2,4-Disubstituted piperidines as selective CC chemokine receptor 3 (CCR3) antagonists: Synthesis and selectivity

Article abstract of DOI:10.1016/j.bmcl.2006.08.012

Linear unselective CCR3 antagonist leads with IC₅₀ values in the 200 nM range were converted into low nM binding compounds selective at CCR3 by moving the piperidine nitrogen substituent to the carbon at the 2-position of the ring. Substitution

Full text of DOI:10.1016/j.bmcl.2006.08.012

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5695–5699
2,4-Disubstituted piperidines as selective CC chemokine receptor 3
(CCR3) antagonists: Synthesis and selectivity
Paul S. Watson,*, Bin Jiang, Kim Harrison, Nao Asakawa, Patricia K. Welch,
Maryanne Covington, Nicole C. Stowell, Eric A. Wadman, Paul Davies,
Kimberly A. Solomon, Robert C. Newton, George L. Trainor, Steven M. Friedman,
Carl P. Decicco and Soo S. Ko
Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ 08542-5400, USA
Received 9 June 2006; revised 31 July 2006; accepted 1 August 2006
Available online 23 August 2006
Abstract—Linear unselective CCR3 antagonist leads with IC50 values in the 200 nM range were converted into low nM binding
compounds selective at CCR3 by moving the piperidine nitrogen substituent to the carbon at the 2-position of the ring. Substitution
of the piperidine nitrogen with simple alkyl and acyl groups was found to improve the selectivity of this new compound class. In
particular, N-{3-[(2S, 4R)-1-(propyl)-4-(4-fluorobenzyl)piperidinyl]propyl}-N0-(3-acetylphenyl)urea exhibited single digit nanomo-
lar IC50 values for CCR3 with >100-fold selectivity against an extensive counter screen panel.
Ó 2006 Elsevier Ltd. All rights reserved.
Infiltration of eosinophils into the airways is a com-
mon feature of allergic asthma and, in animal models,
contributes to airway hyperresponsiveness and remod-
eling. CC chemokine receptor-3 (CCR3) is a principal
mediator of eosinophil migration, and its ligands,
such as eotaxin, are prominently expressed in asth-
matic airways. Moreover, mice in which the CCR3
or eotaxin genes are disrupted exhibit reduced airway
eosinophilia following allergen challenge. Thus, small
molecule antagonists of CCR3 may provide a novel
therapy in asthma by inhibiting eosinophilic
inflammation.
of two of the three key binding motifs (benzyl group and
phenyl urea) were chosen for further exploration. The
results for path A have been described previously.3
The following letter describes the SAR observed by
moving the urea alkyl chain from the nitrogen to the
NH
H
N
H
N
CN
CN
O
O
B
Previously disclosed efforts identified compound 1 as a
lead structure for further optimization (see Fig. 1).1
Compound 1 is moderately potent at CCR3 (binding
IC50 = 200 nM),2 but it exhibited little selectivity against
several other protein targets surveyed (for example, 5-
HT2a). In efforts to improve potency and selectivity,
two chemical series that altered the spacial arrangements
N
N
H
N
H
A
1
O
N
N
H
N
H
CN
Keywords: Eotaxin; CCR3; Antagonist; Piperidine; Eosinophil.
*
Corresponding author. Tel.: +1 919 287 1268; fax: +1 919 941
9177; e-mail: pwatson@inspirepharm.com
Figure 1. Structural modifications explored to address selectivity issues
of compound 1.
Present address: Inspire Pharmaceuticals Inc., 4222 Emperor Bou-
levard, Suite 200, Durham, NC 27703, USA.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.012

5696
P. S. Watson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5695–5699
2-carbon (path B). 2,4-Disubstituted piperidines were
considered interesting targets due to the introduction
of stereochemical diversity (both relative and absolute)
and the availability of the free nitrogen atom as an addi-
tional site of derivation.
improvement in selectivity. Previous observation that
installation of a p-fluoro substituent on the piperidine
benzyl ring was found to improve CCR3 affinity by
10-fold1 led to the preparation of compounds 9 and
10. Unlike the previous chemotypes, the p-fluoro substi-
tuent only provided for a twofold improvement in
potency (compare 8 to 9). The p-fluoro substituent on
the piperidine benzyl ring improved the selectivity for
DAT and NET (compare 8 to 9), but eroded the selectiv-
ity for 5-HTT. As in our other series of CCR3 antago-
nists,3 symmetrical 3,5-disubstitution on the urea
phenyl group improved CCR3 binding affinity by three-
fold (compare 9 to 10). Symmetrical 3,5-disubstitution
on the urea phenyl group was not beneficial to any
transporter selectivity.
