Renin inhibitors for the treatment of hypertension: Design and optimization of a novel series of spirocyclic piperidines

Article abstract of DOI:10.1016/j.bmcl.2011.10.013

The discovery and SAR of a novel series of spirocyclic renin inhibitors are described herein. It was found that by restricting the northern aromatic plate to the bioactive conformation through spirocyclization, increase in renin potency and decrease in hE

Full text of DOI:10.1016/j.bmcl.2011.10.013

Bioorganic & Medicinal Chemistry Letters 21 (2011) 7399–7404
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Renin inhibitors for the treatment of hypertension: Design and optimization
of a novel series of spirocyclic piperidines
⇑
Austin Chen , Renee Aspiotis, Louis-Charles Campeau, Elizabeth Cauchon, Amadine Chefson,
Yves Ducharme, Jean-Pierre Falgueyret, Sébastien Gagné, Yongxin Han, Robert Houle, Sébastien Laliberté,
Guillaume Larouche, Jean-François Lévesque, Dan McKay, David Percival
Merck Frosst Centre for Therapeutic Research, 16711 Trans Canada Highway, Kirkland, Québec, Canada H9H 3L1
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery and SAR of a novel series of spirocyclic renin inhibitors are described herein. It was found
that by restricting the northern aromatic plate to the bioactive conformation through spirocyclization,
increase in renin potency and decrease in hERG affinity could both be realized. When early members
of this series were found to be potent time-dependent CYP3A4 inhibitors, two distinct strategies to
address this liability were explored and this effort culminated in the identification of compound 31 as
an optimized renin inhibitor.
Received 16 August 2011
Revised 30 September 2011
Accepted 4 October 2011
Available online 19 October 2011
Keywords:
Renin
Ó 2011 Elsevier Ltd. All rights reserved.
Inhibitor
Hypertension
CYP
hERG
Spirocyclization
The renin-angiotensin-aldosterone system (RAAS, Fig. 1) is a
tightly regulated biological cascade known to play a pivotal role
in the control of blood pressure.1 With more than a billion people
worldwide suffering from primary hypertension, research towards
the identification of novel agents capable of perturbing this impor-
tant cascade at its various points have the potential to further
expand on our existing anti-hypertensive armatorium.2 Since the
renin enzyme is involved in the first and rate-limiting step of RAAS,
its inhibition should offer the best potential for blood pressure con-
trol as well as end organ protection.3 Furthermore, given that the
only known function of the renin enzyme involves the processing
of the biologically inactive angiotensinogen, renin blockade should
not be plagued by mechanism-based adverse events.4
Recently, we reported the discovery of a novel series of potent
renin inhibitors displaying a piperidine warhead substituted at
the 4-position by a tertiary alcohol and a suitably functionalized
arene (e.g., 1, Fig. 2).5 Despite the fact that these compounds were
found to be substrates of the endogenous human and rat P-gp
transporters, their high cellular permeability allowed these com-
pounds to effectively overwhelm the above efflux mechanism
and as a result, good oral bioavailability in rats was achieved across
all the doses examined.6 However, most of these renin inhibitors
were also found to be potent inhibitors of CYP 3A4 and to exhibit
⇑
Corresponding author.
E-mail address: austin.chihyu.chen@gmail.com (A. Chen).
Figure 1. The renin-angiotensin-aldosterone system.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.10.013

7400
A. Chen et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7399–7404
Figure 2. Representative piperidine-based renin inhibitors.
tuted fluorine to engage in a highly favorable dipole–dipole
interaction with renin’s Trp45 residue. On the other hand, the ter-
tiary alcohol points out into solvent and does not appear to partic-
ipate in any interaction with the renin enzyme. Consequently we
hypothesized that by cyclizing the free alcohol onto the aromatic
plate, one would restrict the molecule to its bioactive conformation
which in turn, can concurrently improve the intrinsic potency
against renin and decrease its affinity for other undesired targets.
