Synthesis and biological evaluation of novel T-type calcium channel blockers

Article abstract of DOI:10.1016/j.bmcl.2006.10.024

3,4-Dihydroquinazoline analogues substituted by N-methyl-N-(5-pyrrolidinopentyl)amine at the 2-position were synthesized and their blocking effects were evaluated for T- and N-type calcium channels. Compound 11b (KYS05080), compared to mibefradil (IC₅₀ = 1.34 ± 0.49 μM), was about 5-fold potent (IC₅₀ = 0.26 ± 0.01 μM) for T-type calcium channel (α_1G) blocking and its selectivity of T/N-type was also improved (7.5 versus 1.4 of mibefradil).

Full text of DOI:10.1016/j.bmcl.2006.10.024

Bioorganic & Medicinal Chemistry Letters 17 (2007) 471–475
Synthesis and biological evaluation of novel T-type calcium
channel blockers
Ja Youn Choi,a Han Na Seo,a Min Joo Lee,a Seong Jun Park,b,c Sung Jun Park,a
Ji Young Jeon,a Joo Hi Kang,a Ae Nim Pae,b Hyewhon Rhimb and Jae Yeol Leea,*
aResearch Institute for Basic Sciences and Department of Chemistry, College of Sciences, Kyung Hee University, 1 Hoegi-Dong,
Seoul 130-701, Republic of Korea
bLife Sciences Division, Korea Institute of Science and Technology, PO Box 131, Cheongryang, Seoul 130-650, Republic of Korea
cDepartment of Chemistry, Korea University, 1-Anamdong, Seoul 136-701, Republic of Korea
Received 9 July 2006; revised 25 September 2006; accepted 10 October 2006
Available online 12 October 2006
Abstract—3,4-Dihydroquinazoline analogues substituted by N-methyl-N-(5-pyrrolidinopentyl)amine at the 2-position were synthe-
sized and their blocking effects were evaluated for T- and N-type calcium channels. Compound 11b (KYS05080), compared to
mibefradil (IC50 = 1.34 0.49 lM), was about 5-fold potent (IC50 = 0.26 0.01 lM) for T-type calcium channel (a1G) blocking
and its selectivity of T/N-type was also improved (7.5 versus 1.4 of mibefradil).
Ó 2006 Elsevier Ltd. All rights reserved.
T-type (low-threshold) calcium channels are known to
be implicated in pathogenesis of epilepsy and neuro-
pathic pain.1,2 In spite of this, the pharmacology of
T-type channels is complex because many drugs have
been found to block T-type currents.3 Unfortunately,
none of these compounds has high selectivity for these
channels. Therefore it is difficult to isolate T-type cur-
rent pharmacologically. Mibefradil has been marketed
worldwide for the treatment of hypertension and angina
for a short period before it was withdrawn due to drug–
drug interactions (Fig. 1). Mibefradil binds to skeletal
muscle L-type calcium channels and brain voltage-gated
sodium channels.4,5 It also can block potassium and
chloride channels. Obviously, this makes it not an ideal
tool for in vitro or in vivo studies on T-type channels.
Studies of calcium channels expressed in oocytes have
identified kurtoxin as a promising tool for functional
and structural studies of low-threshold T-type calcium
channels.6 This peptide, isolated from the venomous
scorpion Parabuthus transvaalicus, inhibits expressed
low-threshold Cav3.1 (a1G) and Cav3.2 (a1H) calcium
channels with relatively high potency and high selectivi-
ty. In thalamic neurons, however, kurtoxin blocks the
composite high-threshold calcium channel to about
50% as well as T-type calcium channel almost complete-
ly in these cells.7 Therefore, only limited progress has
been made to date in the quest to identify both potent
and selective compounds except kurtoxin and mibefradil
for T-type channel blockade.
