial M3 agonist,
whereas 24 is an agonist in both M1 and M3 prepara-
tions. Thus, a high ratio may indicate a full agonist,
lower ratios may indicate selectivity for one receptor
subtype (e.g. 8), a partial agonist or a mixed agonist/
antagonist pro®le (e.g. compounds 9 and 24). The use of
Figure 3. Left: accuracy as measured by the mean number of
consecutive correct responses in a session of 30 trials. Right:
mean time taken to swim to platform. All data mean and stan-
dard deviation: *signi®cant di€erence from HC3. For further
details see Experimental.
Cholinergic properties in mnemonic models
The e€ects of the compounds tested in the two mne-
monic models are particularly interesting. Both models
test ostensibly spatial memory, and both use animals
where performance has been impaired by the depletion
of acetylcholine in the brain by prior administration of
hemicholinium-3 (HC3, see Experimental for further
details). Compounds that have bene®cial e€ects in one
model do not automatically show bene®cial e€ects in
the other (Table 5, Figs 1 and 2). General agonists (i.e.,
compounds with high ecacy at all receptors tested
such as 5, 6, 13, 20, oxotremorine, and arecoline) were
able to reverse the HC3-induced de®cit in the swim
maze, but were unable to signi®cantly reverse the de®cit
in the DMTP. In contrast two compounds (9 and 24)
were able to improve performance in both models.
These latter two entities lack the strong M2 agonist
e€ects apparent in the other compounds (Table 4).

1412
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
binding ratios to predict activity is also limited by the
tissue preparation used. For example, in tissues with a
high receptor reserve, such as the guinea pig ileum,
compounds with slight agonist properties will be readily
identi®ed as agonists. This probably accounts for the
lack of correlation between ecacy and binding ratio
and highlights the importance of choosing the relevant
tissue preparation for pharmacological investigation of
novel chemical entities. Other research has indicated
functional assays to be a more reliable indicator of
selectivity than binding anities.29 These studies indi-
cate the diculties faced in designing selective com-
pounds. Although much work has concentrated on the
production of selective M1 agonists, the few compounds
that have emerged have not yet been shown to be
unequivocally bene®cial in the treatment of mnemonic
loss in Alzheimer's disease. Indeed, some have argued
that M1 agonists may be less e€ective than initially
expected due to a fault in M1 receptor function in Alz-
heimer's disease.16 The addition of some M3 agonism
and M2 antagonism to this pro®le has thus been desired;
both M2 antagonists and M3 agonists have shown posi-
tive e€ects in animal models.21,30,31 However, de®nitive
data from the clinic is still awaited, and the unwanted
activity of at least the current M3 agonists in the per-
iphery limits their therapeutic potential.
reported as d values (parts per million) relative to Me4Si
as an internal standard. Fast atom bombardment (FAB)
mass spectra were recorded with a Finnigan MAT 90
mass spectrometer (Finnigan MAT, Bremen, FRG).
Samples were dissolved in methanol and mixed with the
matrix compounds on standard stainless steel targets.
Exact masses of the protonated molecular ions were
determined with the peak matching technique at a mass
resolution of >8900 (10% valley de®nition) in the
positive ion mode using two reference masses either
from poly(ethylene glycol), average M.M. 400, or
poly(propylene glycol), average M.M. 425. Average
exact masses were calculated from at least 10 computer-
controlled measurements using the bracketing method.
All new compounds gave satisfactory m/z data. Thin-
layer chromatography (TLC) was carried out by using
Merck precoated silicagel F254 plates (thickness
0.25 mm). Spots were visualized with a UV handlamp
and Cl2/tetramethylbenzidine. For column chromato-
graphy Merck silica gel 60 was used. No particular
attempts were made to optimize the reaction conditions
for the reactions described.
The O-alkyl hydroxylamines used were commercially
available, prepared from the appropriate bromine deri-
vative19 or prepared from the corresponding alcohol
derivative20 according to general literature procedures.
General procedure 1: 1,6-dihydro-5-(4H)-pyrimidinone
derivatives from diamino-acetone 1,6-dihydro-5-(4H)-
pyrimidinone O-methyloxime (5). 1,3-Diamino acetone
dihydrochloride monohydrate (2.6 g, 14.6 mmol) and O-
methyl hydroxylamine hydrochloride (1.5 g, 18 mmol)
were dissolved in methanol. The solution was heated to
re¯ux for 48 h. The progress of reaction was monitored
by 1H NMR spectroscopy. 1H NMR: (MeOD,
200 MHz) d 4.04 (s, 3H, OCH3), 3.95 (s, 2¯, CH2), 3.7 8
(s, 2H, CH2). The crude product was dissolved in
methanol and 50 mmol trimethyl-orthoformate was
added. The mixture was heated to re¯ux for 20 h the
solvent was removes in vacuo. After recrystallisation
from methanol/ethylacetate the desired 1,6-dihydro-5-
(4H)-pyrimidinone O-methyloxime 5. was obtained in
Conclusion
A series of muscarinic agonists that di€ered con-
siderably in their anity and pharmacological pro®le
were synthesized based around arecoline. Systematic
variation of R1 yielded compounds, which were full
agonists at the M1, M2, and M3 receptors, selective
agonists at the M3 receptors, or M1/M3 agonists with
partial agonist or no agonistic e€ects at the M2 recep-
tors; substitutions into the heterocyclic ring generally
disrupted the anity and pharmacology. Some com-
pounds with mixed agonist/antagonist pro®les were able
to reverse some cholinergic-related performance de®cits.
However, no compound had a pro®le free of peripheral
side e€ects, accordingly bene®cial e€ects in the mne-
monic models were limited.
1
36% overall yield. H NMR: (MeOD, 200 MHz) d 8.13
(s, 1H, N=CH), 4.30 (s, 2H, CH2), 4.1 0 (s, 2H, CH2),
3.90 (s, 3H, O-CH3). Exact mass calcd for C5H9N3O
[M+H]+ 128.0779, found: 128.0824. Anal. calcd
.
(C5H9N3O HCl): 36.71% C, 6.16% H, 25.68% N.
Found: 36.44% C,6.19% H, 25.21% N.
Experimental
Chemistry
Melting points (Ep) were taken on a Buechi capillary
melting point apparatus and are uncorrected. The ele-
mental analyses were within 0.45 of the theoretical
values. Proton magnetic resonance spectra were mea-
sured on a Bruker WP200, AC200, or AM360 instru-
ment (using standard conditions). Chemical shifts are
According to the general procedure 1 the following
derivatives were prepared
1,6-Dihydro-5-(4H)-pyrimidinone oxime hydrochloride
(4). Starting from diamino-acetone dihydrochloride
.
monohydrate and hydroxylamine HCl the diamino

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1413
1
0
derivative was prepared. H NMR: (MeOD, 200 MHz)
d 3.95 (s, 2H, CH2), 3.85 (s, 2H, CH2). Cyclisation with
trimethyl-orthoformate gave 4 in 54% overall yield. H
6H, CH3 s). Cyclisation with trimethylorthoformate gave
1
9 in 47% overall yield. H NMR: (MeOD, 200 MHz) d
8.13(s, 1H, N=CH), 4.50±4.20 (m, 1H, CH), 4.30 (s, 2H,
CH2), 4.10 (s, 2H, CH2), 1.25 (d, 6H, CH3 s). Exact mass
calcd for C7H11N3O [M+H]+ 156.1113, found:
.
156.1137. Anal. calcd (C7H13N3O HCl): 43.87% C, 7.36%
H, 21.92% N. Found: 43.27% C, 7.01 % H, 21.39% N.
1
0
NMR: (MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 4.30
(s, 2H, CH2), 4.10 (s, 2H, CH2). Exact mass calcd for
C4H7N3O [M+H] 114.0634, found: 114.0667. Anal.
.
calcd (C4H7N3O HCl): 32.11% C, 5.39% H, 28.09% N.
Found: 31.85% C, 5.19% 14, 27.41% N.
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-methyl-2-propyl)-
oxime hydrochloride (10). Starting from diamino acetone
dihydrochloride monohydrate and 2-methyl-2-propyl-
1,6-Dihydro-5-(4H)-pyrimidinone O-ethyloxime hydro-
chloride (6). Starting from diamino acetone dihydro-
20
.
