Novel arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-dioxides as platelet aggregation inhibitors. 2. Optimization by quantitative structure-activity relationships

Article abstract of DOI:10.1021/jm981104b

In the previous paper (Part 1), we described the synthesis and antiplatelet activity of a series of phenyl- and heteroarylpyrazino[2,3- c][1,2,6]thiadiazine 2,2-dioxides. In this paper, we report the optimization of the platelet aggregation inhibitory activity by an iterative sequence of quantitative structure-activity relationship studies which encompassed synthesis and evaluation of the effects of structure variations at the 1-, 6- , and 7-positions of the heterocyclic system. A model has been established that correctly correlates antiplatelet activity in this series with the partial atomic charges calculated by a local density functional ab initio method. As a result of this study, the experimental platelet aggregation inhibitory activity of the lead compound was improved 300-fold.

Full text of DOI:10.1021/jm981104b

J . Med. Chem. 1999, 42, 3279-3288
3279
Novel Ar ylp yr a zin o[2,3-c][1,2,6]th ia d ia zin e 2,2-Dioxid es a s P la telet Aggr ega tion
In h ibitor s. 2. Op tim iza tion by Qu a n tita tive Str u ctu r e-Activity Rela tion sh ip s‡
Nuria Campillo,* Pilar Goya, and J uan A. Pa´ez
Instituto de Quı´mica Me´dica, CSIC, J uan de la Cierva, 3, 28006 Madrid, Spain
Received October 13, 1998
In the previous paper (Part 1), we described the synthesis and antiplatelet activity of a series
of phenyl- and heteroarylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-dioxides. In this paper, we report
the optimization of the platelet aggregation inhibitory activity by an iterative sequence of
quantitative structure-activity relationship studies which encompassed synthesis and evalu-
ation of the effects of structure variations at the 1-, 6-, and 7-positions of the heterocyclic system.
A model has been established that correctly correlates antiplatelet activity in this series with
the partial atomic charges calculated by a local density functional ab initio method. As a result
of this study, the experimental platelet aggregation inhibitory activity of the lead compound
was improved 300-fold.
In tr od u ction
information about the target is available is to modify
the structure by variations of the substituents at
particular sites. Thus, in a first approach, three different
positions of the pyrazinothiadiazine system were con-
sidered responsible for modulating the biological activity
and were subject to modification.
In position 1 different groups such as methyl, ethyl,
tert-butoxy- and methoxycarbonylmethyl, carbamoyl-
methyl, and benzyl were chosen since from our previous
experience the corresponding N(1)H-pyrazino[2,3-c]-
[1,2,6]thiadiazine derivatives are devoid of activity. For
positions 6 and 7, a variety of alkyl, aryl, halogen,
alkoxy, and hydrogen were considered with the restric-
tion that at least one of them should be an aromatic
group which is necessary, according to our preliminary
results for antiplatelet activity.
In a previous paper,1 we described the synthesis and
pharmacological evaluation of a new family of platelet
aggregation inhibitors derived from 6- or 7-phenylpyrazi-
no[2,3-c][1,2,6]thiadiazine 2,2-dioxide. The most promis-
ing compounds of the series were N-1-substituted 6,7-
diphenyl- and 6-phenylpyrazino[2,3-c][1,2,6]thiadiazine
derivatives which showed significant inhibition of plate-
let aggregation. The fact that pyrazinothiadiazines are
in general deprived of significant toxicity, together with
the preliminary results obtained ex vivo of their inhibi-
tion of platelet aggregation induced by different ago-
nists, prompted us to optimize this series and to design
new active compounds by means of a QSAR study.
The general methodology for this kind of work has
been followed establishing a mathematical model which
correlates the pharmacological activity of the compounds
with some parameter that describes the structural
features.2,3 For the measurement of the pharmacological
activity, a quick simple method, i.e., determination of
the EC100, has been used, and as structural descriptors
the partial atomic charges obtained by quantum me-
chanical calculations were used. The initial model was
then used to design compounds with a predicted in-
creased activity which were synthesized and their
experimental antiplatelet properties determined and
confirmed. These results were used to establish the final
model which was validated from a statistical point of
view.
Taking into account all these facts, we selected among
the different pyrazino[2,3-c][1,2,6]thiadiazine 2,2-diox-
ides previously synthesized by us a first training set of
17 derivatives (1-17) which are gathered in Table 1.
P h a r m a cologica l Activity. As pharmacological re-
sponse we have used the effective concentration (EC100
,
µg/mL) which produced a 100% of inhibition of platelet
aggregation (rabbit platelet-rich plasma) induced by
arachidonic acid.4 This relatively simple method for
determining the platelet aggregation inhibitory activity
is sufficient for our mathematical model, and as already
mentioned, we have previously observed that com-
pounds giving positive results in this assay also show
interesting properties ex vivo.
QSAR Stu d y
Str u ctu r a l P a r a m eter s. The standard Hansch pa-
rameters were not suitable for these molecules,2 so we
decided to use data derived from quantum mechanical
ab initio calculations to correctly describe the struc-
ture. For each of the 17 selected compounds, the
Hirshfeld partial charges of each atom of the heterocycle
were calculated by a local density functional (LDF) ab
initio method5-7 using the DMol program.8 The feas-
ability of this method for calculating molecules incor-
porating the N-SO2-N rest had already been proved
Tr a in in g Set. Compound 1, 1-ethyl-6,7-diphenylpyr-
azino[2,3-c][1,2,6]thiadiazine 2,2-dioxide, with an EC100
) 0.3 µg/mL was chosen as the lead compound. One
possible way to optimize a lead when no structural
‡ This paper is dedicated to Professor J ose´ Elquero on his 65th
birthday.
* To whom correspondence should be addressed. Fax: 91 5644853.
E-mail: nuria@suricata.iqm.csic.es.
10.1021/jm981104b CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/10/1999

3280 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Campillo et al.
Ta ble 1. Experimental and Calculated EC100 and Residual of Compounds 1-45 from Eq 1 and Final Model Eq 2
eq 1
eq 2
a
compd
R1
R2
R3
EC100
resid
fitted
pred
resid
fitted
pred
1b,c
2b,c
3b,c
4b,c
5b,c
6b,c
7b,c
8b,c
9b,c
10b,c
11b,c
12b,c
13b,c
14b
15b
16b
17b
18c
19c
20c
21c
22c
23c
24c
25c
26
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Me
Br
Ph
Ph
Ph
Ph
Ph
H
Et
CH2Ph
0.3
3.0
0.1
1.0
3.0
0.3
1.0
1.0
1.0
10
10
3.0
-0.428
0.728
2.171
1.135
1.262
1.350
0.973
0.411
1.747
1.376
10.688
9.349
2.586
•
-0.842
1.142
2.795
1.112
1.302
1.438
1.023
0.450
2.090
1.803
10.345
8.961
2.327
•
0.829
•
0.205
•
CH2CO2Et
-1.035
•
-1.012
•
CH2CO2tBu
-0.262
•
-0.302
•
CH2CONH2
Me
Et
Et
Et
CH2CO2Et
Et
Me
Et
Et
CH2CO2Et
Me
Et
Me
Me
Me
Me
Et
Me
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
1.650
•
1.562
•
-0.673
0.589
•
-0.723
0.550
•
H
•
•
Me
Ph
Ph
Ph
2-pyridyl
2-thienyl
Ph
Ph
2-furyl
Ph
4-MeOPh
H
OMe
4-FPh
4-ClPh
4-BrPh
OEt
-0.747
•
-1.090
•
-0.376
•
-0.803
•
-0.688
•
-0.345
•
Cl
0.651
•
1.039
•
2-pyridyl
2-thienyl
H
0.414
0.076
•
0.674
•
3.0
2.924
•
-0.621
3.621
•
>10
>10
1.0
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
18.436
16.949
2.895
13.247
-1.538
-0.233
-1.459
0.263
2.180
0.303
-1.419
12.423
17.037
-2.512
2.426
6.828
1.104
1.402
1.452
2.999
1.075
0.978
0.360
17.016
17.651
17.584
9.232
1.125
1.740
2.705
3.549
3.475
•
•
17.616
15.947
3.611
12.479
•
H
•
•
2-furyl
Br
4-MeOPh
2-FPh
Ph
4-FPh
4-ClPh
4-BrPh
Ph
4-ClPh
H
3-MeOPh
4-ClPh
2-ClPh
Ph
4-ClPh
Ph
4-NO2Ph
4-ClPh
4-MePh
Ph
•
•
1.0
•
•
0.001
0.03
0.1
0.3
3.0
1.0
0.30
10
0.569
-0.568
0.292
0.150
-0.350
0.616
0.424
0.230
-0.223
-0.262
•
-0.050
•
0.650
•
2.384
0.576
0.070
10.223
•
•
•
OEt
•
4-MePh
3-MeOPh
4-ClPh
2-ClPh
4-ClPh
Ph
>10
>10
NDe
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
16.658
-1.773
2.494
6.438
1.424
1.706
1.240
2.909
1.564
1.368
1.100
16.417
16.270
16.434
8.896
1.268
1.953
2.606
3.039
3.118
27
28d
29d
30d
31d
32d
33d
34d
35d
36d
37d
38d
39d
40d
41d
42d
43d
44d
45d
4-NO2Ph
Ph
4-MePh
4-ClPh
4-HOPh
4-ClPh
4-NO2Ph
3-NO2Ph
4-ClPh
H
H
H
H
H
H
H
H
Cl
Et
Me
Me
Et
Et
Me
4-MePh
4-MeOPh
4-ClPh
4-NO2Ph
3-NO2Ph
a
Activity expressed as the EC100 (µg/mL), effective concentration which produced 100% inhibitory response. ASA, EC100 ) 2.5 µg/mL.
