3-[6-(2-Dimethylamino-1-imidazol-1-yl-butyl)-naphthalen-2-yloxy]-2,2-dimethyl-propionic acid as a highly potent and selective retinoic acid metabolic blocking agent

Article abstract of DOI:10.1016/j.bmcl.2006.02.020

3-[6-(2-Dimethylamino-1-imidazol-1-yl-butyl)-naphthalen-2-yloxy]-2,2-dimethyl-propionic acid and analogs were designed and synthesized as highly potent and selective CYP26 inhibitors, serving as retinoic acid metabolic blocking agents (RAMBAs), with demonstrated in vivo efficacy to increase the half-life of exogenous atRA.

Full text of DOI:10.1016/j.bmcl.2006.02.020

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2729–2733
3-[6-(2-Dimethylamino-1-imidazol-1-yl-butyl)-naphthalen-2-yloxy]-
2,2-dimethyl-propionic acid as a highly potent and selective
retinoic acid metabolic blocking agent
Mark J. Mulvihill,* Julie L. C. Kan, Andrew Cooke, Shripad Bhagwat, Patricia Beck,
Mark Bittner, Cara Cesario, David Keane, Viorica Lazarescu, Anthony Nigro,
Christy Nillson, Bijoy Panicker, Vanessa Smith, Mary Srebernak, Feng-Lei Sun,
Matthew O’Connor, Suzanne Russo, Gia Fischetti, Michael Vrkljan, Shannon Winski,
Arlindo L. Castelhano, David Emerson and Neil W. Gibson
OSI Pharmaceuticals, Inc., 1 Bioscience Park Drive, Farmingdale, NY 11735, USA
Received 21 December 2005; revised 3 February 2006; accepted 6 February 2006
Available online 28 February 2006
Abstract—3-[6-(2-Dimethylamino-1-imidazol-1-yl-butyl)-naphthalen-2-yloxy]-2,2-dimethyl-propionic acid and analogs were
designed and synthesized as highly potent and selective CYP26 inhibitors, serving as retinoic acid metabolic blocking agents
(RAMBAs), with demonstrated in vivo efficacy to increase the half-life of exogenous atRA.
Ó 2006 Elsevier Ltd. All rights reserved.
All-trans retinoic acid (atRA) (1), a naturally occurring
retinoid, is biosynthesized and present in a multitude of
human and mammalian tissues, and performs a critical
role in the regulation of gene expression, cellular differ-
entiation, and proliferation of epithelial cells.1 Retinoids
have proven to be valuable agents in treating skin relat-
ed diseases as well as in tumor therapy.2 Furthermore,
atRA has been successful in treating both acute myelog-
enous leukemia (AML) and acute promyelocytic leuke-
mia (APL), changing the prognosis of APL from a
fatal leukemia to a highly curable disease.3
Inhibitors of CYP26, also known as retinoic acid meta-
bolic blocking agents (RAMBAs), have proven to be
effective in blocking the catabolic effects on atRA.6,7
Additionally, CYP26 inhibitors have the potential to
provide an alternative approach of potentiating endoge-
nous atRA, avoid the frequency and severity of compli-
cations associated with intensive high dose atRA
therapy, and provide an effective means of treatment fol-
lowing relapse in cases where resistance emerges due to
CYP26 upregulation.
To explore the therapeutic potential of increasing
endogenous levels of atRA through small molecule
CYP26 inhibitors, a series of 2,6-disubstituted naphtha-
lenes were designed via a de novo approach. As de-
scribed in our previous report,7 we designed a novel
class of highly potent and selective naphthyl-based
CYP26 inhibitors, which displayed inhibition of
proliferation of both T47D and AT6.1 cells in vitro in
combination with atRA. From our efforts, 3-[6-(2-di-
methylamino-1-imidazol-1-yl-propyl)-naphthalen-2-yloxy]-
2,2-dimethyl-propionic acid 1a emerged as a lead
CYP26 inhibitor (Fig. 1). We report herein, our success-
ful efforts to further optimize lead CYP26 inhibitor 1a
through modifications of R1–R6, affording highly po-
tent, selective, and orally bioavailable CYP26 inhibitors
Despite the effectiveness of atRA, clinical uses have been
significantly hampered by the emergence of resistance.
The oxidative catabolism of atRA to 4-hydroxy-atRA
by CYP26, an inducible cytochrome P450 enzyme,4 is
believed to be one important mechanism of resistance.
