Synthesis and diuretic activity of bicyclic fused heterocycles containing oxime-O-sulfonic acid moiety

Article abstract of DOI:10.1016/S0223-5234(00)00122-7

In order to investigate the origin of the loop-type diuretic activity of M17055 (1), several variants (3-9) were designed and synthesized by modifying the quinolinone skeleton, and their diuretic activities were compared with the lead 1 and furosemide in dogs. It was found that the negative charge distribution pattern afforded by the dispositional arrangement of the 4- oxime-O-sulfonic acid and 1-N-acyl carbonyl moiety attached to the tetrahydropyridine ring system is inevitable for the development of the activity, which strongly supports the previously proposed model for the active site of the Na⁺-K⁺-2Cl^- cotransporter. Also reported is the first synthesis of the dihydrothieno[3,2-b]pyridine-7(4H)-one ring system required in the synthesis of compound 9. (C) 2000 Editions scientifiques et medicales Elsevier SAS.

Full text of DOI:10.1016/S0223-5234(00)00122-7

Eur. J. Med. Chem. 35 (2000) 227−240
© 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved
S0223523400001227/FLA
227
Original article
Synthesis and diuretic activity of bicyclic fused heterocycles containing
oxime-O-sulfonic acid moiety
Kazumi Nishijimaa*, Hidemitsu Nishidaa, Yoshiaki Yamashitaa, Manabu Itoa, Yoshiaki Onukia,
Masahiro Mizotaa, Sotaro Miyanob
aFuji Central Research Laboratory, Mochida Pharmaceutical Co. Ltd., 722 Jinba-aza-Uenohara, Gotemba, Shizuoka 412-8524, Japan
bDepartment of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, Aramaki-Aoba 07, Aoba-ku,
Sendai 980-8579, Japan
Received 6 November 1998; revised 29 July 1999; accepted 5 August 1999
Abstract – In order to investigate the origin of the loop-type diuretic activity of M17055 (1), several variants (3–9) were designed and
synthesized by modifying the quinolinone skeleton, and their diuretic activities were compared with the lead 1 and furosemide in dogs. It was
found that the negative charge distribution pattern afforded by the dispositional arrangement of the 4-oxime-O-sulfonic acid and 1-N-acyl
carbonyl moiety attached to the tetrahydropyridine ring system is inevitable for the development of the activity, which strongly supports the
previously proposed model for the active site of the Na+-K+-2Cl– cotransporter. Also reported is the first synthesis of the dihydrothieno[3,2-
b]pyridine-7(4H)-one ring system required in the synthesis of compound 9. © 2000 Éditions scientifiques et médicales Elsevier SAS
M17055 / oxime-O-sulfonic acid / diuretic activity / structure–activity relationship / computational chemistry
1. Introduction
sulfonyl moiety (-SO3-) of compound 1 and the carboxyl
moiety (-CO2-) of furosemide, while relatively small and
weak ones reside on the acyl carbonyl (>N–CO-) group
of the former and the sulfamoyl (-SO2NH2) group of the
latter. Thus, it was suggested that the charge distribution,
rather than the entity of the functional groups themselves,
is essential for the development of the loop-diuresis. The
structure–activity relationships derived from pharmaco-
logical, as well as computational, studies led us to
propose a model for the active site of the Na+-K+-2Cl–
cotransporter in which the -SO3 function and the acyl
carbonyl function are of particular importance in binding
to the Cl– binding site of the cotransporter and serving as
a receptor for a hydrogen bonding, respectively, with the
aid of the two lipophilic phenyl moieties of the quinoli-
none and the acyl residue assisting the interaction by
fitting to the hydrophobic sites of the cotransporter.
7-Chloro-2,3-dihydro-1-(2-methylbenzoyl)-4-(1H)-qui-
nolinone 4-oxime-O-sulfonic acid potassium salt 1
(M17055) is a novel diuretic of high potency, which is
now under the last stage of phase III trials by Mochida
Pharmaceutical Co. Ltd., Japan [1] (figure 1). Although
the chemical structure of the oxime sulfonate 1 is quite
different from those of the conventional diuretics, its
dominant site of action has been shown to be the loop of
the Henle like loop-type diuretics represented by furo-
semide, as evidenced by a wide range of pharmacological
studies [2].
In the course of our efforts to clarify the origin of the
diuretic activity induced by compound 1, we recently
carried out several computational analyses on these
diuretic materials to show that compound 1 and furo-
semide are closely related to each other from the view
point of electron charge distributions: both compounds
show quite similar electrostatic potential maps, in which
large and strong negative charges are located on the
In the next step, we designed tricyclic compound 2,
which, due to the restriction of the rotational freedom
around the original >N–CO- bond by converting to a
cyclic >N-C=N- linkage, while retaining the molecular
features of the lead 1, should fit more comfortably to the
*Correspondence and reprints: kazumi@mochida.co.jp