With compound 2 as the reference point, we were gratified to
see that by simply tethering the tertiary alcohol to its geminal
phenyl ring with a methylene linker (i.e., 3), a concomitant
improvement in both intrinsic renin potency and off-target profile
(i.e., hERG affinity and CYP3A4 inhibition) could be realized (Table
1). However, the exact consequence observed is intimately depen-
dent on the nature of the spirocycle. Tethering, for example, with a
carbonyl linker (i.e., 4) in fact led to a drop in renin potency,
although its off target profile was again improved versus the
non-spirocyclized analogue 2. Introduction of greater flexibility
to the spirocycle via ring expansion led to a decrease in renin po-
tency when the spirocycle in question is a cyclic ether (i.e., 3 vs
5). On the other hand, ring expansion was found to be beneficial
for renin potency when the spirocycle in question is a cyclic lac-
tone (i.e., 4 vs 7). Unfortunately, compound 7 was a moderate
time-dependent CYP3A4 inhibitor. Although the analogue bearing
an isochromene spirocycle (i.e., 6) proved to be optimal for renin
potency, its off-target profile was found to be inferior to compound
5. Consequently, since the tetrahydrofuran spirocycle 3 appeared
to offer the best compromise between renin potency and off-target
profile, its 5,6-difluoro analogue 8 was synthesized. Indeed, the
Figure 3. Modeling of compound 1 in the renin active site.
worrisome affinity for the hERG channel. Both of these liabilities in
turn rendered their clinical development a highly risky endeavor.
When compound 1 was docked into the renin active site (Fig. 3),
the molecule adopts a conformation where-by the plane of its
3,4-difluoroaromatic plate bisects the piperidine ring to which
the arene is attached. In this orientation, a stabilizing p-stack be-
tween the fluorinated arene and renin’s Tyr83 residue is formed.
Furthermore, this conformation also positioned the meta-substi-
Table 1
SAR of select renin inhibitors: Northern modifications
a
Renin Buffer IC50 (nM)
320
990
5
150
1200
20
6
11
1400
1900
19
7.2
16
390
1200
24
6.4
0
40
230
7
1.1
18
90
290
43
5.6
41
5
a
Renin Plasma IC50 (nM)
33
24
13
15
hERG Ki (lM)
CYP 3A4 reversible inhibition IC50
(lM)
2.6
CYP 3A4 time-dependent inhibition (% activity loss @ 10
l
M after 13
30 min incubation)b
a
Average of at least two replicates. See Ref. 7 for assay protocols.
Calculated as a percentage of the difference in rate in the CYP3A4-mediated conversion of testosterone (250
b
lM) to 6-b-hydroxytestosterone before and after 30 min
incubation period with the compound (10
lM in DMSO). A 0% change corresponds to no measurable time-dependent CYP 3A4 inhibition.

A. Chen et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7399–7404
7401
installation of these fluorine substituents improved the intrinsic
renin potency by 30-fold without sacrificing the gains we have
made in the off-target profile. Furthermore by comparing the key
profiles of compounds 1 and 8, the virtue of spirocyclization was
once again made apparent.
Having settled on the 5,6-difluoro-1,3-dihydro-2-benzofuran
spirocycle, an optimization campaign of the benzyl amide handle
was then initiated. The requisite chiral amide 9 could be accessed
using the chemistry shown in Scheme 1. Commercially available
2-bromo-4,5-difluorobenzoic acid (i.e., 10) was first reduced to
the corresponding alcohol 11 with borane dimethyl sulfide com-
plex and then protected as its TBS ether. Aryl bromide 12 was then
coupled with known boronate 138 using the standard Suzuki con-
ditions.9 After treatment with TBAF, the newly liberated alcohol
underwent spontaneous 5-exo-trig cyclization to furnish ester 15
as a single diastereomer.10 Ester 15 could then be converted to
the corresponding amide by heating with the requisite magnesiate
prepared in situ from isopropyl magnesium chloride and cyclopro-
pylamine.11 Racemic 15 could then be resolved on a chiral AD col-
umn with 80:10:10 (v/v/v) hexanes–ethanol–i-propanol as the
eluent.
