Synthesis, SAR, and X-ray structure of human BACE-1 inhibitors with cyclic urea derivatives

Article abstract of DOI:10.1016/j.bmcl.2008.03.081

We describe synthesis and evaluation of a series of cyclic urea derivatives with hydroxylethylamine isostere. Modification of P3, P1, and P2′ and combination of SAR display a >100-fold increase in potency with good cellular activity (IC₅₀ = 0.1

Full text of DOI:10.1016/j.bmcl.2008.03.081

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2900–2904
Synthesis, SAR, and X-ray structure of human BACE-1
inhibitors with cyclic urea derivatives
a,
a
a
a
Heuisul Park,a,* Kyeongsik Min, Hyo-Shin Kwak, Ki Dong Koo, Dongchul Lim,
*
Sang-Won Seo,a Jae-Ung Choi,a Bettina Plattb and Deog-Young Choia
aDrug Discovery, Research & Development, LG Life Sciences, 104-1 Munji-dong, Yuseong-gu, Daejon 305-380, Republic of Korea
bSchool of Medical Sciences, College of Life Sciences and Medicine, University of Aberdeen, Institute of Medical Sciences,
Foresterhill, Aberdeen AB25 2ZD, Scotland, UK
Received 23 October 2007; revised 28 March 2008; accepted 29 March 2008
Available online 8 April 2008
Abstract—We describe synthesis and evaluation of a series of cyclic urea derivatives with hydroxylethylamine isostere. Modification
of P3, P1, and P20 and combination of SAR display a >100-fold increase in potency with good cellular activity (IC50 = 0.15 lM)
relative to the previously reported compound 3.
Ó 2008 Elsevier Ltd. All rights reserved.
Alzheimer’s disease (AD) is the most common cause
of progressive dementia characterized by pathological
hallmark of intracellular neurofibrillary tangles and
extracellular neuritic plaques primarily composed of
amyloid beta peptide (Ab).1 Ab is a 40- to 42-amino
acid peptide generated by sequential proteolytic cleav-
age of membrane-anchored amyloid precursor protein
(APP) by b-secretase (BACE-1) and c-secretase.2
Although the link between Ab as its partially soluble
or aggregated forms and neuronal dysfunction in
AD is controversial, there are many evidences that
progressive accumulation of Ab in the brain is critical
response for the neurodegeneration in AD.3 Therefore
inhibition of APP proteolysis to lower the concentra-
tion of neurotoxic Ab is an attractive strategy for clin-
ical intervention.
showed no Ab production and a relatively normal phe-
notype.5 Furthermore, crystal structures of BACE-1 in
complex with inhibitors were determined at atomic res-
olution,6 which provided a powerful tool for structure-
based drug design of BACE-1 inhibitors. As such,
BACE-1 is an attractive therapeutic target for the treat-
ment and prevention of AD.
In order to replace the P2 isophthalide ring, 2-oxo-het-
erocyclic compounds were designed and evaluated.7
The carbonyl group of 2-oxo-heterocyclic group as in
compounds 1–3 at P2 serves as a hydrogen acceptor of
Thr232 amide NH enhancing the binding affinity.
Among them, compound 3 was more potent, more selec-
tive against cathepsin D (Cat-D), and cell-permeable
inhibitor (IC50 = 7.3 lM). In this letter, we describe
optimization of compound 3 to further improve potency
and selectivity (Fig. 1).
BACE-1 is a type I membrane-anchored aspartyl prote-
ase identified in 1999.4 The proteolysis of APP by
BACE-1 occurs in the luminal surface of the plasma
membrane and releases the soluble ectodomain of
APP. The remaining membrane-associated C-terminal
domain including Ab is subsequently cleaved to Ab by
c-secretase. BACE-1 is highly expressed in the central
nervous system. BACE-1 knockout homozygote mice
At first, we examined the hydrophobic interaction of the
benzyl group at N-3 position (R1) as well as the substi-
tuent effect of cyclic urea moiety at C-5 position (R2).
Compounds were synthesized according to Scheme 1
and the results are summarized in Table 1. Absence of
the benzyl group (4) and replacement with the n-butyl
group (5) of P3, thereby reducing van der Waals con-
tacts, resulted in nearly 70-fold loss of potency. This re-
veals that the benzyl group at N-3 position (R1) is
essential for binding and the proper size is necessary
to obtain high affinity for BACE-1.
