Design, synthesis, and biological evaluation of tricyclic heterocycle-tetraamine conjugates as potent NMDA channel blockers

Article abstract of DOI:10.1016/j.bmcl.2007.06.069

We have developed a new class of N-methyl-d-aspartate (NMDA) channel blockers having a conjugate structure that consists of a nitrogenous heterocyclic head and a tetraamine tail. Among them, dihydrodibenzazepine-homospermine conjugate (8) exhibited potent

Full text of DOI:10.1016/j.bmcl.2007.06.069

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4729–4732
Design, synthesis, and biological evaluation of tricyclic
heterocycle-tetraamine conjugates as potent
NMDA channel blockers
Hiromitsu Takayama,a,* Yuichi Yaegashi,a Mariko Kitajima,a Xia Han,b
Kazuhiro Nishimura,b Shigeru Okuyamac and Kazuei Igarashib,*
aGraduate School of Pharmaceutical Sciences, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
bGraduate School of Pharmaceutical Sciences, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8675, Japan
cMedicinal Development Research Laboratories, Taisho Pharmaceutical Co., Ltd, 403 Yoshino-cho, Kita-ku, Saitama 331-9530, Japan
Received 22 May 2007; revised 20 June 2007; accepted 22 June 2007
Available online 27 June 2007
Abstract—We have developed a new class of N-methyl-D-aspartate (NMDA) channel blockers having a conjugate structure that
consists of a nitrogenous heterocyclic head and a tetraamine tail. Among them, dihydrodibenzazepine-homospermine conjugate
(8) exhibited potent antagonistic activity at NR1/NR2A or NR1/NR2B NMDA subtype receptors compared with the lead com-
pound, AQ343 (1), or memantine, as well as weak cytotoxicity. Its superior biological profiles compared with known compounds
point to its potential use as therapeutic agents for neurological disorders.
ꢀ 2007 Elsevier Ltd. All rights reserved.
N-Methyl-D-aspartate (NMDA) receptors, which are
ligand-gated cation channels embedded in the cell mem-
brane of neurons, have been implicated in learning and
memory, and may also play a central role in various
conditions leading to neuronal degeneration. Overexci-
tation of NMDA receptors leads to excessive Ca2+ influx
through receptor-associated ion channels, resulting in
neuronal cell injury or death. Therefore, NMDA recep-
tor antagonists could be of therapeutic benefit to a num-
ber of neurological disorders, such as stroke,
Alzheimer’s disease, or Parkinson’s disease. Memantine,
which possesses an aminoadamantane structure, was re-
cently approved by the European Union and US FAD
for the treatment of dementia.1
that of memantine. Further, we have revealed that the
angle between the polycyclic head moiety and the linear
polyamine tail, as well as the length of the polyamine it-
self, is important for interactions with the NMDA chan-
nel and for the activity of these compounds. Based on
these findings, we further investigated new types of
NMDA antagonists by utilizing AQ343 as the lead com-
pound in order to develop more efficient NMDA chan-
nel blockers having potential clinical applications. Here
we report the synthesis and biological evaluation of tri-
cyclic heterocycle-tetraamine conjugates as NMDA
channel blockers.
As the head moiety of new conjugates, we have
adopted tricyclic nitrogenous heterocycles, i.e., 5H-
dibenz[b,f]azepine (A), 10,11-dihydro-5H-dibenz[b,f]aze-
pine (B), phenothiazine (C), and carbazole (D), in
the present study. Tricyclic heterocycle-tetraamine con-
juncts (3–14) were synthesized by coupling carboxylic
acid derivatives of tricyclic heterocycles with tetraam-
ines, including spermine, N1-(3-(4-aminobutylami-
no)propyl)butane-1,4-diamine, and homospermine, i.e.,
N1-(4-(4-aminobutylamino)butyl)butane-1,4-diamine,
as follows.
In our continuing research on NMDA receptor channel
blockers,2 we have found that polycycle-polyamine con-
jugates, such as AQ343 (1) and AQ444 (2), are reversible
and voltage-dependent NMDA blockers, particularly at
NR1/NR2A and NR1/NR2B receptors,3 and their
antagonistic activities (IC50 values) are more potent than
Keywords: NMDA receptor; Channel blocker; Polyamine; Nitroge-
nous heterocycle; Dementia.
*
Starting from 1,4-diaminobutane (15), N1-nitro-
sobenzenesulfonyl-N4-Boc derivative (17)4 was prepared
Corresponding authors. Tel./fax: +81 43 290 2901 (H.T.); e-mail:
takayama@p.chiba-u.ac.jp
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.069