Compounds included in ref 1. c Compounds included in the model 2. Compounds not synthesized. e Not determined.
b
d
by us, in a previous study in which different quantum
platelet aggregation inhibitory activity can be expressed
as a function of the atomic charges of N5, C7, C4a, N1,
and C6 atoms by the following eq 1 (model 1):
mechanical calculations were tested on related sub-
stances.9,10
QSAR An a lysis
EC100 ) -18.04 + 1719.91X1 + 4482.12X2 +
40658.12X3 - 79373.68X4 (1)
The structures and experimental platelet aggregation
inhibitory activities of the training set compounds (1-
17) are gathered in Table 1, and the values of the
calculated charges are in Table 2.
Different models involving combinations of all the
partial atomic charges were tested according to standard
deviation (s) and correlation coefficient (R2) criteria. The
initial models tested were those including the variables
which showed a higher standard deviation and no
significant collinearity. In the best model found, the
n ) 13 (1-13) R2 )0.95 Ra ) 0.92 s ) 0.96
2
F4,8 ) 35.39 p < 0.0001
X1 ) (N5)2 X2 ) (N5)2C7 X3 ) (N5)2C4a
X4 ) (N1)2C6
It should be noted that, although traditional QSAR is

Arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3281
Ta ble 2. Hirshfeld Partial Charges (au) of Candidate Compounds 1-45a
compd
R1
R2
R3
C-4a
N-5
C-6
C-7
N-1
1
2
3
4
5
6
7
8
9
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Me
Br
Ph
Ph
Ph
Ph
Ph
H
Et
CH2Ph
0.01510
0.01250
0.01900
0.01840
0.01780
0.01980
0.02030
0.01490
0.01620
0.01780
0.01930
0.02050
0.01180
0.01650
0.01870
0.01200
0.02000
0.01120
0.02130
0.00300
0.01570
0.01640
0.01630
0.00230
0.03200
0.01350
0.01340
0.01770
0.01970
0.01600
0.01590
0.02090
0.01850
0.01440
0.01550
0.01270
0.01530
0.02180
0.01970
0.01900
0.01900
0.01820
0.02020
0.02360
0.02220
-0.1058
-0.1087
-0.1041
-0.1054
-0.1051
-0.1038
-0.1033
-0.1061
-0.1147
-0.1182
-0.1222
-0.1010
-0.1080
-0.1132
-0.1112
-0.1099
-0.1206
-0.1062
-0.1000
-0.1019
-0.1055
-0.1066
-0.1023
-0.1024
-0.1027
-0.1135
-0.1030
-0.1078
-0.1101
-0.1068
-0.1067
-0.1040
-0.1052
-0.1073
-0.1074
-0.1085
-0.1124
-0.1095
-0.1100
-0.1220
-0.1057
-0.1077
-0.1047
-0.1013
-0.1042
0.0399
0.0401
0.0429
0.0422
0.0427
0.0368
0.0367
0.0362
0.0471
0.0300
0.0461
0.0323
0.0286
-0.0082
-0.0032
0.0293
0.0311
0.0387
0.0366
0.0325
0.0388
0.0383
0.0386
0.0321
0.0306
-0.0090
0.0423
0.0400
0.0314
0.0396
0.0390
0.0418
0.0376
0.0384
0.0398
0.0387
-0.0086
-0.0052
-0.0082
0.0452
0.0374
0.0365
0.0356
0.0350
0.0356
0.0620
0.0656
0.0648
0.0648
0.0649
0.0193
0.0185
0.0701
0.0574
0.0550
0.0558
0.0547
0.0444
0.0582
0.0622
0.0477
0.0561
0.0598
0.0149
0.1247
0.0607
0.0604
0.0613
0.1245
0.1246
0.0579
0.0645
0.0622
0.0538
0.0610
0.0620
0.0605
0.0639
0.0616
0.0605
0.0607
0.0572
0.0578
0.0574
0.0541
0.0185
0.0171
0.0198
0.0231
0.0225
-0.0575
-0.0549
-0.0568
-0.0582
-0.0565
-0.0559
-0.0574
-0.0588
-0.0639
-0.0610
-0.0613
-0.0556
-0.0556
-0.0619
-0.0585
-0.0598
-0.0611
-0.0600
-0.0548
-0.0569
-0.0597
-0.0573
-0.0556
-0.0569
-0.0548
-0.0592
-0.0587
-0.0592
-0.0588
-0.0592
-0.0587
-0.0582
-0.0564
-0.0590
-0.0596
-0.0602
-0.0571
-0.0568
-0.0555
-0.0612
-0.0569
-0.0577
-0.0535
-0.0501
-0.0534
CH2CO2Et
CH2CO2tBu
CH2CONH2
Me
Et
Et
Et
CH2CO2Et
Et
Me
Et
Et
CH2CO2Et
Me
Et
Me
Me
Me
Me
Et
Me
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
Me
Ph
Ph
Ph
2-pyridyl
2-thienyl
Ph
Ph
2-furyl
Ph
4-MeOPh
H
OMe
4-FPh
4-ClPh
4-BrPh
OEt
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
36
37
38
39
40
41
42
43
44
45
Cl
2-pyridyl
2-thienyl
H
H
2-furyl
Br
4-MeOPh
2-FPh
Ph
4-FPh
4-ClPh
4-BrPh
Ph
4-ClPh
H
3-MeOPh
4-ClPh
2-ClPh
Ph
4-ClPh
Ph
4-NO2Ph
4-ClPh
4-MePh
Ph
OEt
4-MePh
3-MeOPh
4-ClPh
2-ClPh
4-ClPh
Ph
4-NO2Ph
Ph
4-MePh
4-ClPh
4-HOPh
4-ClPh
4-NO2Ph
3-NO2Ph
4-ClPh
H
H
H
H
H
H
H
H
Cl
Et
Me
Me
Et
Et
Me
4-MePh
4-MeOPh
4-ClPh
4-NO2Ph
3-NO2Ph
a
Only the values of the charges used in eqs 1 and 2 are given.
usually carried out using a logarithmic form of the
biological activity, in our case, we were not able to find
a correlation between the charges and log EC100. How-
ever, since, as is later dealt with, it was possible with
this initial model to predict the EC100 within this
experimental domain, we decided to use it for our
optimization purposes.
From a chemical point of view it is reasonable to
assume that the charges at the atoms of the heterocyclic
system bearing the different substituents and the
contiguous atoms should be more influenced, and so it
is not surprising that the model involves these particu-
lar atoms and not others.
they showed a residual about or higher than 1.5 the
standard deviation (s). However from this initial model,
although not statistically validated, it was possible to
predict those compounds devoid of activity (see the calcd
EC100 values of derivatives 14 and 15 which show an
exptl EC100 > 10, Table 1). Therefore, we decided to use
this eq to identify new derivatives with an increased
activity.
If we now think of structural variations at positions
1, 6, and 7 and include for these last two differently
substituted aryl rests, the number of possible com-
pounds to be synthesized is very large. This is usually
the case in a QSAR study as this, and therefore, it is
necessary to use some criterion in order to somehow
restrict the experimental domain. In principle, we
included as candidates only those compounds which
could be accessible from a chemical point of view, and
also we limited it to compounds bearing methyl or ethyl
In this first approach, we could see (Table 1) that it
was possible to reproduce the pharmacological activity
of the training set compounds (compare exptl EC100 with
calcd EC100 derivatives 1-13), except in compounds 16
and 17. These data points were omitted from eq 1 since

3282 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Campillo et al.