All the amides shown in Table 2 could be most readily accessed
from 9 using a two-step sequence: N-alkylation with sodium hy-
dride and an appropriate benzyl bromide in THF, followed by re-
Scheme 1. Synthesis of 9: (a) BH3ꢀDMS, THF, reflux, 2.5 h, >99%; (b) TBSCl, Imidazole, DMF, rt, 16 h, 96%; (c) Pd(dppf)Cl2, 2 N aq Na2CO3, 13, 2:1 (v/v) toluene–ethanol, 80 °C,
14 h, 70%; (d) TBAF, THF, rt, 1 h, 49%; (e) iPrMgCl, cyclopropylamine, THF, 50 °C, 16 h, 93% (f) chiral resolution, AD column, 80:10:10 (v/v/v) hexanes–EtOH–iPrOH.
Table 2
SAR of select renin inhibitors: Amide modifications
Compd
R1
R2
R3
R4
Renin Potencya
Buffer (nM) Plasma (nM)
hERG Ki (
l
M)
CYP 3A4 inhibition
Reversible IC50
(lM)
time dependent (% loss)b
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Cl
Cl
Cl
Cl
Me
Me
Cl
Me
H
Cl
H
H
Me
CF3
CF3
Me
—
—
—
—
—
H
H
Cl
H
H
H
H
H
—
—
—
—
—
—
H
H
H
Cl
H
H
H
7.1
0.2
1.3
3.7
0.11
0.3
0.5
0.18
0.16
0.09
<0.01
1.0
45
5.4
39
20
11
5.0
5.6
1.9
1.6
5.0
2.3
3.6
3.5
6.0
1.6
15
92
89
48
76
69
72
80
50
55
48
69
55
9
9.3
1.8
27
6.4
12
10
5.7
12
175
0.5
5.4
11
2.7
0.8
5.2
0.5
1.5
7.1
2.5
(CH2)3OMe
H
(CH2)3OMe
H
—
—
—
H
H
(CH2)3OMe
7.6
>60
55
—
—
—
9.3
31
8.9
4.2
0.7
—
29
69
a
Average of at least two replicates. See Ref. 7 for assay protocols.
Calculated as a percentage of the difference in rate in the CYP3A4-mediated conversion of testosterone (250
b
lM) to 6-b-hydroxytestosterone before and after 30 min
incubation period with the compound (1 0
lM in DMSO). A 0% change corresponds to no measurable time-dependent CYP 3A4 inhibition.

7402
A. Chen et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7399–7404
moval of the BOC protecting group. The latter step could either be
performed in the presence of a large excess of 4 M HCl in dioxane
or with zinc(II) bromide in CH2Cl2.12
renin potency. While other quinoline and isoquinoline isomers of
28 were found to be better inhibitors of renin (i.e., 27, 29), these
analogues were also more potent reversible and time-dependent
inhibitors of CYP3A4.
An equipotent renin inhibitor was obtained when the 3,5-disub-
stituted benzyl amide in 8 was replaced with a simple 2-chloro-
benzyl amide (i.e., 16). However, the resulting compound was
found to inhibit CYP3A4 in a time dependent manner and hence
may demonstrate the potential to perpetrate drug–drug interac-
tions when co-dosed with drugs cleared mainly through CYP3A4
metabolism.13 Although the intrinsic renin potency could be fur-
ther enhanced via the addition of an extra chlorine at either the
3- (i.e., 17), 4- (i.e., 18) or 5-position (i.e., 19), these modifications,
unfortunately, all failed to improve the CYP3A4 profile. When both
chlorines in compound 17 were replaced with simple methyl resi-
dues (i.e., 20), the renin IC50 became much less shifted in the pres-
ence of human plasma. The same improvement could be realized
when the 2,3-dichlorobenzene in 17 was replaced with its known
naphthalene isostere (i.e., 24). However neither variation success-
fully addressed the key time-dependent CYP3A4 inhibition (TDI)
issue. Previously, it has been shown that the judicious incorpora-
tion of a metabolically labile 3-methoxypropyl appendage onto
the benzene ring may help to decrease TDI by diverting the metab-
olism to a more innocuous area in the molecule.13 Unfortunately,
this strategy failed to pay significant dividends in the current series
of renin inhibitors (i.e., 23 vs 20, 25 vs 24, 26 vs 24). The first break-
through in our attempt to dial down TDI was realized when the 1-
naphthalene ring in 24 was changed to a 4-quinoline (i.e., 28). This
improvement however, was accompanied by a significant loss in
To determine whether the observed TDI was due to the forma-
tion of an active metabolite capable of binding irreversibly to the
CYP enzyme, compound 20 was incubated at 1 lM with either hu-
man or rat liver microsomes (0.25 mg/mL, in the presence of
NADPH regenerating system). After 30 min at 37 °C, only 1% and
0%, respectively, of compound 20 remained intact. Subsequent
metabolite identification using freshly harvested rat hepatocytes
(10
l
M of 20, 1 ꢁ 106 cells/mL, 2 h, 37 °C) revealed that the eastern
benzyl amide suffered extensive oxidative metabolism (Fig. 4).