The systemic concentration of atRA is tightly controlled
through such a negative feedback mechanism, limiting
its plasma levels and therefore, its biological efficacy.5
Keywords: All-trans retinoic acid; atRA; Retinoic acid metabolic
blocking agents; RAMBA; CYP26; Cancer.
*
Corresponding author. Tel.: +1 631 962 0787; fax: +1 631 845
5671; e-mail: mmulvihill@osip.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.02.020

2730
M. J. Mulvihill et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2729–2733
the two synthetic pathways used to prepare key inter-
N
mediate 7. Scheme 2 illustrates the synthetic pathway
used to prepare final compounds. The first route
which incorporated the amino moiety (NR3R4) earlier
in the synthesis was utilized to prepare compounds of
series I and began with the treatment of 2-bromo-6-
methoxynaphthalene 2 with magnesium to afford the
naphthylmagnesium bromide Grignard species which
was added to a solution of 2-chloropropanoylchloride
in THF cooled to ꢀ78 °C to afford a-chloroketone 3
(Scheme 1).8 Conversion of chloroketone 3 to the
more reactive iodoketone 4 via the Finkelstein reac-
tion (NaI and acetone) proceeded smoothly and subse-
quent alkylation with amines cleanly afforded
aminoketones 5. Deprotection of methyl ether 5 with
HBr/acetic acid afforded naphthenol 6. Alkylation of
naphthenol 6 under standard Mitsunobu conditions
(PPh3, DIAD, and HOCH2C(R1R2)CO2R7) cleanly
afforded ethers 7. The second route incorporated the
amino moiety (NR3R4) at a later stage in the synthesis
and was utilized in preparing compounds of series II.
The synthesis began with the preparation of naphthyl-
magnesium bromide Grignard species followed by
N
R6
R5
R4
O
N
HO
O
3 R
R2
1 R
1
1a (R1, R2, R3, R4, R5 = CH3 and R6 =H)
Figure 1. Naphthyl-derived CYP26 inhibitor.
with in vivo efficacy resulting in the increase of duration
of exogenous atRA.
Our naphthyl-based CYP26 inhibitor optimization pro-
gram focused SAR modifications to four series: series I:
modifications to R1 and R2, while holding R3–6 con-
stant; series II: modifications to R3 and R4, while hold-
ing R1–2 and R5–6 constant; series III: merging optimized
modifications from series I and II; series IV: modifica-
tion to R5 and R6 while holding R1–4 constant.
reaction with
a
suitable acid chloride (ClC-
OCH(R5R6)) to afford ketone 8. Hydrolytic deprotec-
tion of the methyl ether of compound 8 with HBr/
acetic acid afforded naphthenol 9, which cleanly
underwent Mitsunobu-directed alkylation with suitable
Two synthetic strategies were pursued in order to af-
ford appropriate advanced intermediates which would
be suitable for late stage analoging to afford targeted
compounds (Schemes 1 and 2). Scheme 1 illustrates
O
O
R6
R6
a
c
e
R5
N
R5
8
8
X
RO
3 R
R4
RO
5 (R8 = Me)
O
3 (X = Cl)
4 (X = I)
d
R6
R5
R4
b
6 (R8 = H)
Br
O
8
7
N
RO
O
3 R
RO
R2
7
1 R
2
O
O
e
R6
R6
R5
f
c
O
5 R
Y
8
7
RO
RO
O
R2
1 R
8 (R8 = Me)
9 (R8 = H)
10 (Y = H)
11 (Y = Br)
g
d
Scheme 1. Reagents and conditions: (a) i—Mg, THF; ii—ClCOC(X)(R5)(R6), ꢀ78 °C; (b) NaI, acetone; (c) HNR3R4, MeOH; (d) HBr/AcOH,
120 °C; (e) DIAD, Ph3P, HOCH2C(R1R2)CO2R7; (f) i—Mg, THF; ii—ClCOCH(R5)(R6), ꢀ78 °C; (g) Br2, dioxane, 120 °C.10,11
N
OH
N
R6
R5
R4
R6
R5
R4
b
a
O
O
7
7
7
N
N
RO
O
3 R
RO
O
3 R
R2
R2
1 R
1 R
12
13 (R7 = Me/Et)
1 (R7 = H)
c
Scheme 2. Reagents and conditions: (a) NaBH4, MeOH; (b) CDI, CH3CN, 65 °C; (c) NaOH, MeOH/THF.

M. J. Mulvihill et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2729–2733
2731
Table 2. SAR of R3/R4 modifications (series II)
alcohols (HOCH2C(R1R2)CO2R7) to afford ethers 10.
