Smoothened antagonists for hair inhibition

Article abstract of DOI:10.1016/j.bmcl.2010.06.033

A series of aminomethylpyrazoles were prepared and evaluated using cell-based Smoothened β-lactamase reporter assay and Smoothened binding assay. Potent Smoothened antagonists 10k and 10l were found to inhibit hair growth in vivo in the C3H/HeN mouse hair growth model. The more selective compound 10l was tested negative in the 3T3 NRU assay, indicating a low risk for causing photo-irritation and was efficacious using the C3H/HeN mouse hair growth model although it was slightly less efficacious than that of the reference compound eflornithine (7).

Full text of DOI:10.1016/j.bmcl.2010.06.033

Bioorganic & Medicinal Chemistry Letters 20 (2010) 4932–4935
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Smoothened antagonists for hair inhibition
a
b
b
b
Jie Jack Li a, , Veerabahu Shanmugasundaram , Satya Reddy , Laura L. Fleischer , Zenquan Wang ,
Yvonne Smith a, William G. Harter a, Wen-Song Yue a, Manju Swaroop b, Ling Li b, Christy Xiaodong Ji b,
Danielle Dettling b, Bella Osak c, Laura R. Fitzgerald c, Robert Conradi d,e
*
a Department of Chemistry, Michigan Laboratories, Pfizer Global Research & Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
b Department of Dermatology Biology, Michigan Laboratories, Pfizer Global Research & Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
c Mechanistic and Target Biology, Michigan Laboratories, Pfizer Global Research & Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
d Department of Pharmacokinetics and Drug Metabolism, Michigan Laboratories, Pfizer Global Research & Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
e Department of Pharmaceutical Sciences, Michigan Laboratories, Pfizer Global Research & Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of aminomethylpyrazoles were prepared and evaluated using cell-based Smoothened b-lacta-
mase reporter assay and Smoothened binding assay. Potent Smoothened antagonists 10k and 10l were
found to inhibit hair growth in vivo in the C3H/HeN mouse hair growth model. The more selective com-
pound 10l was tested negative in the 3T3 NRU assay, indicating a low risk for causing photo-irritation and
was efficacious using the C3H/HeN mouse hair growth model although it was slightly less efficacious
than that of the reference compound eflornithine (7).
Received 23 April 2010
Revised 4 June 2010
Accepted 7 June 2010
Available online 11 June 2010
Keywords:
Smoothened antagonists
Hair
Ó 2010 Elsevier Ltd. All rights reserved.
Photo-irritation
Aminomethylpyrazoles
Largely thanks to the pioneering work of Lewis,1 Nüsslein-
Volhard and Wieschaus2 three decades ago, genetic experiments
in Drosophila identified the Hedgehog (Hh) mutant. Mammals
have three Hedgehog homologues: Sonic Hedgehog (Shh), Desert
Hedgehog (Dhh), and Indian Hedgehog (Ihh). The Hh signaling
pathway is implicated in many different processes in vertebrate
development including cartilage differentiation, myotome and
sclerotome specification, hair follicle development, limb morpho-
genesis and the specification of different neuronal cell types.3
Aberrant activation of the Hedgehog pathway has been identified
as the likely cause of a number of tumors in humans such as ba-
sal cell carcinomas (BCCs) and primitive neurectodermal tumors
(PNETs). It is now clear that multiple components of the Hh path-
way can be altered in tumors.4
On the other hand, the seven-transmembrane protein Smooth-
ened (Smo) mediates the cellular response to the Hedgehog protein
signal and is involved in cell growth and differentiation during
embryonic development. Stimulation of the Smo pathway is directly
implicated in tissue maintenance and repair, but over-activation of
Smo could lead to tumorigenesis. Beachy and co-workers at Johns
Hopkins demonstrated that in unstimulated cells, the activity of
Smo is repressed by the protein Patched (Ptc), which appears to be
the receptor for the Shh ligand.5 Therefore, blocking the pseudo-
GPCR Smo would block the Hh signaling pathway.