Ta ble 3. Physicochemical Data for Pyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides
compd
R1
R2
R3
mp (°C)
recryst solvents
formula
Mr
anal.a
18
19
20
21
22
23
24
25
26
27
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
4-MeOPh
2-FPh
Ph
4-FPh
4-ClPh
4-BrPh
Ph
4-MeOPh
H
OMe
4-FPh
4-ClPh
4-BrPh
OEt
Me
Me
Me
Me
Et
Me
Me
Et
Me
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
255-6
245-7
208
224-6
294-6
268-70
249-50
212
MeOH/H2O
EOH
EtOH/H2O
EtOH/H2O
EtOH
EtOH/H2O
EtOH/acetone
EtOH/H2O
EtOH/H2O
MeOH
C20H19N5O4S
C12H10FN5O4S
C13H13N5O2S
C18H13F2N5O2S
425.47
307.31
319.34
401.39
448.33
527.23
343.19
381.84
303.34
425.47
411.43
411.43
420.27
420.27
387.36
509.19
385.83
385.83
396.38
396.38
399.50
399.50
367.38
289.32
309.73
320.28
320.28
344.17
289.32
305.31
309.73
320.28
320.28
293.27
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,Cl,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,Cl,N,S
C,H,Cl,N,S
C,H,N,S
C,H,Br,N,S
C,H,Cl,N,S
C,H,Cl,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,Cl,N,S
C,H,N,S
C,H,N,S
C,H,Cl,N,S
C,H,N,S
C,H,N,S
C,H,Cl,N,S
C,H,N,S
C,H,N,S
C
19H15Cl2N5O2S
C18H13Br2N5O2S
14H15N5O3S
C
4-ClPh
H
OEt
C15H16ClN5O3S
C13H13N5O2S
4-MePh
3-MeOPh
4-MeOPh
3-MeOPh
4-ClPh
2-ClPh
4-FPh
4-BrPh
4-ClPh
Ph
282-4
224-6
277-8
251-3
>350
3-MeOPh
4-MeOPh
3-MeOPh
4-ClPh
2-ClPh
4-FPh
4-BrPh
Ph
4-ClPh
Ph
4-NO2Ph
4-ClPh
4-MePh
Ph
C
20H19N5O4S
C19H17N5O4S
19H17N5O4S
EtOH/H2O
MeOH
C
EtOH/H2O
EtOH/H2O
H2O/EtOH
AcOH/EtOH
EtOH/H2O
EtOH/H2O
MeOH/H2O
C17H11Cl2N5O2S
C17H11Cl2N5O2S
C17H11F2N5O2S
310
>350
>350
C
17H11Br2N5O2S
>350
C17H12ClN5O2S
C17H12ClN5O2S
C17H12N6O4S
C17H12N6O4S
>350
4-NO2Ph
Ph
338-40
>350
4-MePh
4-ClPh
4-HOPh
4-MePh
4-ClPh
4-NO2Ph
3-NO2Ph
4-ClPh
H
H
H
H
H
331-3
>350
>350
345
C
18H14ClN5O2S
MeOH/H2O
EtOH/H2O
EtOH/H2O
EtOH/H2O
EtOH/H2O
EtOH/H2O
MeOH/H2O
MeOH/H2O
EtOH/H2O
H2O/EtOH
EtOH/AcOH
H2O/EtOH
EtOH/H2O
C18H14ClN5O2S
C17H13N5O3S
C12H11N5O2S
C11H8ClN5O2S
C11H8N6O4S
C11H8N6O4S
C11H8Cl2N5O2S
C12H11N5O2S
C12H11N5O3S
C11H8ClN5O2S
C11H8N6O4S
C11H8N6O4S
C11H8FN5O2S
H
H
H
H
>350
>350
280-2
>350
Cl
4-MePh
4-MeOPh
4-ClPh
4-NO2Ph
3-NO2Ph
2-FPh
>350
318-20
>350
C,H,Cl,N,S
C,H,N,S
C,H,Cl,N,S
C,H,N,S
>350
328-30
330-2
H
a
Elemental analyses were within (0.4 of the calculated values for the formulas given.
at N-1, since none of the more complicated substituents
at these positions significantly modified the activity as
was shown in the first part of this study.1 For positions
6 and 7 differently substituted aromatic rings (with
activating and deactivating substituents at various
positions), 6-bromo, 6-chloro, 7-alkoxy, etc., were in
principle plausible.
the initial solution of 13 molecules (variance function
7.45 and G ) 5%). Besides, derivatives 18-20 and 27
were also selected since they were predicted to have an
improved activity according to eq 1 (Table 1). The fact
that some compounds have predicted negative activities
deserves comment. Although these values have no
physical meaning, they lie within the range of the
standard deviation. In any case, experiences with nega-
However, the problem was still the very large number
of possible compounds to be synthesized, and in these
cases, experimental design schemes are of great help
in focusing on the most informative experiments; i.e.,
only those derivatives that are more representative of
the experimental domain should be synthesized.11,12
tive values are indicative of compounds with low EC100
,
therefore high activity and thus potential candidates to
be synthesized.
All the selected derivatives were synthesized (see
Chemistry, Table 3), and their platelet aggregation
inhibitory activity was determined. In addition, com-
pounds 21-23 were also evaluated. Table 1 shows their
experimental EC100 values together with those calcu-
lated for synthesized compounds 18-27 and the pre-
dicted values for the rest of the compounds 28-45 using
eq 1.
The proposed candidate molecules 1-45 (45 com-
pounds) are gathered in Table 1. The experimental
domain was designed including only N-1-substituted
derivatives of parent N(1)H-pyrazinothiadiazines syn-
thesized (Table 3). Having calculated the partial atomic
charges, by the ab initio LDF method used before, and
taking as mathematical model eq 1, it was possible to
apply the experimental design methodology12 for select-
ing the compounds. Thus, derivatives 24-26 were
chosen according to the criteria of minimization of the
variance function and the efficiency G.13 The solution
of 16 candidate molecules showed a variance function
of 0.86 and maximum efficiency (36%) in comparison to
As can be seen, experimental EC100 values for deriva-
tives 18-26 are in good agreement with those predicted
by equation 1. The values calculated or predicted from
the final model failed for compounds 16, 17, and 27. The
elimination of compounds 16 and 17 from the final
model (EC100 for 27 > 10) has to be done on a statistical
basis since, although compound 16 is relatively unstable

Arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3283
Ta ble 4. Equation Corresponding to Final Model (Eq 2) and
Statistical Data
and p value. The F value represents the statistical
significance of the regression model and is calculated
as the ratio between regression and residual variances,
and p is the probability that the correlation may be due
to chance (Table 4).
EC100 - â0 + â1X1 + â2X2 + â3X3 + â4X4
(2)
n ) 21 (1-13, 18-25) R2 ) 0.95 Ra ) 0.94a s ) 0.80
2
F4,16 ) 83.88 p < 0.0001
To know the predictive ability of the final model, to
be sure that it is not misinterpreted, and to prevent the
phenomenon of chance correlation in QSAR studies, use
of validation methods is recommended. In this respect,
the first procedure used to check the predictive ability
of the final model was the leave-one-out method.14 The
statistical results summarized in Table 5 together with
Xi
variableb
const
coefficient
SE
t value
p value
-17.272
1835.312
3885.427
1.662 -10.39 <0.0001
188.951
605.719
X1 (N5)2
9.71 <0.0001
6.41 <0.0001
10.23 <0.0001
-8.35 <0.0001
X2 (N5)2C7
X3 (N5)2C4a
X4 (N1)2C6
29092.447 2841.672
-73862.781 8840.759
the value of R2 (0.95) show that the final model has a
cv
a
2
Ra
: adjusted correlation coefficient (correlation coefficient
high predictive ability.
b
adjusted with the degrees of freedom). N5, C7, C4a, N1, C6:
Hirshfeld partial charges calculated from DMol program.
The second method used was the so-called random-
ization of the responses into an array of reordered
variables.15 In this test, the values in the response
column are shuffled randomly while the predictor
columns are kept constant and a new model is estab-
in solution, there is not sufficient evidence to assume a
different chemical behavior in the pharmacological
assays.