Bromination of compound 10 with bromine in diox-
ane afforded bromoketone 11, which was reacted with
amines to afford aminoketones 7.
N
N
O
Key intermediate 7 underwent reduction with NaBH4,
affording a 4:1 mixture of syn:anti-aminoalcohols 12
(syn isomer).9 The isomers were easily separable by
column chromatography. Based upon our previous
SAR which demonstrated that the syn isomer was more
potent than the anti isomer, the syn isomer 12 was taken
on to its respective syn-imidazolyl product 13 by
reaction with CDI in acetonitrile.9 Hydrolysis of the
ester moiety of 13 afforded final carboxylic acid-derived
products 1.
H
N
HO
O
3 R
R4
Compound
NR3R4
Biochemical IC50 (nM)
CYP26
CYP3A4
1h
1i
NMeEt
NEt2
N(Me)iPr
9.0
5.0
1700.0
580.0
400.0
448.0
2270.0
2150.0
NT
1j
3.0
1k
1l
N(Me)cyclohexyl
N(Me)nBu
125.0
60.0
34.0
583.0
1m
1n
Morpholino
Pyrrolidinyl
The efficacy of compounds 1a–t, as inhibitors of CYP26
was demonstrated and confirmed by a number of phar-
macological in vitro assays.7 The in vitro biochemical
assay was performed using microsomal preparations
from T47D cells induced to express CYP26. Enzymatic
activity was measured as the conversion of the radiola-
beled substrate to its metabolites, 4-OH-atRA (4-hy-
droxy all-trans retinoic acid) and 4-oxo-atRA (4-oxy
retinoic acid) by separation on a C18 HPLC column.
Inhibition of CYP26 activity in the presence of variable
naphthalene analog concentrations was used to deter-
mine the IC50s. Cell activity was determined for most
of the compounds; a subset of the results are as noted
(Table 5).
cluded that the dimethylamino moiety at NR3R4 was
optimal, affording both CYP26 potency and selectivity.
Series III consisted of the merger of the best R1–2 moiety
(cyclopentyl) with the best R3–4 moieties in an attempt
to gain synergy between the two series (Table 3). The hy-
brids were more potent and moderately selective
inhibitors.
Series IV focused on the chiral center at the C2 position
of the propyl appendage (Table 4). Replacement of the
methyl group at R6 with H afforded a significant loss
in potency (compound 1r). However, replacement of H
at R5 with methyl was tolerated (compound 1s). The
most significant breakthrough resulted from the incor-
poration of an ethyl side chain at R6. Compound 1t
was an inhibitor of CYP26 at 1.3 nM and was >2300-
fold selective toward CYP3A4.
For series I, modifications to R1 and R2 while holding
R
3–6 constant from the original lead compound 1a affor-
ded potent and selective CYP26 inhibitors (Table 1).
Cycloalkyl groups such as cyclopentyl and cyclohexyl
were preferred over that of cyclobutyl, cyclopropyl, or
pyranyl.
For series II, modifications to R3 and R4 of the amino
moiety while holding R1–2 and R5–6 constant with small
alkyl groups were generally tolerated (Table 2). Howev-
er, larger groups afforded a decrease in potency as well
as an increase in CYP3A4 activity. It was therefore con-
With the identification of highly potent and selective
CYP26 inhibitors through in vitro screening and medic-
inal chemistry efforts, in vivo profiling was conducted to
determine those with suitable PK properties. Four key
compounds, 1e,f,o, and t, in conjunction with original
lead 1a, met all criteria for potency, selectivity, and
Table 1. SAR of R1/R2 modifications (series I)
N
Table 3. Combining R1/R2 and R3/R4 modifications (series III)
N
N
O
H
N
N
HO
O
H3C
CH3
O
H
N
R2
1 R
HO
O
3 R
R4
R2
1 R
Compound
CR1R2
Biochemical IC50 (nM)
CYP26
CYP3A4
1a
1b
1c
1d
1e
1f
C(CH3)2
20.0
37.0
34.7
40.0
7.3
6300.0
5300.0
7500.0
Compound
NR3R4
CR1R2
Biochemical IC50
(nM)
C(CH2CH3)2
Cyclopropyl
Cyclobutyl
Cyclopentyl
Cyclohexyl
4-Pyranyl
CYP26
CYP3A4
7000.0
8200.0
1o
1p
1q
NMeEt
N(Me)iPr
NEt2
Cyclopentyl
Cyclopentyl
Cyclopentyl
4.2
1.6
1.4
2300
800.0
1010.0
11.2
46.7
8000.0
>20,000.0
1g

2732
M. J. Mulvihill et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2729–2733
Table 4. SAR of R5/R6 modifications (series IV)
RAMBA compounds. Plasma samples were collected
and extracted as above, and analyzed by HPLC-UV
(atRA) and HPLC–MS/MS (RAMBAs) using calibra-
tion and quality control samples prepared in blank
mouse plasma. The efficacies of compounds 1a and t
are illustrated in Graph 1, in which co-administration
of RAMBAs increases the half-life of atRA, resulting
in increases in the level and duration of atRA plasma
exposure (Table 6).