Keywords: Alzheimer’s disease; BACE-1; Cyclic urea.
*
Corresponding authors. Tel.: +82 42 866 2240; fax: +82 42 866
5754; e-mail: renivie@lgls.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.081

H. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2900–2904
2901
Table 1. SAR of P2 and P3
Thr232
NH
S2
O
H
O
X
R2
OH
N
R1
OH
H
Y
H
H
N
H
N
N
N
N
N
S3
N
O
O
S1
BACE-1
Ki = 2.194 μM
Ki = 3.005 μM
Ki = 0.284 μM
Compound R1
R2
Ki (lM)
BACE-1 CatD
Cathepsin D
Ki = 2.053 μM
Ki = 0.810 μM
Ki = 0.612 μM
1: X = N,
2: X = CH,
3: X = N,
Y = CH2
Y = NH
Y = NH
3
4
Benzyl
H
n-Butyl
H
0.284
19.004
39.989
6.800
3.514
6.768
5.001
3.500
8.177
0.612
—
—
H
5
H
Figure 1. Structures of BACE-1 inhibitors.
6
Phenethyl
H
0.310
0.590
0.431
0.080
1.125
1.302
7
Phenoxy-ethyl
Benzyl
Benzyl
H
8
Methyl
Benzyl
H
We synthesized a series of aryl compounds with various
lengths to optimize the phenyl ring (Scheme 2). The 3-
phenyl-2-oxo-imidazoline derivatives (10 and 11) gave
lower affinity compared with benzyl derivative (3). The
longer chain derivatives (6 and 7) also gave lower affinity
and increased the binding affinity for Cat-D. Although
the benzyl group appears to be the best in this series, this
result indicates that the proper position of the phenyl
ring is important for BACE-1 affinity and selectivity.
9
10
11
3,5-Dibromophenyl
Phenyl
H
the phenoxy part is probably related to disordered bind-
ing due to the improper binding site for the phenoxy
part (data not shown).
Filling S2 site was attempted by introducing the methyl
(8) or the benzyl group (9) at C-5 position (R2). Com-
pounds 8 and 9 were synthesized by Scheme 1 and were
selected from two kinds of isomers which showed better
activity. The absolute configuration was confirmed by
X-ray crystallography, respectively. S2 binding site was
known to be favorable to hydrophobic side chains of
Ala and Met8 but the inhibitory activity and the selectiv-
ity of compounds 8 and 9 were approximately 20-fold
decreased compared to compound 3. To determine the
binding mode of 9, co-crystal structure was solved at
The phenyl groups of 10 and 11 might be directed to P4
position of OM99-2, a peptidomimetic inhibitor with
eight residues. The longer phenethyl (6) and the phen-
oxyethyl group (7) were thought to be larger for the
S3 site. The co-crystal structure of BACE-1 in complex
with compound 7 was determined to elucidate the bind-
ing mode but the electron density for the phenoxy part
was absent. However, the orientation of the ethyl part
toward the P4 main-chain backbone of OM99-2 indi-
cates that the phenoxyethyl group is not located in the
binding site of the P3 benzyl group in the BACE-1/com-
pound 3 structure.6,7 The absence of electron density for
13
˚
2.7 A (Fig. 2). The loss of activity is probably due to
O
O
O
Cbz
H
H
OH
N
N
N
H
N
H
N
a
b
HN
HN
HN
O
O
OH
N
N
O
O
O
c
4
O
N
O
O
N
Cbz
H
H
N
OH
N
N
H
N
H
N
R1
R1
R1
N
a
b
O
O
OH
O
3: R1 = Bn
14: R1 = (4-F)Bn
5: R1 =
n
-Butyl
15: R1 = (3,4-diF)Bn
16: R1 = (2,4-diF)Bn
17: R1 = (3,5-diF)Bn
18: R1 = (2,4,5-triF)Bn
19: R1 = (2-CF3)Bn
6: R1 = Ph-(CH2)2
7: R1 = PhO-(CH2)2
d
12: R1 = (3-methoxy)Bn
13: R1 = (3,5-dimethoxy)Bn
20: R1 = (4-CF3)Bn
O
N
O
N
O
N
Cbz
R2
O
H
N
H
N
R2
a
b
R2
OH
OH
N
H
H
R1
R1
R1
N
N
N
O
O
O
8: R1 = Bn R2 = methyl
9: R1 = Bn R2 = Bn
Scheme 1. Reagents and conditions: (a) HBr, rt, 4 h; (b) (2R,3 S)-1-(3-dimethylamino)benzylamino-2-hydroxy-3-amino-4-phenyl-butane, EDC,
HOBt, DMF, Et3N, rt, 4 h; (c) NaH, R1-X, THF, 0 °C to rt, 4 h; (d) LiHMDS, R2-X, THF, 0 °C–rt, 4 h.