228
Figure 1. Synthetic plans for modification of the quinoline skeleton.
Figure 2. Synthetic pathway to 3.

229
Figure 3. Synthetic pathway to 4 and 5.
supposed active site, thereby inducing a larger diuretic
potency than the parent [3]. To our satisfaction, many of
the newly synthesized tricycles 2 actually showed a
comparable or superior activity to that of the lead 1,
among which compound 2a was the most powerful and
has been adopted as the promising candidate to succeed 1
(M17055). Pharmacological and clinical studies on it are
now underway.
2. Chemistry
Figure 2 illustrates the synthetic pathway to compound
3. By referring to a paper describing the synthesis of
1,4-dihydro-3(2H)-quinolinone derivatives [4], 7-chloro-
1,4-dihydro-3(2H)-quinolinone dimethyl acetal 14 was
synthesized from a commercially available 4-chloro-2-
nitrotoluene 10 via the Claisen-type condensation with
dimethyl oxalate to give compound 11 followed by
acetallization to compound 12, reductive cyclization to
quinolinone 13 and finally, reduction of the carbonyl
function with Selectlide. Compound 14 was N-acylated
by 2-methylbenzoyl chloride/pyridine followed by acid-
promoted deacetallization to compound 15, which was
then allowed to react successively with hydroxylamine-
O-sulfonic acid and then with potassium carbonate to
give the desired sulfonic acid salt 3. Hereafter, the
formation of pertinent sulfonate salts by treatment with
(1) 2-methylbenzoyl chloride/pyridine, (2) H2NOSO3H,
and (3) K2CO3 will be denoted as method A.
With the above results in mind, we were tempted to
further substantiate the active site model by demonstrat-
ing that the slight modification of the distribution of the
electron charge density of compound 1 would reduce or
eliminate the diuretic activity. Thus, we synthesized
compounds 3–9, where compounds 3–5 are different
from each other in respect to the relative positions of the
sulfonic acid and the acyl carbonyl function, compounds
6 and 7 are deformed in the three-dimensional structures
from that of compound 1 by ring-enlargement (6) or by
cleavage of the cyclic linkage (7), and compound 8 and 9
have either a pyridine (8) or thiophene ring (9) in place of
the benzene ring of the quinolinone skeleton of com-
pound 1. Then, the diuretic activities of these newly
synthesized compounds were tested in dogs and the
structure–activity relationships discussed with the aid of
computer graphics.
Target compounds 4 and 5 were obtained from the key
intermediates 16 [5, 6] and 18 [7], respectively, which
had been prepared according to procedures mentioned in
the literature (figure 3). Compound 16 was readily acy-
lated to compound 17 and then converted to the

230
Figure 4. Synthetic pathway to 6 and 7.
O-sulfonic acid salt 4 by method A. On the other hand,
compound 18 failed to react with 2-methylbenzoyl
chloride/pyridine due to the reduced reactivity of the
amide. So, compound 18 was first converted to the
thiocarbonyl derivative 19, which was found to undergo
a reaction with the acyl chloride to give compound 20.
Compound 20 was again not active enough to react with
hydroxylamine-O-sulfonic acid, probably due to elec-
tronic as well as, at least in part, to steric reasons.
Luckily, however, it was found that compound 20 reacted
regioselectively with hydroxylamine hydrochloride on
the thiocarbonyl rather than the acyl carbonyl function of
the imide moiety to give compound 21. This was then
converted to sulfonic acid by treatment with sulfur
trioxide/pyridine followed by cation exchange to give the
desired product 5. Hereafter, the formation of the perti-
nent sulfonate salts by treatment with (1) H2NOH HCl/
pyridine, (2) sulfur trioxide/pyridine and (3) K2CO3 will
be denoted as method B.
The routes for preparing compounds 6 and 7 are shown
in figure 4. Referring to a paper describing the synthesis
of 1,2,3,4-tetrahydro-1-benzazepin-5(5H)-one [8], the

231
Figure 5. Synthetic pathway to 8 and 9.
8-chloro counterpart 27 was formed from commercially
available compound 22. Compound 27 was then con-
verted to the target compound 6 by method A. A known
compound 29 [9] was acylated to compound 30, which
was then converted to the target compound 7 via an
oxime 31 according to method B.
Figure 5 shows the synthesis of pyridine and thiophene
derivatives 8 and 9. 7-Chloro-2,3-dihydro-1,8-naphthy-
ridine-4(1H)-one 32 was synthesized referring to a pro-
cedure found in the literature [10], and then it was
converted to the desired compound 8 via N-acyl deriva-
tive 33 by using essentially method A, except that lithium
bis(trimethylsilyl)amide was used instead of pyridine as
the base in the acylation step.
The heteroaromatic ring system of compound 9,
dihydrothieno[3,2-b]pyridine-7(4H)-one nucleus, had
seemed, to the best of our knowledge, to be unprec-
edented, although the [3,2-c] counterpart had been
known. Several trials for the synthesis of compound 9
starting from 3-aminothiophene were unfruitful because
of its incompatibility with the Michael reaction condi-
tions. To our satisfaction, however, we eventually suc-
ceeded in the first synthesis of compound 37, which
should be the key intermediate for the syntheses of the
[3,2-b]-ring class compounds, by finding that introduc-
tion of a 2-methoxycarbonyl substituent to 3-amino-
thiophene to form compound 34 substantially stabilized it
under an alkaline medium, and thus compound 34 was
readily subjected to the reaction with acrylic acid to give
the Michael adduct 35. After compound 35 had been
treated with sodium hydroxide to give the disodium salt
36, it was cyclized by heating at reflux in acetic anhydride
in the presence of sodium acetate to give the Dieckmann-
type ring-closure [11] product 37. Chlorination [12] of
compound 37 by treatment with N-chlorosuccinimide
(NCS) to give the chloride 38 was followed by deacetyl-
ation to 2-chloro-5,6-dihydrothieno[3,2-b]pyridine-
7(4H)-one 39, which was eventually treated using
method A to afford the desired compound 9.
Figure 6 shows the main 1H-NMR spectral data of the
final sulfonic acid derivatives 3–9 along with those of
parent 1. Full details of the 1H-NMR data for compounds

