5-Lipoxygenase inhibitors with histamine H1 receptor antagonist activity

Article abstract of DOI:10.1016/j.bmcl.2004.02.005

A series of novel compounds with both 5-lipoxygenase (5-LO) inhibitory and histamine H₁ receptor antagonist activity were designed for the treatment of asthma. These dual-function compounds were made by connecting 5-LO and H₁ pharmacophores,N-hydroxyureas and benzhydryl piperazines, respectively. A range of in vitro activities was observed, with the furan analog 10 demonstrating both activities in an animal model. The activities observed were compared to single-function drugs.

Full text of DOI:10.1016/j.bmcl.2004.02.005

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2265–2268
5-Lipoxygenase inhibitors with histamine H1 receptor
antagonist activity
Timothy A. Lewis,* Lynn Bayless, Joseph B. Eckman, James L. Ellis, Gurmit Grewal,
Lyn Libertine, Jean Marie Nicolas, Ralph T. Scannell, Bruce F. Wels,
Karen Wenberg and Donna M. Wypij
UCB Research, 840 Memorial Dr., Cambridge MA 02139, USA
Received 4 December 2003; accepted 2 February 2004
Abstract—A series of novel compounds with both 5-lipoxygenase (5-LO) inhibitory and histamine H1 receptor antagonist activity
were designed for the treatment of asthma. These dual-function compounds were made by connecting 5-LO and H1 pharmaco-
phores, N-hydroxyureas and benzhydryl piperazines, respectively. A range of in vitro activities was observed, with the furan analog
10 demonstrating both activities in an animal model. The activities observed were compared to single-function drugs.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Asthma is a complex respiratory disease involving some
50 mediators.1 Targeting more than one of these medi-
ators may provide greater benefits for asthmatics.
Clinical data in asthmatics demonstrated that those
patients treated with a combination of an LTD4 recep-
tor antagonist (zafirlukast) and an antihistamine H1
receptor antagonist (loratidine) responded better than
those patients treated with a single drug.2 Similarly, the
combination of montelukast (LTD4 receptor antagonist)
and cetirizine (H1 receptor antagonist) has proven
equally effective as corticosteroids for asthmatics.3
nation of antihistamines and 5-LO inhibitors, though
mizolastine, an antihistamine, is reported to demon-
strate 5-LO inhibition in vitro.6 Our goal was instead to
discover compounds acting as both 5-lipoxygenase
inhibitors and as H1 receptor antagonists.7;8
Several groups have investigated the use of N-hydroxy-
ureas as 5-LO inhibitors, working to improve the
potency and pharmacokinetics of zileuton. Abbott
studied a number of furans and thiophenes such as ABT-
761,9 while others focused on tetrahydrofurans such as
CMI-977.10
Leukotriene D4 is one of the several peptidoleukotri-
enes, which are potent, pro-inflammatory mediators
responsible for bronchoconstriction during an asthmatic
attack.4 Biosynthesis of the leukotrienes begins with the
oxidation of arachidonic acid by 5-lipoxygenase (5-LO);
hence 5-LO inhibitors (e.g., zileuton) have been inves-
tigated for the treatment of asthma.5 To our knowledge,
no clinical studies have been performed with a combi-
Our strategy to simultaneously target both an intracel-
lular enzyme (5-LO) and a cell surface receptor (H1) was
to link together two distinct pharmacophores. To
maintain 5-LO activity, we retained the ring-butynyl-
hydroxyurea portion of the above 5-LO inhibitors. To
incorporate the H1 receptor antagonist activity, the
fluorophenyl group was removed from the above
O
OH
N
F
NH2
N
NH2
S
O
OH
O
O
F
CMI-977
ABT-761
* Corresponding author. Tel.: +1-617-547-0033x262; fax: +1-617-547-8481; e-mail: timothy.lewis@ucb-group.com
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.005

