Acylguanidine inhibitors of β-secretase: Optimization of the pyrrole ring substituents extending into the S1 and S3 substrate binding pockets

Article abstract of DOI:10.1016/j.bmcl.2007.12.010

Proteolytic cleavage of amyloid precursor protein by β-secretase (BACE-1) and γ-secretase leads to formation of β-amyloid (Aβ) a key component of amyloid plaques, which are considered the hallmark of Alzheimer's disease. Small molecule inhibitors of BACE-1 may reduce levels of Aβ and thus have therapeutic potential for treating Alzheimer's disease. We recently reported the identification of a novel small molecule BACE-1 inhibitor N-[2-(2,5-diphenyl-pyrrol-1-yl)-acetyl]guanidine (3.a.1). We report here the initial hit-to-lead optimization of this hit and the SAR around the aryl groups occupying the S₁ and S_2′ pockets leading to submicromolar BACE-1 inhibitors.

Full text of DOI:10.1016/j.bmcl.2007.12.010

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1063–1066
Acylguanidine inhibitors of b-secretase: Optimization
of the pyrrole ring substituents extending into the S1
and S3 substrate binding pockets
Derek C. Cole,a,* Joseph R. Stock,a Rajiv Chopra,b Rebecca Cowling,a
John W. Ellingboe,a Kristi Y. Fan,c Boyd L. Harrison,c Yun Hu,d Steve Jacobsen,d
Lee D. Jennings,a Guixian Jin,a Peter A. Lohse,e Michael S. Malamas,c Eric S. Manas,f
William J. Moore,f Mary-Margaret O’Donnell,e Andrea M. Olland,b Albert J. Robichaud,c
Kristine Svenson,b JunJun Wu,b Eric Wagnerd and Jonathan Bardd
aChemical & Screening Sciences, Wyeth Research, 401 N. Middletown Road, Pearl River, NY 10965, USA
bChemical & Screening Sciences, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
cChemical & Screening Sciences, Wyeth Research, 865 Ridge Road, Monmouth Junction, NJ 08852, USA
dDiscovery Neuroscience, Wyeth Research, 865 Ridge Road, Monmouth Junction, NJ 08852, USA
eArQule, 19 Presidential Way, Woburn, MA 01801, USA
fChemical & Screening Sciences, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
Received 22 October 2007; revised 5 December 2007; accepted 6 December 2007
Available online 26 December 2007
Abstract—Proteolytic cleavage of amyloid precursor protein by b-secretase (BACE-1) and c-secretase leads to formation of b-amy-
loid (Ab) a key component of amyloid plaques, which are considered the hallmark of Alzheimer’s disease. Small molecule inhibitors
of BACE-1 may reduce levels of Ab and thus have therapeutic potential for treating Alzheimer’s disease. We recently reported the
identification of a novel small molecule BACE-1 inhibitor N-[2-(2,5-diphenyl-pyrrol-1-yl)-acetyl]guanidine (3.a.1). We report here
0
the initial hit-to-lead optimization of this hit and the SAR around the aryl groups occupying the S1 and S2 pockets leading to sub-
micromolar BACE-1 inhibitors.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Alzheimer’s disease (AD) is a progressive neurodegener-
ative disease that is the leading cause of dementia.1
Varying estimates suggest that from 5% to 50% of early
onset AD cases result from a variety of genetic muta-
tions, meaning the remainder are sporadic with the risk
increasing with age.2 b-Amyloid plaques are one of the
key hallmarks of AD and are formed by aggregation
of amyloid fibrils, which in turn are formed from the
neurotoxic amyloid b-peptide (b-amyloid, Ab40,42). Pro-
teolytic cleavage of amyloid precursor protein (APP) by
b-secretase (also referred to as b-site APP cleaving en-
zyme-1, BACE-1, Asp-2, and memapsin-2) generates
the membrane-bound b-C-terminal fragment (bCTF,
also referred to as C99) which in turn is cleaved by
c-secretase to generate Ab.3–5 Thus, BACE-1 inhibitors
may decrease levels of Ab and have therapeutic benefit
in the treatment of AD.6,7 A high throughput screening
campaign identified N-[2-(2,5-diphenyl-pyrrol-1-yl)-
acetyl]guanidine 3.a.1 (Ar1 & Ar2 = Ph, Table 1) as an
inhibitor of BACE-1 with low micromolar activity
(BACE-1 IC50 = 3.7 lM, Kd = 2.8 lM).8 Compound
3.a.1 inhibited Ab formation in a cellular assay with
an IC50 of 8.9 lM and consistent with a BACE-1 medi-
ated inhibition mechanism caused a dose dependent
reduction of bCTF and Ab levels in a radiolabeled
immunoprecipitation assay without affecting a-secretase
amyloid precursor protein (a-sAPP) levels.8 Compound
3.a.1 was selective for BACE-1 over cathepsin D
(IC50 = 60 lM), pepsin (IC50 > 50 lM), and BACE-2
(IC50 > 50 lM).