.
chloride monohydrate and O-ethylhydroxylamine HCl
hydroxylamine HCl the diamino derivative was pre-
the diamino derivative was prepared. 1H NMR:
(MeOD, 200 MHz) d 4.15±4.05 (q, 2H, OCH2), 3.85 (s,
2H, CH2), 3.65 (s, 2H, CH2), 1.25 (t, 3H, CH3). Cycli-
sation with trimethylorthoformate gave 6 in 48 % over-
all yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s, 1H,
N=CH), 4.30 (s, 21-1, CH2),4.10 (s, 2H, CH2), 4.25±
4.10 (q, 2H, OCH2), 1.25 (t, 3H, CH3). Exact mass calcd
for C6H11N3O [M+H]+ 142.0938, found: 142.0980.
pared. 1H NMR: (MeOD, 200 MHz) d 3.95 (s, 2H,
CH2), 3.85 (s, 2H, CH2), 1.35 (s, 9H, CH3 s). Cyclisation
0
with trimethylorthoformate gave 10 in 49% overall
yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s, 1H,
N=CH), 4.85 (s, 2H, CH2), 4.15 (s, 2H, CH2), 1.35 (d,
0
9H, CH3 s). Exact mass calcd for C8H15N3O [M+H]+
.
170.1276, 170.1293. Anal. calcd (C8H15N3O HCl):
46.71% C, 7.84% H, 20.42% N. Found: 46.87% C,
7.63% H, 20.73% N.
.
Anal. calcd (C6H11N3O HCl): 40.57% C, 6.81% H,
23.66% N. Found: 40.59% C, 6.73% H, 23.54% N.
1,6-Dihydro-5-(4H)-pyrimidinone O-(1-methyl-1-propyl)-
oxime hydrochloride (11). Starting from diamino acetone
dihydrochloride monohydrate and 1-methyl-1-propyl-
1,6-Dihydro-5-(4H)-pyrimidinone O-propyloxime hydro-
chloride (7). Starting from diamino acetone dihydro-
chloride monohydrate and O-propylhydroxylamine.
HCl19 the diamino derivative was prepared. 1H
NMR:(MeOD, 200 MHz) d 4.00 (t, 2H, OCH2), 3.95 (s,
21-1, CH2), 3.85 (s, 2H, CH2), 1.75±1.60 (m, 2H, CH,),
1.00 (t, 3H, CH3). Cyclisation with trimethylortho-
formate gave 7 in 45% overall yield. 1H NMR: (MeOD,
200 MHz) d 8.13 (s, 1H, N=CH), 4.30 (s, 2H, CH2),
4.10 (s, 2H, CH2), 4.00 (t, 2H, OCH2), 1.75±1.60 (m,
2H, -CH2), 1.00 (t, 3H, CH3). Exact mass calcd for
C7H11N3O [M+H]+ 156.1112, found: 156.1137.
20
.
hydroxylamine HCl the diamino derivative was pre-
d 4.40±4.10
pared. 1H NMR: (MeOD, 200 MHz)
(m, 1H, CH), 4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 1.80±
1.45 (m, 2H, CH2), 1.20 (d, 3H, CH2), 0.95 (t, 3H, CH3).
Cyclisation with trimethylorthoformate gave 11 in 40%
overall yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s,
1H, N=CH), 4.40±4.10 (m, 1H, CH), 4.35 (s, 2H, CH2),
4.15 (s, 2H, CH2), 1.80±1.45 (m, 2H, CH2), 1.20 (d, 3H,
CH3), 0.95 (t, 3H, CH3). Exact mass calcd for
C8H15N3O [M+H]+ 170.1277, found: 170.1293.
1,6-Dihydro-5-(4H)-pyrimidinone O-butyloxime hydro-
chloride (8). Starting from diamino acetone dihydro-
1,6-Dihydro-5-(4H)-pyrimidinone O-(methylcyclopropyl)-
oxime hydrochloride (12). Starting from diamino acetone
dihydrochloride monohydrate and O-methylcyclopropyi-
20
.
chloride monohydrate and O-butylhydroxylamine HCl
the diamino derivative was prepared. 1H NMR:
(MeOD, 200 MHz) d 4.05 (t, 2H, OCH2), 4.20±4.10 (m,
2H, OCH2), 3.85 (s, 2H, CH2), 1.70±1.55 (m, 2H, CH2),
1.50.1.10 (m, 2H, CH2), 0.90 (t, 3 H, CH3). Cyclisation
with trimethylorthoformate gave 8 in 49% overall yield.
1H NMR: (MeOD, 200 MHz) d 8.13 (s, 1H, N=CH),
4.35 (s, 2H, CH2), 4.20±4.10 (m, 2H, OCH2), 4.15 (s,
2H, CH2), 1.70±1.55 (m, 2H, -CH2), 1.50±1.10 (m, 2H,
CH2), 0.90 (t, 3H, CH3). Exact mass calcd for
C8H15N3O [M+H]+ 170.1288, found: 170.1293.
hydroxylamine HCl the diamino derivative was pre-
19
.
pared. 1H NMR: (MeOD, 200 MHz) d 3.70 (d, 2H,
OCH2), 3.75 (s, 2H, CH2), 3.65 (s, 2H, CH2), 1.30±1.00
(m, 1H, CH2), 0.80±0.60 (m, 2H, CH2), 0.50±0.30 (m,
2H, CH2). Cyclisation with trimethylorthoformate gave
12 in 57% overall yield. 1H NMR: (MeOD, 200 MHz) d
8.10 (s, 1H, N=CH), 4.40 (d, 2H, CH2), 4.20 (s, 2H,
CH2), 4.00±3.90 (d, 2H, OCH2), 1.30±1.00 (m, 1H,
CH2), 0.80±0.60 (m, 2H, CH2), 0.50±0.30 (m, 2H, CH2).
Exact mass calcd for C8H13N3O [M+H]+ 168.1127,
.
found: 168.1137. Anal. calcd (C8H13N3O HCl): 47.18%
1,6-Dihydro-5-(4H)-pyrimidinone O-(1-methylethyl)oxime
hydrochloride (9). Starting from diamino acetone di-
hydrochloride monohydrate and O-1-methyl-ethyl-
C, 6.93% H, 20.63% N. Found:47.35% C, 6.59% H,
20.81% N.
21
.
hydroxylamine HCl the diamino derivative was pre-
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-propenyl)oxime
hydrochloride (13). Starting from diamino acetone di-
hydrochloride monohydrate and 2-propenylhydroxyl-
pared. 1H NMR: (MeOD, 200 MHz) d 4.60±4.3 0 (m,
1H, CH), 3.75 (s, 2H, CH2), 3.60 (s, 2H, CH2), 1.25 (d,

1414
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
20
amine HCl the diamino derivative was prepared. 1H
NMR: (MeOD, 200 MHz) d 6.20±5.80 (m, 1H, =CH),
5.40±5.20 (m, 2H, =CH2), 4.65 (d, 2H, OCH2), 4.00 (s,
2H, CH2), 3.85 (s, 2H, CH2). Cyclisation with tri-
methylorthoformate gave 13 in 57% overall yield. 1H
NMR: (MeOD, 200 MHz) d 8.13 (s, 1 H, N=CH),
6.10±5.80 (m, 1H, =CH), 5.40±5.25 (m, 2H, =CH2-),
4.65 (d, 2H, OCH2), 4.40 (s, 2H, CH2), 4.15 (s, 2H,
CH2). Exact mass calcd for C7H11N3O [M+H]+
154.0969, found: 154.0980.
hydroxylamine HCl the diamino derivative was pre-
pared. 1H NMR: (MeOD, 200 MHz) d 5.00 (m, 2H,
=CH2), 4.60±4.50 (s, 2H, OCH2), 4.05 (s, 2H, CH2),
3.85 (s, 2H, CH2), 1.75 (s, 3H, CH3). Cyclisation with
.
.