2266
T. A. Lewis et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2265–2268
structures and replaced with benzhydryl piperazines, the
H1 pharmacophore of the antihistamines cetirizine11 and
efletirizine.12
Dual-function furans were synthesized as in Scheme 1.
Conversion of either terminal butynol to the corre-
sponding N-hydroxyureas 113 and 2 was performed by
treating the alcohols with N,O-(bisphenoxycar-
bonyl)hydroxylamine 314 under Mitsunobu conditions,
followed by treatment with ammonia. Palladium cata-
lyzed coupling of 1 and 2 to 5-bromo-2-furaldehyde
gave aldehydes 4 and 5, respectively. Aldehyde 6, with
no alkyne, was prepared by reacting 3 with 5-(hydroxy-
methyl)furfural, followed by treatment with ammonia.
Reductive aminations of 4, 5, and 6 were performed
with either benzhydryl piperazine 7 or 8. Enantiomeri-
cally pure 7 has the same stereochemistry as the more
potent isomer of cetirizine.11 Compound 9 was tested as
a mixture of diastereomers.15
Any number of dual-function hybrid compounds are
feasible. Carboxylates were not included in our hybrid
structures for two reasons: as 5-LO is an intracellular
enzyme, N-hydroxyureas need a certain degree of lipo-
philicity to penetrate cells and get to the active site.
Secondly, the carboxylate moiety is not necessary for H1
activity, as illustrated by the potent antihistamine
hydroxyzine.11
R
cetirizine: R = Cl, R' = H, X = COOH
Tetrahydrofuran synthesis began with commercially
available, optically pure 5-(S)-hydroxymethyl-c-bu-
tyrolactone (Scheme 2). The alcohol was converted to
the chloride with thionyl chloride, and the lactone
reduced to the lactol. Condensation with benzenesulfinic
O
X
efletirizine: R = R' = F, X = COOH
N
N
hydroxyzine: R = Cl, R' = H, X = CH2OH
R'
O
O
O
a
b
O
HN
O
OPh
OPh
N
NH2
N
NH2
NH2
O
+
OH
H
OH
OH
O
1
4
5
3
OH
N
OH
N
NH2
O
b
a
OH
O
3
+
H
O
O
2
O
O
O
O
a
H
OH
O
3
+
H
N
NH2
O
6
OH
Cl
N
OH
NH2
O
Cl
N
N
c
N
N
N
NH
+ 4
9
OH
7
N
NH2
O
Cl
Cl
O
c
7 + 5
10
O
O
N
N
N
NH2
c
c
N
N
OH
7 + 6
+ 5
11
F
F
OH
N
NH2
F
O
O
N
NH
12
8
F
Scheme 1. (a) 1. PPh3, DIAD/THF, 2. NH3/water/THF (10–50%); (b) PdCl2(CH3CN)2, CuI, PPh3, Et3N/DMF (25–67%); (c) NaBH3CN, HOAc/
MeOH (20–67%).

