Synthesis and structure-activity relationships of γ-carboline derivatives as potent and selective cysLT1 antagonists

Article abstract of DOI:10.1016/j.bmcl.2009.05.094

A γ-carboline series of cysLT₁ receptor antagonists has been prepared. Some of the compounds show good potencies both, in vitro and in vivo, compared to the standard compounds.

Full text of DOI:10.1016/j.bmcl.2009.05.094

Bioorganic & Medicinal Chemistry Letters 19 (2009) 4299–4302
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and structure–activity relationships of
as potent and selective cysLT1 antagonists
c-carboline derivatives
Josep Bonjoch a, Faïza Diaba a, Lluís Pagès b, Daniel Pérez c, Lidia Soca c, Montserrat Miralpeix b,
Dolors Vilella b, Paquita Anton b, Carles Puig c,
*
a Laboratori de Química Orgànica, Facultat de Farmàcia, Institut de Biomedicina (IBUB), Universitat de Barcelona, Av. Joan XXIII s/n, 08028-Barcelona, Spain
b Almirall SA, Research Center, Laureà Miró 408-410, 08980 St Feliu de Llobregat, Barcelona, Spain
c Almirall SA, Medicinal Chemistry, Treball, 2-4, 08960 St Just Desvern, Barcelona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
A
c
-carboline series of cysLT1 receptor antagonists has been prepared. Some of the compounds show
good potencies both, in vitro and in vivo, compared to the standard compounds.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 27 April 2009
Revised 18 May 2009
Accepted 20 May 2009
Available online 27 May 2009
Keywords:
Cys-LT1 antagonists
c
-Carboline
Carboxylic derivatives
Anti-inflammatory compounds
Asthma is one of the most rapidly growing therapeutic markets,
the disease affecting over 300 million people worldwide.1 Cys-
teinyl leukotrienes (LTC4, LTD4 and LTE4) are products of the 5-
lipoxygenase pathway of arachidonic acid metabolism and play a
crucial role in asthma pathophysiology by causing bronchocon-
striction, mucus production and an increase in vascular permeabil-
ity.2 They represent one of the most effective approaches to the
treatment of asthma3 and several compounds with this mechanism
of action have reached the market.4 In recent years there has been
particular interest in searching for dual H1/cysLT1 antagonists in
the hope of managing asthma by synergistic effects.5
One of the chemical series we have designed in order to achieve
this goal is based on the antiH1 derivative mebhydroline.6 Taking
into account the pharmacophoric model for cysLT1 antagonists,7
we expected that the introduction of a quinoline-type substituent
in the benzyl group of mebhydroline and an acid group branching
from the piperidine moiety (see Fig. 1) would confer a cysLT1
antagonistic character to the resulting structure (e.g. 1, see Fig. 2).
Although the first compounds synthesized in this series (1 and
related tetrahydro-b-carbolines) lacked the parent anti H1 activity,
the anticysLT1 activities were so interesting as to encourage us to
pursue our efforts in this field.8 In this Letter we describe our stud-
ies in developing a new series of cysLT1 antagonists based on our
first mebhydroline derivative, compound 1.
The basic pathway to the target compounds is depicted in
Scheme 1. Starting from the phenylhydrazines 2 the corresponding
tetrahydrocarbazoles 3 were prepared by Fischer indolization of 1-
benzyloxycarbonyl-4-piperidone. Compounds 3 were alkylated at
the indole nitrogen and then the BOC group was removed in acidic
conditions. After alkylation of the carboline nitrogen in 4 and sub-
sequent hydrolysis (or reaction of the corresponding nitrile with
tributyltin azide) the target carbolines 1 and 5–30 were obtained.9
The hexahydrocarboline derivative 33 were prepared through a
similar synthetic pathway from the corresponding compounds
32, which were in turn synthesized by cyanoborohydride reduction
of the indole compound 3 (R1 = H).
The indole derivatives 35 and 36 (Scheme 2) were obtained by
alkylation and subsequent saponification of the butyric esters 34,
which are commercially available (when R = H) or easily prepared
from phenylhydrazine and ethyl 6-oxoheptanoate (when R = Me).
Me
Linker
Linker
N
N
Acid group
N
N
N
H1/cysLT1 antagonist
Figure 1.
Mebhydroline
* Corresponding author. Tel.: +34 932913585; fax: +34 933128635.