2902
H. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2900–2904
O
X
H
NH-Tr
N
a-d
O
HO
N
OH
O
O
X
O
N
X
H
OH
N
H
H
N
N
e
N
O
X
10: X = Br
11: X = H
Scheme 2. Reagents and conditions: (a) SO3ÆPy, Et3N, CH2Cl2, rt, 4 h;
(b) aniline, NaBH(OAc)3, ClCH2CH2Cl, rt, 8 h; (c) Phosgene (40% in
Toluene), 0 °C–rt, 8 h; (d) HBr, rt, 4 h; (e) (2R, 3S)-1-(3-dimethyl-
amino)benzylamino-2-hydroxy-3-amino-4-phenyl-butane,
EDC,
HOBt, DMF, Et3N, rt, 4 h.
Figure 2. X-ray structure of compound 9 (green) in complex with
BACE-1. Overlay with the peptidomimetic inhibitor, OM99-2 (yellow)
from 1FKN. The flap region is removed to clear view of P2.
˚
geometric differences (ꢀ1 A) at P2 Ca position between
compound 9 and OM99-2, which resulted an unfavor-
able interaction of the R2 benzyl group located far away
from S2 site.
(3). To confirm whether the methoxy group oriented
to the S3-sub or not, co-crystal structure of BACE-1
in complex with compound 12 was determined at
Next, we introduced various R1 substituents at the phe-
nyl ring of compound 3 such as F, OMe, CF3 to improve
potency by enhancing hydrophobic interaction and also
improve selectivity by using the S3-subpocket (S3-sub)
(Scheme 1). The results are summarized in Table 2.
The S3-sub was revealed by HTS hit compound/
BACE-1 structure9 and compounds with binding motif
in the S3-sub showed improved potency and selectiv-
ity.10 To access the S3-sub, we synthesized the 3-meth-
oxy and the 3,5-dimethoxy benzyl derivatives (12 and
13). Contrary to our expectation, those compounds gave
lower activity and selectivity than the benzyl derivative
˚
2.3 A resolution (Fig. 3). BACE-1/12 complex confirmed
that the methoxy group was oriented toward the S3-sub.
However, there was a subtle difference in the direction
between the methoxybenzyl group of compound 12
and the S3-sub binding part of compound 27 when they
were superimposed.10,11 A close examination of the
complex structure revealed that positioning of the R3
methoxybenzyl group for optimum fit in the S3-sub
was sterically hindered by the P1 benzyl group. Further-
more, this steric repulsion make the methoxy group to
be unfavorable for the S3-sub. Interestingly, ortho-tri-
Table 2. Optimization of cyclic urea derivatives
O
H
OH
N
R1
H
H
N
N
N
R3
R4
O
R4
Compound
R1
R3
R4
Ki (lM)
SEAP IC50 (lM)
BACE
CatD
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
(3-Methoxy)benzyl
(3,5-Dimethoxy)benzyl
(4-F)Benzyl
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–N(CH3)2
–CF3
H
H
H
H
H
H
H
H
H
H
H
H
H
F
0.449
0.662
0.818
0.272
0.867
0.053
0.260
0.267
4.957
1.256
1.881
0.018
0.009
0.051
0.002
0.195
0.142
0.276
0.354
0.450
0.417
0.520
0.190
0.290
1.071
0.300
0.151
0.135
0.320
0.066
14.22
36.48
11.63
5.74
9.95
1.97
5.59
3.12
—
(3,4-diF)Benzyl
(2,4-diF)Benzyl
(3,5-diF)Benzyl
(2,4,5-triF)Benzyl
(2-CF3)Benzyl
(4-CF3)Benzyl
Benzyl
12.33
21.07
0.58
0.43
1.29
0.15
Benzyl
–OCF3
i-Propyl
(3,5-diF)Benzyl
(3,5-diF)Benzyl
Benzyl
t-Butyl
–N(CH3)2
t-Butyl
(3,5-diF)Benzyl
F

H. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2900–2904
2903
O
H
N
O
F
boc
boc
a-c
d-e
F
O
N
N
O
OH
OH
H
H
N
I
NH2
F
N
boc
f-h
i,j
F
F
F
OH
H
O
N
N
H2N
H
R3
N
R1
F
+
OH
O
F
Figure 3. X-ray structure of compound 12 (green) in complex with
BACE-1. Overlay with the S3-sub binding inhibitor (yellow) from
2B8L (9).