T. A. Lewis et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2265–2268
2267
a
c
b
O
O
O
OH
O
O
Cl
HO
Cl
OTBDMS
OTBDMS
O
d
SO2Ph
O
Cl
Cl
13
O
Cl
Cl
N
e
N
14
OH
N
NH2
O
N
O
f
N
15
OH
N
OH
N
NH2
NH2
O
F
O
Cl
N
O
N
O
N
N
17
16
F
Scheme 2. (a) SOCl2/pyridine (90%); (b) DIBAL-H/CH2Cl2 )78 ꢁC, (84%); (c) PhSO2H, CaCl2/CH2Cl2 (91%); (d) HCCCH2CH2OTBDMS,
EtMgBr, ZnBr2/THF (78%); (e) 7, NaI, K2CO3/DMF/toluene, reflux (36%), column chromatography; (f) 1. TBAF/THF, 2. 3, PPh3, DIAD/THF,
3. NH3/MeOH (36%, three steps).
Table 1. 5-LO and H1 activities of dual-function moleculesa
acid gave the sulfone. The sulfone was displaced with
the acetylenic Grignard reagent derived from TBDMS-
protected 3-butyn-1-ol giving compound 13 as a mixture
of stereoisomers.16 This mixture was treated with amine
7 to give a mixture of cis and trans isomers of 14. These
isomers were separated by column chromatography and
the relative stereochemistry determined by NOE and
COSY 1H NMR (400 MHz) experiments. Previous work
on N-hydroxyureas containing 2,5-disubstituted tetra-
hydrofuran rings consistently demonstrated better 5-LO
activity for the trans isomers than for the cis, hence only
the trans isomer of 14 was carried further.17 The silyl
group was removed with TBAF and the resulting alco-
hol converted to the N-hydroxyurea 15. Compound 13
was also treated with amine 8 and the sequence con-
tinued to give 16. Repeating the entire synthesis with 5-
(R)-hydroxymethyl-c-butyrolactone and amine 7 gave
17, with the opposite (R,R) configuration about the
THF ring.
Compound
H1 binding
(Ki, nM)
5-LO activity
(HWB, IC50, nM)
9
10
190
150
3500 2100 (2)
420 360 (5)
1700
11
12
4
550
170 50 (3)
310 180 (3)
140
15
16
50 20 (5)
660
180
17
Zileuton
Cetirizine
510
873 391 (118)
––
––
14 6 (60)
a Mean SD (n); otherwise n ¼ 1.
The furan series displayed greater variance among the
analogs tested. Structural changes near the N-hydroxy-
urea affected both the 5-LO and the H1 activity of the
molecules. Compound 11 has antihistaminergic activity
similar to cetirizine, but with poor 5-LO activity.
Moving the hydroxyurea moiety further from the furan
ring improves the 5-LO activity with concomitant loss of
H1 activity (11 vs 10 and 12). Compounds 10 and 12,
with the terminal hydroxyurea, both gave increased 5-
LO activity over the a-substituted hydroxyurea 9. The
difluoro analog 12 had the highest H1 binding constant,
but the best 5-LO potency in the furan series. This trend
was seen in the tetrahydrofuran series as well (16 vs 15
Compounds 9–12 and 15–17 were tested in vitro for
both 5-LO inhibitory activity and H1 receptor antago-
nist activity using standard assays (Table 1). 5-LO
activity was tested in a human whole blood (HWB)
assay monitoring the inhibition of LTB4 formation
using zileuton as a reference standard.18 Human H1
receptor binding was performed using CHO-K1 cells
expressing recombinant human H1 receptor with cetiri-
zine as the positive control.19