E-mail address: carlos.puig@almirall.com (C. Puig).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.05.094

4300
J. Bonjoch et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4299–4302
The open chain amino derivatives 37 and 38 were prepared from
tryptamine and indole, respectively, following the same reaction
procedure as above (Scheme 2).
The carbazole 39 was prepared through a synthetic pathway
involving Fries-type and Willgerodt–Kindler transpositions, whilst
the tetrahydrocarbazole 40 was synthesised again through a
Fischer-indole cyclization. The b-carboline derivatives 41 and 42
were prepared from the commercially available tetrahydro-b-carb-
oline (Scheme 3).
All synthesized compounds were tested in a binding assay in
guinea pig lung using [3H]LTD4 as a radioligand,10 and in the inhi-
bition of LTD4-induced airway microvascular permeability at 4 h,
also in guinea pigs.11 First of all, the influence of the carboxylic side
acid 1. By inserting a double C@C bond, a cyclopropyl or a phenyl
group the affinity improved significantly but not the oral activity.
The nature of the linker between the phenyl spacer and the
lipophilic group was also examined (Table 2). Substitutions of the
vinyl bridge for methoxy (13), ethylene (15), or acetylene (16) or
inclusion in a benzofuran ring (14) afforded compounds with
poorer oral activities in all cases. Similar results were obtained
when changing the position of the nitrogen atom to a b-carboline,
modifying the acid group to a tetrazole, saturating the system to
indoline or ‘opening’ the carboline system (Table 3), in spite of
the improvement in some affinity values. The role of the carboline
nitrogen atom was also assessed. In all the analogs lacking it—open
or cyclic derivatives—the oral activities decreased in spite of the
improvement in affinity (Table 4). The improvement of oral activity
chain R in c-carbolines (1, 5–12) was assessed (Table 1). All com-
pounds tested showed moderate to good in vitro activity, while
related to the x-aminoalkyl carboxylic chain motif has also been
the best in vivo compound was our first derivative, the propionic
observed with many of the 2nd generation antiH1 derivatives. In
this type of compounds the presence of a carboxylic chain attached
to the nitrogen atom keeps or sometimes enhances oral activity
compared to the parent amine compound, even though in some
cases the H1 affinity decreases.12 The next point of variation con-
cerns the lipophilic moiety, for which a number of substituted
quinolines and other heterocycles were tested (Table 5). In this
case a significant improvement was achieved with the 6,7-diflu-
oroquinoline derivative, a compound that shows an ED50 value
for inhibition of LTD4 extravasation of 0.04 mg/kg. Finally, a few
substitutions at the carboline phenyl system were tested. A meth-
oxy group did not improve oral activity on parent compound 1, but
CO2H
N
N
Cl
N
1
Figure 2.
CO2Et
R3
H
Boc
N
N
N
N
d,e
a,b,c
f,g
+
N
N
H
N
NHNH2
R1
R1
R1
R
R2
R2
O
2
3
4
1, 5-30
CO2H
N
N
Boc
Boc
h
N
N
e,f,g
i
3
N
N
H
N
Cl
R2
31
32
33
Scheme 1. Reagents and conditions: (a) AcOH, rfx, 3 h (88%); (b) KOH/IPA, rfx, 16 h (87%); (c) (Boc)2O, THF, 5 °C to rt (98%); (d) NaH, DMF; then R2X (69–95%); (e) TFA, CH2Cl2,
1 h rt or aq HCl, EtOH, rfx, 1 h (69–97%); (f) ethyl acrylate or acrylonitrile, EtOH, rfx 1 h or R3X, K2CO3, MIK, rfx, 16 h; (g) 5 N NaOH, EtOH or LiOH, THF, H2O, 1 h, rt or SnBu3N3,
110 °C, 3 h (28–97% for the two steps); (h) NaCNBH3, AcOH, rt (30%); (i) (E)-7-chloro-2-[3-(chloromethyl)styryl]quinoline, Et3N, DMF, 60 °C (40%).
indole
tryptamine
c,d
e
(CH2)3CO2R
Me
N
CO2Et
CO2Et
N
N
H
R
Me
N
H
34
N
a,b
a,b
H
a,b
Me
N
CO2H
CO2H
N
CO2H
N
N
N
R
Me
N
Cl
N
Cl
Cl
N
35, R = H
36, R = Me
37
38
Scheme 2. Reagents and conditions: (a) NaH/DMF, then (E)-7-chloro-2-(3-(chloromethyl)styryl)quinoline; (b) NaOH 5 N/EtOH or LiOH/THF/H2O, 1 h, rt (13–86% overall
yield); (c) ethyl acrylate, EtOH, rfx, 1 h (quantitative); (d) MeI, K2CO3, CH3CN, rt, 5 h (51%); (e) ethyl 3-methylaminopropanoate/HCHO, MeOH, 60 °C, 5 h (36%).