In previous studies, we have tried to identify new com-
pounds with higher potency and selectivity for T-type
channel and reported some lead-like compounds contain-
ing 3,4-dihydroquinazoline ring as novel scaffold.8–10
Among them, KYS05044 and KYS05050 were found to
be superior to mibefradil with respect to both potency
and selectivity for T-type channel (Fig. 1).9
However, both compounds have an innate synthetic
problem for the further development as clinically appli-
cable drug due to the difficulty in separating two regioi-
somers (iii and iv) stemmed from a primary amine in the
synthetic procedure (Fig. 2).9 In order to obtain a de-
sired compound (iii) among two regioisomers, therefore,
an attacking nucleophile should be a secondary amine.
On the other hand, a computational mapping of com-
pound KYS05050 (most active among this series com-
pounds) into 6-feature 3D pharmacophore hypothesis,
which was generated from our previous results,11
hypothesized that its blocking activity would increase
when one phenyl group is removed from the parent
Keywords: T-type calcium channel; Channel blockers; 3,4-Dihydro-
quinazoline; Regioisomers.
*
Corresponding author. Tel.: +82 2 961 0726; fax: +82 2 966 3701;
e-mail: ljy@khu.ac.kr
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.024

472
J. Y. Choi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 471–475
O
O
OCH3
H3C
CH3
O
R2
H
R1
R1
N
O
N
N
N
N
CH3
F
N
N
NH2
N
N
N
H
CH3
Mibefradil
KYS05044: R1 = OCH3
Target compounds
R
R
KYS05050: R1 = NHCH2Ph
1 = OCH3 or NHCH2Ph
2 = H or Ph
Figure 1. Mibefradil and 3,4-dihydroquinazoline derivatives.
O
H3CO
Ra
N
A-pathway
B-pathway
N
NHRb
O
O
A
iii
NH2Rb
OCH3
OCH3
N
C
N Ra
N C
NHRa
B
NHRb
O
i
ii
H3CO
Rb
N
N
NHRa
iv
Figure 2. Formation of two regioisomers due to a primary amine nucleophile.
molecule and/or a hydrophobic residue is introduced
into its terminal amine group (Fig. 3). Such implication
prompted us to design the subsequent modification of
compounds KYS05050. Herein, we report on the synthe-
sis and brief structure–activity relationship (SAR) of
target compounds as KYS05050 analogues (Fig. 1).
The in vitro calcium channel blocking activities of 3,4-
dihydroquinazoline derivatives (9a–b and 11a–b) were
determined in T-type channels stably expressed in
Xenopus oocytes (a1H) and HEK293 cells (a1G). First,
all compounds were evaluated for their blocking effects
on a1H T-type calcium channels expressed in Xenopus
oocytes by a two-electrode voltage clamp method at
100 lM concentration.16 They were again re-evaluated
for the blocking effects on a1G T-type calcium channels
expressed in HEK293 cells at 10 lM concentration by
whole-cell patch clamp methods.17 The molar concen-
trations (IC50) of test compounds required to produce
50% inhibition of a1G T-type currents were determined
from fitting raw data into dose–response curves. Their
in vitro blocking data are summarized in Table 1 and
the data of parent compound (KYS05044 and
KYS05050) were also inserted for comparison. New syn-
thetic compounds showed high blocking activities
against a1H T-type calcium channel (Xenopus oocyte) ex-
cept for compound 9a (ca. 48%), which has a mono-
phenyl group at 3-position and methyl ester at 4-posi-
tion of 3,4-dihydroquinazoline ring. Among them, com-
pound 11b possessing a biphenyl group at 3-position
and a benzyl amide group at 4-position was found to
be most active (96%) and also comparable to the parent
compound KYS05044 (97%). Against a1G T-type calci-
um channel (HEK293 cell), additionally, compound 9a
and 11a containing a simple phenyl group at 3-position
The key intermediate, N-methyl-N-(5-pyrrolidinopen-
tyl)amine (6), was prepared starting from 1,5-dibrom-
opentane 1 (Scheme 1). Compound 1 was treated with
potassium phthalimide to afford mono-phthalimide
2,12 which was treated with pyrrolidine to provide com-
pound 3 under the basic condition (K2CO3, EtOH).