4730
H. Takayama et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4729–4732
O
O
O
by successive treatment with 2-nitrobenzenesulfonyl
H
chloride5 (0.33 eq.) and Boc2O (Scheme 1). Then, amina-
tion of 1,3-dibromopropane (18) or 1,4-dibromobutane
(19) with 17 afforded linear polyamine derivative (20)
or (21), respectively, which was then treated with triflu-
oroacetic acid to remove the Boc group, giving terminal
amine-free tetraamine derivatives (22 and 23) in good
yields (Fig. 1).
N
NH2
N
H
N
H
O
O
AQ343 (1)
H
N
NH2
N
H
N
H
O
Next, coupling of the heterocyclic head moiety with
the polyamine tail was carried out. Spermine-conju-
gated dibenzazepine, dihydrodibenzazepine, and phe-
nothiazine compounds (3, 6, 9) were, respectively,
prepared by condensation of their carbamoyl chlorides
(24–26) and a large excess of spermine itself. In the
case of compound 12, due to the high reactivity of
the carbamoyl chloride derivative of carbazole (D),
chloride 27 was initially converted into p-nitrophenyl
carbamate (28) and then condensed with a large excess
of spermine. In turn, N1-(3-(4-aminobutylamino)pro-
pyl)butane-1,4-diamine- or N1-(4-(4-aminobutylami-
no)butyl)butane-1,4-diamine-conjugated dibenzazepine,
dihydrodibenzazepine, and phenothiazine compounds
were, respectively, prepared by coupling their carbam-
oyl chlorides (24–26) with protected tetraamines (22 or
23), followed by the removal of Ns groups using
thiophenol in the presence of Cs2CO3 or K2CO3 in
DMF.5 As in the preparation of conjugate 12, the
p-nitrophenyl carbamate of carbazole (28) was used
for the syntheses of tetraamines (13 and 14). The
structures of all the synthetic compounds were con-
firmed by analysis of spectroscopic data (1H, 13C
NMR, MS, HRMS, UV, IR).
AQ444 (2)
S
N
H
N
H
N
H
N
H
(B)
(C)
(A)
(D)
Figure 1. Structures of AQ343, AQ444, and tricyclic heterocycles.
15: R1=R2=H
NHR2
i
16: R1=Ns, R2=H
R1HN
ii
17: R1=Ns, R2=BOC
18
19
Br
Br
Br
or
iii
Br
R1
R1
N
N
n
R2HN
NHR2
20: n=3, R1=Ns, R2=Boc
22: n=3, R1=Ns, R2=H
21: n=4, R1=Ns, R2=Boc
23: n=4, R1=Ns, R2=H
iv
iv
The antagonistic effects of prepared compounds (3–14)
were studied using recombinant NMDA receptors ex-
pressed in Xenopus laevis oocytes. Most NMDA recep-
tors in the adult central nervous system contain
combinations of NR1 and NR2, with NR2A and
Scheme 1. Reagents and conditions: (i) NsCl (0.33 equiv), EtOH, 0 ꢁC,
30 min, 62%; (ii) Boc2O, CH2Cl2, rt, 2 h, quant.;(iii)18 or 19, DMF, 60 ꢁC,
15 h, 96% for 20, 95% for 21; (iv) TFA, CH2Cl2, rt, 1.5 h, quant.
S
N
N
N
Cl
Cl
24 (A)
O
O
Cl
O
26 (C)
25 (B)
i
iii
N
O
Cl
Ns
N
Ns
N
H
27 (D)
ii
R
N
NH2
iv
m
n
m
O
R
A
B
C
D
m=4, n=3 m=4, n=4
H
H
H
iii
R
N
N
N
NH2
29
30
31
32
33
34
35
36
m
n
m
O
R
m=3, n=4 m=4, n=3 m=4, n=4
N
NO2
A
B
C
D
3
6
4
7
10
13
5
8
11
14
i
O
O
9
12
28
Scheme 2. Reagents and conditions: (i) spermine (4.8–5 equiv), CH2Cl2, rt (for 3, 6) or 0 ꢁC (for 9, 12); (ii) p-nitrophenol, Et3N, CH2Cl2, rt; (iii)
tetraamine 22 (2 equiv) or 23 (2 equiv), CH2Cl2, rt, or 40 ꢁC (for 36); (iv) PhSH, Cs2CO3 or K2CO3 (for 5, 13, 14), DMF, rt.