The CCR3 binding affinities of the cis and trans-2,4-di-
substituted urea diastereomers are shown in Table 1.
The propyl cis-diastereomer (2) showed twofold less
binding affinity to CCR3 with regard to linear achiral
1. However, the propyl trans-diastereomer (3) demon-
strated a 10-fold improvement in binding affinity to
CCR3. Separation of the enantiomers of the propyl
trans-racemic mixture by HPLC provided both antipo-
des. The binding results for 4 and 5 established that
the affinity for CCR3 resides in the (+)-trans-diastereo-
mer. Evaluation of the ‘‘linker’’ chain length (at the level
of the racemate) found the ethyl and butyl trans-diaste-
reomers (compounds 6 and 7, respectively) to be 10-fold
less active than the propyl trans-diastereomer (3). The
optimal length of the linker chain in this new series
was in complete agreement with our previous work on
other related series.1,3
In an attempt to address the transporter selectivity issue,
we examined substitution on the piperidine nitrogen.
The 3- and 3,5-diacetyl phenyl urea substituents are used
to illustrate these general SAR trends (Table 3). As ob-
served previously in this and other series,3 disubstitution
on the urea phenyl group was again found to be more
potent than monosubstitution (cf. 12–16, 14–17, and
15–18, respectively). Substitution at the piperidine nitro-
gen was found to be well tolerated and did not signifi-
cantly alter binding affinity to CCR3 in most
derivatives. However, substitution with a 2-trifluoroeth-
yl substituent (compound 25) eliminated all activity for
CCR3. Fortunately, simple alkyl substitution of the
piperidine nitrogen provided selectivity over the CNS
transporters. All nitrogen substituents tested were found
to provide reasonable (at most occurrences >100-fold)
selectivity over NET and DAT. In regard to 5-HTT
selectivity, simple alkyl groups (e.g., compounds 11–14
and 16–19) provided 40- to 300-fold selectivity. The
presence of an oxygen atom c to the piperidine nitrogen
(compounds 15 and 18) was found to significantly re-
duce the 5-HTT selectivity. Notably, acylation of the
nitrogen (compounds 26 and 27) yielded selective com-
pounds across all three CNS transporters with only
about 5-fold loss of the CCR3-binding affinity (compare
to compounds 10 and 16). The maintenance of the
CCR3 binding affinity in these derivatives could be spec-
ulated to derive from the enhancement of a hydrogen
bond or the rigidifying effect of the nitrogen substituent
(pseudo A1,3 strain) on the overall conformation of the
system. This result is quite interesting, given that com-
pound 25 is poorly active at CCR3. Finally, urea 12
Concurrent with the discovery that the propyl trans-dia-
stereomer was a potent CCR3 antagonist was the obser-
vation that this unsubstituted piperidine structural motif
also demonstrated potent inhibitory activity for norepi-
nephrine (NET), dopamine (DAT), and serotonin reup-
take transporters (5-HTT). Table 2 illustrates that both
enantiomers of the propyl trans-diastereomer (4 and 5)
are very potent inhibitors of all three transporters. Sub-
stitution of the 3-cyano group with a 3-acetyl function-
ality (compare compounds 4 and 8) provided no
Table 1. CCR3 binding affinities of 2,4-disubstituted ureas
NH
H
H
N
N
CN
Bn
n
O
Compound
Optical purity
Isomer
CCR3 IC50 (nM)
n
2
3
4
5
6
7
( )
( )
(+)
(ꢀ)
( )
( )
cis
386
22
1
1
1
1
0
2
trans
trans
trans
trans
trans
13
644
185
142