J. Bonjoch et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4299–4302
4301
O
CO2H
CO2H
a
b
c,d
N
N
H
N
N
H
Cl
N
O
39
CO2H
CO2Et
N
CO2Et
d
e
+
Cl
N
N
H
NHNH2
O
40
CO2H
m
d
f
N
N
N
NH
N
H
(CH2)nCO2Et
N
H
Cl
N
41 m = 0
42 m = 1
Scheme 3. Reagents and conditions: (a) AlCl3 neat, 115 °C, 2 h; (b) S8, morpholine, rfx, 16 h, then KOH/EtOH, rfx, 5 h (38%); (c) EtOH/HCl rfx, 5 h; (d) NaH, (E)-7-chloro-2-(3-
(chloromethyl)styryl)quinoline; DMF, rt; then NaOH 5 N, EtOH, THF, rt (23–42% overall yield); (e) ethyl (4-oxocyclohexyl)acetate, AcOH, 100 °C, 1 h (32%); (f) ethyl acrylate,
EtOH, rfx, 1 h (quantitative) for 41; Br(CH2)3CO2Et, K2CO3, KI, MIK, 90 °C, 2 h (62%) for 42.
Table 1
Table 2
Binding and oral activities for different acid side-chains of
c
-carbolines (1, 5–12)
Binding and oral activities for different quinoline/spacer linkers of c-carbolines (13–
16)
R
linker
CO2H
N
1
5
6
7
CH2CH2CO2H
CH2CO2H
(CH2)3CO2H
N
R
OCH2
14 CH=C
13
N
N
Cl
(CH2)4CO2H
N
O
Cl
CH2CH2
15
16
8
N
C
C
H2C
CH2CO2H
9
COCH2CO2H
Inhibition of LTD4 extravasationb
a
Compd
LTD4 binding IC50 (nM)
10 (E) CH2CH=CHCO2H
13
14
15
16
42 (13)
14 (4)
25 (1)
24 (0.1); 55 (1)
15 (0.1); 30 (1)
7 (0.1); 19 (1)
0 (0.1); 60 (1)
CO2H
11
R
N
F
1.3 (0.08)
CH2
N
a,b
See footnotes at Table 1.
F
CO2H
CH2CH2
12
N
O
performed at 1 and 8 h for some selected compounds (see results
in Table 7). The assayed compounds showed a sustained duration
of action in vivo, similar to that observed for Zafirlukast or Mont-
elukast, with potencies lying between those observed for these
two standard LTD4 antagonists. The pharmacokinetics of 21 in rat
a
Compd
LTD4 binding IC50 (nM)
Inhibition of LTD4 extravasationb
1
5
6
7
8
9
10
11
12
16 (6)
17 (6)
30 (12)
16 (6)
3.2 (1.1)
26 (11)
4.1 (1.5)
3.0 (1.8)
0.91 (0.43)
72 (0.1); 100 (1)
22 (0.1); 43 (1)
25 (0.1); 80 (1)
13 (0.1); 77 (1)
20 (0.1); 64 (1)
1 (0.1); 26 (1)
22 (0.1); 52 (1)
34 (0.1); 79 (1)
0 (0.1); 46 (1)
Table 3
Binding and oral activities for miscellaneous compounds (17, 33, 37, 38, 41, 42)
a
Compd
LTD4 binding IC50 (nM)
Inhibition of LTD4 extravasationb
a
Values are means of three experiments, standard deviation is given in
41
42
17c
33
37
38
20 (5.4)
15 (3.4)
1.2 (1.7)
12 (4.5)
9.3 (2.7)
61 (19)
10 (0.1); 43 (1)
16 (0.1); 39 (1)
3 (0.1); 14 (1)
22 (0.1); 66 (1)
55 (0.1); 67 (1)
—
parentheses.
b
% Inhib (dose mg/kg).
a fluorine atom at 9 position on parent difluoro derivative 21 did
(Table 6).
In order to assess the duration of action of this type of com-
pounds, the inhibition of LTD4 extravasation in guinea-pig was also
a,b
See footnotes at Table 1.
c
Compound 17 has a structure as 1 exchanging the carboxyl by a 5-tetrazolyl
group.

