Synthesis, SAR and evaluation of [1,4′]-bipiperidinyl-4-yl-imidazolidin-2-one derivatives as novel CCR5 antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.04.077

Elaboration of our previously disclosed spiropiperidine template led to the development of a series of novel CCR5 antagonists. Results of SAR exploration and preliminary lead characterization are described.

Full text of DOI:10.1016/j.bmcl.2010.04.077

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3219–3222
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis, SAR and evaluation of [1,40]-bipiperidinyl-4-yl-imidazolidin-2-one
derivatives as novel CCR5 antagonists
*
David M. Rotstein , Stephen D. Gabriel, Nicole Manser, Lubov Filonova, Fernando Padilla, Surya Sankuratri,
Changhua Ji, Andre deRosier, Marianna Dioszegi, Gabrielle Heilek, Andreas Jekle, Paul Weller, Pamela Berry
Roche Palo Alto LLC, 3431 Hillview Avenue, Palo Alto, CA 94304, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Elaboration of our previously disclosed spiropiperidine template led to the development of a series of
novel CCR5 antagonists. Results of SAR exploration and preliminary lead characterization are described.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 22 March 2010
Revised 16 April 2010
Accepted 19 April 2010
Available online 22 April 2010
Keywords:
CCR5 Antagonist
Chemokine
Anti-HIV-1
Antiviral
The utility of CCR5 antagonists for the treatment of HIV1 and for
their potential use in a variety of autoimmune indications2 remains
an area of interest. We have described our earlier efforts in this
area which resulted in the discovery of a novel spiropiperidine
template 1.3 In a continuation of our program, analysis of a series
of CCR5 antagonists disclosed by AnorMed4 (represented by com-
pound 2) suggested replacement of the Roche spiro group with
the imidazolidinone fragment from the AnorMed template. This
strategy is illustrated in Figure 1, affording a novel new series rep-
resented by compound 3.5
As seen in Table 1, compound 3 showed promising activity in
binding and antiviral assays.6 Assessment in commonly used
in vitro pharmacokinetic assays demonstrated a number of issues
which could affect bioavailability. Screening in human liver micro-
somes (HLM) showed medium to low metabolic stability7 with a
high lipophilicity exhibit the best potency. Unfortunately, highly
lipophilic templates frequently suffer from low metabolic stability
and high clearance issues. A strategy to reduce lipophilicity and
block metabolism by incorporation of polar functionalities can lead
to improved metabolic stability and lower clearance. However, as
these analogs become more hydrophilic, antiviral potency typically
decreases and issues with permeability and efflux arise. Thus the
key challenge is to balance the opposing trends of antiviral potency
and permeability vs. metabolic stability and efflux.
A targeted SAR study of the lead template was thus initiated to
address these issues. For convenience, the right hand side of the
Cl
N
clearance of 31 lL/min/mg protein, which corresponds to pro-
N
jected in vivo clearance close to hepatic blood flow.8 As Table 1 also
shows, screening of compound 3 in Caco-2 cells9 showed low
intrinsic permeability10 and high P-glycoprotein (P-gp)11 mediated
active efflux, additional factors which could limit in vivo exposure.
Thus, the goal was to improve antiviral activity and address
metabolic and efflux issues with this template.
N
N
N
O
N
N
N
O
O
O
N
O
O
O
N
N
1
2
N
Our experience working with CCR5 antagonists has taught us
that meeting this goal can be quite challenging. The CCR5 receptor
binding pocket is largely lipophilic in nature and antagonists with
N
O
N
N
N
N
O
O
3
* Corresponding author. Tel.: +1 408 732 9030.
E-mail address: david.rotstein@comcast.net (D.M. Rotstein).
Figure 1.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.04.077

3220
D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3219–3222
Table 1
Table 2
SAR of lead compound 3
SAR of tail substituents
O
R1
O
N
N
N
N
N
N
N
N
R2
R2
N
O
O
N
#
R1
R2
R3
Binding
IC50
Antiviral HLMc Cacod
b
#
R2
Binding
Antiviral HLMa clog P Cacoa
a
IC50
AB/ER
a
a
IC50
IC50
AB/ER
O
3
4
5
rac 27
(R) 20
(S) >500
47
31
0.4/33
N
N
N
4
6
O
20
32
16
96
1.6
2.1
0.5/29
N
N
N
O
8
25
0.8/24
O
O
32
16
0.5/29
NDe
7
O
90
26
34
>625
26
7
8
155
224
13
0.9
2.1
1.2
1.3
2.5
2.8
1.9
0.6/28
0.3/56
0.2/38
0.1/13
1.4/13
3/4
>650
NDe
8
N
N
N
CO2Me
SO2Me
a
9
Competitive binding evaluated versus RANTES with IC50 values in nM, as mean
of two experiments.
b
10
11
12
13
CONCH3 32
99
8
Antiviral IC50 values in nM as mean of two experiments.