Thus, by means of this methodology, it was possible
to predict compounds of this series with a considerably
improved activity, and indeed compounds 18 and 19
once synthesized did show experimental EC100 of 0.001
and 0.03, respectively. In fact, compound 18 represents
a 300-fold improvement of the starting lead compound
1. To refine our initial model these values were intro-
duced providing a final model from 21 compounds
(equation 2) which correctly correlates antiplatelet
activity of this series with the atomic charges of certain
atoms (Table 4).
Sta tistica l An a lysis a n d Va lid a tion of th e Mod el.
The multiple regression eq provides information on the
statistical quality of our model which should be evalu-
ated not only by how well it fits the data but also by
how well it predicts the data.
lished. This process is repeated a number of times, and
2
the estimates of Ra and R2 are calculated. The best
cv
results obtained of 10 runs of the randomization test
are given in Table 6. These values provide an estimate
of the relevance of the real QSAR model. Looking at
2
Table 6 we can see that the Ra and R2 values of the
cv
random models (a-e) are smaller and the PRESS
(predictive residual sum of squares) values increase
dramatically. This result shows that the final model 2
is very reliable for predicting data and does not imply
chance correlation.
Ch em istr y
The N-substituted compounds synthesized for the
QSAR study are gathered in Scheme 1. The substitution
at position 1 can be performed by different procedures
such as alkyl sulfates in water or alkyl halides in
The final model 2 is statistically significant with a
degree of confidence of 95% according to the F statistic
Ta ble 5. Summary of Statistical Results of Leave-One-Out Test
predicted
leave-one-out
compd
R1
R2
R3
EC100 exptl
EC100 calcd (eq 2)
PRESSa
residual
1
2
3
4
5
6
7
8
9
10
11
12
13
18
19
20
21
22
23
24
25
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Me
Br
Ph
Ph
Ph
Ph
Ph
H
Et
CH2Ph
0.3
3.0
0.1
1.0
3.0
0.3
1.0
1.0
1.0
10
1.142
2.795
1.112
1.302
1.438
1.023
0.450
2.090
1.803
10.345
8.961
2.327
3.621
-0.568
-0.262
-0.050
0.650
2.384
0.576
0.070
10.223
1.194
2.774
1.243
1.340
1.299
1.192
0.320
2.150
2.376
10.697
7.976
2.218
3.963
-0.643
-0.363
-0.140
0.680
2.353
0.542
-0.083
11.517
9.463
10.169
9.059
10.113
7.559
9.570
9.841
8.962
9.111
9.975
8.113
9.688
9.618
9.849
10.100
10.179
10.082
9.816
10.021
10.127
9.877
-0.894
0.226
-1.143
-0.340
1.701
-0.892
0.680
-1.150
-1.376
-0.967
2.024
CH2CO2Et
CH2CO2tBu
CH2CONH2
Me
Et
Et
Et
CH2CO2Et
Et
Me
Et
Me
Me
Me
Me
Et
Me
Me
Et
H
Me
Ph
Ph
Ph
2-pyridyl
2-thienyl
4-MeOPh
H
Cl
10
2-pyridyl
2-thienyl
4-MeOPh
2-FPh
Ph
4-FPh
4-ClPh
4-BrPh
Ph
3.0
3.0
0.001
0.03
0.1
0.3
3.0
1.0
0.30
10
0.782
-0.963
-0.106
0.393
0.240
-0.380
0.647
0.458
0.383
-1.517
OMe
4-FPh
4-ClPh
4-BrPh
OEt
4-ClPh
OEt
a
PRESS: predictive sum of squares of final model (eq 2) with the ith compound excluded from the regression.

3284 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Campillo et al.
Sch em e 1
Sch em e 2
Ta ble 6. Summary of Results of the Randomization Test
acetone.1 Compounds 18-27 were obtained from the
corresponding N(1)H-pyrazinothiadiazines with alkyl
halides in acetone since the N(1)H-aryl derivatives are
very insoluble in water.
2 a
c
model
Ra
PRESSb
R2
cv
2
a
b
c
d
e
0.94
0.69
0.55
0.53
0.51
0.43
10.21
37.81
60.28
47.17
138.44
161.28
0.95
0.83
0.73
0.79
0.60
0.54
In Table 3 are shown the novel N-alkyl-substituted
pyrazino[2,3-c][1,2,6]thiadiazines 18-27 and the N(1)H-
derivatives 48-71 synthesized for the QSAR study.
Reaction of 49 and 52 with methyl iodide in the
presence of potassium carbonate afforded compounds 21
and 27, respectively. The 1-methyl (18-20, 23, 24, 26)
and 1-ethyl (22, 25) derivatives were prepared from the
corresponding N(1)H-compounds with methyl iodide and
ethyl iodide, respectively, using triethylamine to avoid
the precipitation of the corresponding potassium salt in
acetone of the arylpyrazinothiadiazine derivative.
The synthesis of N(1)H-pyrazinothiadiazines was
carried out from 4616 with 1,2-dicarbonyl compounds,
R-keto aldehydes, and R-hydroxyimino ketones (Table
3). Reaction of 46 with 4,4′-dimethoxybenzil, 3,3′-
dimethoxybenzil, 4,4′-dichlorobenzil, 2,2′-dichlorobenzil,
4,4′-difluorobenzil, and 4,4′-dibromobenzil (47a -f) af-
forded the corresponding 6,7-bisarylpyrazinothiadi-
azines 48-53 (Scheme 2).
a
2
Ra
: adjusted correlation coefficient (correlation coefficient
b
adjusted with the degrees of freedom). PRESS: predictive re-
sidual sum of squares. c R2cv: cross-validated correlation coefficient
R2 (a R2 value of 1.0 corresponds to perfect prediction).
cv
obtained, and reaction of 46 with 1-(4-chlorophenyl)-2-
phenyl-1,2-ethanedione, 1-(4-nitrophenyl)-2-phenyl-1,2-
ethanedione, and 1-(4-chlorophenyl)-2-(4-methylphenyl)-
1,2-ethanedione (47g-i) afforded a mixture of the
corresponding pyrazinothiadiazines derivatives 54-55,
56-57, and 58-59, respectively. However, when the
reaction was carried out with 1-(4-hydroxyphenyl)-2-
phenyl-1,2-ethanedione (47j) the regioselectivity is high
and only one compound 60 was isolated (Scheme 2).
To obtain 6-unsubstituted arylpyrazinothiadiazines,
reaction of 46 with R-keto aldehydes was carried out.