1
trimethylorthoformate gave 17 in 9% overall yield. H
NMR: (MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 5.00
(m, 2H, =CH2), 4.60±4.50 (s, 2H, OCH2), 4.35 (s, 2H,
CH2), 4.15 (s, 2H, CH2), 1.75 (s, 3H, CH3). Exact mass
calcd for C8H13N3O [M+H]+ 168.1108, found: 168.1137.
1,6-Dihydro-5-(4H)-pyrimidinone O-(3-methyl-2-butenyl)-
oxime hydrochloride (18). Starting from diamino acetone
dihydrochloride monohydrate and 3-methyl-2-butenyl-
(E)-1,6-Dihydro-5-(4H)-pyrimidinone O-(2-butenyl)oxime
hydrochloride (14). Starting from diamino acetone di-
hydrochloride monohydrate and (E)-2-butenylhydroxyl-
hydroxylamine HCl20 in tert-butanol the diamino deriva-
.
20
1
amine HCl the diamino derivative was prepared. 1H
tive was prepared. H NMR: (MeOD, 200MHz) d 5.50
(m, 1H, =CH), 4.80 (d, 2H, OCH2), 4.05 (s, 2H, CH2),
.
NMR: (MeOD, 200 MHz)
CH=CH), 4.50 (d, 2H, OCH2), 4.05 (s, 2H, CH2),
d
5.90±5.50 (m, 2H,
3.85 (s, 2H, CH2), 1.75 (d, 6H, =(CH3)2). Cyclisation with
1
3.85(s, 2H, CH2), 1.70 (d, 3H, CH3). Cyclisation with
1
trimethylorthoformate gave 18 in 44% overall yield. H
trimethylorthoformate gave 14 in 16% overall yield. H
NMR: (MeOD, 200MHz) d 8.05 (s, 1H, N=CH), 5.40
(m, 1H, =CH), 4.60 (d, 2H, OCH2), 4.30 (s, 2H, CH2),
4.10 (s, 2H, CH2), 1.75 (d, 6H, =(CH3)2). Exact mass
calcd for C9H15N3O [M+H]+ 182.1255, found: 182.1293.
NMR: (MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 5.90±
5.50 (m, 2H, CH=CH), 4.50 (d, 2H, OCH2), 4.35 (s, 2H,
CH2), 4.15 (s, 2H, CH2), 1.70 (d, 3H, CH3). Exact mass
calcd for C8H13N3O [M+H]+ 168.1128, found: 168.1137.
1,6-Dihydro-5-(4H)-pyrimidinone O-(2,4-hexendienyl)-
oxime hydrochloride (19). Starting from diamino acetone
dihydrochloride monohydrate and 2,4-hexendienyl-
(Z)-1,6-Dihydro-5-(4H)-pyrimidinone O-(2-butenyl)oxime
hydrochloride (15). Starting from diamino acetone di-
hydrochloride monohydrate and (Z)-2-butenythydroxyl-
20
.
hydroxylamine HCl
the diamino derivative was
amine HCl the diamino derivative was prepared. 1H
prepared. 1H NMR: (MeOD, 200 MHz) d 6.30±6.20 (m,
1H, =CH), 6.15±6.00 (t, 1H, =CH), 5.90±5.65 (m, 2H,
=CH), 4.75 (d, 2H, OCH2), 4.00 (s, 2H, CH2), 3.80 (s,
2H, CH2), 1.75 (d, 3H, CH3). Cyclisation with tri-
methylorthoformate gave 19 in 37% overall yield. 1H
NMR: (MeOD, 200 MHz) d 8.10 (s, 1H, N=CH), 6.30±
6.00 (m, 2H, HC=CH), 5.85±5.60 (m, 2H, HC=CH),
4.65 (d, 2H, OCH2), 4.30 (s, 2H, CH2), 4.10 (s, 2H,
CH2), 1.75 (d, 3H, CH3). Exact mass calcd for
C10H15N3O [M+H]+ 194.1263, found: 194.1293.
20
.
NMR: (MeOD, 200 MHz) d 5.70±5.30 (m, 2H,CH=CH),
4.60 (d, 2H, OCH2), 4.05 (s, 2H, CH2), 3.85(s, 2H,
CH2), 1.50 (d, 3H, CH3). Cyclisation with trimethyl-
1
orthoformate gave 15 in 32% overall yield. H NMR:
(MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 5.70±5.30
(m, 2H, CH=CH), 4.50 (d, 2H, OCH2), 4.35 (s, 2H,
CH2), 4.15 (s, 2H, CH2), 1.50 (d, 3H, CH3). Exact mass
calcd for C8H13N3O [M+H]+ 168.1116, found:
168.1137.
1,6-Dihydro-5-(4H)-pyrimidinone O-(1-methyl-2-propenyl)-
oxime hydrochloride (16). Starting from diamino acet-
one dihydrochloride monohydrate and 1-methyl-2-pro-
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-propynyl)oxime
hydrochloride (20). Starting from diamino acetone dihy-
drochloride monohydrate and 2-propynylhydroxyl-
20
penylhydroxylamine HCl the diamino derivative was
amine HCl the diamino derivative was prepared. 1H
.
.
prepared. 1H NMR: (MeOD, 200 MHz) d 6.00±5.80 (m,
1H, CH=), 5.30±5.10 (m, 2H, =CH2), 4.80±4.60 (m,
1H, OCH), 4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 1.35 (d,
3H, CH3). Cyclisation with trimethylorthoformate gave
NMR: (MeOD, 200 MHz) d 4.80 (d, 2H, OCH2), 4.05
(s, 2H, CH2), 3.85 (s, 2H, CH2), 3.40 (t, 1H, (CH).
Cyclisation with trimethylorthoformate gave 20 in 36%
overall yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s,
1H, N=CH), 4.75 (d, 2H, OCH2) 4.3 5 (s, 2H, CH2),
4.15 (s, 2H, CH2), 2.95(t, 3H, ꢂCH). Exact mass calcd
for C7H9N3O [M+H]+ 152.0800, found: 152.0824.
1
16 in 8% overall yield. H NMR: (MeOD, 200 MHz) d
8.13 (s, 1H, N=CH), 6.00±5.80 (m, 1H, HC=), 5.30±
5.10 (m, 2H, =CH2), 4.80±4.60 (m, 1H, OCH), 4.35 (s,
2H, CH2), 4.15 (s, 2H, CH2), 1.35 (d, 3H, CH3). Exact
mass calcd for C8H13N3O [M+H]+ 168.1107, found:
168.1137.
.
Anal. caled (C7H9N3O HCl): 44.81% C, 5.37% H,
22.40% N. Found: 45.05% C, 5.22% H, 22.23% N.
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-butynyl)oxime
hydrochloride (21). Starting from diamino acetone di-
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-methyl-2-propenyl)-
oxime hydrochloride (17). Starting from diamino acetone
dihydrochloride monohydrate and 2-methyl-2-propenyl-
hydrochloride monohydrate and 2-butynylhydroxyl-
20
.
amine HCl
the diamino derivative was prepared.

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1415
1H NMR: (MeOD, 200 MHz) d 4.65 (d, 2H, OCH2),
4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 1.80 (s, 3H, CH3).
Cyclisation with trimethylorthoformate gave 21 in
35% overall yield. 1H NMR: (MeOD, 200 MHz) d
8.13 (s, 1H, N=CH), 4.65 (d, 2H, OCH2) 4.35 (s,
2H, CH2), 4.15 (s, 2H, CH2), 1.80 (s, 3H, CH3).
Exact mass calcd for C8H11N3O [M+H]+ 166.0966,
found: 166.0980.
4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 1.85 (d, 3H, CH3).
Cyclisation with trimethylorthoformate gave 25 in 32%
overall yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s,
1H, CH=N), 6.00 (t, 1H, =CH), 4.65 (d, 2H, OCH2),
4.35 (s, 2H, CH2), 4.15 (s, 2H, CH2), 1.85 (d, 3H, CH3).
Exact mass calcd for C10H13N3O [M+H]+ 192.1120,
found: 192.1137.
