Novel hexahydropyrrolo[3,4-c]pyrrole CCR5 antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.03.095

Starting with a high-throughput screening lead, a novel series of CCR5 antagonists was developed utilizing an information-based approach. Improvement of pharmacokinetic properties for the series was pursued by SAR exploration of the lead template. The synthesis, SAR and biological profiles of the series are described.

Full text of DOI:10.1016/j.bmcl.2010.03.095

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3116–3119
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel hexahydropyrrolo[3,4-c]pyrrole CCR5 antagonists
*
David M. Rotstein , Chris R. Melville, Fernando Padilla, Dick Cournoyer, Eun K. Lee, Remy Lemoine,
Ann C. Petersen, Lina Q. Setti, Jutta Wanner, Lijing Chen, Lubov Filonova, David G. Loughhead, Jason Manka,
Xiao-Fa Lin, Shelley Gleason, Surya Sankuratri, Changhua Ji, Andre deRosier, Marianna Dioszegi,
Gabrielle Heilek, Andreas Jekle, Pamela Berry, Cheng-I Mau, Paul Weller
Roche Palo Alto LLC, 3431 Hillview Avenue, Palo Alto, CA 94304, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Starting with a high-throughput screening lead, a novel series of CCR5 antagonists was developed utiliz-
ing an information-based approach. Improvement of pharmacokinetic properties for the series was pur-
sued by SAR exploration of the lead template. The synthesis, SAR and biological profiles of the series are
described.
Received 26 February 2010
Revised 23 March 2010
Accepted 26 March 2010
Available online 30 March 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
CCR5 antagonist
Chemokine
Anti-HIV-1
Antiviral
Despite remarkable advances in the treatment of HIV/AIDS, a
major challenge is the rapid evolution of resistant viral strains.
Thus, there remains a need for the development of new therapeutic
agents.1 As one of the two key co-receptors for HIV entry, blockade
of CCR5 continues to be an area of therapeutic interest.2 In a con-
tinuation of our efforts towards the discovery of CCR5 antagonists
for the treatment of HIV/AIDS3 and potentially for a variety of
autoimmune indications,4 studies of new, novel templates were
pursued.
Examination of hits from high-throughput screening of the
Roche library versus the CCR5 receptor revealed compound 1 as
a low potency antagonist. Comparison of this hit to the Pfizer clin-
ical candidate, maraviroc 2,5 suggested that the hexahydropyrrol-
o[3,4-c]pyrrole group present in 1 might serve as a bioisostere
for the tropane subunit in maraviroc. A similar comparison to the
Schering series of CCR5 antagonists,6 represented by structure 3,
suggested replacement of the dichlorobenzyl group of 1 with the
2,6-dimethyl benzamide group of 3. Figure 1 illustrates this strat-
egy and the first example of our morphed template, compound 4.
As can be seen in Table 1, compound 4 demonstrated excellent
binding and antiviral activity with IC50 values of 6 and 7 nM, respec-
tively. These are comparable to both Pfizer’s maraviroc 2 (binding
and antiviral IC50 of 14 and 3 nM, respectively) and Schering analog
3 (binding and antiviral IC50 of 6 and 4 nM, respectively).7
A major issue to be addressed with this lead compound was the
high intrinsic clearances observed in human (HLM) and rat (RLM)
liver microsomes: 256 and 312 l
l/min/mg, respectively.8 These
in vitro values corresponded to projected in vivo clearances close
to hepatic blood flow.9 Subsequent rat PK experiments were
F
F
O
O N
N
N
N
N
O
N
N
N
Br
3
2
O
N
H
N
N
O
H
4
H
N
N
O
N
H
1
Cl
Cl
* Corresponding author. Tel.: +1 408 7329030.
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.2010.03.095

D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3116–3119
3117
Table 1
Table 2
Tail SAR of lead compound 4
Head SAR of lead compound 4
R1
R1
O
N
O
N
N
N
N
O
R2
N
O
R2
#
clog P R1
R2
Binding Antiviral HLM/
Cacoc
AB/ER
RLMb
#
R1
R2
Bindinga IC50
6
Antiviral/b IC50
7
a
a
IC50
IC50
256/
322
4
4
4.7
6
465
9
7
4/5
ND
ND
13 2.1
14
Me
>1000
3
ND
5
6
CH3
15
11
7
132
100
39
264/
ND
5
Cl
Cl
389/
ND
15 5.2
14
6
92
3
ND
ND
7
Cl
8
3
30
236/
517
16
17
5
5
F
F
F
9
4
9
241/
236
26
0.4
33/2
10
11
12
8
124
85
MeO
18
15
O
O
18 2.8
19 3.6
19
7
545
7
ND
ND
N
O
101
67/89
4/17
N
a
Competitive binding evaluated versus RANTES with IC50 values in nM, as mean
of two experiments.
b
20 2.3
21 2.7
22 2.5
44
14
30
19
27
6
6/12
.1/300
1.3/17
0.5/5
N+
Antiviral IC50 values in nM as mean of two experiments.
