Synthesis and SAR of bis-statine based peptides as BACE 1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.04.068

A new series of bis-statine based peptidomimetic inhibitors of human β-secretase (BACE 1) was developed by structure-based modification of the three regions to the initial lead 3: an N-terminus, a central bis-statine core, and a C-terminus. Introduction o

Full text of DOI:10.1016/j.bmcl.2004.04.068

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3457–3460
Synthesis and SAR of bis-statine based peptides as BACE 1
inhibitors
Baihua Hu,a,* Kristi Yi Fan,b Kristie Bridges,c Rajiv Chopra,c Frank Lovering,c
Derek Cole,d Ping Zhou,b John Ellingboe,d Guixiang Jin,d Rebecca Cowlingd
and Jonathan Barde
aDepartment of Chemical and Screening Sciences, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
bDepartment of Chemical and Screening Sciences, Wyeth Research, PO Box CN 8000, Princeton, NJ 08543, USA
cDepartment of Chemical and Screening Sciences, Wyeth Research, 85 Bolton Street, Cambridge, MA 02140, USA
dDepartment of Chemical and Screening Sciences, Wyeth Research, 401 N. Middletown Road, Pearl River, NY 10965, USA
eDepartment of Neuroscience, Wyeth Research, PO Box CN 8000, Princeton, NJ 08543, USA
Received 24 March 2004; accepted 20 April 2004
Available online
Abstract—A new series of bis-statine based peptidomimetic inhibitors of human b-secretase (BACE 1) was developed by structure-
based modification of the three regions to the initial lead 3: an N-terminus, a central bis-statine core, and a C-terminus. Introduction
of a 4-aminomethylbenzoic acid on the C-terminus resulted in a potent BACE 1 inhibitor with an IC50 value of 21 nM. The general
requirements for the optimal substrate–enzyme interaction are disclosed herein.
Ó 2004 Elsevier Ltd. All rights reserved.
b-Secretases, or b-amyloid converting enzyme (BACE 1),
is a membrane-associated aspartic protease that cleaves
the b-amyloid precursor protein (APP) and generates the
O
O
S
O
H
H
amyloid-b protein Ab, a major pathological feature of
Alzheimer’s disease (AD). BACE 1 knockout mice de-
velop normally, show no consistent phenotypic differ-
ences from their wild-type littermates, and their Ab levels
in the brains are reduced to negligible levels. Therefore,
BACE 1 inhibitors might be beneficial for the treatment
of AD.1 Earlier discovered potent BACE 1 inhibitors,
such as Elan’s Boc-VM-[Sta]-V-4-aminomethylbenzoic
acids (1) and Ac-VM-[Sta]-VAEF (2), are mono-statine
transition state analogues based on the amino acid se-
quence of APP around the BACE 1 cleavage site.2;3 Our
initial bis-statine lead 3, with substantially reduced
molecular size, was generated by peptide spot synthesis.4
It is a modestly potent BACE 1 inhibitor with IC50 of
0.5 lM. After independent modification of the central
core (statine #1 and statine #2), the C-terminus and the
N-terminus of 3, a new lead 9 with IC50 of 21 nM was
generated, which is described in this communication.5;6
N
N
N
N
N
O
H
H
H
OH
O
OH
O
O
Boc-VM-[Sta}-V-4-aminmethylbenzoicacid (1)
O
O
O
O
H
H
H
OH
N
N
N
N
N
N
N
H
H
H
H
O
O
O
OH
O
O
S
OH
µ
Ac-VM-[Sta}-VAEF (2) IC50 0.3 M
S
O
OH
O
O
H
H
N
N
N
N
NH2
H
H
O
OH
O
statine #2
C-terminus
statine #1
N-terminus
* Corresponding author. Tel.: +1-484-865-7867; fax: +1-484-865-9399;
e-mail: hub@wyeth.com
3
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.04.068