(Z)-1,6-Dihydro-5-(4H)-pyrimidinone O-(3-methyl-2-
penten-4-ynyl)oxime hydrochloride (26). Starting from
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-pentynyl)oxime
hydrochloride (22). Starting from diamino acetone di-
hydrochloride monohydrate and 2-pentynylhydroxyl-
diamino acetone dihydrochloride monohydrate and
20
.
(Z)-3-methyl-2-penten-4-ynylhydroxylamine HCl
the
amine HCl the diamino derivative was prepared. 1H
diamino derivative was prepared. 1H NMR: (MeOD,
200 MHz) d 6.00 (t, 1H, =CH), 4.85 (d, 2H, OCH2),
4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 1.85 (d, 3H, CH3).
Cyclisation with trimethylorthoformate gave 26 in 42%
overall yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s,
1H, CH=N), 6.00 (t, 1H, =CH), 4.85 (d, 2H, OCH2),
4.35 (s, 2H, CH2), 4.15 (s, 2H, CH2), 1.85 (d, 3H, CH3).
Exact mass calcd for C10H13N3O [M+H]+ 192.1132,
found: 192.1137.
20
.
NMR: (MeOD, 200 MHz) d 4.70 (t, 2H, OCH2), 4.05 (s,
2H, CH2), 3.85 (s, 2H, CH2), 2.30±2.10 (m, 2H, ꢂC-
CH2), 1.10 (t, 3H, CH3). Cyclisation with trimethyl-
1
orthoformate gave 22 in 40% overall yield. H NMR:
(MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 4.70 (t, 2H,
OCH2) 4.35 (s, 2H, CH2), 4.15 (s, 2H, CH,), 2.30±2.10
(m, 2H, ꢂC-CH2), 1.10 (t, 3H, CH3). Exact mass calcd
for C9H13N3O [M+H]+ 180.1120, found: 180.1137.
1,6-Dihydro-5-(4H)-pyrimidinone O-(2-hexynyl)oxime
hydrochloride (23). Starting from diamino acetone dihy-
1,6-Dihydro-5-(4H)-pyrimidinone O-(l-methyl-2-propynyl)-
oxime hydrochloride (27). Starting from diamino acetone
dihydrochloride monohydrate and 1-methyl-2-propynyl-
drochloride monohydrate and 2-hexynylhydroxyl-
20
amine HCl the diamino derivative was prepared. 1H
hydroxylamine HCl the diamino derivative was pre-
20
.
.
NMR: (MeOD, 200 MHz) d 4.55 (m, 2H, OCH2), 4.05
(s, 2H, CH2), 3.85 (s, 2H, CH2), 2.20 (m, 2H, ꢂCH),
1.50 (m, 2H, -CH2-), 1.00 (t, 3H, -CH3). Cyclisation
with trimethyforthoformate gave 23 in 52% overall
yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s, 1H,
N=CH), 4.55 (m, 2H, OCH2) 4.35 (s, 2H, CH2), 4.15 (s,
2H, CH2), 2.20 (m, 2H, ꢂCH), 1.50 (m, 2H, -CH2-),
1.00 (t, 3H, CH3). Exact mass calcd for C10H15N3O
[M+H]+ 194.1303, found: 194.1293.
pared. 1H NMR: (MeOD, 200 MHz) d 5.00 (m, 1H,
OCH), 4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 2.90 (m,
1H, ꢂCH). Cyclisation with trimethylorthoformate
gave 27 in 10% overall yield. 1H NMR: (MeOD,
200 MHz) d 8.13 (s, 1H, CH=N), 5.00 (m, 1H, OCH),
4.35 (s, 2H, CH2),4.15 (s, 2H, CH2), 2.90 (m, 1H, ꢂCH).
Exact mass calcd for C8H11N3O [M+H]+ 166.0947,
found: 166.0980.
1,6-Dihydro-5-(4H)-pyrimidinone O-(phenylmethyl)oxime
hydrochloride (28). Starting from diamino acetone di-
hydrochloride monohydrate and O-phenylmethylhydrox-
(E)-(1,6-Dihydro-5-(4H)-pyrimidinone O-(2-penten-4-ynyl)-
oxime hydrochloride (24). Starting from diamino acetone
dihydrochloride monohydrate and (E)-2-penten-4-ynyl-
ylamine HCl the diamino derivative was prepared. 1H
.
19
.
hydroxylamine HCl the diamino derivative was pre-
NMR: (MeOD, 200 MHz) d 7.50±7.00 (m, 5H, Ar.),
5.20 (s, 2H, OCH2), 3.95 (s, 2H, CH2), 3.85 (s, 2H,
CH2). Cyclisation with trimethylorthoformate gave 28
in 60% overall yield. 1H NMR: (MeOD, 200 MHz) d
8.13 (s, 1H, N=CH), 7.50±7.20 (m, 5 H, Ar.), 5.20 (s,
2H, OCH2), 4.35 (s, 2H, CH2), 4.10 (s, 2H, CH2). Exact
pared. 1H NMR: (MeOD, 200 MHz) d 6.40 (m, 1H,
=CH), 5.90±5.80 (m, 1H, =CH), 4.70 (d, 2H, OCH2),
4.05 (s, 2H, CH2), 3.85 (s, 2H, CH2), 3.20 (d, 1H, (CH).
Cyclisation with trimethylorthoformate gave 24 in 58%
overall yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s,
1H, CH=N), 6.40 (m, 1H, =CH), 5.80±5.70 (m, 1H,
=CH), 4.65 (d, 2H, OCH2), 4.35 (s, 2H, CH2), 4.15 (s,
2H, CH2), 3.30 (d, 1H, (CH). Exact mass calcd for
C9H11N3O [M+H]+ 178.0991, found: 178.0980.
mass calcd for C11H13N3O [M+H]+ 204.1137, found:
11 13
.
204.1142. Anal. calcd (C H N3O HCl): 55.12% C, 5.8
9% H, 17.5 3 % N. Found: 54.84% C, 5.85% H,
17.81% N.
(E)-(1,6-Dihydro-5-(4H)-pyrimidinone-O-(3-methyl-2-
penten-4-ynyl)oxime hydrochloride (25). Starting from
diamino acetone dihydrochloride monohydrate and (E)-
1,6-Dihydro-5-(4H)-pyrimidinone O-[(2-chlorophenyl)-
methylloxime hydrochloride (29). Starting from diamino
acetone dihydrochloride monohydrate and O-(2-chloro-
20
20
.
.
3-methyl-2-penten-4-ynylhydroxylamine HCl
the
phenyl)methylhydroxylamine HCl the diamino deriv-
diamino derivative was prepared. 1H NMR: (MeOD,
ative was prepared. 1H NMR: (MeOD, 200 MHz) d
7.50±7.00 (m, 4H, Ar.), 5.20 (s, 2H, OCH2), 3.95 (s, 2H,
200 MHz) d 6.00 (t, 1H, =CH), 4.75 (d, 2H, OCH2),

1416
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
CH2), 3.85 (s, 2H, CH2). Cyclisation with trimethyl-
1
H's), 5.00 (s, 2H, OCH2), 4.40 (s, 2H, CH2), 4.15 (s, 2H,
CH2), 3.80 (s, 3H, OCH3). Exact mass calcd for
C14H15N3O2 [M+H]+ 258.1242, found: 258.1239.
orthoformate gave 29 in 26% overall yield. H NMR:
(MeOD, 200 MHz) d 8.13 (s, 1H, N=CH) 7.50±7.00 (m,
4H, Ar), 5.20 (s, 2H, OCH2), 4.35 (s, 2H, CH2), 4.10 (s,
2H, CH2). Exact mass calcd for C11H12ClN3O [M+H]+
238.0716, found: 238.0747.