4302
J. Bonjoch et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4299–4302
Table 4
Table 7
Binding and oral activities for compounds lacking the carboline N-atom (35, 36, 39,
40)
Duration of action in vivo of selected compounds
Compound
Oral activitya
4 h
a
Compd
LTD4 binding IC50 (nM)
Inhibition of LTD4 extravasationb
1 h
8 h
39
40
35
36
4.8 (1.4)
4.9 (2.1)
4.4 (2.8)
3.4 (3)
29 (0.1); 52 (1)
14 (0.1); 66 (1)
24 (0.1); 31 (1)
0 (0.1); 15 (1)
1
0.06
0.02
0.014
0.14
0.26
0.027
0.041
0.04
0.02
0.43
2.0
0.075
0.03
0.05
0.39
1.0
21
30
Ab
Bb
Cb
a,b
See footnotes at Table 1.
0.008
0.006
a
Inhibition of LTD4 extravasion: ED50, mg/kg.
A:{3-[2-Methoxy-4-toluene-2-sulfonylaminocarbonyl)benzyl]-1-methyl-1H-
b
Table 5
Binding and oral activities for different lipophilic moieties (1, 18–24)
indol-5-yl}carbamic acid cyclopentyl ester (zafirlukast); B: 8-[4-(4-phenylbu-
tyloxy)benzoyl]amino-2-(tetrazol-5-yl)-4-oxo-4H-1-benzopyran (pranlukast); C:
[1-(1-{3-[2-(7-chloroquinolin-2-yl)-vinyl]-phenyl}-3-[2-(1-hydroxy-1-methyl-ethyl)-
phenyl]-propylsulfanylmethyl)-cyclopropyl]-acetic acid (montelukast).
Het
CO2H
N
R1 R2
N
R1
R2
1
18
H
H
Cl
F
Het
19 Cl Cl
Acknowledgments
N
20
21
F
F
Cl
F
We would like to thank D. Pérez, F. Biosca, I. Pagan, R. Ortiz, J.
Prieto and E. Villanova for their excellent work.
N
S
N
S
tBu
Cl
S
Cl
N
Cl
References and notes
22
23
24
a
1. Braman, S. S. Chest 2006, 130, 4S.
Compd
LTD4 binding IC50 (nM)
Inhibition of LTD4 extravasationb
2. (a) Bailey, J. M. Prostaglandins, Leukotrienes, and Lipoxins: Biochemistry,
Mechanism of Action, and Clinical Applications; Plenum Press: New York, 1985;
(b) Frank, A. K. Nat. Immunol. 2008, 9, 113.
3. Busse, W.; Kraft, M. Chest 2005, 127, 1312.
4. Rodger, I. W. Am, J. Respir. Crit. Care Med. 2000, 161, S7.
5. (a) Zhang, M.-Q.; Timmerman, H. Inflamm. Res. 1997, 46, 593; (b) Ruck, L. M.;
Rizzo, C. A.; Anthes, J. C.; Eckel, S.; Egan, R. W.; Cuss, F. M.; Hey, J. A. Life Sci.
2001, 68, 2825.
1
16 (6)
72 (0.1); 100 (1)
42 (0.1); 100 (1)
0 (0.1); 86 (1)
58 (0.1); 97 (1)
18
19
20
21
22
23
24
3.4 (2.2)
21 (6.8)
1.7 (0.5)
1.0 (0.4)
37 (9)
48 (0.03); 96 (0.1); 100 (1)
—
—
—
112 (52)
50 (29)
6. Hörlein, U. Chem. Ber. 1954, 87, 463.
7. (a) Zwaagstra, M. E.; Schoenmakers, S. H. H. F.; Nederkoorn, P. H. J.; Gelens,
E.; Timmerman, H.; Zhang, M.-Q. J. Med. Chem. 1998, 41, 1439; (b) Palomer,
A.; Pascual, J.; Cabré, F.; Garcia, M.-L.; Mauleón, D. J. Med. Chem. 2000, 43,
392.
a,b
See footnotes at Table 1.
8. Other indole anti LTD4 series have been described hitherto: (a) Matassa, V. G.;
Maduskuie, T. P.; Shapiro, H. S.; Hesp, B.; Snyder, D. W. J. Med. Chem. 1990, 33,
1781; (b) Boot, J. R.; Bond, A.; Thomas, K. H.; O’Brien, A.; Gilmore, J.; Todd, A.