Deprotection of phthalimide group of compound 3 via
Gabriel condition afforded a free amine 4.13,14 Protec-
tion of amine group of compound 4 with tert-butyloxy-
carbonyl group followed by reduction using lithium
aluminum hydride provided the key intermediate 6 as
a secondary amine.15
This intermediate 6 was coupled with cinnamate-derived
carbodiimide 8 to provide 3,4-dihydroquinazoline deriv-
ative 9 as a single regioisomer via tandem nucleophilic
addition and intramolecular conjugate addition. Finally,
hydrolysis of compound 9 with LiOH followed by cou-
pling reaction of compound 10 with benzylamine, EDC,
and HOBt afforded the target compound 11 by using
our previous procedure (Scheme 2).10

J. Y. Choi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 471–475
473
Figure 3. The mapping of compound KYS05050 into 6-feature 3D pharmacophore hypothesis: blue color is hydrophobic region; red color is positive
ionizable region; green color is hydrogen bonding region.
O
O
N
N
N
Br
i
ii
iii
Br
Br
O
O
1
4
2
3
H3C
Boc
N
N
v
iv
H2N
N
N
H
N
H
5
6
Scheme 1. Reagents and conditions: (i) potassium phthalimide, DMF, rt, 50%; (ii) pyrrolidine, K2CO3, NaI, EtOH, reflux, 35%; (iii) NH2NH2/H2O,
EtOH, reflux, 81%; (iv) (tBOC)2O, CH2Cl2, rt, 86%; (v) LiAlH4, THF, reflux, 91%.
O
R2
O
O
H3CO
i
ii
OCH3
OCH3
N
N
C
N
R2
NO2
N
N
N
CH3
8a: R2 = H
b: R2 = Ph
9a: R2 = H (85%)
7
b: R2 = Ph (91%)
O
O
R2
R2
HO
N
H
iii
iv
N
N
N
N
N
N
N
N
CH3
CH3
10a: R2 = H (quantitative)
b: R2 = Ph (quantitative)
11a: R2 = H (48%)
b: R2 = Ph (52%)
Scheme 2. Reagents and conditions: (i) Ref. 10; (ii) compound 6, benzene, rt; (iii) LiOH, THF–H2O (1:1), 70 °C; (iv) PhCH2NH2, HOBt, EDC,
THF–CH2Cl2 (1:1), rt.

474
J. Y. Choi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 471–475
Table 1. In vitro calcium channel blocking effects of 3,4-dihydroquinazoline derivatives
O
R2
R1
N
N
N
R3
R4
Compound
(library code)
R1
R2 R3
R4
Xenopus oocyte
HEK293 cell (T-type: a1G
)
HEK293 cell
(N-type: a1B)
Selectivity
(T/N-type)
at 10 lM
T-type (a1H
)
% inhibition (100 lM) % Inhibitiona IC50 b(lM) % inhibitiona (10 lM)
(10 lM)
KYS05044
KYS05050
OCH3 Ph
NHBn Ph
H
H
NH2
NH2
97.0
80.1
82.5 0.7
83.8 1.4
0.56 0.10 No blockingc
0.13 0.01 8.3 1.8
>100
10.1
9a (KYS05076)
9b (KYS05079)
OCH3
H
CH3
47.7
88.0
82.7
59.5 1.1
94.9 1.2
68.1 0.5
5.84 0.44 10.7 2.3
0.34 0.04 36.5 0.2
4.20 0.20 46.3 0.3
5.6
2.6
1.5
N
N
N
N
OCH3 Ph CH3
CH3
11a (KYS05077) NHBn
H
11b (KYS05080) NHBn Ph CH3
Mibefradil
95.5
86.0
88.1 1.7
95.9 1.7
0.26 0.01 11.7 5.8
1.34 0.49 67.6 1.2
7.5
1.4
a % inhibition value ( SE) was obtained by repeated procedures (n P 4).
b IC50 value was determined from the dose–response curve.
c ‘No blocking’ means the inhibition was less than 1%.