1,6-Dihydro-5-(4H)-pyrimidinone derivatives from dichloro-
acetone (E)-1,3-Dichloroacetone O-(2-penten-4-ynyl)-
oxime (2a). (E)-(2-Penten-4-ynyl)hydroxylamine (B)
(6.7 g, 50 mmol) and 1,3-dichloroacetone (6.1 g,
48 mmol) were dissolved in 250 mL dry methanol and
heated to 40 ꢀC. After 2 h the solvent was evaporated
and the residue dissolved in 100 mL water. The solution
was extracted three times with 75 mL ethylacetate. The
combined organic layers were dried (MgSO4) and eva-
1,6-Dihydro-5-(4H)-pyrimidinone O-[(2-¯uorophenyl)
methylloxime hydrochloride (30). Starting from diamino
acetone dihydrochloride monohydrate and O-(2-¯uoro-
20
.
phenyl)methylhydroxylamine HCl
the
diamino
1
derivative was prepared. H NMR: (MeOD, 200 MHz)
d 7.50±7.00 (m, 4H, Ar), 5.20 (s, 2H, OCH2), 3.95 (s,
2H, CH2), 3.85 (s, 2H, CH2). Cyclisation with tri-
methylorthoformate gave 30 in 50% overall yield. 1H
NMR: (MeOD, 200 MHz) d 8.13 (s, 1H, N=CH) 7.50±
7.00 (m, 4H, Ar.), 5.20 (s, 2H, OCH2), 4.3 5 (s, 2H,
CH2), 4.10 (s, 2H, CH2). Exact mass calcd for
C11H12FN3O [M+H]+ 222.1042, found: 222.103 7.
1
porated and gave 2a quantitatively as an oil. H NMR:
(CDCl3, 200 MHz) d 6.30 (dt, 1H, =CH), 5.70 (dd, 1H,
=CH), 4.70 (d, 2H, OCH2), 4.35 (s, 2H, CH2Cl), 4.25
(s, 2H, CH2Cl), 2.95 (s, 1H, (CH).
(E)-1,3-Diphthaloylacetone O-(2-penten-4-ynyl)oxime (3a).
Potassiumphthalimide (20.4 g, 110 mmol) was added to
a solution of 2a (50 mmol) in 75 mL dry DMF. The
mixture was heated to 100 ꢀC for 3 h. After cooling, the
solution was diluted with 150 mL water and extracted
three times with 150 mL dichloromethane. The combined
organic layers were dried (MgSO4) and the solvent was
evaporated. The residue was subjected to column chro-
matography (Silicalgel 60, toluene:ethanol, 9:1) to give
the crude product. Recrystallisation from heptane/
ethylacetate gave 3a in 43% yield. 1H NMR: (CDCl3,
200 MHz) d 7.90±7.65 (m, 8H, AR H's), 6.25 (dt, 1H,
=CH), 5.55 (d, 1H, =CH), 4.60 (s, 2H, CH2N), 4.55 (d,
2H, OCH2), 4.45 (s, 2H, CH2N), 2.85 (s, 1H, (CH).
1,6-Dihydro-5-(4H)-pyrimidinone O-[(2-tri¯uoromethyl-
phenyl)methylloxime hydrochloride (31). Starting from
diamino acetone dihydrochloride monohydrate and
20
.
O-(2-tri¯uoromethylphenyl)methylhydroxylamine HCl
the diamino derivative was prepared. 1H NMR:
(MeOD, 200 MHz) d 7.60±7.20 (m, 4H, ArH's), 5.20 (s,
2H, OCH2), 3.95 (s, 2H, CH2), 3.85 (s, 2H, CH2).
Cyclisation with trimethylorthoformate gave 31 in 7%
overall yield. 1H NMR: (MeOD, 200 MHz) d 7.95 (s,
1H, N=CH) 7.607.20 (m, 4H, Ar.), 5.20 (s, 2H, OCH2),
4.20 (s, 2H, CH2), 3.95 (s, 2H, CH2). Exact mass calcd
for C12H12F3N3O [M+H]+ 272.1010, found: 272.1002.
1,6-Dihydro-5-(4H)-pyrimidinone O-(3-phenyl-2-propynyl)-
oxime hydrochloride (32). Starting from diamino acetone
dihydrochloride monohydrate and 3-phcnyl-2-propynyl-
(E)-1,6-Dihydro-5-(4H)-pyrimidinone O-(2-penten-4-ynyl)-
oxime (Z)-2-butenedioate (24). Sodium (1.7 g, 74 mmol)
was dissolved in 150 mL dry methanol and hydroxyl-
20
.
.
amine HCl (2.66, 37 mmol) was added. After 1 h stirring
hydroxylamine HCl
the diamino derivative was
prepared. 1H NMR: (MeOD, 200 MHz) d 7.50±7.30 (m,
5H, Ar H's), 5.10 (s, 2H, OCH2), 4.05 (s, 2H, CH2), 3.90
(s, 2H, CH2). Cyclisation with trimethylorthoformate
gave 32 in 55% overall yield. 1H NMR: (MeOD,
200 MHz) d 8.05 (s, 1H, N=CH) 7.50±7.30 (in, 5H, Ar
H's), 5.00 (s, 2H, OCH2), 4.40 (s, 2H, CH2), 4.15 (s, 2H,
CH2). Exact mass calcd for C13H13N3O[M+H]+
228.1089, found: 228.1137.
at room temperature, a solution of 3a in 100 mL dry
dichloromethane was added slowly. A dark red suspen-
sion appeared. After 2 h, maleic acid (8.6 g, 74 mmol)
was added. The solvent of the now white suspension was
evaporated. Recrystallisation of the crude residue in
(ethylacetate:ethanol, 9:1) gave the diamino oxime deri-
vative as a crude product. Cyclisation with trimethyl-
orthoformate gave 24 as a maleic acid salt. 1H NMR:
(CDCl3, 200 MHz) d 8.15 (s, 1H, CH=N), 6.25 (m, 1H,
=CH), 6.25 (s, 2H, mal), 5.75 (d, 1H, =CH), 4.70 (d,
2H, OCH2), 4.35 (s, 2H, CH2N), 4.10 (s, 2H, CH2N),
3.30 (s, 1H, CH).
1,6-Dihydro-5-(4H)-pyrimidinone O-[3-(3-methoxyphenyl)-
2-propynylloxime hydrochloride (33). Starting from
diamino acetone dihydrochloride monohydrate and 3-
20
.
(3-methoxyphenyl)-2-propynylhydroxylamine HCl the
diamino derivative was prepared. 1H NMR: (MeOD,
200 MHz) d 7.30±6.90 (m, 4H, Ar H's), 5.10 (s, 2H,
OCH2), 4.00 (s, 2H, CH2), 3.85 (s, 2H, CH2), 3.80 (s,
3H, OCH3). Cyclisation with trimethylorthoformate
gave 33 in 44% overall yield. 1H NMR: (MeOD,
200 MHz) d 8.15 (s, 1H, N=CH) 7.30±6.90 (m, 4H, Ar
(E)-1,3-Dichloroacetone O-(3-phenyl-2-propenyl)oxime
(2b). Starting from 1,3-dichloro-acetone and (E)-3-phe-
.
nyl-2-propenylhydroxylamine HCl the dichloro deriva-
tive 2b was prepared according the procedure described
for 2a in 100% crude yield. 1H NMR: (CDCl3,
200 MHz) d 7.50±7.20 (m, 5H, Ar H's), 6.70 (d, 1H,
1

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1417
=CH), 6.40 (dt, 1H, =CH), 4.70 (d, 2H, OCH2), 4.45
(s, 2H, CH2Cl), 4.35 (s, 2H, CH2Cl).
the reaction product crystallized spontaneously. Crys-
tallisation from methanol/ethylacetate gave 38 in 18%
yield. 1H NMR: (MeOD, 200 MHz) d 8.13 (s, 1H,
N=CH), 4.25 (m, 1H, CH), 3.85 (d, 4H, CH2). Exact
mass calcd for C4H9N3O [M+H]+ 116.0820, found:
.
116.0824. Anal. calcd C4H9N3O 2HC1): 25.54% C,
5.89% H, 22.34% N. Found: 25.95% C, 5.64% H,
22.45% N.
(E)-1,3-Diphthaloylacetone O-(3-phenyl-2-propenyl)oxime
(3b). Starting from potassium phthalimide and 2b the
diphthaloyl derivative 3b was prepared according to the
procedure described for 3a in 58% yield. 1H NMR:
(CDCl3, 200 MHz) d 7.85±7.60 (m, 8H, Ar H's), 7.50±
7.20 (m, 5H, Ar H's), 6.45 (d, 1H, =CH), 6.20 (dt, 1H,
=CH), 4.70 (d, 2H, OCH2), 4.60 (s, 2H, CH2N), 4.50 (s,
2H, CH2N).