Human liver microsomal intrinsic clearance (lL/min/mg protein).
Permeability in Caco-2 cells AB, apical to basolateral and BA, basolateral to
c
d
OMe
OEt
OH
62
6
69
apical movement of test compound in 21 day cultured Caco-2 cells (cm/s ꢀ 10ꢁ6)
.
ER, efflux ratio of BA to AB.
3
47
e
Not determined.
20
8
12
0.2/61
a
Defined as in Table 1.
template, containing the 2,6-dimethyl pyrimidine group is referred
to as the head region and the left hand side, containing the substi-
tuted imidazolidinone is called the tail region.
As a first step, the individual enantiomers of racemate 3, (R) and
(S) isomers 4 and 5 were prepared. As can be seen in Table 1, all
antiviral activity resided in the (R) enantiomer 4 and subsequent
analogs maintained this stereochemistry.
Table 3
SAR of analogs containing cyanopyridine head
R3
O
As seen in Table 2, elongation of the tail linker of analog 4 to
two carbons afforded compound 6, which demonstrated compara-
ble antiviral activity, but at a cost of decreased metabolic stability.
Conversely, shortening of the tail linker in analog 7 resulted in loss
of activity, but did afford improved metabolic stability. Next, a
variety of substituted piperidine analogs were prepared and char-
acterized. Carbamate analog 8 maintained antiviral potency while
sulfonamide 9 demonstrated improved activity, but both were sig-
nificantly less stable. Incorporation of a polar urea functionality in
compound 10 improved metabolic stability, but at a price of anti-
viral potency. Improved permeability and decreased efflux was ob-
served with trans-4-methoxy and ethoxy cyclohexyl analogs 11
and 12, along with excellent antiviral activity. While the higher
lipophilicity for these compounds may be driving these improved
permeability characteristics, it also contributes to their unfavor-
able metabolic stability. The corresponding trans-4-hydroxy-cyclo-
hexyl analog 13 did have a favorable combination of antiviral
activity and metabolic stability. Unfortunately, limited permeabil-
ity and increased efflux are significant problems with this
compound.
Building on the trend that a number of analogs had promising
metabolic stability parameters but lacked the necessary antiviral
potency, a series of compounds were prepared by introducing a
cyanopyridine as a replacement for the pyrimidine head group.
Our previous studies with spiropiperidine CCR5 antagonists had
demonstrated that incorporation of this group had a beneficial ef-
fect on antiviral potency.
As can be seen in Table 3, replacement of the pyrimidine group
in compounds 4, 7, 8 and 13 by the corresponding cyanopyridine
head group (compounds 14–17) resulted in improved antiviral po-
tency of 4–90-fold. Even compound 18, where the tail group had
N
N
N
N
R2
N
O
CN
#
R2
R3 Binding
Antiviral
HLMa Cacoa
AB/ER
a
a
IC50
IC50
14
15
16
17
O
N
Ph 12
Ph 21
2
7
57
15
53
1/20
O
1/12
CONCH3 Ph 26
25
0.7/10
OH
Ph 33
Ph 24
3
34
44
0.4/47
0.6/25
18 Me
22
a
Defined as in Table 1.
been trimmed down to a simple methyl group demonstrated good
potency with an IC50 of 22 nM. Although inclusion of the cyanopyr-
idine group generally gave a slight decrease in metabolic stability,
analogs such as 15 had a good balance of metabolic stability and
antiviral activity.
As a next step, modifications of the tail side chain group were
investigated. Substitution on the phenyl ring of 4 with a variety
of groups such as chloro, fluoro and cyano all resulted in a dramatic
loss of antiviral activity. Similar loss of activity was observed by
replacement of the phenyl group with 2-pyridyl, 4-pyran or cyclo-
hexyl (data not shown). Recovery of good antiviral potency and
reasonable metabolic stability could be achieved by combination
of the pyridyl side chain and the cyanopyridine head group as seen

D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3219–3222
3221
Table 5
in Table 4 (compounds 19 and 20). Replacement of the phenyl
Pharmacokinetic profiles in rata,b
group with n-butyl resulted in a loss of antiviral potency as seen
with compounds 21 and 22. Introduction of the cyanopyridine
head group in 23 did bring back reasonable antiviral activity. The
poor correlation between CCR5 binding inhibition and antiviral
activity observed in analogs such as 7, 21 and 22 is not unusual
and has been previously observed.3,12 It most likely results from
differences in the allosteric modes of inhibition for chemokine ver-
sus viral binding.
A number of compounds with promising in vitro properties
were further profiled in single dose pharmacokinetic experiments
in rat and monkey as seen in Tables 5 and 6. In rat PK, analogs 3
and 15 (screened as the corresponding racemate) showed the most
promising bioavailability with moderate clearance and half life.