Keywords: BACE-1; b-Secretase; Inhibitor; Acylguanidine.
Corresponding author. Tel.: +1 845 602 4619; fax: +1 845 602
5561; e-mail: coledc@wyeth.com
*
An X-ray structure of 3.a.1 with BACE-1 was solved
and revealed key binding interactions between the
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.12.010

1064
D. C. Cole et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1063–1066
Table 1. BACE-1 inhibitory activities of compounds 3.(a–l).(1–13)
NH2
HN
NH
O
Ar2
N
Ar1
Ar1
Ar2
Inhibition (%) at Ph
a
b
c
d
e
f
g
h
i
j
k
l
2-Cl- 2-MeO- 3-F- 3-Cl- 3-MeO- 3-CN- 4-F- 4-MeO- 4-CF3- 1-Naphthyl 3-Benzothienyl
10 lM/IC50 (lM)
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
1
2
Ph
3.7
3
51%
56% 62%
65%
69%
56%
47%
63%
70%
58%
68%
1.4
78%
39%
37%
34%
52%
38%
41%
22%
3%
66%
2
63%
81%
75%
64%
71%
66%
43%
92%
0.6
56%
35%
34%
25%
37%
61%
9%
2-Cl-Ph
3-Cl-Ph
3-Br-Ph
3-Me-Ph
4-Me-Ph
4-MeO-Ph
2,5-di-Cl-Ph
4-PhO-Ph
85%
40%
28%
9%
3
69% 58% 51%
68% 69% 46%
66% 59%
63% 66%
72% 41%
65% 39%
53% 12%
4
5
4.6 70% 56%
75% 43% 48%
57%
62%
61%
65%
1.5
38%
6
7
53%
3.5
32%
48% 78%
23%
60%
0.6
16%
8
2
83%
0.5
51%
9
1.3
1.1
0.7
1.3
1.6
1.3
1.4
1.3
47%
0.7
10
11
12
13
4-(4-Ac-PhO)-Ph 1.0
1.2
1.3
0.8
0.6
0.9
4-BnO-Ph
2-Naph
6-Me-2-Naph
64% 84%
56% 70% 53%
60% 84% 53%
58%
50%
61%
79%
64%
79%
41%
39%
46%
25%
72%
49%
7%
40%
50% 44%
58% 17%
29%
acylguanidine moiety and the two catalytic aspartic
acids, Asp32 and Asp228 (Fig. 1).8–10 The pyrrole ring
points toward the flap region, making a p-edge stacking
interaction with Tyr71 and stabilizing the flap in an
‘open’ conformation relative to published peptidomi-
metic co-structures.11 The two phenyl groups on the
tively with molecular modelers and was based on the
anticipated ability of different substituent groups to be
0
accommodated into the S1–S3 and/or S2 pockets. Ana-
logs were synthesized from a-bromomethyl ketones
and methyl ketones as starting materials as shown in
Scheme 1.
0
pyrrole ring extend into the S1 and S2 substrate binding
pockets. The para-position of the P1 phenyl group
projects directly toward the unoccupied S3 pocket, indi-
cating an opportunity to add substituents to the P1 phe-
nyl extending into the S3 pocket and thereby potentially
increasing binding affinity. This manuscript describes
the initial hit-to-lead optimization of this series of
BACE-1 inhibitors.
The 1,4-diarylbutane-1,4-diones 1.(a–l).(1–13) were pre-
pared in a one-step cross-coupling reaction under the ac-
tion of ZnCl2 Æ t-BuOH Æ Et2NH using the procedure of
Kulinkovich.12 Reactions were carried out over 3–5 days
at ambient temperature and gave modest yields of the
diketones (36–50%). The crude diketones were coupled
with glycine in an acid catalyzed condensation and sub-
sequent CDI mediated coupling of the resultant pyrrole
acetic acids with guanidine hydrochloride gave the tar-
get acylguanidines.