Eur. J. Pharmacol. 1993, 231, 83; (c) Sawyer, J. S.; Thrasher, K. J.; Bach, N. J.;
Stengel, P. W.; Cockerham, S. L.; Silbaugh, S. A.; Roman, C. R.; Froelich, L. L.;
Fleisch, J. H. Bioorg. Med. Chem. Lett. 1996, 6, 249; (d) Merschaert, A.; Boquel, P.;
Van Hoeck, J.-P.; Gorissen, H.; Borghese, A.; Bonnier, B.; Mockel, A.; Napora, F.
Org. Process Res. Dev. 2006, 10, 776. and references cited therein.
9. In the case of the isomeric fluoro derivatives 28 and 30 the separation was
achieved through chromatography on silica gel of the mixture of the
corresponding amino antecedents 4 using a mixture of CH2Cl2/MeOH/NH4OH
(40:2.5:0.1).
Table 6
Binding and oral activities for some phenyl-substituted c-carbolines (1, 21, 25–30)
R1
R2
9
CO2H
8
1
H
7«-Cl
N
R1
R2
6«
7«
25 6-OEt 7«-Cl
26 7-OMe 7«-Cl
27 8-OMe 7«-Cl
7
N
6
N
21
H
6«,7«-diF
28 7-F 6«,7«-diF
29 8-F 6«,7«-diF
30 9-F 6«,7«-diF
10. The assays were performed using guinea-pig lung membranes as a source of
receptors and [3H]LTD4 from NEN in a modification of the previously described
method (Aharony D.; Falcone, R. C.; Krell, R. D. J. Pharmacol. Exp. Ther. 1987,
Compd
Inhibition of LTD4 extravasationb
a
243, 921). The assay mixture contained 200
final volume of 10 mM PIPES pH 7.5 containing 10 mM CaCl2, 50 mM NaCl,
2 mM
-cysteine, 2 mM Glycine and 300 pM [3H]LTD4. Non specific binding
was determined in the presence of zafirlukast 10 M. The assays were directly
performed on GF/B Millipore Multiscreen 96 well plates, pre-soaked with
200 l/well of assay buffer, filtered, washed three times with 175 l of 10 mM
lg of membranes per assay in a
LTD4 binding IC50 (nM)
1
16 (6)
150 (74)
16 (5)
7.5 (2,5)
1.0 (0.1)
4.5 (3.4)
2.5 (0.6)
1.2 (0.8)
72 (0.1); 100 (1)
L
25
26
27
21
28
29
30
—
l
29 (0.1); 53 (1)
45 (0.1); 74 (1)
l
l
TRIS buffer pH 7.5 containing 100 mM NaCl at 4 °C, dried and read in a TRILUX
Microbeta Counter of Wallac.
0.04
0.05
0.07
0.02
11. Male Dunkin–Hartley guinea-pigs were administered with the test compounds
by oral gavage at the indicated time-points before being anesthetized. Then,
the left jugular vein was cannulated and the animals received the Evans blue
dye intravenously. After 5 min, LTD4 was administered to the animals in order
to induce airway microvascular leakage. After another period of 5 min, animals
were exsanguinated and the vascular bed was rinsed. Then the trachea was
excised and incubated in formamide for 20 h at 55 °C to extract the Evans blue
dye from the tissue. Microvascular permeability was determined by light
spectrophotometry at 620 nm. The number of animals was 6 per dose or 8 in
the case of the complete dose–response curve.
a
See footnotes at Table 1.
% Inhib (dose mg/kg) or ED50, mg/kg.
b
shows a 47% of bioavailability, an intermediate value between the
63% for montelukast and the 25% for zafirlukast.
In conclusion, we have developed a SAR around the tetrahyd-
rocarbazole scaffold, which led to the selection of compound 21.
Further studies will provide a better understanding of its long term
efficacy and potential for human treatment.
12. Inter alia, see: (a) Iwasaki, N.; Sakaguchi, J.; Ohashi, T.; Takahara, E.; Ogawa, N.;
Yasuda, S.; Koshinaka, E.; Kato, H.; Ito, Y.; Sawanishi, H. Chem. Pharm. Bull.
1994, 42, 2276; (b) Iwasaki, N.; Sakaguchi, J.; Ohashi, T.; Yamazaki, M.; Ogawa,
N.; Yasuda, S.; Koshinaka, E.; Kato, H.; Ito, Y.; Sawanishi, H. Chem. Pharm. Bull.
1994, 42, 2285.