Since M2, the putative acid metabolite of compound 20, was most
likely formed from alcohol M1 or M3 by way of an reactive alde-
hyde intermediate, the suspected metabolic soft spots were then
blocked with halogens to afford 22. These alterations not only
led to a drop in renin potency but they offered no tangible benefit
in terms of TDI (Table 2). Consequently, compound 22 was also
incubated with human liver microsomes. While we were gratified
to see that half of compound 22 now survived the incubation intact
(vs 1% for compound 20), the site of metabolism has shunted from
the benzyl amide to the tetrahydrofuran spirocycle. Furthermore
two of the metabolites identified, namely M10 and M30,14 appear
to possess the potential to bind irreversibly to CYPs. As a result,
the lactone analogue of 22 (i.e., 30) was synthesized and profiled.
Once again, this modification failed to improve either renin po-
tency (renin IC50 = 3.9 nM (buffer), 49 nM (plasma)) nor off-target
Figure 4. In vitro metabolic profile of select renin inhibitors.

A. Chen et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7399–7404
7403
Table 3
SAR of select renin inhibitors: incorporation of polarity
Compound
R4
Renin Potencya
Buffer (nM) Plasma (nM)
hERG Ki (lM)
CYP 3A4 inhibition
Reversible IC50 (lM)
Time dependent(% loss)b
31
32
0.07
0.3
0.4
0.6
23
12
2.6
5.8
18
17
33
34
35
0.08
0.1
0.4
0.3
0.5
8.2
22
2.2
2.0
2.4
17
22
36
0.05
7.7
36
0.5
0.9
47
2.8
66
a
Average of at least two replicates. See Ref. 6 for assay protocols.
Calculated as a percentage of the difference in rate in the CYP3A4-mediated conversion of testosterone (250
b
l
M) to 6-b-hydroxytestosterone before and after 30 min
incubation period with the compound (10
lM in DMSO). A 0% change corresponds to no measurable time-dependent CYP 3A4 inhibition.
quently, analogues bearing a variety of polarity-bearing appendage
Table 4
Pharmacokinetic profiles of compounds 31
were synthesized and profiled (Table 3). We were gratified to see
that most of these modifications (i.e., 31–35) were indeed success-
ful at lowering the extent of time-dependent CYP3A4 inhibition to
a more reasonable level while maintaining potency against renin,
although an exception does exist (i.e., 36).
Since compound 31 appeared to offer the optimum balance be-
tween renin potency and off-target profile, it was selected for fur-
ther characterization (Table 4). While the oral bioavailability was
found to be moderate in SD rats when compound 31 was dosed
at 3 mpk as its bis HCl salt (F = 23%), the plasma exposure achieved
was however unacceptably low in both beagle dogs (F = 1%) and
rhesus monkeys (F = 1%).16 Furthermore, because this dichotomous
pharmacokinetic behavior between rodents and higher order spe-
cies was also seen with other polarity-bearing analogues (data
not shown), this series has been put on hold.