N
N
R6
R5
O
N
HO
O
Compound
R5
R6
Biochemical IC50 (nM)
CYP26
CYP3A4
In conclusion, a series of [2-imidazol-1-yl-2-(6-alkoxy-
naphthalen-2-yl)-1-methyl-ethyl]-dimethylamines were
designed and synthesized as CYP26 inhibitors, serving
as retinoic acid metabolic blocking agents (RAMBAs).
Building upon the SAR from compound 1a, a series of
modifications were made to R1–6 of compound 1 and
ultimately afforded highly potent and selective CYP26
inhibitors 1e,f,o, and t. The lead compounds main-
tained excellent pharmacokinetic and physiochemical
properties and demonstrated the ability to suppress
CYP26-mediated catabolism of atRA, improving the
in vivo half-life of atRA and its overall plasma concen-
trations. This approach provides a means for modulat-
1r
1s
1t
H
H
CH3
403.0
26.0
1.30
>20,000.0
2100
3300.0
CH3
H
CH2CH3
bioavailability (Table 5). Compounds 1a and t were pro-
gressed to in vivo efficacy studies, designed to determine
whether co-administration of atRA and an orally bio-
available CYP26 inhibitor afforded an increase in the
overall plasma exposure of exogenous atRA. The in vivo
studies used to determine the pharmacokinetics of the
lead compounds as well as in vivo efficacy were as fol-
lows: dose solutions for oral administration were pre-
pared by dissolving compounds at 0.5 mg/mL (of the
free base) in water for injection, and for intravenous
administration at 0.1 mg/mL (of the free base) in water
for injection. Female CD-1 mice (approximately six
weeks old) were dosed with compounds at 10 mL/kg
(5 mg/kg) by oral gavage and at 10 mL/kg (1 mg/kg)
by intravenous injection into the tail vein. Three mice
were dosed at each timepoint, terminal blood samples
in EDTA were taken by cardiac puncture and plasma
samples were analyzed by HPLC–MS/MS using calibra-
tion and quality control samples prepared in blank
mouse plasma. All plasma samples were extracted by
protein precipitation with acetonitrile (50 lL plas-
ma + 250 lL of acetonitrile) followed by centrifugation
(10,000g for 10 min at 4 °C). Pharmacokinetic parame-
ters were obtained by noncompartmental modeling of
the median (n = 3) concentration–time data.
8
7
5mg/kg atRA
6
5mg/kg atRA +
5
10mg/kg Compound 1a
4
5mg/kg atRA +
10mg/kg Compound 1t
3
2
1
0
0
2
4
6
8
10
Time (Hrs)
Graph 1. Plasma concentrations of atRA (lM) at 5 mg/kg alone and in
combination with compounds 1a and t.
Table 6. PK parameters for atRA at 5 mg/kg alone and in combina-
tion with compounds 1a and t
The in vivo efficacy studies were carried out similarly,
with the RAMBA compounds dosed at 10 mg/kg in
water for injection by oral gavage. Dose solutions of
atRA for oral administration were prepared by dissolv-
ing the compound at 0.5 and 2.0 mg/mL in 1.5% v/v
DMSO in safflower oil and were dosed by oral gavage
at 10 mL/kg (5 and 20 mg/kg) immediately prior to the
Dose schedule
atRA alone atRA + 1a atRA + 1t
C2h (lM)
C8h (lM)
AUC0–last (ng h/mL)
AUC0–last (fold increase
vs atRA alone)
5.24
<0.01
1686
1
6.66
0.04
6472
3.8
7.72
2.73
9896
5.9
Table 5. Profile of compounds 1a,e,f,o, and t
Properties
1a
1e
1f
1o
1t
Biochemical potency (nM)
Cell activity (nM)
CYP3A4 (nM)
20
125
7.3
11.2
3.9
1.3
<5
<100
8200
>5000
>5000
>5000
5.5
<100
8000
>5000
>5000
>5000
23
<100
2300
>5000
>5000
>5000
13
6300
>5000
>5000
>5000
6.0
3300
>5000
>5000
>5000
6.7
CYP1A2 (nM)
CYP2D6 (nM)
CYP2C9 (nM)
Cmax 5 mg/kg po (lM)
AUC0–1 5 mg/kg po (ng h/mL)
Terminal t1/2 1 mg/kg iv (h)
Oral bioavailability (F %)
4919
2.5
5799
2.49
17,492
2.46
99
18,946
7.1
99
5434
2.76
73
99
77

M. J. Mulvihill et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2729–2733
2733
5. Muindi, J.; Frankel, S. R.; Miller, W. H., Jr.; Jakubowski,
A.; Scheinberg, D. A.; Young, C. W.; Dmitrovski, E.;
Warrell, R. P., Jr. Blood 1992, 79, 299.