Last decade saw great strides in the Smo field. As shown in
Figure 1, using the first known Smo antagonist cyclopamine (1)
as the starting point, Infinity Pharmaceuticals discovered IPI-926
(2), a potent and orally active Hh pathway antagonist now in phase
I clinical trials in cancer patients.6 Collaborating with Curis Inc., a
pioneer in the Smo field, Genentech discovered Smo antagonist
GDC-0449 (3).7 On the basis of positive results in phase I clinical
trials, 3 has been recently advanced to phase II trials to treat pa-
tients with locally advanced or metastatic solid tumors and BCC.
Beachy’s group also discovered potent Smo antagonists including
SANT-1 (4) and SANT-2 (5), as well as agonists such as SAG (6).8
Interestingly, Hh signaling is also closely related to hair cycle
regulations.9–14 Shh-dependent activation of Gli2 is essential for
embryonic hair follicle development.10 Sustained Hh signaling is
required for BCC proliferation and survival: conditional skin
tumorigenesis recapitulates the hair growth cycle.13 In knock-out
(KO) mice (ShhÀ/À), hair follicle development was severely im-
paired in mutant animals.14 The most compelling evidence
emerged when small molecule agonists of the Hh signaling path-
way was shown to modulate hair growth.15
* Corresponding author at present address: Discovery Chemistry, Bristol–Myers
Squibb Company, 5 Research Parkway, Wallingford, CT 06492, USA. Tel.: +1 203 677
7225.
Currently, the only prescription treatment for unwanted hair,
eflornithine (7, VaniqaÒ, Fig. 2), is only moderately efficacious in
a small percentage of subjects.16 Novel mechanisms with robust
E-mail address: jack.li@bms.com (J. J. Li).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.06.033

J. J. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4932–4935
4933
H
O
H
O
NH
H
N
H
H
H
H
H
H
H
H
MeO2SHN
HO
Cl
H
cyclopamine (1)
IPI-926 (2)
N
N
N
N
N
O
Cl
N
N
H
SANT-1 (4)
S
N
O2
GDC-0499 (3)
Cl
O
N
H
N
Cl
EtO
EtO
EtO
HN
O
N
S
Figure 3. Day 14 for compounds 4 and 7 in the in vivo efficacy tests using the
C3H/HeN mouse hair growth model.
SANT-2 (5)
NH
SAG (6)
SANT-1 (4) to develop our own Smo antagonists due to its robust
in vivo efficacy in the mouse hair growth model. It was decided
to replace the hydrazone functionality on 4 with amine due to
the known hepatotoxicity and neurotoxicity associated with
hydrazone-containing drugs.20 Furthermore, the potential hydra-
zone-associated toxicities could be even more acute and pro-
nounced for dermatological indications when the drug is exposed
to sunlight. As shown in the Scheme 1, our Smo antagonists 10
were readily assembled by reductive amination of pyrazole-alde-
hyde 8 with amine 9.
Figure 1. Known Smo antagonists 1–5 and Smo agonist 6.
F
F
NH2
OH
H2N
7
O
Figure 2. Eflornithine (7, VaniqaÒ).
The structure–activity relationship (SAR) of the first-batch
derivatives is listed in Table 1. All tri-substituted phenyl-pyrazoles
(10aÀ10e) have significantly weaker activities in the cell-based as-
say in comparison to hydrazone 4. Remarkably, while the para-
methoxy-phenyl pyrazole 10b has an IC50 of 2000 nM in the Smo
binding assay, the corresponding meta-methoxy-phenyl pyrazole
10b has an IC50 of 20 nM, a 100-fold increase! This observation
suggests that the meta-methoxyl-phenyl moiety may represent a
favorably tight binding to a particular pocket at the Smo receptor.
Chloro-pyrazole 10d is also very potent in the binding assay with
an IC50 of 10 nM, but it was not further explored because it was
tested positive in the 3T3 neutral red uptake (NRU) assay,21,22 indi-
cating a high risk for causing photo-irritation.