Table 2. Transporter selectivity issues of 2,4-disubstituted piperidine ureas
R
NH
H
N
H
N
R1
O
R2
Compound
R
R1
R2
CCR3 IC50 (nM)
13.0
644
5-HTT Ki (nM)
DAT Ki (nM)
NET Ki (nM)
4
5
(+)
(ꢀ)
( )
(+)
(+)
H
H
H
F
CN
CN
Ac
H
H
H
H
Ac
35
54
42
18
3.5
34
10
37
46
8
6.0
1.4
0.5
32
39
9
10
Ac
Ac
219
119
233
95
F

P. S. Watson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5695–5699
5697
Table 3. Binding affinities and transporter selectivity profiles for selected N-substituted piperidine ureas
F
R2
N
H
H
N
N
R1
O
Ac
Compound
R1
R2
CCR3 IC50 (nM)
5-HTT, Ki (nM)
DAT, Ki (nM)
NET, Ki (nM)
9
11
12
13
14
15
10
16
17
18
19
20
21
22
23
24
25
26
27
28
H
H
Me
1.4
1.7
18
572
296
117
217
50
219
96
233
257
H
H
Pr
Cyclopropyl methyl
Allyl
2.0
2.5
756
251
855
2240
119
157
297
451
609
754
2659
651
621
931
NT
1224
2082
31
4168
3449
1230
1820
95
H
H
4.1
7.0
H
(CH2)2OH
H
Pr
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
0.5
0.7
3.5
166
69
1480
876
Allyl
1.0
0.9
(CH2)2OH
(CH2)3OH
CH2C(O)Me
CH2C(O)NH2
CH2CH2F
Propargyl
CH2CHF2
CH2CF3
C(O)CH3
C(O)CH2CH3
C(@NH)NH2
40
1390
497
1.3
0.9
212
11
381
3.6
2.3
710
49
5007
317
1.9
305
328
NT
5465
224
19
524
12.4
1380
3.9
1052
NT
1379
9261
388
5.8
4.3
(IS811) was found to be 170-fold selective over 5-HT2a,
a potential liability previously mentioned for this series
of compounds.3
properties to warrant further preclinical in vivo studies.
As shown in Figure 2, IS811 dose-dependently inhibited
eotaxin-induced eosinophil influx to the lung.5,6
A representative set of compounds was further charac-
terized in a secondary calcium mobilization assay and
in their ability to block the eotaxin-induced human
eosinophil chemotaxis assay (Table 4).4 All 2,4-disubsti-
tuted piperidine urea analogs proved to be antagonists
of eotaxin-induced calcium mobilization. Consistent
with previous data, 3,5-disubstitution proved to be opti-
mal in both functional assays (compare compounds 12
to 16). Notably, compound 26, which lacks a basic
nitrogen and exhibits an IC50 of 3.9 nM in the CCR3
binding assay, proved much less potent in the secondary
functional assays. Compounds 12 and 16 proved to be
potent inhibitors of chemotaxis and the potency corre-
lated well to the binding assay results.
The synthesis of the required urea analogs (2–5) was ini-
tially performed according to the sequence described in
Scheme 1. Starting from N-Boc-4-benzyl piperidine (29),
metalation with s-BuLi and alkylation of the resulting
Table 5. Pharmacokinetic properties of compound 12 (IS811) in
mouse and Cynolgus monkey
Species and route
of administration
Dose
(mg/kg) (L/h/kg) (L/kg) (h) (nM)
Cl
VD
T1/2 Cmax %F
Mouse (IV)
Mouse (PO)
Cyno (IV)
Cyno (PO)
2
10
2
10.6
—
8
— —
1.4 274 36
—
3.1
0.4
—
— —
989 18
5
9
Compound 12 (IS811) was further evaluated in two ani-
mal PK models (Table 5) and found to have suitable
7.5
5.0
2.5
0.0
Table 4. Calcium mobilization and chemotaxis values for selected
analogs
F
R2
N
H
H
N
N
R1
O
Ac
*
*
Compound R1 R2 CCR3
Ca+
IC50 (nM) IC50(nM) EC50 (nM)
Chemotaxis
50
2
20
0
10
1
Dose IS811 (mg/kg)
12
16
26
H Pr
Ac Pr
2.0
0.7
26
1.2
133
19
4.7
>100
Figure 2. Inhibition of eotaxin-induced eosinophil influx into the lung
by compound 12 (IS811).6
Ac Ac 3.9

5698
P. S. Watson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5695–5699
hol 31 could be displaced with sodium cyanide and re-
Bn
a
b
NBoc
30
NBoc
duced to give amine 36. Subsequent treatment with the
isocyanate, acidic cleavage, and separation by HPLC
provided trans-urea 7.