232
1
Figure 6. Main H-NMR spectral data of variant compounds 3–9 and 1 (M17055).
3–9 and synthesized intermediates are summarized in
tables I and II.
On the other hand, pyridine derivative 8 retained
substantial activity of the lead 1, while the thiophene
counterpart 9 was largely lost but still retained diuretic
activity. These results show that here again, as long as the
1,4-relationship of the N-(2-methylbenzoyl)amino group
and the O-sulfonic acid function is retained on the
tetrahydropyridine ring system, the adjacent phenyl
nucleus of the parent quinolinone skeleton may be
substituted by another heteroaromatic ring.
3. Diuretic activity of compounds 3–9
The diuretic activities of the newly synthesized com-
pounds 3–9 were determined as reported in the previous
paper [1] based on the ratio of increase in urine volume
after their administration to dogs, via the intrarenal artery
(i.r.a.) or intravenously (i.v.), to that after furosemide
administration to the same dogs at the same dose in the
same manner, the results are summarized in table III. The
diuretic activities of 1 (M17055) and furosemide are also
included in the table for the convenience of comparison.
It can be seen that compounds 3–5 completely lost
diuretic activity, showing that the 1,4-relationship of the
N-(2-methylbenzoyl)amino group and the O-sulfonic acid
group on the tetrahydropyridine ring system is crucial to
induce potency. Seven-membered heterocycle 6 and ring-
opened compound 7 did not show any activity, suggesting
that a slight deformation of the three-dimensional con-
formation of the tetrahydopyridine ring structure of the
lead 1 is detrimental.
4. Discussion
The structure–activity relationships of the newly syn-
thesized compounds 3–9 were considered with the aid of
computer graphics. As stated above, compounds 3–5
completely lost the diuretic activity of the lead 1. Fig-
ure 7 compares the computer-generated electrostatic po-
tential map of 1 (M17055) with those of compounds 3–5.
It can be seen that the alternation of the relative disposi-
–
tion of the -SO3 and N-acyl function induced various
changes in the negative charge distribution. Compounds 3
and 4 still retained the large negative charges and the
small ones located on the relevant functional groups as in

233
Table I. Physical data of the intermediates.
a
Compound M.p. (°C)
Formula
1H-NMR (ppm) [Solvent]
[recryst. solv.]
11
12
111–113
95–98
C10H8ClNO5 [CH2Cl2-Hexane]
3.94 (3H, s), 6.78 (1H, d, J = 14.2), 6.80 (1H, d, J = 14.2), 7.55 (1H, dd,
J = 8.6, J = 2.3), 7.89 (1H, d, J = 2.3), 8.21 (1H, d, J = 8.6) [CDCl3]
3.31 (6H, s), 3.52 (2H, s), 3.63 (3H, s), 7.20–7.60 (2H, m), 7.78 (1H, d,
J = 1.7) [CDCl3]
C12H14ClNO6 [CH2Cl2-Hexane]
13
14
125–127
116–117
C11H12ClNO4 [MeOH-Et2O]
C11H14ClNO2 [CH2Cl2-Hexane]
3.16 (2H, s), 3.36 (6H, s), 7.00–7.40 (3H, m), 8.70 (1H, br) [CDCl3]
2.91 (2H, s), 3.30 (8H, s), 3.90 (1H, br), 6.50–6.70 (2H, m), 6.86 (1H, d,
J = 7.9) [CDCl3]
15
17
135–137
99–102
C17H14ClNO2 [CH2Cl2-Hexane]
C17H14ClNO2 [CH2Cl2-Hexane]
2.30 (3H, s), 3.66 (2H, s), 4.39 (2H, s), 7.10–7.30 (7H, m) [CDCl3]
2.09 (3H, s), 4.40 (2H, s), 4.85 (2H, s), 6.86–7.68 (6H, m), 7.76–8.00 (1H,
m) [DMSO-d6]
20
21
24
120–122
165–168
134–135
C17H14ClNOS [CH2Cl2-Hexane]
C17H15ClN2O2 [CH2Cl2-Hexane]
C21H24ClNO6S [EtOH]
2.54 (3H, s), 3.20 (2H, t, J = 6.3), 4.13 (2H, t, J = 6.3), 6.92–7.40 (6H, m),
8.40 (1H, d, J = 9.2) [CDCl3]
2.26 (3H, s), 3.04 (2H, t, J = 6.4), 4.14 (2H, t, J = 6.4), 6.80–7.36 (6H, m),
7.68 (1H, br), 8.31 (1H, d, J = 9.2) [CDCl3]
1.23 (3H, t, J = 7.2), 1.80–2.10 (2H, m), 2.42 (2H, t, J = 7.2), 2.43 (3H,
s), 3.40–3.80 (2H, m), 3.82 (3H, s), 4.10 (2H, q, J = 7.2), 7.20–7.60 (6H,
m), 7.83 (1H, d, J = 8.6) [CDCl3]
26
27
28
30
31
33
37
38
39
40
83–84
99–103
138–139
121–123
189–194
174–177
97–98
C17H16ClNO3S [CH2Cl2-Hexane]
C10H10ClNO [Et2O-hexane]
C18H16ClNO2 [EtOH]
1.80–2.20 (2H, m), 2.43 (2H, t, J = 5.8), 2.44 (3H, s), 3.83 (2H, t, J = 6.4),
7.20–7.40 (3H, m), 7.50–7.80 (4H, m) [CDCl3]
1.90–2.40 (2H, m), 2.80 (2H, t, J = 6.3), 3.00–3.40 (2H, m), 4.69 (1H, br),
6.50–6.90 (2H, m), 7.63 (1H, d, J = 8.6) [CDCl3]
1.90–2.30 (2H, m), 2.41 (3H, s), 2.90 (2H, t, J = 6.3), 3.80–4.20 (2H, m),
6.70–7.10 (6H, m), 7.84 (1H, d, J = 8.0) [CDCl3]
2.55 (3H, s), 2.64 (3H, s), 7.00–8.00 (6H, m), 9.06 (1H, s), 12.19 (1H, s)
[CDCl3]
2.22 (3H, s), 2.44 (3H, s), 7.12–7.76 (6H, m), 8.54 (1H, d, J = 2.2), 11.32
(1H, s), 11.54 (1H, s) [DMSO-d6]
2.38 (3H, s), 2.91 (2H, t, J = 6.5), 4.40 (2H, t, J = 6.5), 6.72–7.44 (5H, m),
8.12 (1H, d, J = 8.4) [CDCl3]
2.34 (3H, s), 2.71 (2H, t, J = 6.6), 4.15 (2H, t, J = 6.6), 7.76 (1H, d,
J = 5.5), 8.00 (1H, d, J = 5.5) [DMSO-d6]
2.33 (3H, s), 2.72 (2H, t, J = 6.6), 4.15 (2H, t, J = 6.6), 7.80 (1H, s)
[DMSO-d6]
2.43 (2H, t, J = 7.1), 3.65 (2H, t, J = 7.1), 7.44 (1H, br), 7.00 (1H, s)
[DMSO-d6]
2.30 (3H, s), 2.73 (2H, t, J = 6.5), 3.96 (2H, t, J = 6.5), 7.04–7.52 (5H, m)
[DMSO-d6]
C16H14ClNO2 [CH2Cl2-Hexane]
C16H15ClN2O2 [CH2Cl2-Hexane]
C16H13ClN2O2 [CH2Cl2-Hexane]
C9H9NO2S [CH2Cl2-Hexane]
C9H8ClNO2S [CH2Cl2-Hexane]
C7H6ClNOS [CH2Cl2-Hexane]
C15H12ClNO2S [CH2Cl2-Hexane]
65–68
128–129
118–121
aCompounds were analysed by HR-MS.
Table II. Physical date of compounds 3–9.
Compound
M.p. (°C)a (dec)
Formulab
1H-NMR (DMSO-d6, ppm)
3
4
5
138–140
193–195
172–173
C17H14ClKN2O5S 2.25 (3H, s), 3.71 (2H, s), 4.51 (2H, s), 7.20–7.80 (7H, m)
C17H14ClKN2O5S 2.09 (3H, s), 4.40 (2H, s), 4.85 (2H, s), 6.86–7.68 (6H, m), 7.76–8.00 (1H, m)
C17H14ClKN2O5S 2.44 (3H, s), 3.08 (2H, t, J = 6.4), 4.02 (2H, t, J = 6.4), 6.72–7.52 (6H, m), 8.11 (1H,
d, J = 8.6)
6
114–115
C18H16ClKN2O5S 1.50–2.00 (2H, m), 2.26 (3H, s), 2.40–2.80 (2H, m), 3.50–4.00 (2H, m), 6.80–7.50
(7H, m)
7
8
9
171–174
130–133
172–175
C16H14ClKN2O5S 2.23 (3H, s), 2.45 (3H, s), 7.04–7.72 (6H, m), 8.48 (1H, d, J = 1.8), 10.93 (1H, s)
C16H13ClKN3O5S 2.27 (3H, s), 2.88–4.16 (4H, m), 6.80–7.36 (5H, m), 8.19 (1H, d, J = 7.9)
C15H12ClKN2O5S2 2.26 (3H, s), 2.64–2.92 (2H, m), 3.48–3.92 (2H, m), 7.20–7.44 (5H, m)
aRecrystalisation solvent CH2Cl2-MeOH. bCompounds 3 and 5 were analysed for C, H and N. Analytical results obtained for these elements
were within ± 0.4% of the calculated values for the formulae shown. Compounds 4 and 6–9 were analysed by HR-FAB-MS.