SD Rat (3 mpk po, 0.5 mpk iv)a
F (%)
po AUC (lM h)
Cl (mL/min/kg)
T1/2 (h)
Vdss (L/kg)
F (%)
po AUC (
Cl (mL/min/kg)
T1/2 (h)
Vdss (L/kg)
F (%)
23
0.37
47
5.3
24
1
0.01
73
10
73
1
Beagle Dog (3 mpk po, 1 mpk iv)a
Rhesus Monkey (3 mpk po, 1 mpk iv)a
l
M h)
po AUC (
Cl (mL/min/kg)
T1/2 (h)
Vdss (L/kg)
l
M h)
0.01
66
5
45
a
Compound 31 was dosed as the HCl salt in either 0.5% aqueous methocel (5 mL/
In summary, we have described the discovery and SAR of a no-
vel series of potent, spirocyclic renin inhibitors. Although early
members of this series were found to be potent time-dependent
inhibitors of CYP3A4, the judicious incorporation of polarity
proved to be successful at dialing down this liability. Unfortu-
nately, these latter compounds were found to be poorly bioavail-
able in dogs and monkeys.
kg, po) or 60% aqueous PEG200 (1 mL/kg, iv).
profile (hERG Ki = 13
lM, reversible CYP3A4 inhibition IC50 =
0.7 M, 87% CYP3A4 activity lost after 30 min incubation period).
l
Finally when compound 30 was incubated under the same condi-
tion as described above, we discovered that the site of metabolism
has shifted once more; this time to the piperidine warhead. It also
warrants mentioning that with the lactone spirocycle, the two aryl
fluorides are now susceptible to hydrolysis to afford the corre-
sponding catechol (i.e., M500).
At this juncture, a different strategy was required. Rather than
continuing to chase down each and every metabolic soft spot in
this series of renin inhibitors, we instead felt it was more beneficial
to decrease the overall metabolic stress on these compounds
through the incorporation of polarity. Previously, we have shown
that a small, solvent-exposed channel is present in the renin S3
pocket through which a linear aliphatic chain of appropriate length
can be threaded.15 Furthermore, this chain can be capped with a
polar group with no adverse impact on the renin potency. Conse-
References and Notes
1. Zaman, M. A.; Oparil, S.; Calhoun, D. A. Nat. Rev. Drug Discovery 2002, 1, 621.
2. Rosenburg, S. H.; Boyd, S. A. In Antihypertensive Drugs; Van Zweiten, P. A.,
Greelee, W. J., Eds., 1998; pp 77–111.
3. Brunner, H. R. Am. J. Hypertens. 1992, 5, 243S.
4. Cooper, M. E. Am. J. Hypertens. 2004, 17, 16S.
5. Chen, A.; Cauchon, E.; Chefson, A.; Dolman, S.; Ducharme, Y.; Dubé, D.;
Falgueyret, J.-P.; Fournier, P.-A.; Gagné, S.; Gallant, M.; Grimm, E.; Han, Y.;
Houle, R.; Huang, J.-Q.; Hughes, G.; Jûteau, H.; Lacombe, P.; Lauzon, S.;
Lévesque, J.-F.; Liu, S.; MacDonald, D.; Mackay, B.; McKay, D.; Percival, M. D.;
St-Jacques, R.; Toulmond, S. Bioorg. Med. Chem. Lett. 2011, 21, 3976.
6. Lévesque, J.-F.; Bleasby, K.; Chefson, A.; Chen, A.; Dubé, D.; Ducharme, Y.;
Fournier, P.-A.; Gagné, S.; Gallant, M.; Grimm, E.; Hafey, M.; Han, Y.; Houle, R.;
Lacombe, P.; Laliberté, S.; MacDonald, D.; Mackay, B.; McKay, D.; Papp, R.;
Tschirret-Guth, R. Bioorg Med. Chem. Lett. 2011, manuscript accepted.