showed the diminished activity (both of % inhibition
and IC50 values) irrespective of a kind of 4-position
group when compared to parent compounds
(KYS05044 and KYS05050). Compound 9b possessing
a biphenyl group at 3-position and a methyl ester group
at 4-position was most active and comparable to mibe-
fradil (96%). With respect to IC50 values, compound
9b and 11b were more or less active than the parent com-
pounds KYS05044 and KYS05050, but they were found
to be 4- to 5-fold higher than mibefradil. With respect to
the channel selectivity (T/N-type channel), all com-
pounds were less selective than the parent compounds
although they still showed the improved selectivity when
compared to mibefradil. Therefore, these biological data
imply that the biphenyl group may be an essential
pharmacophore for the blocking activity of this series
compounds opposite to the above 6-feature 3D pharma-
cophore hypothesis. Additionally, the introduction of
methyl and/or hydrophobic residue in terminal area
resulted in the decreased channel selectivity although
of less effect on the blocking activity.
Acknowledgment
This study was supported by Vision 21 Program from
Korea Institute of Science and Technology.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.10.024.
References and notes
1. Nelson, M. T.; Todorovic, S. M. Curr. Pharm. Des. 2006,
12, 2189.
2. Perez-Reyes, E. Physiol. Rev. 2003, 83, 117.
3. Lacinova, L. Curr. Drug Targets CNS Neurol. Disord.
2004, 3, 105.
4. Eller, P.; Berjukov, S.; Wanner, S.; Huber, I.; Hering, S.;
Knaus, H. G.; Toth, G.; Kimball, S. D.; Striessnig, J. Br.
J. Pharmacol. 2000, 130, 669.
In summary, new KYS05050 analogues were prepared
for promoting the synthetic efficiency and studying a
brief structure–activity relationship via the molecular
modeling. Among them, compound 11b (KYS05080)
was found to be more active and selective than mibefra-
dil although more or less active and selective than their
parent compounds.18 However, it is conceivable that
compound 11b (KYS05080) can be further modified to
exhibit better blocking activity and selectivity. Further
studies to acquire more information about structure–
activity relationships are in progress in our laboratory.
5. Strege, P. R.; Bernard, C. E.; Ou, Y.; Gibbons, S. J.;
Farrugia, G. Am. J. Physiol. Gastrointest. Liver Physiol.
2005, 289, G2493.
6. Chuang, R. S.; Jaffe, H.; Cribbs, L.; Perez-Reyes, E.;
Swartz, K. J. Nat. Neurosci. 1998, 1, 668.
7. Sidach, S. S.; Mintz, I. M. J. Neurosci. 2002, 22,
2023.
8. Lee, Y. S.; Lee, B. H.; Park, S. J.; Kang, S. B.; Rhim, H.;
Park, J.-Y.; Lee, J.-H.; Jeong, S.-W.; Lee, J. Y. Bioorg.
Med. Chem. Lett. 2004, 14, 3379.
9. Rhim, H.; Lee, Y. S.; Park, S. J.; Chung, B. Y.; Lee, J. Y.
Bioorg. Med. Chem. Lett. 2005, 15, 283.

J. Y. Choi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 471–475
475
10. Park, S. J.; Park, S. J.; Lee, M. J.; Rhim, H.; Kim, Y.; Lee,
J.-H.; Chung, B. Y.; Lee, J. Y. Bioorg. Med. Chem. 2006,
14, 3502.
11. Doddareddy, M. R.; Jung, H. K.; Lee, J. Y.; Lee, Y. S.;
Cho, Y. S.; Koh, H. Y.; Pae, A. N. Bioorg. Med. Chem.
2004, 12, 1605.
CDCl3) 7.56–7.39 (6H, m, Ph), 7.34–7.26 (1H, m, Ph),
7.23–7.14 (4H, m, Ph), 7.00–6.88 (2H, m, Ph), 5.16 (1H,
dd, J = 10.8 and 4.5 Hz, –CO–CH2–CH–N–), 3.79 (3H,
s, CH3O–), 3.58 (1H, br s, –NH–CH2–C4H8–), 3.16 (1H,
br s, –NH–CH2–C4H8–), 2.93–2.85 (4H, m, –CO–CH2–
CH–N–, –NCH3–C5H10–), 2.58–2.41 (7H, m, –CO–CH2–
CH–N–, –C4H8–CH2-pyrrolidinyl-H2, H5), 1.80–1.79
(4H, m, –pyrrolidinyl-H3, H4), 1.57–1.53 (4H, m,
12. Jourdain, F.; Caron, M.; Pommelet, J. C. Synth. Commun.
1999, 29, 1785.