234
Figure 7. Electrostatic potential maps of 1 (M17055), 3, 4 and
5. The thick line is the contour at –30 kcal/mol and the thin line
at 10 kcal/mol of the electrostatic potential energy.
Figure 8. Electrostatic potential maps of 1 (M17055), 6 and 7.
The thick line is the contour at –30 kcal/mol and the thin line at
10 kcal/mol of the electrostatic potential energy.
the case of the lead 1, but their disposition patterns
apparently differ to some extent from that of the lead 1.
Furthermore, in the case of compound 5, the disposition
of the sulfonyl and acyl carbonyl function in close
vicinity brought about the collapse of the two negative
charges to form only one domain of a distinct negative
charge. These results may support the assumption that the
negative charge distribution on the tetrahydropyridine
backbone is the main determinant for effective interaction
with the Na+-K+-2Cl– cotransporter to induce diuretic
activity.
Interestingly, it was found that ring-enlarged 6 and
ring-opened compound 7 showed rather similar electro-
static potential maps to that of the lead 1, as shown in
figure 8, in spite of the fact that the two did not show any
activity at all (table III). Consideration of the three-
dimensional structures of these compounds seems to
answer the apparent exceptions to the active site model
hypothesis: figure 9 shows the molecular graphics of 1
(M17055) (in black) and compounds 6 and 7 (both in
grey), in which the main parts of the molecule containing
the chlorobenzene ring and the oxime sulfonic acid
moiety are superimposed on each other. The relevant
pictures show clearly that the 2-methylbenzoyl moieties
of both compounds 6 and 7 are shifted drastically from
their original position in the lead 1 to disturb effective
interaction with the active site of the cotransporter.
Figure 10 shows the electrostatic potential maps of
compounds 8 and 9 as compared with 1 (M17055). It can
be seen that the negative charge around the N-acyl
carbonyl function of compound 8 is slightly strengthened,
apparently due to the nearby nitrogen atom of the
pyridine ring, but retains almost the same characteristic
pattern of the negative charge distribution of the lead 1.
Thus, it may be reasonable to suppose that compound 8
retains ca. 70% of the activity of 1 (M17055).
In the case of compound 9, however, it is seen that the
charge distribution is somewhat disturbed from that of the
lead 1, which should result in the significant loss of the
activity. The perturbation of the charge distribution seems
to be brought about by the conformational change of the
2-methylbenzoyl moiety of compound 9 from that of the
lead 1, as also indicated by AM 1 calculation (figure 11).