H. Takayama et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4729–4732
4731
Figure 2. Effect of dihydrodibenzazepine-homospermine 8 on NR1/NR2A and NR1/NR2B receptors. Channel blocking activity of compound 8 was
measured using an oocyte voltage clamped at À70 mV. Receptors were activated by superfusing glycine (Gly) and glutamate (Glu) (10 lM).
Macroscopic currents were recorded with a two-electrode voltage clamp using a Gene Clamp 500 amplifier (molecular devices). (a) Representative
traces are shown to illustrate the effect of 1 lM compound 8. (b) Concentration–inhibition curves were determined at NR1/NR2A and NR1/NR2B
receptors.
NR2B predominating in forebrain areas such as the
cerebral cortex.6 Thus, IC50 values were obtained from
concentration–inhibition curves with five or six concen-
trations of each compound at NR1/NR2A or NR1/
NR2B NMDA receptors.7 Representative traces of the
effect of 1 lM compound 8 on NR1/NR2A and NR1/
NR2B receptors are shown in Figure 2a, and concentra-
tion and inhibition curves of compound 8 are shown in
Figure 2b.
Compared to memantine and the anthraquinone deriva-
tive, AQ343 (1), almost all the conjugates (3–14) pre-
pared in the present study were potent antagonists of
recombinant NMDA receptors (Table 1). Particularly,
dihydrodibenzazepine-tetraamine (8) (IC50 0.072 lM at
NR1/NR2A), phenothiazine-tetraamine (10) (IC50
0.078 lM at NR1/NR2A), and carbazole-spermine (12)
(IC50 0.083 lM at NR1/NR2B) were the most effective
compounds of this series, showing approximately
Table 1. Biological assay for memantine and compounds 1 and 3–14
H
H
H
R
N
N
N
NH2
m
n
m
O
Cytotoxicityb
Selectivity index
a
Compound
R
m
n
Channel block IC50 (lM)
NR1/NR2A
NR1/NR2B
IC50 (lM)
Cytotoxicity/channel blockc
Memantine
1 (AQ343)
—
—
3
—
4
0.911 0.043
0.392 0.040
0.203 0.032
0.198 0.013
0.240 0.045
1.02 0.077
0.491 0.060
0.199 0.021
0.335 0.103
0.293 0.063
147 10.3
14.9 0.38
97.6 1.49
115 2.89
132 8.77
152
33.7
486
432
495
3
4
5
3
4
4
3
N
N
4
4
6
7
8
3
4
4
4
3
4
0.958 0.064
0.117 0.013
0.072 0.017
0.442 0.063
0.270 0.015
0.142 0.023
102 4.21
169 15.9
133 8.15
146
873
1242
S
N
9
3
4
4
4
3
4
0.215 0.009
0.078 0.021
0.127 0.019
0.348 0.032
0.108 0.009
0.108 0.030
57.1 4.60
32.5 3.81
40.4 2.69
202
349
344
10
11
12
13
14
3
4
4
4
3
4
0.145 0.051
0.211 0.013
0.111 0.029
0.083 0.019
0.189 0.043
0.166 0.033
31.3 9.48
42.3 4.21
84.2 2.17
275
212
607
N
a Oocytes were injected with NR1 plus NR2 cRNAs in a ratio of 1:5 (0.1–4 ng of NR1 plus 0.5–20 ng of NR2). Microscopic currents were recorded at
À70 mV. Values are means SEM of four oocytes.7
b FM3A cells (1 · 104) were cultured in ES medium in the presence or absence of various concentrations of each compound at 37 ꢁC for 48 h. The
number of cells was counted in the presence of 0.05% Trypan blue. IC50 values are means SD of triplicate determinations.8
c Selectivity index (cytotoxicity/channel block) was determined using the mean value of IC50 of NR1/N2A and NR1/NR2B NMDA receptors.8