7404
A. Chen et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7399–7404
7. Buffer assay: Human recombinant renin (Proteos) at 100 pM was incubated in
the presence or absence of renin inhibitors and 6 M of Q-FRET substrate 9
14. The structure was confirmed through independent synthesis:
l
DNP-Lys-His-Pro-Phe-His-Leu-Val-Ile-His-D,L-Amp in 50 mM MOPS, 100 mM
NaCl, pH 7.4, 0.002% Tween. The reactions take place in a Costar 384 well
black plate at 37 °C for 3 h. Fluorescence was measured at times 0 and 3 h in a
SpectraMax Gemini EM reader with excitation and emission filters at 328 and
388 nm, respectively. Plasma assay: Frozen human EDTA-plasma was rapidly
thawed in warm water and centrifuged at 2900 g for 15 min at 40 °C. The
supernatant was collected and recombinant human renin (Proteos) added at
1 nM nominal concentration. The plasma was transferred to Costar black 384
well plates, renin inhibitors added and the mixture pre-incubated at 37 °C for
10 min. The renin Q-FRET substrate QXL520-Lys-His-Pro-Phe-His-Leu-Val-Ile-
His-Lys-(5-FAM) (Proteos), diluted in 3 M Tris/200 mM EDTA, pH 7.2 was added
to the plasma with final concentrations of 342 mM Tris, 23 mM EDTA and
6.8 lM substrate. The plate was incubated at 37 °C for 1 h and the plate read in
a SpectraMax Gemini EM reader with excitation and emission filters at 490 and
520 nm, respectively, at time 0 and 1 h.
8. Chen, A.; Campeau, L. C.; Cauchon, E.; Chefson, A.; Ducharme, Y.; Dubé, D.;
Falgueyret, J.-P.; Fournier, P.-A.; Gagné, S.; Grimm, E.; Han, Y.; Houle, R.; Huang,
J.-Q.; Lacombe, P.; Laliberté, S.; Lévesque, J.-F.; Liu, S.; MacDonald, D.; Mackay,
B.; McKay, D.; Percival, D.; Regan, C.; Regan, H.; St-Jacques, R.; Toulmond, S.
Bioorg. Med. Chem. Lett. 2011, 21, 3970.
(a) 0.9 equiv NaBrO3, 0.03 equiv RuCl3, 2:1 (v/v) water–MeCN, rt, 8 h, 65%; (b)
NaH (60% dispersion in paraffin oil), 2-chloro-3-trifluorobenzyl bromide, DMF,
45 °C, 12 h, 39%; (c) LiBH4, THF, rt, 8 h, 23%; (d) HCl (4.0 M in dioxane), rt,
40 min, 5%.
9. Nigam, S. C.; Mann, A.; Taddei, M.; Wermuth, C.-G. Synth. Commun. 1989, 19,
3139.
10. The only other product observed during the cyclization step was derived from
15. Chen, A.; Bayly, C.; Bezençon, O.; Richard-Bildstein, S.; Dubé, D.; Dubé, L.;
Gagné, S.; Gallant, M.; Gaudreault, M.; Grimm, E.; Houle, R.; Lacombe, P.;
Laliberté, S.; Lévesque, J. F.; Liu, S.; MacDonald, D.; Mackay, B.; Martin, D.;
intramolecular
transesterification
to
furnish
the
corresponding
ˇ
McKay, D.; Powell, D.; Remen, L.; Soisson, S.; Toulmond, S. Bioorg. Med. Chem.
benzoxepinone. All attempts to convert this to the desired product were
unsuccessful.
Lett. 2010, 20, 220.
16. Incubations studies (37 °C, 30 min) of compound 31 (1 lM) with rat, dog,
11. All other attempts to furnish the desired amide were unsuccessful as ester 15
readily underwent retro-Michael reaction prior to amide formation. Attempts
rhesus and human liver microsomes (0.25 mg/mL in the presence of NADPH
regenerating system) were performed. The % parent remaining was calculated
to be 53% (rat), 43% (dog), 10% (rhesus) and 51% (human).
to coax similar ring closure on the resulting
unsuccessful.
a,b-unsaturated amide were also
12. Hess, P.; Clozel, M.; Clozel, J.-P. J. Appl. Physiol. 1996, 81, 1027.
13. Chen, A.; Dubé, D.; Dubé, L.; Gagné, S.; Gallant, M.; Gauldreault, M.; Grimm, E.;
Houle, R.; Lacombe, P.; Laliberté, S.; Liu, S.; MacDonald, D.; Mackay, B.; Martin,
D.; McKay, D.; Powell, D.; Lévesque, J.-F. Bioorg. Med. Chem. Lett. 2010, 20, 5074.