O
H
N
OH
H
N
H
N
R1
N
k
R3
F
O
O
N
OH
H
H
N
N
H2N
R1
R3
+
OH
F
O
25: R1 = Bn
R3 = -N(CH3)2
-Butyl
26: R1 = (3,5-diF)Bn R3 =
t
O
N
H
OH
N
H
Scheme 4. Reagents and conditions: (a) LiHMDS, THF, 0 °C–rt, 4 h
(92 %) ; (b) 2 N HCl, MeOH, reflux, 6 h, (92%); (c) (BOC)2O, Et3N,
CH2Cl2, rt, 6 h (92%); (d) Chloroiodomethane, LDA, THF, 0 °C–rt,
4 h (62 %); (e) NaBH4, THF, 0 °C–rt, 4 h (80 %); (f) KOH, EtOH, rt,
1 h, (88%); (g) NaN3, NH4Cl, H2O, i-PrOH, 60 °C, 8 h (95%); (h) Pd/
C, H2 (Parr reactor, 45 psi), MeOH, rt, 4 h, (92 %); (i) Benzaldehyde,
NaBH(OAc)3, ClCH2CH2Cl, rt, 4 h (90–95%); (j) TFA, CH2Cl2, rt,
2 h (95%); (k) EDC, HOBt, Et3N, DMF, rt, 4 h.
H
R1
a
N
N
R3
O
21: R1 = Bn
22: R1 = Bn
R3 = -CF3
R3 = -OCF3
23: R1 = (3,5-diF)Bn R3 =
i
t
-Propyl
-Butyl
24: R1 = (3,5-diF)Bn R3 =
Scheme 3. Reagents and conditions: (a) EDC, HOBt, Et3N, DMF, rt,
4 h.
tween the flap and ligand. Finally, we optimized P1 site
simply by introducing F at 3 and 5 positions of the phe-
nyl ring, which was synthesized by Scheme 4. The 3,5-
difluorobenzyl group at P1 site (25 and 26) resulted in
excellent enzyme potency and cellular activity, which is
presumably due to the enhanced hydrophobic
interaction.
fluoromethyl substitution of the R1 benzyl group (19)
maintained BACE-1 potency but showed a 3-fold in-
crease in the potency for Cat-D while para-substitution
(20) resulted in a 17-fold decrease of BACE-1 potency.
These results suggest that the hydrophobic ortho-substit-
utent is more favorable for the potency for Cat-D and
the para-substitutent is unfavorable for the potency for
BACE-1 due to limited room for its binding.
In summary, we have described the optimization and
SAR of cyclic urea derivatives starting from compound
3. Although a co-crystal structure of compound 12
bound to BACE-1 shows limits for access to the S3-
sub to improve potency and selectivity, modification of
P3, P1, and P20 with hydroxyethylamine isosteres in-
creased potency by more than 100-fold. This compound
exhibited 33-fold selectivity against Cat-D and good cel-
lular activity.
Motivated by reports that fluorinated benzyl groups en-
hanced enzymatic and cellular activity,12 a series of fluo-
rinated benzyl compounds were investigated (14–18).
Among them, the 3,5-difluorobenzyl group (17) is the
most favorable substituent to give excellent enzyme po-
tency, cellular activity, and selectivity over Cat-D.
We also optimized the prime site by varying R3 substitu-
ent at phenyl ring (Scheme 3). While substitution of the
trifluoromethyl (21) and the trifluoromethoxy groups
(22) at meta position gave low potency, simple replace-
ment with the isopropyl (23) and the tert-butyl (24)
groups increased a 3- and 6-fold of potency respectively
and cellular activity. It is possible that those bulky
hydrophobic groups could enhance the interaction be-
Acknowledgments
We thank the staff at 4A Macromolecular Crystallogra-
phy Wiggler Beamline at Pohang Accelerator Labora-
tory, Korea for assistance during X-ray data
collection. This study was supported by a grant of the
Korea Health 21 R&D Project, Ministry of Health
and Welfare, Republic of Korea (A040147).