Thus, compounds 61-64 were prepared from 4-methyl-
phenylglyoxal, 4-chlorophenylglyoxal, 4-nitrophenylgly-
When the dicarbonyl compounds are not symmetric,
the two different isomers in positions 6 and 7 can be

Arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3285
Sch em e 3
Ta ble 7. 13C NMR Data for Compounds 18-27 and 48-71
compd
C-4
C-7
C-8a
C-6
C-4a
other signals
18
19
20
21
22
23
24
25
26
27
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
158.6
158.3
158.4
158.3
158.3
157.9
158.6
158.5
157.9
158.3
158.8
158.5
158.3
158.0
158.4
158.3
158.6
158.5
158.3
158.3
158.3
158.3
158.9
158.4
158.3
158.1
158.1
157.2
158.6
158.8
158.5
158.2
158.3
158.4
154.6
148.2
159.5
154.2
154.1
153.9
159.3
159.2
154.2
155.0
155.3
155.8
154.8
155.5
155.0
154.9
154.2
156.2
156.5
153.9
154.5
155.9
156.3
154.9
153.6
152.6
152.4
155.3
145.7
145.6
145.9
146.8
146.5
148.5
146.7
147.0
147.5
146.9
146.5
147.0
147.8
146.9
148.3
147.0
147.9
147.0
147.1
147.6
147.9
147.1
147.2
147.1
147.1
147.5
146.7
147.0
147.2
148.6
148.6
148.6
148.5
147.5
147.6
147.4
148.2
148.5
148.4
147.6
144.2
140.4
134.4
143.3
143.0
142.9
134.8
135.0
133.1
144.2
145.5
145.4
144.2
144.5
144.5
144.2
145.6
144.4
143.8
145.6
145.4
144.0
145.7
134.8
136.3
135.7
135.2
137.1
144.8
144.8
143.3
142.0
142.3
141.2
120.2
122.6
114.6
120.6
120.9
120.9
114.6
114.8
121.2
120.8
119.0
120.0
120.2
120.7
119.9
120.2
120.3
120.1
120.8
120.1
120.6
119.6
119.3
120.5
121.1
122.2
121.9
120.1
121.0
129.0
121.1
121.5
121.3
121.9
160.6; 159.6; 131.5; 131.1; 129.6; 113.9; 113.7; 55.3; 55.4; 28.1
161.6; 131.9; 131.4; 125.1; 116.3; 124.8; 28.5;
134.1; 129.1; 128.9; 128.1; 55.1; 28.2
165.0; 160.0; 133.6; 133.3; 132.2; 132.0; 131.7; 115.5; 115.1; 114.8; 29.1
136.0; 135.7; 134.8; 133.6; 131.6; 128.5; 128.2; 37.9; 13.9
136.1; 135.9; 131.5; 131.1; 130.9; 123.4; 122.7; 27.9
134.4; 129.4; 129.2; 128.4; 64.5; 14.2; 28.5
133.1; 132.9; 130.7; 128.1; 64.2; 38.0;14.2
141.3; 131.6; 129.4; 127.3; 20.5; 27.5
158.9; 158.8; 138.5; 138.3; 129.3; 129.0; 121.9; 115.2; 114.2; 28.1; 55.0
160.2; 159.3; 131.2; 130.9; 130.0; 113.6; 113.5; 55.2
158.7; 159.0; 138.5; 138.4; 129.0; 129.2; 122.0; 121.9; 115.1; 115.0; 55.0
153.8; 133.7; 133.1; 132.9; 127.2; 127.0; 117.4
135.3; 135.2; 132.5; 132.1; 131.6; 131.3; 130.9; 130.6; 129.3; 129.2; 126.6
164.9; 160.0; 133.6; 132.0; 131; 9; 115.1; 115.0
136.3; 136.2; 131.8; 131.7; 131.2; 131.0; 123.4; 122.2
137.1; 136.1; 134.5; 131.6; 129.8; 129.7; 128.2
137.0; 136.2; 133.5; 131.5; 129.7; 128.7; 128.1
147.7; 136.7; 131.0; 129.8; 128.7; 128.1
147.0; 143.8; 136.5; 129.7; 128.3
129.3; 136.2; 134.0; 133.2; 131.2; 129.4; 128.5; 127.2
138.0; 134.2; 134.0; 131.3; 129.4; 128.6; 128.1
159.1; 138.0; 131.8; 129.8; 128.7; 127.9; 115.3
141.5; 131.7; 129.7; 127.5; 20.9
134.3; 133.3; 129.3; 129.2
148.8; 140.5; 128.9; 124.0; 122.2
148.3; 136.2; 133.7; 130.8; 125.6; 122.1
135.3; 133.9; 133.4; 128.3
139.3; 131.9; 129.4; 126.4; 20.8
160.5; 128.1; 127.3; 114.3; 55.3
134.5; 133.7; 128.8; 128.3
140.0; 147.8; 127.6; 123.9
148.6; 136.6; 132.8; 130.3; 123.9; 121.0
159.8; 131.6; 131.1; 122.8; 124.9; 116.2
oxal, and 3-nitrophenylglyoxal (47k -n ). The 6-chloro-
7-(4-chlorophenyl) derivative 65 was obtained by chlo-
rination of compound 62 with N-chlorosuccinimide in
DMF (Scheme 3).
60, the structure was assigned using an HMQC experi-
ment.
Con clu sion s
Another procedure to prepare compounds with dif-
ferent substituents at 6- and 7-positions involves the
use of R-hydroxyimino ketones as starting materials.
Thus, reaction of 46 with hydroxyiminoacetophenones
47o-t afforded the corresponding 6-arylpyrazinothia-
diazines 66-71 (Scheme 4).
In a series of novel arylpyrazino[2,3-c][1,2,6]thia-
diazines, a statistically significant correlation between
the platelet aggregation inhibition (EC100) and partial
Hirshfeld charges has been established. Although this
is not a conventional QSAR analysis in the sense that
no correlation could be found when using logarithmic
values of the pharmacological activity and that two
outliers had to be removed, it has been possible to use
the model for predictive purposes. Thus, new compounds
were designed and synthesized, and among them,
4-amino-1-methyl-6,7-bis(4-methoxyphenyl)pyrazino-
[2,3-c][1,2,6]thiadiazine 2,2-dioxide (18), when evaluated
as a platelet aggregation inhibitor, showed high activity
The structures of all the new compounds synthesized
have been established on the basis of their analytical
1
and spectroscopic data. The 13C and H NMR chemical
shifts are gathered in Tables 7 and 8. The positions of
the 6- and 7-substituents were established on the basis
of the long-range coupling constants and chemical shifts
of the C-7 and C-6 carbons. In the case of compound

3286 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Ta ble 8. 1H NMR (DMSO-d6) Data for Compounds 18-27 and 48-71
Campillo et al.
compd
NH2
R1, R2
R3
18
19
20
21
22
23
24
25
26
27
48
8.90; 8.77 7.51 (d, 4H, Ar); 6.96 (m, 4H, Ar); 3.80 (s, 3H, Me); 3.79 (s, 3H, Me)
9.11; 8.98 9.11 (s, 1H, H-7); 8.31 (s, 1H, Ar); 7.45-7.35 (m, 3H, Ar)
8.66; 8.56 8.16-7.44 (m, 5H, Ar); 4.11 (s, 3H, Me)
3.47 (s, 3H, Me)
3.44 (s, 3H, Me)
3.45 (s, 3H, Me)
3.45 (s, 3H, Me)
4.15 (q, 2H, CH2); 1.35 (t, 3H, Me)
3.46 (s, 3H, Me)
5.17 (s, 2H, CH2)
4.59 (q, 2H,CH2); 1.45 (t, 3H, Me)
3.52 (s, 3H, Me)
8.95; 8.86 7.17-5.58 (m, 8H, Ar)
8.91; 8.83 7.58-7.41 (m, 8H, Ar)
8.97; 8.88 7.57-7.49 (m, 8H, Ar)
8.70; 8.57 8.11-8.08 (m, 2H, Ar); 7.51-7.25 (m, 8H, Ar); 3.95 (s, 3H, Me)
8.63; 8.57 8.25 (d, 2H, Ar); 7.55 (d, 2H, Ar); 4.05 (q, 2H, CH2); 1.37 (t, 3H, Me)
8.95; 8.61 8.17 (d, 2H, Ar); 7.40 (d, 2H, Ar); 2.42 (s, 3H, Me)
8.87
6.91-7.31 (m, 8H, Ar); 3.68 (s, 3H, Me)
3.64 (s, 3H, Me)
12.55 (br s, 1H, NH)
8.62; 8.32 7.46 (d, 2H, Ar); 7.40 (d, 2H, Ar); 6.94 (d, 2H, Ar) 6.90 (d, 2H, Ar);
3.78 (s, 3H, Me); 3.77 (s, 3H, Me)
49
50
51
52
53
54
55
56
57
58
59
60
8.65; 8.45 7.31-6.87 (m, 8H, Ar); 3.67 (s, 3H, OMe); 3.63 (s, 3H, OMe)
8.71; 8.63 7.54 (d, 4H, Ar); 7.42 (d, 4H, Ar)
8.73; 8.55 7.48-7.26 (m, 8H, Ar)
12.40 (br s, 1H, NH)
12.49 (br s, 1H, NH)
12.62 (br s, 1H, NH)
12.41 (br s, 1H, NH)
12.52 (br s, 1H, NH)
12.41 (br s, 1H, NH)
12.41 (br s, 1H, NH)
12.40 (br s, 1H, NH)
12.40 (br s, 1H, NH)
12.37 (br s, 1H, NH)
12.37 (br s, 1H, NH)
8.69; 8.60 8.56-5.14 (m, 8H, Ar)
8.73; 8.71 7.65-7.36 (m, 8H, Ar)
8.70; 8.61 7.53-7.32 (m, 9H, Ar)
7.58
7.53-7.32 (m, 9H, Ar)
8.78; 8.72 8.16-7.26 (m, 9H, Ar)
8.78; 8.72 8.16-7.26 (m, 9H, Ar)
8.69; 8.55 7.60-7.13 (m, 8H, Ar); 2.31 (s, 3H, Me)
8.66; 8.52 7.60-7.13 (m, 8H, Ar); 2.30 (s, 3H, Me)
8.63; 8.49 9.95 (br s, 1H, OH); 7.52 (m, 2H, Ar); 7.49-7.35 (m, 3H, Ar); 7.27 (d, 2H, Ar); 12.29 (br s, 1H, NH)
6.71 (d, 2H, Ar)
61
62
63
64
8.12
8.63
8.70
8.76
8.96 (s, 1H, H-6); 8.12 (d, 2H, Ar); 7.40 (d, 2H, Ar); 2.40 (s, 3H, Me)
9.01 (s, 1H, H-6); 8.24 (d, 2H, Ar); 7.67 (d, 2H, Ar)
9.10 (s, 1H, H-6); 8.49-8.44 (m, 5H, Ar)
9.18 (s, 1H, H-6); 9.00 (s, 1H, Ar); 8.71 (d, 1H, Ar); 8.46 (dd, 1H, Ar);
7.92 (t, 1H, Ar)
12.21 (br s, 1H, NH)
12.31 (br s, 1H, NH)
12.46 (br s, 1H, NH)
12.52 (br s, 1H, NH)
65
66
67
68
69
70
8.74. 8.65 7.80 (d, 2H, Ar); 7.63 (d, 2H, Ar)
12.63 (br s, 1H, NH)
12.28 (br s, 1H, NH)
12.45 (br s, 1H, NH)
12.40 (br s, 1H, NH)
8.74; 8.71 9.21 (s, 1H, H-7); 8.21 (d, 2H, Ar); 7.31 (d, 2H, Ar); 2.36 (s, 3H, Me)
8.44; 8.68 9.19 (s, 1H, H-7); 8.27 (d, 2H, Ar); 7.05 (d, 2H, Ar); 3.76 (s, 3H, OMe)
8.81; 8.71 9.26 (s, 1H, H-7); 8.36 (d, 2H, Ar); 7.57 (d, 2H, Ar)
8.93; 8.80 9.38 (s, 1H, H-7); 8.80 (d, 2H, Ar); 8.62 (d, 2H, Ar)
8.98; 8.75 9.39 (s, 1H, H-7); 9.08 (m, 1H, Ar); 8.72 (dd, 2H, Ar); 7.30 (dd, 1H, Ar);
7.81 (t, 1H, Ar)
71
8.58; 8.42 8.99 (d, 1H, H-7); 8.29 (dt, 1H, Ar); 7.56-7.52 (m, 1H, Ar); 7.43 (m, 2H, Ar)
12.44 (br s, 1H, NH)
trometer and are reported in ppm on the δ scale. The signal
of the solvent was used as reference. Mass spectra (electron
impact, 70 eV) were obtained on a VG 12-250 (VG Masslab).