According to the general procedure II the following
derivatives were prepared
(E)-1,6-Dihydro-5-(4H)-pyrimidinone O-(3-phenyl-2-pro-
penyl)oxime (Z)-2-butenedioate (34). According to the
procedure described for 24 the dihydropyrimidine deri-
1,4,5,6-Tetrahydro-N-methoxy-5-pyrimidinamine dihydro-
chloride (39). Starting from
imidinamine 39 was prepared in 38% yield. H NMR:
(MeOD, 200 MHz) d 8.13 (s, 1H, N=CH), 4.10 (m, 1H,
CH), 3.90 (s, 3H, OCH3), 3.80±3.65 (m, 4H, CH2).
Exact mass calcd for C5H11N3O [M+H]+ 130.0991,
.
found: 130.0980. Anal. calcd (C5H11N3O 2HC1):
29.72% C, 6.48% H, 20.79% N. Found: 28.95% C,
6.57% H, 19.86% N.
5 the tetrahydropyr-
1
1
vative 34 was prepared in 39% overall yield. H NMR:
(MeOD, 200 MHz) d 8.15 (s, 1H, CH=N), 7.50±7.70
(m, 5H, Ar H's), 6.65 (d, 2H, =CH), 6.40 (dt, 2H,
=CH2), 6.25 (s, 2H, mal), 4.80 (d, 2H, OCH2), 4.30 (s,
2H, CH2), 4.10 (s, 2H, CH2). Exact mass calcd for
C13H15N3O [M+H]+ 230.1277, found: 230.1293.
(E)-1,6-Dihydro-2-methyl-5-(4H)-pyrimidinone O-(3-
phenyl-2-propenyl)oxime hydrochloride (35). Starting
1,4,5,6-Tetrahydro-N-ethoxy-5-pyrimidinamine dihydro-
chloride (40). Starting from 6 the tetrahydropyr-
imidinamine 40 was prepared in 76% yield. H NMR:
(D2O, 200 MHz) d 8.05 (s, 1H, N=CH), 4.20±4.00 (m,
1H, CH), 3.80±3.65 (d, 4H, CH2), 1.25 (t, 3H, CH3).
Exact mass calcd for C6H13N3O [M+H]+ 144.1949,
.
found: 144.1961. Anal. calcd (C6H13N3O 2HC1):
33.34% C, 6.99% H, 19.44% N. Found: 32.40% C,
6.85% H, 18.68% N.
1
from diamino acetone dihydrochloride monohydrate
.
and 2-propenylhydroxylamine HCl the diamino deriva-
tive was prepared. 1H NMR: (MeOD, 200 MHz) d 6.20±
5.80 (m, 1H, =CH), 5.40±5.20 (m, 2H, =CH2), 4.65 (d,
2H, -OCH2), 4.00 (s, 2H, CH2), 3.85 (s, 2H, CH2).
Treatment with trimethylorthoacetate gave 35 in 32%
1
overall yield. H NMR: (MeOD, 200 MHz) d 6.10±5.85
(m, 1H, =CH), 5.30±5.10 (m, 2H, =CH2), 4.60 (d, 2H,
OCH2), 4.30 (s, 2H, CH2), 4.10 (s, 2H, CH2), 2.25 (s,
3H, CH3). Exact mass calcd for C,H,N,O [M+H]+
168.1129, found: 168.1137.
In Vitro Studies
Agonist and antagonist binding studies
(E)-1,6-Dihydro-4-methyl-5-(4H)-pyrimidinone
O-(3-
Binding of [N-methyl-3H]-oxotremorine-M acetate (3H-
Oxo-M) in homogenates of frontal cortex. The rapid ®l-
tration method of Freedman et al.32 was used to mea-
sure the agonist character of muscarinic cholinergic
drugs in rat cerebral cortex homogenates. For routine
measurements the concentration of [3H]-Oxo-M was
0.5 nM, tissue concentration was about 1 mg/mL origi-
nal tissue and incubation was for 40 min at 30 ꢀC. Non-
speci®c binding was de®ned as the amount of binding of
[3H]-Oxo-M in the presence of 2 mM atropine sulphate
and represented about 10% of total binding.
phenyl-2-propenyl)oxime hydrochloride (37). Starting
from 1,3-diamino-1-methylacetone and O-ethylhydroxyl-
amine the diamino derivative 36 was prepared. Cyclisa-
tion with trimethyl orthoformate gave 37 in 45% overall
yield. 1H NMR: (MeOD, 200 MHz) d 7.95 (s, 1H,
CH=N), 4.15 (s, 2H, -CH2-), 4.00 (q, 2H, OCH2), 1.30
(d, 3H, CH3), 1.05 (t, 3H, CH3). Exact mass calcd for
C7H13N3O [M+H]+ 156.1110, found: 156.1137.
General procedure II: Tetrahydro-N-hydroxy-5-
pyrimidinamines
Binding of [N-methyl-3H]-pirenzepine ([3H]-PZ) in
homogenates of rat forebrain. The rapid ®ltration
method of Freedman et al.32 was used to characterize
M1-muscarinic cholinergic properties of drugs in rat
forebrain membranes. For routine measurements the
concentration of [3H]-Pz was 1 nM, tissue concentration
was about 10 mg/mL original tissue and incubation was
for 60 min at 25 ꢀC. Nonspeci®c binding was de®ned as
3,4,5,6-Tetrahydro-N-hydroxy-5-pyrimidinamine dihydro-
chloride (38). A solution of 1,6-dihydro-5-(4H)pyr-
imidinone oxime hydrochloride 4 (0.980 g, 6.6 mmol) in
dry methanol was added dropwise to a stirred solution
of HC1/methanol (6.5 mL of a 5.2 N solution) and tri-
methylamine borohydride (0.525 g, 7.2 mmol) at room
temperature under a nitrogen atmosphere. After 24 h

1418
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
the amount of binding of [3H]-Pz in the presence of
1 mM atropine sulphate and represented about 20% of
total binding.
Glass recording electrodes are pulled from 1.0 or
1.5 mm boro-silicate capillaries with ®lament, using a
conventional horizontal puller. These electrodes are ®l-
led with NaCl solution in water (4 mol/L) resulting in an
electrode resistance of approx. 5 MOhm. The recording
electrode is placed approx. 200 mm below the top surface
of the slice and at a distance of approx 600 mm from the
stimulation electrode in the stratum radiatum of CA1.
Synaptic responses are elicited by 100 ms, constant cur-
rent, monophasic square wave pulses at a frequency of
0.033 Hz. The slice is stimulated at submaximal intensity
until a stable response (1±3 mV fEPSP amplitude) is
obtained. The stimulus intensity is then adjusted so that
the evoked fEPSP apparently does not trigger post-
synaptic cells to ®re an action potential, i.e. no `Popu-
lation Spike' is observed. Signals are ampli®ed (gain
1000), ®ltered (DC-5 kHz), monitored on an oscillo-
scope and taped for o€-line analysis using a digital tape
recorder (DTR 1800).
Evaluation of the data. Displacement curves were
obtained for the various compounds by measuring the
speci®c binding in the present of at least four di€erent
concentrations and IC50 values were obtained using a
four parameter ®tting procedure. Ki values were
obtained from the IC0 values by using the Cheng±Prus-
o€ equation Ki=IC50/1+C/Kd in which equals the
radiolabelled ligand concentration and Kd equals the
dissociation constant for the radiolabelled ligand. Kd
values used for these calculations were as follows: [3H]-
Oxo-M binding: Kd=0.7 nM; [3H]-Pz binding:
Kd=8.3 nM.
Interactions with muscarinic subtype mediated responses
in isolated organs
M1-mediated activity in the hippocampal slice. Muscari-
nic cholinergic agonists e€ectively suppress the elec-
trically evoked ®eld excitatory postsynaptic potential
(FEPSP) in the rat hippocampal slice.33 This e€ect is
due to a decrease in release of excitatory amino acids
from the Scha€er collateral/commissural ®bers, prob-
ably mediated by a presynaptic M1 muscarinic recep-
tor.33 Carbachol (VI) (3E-5 mol/L 1) causes a 100%
decrease (intrinsic activity=1.0) in the amplitude of the
electrically evoked FEPSP an e€ect can be fully antag-
onized by the M1 selective antagonist pirenzepine.