O
consistent with this projection: only 3% bioavailability was ob-
tained upon PO dosing at 10 mg/kg. Thus, a SAR study of this
template was initiated, seeking to improve our lead’s DMPK
characteristics, while maintaining good antiviral activity.
35/51
25/25
N
N
N
In describing results from our studies, the cyclopentyl and
benzamide groups are typically referred to as the tail and head
groups of the template, respectively. In our series, as has previously
observed,10 good antagonist binding to the CCR5 receptor was
maintained across a range of structural modifications. In contrast,
antiviral potency in cells was more sensitive to structural changes.
Initial SAR studies of the tail group show a preference for cyclic
aliphatic rings. As seen in Table 1, replacement of the cyclopentyl
tail group with methyl or isopropyl, as in 5 and 6, led to significant
loss in antiviral potency. Shrinkage of cyclopentyl to cyclobutyl
or enlargement to cyclohexyl (analogs 7 and 8) led to ꢀ5-fold
loss in antiviral activity. However, incorporation of a 4,4-dif-
luorocyclohexyl tail restored antiviral activity as seen with analog
9. Compounds 10–12 demonstrate that aryl, heteroaryl and hetero-
cycle functionalities were not well tolerated in the tail region. Met-
abolic stability assessment of analogs from Table 1 showed no
improvement in intrinsic clearance (data not shown).
Table 2 describes results from studies of head group SAR. Dele-
tion of the benzamide group, as in acetyl analog 13, resulted in
complete loss of binding and antiviral activity. Removal of either
or both of the 2,6-dimethyl flanking groups gave a similar result
(data not shown), however, replacement with halogen was toler-
ated as can be seen with compound 14. The 2,6-dichloro analog
15 maintained binding activity but lost significant antiviral po-
tency. Additional 4-substitution on the aryl ring was tolerated, as
demonstrated by 4-fluoro and 4-methoxy analogs (16 and 17). Aryl
replacement with a five member heteroaromatic such as isoxazole
18, resulted in a loss of antiviral activity. A variety of six member
O
227/
144
23 3.4
12
0.4
3/8
N
NC
a
b
c
Defined as in Table 1.
Human liver microsomal intrinsic clearance (ll/min/mg protein).
Permeability in Caco-2 cells (cm/sec  10EÀ6) apical to basolateral (AB) and
basolateral to apical (BA) movement of test compound in 21-day cultured Caco-2
cells. ER = efflux ratio of BA to AB.
heteroaromatic rings such as pyridine, pyridine N-oxide, pyrimi-
dine, pyridone and 6-cyanopyridine maintained good activity
(19–23). Another beneficial feature of these more polar analogs is
that the heteroatom substitution improved metabolic stability by
lowering overall lipophilicity and blocking specific sites of metab-
olism. However, a downside of the decreased lipophilicity is that,
in general, the resultant molecules demonstrated decreased per-
meability and enhanced efflux in Caco-2 cells (contrast analogs 4
vs 19 and 20).
A number of these analogs were screened to assess their in vivo
PK in rat. As can be seen in Table 3, the increased metabolic stabil-
ity of analogs such as 19 and 21 resulted in somewhat improved
bioavailability relative to our starting point with compound 4.
However, no effect was seen on the high systemic clearance as
the structural modifications have not addressed the primary driver
of clearance, PgP mediated efflux.