3458
B. Hu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3457–3460
O
O
NH2
d,e,f
H
NH2HCl
a,b,c
HN
N
H2N
N
H
MeO2C
MeO2C
OMe
OH
O
4
5
6
O
g,h,i
or j, k
l
O
OH
O
O
OH
O
O
H
H
H
N
N
N
H2N
N
N
H
N
H
N
N
H
H
H
O
OH
O
OMe
OH
O
OMe
8
O
O
7
m
O
OH
O
O
O
OH
O
O
H
H
H
H
N
N
N
N
N
N
H
N
H
N
N
N
H
H
H
H
O
OH
O
OH
O
OH
10
O
OH
O
O
9
Scheme 1. Initial approach: Reagents and conditions: (a) N-Boc-Val, EDC; (b) TFA; (c) aq NaHCO3, 80% for steps (a)–(c); (d) N-Boc-phenyl
statine, EDC; (e) TFA; (f) aq NaHCO3, 37% for steps (d)–(f); (g) N-Boc-statine, EDC; (h) TFA; (i) aq NaHCO3, 22% for steps (g)–(i); (j) N-Fmoc-
statine, EDC, HOBt; (k) piperidine, 51% for steps (j)–(k); (l) N-Ac-Leu, EDC, HOBt; (m) LIOH, 37% for steps (l)–(m).
The representative synthesis of various inhibitors con-
taining the bis-statine core is outlined in Scheme 1. Boc-
protected valine was coupled to 4-aminomethyl-benzoic
acid methyl ester (4) using N-ethyl-N0-(dimethylamino-
propyl)carbodiimide hydrochloride (EDC) as the cou-
pling agent and N-methyl-morpholine as the base. The
Boc-group was removed by treatment with trifluoro-
acetic acid (TFA) in CH2Cl2 and the resulting free amine
5 was allowed to couple subsequently with N-Boc-phen-
yl statine (4(S)-amino-3(S)-hydroxy-5-phenyl pentanoic
acid) and N-Boc-statine using the same coupling/de-
protection sequence to give the amine 7. However, the
yield for the statine coupling/deprotection sequence was
only 22% and the amine 7 was isolated with unidentified
impurities. This low yield was resulted from the
incomplete conversion of 6 and acidic instability of
the amine 7. Modifications were therefore made to use
the Fmoc-protected statine instead of Boc-protected
statine and 1-hydroxybenzotriazole (HOBt) in addi-
tional EDC. The Fmoc-group was then removed by
treatment with piperidine to give the pure amine 7 in
55% yield for the coupling/deprotection steps. The N-
Ac-terminated leucine inhibitor 9 was successfully pre-
pared by the improved peptide coupling conditions
(EDC/HOBt) followed by basic hydrolysis. Other
inhibitors (such as 4,5-dehydro-leucine derivative 10) in
Tables 1 and 2 were prepared by analogous sequence as
described above.
Table 1. SARs on P1 and P2 modifications
O
OH
O
P1
O
H
H
N
N
N
N
N
H
H
H
O
P2
OH
O
OH
O
Compound P1
P2
IC50 (lM)a
9
Benzyl
2-Methylbutyl
2-Methylbutyl
Benzyl
0.021
0.65
0.46
0.05
11
12
13
14
Methylsulfanyl ethyl
c-Hexyl methyl
Methylsulfinyl ethyl
2-Methylbutyl
Benzyl
2-Methylbutyl 34.0
a Concentration necessary to inhibit 50% of enzyme activity in BACE 1
FRET assay.
optimized activity. Based on a SAR study of a previ-
ously described mono-statine based BACE 1 inhibi-
tors,2;3 it was anticipated that the introduction of
4-aminomethylbenzoic acid at the C-terminus would be
optimal for BACE 1 inhibition. To our delight, an
aminomethylbenzoic acid bis-statine derivative 9 indeed
resulted in 24-fold increase in potency compared to 3. It
has an IC50 value of 21 nM. Replacement of the iso-
propyl substituent with isobutene (4,5-dehydroleucine
derivative 10) resulted in a minor increase in BACE 1
inhibition (IC50 value of 16 nM for 10 vs 21 nM for 9). A
preliminary study to investigate the amino methyl ben-
zoic acid replacement by peptide spot synthesis had been
shown to yield a significant loss of affinity.4 On the basis
of the results of peptide spot synthesis study and the
reported mono-statine SARs,3 we did not manipulate
the C-terminus further. Hence, our efforts were directed
toward the modification of the center core (P1/P2 side
chains). As shown in Table 1, a large substituent on P1
position is preferred for the inhibitory activity.
The structure7 of various bis-statine based peptides and
their inhibitory potencies against BACE 1 FRET
(fluorescence resonance energy transfer) assay8 are
shown in Tables 1 and 2. As illustrated, our initial bis-
statine inhibitor 3 was conceptually subdivided into
three regions for discussion purposes: an N-terminus, a
central core (statine #1, statine #2), and a C-terminus.
These regions were individually modified to probe the