Procedure. At least two control cycles consisting of four
stimulation pulses (1/15 Hz) are run before drug admin-
istration. If the control cycles are comparable, the test
or reference compound is applied to the slice by switch-
ing between the non drug and drug containing solutions
using a miniature Hamilton valve. The slice is exposed
to a drug concentration for a maximum of 10 min. At 2,
5, and 8 min of application a test cycle consisting of four
stimulation pulses (1/15 Hz) is run. Each test or refer-
ence compound is initially tested at successively-admi-
nistered concentrations of 1E-6, 1E-5, and IE-4 mol/L in
one slice to obtain a cumulative concentration±response
curve. Dependent on the activity found the concentra-
tion steps are adapted for a follow up experiment in
another slice. If the fEPSP is maximal suppressed or the
concentration reaches 3E-4 mol/L the slice is replaced.
Each test compound is tested in at least two slices and a
washout response is generally obtained.
Preparation of hippocampal slices. Rats (200±350 g;
strain: Hsd/Cpb:WU, SPF bred by Harlan CPB, Zeist,
The Netherlands) are killed by decapitation using a
small guillotine. The brain is taken from the skull and
rinsed in ice-cold (1±2 ꢀC) arti®cial CSF gassed with
carbogen (95% 02+5% CO2). The hippocampus is
taken from either hemisphere and transverse slices
(thickness 500 mm) are chopped using a McIlwain tissue
chopper. Slices from the middle one third of the hippo-
campus are separated using two ®ne nylon paint brushes
and subsequently transferred to a slice holding chamber
containing ACSF at room temperature (approx. 22 ꢀC).
Evaluation of responses. The e€ects of the test com-
pounds are evaluated on the basis of amplitude mea-
surements of the fEPSP before (control) and during their
application. The fEPSP amplitude during application is
expressed as a percentage of the fEPSP amplitude before
application. A line is ®tted through the data points of
the log drug concentration±e€ect relationship and the
pD2 value (negative logarithm of agonist concentration
causing 50% inhibition) and a (ecacy) calculated.
Electrical stimulation and recording. After a recovery
period of at least 1 h a single slice is placed in the tem-
perature controlled recording chamber (capacity
approx. 0.8 mL, T=34±35 ꢀC) and sandwiched between
two nylon meshes by adjusting the top mesh in height
with a micromanipulator. The slice is continuously
gravity fed superfused with aCSF (gassed with carbo-
gen) resulting in 3±4 turnovers per min. The e‚uent is
removed by vacuum. A bipolar concentric stimulation
electrode is placed in the stratum radiatum of CA1 near
the border with CA2/CA3.
M2-mediated e€ects. Interactions with M2-muscarinic
cholinergic receptors were studied in the isolated left
atrium of the rat. An automated assay was used similar
to the (1-adrenoreceptor described previously34 but with
the following adaptation. The inhibition by cholinergic
agonists of the electrical stimulation evoked switch of
the atrium was measured using carbachol (VI) as a

R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
1419
reference. Cholinergic agonistic activity was compared
to carbachol. Potential M2 activity was veri®ed via
antagonism in the presence of the M2-selective antago-
nist AFDX 1 16. Antagonistic activity was measured as a
shift to the right of the dose±response curve to carbachol
in the presence of the compound as described above.
PC. Training in the swim maze followed the procedure
described in detail;37 the same 2.1 m diameter water
maze, cues and islands used by Hagan et al. were used in
this study. After stabilization training, the rats were
implanted with guide cannula into the lateral ventricle
as described previously.37,38 For the delayed matching
experiment, rats were infused with 5 mL of merlys, or
5 mL of merlys containing 1 mg of the cholinergic
depleting agent hemicholinium-3, 1 h before testing in
the operant chambers. Rats in the swim maze received
the same vehicle and infusion volume, but a concentra-
tion of 5 mg of hemicholinium-3. Hemicholinium-3 leads
to a central depletion of acetylcholine and leaves both
central and peripheral cholinoreceptors open for stimu-
lation and may therefore allow a more relevant test of
the cognitive e€ects of cholinergic agonist: Alzheimer's
patients su€er from loss of transmitter, not blockade of
receptors. Thirty minutes before the session, rats were
administered the test drug sc in a saline vehicle. For
both the delayed matching and swim maze experiments
®ve drug treatments were given to groups of 8±10 rats in
a latin square design: typically: placebo icv and sc;
hemicholinium-3 icv and placebo sc; and hemi-
cholinium-3 icv and three separate doses of the test
compound. Several parameters were recorded in both
procedures. Data for correct responses, trials completed
and response time latencies for both swim maze and
delayed matching data were analyzed using ANOVA
and, where signi®cance was attained, by appropriate
post hoc testing.
M3-mediated e€ects. Interactions with M3-muscarinic
cholinergic receptors were measured in the isolated gui-
nea pig ileum. A fully automated method was used as
described previously.34 Contractions induced with acet-
ylcholine as an agonist (pD2-values between 6 and 7)
were used to evaluate the potency of cholinergic
antagonist. Antagonistic activity was measured as
described above.
In vivo studies
Miosis. Antagonism of clonidine-induced mydriasis was
studied using the method described by Hagan et al.35
Male Wistar rats (250±300 g, Hsd/Cpb:WU, Harlan,
Zeist, Netherlands) were anaesthetized with 60 mg/kg of
Nembutal, placed on a heated blanket and the pupil
diameter measured. Clonidine (0.3 mg/kg) or placebo is
then administered sc and 20 min later the pupil diameter
measured. Immediately following this measurement
separate groups of rats (n=6) are administered placebo
or one of the drug doses in a 20 mL drop applied directly
onto the eye. Thereafter the pupil is measured at 10, 30,
and 60 min after application. Changes in pupil diameter
are calculated with respect to levels induced by clonidine.
E€ects on sleep electroencephalograph (EEG) activity
Salivation. The induction of salivation was measured
using mice. Male mice (30 g, CD-1, Charles Rivers,
Germany) were anaesthetized using Avertine. After
10 min separate groups of mice (n=6) were injected sc
with several doses of the test drug in a saline/mulgofen
vehicle (1±10 mg/kg). The mice were then placed on a
®lter paper for 20 min. After this time the ®lter paper
was removed and the extent of any stain due to saliva-
tion measured. A dose for which the resulting stain
equated 1 cm2 was recorded as the minimal e€ective dose.
Di€erent stages of sleep±waking behaviour (active wak-
ing, quiet waking, quiet sleep, deep sleep, pre-REM
sleep and REM sleep) can be identi®ed on the basis of
EEG recordings from the cerebral cortex, a measure of
muscle tone and a movement index. Most psychotropic
compounds alter sleep in a speci®c way and can be
classi®ed either according to drug class (anxiolytic)
or speci®c drug e€ect (benzodiazepine); see ref. 38 for
discussion.
Reversal of hemicholinium-3-induced memory de®cit.
Drug e€ects on rat short-term spatial memory were
assessed using a minor variation of the delayed match-
ing to position procedure reported by Dunnett36 and the
two island swim task in the Morris swim maze as used
by Hagan et al.37 For both tasks separate groups of
male rats (350 g, Hsd/Cpb: Long Evans, Harlan, Zeist,
Netherlands) were used. Training in the delayed match-
ing to position task has been described in detail else-
where.37 Rats were trained on the spatial matching task
in standard operant chambers (Coulbourn Instruments,
PA, USA) connected to a MEDLab interface (Med
Associates, VT, USA) and controlled by an IBM PS2
Full details of the procedure have been described in
detail elsewhere.39 Brie¯y, epidural screw electrodes
were implanted over the parietal-occipital cortex of male
rats (Hsd/Cpb:Wu, Harlan, Zeist, Netherlands; 250±
300 g) for recording of EEG against a frontal reference
electrode. Stainless steel electrodes were inserted into
the dorsal neck muscles for recording of electro-
myogram. Twenty-nine hour EEG and EMG recordings
were made in sound attenuated Faraday cages. Move-
ments were detected as capacitative arterartefacts
generated in an open-ended wire of the cable connecting
the rats to the swivel commutator. The swivel was con-
nected to ampli®cation and A/D conversion units

1420
R. Plate et al./Bioorg. Med. Chem. 6 (1998) 1403±1420
attached to a dedicated PDP-11/83 minicomputer sys-
tem for on-line spectral EEG and data compression.