Initial SAR investigation of this lead began with the syn-
thesis and assay of a 12 · 13 combinatorial library. The
selection of target molecules was carried out interac-
All compounds were assayed for BACE-1 inhibition in a
fluorescence resonance energy transfer (FRET) assay8 at
a concentration of 10 lM, and IC50’s were determined
O
O
O O
a
Ar1
Ar2
Br
+
Ar1
Ar2
1.(a-l).(1-13)
NH2
OH
HN
NH
O
Ar2
b
c
N
Ar1
O
N
Ar1
Ar2
2.(a-l).(1-13)
3.(a-l).(1-13)
Figure 1. Complex of 3.a.1 with BACE-1 showing hydrogen bonding
interactions between the acylguanidine moiety and Asp32 and Asp228
Scheme 1. Reagents and conditions: (a) Et2NH, t-BuOH, ZnCl2,
toluene, rt, 2–5 days; (b) glycine, p-TSA, EtOH, 80 ꢁC, 3 days; (c) i—
CDI, DMF, rt, 1 h; ii—guanidineÆHCl, Et3N, rt, 5 h.
, the p-edge stacking interaction between the pyrrole ring and Tyr71
10
0
and the two phenyl groups projecting into S1 and S2 .

D. C. Cole et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1063–1066
1065
for compounds showing > 80% inhibition (Table 1).
NH2
NH
For acyl guanidines with smaller phenyl ring substitu-
ents (Table 1, rows 1–8), BACE-1 inhibition activity
was not appreciably higher than for 3.a.1 (3.7 lM,
66% inhibition @10 lM). Because of the pseudo-sym-
metry of molecules with smaller phenyl ring substitu-
ents, the phenyl rings may be accommodated in either
HN
O
H
N
a, b, c,
d, e, f
O
R1
L
N
Ph
O2N
4 (meta-, L = CO)
5 (meta-, L = SO2)
6 (para-, L = CO)
7 (para-, L = SO2)
0
the S1 and S2 pockets making interpretation of the
SAR difficult. Improved potency was seen in all exam-
ples with the 4-phenoxyphenyl and 4-(4-acylphen-
oxy)phenyl substituent groups (Table 1, rows 9 and
10). These functional groups were included in the library
design with the expectation that the phenoxy group on
the P1 phenyl would extend into the S3 pocket. para-
Benzyloxyphenyl analogs were poorly tolerated indicat-
ing this group may be too large to be accommodated
(Table 1, row 11). Similarly the bicyclic ring systems
were poorly tolerated (Table 1, row 12 and 13, and col-
umns k and l). When Ar1 is 4-(PhO)Ph or 4-(AcPhO)-
Ph, the arylethers are locked into the S1–S3 pocket facil-
itating interpretation of the Ar2 SAR. Somewhat sur-
prisingly, no clear preference for ortho, meta, or para-
NH2
HN
NH
O
a, b, c,
g, h, f
O
O
R2
N
Ph
HO
8 (meta-)
9 (para-)
Scheme 2. Reagents and conditions: (a) Et2NH, t-BuOH, ZnCl2,
toluene, rt, 2–5 days; (b) glycine, AcOH, 120 ꢁC, 5 h; (c) TMSCl,
MeOH, 70 ꢁC, 2 h; (d) 10% Pd/C, MeOH–THF, 1 h; (e) R1COCl or
R1SO2Cl, Et3N, DCM, 16 h; (f) guanidine, DMSO, rt, 2 h; (g) BBr3,
DCM, ꢀ78 ꢁC, then rt, 5 h; (h) R2Br, K2CO3, DMF, 65 ꢁC, 16 h.
0
substitution of the P2 phenyl was observed in this li-
brary (see Table 1, rows 9 and 10).
ents (8–9) were prepared by modifications of the
general route outlined in Scheme 1 (see Scheme 2).
A co-crystal structure of compound 3.b.10 (BACE-1
IC50 = 0.7 lM), with BACE-1 (Fig. 2) shows the acyl-
guanidine forming key H-bonding interactions with
the two aspartic acids and the pyrrole ring p-stacking
with the Tyr71 of the flap region.10 The 4-(4-acetylphen-
oxy)-phenyl group occupies the S1 and S3 substrate
binding pockets as predicted by our models and the 2-
0
chloro group on the P2 phenyl is oriented toward the
0
back of the S2 pocket, although this substituent does
not appear to contribute significantly to potency (com-
pare 3.a.10 and 3.b.10, Table 1).