Subsequently, analog 4 (R enantiomer of 3) and 15 were screened
in monkey PK. Although compound 15 demonstrated a superior
profile, permeability and efflux are still limiting factors on
bioavailability.
*
PK
RLM7 Cmax (ng/
AUC (ng h/
Cl (mL/min/
kg)
T1/2
(h)
F%
mL)
mL)
3
14
56
21
43
38
189
49
44
65
16
29
1.5
0.8
5.7
1
27
10
23
19
12 89
15c 55
19 87
85
187
a
Doses were 1 mg/kg IV and PO.
Cmax, AUC and %F was determined after oral dosing. Cl, Vdss, t1/2 were deter-
b
mined from the iv dose.
c
Racemate.
Table 6
Pharmacokinetic profiles in cynomologous monkeya,b
*
PK
kFLM7 Cmax (ng/
AUC (ng h/
mL)
Cl (mL/min/
kg)
T1/2
(h)
F%
mL)
4
65
16
55
418
39
17
1.9
1.9
14
42
A particular off-target issue associated with CCR5 antagonists is
affinity for the hERG channel leading to a potential for cardiac
arrhythmias. Screening a set of representative examples from our
series in a PatchXpress clamp system13 showed a range of hERG
15c 33
133
a
Doses were 1 mg/kg IV and PO.
Cmax, AUC and %F was determined after oral dosing. Cl, Vdss, t1/2 were deter-
b
mined from the iv dose.
IC20 values from 0.3 to 30
hERG IC20 values of 22 and 3.5
l
M. Analogs 3 and 15 (racemate) showed
M, respectively.
c
Racemate.
l
A representative synthetic sequence for the preparation of our
series is outlined in Scheme 1 for the preparation of analog 4.14
Deprotection of ketal 2415 under acidic conditions afforded ke-
tone 25. Debenzylation of 25 under hydrogenolysis conditions in
the presence of Boc anhydride gave intermediate 26. The imidazo-
lidinone portion of the molecule was prepared starting from com-
mercially available (R)-(ꢁ)-2-amino-2-phenyl ethanol 27. Amine
protection with Boc anhydride afforded compound 28. Mesylation
followed by treatment with sodium azide gave the corresponding
azide 29. Removal of the Boc group of 29 with trifluoroacetic acid
followed by reductive amination of ketone 26 with sodium triacet-
oxyborohydride afforded compound 30. Catalytic hydrogenation of
azide 30 gave the corresponding amine 31, which was subjected to
reductive alkylation with 4-tetrahydropyran carboxaldehyde fol-
lowed by subsequent cyclization with triphosgene to give the
substituted imidazololidinone 32. Acidic treatment of 32 to remove
the Boc protecting group followed by amide formation with 4,6-
dimethylpyrimidine-5-carboxylic acid afforded target compound 4.
In summary, we have discovered a novel series of potent CCR5
small molecule antagonists, exemplified by analogs 4 and 15, pos-
sessing a promising antiviral, pharmacokinetic and off-target
profile.
Table 4
SAR of tail side chain analogs
R3
O
N
N
N
N
R1
R2
O
R1
R2
R3
Binding IC50
Antiviral IC50
22
HLMa
18
Cacoa AB/ER
0.1/77
a
a
#
O
N
N
19
12
NC
(R)
OEt
N
N
20
25
9
40
0.5/30
NC
(R)
O
21
22
nBu (rac)
43
84
335
36
9
0.4/40
0.4/52
N
N
N
N
O
O
nBu (rac)
>625
N
23
nBu (rac)
65
41
20
1/14
NC
a
Defined as in Table 1.

3222
D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3219–3222
O
O
N
Acknowledgements
O
O
The authors thank Gary Cooper, Ferenc Makra and Ann Petersen
a
b
N
N
N
for synthetic support; Rama Kondru for computational support;
Yanzhou Liu for expert NMR analysis; Marquis L. Cummings Jr.
for chiral HPLC purifications; Naina Patel for formulation develop-
ment; Joe Muchowski, Counde Oyang, David Goldstein, David
Swinney, Denis Kertesz and Eric Sjogren for helpful discussions.
N
boc
N
CH2Ph
CH2Ph
25
26
24
d
c
Boc-HN
Ph
H
References and notes
H2N
Ph
H
OH
OH
1. Kuritzkes, D.; Kar, S.; Kirkpatrick, P. Nat. Rev. Drug Disc. 2008, 7, 15.
2. Turner, J. E.; Steinmetz, O. M.; Stahl, R. A.; Panzer, U. Mini-Rev. Med. Chem. 2007,
7, 1089.