Additional SAR studies were performed around advanced lead
21 to address these DMPK issues and optimize potency. As can

3118
D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3116–3119
Table 3
Table 5
Pharmacokinetic profile in rata
Pharmacokinetic profile in rat and dog
Compd #
AUC PO (ngÁh/mL)
31
65
Cl IV (mL/min/kg)
%F
2.7
6.4
17
Pharmacokineticsa
24
215
99
15
28
4
19
21
134
151
131
Rat AUC (ng h/mL)
Rat Cl (ml min/kg)
Rat %F
108
105
18
309
Dog AUC (ng h/mL)
Dog Cl (ml min/kg)
Dog %F
4210
44
103
307
36
72
a
Doses were 1 mg/kg IV and 10 mg/kg PO, AUC and %F were determined after the
oral dose and Cl was determined from the IV dose.
a
IV/PO doses 1/10 mg/kg except for compound 28 in dog where IV/PO doses
be seen in Table 4, replacement of the cyclopentyl tail of 21, with
4,4-difluoro cyclohexyl or 3,3-difluorocyclobutyl afforded analogs
24 and 25, both of which showed improved antiviral activity, but
at a cost of microsomal stability. Substitution of a 3-fluoro or
3-chloro on the phenyl tail also increased antiviral potency (26–
29). However, the 3,5-difluorophenyl or 3-cyanophenyl analogs,
30–31, showed decreased antiviral activity.
As detailed in Table 5, screening of analogs 24 and 28 in rat PK
failed to show significant improvement in exposure. Interestingly,
when these compounds were administered to dogs, much im-
proved exposure was observed compared to rat despite similar
were 1/1 mg/kg. AUC and %F were determined after the oral dose and Cl was
determined from the IV dose.
paraformaldehyde gave 2,5-dibenzyltetrahydropyrrolo[3,4-c]pyr-
role-1,3-dione 34. Selective debenzylation with 1-chloroethyl
chloroformate, followed by imide reduction with lithium aluminum
hydride afforded 2-benzyloctahydro-pyrrolo[3,4-c]pyrrole 35.
((S)-3-Oxo-1-phenylpropyl)-carbamic acid tert-butyl ester 37
was prepared by BOC protection of (S)-3-amino-3-phenylpropionic
acid ethyl ester 36 using BOC anhydride, followed by chemoselec-
tive reduction of the ester to the corresponding aldehyde 37.
Reductive amination of amine 35 with aldehyde 37 using so-
dium triacetoxyborohydride gave intermediate 38. BOC removal
with HCl was followed by amide coupling with cyclopentanecarb-
oxylic acid, EDCI, HOBT and diisopropylethylamine to yield 39.
Debenzylation of 39 under catalytic hydrogenolysis conditions
afforded the respective amine, which was followed by amide for-
mation with 2,6-dimethylbenzoic acid to afford the desired prod-
uct, cyclopentanecarboxylic acid {(S)-3-[5-(2,6-dimethylbenzoyl)
hexahydropyrrolo[3,4-c]pyrrol-2-yl]-1-phenylpropyl}amide, 4.
In summary, building from an in-house lead, an information-
based approach led to the rapid discovery of a series of novel
CCR5 antagonists with potent antiviral activity. Extensive SAR
studies afforded analogs with improved in vitro and in vivo DMPK
properties.
intrinsic microsomal stabilities (DLM clearance is 30 and 15 ll/
min/mg, respectively, for 24 and 28). One potential explanation
for this difference is the increased contribution of paracellular
absorption through the larger pore junctions in the dog gastro-
intestinal tract.
An example of the synthetic route utilized to prepare our series,
the synthesis of analog 4, is shown in Scheme 1.11
A [2,3]-dipolar cycloaddition between N-benzylmaleimide 32
and the imine ylide derived from N-benzylglycine 33 and
Table 4
Tail SAR of lead compound 21
R1
O
N
N
N
O
R2
R3
O
N Bz
O
N Bz
a
CO2H
R1
R2
R3
Binding
IC50
Antiviral
IC50
HLM/
RLM
Bz
N
Bz
N
+
#
a
a
O
32
O
33
34
H
b
21
24
25
26
27
28
29
30
31
H
14
27
20
13
12
9
27
6
35/51
83/46
60/61
129/17
113/24
28/29
N
N
N
N
N
N
N
N
N
N
N Bz
35
N
N
N
N
N
N
N
N
N
H
F
F
H
NH HCl
2
NHBoc
CHO
CO Et
c
2
F
37
H
3
36
F
NHBoc
e
d
H
N
35
37
+
3-Fl
3-Fl
3-Fl
3-Cl
3.6
1.1
0.8
2
38
N
N
Bz
H
F
F
f
O
N
F
O
N
H
H
F
N
N
O
N
H
F
F
Bz
H
182/
111
4
39
25
30
34
F
F
3,5-
Di-Fl
Scheme 1. Reagents and conditions: (a) (CH2O)n, toluene, reflux, 86%; (b) (i) 1-
chloroethyl chloroformate, CH2Cl2, reflux, 91%, (ii) LAH, THF, 70 °C, 95%; (c) (i)
NaOH, (BOC)2O, H2O, THF, (ii) DIBAL, CH2Cl2, À78 °C, 87%; (d) sodium triacetoxy-
borohydride, CH2Cl2, rt, 88%; (e) (i) HCl, MeOH, 50 °C, (ii) cyclopentanecarboxylic
acid, EDCI, HOBt, i-Pr2NEt, CH2Cl2, rt, 94% for two steps; (f) (i) ammonium
carbonate, Pd(OH)2, EtOH, reflux, (ii) 2,6-dimethylbenzenecarboxylic acid, EDCI,
HOBT, i-Pr2NEt, CH2Cl2, 56% for two steps.