The amides and sulfonamides linked to the P1 phenyl at
the meta-position (4 and 5) all had weak activity (data
not shown), while the para-alkyl, heteroaryl, and aryla-
mides (Table 2, 6a–j) generally had low lM activity with
6g being the most potent analog (IC50 = 0.6 lM). The
corresponding para-sulfonamides (7) had much weaker
activity. Similarly, the set of meta-alkoxyphenyl analogs
(8) had low activity (data not shown) while the para-alk-
oxyphenyl analogs (Table 3, 9a–g) were surprisingly
Because of the significant increase in potency observed
for acyl guanidines with 4-phenoxy substituents on the
P1 phenyl, additional analogs were designed to further
investigate substituents projecting into the S3 pocket.
Meta and para-acylaminophenyl and sulfonylaminophe-
nyl (4–7), and meta and para-alkyloxyphenyl substitu-
Table 2. BACE-1 inhibitory activities of compounds 6 and 7
NH2
HN
NH
O
N
HN
R1 L
R1
L
% Inhibition at 10 lM/IC50 (lM)
6a
6b
6c
6d
6e
6f
Me
Et
–CO–
–CO–
40%
71%
42%
79%
26%
1.2
Cyclopropyl –CO–
t-Bu
–CO–
–CO–
–CO–
–CO–
–CO–
–CO–
–CO–
–SO2–
–SO2–
–SO2–
–SO2–
–SO2–
–SO2–
–SO2–
cyclohexyl
2,4-di-Cl-Ph
3-Br-Ph
3-MeO-Ph
4-Br-Ph
4-F-Bn
Me
6g
6h
6i
0.6
1.8
1.4
6j
54%
65%
75%
71%
86%
50%
65%
43%
7a
7b
7c
7d
7e
7f
Et
n-Propyl
Bn
3-MeO-Ph
3-Br-Ph
3-Me-Ph
Figure 2. Complex of 3.b.10 with BACE-1. Polar/charged residues are
shown in blue and lipophilic residues are shown in red.10
7g

1066
D. C. Cole et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1063–1066
Table 3. BACE-1 inhibitory activities of compounds 9
Table 4. IC50 values for inhibition of BACE-1, BACE-2, cathepsin-D
and cellular Ab production by 3.b.10 and 6g
NH2
BACE-1
BACE-2
CathD
Cellular
IC50, lM IC50, lM IC50, lM Abtotal ED50, lM
HN
NH
3.b.10 0.7
6g 0.6
2.3
0.5
18.9
2.1
3.8
5.6
O
N
O
R2
ers. Future studies will focus on optimization of the
interactions with S3 while reducing the strain energy of
the S1–S3 linker, and also incorporation of substituents
R2
% Inhibition at 10 lM/IC50 (lM)
9a
9b
9c
9d
9e
9f
Et
allyl
1.6
1.7
0
on the acylguanidine extending into the S1 pocket.
Propargyl
3-CN-Propyl
Butyl
83%
0.9
1.8
1.9
References and notes
4-CN-Bn
Phenethyl
9g
68%
1. Alzheimer’s Association. http://www.alz.org.
2. Selkoe, D. J. Physiol. Rev. 2001, 81, 741.
3. Li, Y. M.; Xu, M.; Lai, M. T.; Huang, Q.; Castro, J. L.;
DiMuzio-Mower, J.; Harrison, T.; Lellis, C.; Nadin, A.;
Neduvelil, J. G.; Register, R. B.; Sardana, M. K.;
Shearman, M. S.; Smith, A. L.; Shi, X. P.; Yin, K. C.;
Shafer, J. A.; Gardell, S. J. Nature 2000, 405, 689.
4. Sinha, S.; Anderson, J. P.; Barbour, R.; Basi, G. S.;
Caccavello, R.; Davis, D.; Doan, M.; Dovey, H. F.;
Frigon, N.; Hong, J.; Jacobson-Croak, K.; Jewett, N.;
Keim, P.; Knops, J.; Lieberburg, I.; Power, M.; Tan, H.;
Tatsuno, G.; Tung, J.; Schenk, D.; Seubert, P.; Suo-
mensaari, S. M.; Wang, S.; Walker, D.; Zhao, J.;
McConlogue, L.; John, V. Nature 1999, 402, 537.