2268
T. A. Lewis et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2265–2268
3. Wilson, A. M.; Orr, L. C.; Sims, E. J.; Dempsey, O. J.;
Lipworth, B. J. Am. J. Respir. Crit. Care Med. 2000, 162,
1297.
4. Bernstein, P. R.; Bird, T. G.; Brewster, A. G. In BurgerÕs
Medicinal Chemistry and Drug Discovery, 5th ed.; Wolff,
M. E., Ed.; John Wiley & Sons: New York, 1997; Vol. 5,
pp 405–488.
and 17). Compound 15, with the same (S,S) stereo-
chemistry about the THF ring as CMI-977, had better
activities than its diastereoisomer 17. For comparison,
the reported 5-LO IC50Õs in human whole blood for
ABT-761 and for CMI-977 are 160 nM9 and 117 nM,10
respectively.
5. Carter, G. W.; Young, P. R.; Albert, D. H.; Bouska, J.;
Dyer, R.; Bell, R. L.; Summers, J. B.; Brooks, D. W.
J. Pharmacol. Exp. Ther. 1991, 256, 929.
6. Sudo, K.; Nagai, K.; Yamada, N. Jpn. Pharmacol. Ther.
1998, 26(Suppl), 155.
7. For an appraisal of dual acting anti-inflammatory drugs
Bertolini, A.; Ottani, A.; Sandrini, M. Pharmacol. Res.
2001, 44, 437.
8. For other H1/5–LO molecules see Kawasaki, N.; Miya-
taka, H.; Matsumoto, H.; Yamashita, N.; Sakane, T.;
Mizushima, Y.; Satoh, T. Bioorg. Med. Chem. Lett. 1999,
9, 19.
9. Brooks, C. D. W.; Stewart, A. O.; Basha, A.; Bhatia, P.;
Ratajczyk, J. D.; Martin, J. G.; Craig, R. A.; Kolasa, T.;
Bouska, J. B.; Lanni, C.; Harris, R. R.; Malo, P. E.;
Carter, G. W.; Bells, R. L. J. Med. Chem. 1995, 38, 4768.
10. Gurjar, M. K.; Murugaiah, A. M. S.; Padhakrishna, P.;
Ramana, C. V.; Chorghade, M. S. Tetrahedron: Asymme-
try 2003, 14, 1363.
11. Gillard, M.; Van Der Perren, C.; Moguilevsky, N.;
Massingham, R.; Chatelain, P. Mol. Pharmacol. 2002,
61, 391.
12. Leeson, G. A.; Castaner, R. Drugs Future 1997, 22, 626.
13. Thomas, A. V.; Patel, H. H.; Reif, L. A.; Chemburkar, S.
R.; Sawick, D. P.; Shelat, B.; Balmer, M. K.; Patel, R. R.
Org. Process Res. Dev. 1997, 1, 294.
14. Stewart, A. O.; Brooks, D. W. J. Org. Chem. 1992, 57,
5020.
Compounds 10 and 12 were further tested for 5-LO
inhibitory activity using an RBL-2H3 cell lysate 5-LO
assay;5;18 both had IC50Õs <200 nM. Compound 17 was
tested using human recombinant 5-lipoxygenase,20 the
IC50 of 420 nM agrees well with its HWB activity
(510 nM). These results further confirm that the com-
pounds are binding to the lipoxygenase active site.
Analogs were then screened for their in vitro pharma-
cokinetic properties. Compound 10 showed an accept-
able metabolic stability when incubated with NADPH-
fortified rat liver microsomes (Clint 33 lL/min/mg pro-
tein) and a high permeability in Caco-2 cells (Papps:
6 · 10ꢀ6 cm/sec). Because of its overall promising in vitro
profile, compound 10 was selected for testing in 5-LO
and H1 animal models (guinea pig, 2 mg/kg, po, t ¼ 3 h,
n P 4). The 5-LO ex vivo activity was evaluated by
monitoring calcium ionophore-induced LTB4 formation
in whole blood samples,21 46% inhibition was observed.
In the same model, zileuton demonstrated 82% inhibi-
tion at 5 mg/kg. H1 activity of 10 was evaluated by
monitoring the inhibition of histamine-induced bron-
choconstriction using the Konslett–Rossler protocol,22
86% inhibition was observed. Cetirizine displayed
essentially complete inhibition, 96%, at a lower dose
level (0.5 mg/kg). Overall, the oral activity of compound
10 on both targets agrees with its in vitro profile.
15. All compounds tested were characterized by 1H NMR,
mass spectral, and IR analysis; purity was >95% as
determined by HPLC analysis with UV detection at
254 nM and tandem mass spectral detection.
16. Brown, D. S.; Bruno, M.; Davenport, R. J.; Ley, S. V.
Tetrahedron 1989, 45, 4293.
17. Unpublished results, CytoMed Inc.
18. Sirois, P.; Borgeat, P.; Lauziere, M.; Dube, L.; Rubin, P.;
Kesterson, J. Agents Action 1991, 34, 117.
In conclusion, starting with known 5-LO and H1 phar-
macophores, we have made a novel, orally active, dual-
function compound.23 Further optimization of the
spacer linking these two pharmacophores led to com-
pounds with improved in vitro and in vivo activities.
These results shall be reported soon.
19. Moguilevski, N.; Varsalona, F.; Noyer, M.; Gillard, M.;
Guillaume, J. P.; Garcia, L.; Szpirer, C.; Szpirer, J.;
Bollen, A. Eur. J. Biochem. 1994, 224, 489.
20. Zhang, Y. Y.; Lind, B.; Radmark, O.; Samuelsson, B.
J. Biol. Chem. 1993, 268, 2535.
References and notes
21. Konsett, H.; Rossler, R. Arch. Exp. Pathol. Pharmakol.
1940, 195, 71.
1. Barnes, P. J.; Chung, K. F.; Page, C. P. Pharmacol. Rev.
1998, 50, 515.
2. Roquet, A.; Dahlen, B.; Kumlin, M.; Ihre, E.; Gudrun, A.;
Binks, S. Am. J. Respir. Crit. Care. Med. 1997, 155,
1856.
22. Spaethe, S. M.; Snyder, D. W.; Pechous, P. A.; Clarke, T.;
VanAlstyne, E. L. Biochem. Pharmacol. 1992, 43, 377–382.
23. This work was presented by T. Lewis at the ACS224th
National Meeting, Boston, MA, August 18, 2002, MEDI
15.