A breakthrough was made when di-substituted phenyl-pyra-
zole analogues were prepared in place of the tri-substituted phe-
nyl-pyrazoles. As shown in Table 2, although 3-methyl-1-
phenylpyrazoles 10e and 10f are inactive in both Smo cell-based
assay and Smo binding assay, 3-methyl-1-phenylpyrazoles 10g–
10l showed significantly improved potency in both assays. Once
again, meta-methoxyl analog 10j is remarkably potent with an
IC50 of 5 nM in the binding assay and 64.7 nM in the cellular assay.
When tested in vivo using the C3H/HeN mouse hair growth model,
efficacy and long-term safety are still needed for this indication.
We sought topical Smo antagonists to prevent or slow growth of
unwanted hair.
At the outset of the project, the proof of concept (POC) was
tested with known Smo antagonists for the treatment of unwanted
hair. In the C3H/HeN mouse hair growth model17 and using 30:70
propylene glycol/ethanol as the vehicle (Fig. 3), 3% of SANT-1 (4)
showed spectacular efficacy in inhibiting hair growth while
SANT-2 (5) was not active. In comparison, 13.9% of reference com-
pound 7 showed only moderate efficacy after 14 days.
We at first developed a robust and sensitive functional cell-
based assay that measures the activity of endogenous Smo using
a b-lactamase transcriptional readout.18 This is the first Smo repor-
ter assay that utilizes b-lactamase reporter technology. This assay
type has distinct advantages over other reporter technologies and
can be used in a high-throughput mode to search for therapeuti-
cally relevant downstream Smo target effectors. The assay also al-
lows a quick confirmation of the results by visual inspection
without any special instrumentation.
With functional cell-based b-lactamase assay as the primary as-
say, a Smo binding assay was also developed as the secondary assay
to evaluate if the compound binds to the Smo receptor. To that end,
a construct of truncated Smo over-expressing cells mammalian
expression system was prepared to gauge the compound’s influ-
ence on the specific interaction between Smo and tritiated-SANT-
2 (5).19
R1
R1
R2
R2
H
N
R5
9
R5
R4
CHO
R3
N
R4
N
N
N
N
NaBH(OAc)3
R3
In addition to screening our compound library, an effort was
undertaken to ‘parachute-in’ around known Smo antagonist
10
Scheme 1. Synthesis via reductive amination.
8

4934
J. J. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4932–4935
Table 1
Table 2
SAR of the tri-substituted phenyl-pyrazole analogues
SAR of the di-substituted phenyl-pyrazole analogues
Compd
SMOTTa
SMOBAb
Compd
SMOTTa
SMOBAb
IC50 (nM)
IC50 (nM)
IC50 (nM)
IC50 (nM)
4
4.74
25
N
N
N
N
N
N
>5000
>5000
328
>5000
>5000
333
N
O
N
N
1870
>5000
N
10e
O
10a
N
N
F
F
N
N
N
N
1690
823
2000
10f
F
O
10b
N
N
N
CF3
N
N
N
10g
20
N
O
N
N
10c
N
N
N
105
126
Cl
N
N
10h
N
N
O
1860
10
N
N
N
N
F
10d
78.4
64.7
96
5
a
SMOTT, Smoothened transient transfection cell-based assay.
SMOBA, Smoothened binding assay.
b
10i
F
N
N
N
it was found to be equipotent to reference compound 7. Due to
perceived toxicological liabilities that anilines are associated with
drugs in general and topical drugs in particular (although the nitro-
gen atom of aniline 10j here is in a piperazine ring), we sought to
eliminate this potential toxicity. We found our answer in analogs
10k and 10l, both of which have good potency in cellular and bind-
ing assays. They were subsequently tested efficacious to inhibit
hair growth in the C3H/HeN mouse hair growth model with a min-
imal efficacious dose (MED) of 3%.