OH
NBoc
29
Bn
Bn
31
c,d,e
NBoc
NR
During the evolution of this work, a growing body of evi-
dence suggested that an asymmetric synthesis of this
structural moiety that could incorporate a p-fluoro sub-
stituent on the piperidine benzyl group would be benefi-
cial. Therefore, incorporation of these two features
became part of the primary focus of an improved synthe-
sis. A racemic version of the following sequence has been
described in a different publication.7 Toward this end,
olefin 40 (Scheme 3) was targeted as a versatile interme-
diate. Addition of allyl magnesium chloride to the chiral
acyliminium ion derived from pyridine 37 provided en-
one 38 after acid catalyzed removal of the TIPS group.
The absolute stereochemistry of the 2-position was as-
signed as S using previously reported models.8 Base cat-
alyzed removal of the chiral auxiliary, protection of the
vinylogous amide, and conjugate reduction of the enone
yielded ketone 39. Wittig olefination and dissolving met-
al reduction provided aniline 41 in good yield.
H
N
H
f
N
CN
Bn
NH2
Bn
O
32
R=Boc, 33
R=H, 2-5
g
Scheme 1. Reagents and conditions: (a) s-BuLi, THF, 0 °C then allyl
iodide; (b) 9-BBN, THF then 30% H2O2; (c) TsCl, pyridine, 0 °C; (d)
NaN3, DMSO; (e) H2, 10% Pd/C, MeOH; (f) m-CN–Ph–NCO, THF;
(g) TFA, CH2Cl2.
a
NBoc
NBoc
NBoc
NBoc
6
7
Bn
OH
Bn
34
36
30
b, c
OTs
Bn
NH2
Bn
35
Previous studies indicated that hydrogenation of the
Wittig adduct was not selective for the trans-diastereo-
mer.7 In addition, conditions that did provide high levels
of selectivity (dissolving metal reduction) were also
found to remove the desired p-fluoro substituent. Hence,
a modified procedure that utilizes an aniline as a masked
fluorine substituent was developed. To this end, conver-
sion of the aniline to the fluoride via the in situ genera-
tion of the diazonium salt required the reprotection of
the piperidine nitrogen after crude work-up. This proce-
dure provided olefin 41 in gram quantities. Olefin 41 was
utilized according to procedures outlined in Scheme 1.
Subsequent synthetic derivatives could then be deprotec-
ted (TFA) and substituted on the piperidine nitrogen by
a number of sequences (reductive amination, acylation,
and acylation/reduction).
Scheme 2. Reagents and conditions: (a) O3, CH2Cl2, ꢀ78 °C then
DMS, 50%; (b) NaCN, DMSO, 50 °C, 80%; (c) H2, 10% Pd/C, MeOH,
90%.
anion with allyl iodide provided olefin 30. Hydrobora-
tion of 30 with 9-BBN upon oxidative workup yielded
alcohols 31. Three-step conversion of alcohols 31 (tosy-
lation, azidation, and reduction) led to amines 32. Addi-
tion of amines 32 to 3-cyanophenyl isocyanate provided
a 2:1 mixture of protected ureas 33. Separation of the cis
and trans-diastereomers was accomplished by HPLC.
Subsequent deprotection (TFA) led to racemic mixtures
of each urea 2 and 3. In a similar manner, other deriva-
tives differing by the extent of substitution on the urea
phenyl ring were prepared using the corresponding
isocyanates.
In conclusion, we have demonstrated that moderately
potent nonselective 1,4-disubstituted benzyl piperidines
can be converted into potent selective CCR3 antagonists
by moving the N-substituent to the 2-position and
substituting the free nitrogen position with small alkyl
Amine 30 could also be treated with ozone upon reduc-
tive workup to provide alcohol 34 (Scheme 2). Conver-
sion of 34 to trans-urea 6 was performed in a similar
manner to 2. Finally, the tosylate 35 derived from alco-
OMe
Boc
CO2-(+)-TCC
a,b
c,d,e
N
N
S
TIPS
O
O
N
37
38
39
f,g
H2N
Boc
F
Boc
h
N
N
S
R
40
41
Scheme 3. Reagents and conditions: (a) (+)-TCC–C(O)Cl, toluene/THF, ꢀ30 °C then CH2@CHCH2MgBr, ꢀ78; 10% HCl, rt, 68%; (b) TFA,
CH2Cl2, 100%; (c) K2CO3, MeOH, reflux; (d) Boc2O, DMAP, CH3CN, 92% for two steps; (e) Zn, HOAc, 50 °C, 90%; (f) 4-(H(Z)N)-PhCH2P(Ph)3Cl,
t-BuOK, THF, 90%; (g) Li, NH3, THF, ꢀ78 to ꢀ30 °C, 90%; (h) 70% HF/pyridine, ꢀ78 to ꢀ30 °C then NaNO2, 10 °C then urea, 0 °C; Boc2O,
NaOH, 1,4-dioxane, 20 °C, 80%.