5. Wolfe, M. S.; Xia, W.; Ostaszewski, B. L.; Diehl, T. S.;
Kimberly, W. T.; Selkoe, D. J. Nature 1999, 398, 513.
6. Ghosh, A. K.; Bilcer, G.; Hong, L.; Koelsch, G.; Tang, J.
Curr. Alzheimer Res. 2007, 4, 418.
7. For recent reviews of small molecule BACE-1 inhibitors,
see: (a) Durham, T. B.; Shepherd, T. A. Curr. Opin. Drug
Discov. Dev. 2006, 9, 776; (b) Hills, I. D.; Vacca, J. P. Curr.
Opin. Drug Discov. Dev. 2007, 10, 383.
Figure 3. Complex of 6g with BACE-1. Polar/charged residues are
shown in blue and lipophilic residues are shown in red.10
8. Cole, D. C.; Manas, E. S.; Stock, J. R.; Condon, J. S.;
Jennings, L. D.; Aulabaugh, A.; Chopra, R.; Cowling, R.;
Ellingboe, J. W.; Fan, K. Y.; Harrison, B. L.; Hu, Y.;
Jacobsen, S.; Jin, G.; Lin, L.; Lovering, F. E.; Malamas,
M. S.; Stahl, M. L.; Strand, J.; Sukhdeo, M. N.; Svenson,
K.; Turner, M. J.; Wagner, E.; Wu, J.; Zhou, P.; Bard, J.
J. Med. Chem. 2006, 49, 6158.
9. For other BACE-1 inhibitors with basic amidine-like‘war-
heads’ see: Gerritz, S.; Zhai, W.; Shi, S.; Zhu, S.; Good, A.
C.; Thompson, L. A. III, WO-2007002214 (2007); Mjalli,
A. M.; Jones, D.; Gohimmukkula, D. R.; Huang, G.; Zhu,
J.; Rao, M.; Andrews, R.C.; Ren, T. WO-2006099379
(2006); Murray, C. W.; Callaghan, O.; Chessari, G.;
Cleasby, A.; Congreve, M.; Frederickson, M.; Hartshorn,
M. J.; McMenamin, R.; Patel, S.; Wallis, N. J. Med.
Chem. 2007, 50, 1116; Congreve, M.; Aharony, D.; Albert,
J.; Callaghan, O.; Campbell, J.; Carr, R. A. E.; Chessari,
G.; Cowan, S.; Edwards, P. D.; Frederickson, M.;
McMenamin, R.; Murray, C. W.; Patel, S.; Wallis, N. J.
Med. Chem. 2007, 50, 1124.
active, indicating that sub-lM activity is possible with a
smaller alkyl group occupying the S3 pocket (see 9d).
The crystal structure of the complex of acylguanidine 6g
with BACE-1 was subsequently solved and revealed that
the 3-bromobenzoyl substituent was tightly packed into
the S3 pocket (Fig. 3).10 Accommodation of this rather
large substituent into the pocket was only possible with
twisting of the amide bond out of plane with respect to
the P1 group. Presumably this twisting of the amide
bond is unfavorable leading to reduced potency. In con-
trast, a model of 9d indicates that the cyanopropylether
extends into the S3 pocket with minimal strain achieving
similar potency with greater atom efficiency.
In general, these inhibitors are selective for BACE-1
over cathepsin-D (usually >50-fold, although 6g is only
4-fold) and pepsin (data not shown) and have limited
selectivity (<2-fold) over BACE-213 (Table 3). Most
sub-micromolar BACE-1 inhibitors had low micromolar
activity in the cellular Abtotal lowering ELISA (Table 4).
10. Coordinates for the BACE-1 complexes with 3.a.1, 3.b.10,
and 6g have been deposited in the Protein Data Bank
under accession codes 2QU2, 2ZDZ, and 2ZE1, together
with the corresponding structure.
11. Hong, L.; Koelsch, G.; Lin, X.; Wu, S.; Terzyan, S.;
Ghosh, A. K.; Zhang, X. C.; Tang, J. Science 2000, 290,
150.
12. Nevar, N. M.; Kel’in, A. V.; Kulinkovich, O. G. Synthesis
2000, 1259.
13. Chou, K. C. J. Proteome Res. 2004, 3, 1069.
In summary, in this initial hit-to-lead optimization effort
we have achieved up to 6-fold increase in potency and
demonstrated the ability to extend directly from the S1
pocket into the S3 pocket with groups attached via the
para-position of the P1 group and with a variety of link-