14. Levey, A. I. Life Sci. 1993, 52, 441.
15. Freedman, S. B.; Dawson, G. R.; Iversen, L. L.; Baker, R.;
Hargreaves, R. J. Life Sci. 1993, 52, 489.
O€-line sleep staging was done for 2s epochs based on
®ve spectral EEG band values,39 the integrated EMG
and movement level. A ®rst sleep stage assignment per
epoch is done by application of a discriminant function
to these epoch values. A moving average EEG smooth-
ing procedure and a set of syntactic classi®cation rules
are then used to give the ®nal sleep stage assignments to
each speci®c EEG epoch. Six sleep±wake states are dis-
tinguished including two waking states: active waking,
characterized by movement, theta activity and high
EMG, and quiet waking, without movement. Four sleep
stages are discriminated: quiet sleep, characterized by
EEG spindles; deep slow wave sleep with prominent
delta activity; preREM sleep with spindles against a
background of theta activity and low EMG; REM sleep
with theta activity and low EMG.
16. Flynn, D. D.; Ferrari-DiLeo, G.; Levey, A. I.; Mash, D. C.
Life Sci. 1995, 56, 869.
17. Dawson, G. R.; Bayley, P.; Channell, S.; Iversen, S. D.
Psychopharmacology 1994, 113, 361.
18. Sedman, A. J.; Bockbrader, H.; Schwarz, R. D. Life Sci.
1995, 56, 877.
19. Bymaster, F. P.; Whitesitt, C. A.; Shannon, H. E.; Delapp,
N.; Ward, J. S.; Calligaro, D. O., Shipley, L. A.; Buckelsam, J.
L.; Bodick, N. C.; Farde, L.; Sheardown, M. J.; Olesen, P. H.;
Hansen, K. T.; Suzdak, P. D.; Swedberg, M. D. B.; Sauerberg,
P.; Mitch, C. H. Drug Devel. Res. 1997, 40, 158.
20. Bodick, N. C.; O€en, W. W.; Shannon, H. E.; Satterwhite,
J.; Lucas, R.; Vanlier, R.; Paul, S. M. A.D. Assoc. Disord. 1997,
11, S16.
21. Plate, R.; Plaum, J. M.; Boer, Th. de; Andrews, J. S.
Bioorg. Med. Chem. 1996, 4, 239.
22. Plate, R. Eur. Pat. 93200553.1, 1993.
23. Messer, W. S., Abuh, Y. F., Liu, Y., Periyasami, S., Ngur,
D. O., Edgar, M. A. N., El-Assadi, A. A., Sbeih, S., Dunbar, P.
G., Roknich, S., Rho, T., Fang, Z., Ojo, B., Zang, H., Huzl, J.
J., Nagy, P. I. J. Med. Chem. 1997, 40, 1230; and references
cited therein.
Drugs are administered at the beginning of the light
cycle of the rats. Drug e€ects are assessed on a number
of parameters including percentage time spent in each of
the sleep stages per 30 min period, allowing a pro®le of
changes over sleep stages and over time.
24. Grochowski, E.; Jurczak, J. Synthesis 1976, 682.
25. Fujii, T.; Wu, C.; Yamada, S. Chem. Pharm. Bull. 1967, 345.
References
26. Henneke, K. W.; Schollkopf, U.; Neudecker, T. Just. Lieb.
Ann. 1979, 1370.
1. Perry, E. K.; Tomlinson, B. E.; Blessed, G.; Bergmann, K.;
Gibson, P. H.; Perry, R. H. Brit. Med. J. 1978, 2, 1457.
27. Walland, A.; Burkard, S.; Harnmer, R.; Troger, W. Life
Sci. 1997, 60, 977.
2. Coyle, J. T.; Price, D. L.; Delong, M. R. Science 1983, 217,
1184.
28. Ensinger, H. A.; Bechtel, W. D.; Birke, F. W.; Mendla, K.
D.; Mierau, J.; Speck, G.; Troger, W. Drug Devel. Res. 1997,
144.
3. Mash, D. C.; Flynn, D. D.; Potter, L. T. Science 1985, 228,
1115.
29. Schwarz, R. D.; Callahan, M. J.; Davis, R. E.; Jaen, J. C.;
Tecle, H. Drug Devel. Res. 1997, 40, 133.
4. Araujo, D. M.; Lapchak, P. A.; Robitaille, Y.; Gauthier, S.;
Quirion R. Naunyn-Schmiedeberg's Arch. Pharmacol. 1988,
342, 625.
30. Plate, R.; Plaum, J. M.; Boer, Th de; Andrews, J. S; Rae,
D. R.; Gibson, S. Bioorg. Med. Chem. 1996, 4, 227.
5. Eagger, S. A.; Levy, R.; Sahakian, B. J. Lancet 1991, 337,
989.
31. Doods, H. N.; Quiron, R.; Mihn, G.; Engel, W.; Rudolf,
K.; Enzeroth, M.; Schiavi, G. B.; Ladinsky, H.; Bechtel, W. D.;
Ensinger, H. A.; Mendla, K. D; Eberlein, W. Life Sci. 1993, 52,
497.
6. Lamy, P. P. C N S Drugs 1994, 1, 146.
7. Maltby, N.; Broe, G. A.; Creasey, H.; Jorm, A. F.; Chris-
tensen, H.; Brooks, W. S. Brit. Med. J. 1994, 308, 879.
32. Freedman, S. B.; Beer, M. S.; Harley, E. A. Eur. J. Phar-
macol. 1988, 156, 133.
8. Asthana, S.; Ra€aele, K. C.; Berardi, A.; Greig, N. H.;
Haxby, J. V.; Schapiro, M. B.; Soncrant T. T. A.D. Assoc.
Disord. 1995, 9, 223.
33. Sheridan, R. D., Sutor, B. Neurosci. Lett. 1990, 108, 273.
34. De Graaf, J. S.; De Vos, C. J.; Steenbergen, H. J. J. Phar-
macol. Meth. 1983, 10, 113.
9. Bryson, H. M.; Ben®eld, P. Drugs Aging 1997, 10, 234.
10. Davis, K. L.; Hollander, E.; Davidson, M.; Davis, B. M.;
Mohs, R. C.; Hovath, T. B. Am. J. Psych. 1987, 144, 468.
35. Hagan, J. J.; van der Heijden, B.; Broekkamp, C. L. E.
Naunyn-Schiederberg's Arch. Pharmacol. 1987, 338, 476.
11. Hollander, E.; Davidson, M.; Mohs, R. C.; Horvath, T. B.;
Davis, B. M.; Zemishlany, Z.; Davis, K. L. Biol. Psych. 1987,
22, 1067.
36. Dunnett, S. B. Psychopharmacology 1985, 87, 357.
37. Hagan, J. J.; Jansen J. H. M.; Broekkarnp, C. L. E. Psy-
chopharmacology 1989, 98, 347.
12. Soncrant, T. T.; Ra€aele, K. C.; Asthana, S.; Beradi, A.;
Morris, P. P.; Haxby, J. V. Psychopharmacology 1993, 112, 421.
38. Andrews, J. S.; Jansen, J. H. M.; Linders, S.; Princen, A.
Psychopharmacology 1994, 115, 485.
13. Asthana, S.; Ra€aele, K. C.; Berardi, A.; Greig, N. H.;
Morris, P. P.; Schapiro, M. B.; Rapoport, S. I.; Blackman, M.
R.; Soncrant, T. T. Psychoneuroendocrinology 1995, 20, 623.
39. Ruigt, G. S. F.; Engelen, S.; Gerrits, A.; Verbon, F.
Neuropsychobiology 1993, 28, 138.