235
Figure 9. Superimposition of active compound 1 (M17055) and inactive compounds 6 and 7.
The calculation indicated that the most stable conforma-
tion of compound 9 is the one in which the carbonyl
function is orientated in the opposite direction to that of
the lead 1. Therefore, it may be said that, in spite of the
appreciable change in the negative charge distribution,
compound 9 still retains the activity because the 1,4-
relationship of the N-acyl and the sulfonic acid moiety on
the tetrahydropyridine ring system is retained.
it has been shown that the pattern of electron potential
distributions afforded by the above molecular array is
Table III. Diuretic activities of compounds 1 and 3–9 and furo-
semide.
Compound
Diuretic activity
i.r.a.a
4.2
i.v.b
1 (M17055)
4.2
3
4
5
6
7
8
no activity
no activity
no activity
no activity
no activity
2.9
no activity
no activity
no activity
no activity
no activity
2.6
5. Conclusion
In order to shed more light on the origin of the diuretic
activity induced by 1 (M17055), several derivatives (3–9)
were synthesized by modifying the quinolinone ring
skeleton, and their activities in dogs were compared with
the lead 1. The pharmacological studies showed that the
molecular structure provided by the 4-oxime-O-sulfonic
acid and 1-N-acyl carbonyl moiety attached to a tetrahy-
dropyridine ring system is the determinant for the devel-
opment of the activity. With the aid of computer graphics
9
0.2
1.0
0.4
1.0
furosemide
aTest compounds were injected into the renal artery of dogs and
b
activity was expressed relative to that of furosemide. Test com-
pounds were administered intravenously into dogs and activity was
expressed relative to that of furosemide.

236
Figure 10. Electrostatic potential maps of 1 (M17055), 8 and
9. The thick line is the contour at –30 kcal/mol and the thin line
at 10 kcal/mol of the electrostatic potential energy.
Figure 11. Dihedral angle/energy curve of 1 (M17055) and 9.
1H-NMR spectra were taken on a JEOL FX-90A spec-
trometer with Me4Si as internal standard. Signal multi-
plicities are represented by s (singlet), d (doublet), dd
(double doublet), t (triplet), brs (broad singlet), and m
(multiplet). Chemical shifts were expressed in ppm and
coupling constants (J) in hertz (Hz). Mass spectra (EI-
MS) and high-resolution mass spectra (HRMS or HR-
FAB-MS) were obtained on a JEOL DX-300 and a JEOL
SX-102A mass spectrometer. Elemental analysis was
carried out with a Carlo Erba model 1106 analyzer and
the results were within ± 0.40% of the calculated values.
For column chromatography, silica gel (Kieselgel 60,
70–230 mesh, Merck) was used.
crucial for the development of the activity, which sub-
stantiates the previously proposed active site model. Of
special importance in a series of this study is the finding
in the previous paper [3] that the C=N double bond of
tricyclic compounds 2 plays the same role as the C=O
function of compound 1, and type 2 compounds can
induce diuretic activity comparable to, or even better
than, type 1 bicyclic compounds. These facts can reason-
ably be explained by assuming that coplanarity of type 2
compounds represents the most suitable conformation of
1 (M17055) to fit to the proposed Na+-K+-2Cl– cotrans-
porter active site, having lead us to adopt compound 2a as
the promising candidate to succeed parent 1 (M17055)
for pharmacological and clinical studies.
1
Melting points, formulae and H-NMR data for syn-
thesized intermediates and final sulfonic acid derivatives
are summarized in tables I and II.
6. Experimental protocols
6.1.1. Methyl-4-chloro-2-nitrophenylpyruvate 11
A mixture of 4-chloro-2-nitrotoluene 10 (171.6 g,
1 mol), dimethyl oxalate (118.1 g, 1 mol) and sodium
methoxide (1 mol) in methanol (350 mL) was refluxed
for 2 h. The reaction mixture was poured into H2O,
6.1. Chemistry
Melting points were determined on a Mettler FP-800
hot stage melting point apparatus and are uncorrected.