2904
H. Park et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2900–2904
Makara, G. M.; Annis, D. A.; Tadikonda, P. K.; Nash, H.
References and notes
M.; Vacca, J. P.; Wang, T. J. Med. Chem. 2004, 47, 6117.
10. Stachel, SJ.; Coburn, C. A.; Steele, T. G.; Crouthamel, M.
C.; Pietrak, B. L.; Lai, M. T.; Holloway, M. K.; Munshi,
S. K.; Graham, S. L.; Vacca, J. P. Bioorg. Med. Chem.
Lett. 2006, 16, 641.
1. Selkoe, D. J. Physiol. Rev. 2001, 81, 741.
2. Sinha, S.; Lieberburg, I. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 11049.
3. Jorge, G.; Blas, F. Adv. Drug Delivery Rev. 2002, 54, 1539.
4. Vassar, R.; Bennett, B. D.; Babu-Khan, S.; Kahn, S.;
Mendiaz, E. A.; Denis, P.; Teplow, D. B.; Ross, S.;
Amarante, P.; Loeloff, R.; Luo, Y.; Fisher, S.; Fuller, J.;
Edenson, S.; Lile, J.; Jarosinski, M. A.; Biere, A. L.;
Curran, E.; Burgess, T.; Louis, J.-C.; Collins, F.; Treanor,
J.; Rogers, G.; Citron, M. Science 1999, 286, 735.
5. Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.;
Branstetter, D. G.; Chen, K. S.; Freedman, S.; Frigon, N.
L.; Games, D.; Hu, K.; Johnson-Wood, K.; Kappenman,
K. E.; Kawabe, T.; Kola, I.; Kuehn, R.; Lee, M.; Liu, W.;
Motter, R.; Nichols, N. F.; Power, M.; Robertson, D. W.;
Schenk, D.; Schoor, M.; Shopp, G. M.; Shuck, M. E.;
Sihna, S.; Svensson, K. A.; Tatsuno, G.; Tintrup, H.;
Wijsman, J.; Wright, S.; McConlogue, L. Hum. Mol.
Genet. 2001, 10, 1317.
6. Hong, L.; Koelsch, G.; Lin, X.; Wu, S.; Terzyan, S.; Ghosh,
A. K.; Zhang, X. C.; Tang, J. Science 2000, 290, 150.
7. Min, K. et al.; manuscript in preparation.
8. Hu, J.; Cwi, C. L.; Smiley, D. L.; Timm, D.; Erickson, J.
A.; McGee, J. E.; Yang, H.-C.; Mendel, D.; May, P. C.;
Shapiro, M.; McCarthy, J. R. Bioorg. Med. Chem. Lett.
2003, 13, 4335.
11. Compound structure in PDB ID 2B8L; refer to Ref. 9.
O
O
S
N
S3sub
S3
OH
H
N
H
N
H
N
O
O
27
12. (a) Hom, R. K.; Fang, L. Y.; Mamo, S.; Tung, J. S.;
Guinn, A. C.; Walker, D. E.; Davis, D. L.; Gailunas, A.
F.; Thorsett, E. D.; Sinha, S.; Knops, J. E.; Jewett, N. E.;
Anderson, J. P.; John, V. J. Med. Chem. 2003, 46, 1799;
(b) Pyring, D.; Lindberg, J.; Rosenquist, A .; Zuccarello,
G.; Kvarnstro¨m, I.; Zhang, H.; Vrang, L.; Unge, T.;
Classon, B.; Hallberg, A.; Samuelsson, B. J. Med. Chem.
2001, 44, 3083.
13. Coordinated for the BACE-1 complexes with com-
pounds 9 and 12 have been deposited in the Protein
Data Bank (www.rcsb.org) under Accession Codes
3CKP and 3CKR.
9. Coburn, C. A.; Stachel, S. J.; Li, Y.-M.; Rush, D. M.;
Steele, T. G.; Chen-Dodson, E.; Holloway, M. K.; Xu, M.;
Huang, Q.; Lai, M.-T.; DiMuzio, J.; Crouthamel, M.-C.;
Shi, X.-P.; Sardana, V.; Chen, Z.; Munshi, S.; Kuo, L.;