Elemental analyses were performed on a Heraeus CHN-O-
Rapid analyzer. Column chromatography was carried out on
silica gel (Merck, particle size 70-230 mesh).
Sch em e 4
The R-keto aldehydes were prepared from the procedures
reported by Cava17 (47k ,i) and Steinbach18 (47m ,n ). In the case
of R-hydroxyimino ketones the compounds 47o-t were syn-
thesized following the procedure of Slater.19
Th eor etica l Ca lcu la tion s. The studied compounds were
built with standard bond lengths and angles by using the
molecular modeling package Insight II.20 All the structures
were fully optimized without any symmetry restrictions in the
gas phase. The LDF calculations were carried out using the
DMol program8 distributed by MSI. A double-ú numerical basis
set with polarization functions in all the atoms and the J anak-
Moruzzi-Williams (J MW) exchange correlation potential21
were used. The geometry of the molecules was optimized until
the gradient was smaller than 0.001 au. Then, the partial
Hirshfeld charges were calculated.7
Sta tistica l An a lysis. Multiple regression analysis were
carried out with the StatView, SPSS, and NEMROD pro-
grams.13 The collinearity of the variables in the experimental
domain is not significant with a correlation coefficient between
-0.04 and 0.40 which indicates that our descriptors are not
intercorrelated. The experimental design methodology was
performed using an exchange algorithm included in the
DOPTDC program.13
(EC100 ) 0.001 µg/mL). This value represents a 300-
fold improvement over the starting lead compound and
a 2500-fold improvement over aspirin in the same
experimental conditions. On the basis of these results,
further research in order to study the mechanism of
action and to evaluate potential use as antiplatelet
agents of this series is in progress.
Exp er im en ta l Section
In Vitr o P la telet Aggr ega tion Stu d ies. The arylpyrazi-
nothiadiazine derivatives were screened for their platelet
aggregation inhibitory activity in vitro on citrated rabbit
platelet-rich plasma in which aggregation was induced with
arachidonic acid.1,4
Ch em ist r y. Melting points were determined with
a
Reichert-J ung Thermovar micromelting point apparatus and
are uncorrected. 1H NMR spectra (300 MHz) and 13C NMR
spectra (75 MHz) were recorded on a Varian XL-300 spec-

Arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17 3287
Gen er a l P r oced u r e for th e P r ep a r a tion of N(1)-Alk yl-
p yr a zin o[2,3-c][1,2,6]th ia d ia zin e 2,2-Dioxid es 18-27. To
the corresponding 4-amino-1H-pyrazino[2,3-c][1,2,6]thiadiazine
2,2-dioxide derivative in acetone, and either potassium carbon-
ate or triethylamine, was added the alkyl halide. The reaction
mixture was refluxed and evaporated to dryness, and water
was added to the residue. The precipitate was filtered and
recrystallized from the appropriate solvent.
4-Am in o-6,7-bis(4-m eth oxyp h en yl)-1-m eth ylp yr a zin o-
[2,3-c][1,2,6]th ia d ia zin e 2,2-d ioxid e (18): from 48 (2.90 g,
4.02 mmol), triethylamine (0.5 mL, 4.0 mmol), methyl iodide
(0.8 mL, 12.1 mmol), and acetone (200 mL); reaction time 72
h; yield (2.10 g, 73%).
reaction time 24 h; yield (1.60 g, 61%); previously reported from
compound 46 and 4-methoxybenzaldehyde.24
4-Am in o-6,7-bis(3-m eth oxyp h en yl)-1H-p yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (49): from 46 (2.00 g, 11.3
mmol), methanol (60 mL), concentrated hydrochloric acid (1.0
mL), and 3,3′-dimethoxybenzil (47b) (3.50 g, 12.3 mmol);
reaction time 72 h; yield (1.91 g, 41%).
4-Am in o-6,7-b is(4-ch lor op h en yl)-1H -p yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (50): from 46 (2.50 g, 14.1
mmol), ethanol (120 mL), concentrated hydrochloric acid (1.0
mL), and 4,4′-dichlorobenzil (47c) (4.30 g, 15.5 mmol); reaction
time 72 h; yield (1.85 g, 40%); previously reported from
compound 46 and 4-chlorobenzaldehyde.24
4-Am in o-6-(2-flu or op h en yl)-1-m et h ylp yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (19): from 71 (0.90 g, 3.1
mmol), triethylamine (0.4 mL, 3.1 mmol), methyl iodide (0.6
mL, 9.3 mmol), and acetone (80 mL); reaction time 72 h; yield
(0.47 g, 54%).
4-Am in o-6,7-b is(2-ch lor op h en yl)-1H -p yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (51): from 46 (2.50 g, 14.1
mmol), ethanol (120 mL), concentrated hydrochloric acid (1.0
mL), and 2,2′-dichlorobenzil (47d ) (4.30 g, 15.5 mmol); reaction
time 72 h; yield (1.85 g, 40%).
4-Am in o-7-m eth oxy-1-m eth yl-6-p h en ylp yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (20): from 4-amino-7-meth-
oxy-6-phenylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-dioxide22 (1.70
g, 5.7 mmol), triethylamine (0.8 mL, 5.7 mmol), and methyl
iodide (0.45 mL, 6.8 mmol); reaction time 24 h; yield (1.30 g,
74%).
4-Am in o-6,7-bis(4-flu or op h en yl)-1-m eth ylp yr a zin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-d ioxid e (21): from 52 (1.50 g, 3.8
mmol), potassium carbonate (0.30 g, 1.9 mmol), methyl iodide
(1.2 mL, 19.8 mmol), and acetone (150 mL); reaction time 24
h; yield (1.14 g, 73%).
4-Am in o-6,7-bis(4-ch lor op h en yl)-1-eth ylp yr a zin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-d ioxid e (22): From 50 (1.10 g, 6.7
mmol), triethylamine (0.3 mL, 2.6 mmol), ethyl iodide (0.4 mL,
5.4 mmol), and acetone (50 mL); reaction time 200 h; yield
(0.69 g, 62%).