13. Robarge, M. J.; Husbands, S. M.; Kieltyka, A.; Brodbeck, R.;
Thurkauf, A.; Newman, A. H. J. Med. Chem. 2001, 44, 3175.
14. Turner, S. C.; Esbenshade, T. A.; Bennani, Y. L.;
Hancock, A. A. Bioorg. Med. Chem. Lett. 2003, 13, 2131.
15. Phuan, P.-W.; Kozlowski, M. C. J. Org. Chem. 2002, 67,
6339.
–CH2–CH2–CH2–CH2–CH2–),
1.29–1.28
(2H,
m,
–C2H4–CH2–C2H4–); 13C NMR (75 MHz, CDCl3) d
172.2, 153.6, 145.7, 144.5, 140.6, 137.0, 129.0, 128.7,
128.1, 127.3, 127.0, 125.7, 125.0, 123.2, 122.9, 122.4,
61.5, 56.8, 54.5, 52.2, 50.2, 39.8, 35.7, 29.0, 27.6, 25.4,
23,7. Compound 11b (KYS05080): 1H NMR (300 MHz,
CDCl3) 7.58–7.41 (6H, m, Ph), 7.36–7.11 (10H, m, Ph),
7.02–6.89 (2H, m, Ph), 5.33 (1H, dd, J = 9.6 and 5.1 Hz,
–CO–CH2–CH–N–), 4.51 (2H, dd, J = 5.4 and 3.3 Hz,
Ph–CH2–NH–), 3.36 (1H, br s), 2.76–2.56 (10H, m,
–NCH3–CH2–C4H8–, –NCH3–CH2–, –C4H8–CH2-pyrro-
lidinyl-H2, H5, –CO–CH2–CH–N–), 2.47 (1H, dd,
J = 14.4 and 5.1 Hz, –CO–CH2–CH–N–), 1.80 (4H, br
s, –pyrrolidinyl-H3, H4), 1.52 (4H, br s, –CH2–CH2–
CH2–CH2–CH2–), 1.35–1.29 (2H, m, –C2H4–CH2–C2H4–
); 13C NMR (75 MHz, CDCl3) d 170.4, 154.2, 145.7,
144.2, 140.7, 138.7, 136.9, 129.0, 128.8, 128.3, 128.1,
127.6, 127.3, 127.0, 126.5, 125.4, 123.1, 122.6, 122.4,
61.4, 56.6, 54.3, 43.9, 41.9, 35.4, 30.0, 28.1, 27.2, 24.9,
23.5.
16. Lee, J.-H.; Gomora, J. C.; Cribbs, L. L.; Perez-Reyes, E.
J. Biophys. 1999, 77, 3034.
17. Monteil, A.; Chemin, J.; Bourinet, E.; Mennessier, G.;
Lory, P.; Nargeot, J. J. Biol. Chem. 2000, 275, 6090.
18. Spectral data of selected compounds. Compound 6: 1H
NMR (300 MHz, CDCl3) 2.41 (2H, t, J = 6.9 Hz, CH3–
N–CH2–C4H8–), 2.34–2.24 (9H, m, CH3–N–CH2–C4H8–,
–C4H8–CH2-pyrrolidinyl-H2, H5), 1.63–1.59 (4H, m,
pyrrolidinyl-H3, H4), 1.41–1.31 (4H, m, –N–CH2–CH2–
CH2–CH2–CH2–), 1.27–1.17 (2H, m, –N–C2H4–CH2–
C2H4–); 13C NMR (75 MHz, CDCl3) d 56.7, 54.3, 52.2,
36.7, 30.0, 29.2, 25.5, 23.5; MS(GC) m/z (relative
intensity, %) 170 (M+, 16), 140(23), 110(31), 84(100),
1
55(42). Compound 9b (KYS05079): H NMR (300 MHz,