With regard to selectivity, compound 10k is less selective than
10l. Among the GPCRs tested, 10k is also active against 5-HT1A
(Ki = 103 nM), 5-HT2B (Ki = 180 nM), 5-HT7 (Ki = 15 nM), although
it is not active against b1, b2, D1, D2, 5-HT2A and calcium channel
N
O
10j
F
N
N
N
40.8
8.27
216
124
N
N
10k
O
O
O
O
N
10l
F
with Ki’s all greater than 1
against 1A (Ki = 76 nM) and D4 (Ki = 148 nM), but inactive against
5-HT1A, 5-HT2A, D4, 1B, 2d and dofetilide with Ki’s all greater
than 1 M.
lM. On the other hand, 10l is active only
a
SMOTT, Smoothened transient transfection cell-based assay.
SMOBA, Smoothened binding assay.
b
a
a
a
l
Ideally, topical drugs exert their desired effects locally but are
rapidly inactivated via metabolism once they reach the systemic
circulation, thereby reducing unwanted systemic effects. Prelimin-
ary pharmaco-kinetics data predicts that 10l will be rapidly metab-
olized systematically because its half-life (t1/2) is 9.7 min in rat
liver microsomes (RLM) and 30.2 min in human liver microsomes
(HLM). Its cell membrane penetration PappAB was 11.3 Â 10
although it was slightly less efficacious than that of the reference
compound eflornithine (7).
In summary, a series of aminomethylpyrazoles were prepared
and evaluated using cell-based Smo b-lactamase reporter assay
and Smo binding assay. Potent Smo antagonists 10k and 10l were
found to inhibit hair growth in vivo in the C3H/HeN mouse hair
growth model. The more selective compound 10l was tested nega-
tive in the 3T3 NRU assay, indicating a low risk for causing photo-
irritation. More significantly, 10l was tested efficacious in the C3H/
HeN mouse hair growth model although with slightly lower effi-
cacy than that of the reference compound eflornithine (7).
À6 cm/s at 10
lM. Unlike oral drugs, topical drugs are required to
penetrate the skin to exert their pharmacological effects. We eval-
uated skin penetration of 10l using human cadaver skin mounted
onto Franz diffusion chambers.23 Compound 10l has high skin pen-
etration with a ratio of 0.52 versus isobutylmethylxanthine (IBMX),
a reference compound for skin penetration. It also has a good sol-
References and notes
ubility of 79
resulted in no significant binding and it had only 56% hERG inhibi-
tion at 3 M concentration, indicating a low risk for QTC prolonga-
lg/mL. Screening 10l against a panel of CYP enzymes
1. Lewis, E. B. Nature 1978, 276, 565.
2. Nüsslein-Volhard, C.; Wieschaus, E. Nature 1980, 287, 795.
3. For a review, see: Ingham, P. W. EMBO J. 1998, 17, 3505.
4. For a review, see: Borzillo, G. V.; Lippa, B. Curr. Top. Med. Chem. 2005, 5, 147.
5. Chen, J. K.; Taipale, J.; Cooper, M. K.; Beachy, P. A. Genes Dev. 2002, 16, 2743.
l
tion, and thus cardiac liability. More significantly, 10l was tested
efficacious in vivo using the C3H/HeN mouse hair growth model

J. J. Li et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4932–4935
4935
6. Tremblay, M. R.; Lescarbeau, A.; Grogan, M. J.; Tan, E.; Lin, G.; Austad, B. C., et al
J. Med. Chem. 2009, 52, 4400.
7. Robarge, K. D.; Brunton, S. A.; Castanedo, G. M.; Cui, Y.; Dina, M. S.; Goldsmith,
R., et al Bioorg. Med. Chem. Lett. 2009, 19, 5576.
encoding mouse or human Smo. Membranes were diluted in binding buffer
(50 M Tris pH 7.5, 25 mM MgCl2, 1 M EDTA, 0.1% BSA, and 100 mM NaCl).