4-Am in o-6,7-bis(4-br om oph en yl)-1-m eth ylpyr azin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-d ioxid e (23): from 53 (1.30 g, 2.5
mmol), triethylamine (0.32 mL, 2.5 mmol), methyl iodide
(excess), and acetone (60 mL); reaction time 72 h; yield (0.80
g, 60%).
4-Am in o-7-et h oxy-1-m et h yl-6-p h en ylp yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (24): from 4-amino-7-ethoxy-
6-phenyl-1H-pyrazino[2,3-c][1,2,6]thiadiazine 2,2-dioxide23 (1.50
g, 4.7 mmol), triethylamine (0.7 mL, 4.7 mmol), methyl iodide
(0.6 mL, 9.4 mmol), and acetone (100 mL); reaction time 90 h;
yield (1.11 g, 69%).
4-Am in o-6-(4-ch lor op h en yl)-7-eth oxy-1-eth ylp yr a zin o-
[2,3-c][1,2,6]t h ia d ia zin e 2,2-d ioxid e (25): from 4-amino-
6-(4-chlorophenyl)-7-ethoxy-1H-pyrazino[2,3-c][1,2,6]thiadi-
azine 2,2-dioxide23 (3.60 g, 4.2 mmol), triethylamine (0.6 mL,
4.2 mmol), ethyl iodide (0.6 mL, 9.4 mmol), and acetone (300
mL); reaction time 90 h; yield (2.12 g, 60%).
4-Am in o-1-m e t h yl-7-(4-t olyl)p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (26): from 61 (1.60 g, 5.5 mmol),
triethylamine (0.7 mL, 5.5 mmol), methyl iodide (0.5 mL, 8.3
mmol), and acetone (120 mL); reaction time 120 h; yield (0.77
g, 57%).
4-Am in o-6,7-bis(3-m eth oxyp h en yl)-1-m eth ylp yr a zin o-
[2,3-c][1,2,6]th ia d ia zin e 2,2-d ioxid e (27): from 49 (1.50 g,
3.6 mmol), potassium carbonate (0.25 g, 1.8 mmol), methyl
iodide (5.0 mL, 70 mmol), and acetone (150 mL); reaction time
6 h; yield (1.04 g, 66%).
4-Am in o-6,7-b is(4-flu or op h en yl)-1H -p yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (52): from 46 (1.50 g, 6.2
mmol), methanol (50 mL), concentrated hydrochloric acid (0.6
mL), and 4,4′-difluorobenzil (47e) (1.50 g, 6.2 mmol); reaction
time 72 h; yield (2.0 g, 36%).
4-Am in o-6,7-b is(4-b r om op h en yl)-1H -p yr a zin o[2,3-c]-
[1,2,6]t h ia d ia zin e 2,2-Dioxid e (53). To a solution of 4,4′-
dibromobenzil (47f) (7.40 g, 20.1 mmol) in dimethyl sulfoxide
(200 mL) at 80 °C was added compound 46 (3.00 g, 16.8 mmol)
and the mixture stirred for 24 h at 80 °C. Then, the solvent
was evaporated to dryness, and water was added to the
residue. The precipitate was filtered and recrystallized from
acetic acid/ethanol to yield 53 (2.30 g, 25%).
4-Am in o-7-(4-ch lor op h en yl)-6-p h en yl-1H-p yr a zin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-d ioxid e (54) a n d 4-a m in o-6-(4-
c h lo r o p h e n y l)-7-p h e n y l-1H -p y r a z i n o [2,3-c ][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (55): from 46 (1.00 g, 5.6 mmol), 1-(4-
chlorophenyl)-2-phenyl-1,2-ethanedione (47g) (1.50 g, 6.2 mmol),
methanol (40 mL), and concentrated hydrochloric acid (0.6
mL); reaction time 24 h; yield (0.90 g, 43%); separated by
chromatography on silica gel using CHCl3/AcOH (100/3) as
eluent.
4-Am in o-7-(4-n itr op h en yl)-6-p h en yl-1H-p yr a zin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-d ioxid e (56) a n d 4-a m in o-6-(4-
n i t r o p h e n y l )-7 -p h e n y l -1H -p y r a z i n o [2 ,3 -c ][1 ,2 ,6 ]-
th ia d ia zin e 2,2-d ioxid e (57): from 46 (1.00 g, 5.6 mmol), 1-(4-
nitrophenyl)-2-phenyl-1,2-ethanedione (47h ) (1.50 g, 6.2 mmol),
methanol (40 mL), and concentrated hydrochloric acid (0.6
mL); reaction time 48 h; yield (0.89 g, 41%); separated by
chromatography on silica gel using CH2Cl2/MeOH (40/1) as
eluent.
4-Am in o-7-(4-ch lor op h en yl)-6-(4-t olyl)-1H -p yr a zin o-
[2,3-c][1,2,6]th ia d ia zin e 2,2-d ioxid e (58) a n d 4-a m in o-6-
(4-ch lor op h e n yl)-7-(4-t olyl)-1H -p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (59): from 46 (1.00 g, 5.6 mmol), 1-(4-
chlorophenyl)-2-(4-methylphenyl)-1,2-ethanedione (47i) (1.60
g, 6.2 mmol), methanol (80 mL), and concentrated hydrochloric
acid (0.6 mL); reaction time 72 h; yield (1.30 g, 57%); separated
by chromatography on silica gel using CH2Cl2/MeOH (40/1)
as eluent.
4-Am in o-7-(4-h yd r oxyp h en yl)-6-p h en yl-1H -p yr a zin o-
[2,3-c][1,2,6]th ia d ia zin e 2,2-d ioxid e (60): from 46 (0.68 g,
3.8 mmol), 1-(4-hydroxyphenyl)-2-phenyl-1,2-ethanedione (47j)
(1.00 g, 4.2 mmol), methanol (30 mL), and concentrated
hydrochloric acid (0.15 mL); reaction time 48 h; yield (1.40 g,
68%); EM m/e 367 (M+).
4 -Am i n o -7 -( 4 -t o l y l ) -1 H -p y r a z i n o [ 2 ,3 -c ] [ 1 ,2 ,6 ] -
th ia d ia zin e 2,2-d ioxid e (61): from 46 (8.30 g, 46.5 mmol),
4-methylglyoxal (47k ) (9.90 g, 67.0 mmol), and acetic acid (170
mL); reaction time 48 h; yield (7.30 g, 55%).
4-Am in o-7-(4-ch lor op h en yl)-1H-p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (62): from 46 (2.60 g, 14.5 mmol),
4-chlorophenylglyoxal (47l) (3.30 g, 18.2 mmol), and acetic acid
(50 mL); reaction time 48 h yield (1.18 g, 46%).
Gen er a l P r oced u r e for th e Syn th esis of Com p ou n d s
48-52 a n d 54-64 fr om Dica r bon yl Com p ou n d s. To a
suspension of 46 (1.0 mmol) in methanol or ethanol and either
concentrated hydrochloric acid or acetic acid was added the
corresponding carbonyl compound (1.1 mmol), and the mixture
refluxed. The reaction mixture was evaporated to dryness, and
water was added to the residue. The precipitate was filtered,
washed with dichloromethane, and recrystallized from the
appropriate solvent.
4-Am in o-6,7-bis(4-m eth oxyp h en yl)-1H-p yr a zin o[2,3-c]-
[1,2,6]th ia d ia zin e 2,2-d ioxid e (48): from 46 (4.0 g, 22.6
mmol), methanol (300 mL), concentrated hydrochloric acid (1
mL), and 4,4′-dimethoxybenzil (47a ) (7.1 g, 25.0 mmol);
4-Am in o-7-(4-n it r op h en yl)-1H -p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (63): from 46 (0.70 g, 4.4 mmol),

3288 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 17
Campillo et al.
(2) Hansch, C.; Leo, A. Exploring QSAR: Fundamentals and
Applications in Chemistry and Biology; ACS Professional Refer-
ence Book; Editorial Advisory Board: Washington, DC, 1995.
(3) Fujita, T. The Extrathermodynamic Approach to Drug Design.
In Comprehensive Medicinal Chemistry. Quantitative Drug
Design; Hansch, C., Ed.; Pergamon Press: Oxford, 1990; pp 497-
560.
(4) Bertele, V.; Falanga, A.; Tomasiak, M.; Dejana, E.; Carletti, C.;
De Gaetano, G. Platelet Thromboxane Synthetase Inhibitors
with Low Doses of Aspirin: Possible Resolution of the “Aspirin
Dilemma”. Science 1983, 220, 517-519.