6. (a) Njar, V. C. O.; Brodie, A. M. Drugs 1999, 58, 233; (b)
Miller, V. A.; Rigas, J. R.; Muindi, J. R.; Tong, W. P.;
Venkatraman, E.; Kris, M. G.; Warrell, R. P., Jr. Cancer
Chemother. Pharmacol. 1994, 34, 522; (c) De Coster, R.;
Wouters, W.; Wan Ginckel, R.; End, D.; Krekels, M.;
Coene, M. C.; Bowden, C. J. Steroid Biochem. Mol. Biol.
1992, 43, 197.
7. Mulvihill, M. J.; Kan, J. L. C.; Beck, P.; Bittner, M.;
Cesario, C.; Cooke, A.; Keane, D. M.; Nigro, A. I.;
Nillson, C.; Smith, V.; Srebernak, M.; Sun, F.-L.;
Vrkljan, M.; Winski, S. L.; Castelhano, A. L.; Emerson,
D.; Gibson, N. Bioorg. Med. Chem. Lett. 2005, 15,
1669.
ing endogenous or exogenous atRA exposure in vivo
and therefore, the potential to improve atRA related
responses, resulting in a potentially therapeutic antican-
cer modality.
Acknowledgments
We thank Professors Bernhard Golding, Marvin J.
Miller, Iwao Ojima, and Victor Snieckus for consultation.
References and notes
1. Ross, A. C. J. Nutr. 2003, 133, 291.
8. Sonawane, H. R.; Bellur, N. S.; Kulkarni, D. G.; Ayangar,
N. R. Tetrahedron 1994, 50, 1243.
2. (a) Miller, W. H., Jr. Cancer 1998, 83, 1471; (b) Altucci,
L.; Rossin, A.; Raffelsberger, W.; Reitmair, A.; Chomi-
enne, C.; Gronemeyer, H. Nat. Med. 2001, 7, 680; (c)
Altucci, L.; Gronemeyer, H. Nat. Rev. Cancer 2001, 1,
181; (d) Kuijpers, A. L.; Van Pelt, P. T.; Bergers, M.;
Boegheim, P. J.; Den Bakker, J. E.; Siegenthaler, G.; Van
de Kerkhof, P. C.; Schalkwijk, J. Br. J. Dermatol. 1998,
139, 380.
3. Clinical Oncology; Rubin, P., Ed., 8th ed.; W.B. Saundersf
Company: Philadelphia, 2001; pp 577–578.
4. (a) Aelterman, W.; Lang, Y.; Willemsens, I.; Vervest, I.;
Leurs, S.; De Knaep, F. Org. Proc. Res. Dev. 2001, 5, 467;
(b) Sonneveld, E.; van der Saag, P. T. Int. J. Vitam. Nutr.
Res. 1998, 68, 404.
9. Mulvihill, M. J.; Cesario, C.; Smith, V.; Beck, P.; Nigro,
A. J. Org. Chem. 2004, 69, 5124.
10. Compound 5, where R3–6,8 = Me, was synthesized by
treating compound 2 (R8 = Me) with tBuLi, followed by
reaction with 2-dimethylamino-2-methylpropionitrile and
subsequent quenching with aqueous H2SO4.
11. HOCH2C(R1R2)CO2R7 was synthesized by reacting
C(R1R2)(CO2R7)2 with LiAl(OtBu)3 in THF to afford
HOCH2C(R1R2)CO2R7. In the case where CR1R2 equals
cyclopentyl, CH2(CO2R7)2 was reacted with NaH and
I(CH2)4I in THF to afford C(R1R2)(CO2R7)2 where
C(R1R2) equaled cyclopentyl.