237
acidified with 2 N HCl (aq.) and extracted with AcOEt.
The extract was washed with water and brine, dried over
Na2SO4 and evaporated. The residue was purified by
silica gel column chromatography with AcOEt-hexane
(1:5) as an eluent to give compound 11 (103.2 g, 40%).
with AcOEt/hexane (1:5) as an eluent to give 7-chloro-
1,4-dihydro-1-(2-methylbenzoyl)-3(2H)-quinolinone di-
methyl acetal (1.5 g, 98%). To a solution of the dimethyl
acetal (1.5 g, 4.34 mmol) in THF (40 mL) was added 1 N
HCl (aq.) (20 mL) with stirring. The mixture was stirred
at room temperature for 20 h, poured into H2O and
extracted with Et2O. The extract was washed with
NaHCO3 (aq.) and water, dried over Na2SO4 and evapo-
rated. The residue was purified by silica gel column
chromatography with AcOEt/hexane (1:5) as an eluent to
give compound 15 (0.6 g, 45%).
6.1.2. Methyl 4-chloro-2-nitrophenylpyruvate
dimethyl acetal 12
To a mixture of compound 11 (95 g, 0.369 mol),
trimethyl orthoformate (413 g, 3.89 mol) and MeOH
(700 mL) was added boron trifluoride etherate (140 g,
0.98 mol) at room temperature. The reaction mixture was
refluxed for 10 h, and then the solvent was removed. The
residue was poured into H2O and extracted with AcOEt.
The extract was washed with water and brine, dried over
Na2SO4 and evaporated. The residue was purified by
silica gel column chromatography with AcOEt-hexane
(1:4) as an eluent to give compound 12 (58.5 g, 52%).
6.1.6. 7-Chloro-1,4-dihydro-1-(2-methylbenzoyl)-3(2H)-
quinolinone 3-oxime-O-sulfonic acid potassium salt 3
To a mixture of compound 15 (1.0 g, 3.34 mmol),
MeOH (20 mL) and CH2Cl2 (10 mL) was added
hydroxylamine-O-sulfonic acid (0.6 g, 5.31 mmol) at
room temperature. The mixture was stirred at room
temperature for 30 min, and an aqueous solution of
K2CO3 (0.73 g in 1 mL of H2O, 5.31 mmol) was added.
The reaction mixture was stirred at room temperature for
5 h, and then the solvent was removed. The residue was
purified by silica gel column chromatography with
CH2Cl2/MeOH (5:1) as an eluent to give a white solid,
which was recrystallized from MeOH/CH2Cl2 to give
compound 3 (0.38 g, 26%).
6.1.3. 7-Chloro-3,4-dihydro-3,3-dimethoxy-
1-hydroxy-2(1H)-quinolinone 13
A solution of compound 12 (55.5 g, 0.183 mol) in
MeOH (200 mL) was stirred with 5% Pd-C (2.2 g) in an
atmosphere of H2 at room temperature for 20 h. The
catalyst was filtered off and the filtrate was evaporated.
Recrystallization from MeOH/Et2O gave compound 13
(29.5 g, 62%).
6.1.4. 7-Chloro-1,4-dihydro-3(2H)-quinolinone
dimethyl acetal 14
6.1.7. 7-Chloro-2,3-dihydro-2-(2-methylbenzoyl)-
4(1H)-isoquinolinone 17
To a solution of compound 13 (27.5 g, 0.107 mol) in
benzene (200 mL) was added sodium bis(2-methoxy-
ethoxy)aluminum hydride (ca. 3.4 M in toluene: 80 mL,
0.272 mol) in benzene (400 mL) at 0 °C with stirring.
After the mixture had been stirred at room temperature
for 22 h, 1 N NaOH (aq.) was added. An insoluble
material was filtered off and the filtrate was extracted
with Et2O. The extract was washed with water and brine,
dried over Na2SO4 and evaporated. The residue was
purified by silica gel column chromatography with
AcOEt/hexane (1:10) as an eluent to give compound 14
(4.1 g, 17%).
This compound (85%) was prepared from compound
16 by the same procedure as for compound 15.
6.1.8. 7-Chloro-2,3-dihydro-2-(2-methylbenzoyl)-4(1H)-
isoquinolinone 4-oxime-O-sulfonic acid potassium salt 4
This compound (22%) was prepared from compound
17 by the same procedure as for compound 3.
6.1.9. 6-Chloro-3,4-dihydro-2-(2-methylbenzoyl)-
1(2H)-isoquinolinethione 20
To a solution of 6-chloro-3,4-dihydro-1(2H)-isoqui-
noline 18 (4.65 g, 25.6 mmol) in CHCl3 (60 mL) was
added phosphorous pentasulfide (7.4 g, 33.3 mmol). The
reaction mixture was stirred at room temperature for 13 h
and refluxed for 2 h. An insoluble matter was filtered off
and the solvent was removed. The residue was purified by
silica gel column chromatography with CH2Cl2 as an
eluent to give 6-chloro-3,4-dihydro-1(2H)-isoquinoline-
thione 19 (4.35 g, 86%). Compound 20 (83%) was
prepared from compound 19 by the same procedure as for
compound 15.
6.1.5. 7-Chloro-1,4-dihydro-1-(2-methylbenzoyl)-
3(2H)-quinolinone 15
To a mixture of compound 14 (1.0 g, 4.40 mmol),
pyridine (0.83 g, 10.5 mmol) and CH2Cl2 (10 mL) was
added 2-methylbenzoyl chloride (0.82 g, 5.29 mmol) at
room temperature with stirring. The mixture was refluxed
for 4 h. The mixture was poured into H2O and extracted
with CH2Cl2. The extract was washed with 1 N HCl (aq.)
and water, dried over Na2SO4 and evaporated. The
residue was purified by silica gel column chromatography