(5) Becke, A. Density-Functional Exchange-Energy Approximation
with Correct Asymptotic Behavior. Phys. Rev. A 1988, 38, 3098-
3100.
(6) Delley, B. An All-Electron Numerical Method for Solving the
Local Density Functional for Poliatomic Molecules. J . Chem.
Phys. 1990, 92, 508-517.
(7) Hirshfeld, F. L. Bonded Atom-Fragments for Describing Molec-
ular Charges Densities. Theor. Chim. Acta 1977, 44, 129.
(8) DMol, version 95.0, 3.0.0; MSI, San Diego, CA.
(9) Alkorta, I.; Garc´ıa-Go´mez, C.; Pa´ez, J . A.; Goya, P. Theoretical
and Experimental Analysis of Properties in Heterocycles Con-
taining the Aminosulphonylamino Moiety. J . Phys. Org. Chem.
1996, 9, 203-211.
(10) Alkorta, I. Comparison of Methods to Estimate Geometric and
Electronic Properties on Sulfur Containing Compounds. Theor.
Chim. Acta 1994, 89, 1-12.
(11) Pleiss, M. A.; Unger, S. H. The Design of Test Series and the
Significance of QSAR Relationships. In Comprehensive Medicinal
Chemistry. Quantitative Drug Design; Hansch, C., Ed.; Perga-
mon Press: Oxford, 1990; pp 561-587.
(12) Atkinson, A. G.; Donev, A. N. Optimum experimental designs;
Oxford Science Publications, Clarendon Press: Oxford, 1992; p
116.
4-nitrophenylglyoxal (47m ) (0.90 g, 5.2 mmol), and acetic acid
(40 mL); reaction time 72 h; yield (0.33 g, 23%).
4-Am in o-7-(3-n it r op h en yl)-1H -p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (64): from 46 (5.00 g, 28.0 mmol),
3-nitrophenylglyoxal (47n ) (6.00 g, 33.5 mmol), and acetic acid
(100 mL); reaction time 48 h; yield (3.05 g, 37%).
Gen er a l P r oced u r e for th e Syn th esis of Com p ou n d s
66-71 fr om a -Hyd r oxyim in o Keton es. To a suspension of
46 (1.0 mmol) in methanol and water was added the corre-
sponding R-hydroxyimino ketone (1.2 mmol), and the mixture
was stirred. The reaction mixture was evaporated to dryness,
and water was added to the residue. The precipitate was
filtered, washed with dichloromethane, and recrystallized from
the appropriate solvent.
4-Am in o-6-(4-t olyl)-1H -p yr a zin o[2,3-c][1,2,6]t h ia d i-
a zin e 2,2-d ioxid e (66): from 46 (9.30 g, 52.5 mmol), 4-meth-
yl-2-(hydroxyimino)acetophenone (47o) (12.00 g, 78.0 mmol),
methanol (60 mL), and water (60 mL); reaction time 48 h; yield
(5.92 g, 48%).
4-Am in o-6-(4-m eth oxyph en yl)-1H-pyr azin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (67): from 46 (4.20 g, 23.6 mmol),
4-methoxy-2-(hydroxyimino)acetophenone (47p ) (5.70 g, 34.3
mmol), methanol (90 mL), and water (90 mL); reaction time
120 h; yield (3.40 g, 47%).
4-Am in o-6-(4-ch lor op h en yl)-1H-p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (68): from 46 (8.30 g, 47.0 mmol),
4-chloro-2-(hydroxyimino)acetophenone (47q) (11.30 g, 61.5
mmol), methanol (70 mL), and water (70 mL); reaction time
120 h; yield (6.96 g, 46%).
4-Am in o-6-(4-n it r op h en yl)-1H -p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (69): from 46 (12.00 g, 0.07 mol),
4-nitro-2-(hydroxyimino)acetophenone (47r ) (14.70 g, 0.09
mol), methanol (100 mL), and water (100 mL); reaction time
48 h; yield (5.42 g, 25%).
(13) Mathieu, D.; Phan Tan Luu, R. NEMROD (New Efficient
Methodology of Research Using Optimal Design). I.U.T., Uni-
versite´ d’Aix-Marseille, 1995.
(14) Cramer, R. D.; Patterson, D. E.; Bunce, J . D. Comparative
molecular field analysis (CoMFa). Effect of shape of binding of
steroids to carrier proteins. J . Am. Chem. Soc. 1988, 110, 5959-
5967.
(15) Topliss, J . G.; Edwards, R. P. Chance Factors in Studies of
Quantitative Structure-Activity Relationships. J . Med. Chem.
1979, 22, 1238-1244.
(16) Garc´ıa-Mun˜oz, G.; Madron˜ero, R.; Ochoa, C.; Stud, M. Synthesis
of Purine-Like Ring System Derived From 1,2,6-Thiadiazine 1,1-
Dioxide. J . Heterocycl. Chem. 1976, 13, 793-796.
(17) Cava, M. P.; Litle, R. L.; Napier, D. R. Condensed Cyclobutane
Aromatic Systems. V. The Synthesis of Some R-Diazoinda-
zones: Ring Contraction in the Indone Series. J . Am. Chem. Soc.
1958, 80, 2257-2263.
(18) Steinbach, L.; Becker, E. I. A Synthesis for â-Aroylacrylic Acids
Substitute with Electron-withdrawing Groups. J . Am. Chem.
Soc. 1954, 76, 5808-5810.
4-Am in o-6-(3-n it r op h en yl)-1H -p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (70): from 46 (3.00 g, 16.8 mmol),
3-nitro-2-(hydroxyimino)acetophenone (47s) (3.70 g, 18.9 mmol),
methanol (60 mL), and water (60 mL); reaction time 48 h; yield
(0.98 g, 30%).
4-Am in o-6-(2-flu or op h en yl)-1H-p yr a zin o[2,3-c][1,2,6]-
th ia d ia zin e 2,2-d ioxid e (71): from 46 (1.00 g, 5.6 mmol),
2-fluoro-2-(hydroxyimino)acetophenone (47t) (1.00 g, 7.0 mmol),
methanol (30 mL), and water (30 mL); reaction time 48 h; yield
(1.06 g, 64%).
4-Am in o-6-ch lor o-7-(4-ch lor op h en yl)-1H-p yr a zin o[2,3-
c][1,2,6]th ia d ia zin e 2,2-Dioxid e (65). To a solution of 62
(1.50 g, 5.0 mmol) in DMF (30 mL) was added N-chlorosuc-
cinimide (1.76 g, 10.1 mmol). The reaction mixture was stirred
to room temperature for 24 h, and then the solvent was
evaporated to dryness and water was added to the residue.
The precipitate was filtered and recrystallized from MeOH/
H2O to give 65 (0.80 g, 47%): EM m/e 343 (M+).
(19) Slater, W. K. Experiments on the Preparation of Oximinoderiva-
tives. J . Chem. Soc. 1920, 117, 587-591.
(20) Insight II, 95.0/3.0.0 version; Biosym/MSI, San Diego, CA, 1995.
(21) Moruzzi, V. L.; J anak, J . F.; Williams, A. R. Calculated Electronic
Properties of Metals; Pergamon Press: New York, 1978.
(22) Alkorta, I.; Goya, P.; Pa´ez, J . A.; Pfleiderer, W. Synthesis and
Physico-Chemical Properties of 6- and 7-Monosubstituted Pyrazi-
no[2,3-c][1,2,6]thiadiazine 2,2-Dioxides. Pteridines 1990, 2, 3-7.
(23) Campillo, N.; Pa´ez, J . A.; Goya, P. Unpublished results.
(24) Herrero, A.; Ochoa, C. Reaction Between o-Diamino-1,2,6-
Thiadiazine 1,1-Dioxides and Aldehydes to Give Fused Imidazo
and Pyrazino Derivatives. J . Heterocycl. Chem. 1988, 25, 891-
893.
Ack n ow led gm en t. The Sociedad Espan˜ola de Qu´ı-
mica Terapeu´tica awarded the Almirall Prize, 1997, to
part of this work presented by Nuria Campillo. The
authors are grateful to E. Carrasco and M. Grau for
their collaboration.
Refer en ces
(1) Campillo, N.; Garc´ıa, C.; Goya, P.; Pa´ez, J . A.; Carrasco, E.; Grau,
M. Novel Arylpyrazino[2,3-c][1,2,6]thiadiazine 2,2-Dioxides as
Inhibitors of Platelet Aggregation. 1. Synthesis and Pharmaco-
logical Evaluation. J . Med. Chem. 1999, 42, 1698-1704.
J M981104B