3. Rotstein, D. M.; Gabriel, S. D.; Makra, F.; Filonova, L.; Gleason, S.; Brotherton-
Pleiss, C.; Setti, L. Q.; Trejo- Martin, A.; Lee, E. K.; Sankuratri, S.; Ji, C.; deRosier,
A.; Dioszegi, M.; Heilek, G.; Jekle, A.; Berry, P.; Weller, P.; Maul, C.-I. Bioorg. Med.
Chem. Lett. 2009, 19, 5401.
4. (a) Zhou, Y.; Bourque, E.; Zhu, Y.; Langille, J.; Metz, M.; Yang, W.; Mceachern, E.
J.; Harwig, C.; Baird, I. R.; Li, T-S.; Skerlj, R. T. WO2006138350.; (b) Zhou, Y.;
Bourque, E.; Zhu, Y.; McEachern, E. J.; Harwig, C.; Skerlj, R. T.; Bridger, G. J.; Li,
T.-S.; Metz, M. WO2007022371.
5. Subsequent to our investigative efforts, related templates have been reported:
Bourque, E.; Metz, M.; Baird, I. R.; Yang, W.; Bridger, G.; Skerlj, R. T.
WO2008070758.
28
27
N3
NH2
Ph
H
N
H
N
Ph
Boc-NH
Ph
H
e
f
N3
+
N
N
29
N
boc
N
boc
6. For binding and antiviral assay protocols see: Melville, C. R.; Rotstein, D. M.
WO2008119663.
7. The thresholds above which intrinsic microsomal clearance values were
30
31
considered to be high were P35
ll/min/mL for HLM, P90 ll/min/mL for
RLM and P43 l/min/mL for kfLM (cynomologous monkey). These scaled to
O
l
O
g
N
N
in vivo clearance values P70% of liver blood flow using the well-stirred model.
Intrinsic microsomal clearance values corresponding to 630% of liver blood
N
Ph
Ph
O
H
O
N
H
flow (66.5
for kFLM) were considered to be low.
8. Obach, R. S. Curr. Opin. Drug Disc. Dev. 2001, 4, 36.
ll/min/mL for HLM, 615 ll/min/mL for RLM and (614 ll/min/mL
h
N
N
9. Van Breemen, R. B.; Li, Y. Expert Opin. Drug Metab. Toxicol. 2005, 1, 175.
10. The thresholds above which permeability was considered to be high in the
Caco-2 assay was >2 ꢀ 10ꢁ6 cm/s. Permeability was considered low for values
<0.5 ꢀ 10ꢁ6 cm/s.
N
N
boc
11. Screening in Caco-2 cells with Pgp inhibitors such as elacridar (gf12098) to
suppress active transport and assess intrinsic permeability is well established
(cf. Hochman, J. H.; Yamazaki, M.; Ohe, T.; Lin, J. H. Curr. Drug Metab. 2002, 3,
257.) Screening of our analogs in this way resulted in restoration of good
intrinsic permeability and elimination of efflux.
N
O
N
4
32
12. Ernst, J.; Dahl, R.; Lum, L.; Sebo, L.; Urban, J.; Miller, S. G.; Lundstrom, J. Bioorg.
Med. Chem. Lett. 2008, 18, 1498.
Scheme 1. Reagents and conditions: (a) (i) concd HCl, MeOH, reflux; (ii) 1 N HCl,
reflux 65%; (b) H2, Pd(OH)2, (Boc)2O, MeOH, 87%; (c) (Boc)2O, 1 M NaOH, Et2O, 97%;
(d) (i) MsCl, TEA, CH2Cl2; (ii) NaN3, DMF, 60 °C, 67% for two steps; (e) (i) TFA,
CH2Cl2; (ii) 26, NaBH(O2CCH3)3, ClCH2CH2Cl, 76% for two steps; (f) H2, 5% Pd/C,
MeOH, 98%; (g) (i) 4-tetrahydropyrancarboxaldehyde, NaBH(O2CCH3)3,
ClCH2CH2Cl; (ii) (CCl3O)2CO, pyridine, CH2Cl2, 79% for two steps; (h) (i) TFA, CH2Cl2,
99%; (ii) 4,6-dimethylpyrimidine-5-carboxylic acid, HOBT, EDAC, DIEA, DMF, 60%.
13. Ly, J. Q.; Shy, G.; Misner, D. L. Clin. Lab. Med. 2007, 27, 201.
14. For more detailed experimental information, see: Gabriel, S. D.; Lemoine, R.;
Rotstein, D. M.; Wanner, J. WO2009013171.
15. Thoma, G.; Nuninger, F.; Schaefer, M.; Akyel, K. G.; Albert, R.; Beerli, C.;
Bruns, C.; Francotte, E.; Luyten, M.; MacKenzie, D.; Oberer, L.; Streiff, M. B.;
Wagner, T.; Walter, H.; Weckbecker, G.; Zerwes, H.-G. J. Med. Chem. 2004,
47, 1939.