238
6.1.10. 6-Chloro-3,4-dihydro-2-(2-methylbenzoyl)-
1(2H)-isoquinolinone 1-oxime 21
(aq.) (60 mL), and the mixture was refluxed for 16 h. The
reaction mixture was poured into H2O and extracted with
AcOEt. The extract was washed with water and NaHCO3
(aq.), dried over Na2SO4 and evaporated. The residue was
purified by silica gel column chromatography with
AcOEt/hexane (1:4) as an eluent to give compound 26
(7.3 g, 40%).
To a mixture of compound 20 (3.4 g, 10.8 mmol),
pyridine (2.60 g, 32.4 mmol) and EtOH (150 mL) was
added hydroxylamine hydrochloride (1.13 g, 16.3 mmol),
and the mixture was refluxed for 2 h. The mixture was
poured into 1 N HCl (aq.) and extracted with AcOEt. The
extract was washed with water and brine, dried over
Na2SO4 and evaporated. The residue was purified by
silica gel column chromatography with MeOH/CH2Cl2
(1:19) as an eluent to give compound 21 (0.58 g, 17%).
6.1.14. 8-Chloro-1,2,3,4-tetrahydro-
1-benzazepin-5(5H)-one 27
To a mixture of compound 26 (17 g, 48.6 mmol),
anisole (10 mL) and trifluoroacetic acid (30 mL) was
added trifluoromethanesulfonic acid (5 mL), and stirring
was continued for 2 h at room temperature. The precipi-
tated crystals were separated by filtration. To a mixture of
the product and AcOEt (50 mL) was added NaHCO3
(aq.), and then extracted with AcOEt. The extract was
washed with water and brine, dried over Na2SO4 and
evaporated. Recrystallization from Et2O/hexane gave
compound 27 (7.3 g, 77%).
6.1.11. 6-Chloro-3,4-dihydro-2-(2-methylbenzoyl)-
1(2H)-isoquinolinone 1-oxime-O-sulfonic acid
potassium salt 5
To a solution of compound 21 (12.0 g, 38.1 mmol) in
CH2Cl2 (250 mL) was added pyridine-sulfur trioxide
complex (7.1 g, 44.6 mmol). The reaction mixture was
stirred at room temperature for 24 h, and the solvent was
removed. To the residue were added MeOH (200 mL) and
then an aqueous solution of K2CO3 (6.2 g in 10 mL of
H2O, 45.0 mmol). The reaction mixture was stirred at
room temperature for 5 h, and then the solvent was
removed. The residue was purified by silica gel column
chromatography with CH2Cl2/MeOH (5:1) as an eluent to
give a white solid, which was recrystallized from MeOH/
CH2Cl2 to give compound 5 (6.1 g, 37%).
6.1.15 8-Chloro-1-(2-methylbenzoyl)-1,2,3,4-
tetrahydro-1-benzazepin-5(5H)-one 28
This compound (2.9 g, 90%) was prepared from com-
pound 27 by the same procedure as for compound 15.
6.1.16. 8-Chloro-1-(2-methylbenzoyl)-1,2,3,4-
tetrahydro-1-benzazepin-5(5H)-one 5-oxime-
O-sulfonic acid potassium salt 6
This compound (44%) was prepared from compound
28 by the same procedure as for compound 3.
6.1.12. Ethyl 4-[N-(5-chloro-2-methoxycarbonylphenyl)-
N-p-toluenesulfonylamino]butyrate 24
A mixture of NaH (60% dispersion in mineral oil; 16 g,
0.4 mol), dry DMF (800 mL) and methyl 4-chloro-2-(p-
toluenesulfonylamino)benzoate 23 (135 g, 0.398 mol),
which had been prepared from methyl 2-amino-4-
chlorobenzoate 22, was stirred at room temperature for
0.5 h. To the mixture was added ethyl 4-bromobutyrate
(93 g, 0.475 mol) and the mixture was stirred at 90 °C for
5 h. The mixture was poured into H2O and extracted with
CH2Cl2. The extract was washed with water and brine,
dried over Na2SO4 and evaporated. Recrystallization
from EtOH gave compound 24 (115 g, 64%).
6.1.17. 1-[4-Chloro-2-(2-methylbenzoyl)aminophenyl]-
ethanone 30
This compound (62%) was prepared from 1-[2-amino-
4-chlorophenyl]ethanone 29 by the same procedure as for
compound 15.
6.1.18. 1-[4-Chloro-2-(2-methylbenzoyl)aminophenyl]-
ethanone oxime 31
This compound (58%) was prepared from compound
30 by the same procedure as for compound 21.
6.1.13. 8-Chloro-1-(p-toluenesulfonyl)-1,2,3,4-
tetrahydro-1-benzazepin-5(5H)-one 26
6.1.19. 1-[4-Chloro-2-(2-methylbenzoyl)aminophenyl]-
ethanone oxime-O-sulfonic acid potassium salt 7
This compound (24%) was prepared from compound
31 by the same procedure as for compound 5.
To a refluxed solution of potassium tert-butoxide
(12.4 g, 110 mmol) in toluene (500 mL) was added a
toluene (300 mL) solution of compound 24 (24 g,
52.9 mmol), and stirring was continued for 1 h. The
reaction mixture was poured into H2O, acidified with 1 N
HCl (aq.) and extracted with AcOEt. The extract was
washed with water and brine, dried over Na2SO4 and
evaporated. To a mixture of the crude product 25 in
AcOH (180 mL) and EtOH (60 mL) was added 6 N HCl
6.1.20. 7-Chloro-2,3-dihydro-1-(2-methylbenzoyl)-
1,8-naphthyridine-4(1H)-one 33
To a cooled (–70 to –75 °C) solution of 7-chloro-
2,3-dihydro-1,8-naphthyridine-4(1H)-one 32 (0.6 g,
3.29 mmol) in dry THF (30 mL) was added lithium

239
bis(trimethylsilyl)amide (ca. 1.0 M in hexane: 3.3 mL,
3.30 mmol), and stirring was continued for 0.5 h. A
solution of 2-methylbenzoyl chloride (0.6 g, 3.83 mmol)
in THF (5 mL) was added over a 0.5 h period. The
reaction mixture was then warmed to –25 °C, poured into
H2O and extracted with AcOEt. The extract was washed
with water and brine, dried over Na2SO4 and evaporated.
Recrystallization from CH2Cl2/hexane gave compound
33 (0.11 g, 10%).
purified by silica gel column chromatography with
CH2Cl2/hexane (3:1) as an eluent to give compound 39
(1.3 g, 58%).
6.1.25. 2-Chloro-5,6-dihydro-4-(2-methylbenzoyl)-
thieno[3,2-b]pyridine-7(4H)-one 40
This compound (32%) was prepared from compound
39 by the same procedure as for compound 15.
6.1.26. 2-Chloro-5,6-dihydro-4-(2-methylbenzoyl)-
thieno[3,2-b]pyridine-7(4H)-one 7-oxime-O-
sulfonic acid potassium salt 9
This compound (29%) was prepared from compound
40 by the same procedure as for compound 3.
6.1.21. 7-Chloro-2,3-dihydro-1-(2-methylbenzoyl)-
1,8-naphthyridine-4(1H)-one 4-oxime-O-
sulfonic acid potassium salt 8
This compound (23%) was prepared from compound
33 by the same procedure as for compound 3.
6.1.22. 4-Acetyl-5,6-dihydro thieno[3,2-b]-
pyridine-7(4H)-one 37
6.2. Methods of computational chemistry
A mixture of an aqueous solution of NaOH (12.0 g in
50 mL of H2O, 0.3 mol) and 3-(2-methoxycarbonyl-3-
thienylamino)propionic acid 35 (35 g, 0.153 mol), which
had been prepared from methyl 3-amino-2-
thiophenecarboxylate 34 and acrylic acid, was refluxed
for 1 h, and then the water was removed. To a solution of
the crude compound 36 in Ac2O (150 mL) was added
AcONa (25 g, 0.305 mol), and the mixture was refluxed
for 1 h. The reaction mixture was poured into H2O and
extracted with Et2O. The extract was washed with water
and NaHCO3 (aq.), dried over Na2SO4 and evaporated.
The residue was purified by silica gel column chroma-
tography with CH2Cl2/hexane (3:2) as an eluent to give
compound 37 (11.0 g, 37%).
6.2.1 Computer programs
The ab initio molecular orbital calculation program
GAUSSIAN 92 (GAUSSIAN Inc.), semi-empirical mo-
lecular orbital calculation program MOPAC 6.0 (JCPE),
and molecular modelling package software SYBYL 6.0
(TRIPOS Inc.) were run on a Indigo 2 work station
(Silicon Graphics Inc.).
6.2.2. Molecular modelling
The starting geometries of compounds 3–9 were con-
structed from the X-ray crystal structure of compound 1
(M17055) and modified where necessary using the frag-
ment library of SYBYL 6.0. Those geometries were
optimized by the semi-empirical molecular orbital AM 1
method in MOPAC 6.0. The molecular geometry of
furosemide was provided by the Cambridge Structural
Database (CSD).
6.1.23. 2-Chloro-4-acetyl-5,6-dihydrothieno-
[3,2-b]pyridine-7(4H)-one 38
To a solution of compound 37 (7.7 g, 39.5 mmol) in
AcOH (60 mL) was added N-chlorosuccinimide (6.3 g,
47.2 mmol). The reaction mixture was refluxed for 1 h
and the solvent was removed. To the residue was added
NaHCO3 (aq.) which was then extracted with AcOEt. The
extract was washed with water and brine, dried over
Na2SO4 and evaporated. The residue was purified by
silica gel column chromatography with CH2Cl2/hexane
(2:3) as an eluent to give compound 38 (3.1 g, 34%).
6.2.3. Electrostatic potential contour map preparation
Electrostatic potential was calculated using the classi-
cal Coulomb’s equation, charges were estimated by the
CHELP method using the 3-21G* basis set provided with
GAUSSIAN 92. Contour maps of electrostatic potential
were graphically represented by the isosurface of specific
energy (–30 or 10 kcal/mol).
6.1.24. 2-Chloro-5,6-dihydrothieno[3,2-b]-
pyridine-7(4H)-one 39
6.2.4. Dihedral angle/energy curve preparation
(coordinate driving method)
The energy (heat of formation; kcal/mol) was calcu-
lated against the dihedral angle φ (C–N–C=O) by rotating
the dihedral angle in 0–360 degrees in 10 steps by using
the AM 1 semi-empirical molecular orbital method at
every point.
A mixture of compound 38 (2.8 g, 12.2 mmol), 2 N
HCl (aq.) (6 mL) and AcOH (40 mL) was refluxed for 2 h
and then the solvent was removed. To the residue was
added NaHCO3 (aq.), which was then extracted with
AcOEt. The extract was washed with water and brine,
dried over Na2SO4 and evaporated. The residue was