Test compounds were also prepared by making appropriate dilutions in DMSO
using a final concentration of 1 M. In a 96-well costar assay plate, 5 g of
l
l
l
l
8. Chen, J. K.; Taipale, J.; Young, K. E.; Maiti, T.; Beachy, P. A. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 14071.
membranes were incubated with 4.0 nM of tritiated-SANT-2 (5) in the
presence of vehicle, cold competitor or test compound in binding buffer in a
9. For a review, see: Dlugosz, A. J. Clin. Invest. 1999, 104, 851.
10. Mill, P.; Mo, R.; Fu, H.; Grachtchouk, M.; Kim, P. C. W.; Dlugosz, A. A.; Hui, C. C.
Genes Dev. 2003, 17, 282.
final volume of 250
mixtures were harvested and washed with 6 mL of wash buffer (50
0.1% BSA) into pre coated unifilter GF/B filter plates (Perkin–Elmer) using a
semi-automated 96-well harvester Mach III (Tomtec). Plates were air dried.
l
L. Reactions were incubated at rt for 2 h. Then reaction
M Tris,
l
11. Oro, A. E.; Higgins, K. Dev. Biol. 2003, 255, 238.
12. Brancaccio, A.; Minichiello, A.; Grachtchouk, M.; Antonini, D.; Sheng, H.;
Parlato, R.; Dathan, N.; Dlugosz, A. A.; Missero, C. Hum. Mol. Genet. 2004, 13,
2595.
50 lL of Microscint-20 (Perkin–Elmer) was added to each well, and counted for
3 min using Packard top counter. Specific binding was calculated by
subtracting non specific binding from total binding. Initially, compounds
13. Hutchin, M. E.; Kariapper, M. S. T.; Grachtchouk, M.; Wang, A.; Wei, L.;
Cummings, D.; Liu, J.; Michael, L. E.; Glick, A.; Dlugosz, A. A. Genes Dev. 2005, 19,
214.
14. Sagai, T.; Masuya, H.; Tamura, M.; Shimizu, K.; Yada, Y.; Wakana, S.; Gondo, Y.;
Noda, T.; Shiroishi, T. Mamm. Genome 2004, 15, 23.
15. Paladini, R. D.; Saleh, J.; Qian, C.; Xu, G.-X.; Rubin, L. L. J. Invest. Dermatol. 2005,
125, 638.
16. Jackson, J.; Caro, J. J.; Caro, G.; Garfield, F.; Huber, F.; Zhou, W.; Lin, C.-S.;
Shander, D.; Schrode, K. Int. J. Dermatol. 2007, 46, 976.
17. C3H/HeN mouse hair growth model: Jiang, J.; Tsuboi, R.; Kojima, Y.; Ogawa, H.
Dermatology 2005, 32, 243.
were screened at 10
were titrated from 10 to 0.01
l
M in duplicate plates. Compounds with >50% inhibition
lM in triplicate to determine IC50
.
20. (a) Ortiz de Montellano, P. R.; Augusto, O.; Viola, F.; Kunze, K. L. J. Biol. Chem.
1983, 258, 8623; (b) Kalyanaraman, B.; Sinha, B. K. Environ. Health Perspect.
1985, 64, 179; (c) Nelson, S. D. J. Med. Chem. 1982, 25, 753; (d) Streeter, A. J.;
Timbrell, J. A. Drug Metab. Dispos. 1983, 11, 179; (e) Timbrell, J. A. Drug Metab.
Rev. 1979, 10, 125; (f) Prough, R. A.; Coomes, M. W.; Cummings, S. W.; Wiebkin,
P. Adv. Exp. Med. Biol. 1982, 136B, 983.
21. Nam, C.; An, S.; Lee, E.; Moon, S.; Kang, J.; Chang, I. ATLA, Altern. Lab. Anim. 2004,
32, 693.
22. Spielmann, H.; Presented at Alternatives to Animal Testing II, Proceedings of the
Second International Scientific Conference, Brussels, 1999, 15.
23. Franz diffusion chambers: Antille, C.; Tran, C.; Sorg, O.; Saurat, J.-H. Exp.
Dermatol. 2004, 13, 558.
18. Osak, B.; Fitzgerald, L. R. Assay Drug Dev. Technol. 2007, 5, 785.
19. Experimental procedure for the Smo filtration binding assay: Smo membranes
were prepared from HEK 293 cells that were stably transfected with cDNA