P. S. Watson et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5695–5699
5699
2. For binding assay details, see footnote 16 in reference 1.
and acyl groups. We have discovered that the potency of
these derivatives lies in the 2S, 4R-trans diastereomer.
We have also found that low binding affinities to
CCR3 can be maintained without the presence of a
key basic nitrogen atom. These derivatives were found
to possess very good in vivo PK characteristics in both
mouse and monkey models.
3. Varnes, J. G.; Gardner, D. S.; Santella, J. B., III; Duncia, J.
V.; Estrella, M.; Watson, P. S.; Clark, C. M.; Ko, S. S.;
Welch, P. K.; Covington, M.; Stowell, N. C.; Wadman, E.
A.; Davies, P.; Solomon, K. A.; Newton, R. C.; Trainor, G.
L.; Decicco, C. P.; Wacker, D. A. Bioorg. Med. Chem. Lett.
2004, 14(7), 1645.
4. For calcium mobilization assay conditions, see footnote 19
in reference 1. Readers are referred to US Patent 6,331,544
(Ko, S. S.; DeLucca, G. V.; Duncia, J. V.; Santella, J. B.,
III; Wacker, D.A.) for details regarding chemotaxis assay
conditions.
Acknowledgments
5. De Lucca, G. V.; Kim, U. T.; Vargo, B. J.; Duncia, J. V.;
Santella, J. B., III; Gardner, D. S.; Zheng, C.; Liauw, A.;
Wang, Z.; Emmett, G.; Wacker, D. A.; Welch, P. K.;
Covington, M.; Stowell, N. C.; Wadman, E. A.; Das, A.
M.; Davies, P.; Yeleswaram, S.; Graden, D. M.; Solomon,
K. A.; Newton, R. C.; Trainor, G. L.; Decicco, C. P.; Ko,
S. S. J. Med. Chem. 2005, 48(6), 2194.
We gratefully acknowledge the help of A. J. Mical, K.
A. Rathgeb, D. R. Wu, and N. C. Caputo of the Sepa-
rations Group; T. H. Scholz and G. A. Nemeth of the
Spectroscopy Group; and Percy Carter for his critical re-
view and helpful suggestions during the preparation of
this manuscript.
6. Das, A. M.; Vaddi, K. G.; Solomon, K. A.; Krauthauser,
C.; Jiang, X.; Mcintyre, K.; Wadman, EA.; Welch, P. K.;
Covington, M.; Graden, D. M.; Yeleswaram, S.; Trzas-
kos, J. M.; Newton, R. C.; Mandlekar, S.; Ko, S. S.;
Carter, P. H.; Davies, P. J. Pharmacol. Exp. Ther. 2006,
318, 411.
7. Watson, P. S.; Jiang, B.; Scott, B. Org. Lett. 2000, 2(23),
3679.
8. Comins, D. L.; Goehring, R. R.; Joseph, S. P.; O’Connor,
S. J. Org. Chem. 1990, 55, 2574.
References and notes
1. Wacker, D. A.; Santella, J. B., III; Gardner, D. S.; Varnes,
J. G.; Estrella, M.; DeLucca, G. V.; Ko, S. S.; Tanabe, K.;
Watson, P. S.; Welch, P. K.; Covington, M.; Stowell, N. C.;
Wadman, E. A.; Davies, P.; Solomon, K. A.; Newton, R.
C.; Trainor, G. L.; Friedman, S. M.; Decicco, C. P.;
Duncia, J. V. Bioorg. Med. Chem. Lett. 2002, 12(13), 1785.