240
6.3. Pharmacology
Increase in urine output in 90 min (∆UV90) was
determined as follows:
∆UV90 = (urine output in 90 min after the drug
injection) – [(urine output in 30 min before the drug
injection) × 3]
Diuretic activity was expressed as the ratio of ∆UV90
to that for furosemide administered at the same dose to
the same dog.
6.3.1. Injection via renal artery (i.r.a.)
Mongrel dogs weighing 7–15 kg were used after over-
night fasting with free access to H2O. They were anaes-
thetized with pentobarbital (30 mg/kg, i.v.) and venti-
lated. Following a left flank incision, the left ureter was
cannulated for urine collection and an L-shaped needle
connected to polyethylene tubing was inserted into the
left renal artery for drug administration. The drug injec-
tion route was maintained by infusing 0.9% aq. NaCl
(saline) at 0.05 mL/kg/min. Following the operation,
prime 3 mL/kg saline was given initially and saline was
continuously infused at 0.1 mL/kg/min from a catheter in
the femoral vein. After an equilibration period of 1–2 h,
urine was collected every 5 min. All compounds were
dissolved in alkaline solution prior to left renal artery
injection at 0.01 mg/kg. Administrations were conducted
at appropriate intervals.
Increase in urine output in 20 min (∆UV20) was
computed as follows:
∆UV20 = (urine output in 20 min after the drug
injection) – (urine output in 20 min before the drug
injection).
Diuretic activity was expressed as the ratio of ∆UV20
to that for furosemide injected in the same dog at the
same dose.
Acknowledgements
We wish to thank members of the Mochida diuretic
project team for their contributions to this work.
References
[1] Nishijima K., Shinkawa T., Kato K., Sato N., Nishida H., Yamashita
Y. et al., Eur. J. Med. Chem. 33 (1998) 267–277.
[2] Cragoe E.J. (Ed.), Chemistry and Pharmacology of Drugs. Diuretics:
Chemistry, Pharmacology and Medicine, John Wiley and Sons Inc.,
New York, 1983.
[3] Nishijima K., Shinkawa T., Ito M., Nishida H., Yamamoto I., Onuki
Y., Inaba H., Miyano S., Eur. J. Med. Chem. 33 (1998) 763–774.
[4] Kawahara N., Nakajima T., Itoh T., Ogura H., Heterocycles 16
(1981) 729–731.
[5] Grethe G., Lee H.L., Uskokovic M., Brossi A., J. Org. Chem. 33
(1968) 491–503.
[6] Grethe G., Toome V., Lee H.L., Uskokovic M., Brossi A., J. Org.
Chem. 33 (1968) 504–508.
6.3.2. Injection intravenously (i.v.)
Experimental procedures were essentially as for i.r.a.
The few exceptions are as follows:
– No needle was attached to the renal artery.
– Infusion rate of saline into a femoral vein was
always 0.15 mL/kg/min.
[7] Tomita M., Minami S., Uyeo S., J. Chem. Soc. C (1969) 183–188.
[8] Bell W.H., Hannah E.D., Proctor G.R., J. Chem. Soc. (1964)
4926–4930.
[9] Leonard N.J., Boyd S.N., J. Org. Chem. 11 (1946) 405–418.
[10] Settimo A.D., Biagi G., Primofiore G., Ferrarini P.L., Livi O., Marini
A.M., J. Heterocycl. Chem. 17 (1980) 1225–1229.
– Urine was collected every 10 min.
– Compound dosage into the femoral vein was
0.1 mg/kg.
[11] Bekhli A.F., Khim. Geterotsikl. Soedin. 65–67; Chem. Abstracts 72
(1970) 121309c.
[12] Gronowitz S., Laguna J.F., Acta. Pharm. Suecica. 5 (1968) 563–578.