41
203
45/ND
30/31
F
F
3-CN
a
Defined as in Tables 1 and 2.

D. M. Rotstein et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3116–3119
3119
4. Turner, J. E.; Steinmetz, O. M.; Stahl, R. A.; Panzer, U. Mini-Rev. Med. Chem. 2007,
7, 1089.
Acknowledgements
5. MacArthur, R. D.; Novak, R. M. Rev. Anti-Infect. Agents 2008, 47, 236.
6. Palani, A.; Shapiro, S.; Clader, J. W.; Greenlee, W. J.; Cox, K.; Strizki, J.; Endres,
M.; Baroudy, B. M. J. Med. Chem. 2001, 44, 3339.
The authors thank Gary Cooper, Felicia Thai, Bill Bingenheimer
and Eric Humphreys for synthetic support; Rama Kondru, Nidhi
Arora and Taraneh Mirzadegan for computational support; Yanz-
hou Liu for expert NMR analysis; Marquis L. Cummings Jr. for chiral
HPLC purifications; Naina Patel for formulation development; Joe
Muchowski, Counde Oyang, David Goldstein, David Swinney, Denis
Kertesz and Eric Sjogren for helpful discussions.
7. Briefly, for the antiviral assay, pseudo-type particles were generated by co-
transfection of envelope deleted HIV-1 genomic plasmid (derived from pNL4-3)
and an envelope expression plasmid carrying viral envelope from NL-Bal into
293T cells. Culture supernatant containing pseudo-typed viruses was
harvested after 2 days and used to infect JC53-BL reporter cells. Luciferase
reporter gene activity was measured 3 days post-infection. For more detailed
assay protocols see: Melville, C.R.; Rotstein, D.M. WO2008119663.
8. The thresholds above which intrinsic mircrosomal clearance values were
considered to be high were P90
ll/min/mL for RLM and P35 ll/min/ml for
HLM. These scaled to in vivo clearance values P70% of liver blood flow using
References and notes
the well-stirred model. Intrinsic microsomal clearance values corresponding to
630% of liver blood flow (615 ll/min/ml for RLM and 66.5 ll/min/ml for HLM)
were considered to be low.
1. Perno, C.-F.; Moyle, G.; Tsoukas, C.; Ratanasuwan, W.; Gatell, J.; Schechter, M. J.
Med. Virol. 2008, 80, 565.
2. Palani, A.; Tagat, J. R. J. Med. Chem. 2006, 49, 2851.
9. Obach, R. S. Curr. Opin. Drug Discovery Dev. 2001, 4, 36.
3. Rotstein, D. M.; Gabriel, S. D.; Makra, F.; Filonova, F.; Gleason, S.; Brotherton-
Pleiss, C.; Setti, L. Q.; Trejo-Martin, A.; Lee, E. K.; Sankuratri, S.; Ji, C.; de Rosier,
A.; Dioszegi, M.; Heilek, G.; Jekle, A.; Berry, P.; Weller, P.; Mau, C.-I. Bioorg. Med.
Chem. Lett. 2009, 19, 5401.
10. Ernst, J.; Dahl, R.; Lum, C.; Urban, J.; Miller, S. G.; Lundström J. Bioorg. Med.
Chem. Lett. 2008, 18, 1498.
11. For more detailed experimental information, see: Lee, E.K.; Melville, C. R.;
Rotstein, D.M. WO2005121145.