4732
H. Takayama et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4729–4732
12-fold stronger activity than memantine. Furthermore,
it was found that the optimal length of the tetraamine
part (343, 434, or 444) required to exhibit potent activity
depended on the structure of the head moiety (Scheme 2).
Acknowledgments
This work was supported by a Grant-in-Aid for Scien-
tific Research from the Ministry of Education, Culture,
Science and Technology, Japan; the Japan Science and
Technology Agency; and the Tokyo Biochemical Foun-
dation, Japan.
The degree of inhibition of cell growth (IC50) was then
examined using mouse mammary carcinoma FM3A
cells and five or six concentrations of each compound.8
Compounds (3–8) having
a
dibenzazepine or
References and notes
dihydrodibenzazepine head showed weak cytotoxicity
similar to memantine, which was nearly 10-fold weaker
than our lead compound, AQ343 (1). Because of the
decrease in cytotoxicity of the newly synthesized
compounds, the ratio of channel blocking activity to
cytotoxicity (see selectivity index in Table 1) of our com-
pounds was superior to that of memantine. In particular,
the selectivity index of dihydrodibenzazepine-homosper-
mine (8) was 8-fold higher than that of memantine.
1. Lipton, A. A. Nat. Rev. Drug Discov. 2006, 5, 160.
2. Kashiwagi, K.; Tanaka, I.; Tamura, M.; Sugiyama, H.;
Okawara, T.; Otsuka, M.; Sabado, T. N.; Williams, K.;
Igarashi, K. J. Pharmacol. Exp. Ther. 2004, 309, 884.
3. Jin, L.; Sugiyama, H.; Takigawa, M.; Katagiri, D.; Tom-
itori, H.; Nishimura, K.; Kaur, N.; Phanstiel, O., IV;
Kitajima, M.; Takayama, H.; Okawara, T.; Williams, K.;
Kashiwagi, K.; Igarashi, K. J. Pharmacol. Exp. Ther. 2007,
320, 47.
4. Wang, T.; An, H.; Vickers, T. A.; Bharadwaj, R.; Cook, P.
D. Tetrahedron 1998, 54, 7955.
5. Hidai, Y.; Kan, T.; Fukuyama, T. Chem. Pharm. Bull.
2000, 48, 1570.
6. Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F.
Pharmacol. Rev. 1999, 51, 7.
7. Igarashi, K.; Shirahata, A.; Park, A. J.; Kashiwagi,
K.; Williams, K. J. Pharmacol. Exp. Ther. 1997, 283,
533.
In summary, we have succeeded in developing a new class
of NMDA channel blockers having a conjugate structure
that consists of a nitrogenous heterocyclic head and a tet-
raamine tail. Some of the compounds have potent antag-
onistic activities compared with the lead compound,
AQ343, or memantine. Further, several compounds have
appropriate biological profiles, i.e., weak cytotoxicity,
for the development of therapeutic agents for neurologi-
cal disorders. Investigations, including in vivo experi-
ments using these compounds, are under way.
8. Nishimura, K.; Murozumi, K.; Shirahata, A.; Park, M. H.;
Kashiwagi, K.; Igarashi, K. Biochem. J. 2005, 385, 779.