B. Hu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3457–3460
3459
Table 2. SARs on N-terminus modifications
OH
O
O
H
H
N
N
R
N
N
H
H
OH
O
OH
O
Compound
R
IC50 (lM)
Compound
R
IC50 (lM)a
13
15
16
17
N-Ac-Leu
N-Ac-Val
N-Ac-Gly
N-Ac-Phe
0.05
0.67
9.0
18
19
20
21
N-Ac-Glu
Pro
42
1.3
H
Boc
>100
>100
5.6
a Concentration necessary to inhibit 50% of enzyme activity in BACE 1 FRET assay.
Replacement of the large benzyl group of phenyl statine
at P1 with a smaller group, such as a 2-methylbutyl
group of statine or a methylsulfanyl group of methio-
nine statine (4(S)-amino-3(S)-hydroxy-6-methylsulfan-
yl-hexanoic acid), resulted in more than 20-fold decrease
in potency (9 vs 11 and 12). The potency lost by a small
P1 site of 11 can be recovered back to the level (13 vs 11)
seen with 9 by a larger size P1 site of cyclohexyl statine
(4(S)-amino-5-cyclohexyl-3(S)-hydroxy-pentanoic acid).
However, the size of substituent on P2 seems to be not as
crucial as P1 for the activity (9, 13 vs 11, 12). Interest-
Figure 1. The modeled BACE 1 structure in complex with inhibitor 9.
ingly, oxidation of the methyl sulfide of 12 resulted in a
significant loss (74-fold) of potency (12 vs 14).
In the course of our study on the N-terminus (Table 2),
we observed that the N-Ac leucine is the optimal residue
in the S3 pocket. A smaller residue, such as N-Ac valine
(15) or N-Ac glycine (16), significantly reduced the
potency by 13.5-fold and 180-fold, respectively. On the
other hand, a larger residue, such as N-Ac phenylalanine
derivative 17, also resulted in loss more than 100-fold of
potency. Furthermore, a polar carboxylic acid analog 18
or a conformationally constrained proline analog 19 was
also much less active. Truncation of the N-terminus
portion of inhibitor was tried to determine the minimal
size requirement for the peptide inhibitor. However,
removal of the N-Ac leucine from 9 resulted in complete
loss of potency (20, IC50 > 100 lM). Surprisingly, N-
terminus Boc protected statine analogs,3 which have
been demonstrated to be potent BACE 1 inhibitors,
resulted in an essentially inactive inhibitor 21 in our
series. This suggests that peptapeptidic structures are
required for decent potency.
benzyl or cyclohexyl methyl, while the S2 is the mixed
hydrophobic and polar pocket that seemed to accom-
modate both large and small residues. The N-Ac leucine
on the N-terminus had close contact with S3 pocket and
formed three hydrogen bonds with the BACE 1 site
including Q73 according to the hypothetical model. The
importance of the interaction could be reflected in the
substantial loss of the activity by truncation of the N-
terminus portion discussed earlier. The C-terminal
portion of the inhibitor 9 showed a favorable hydrogen
bond interaction between the benzoic acid and lysine
K224.
In summary, a new series of peptide-based BACE 1
inhibitors have been developed, characterized by the
presence of a bis-statine central core. After independent
modifications on the C-terminal, N-terminal, and the
central core of the initial lead, the spacer requirements
for the preferred substrate–enzyme interaction have
been generally defined. This information can serve as a
convenient starting point from which small molecular-
based BACE 1 inhibitors can be developed.
Based on the SAR from these peptidic inhibitors and the
in-house crystallographic results,4 a model for inhibitor
9 bound in the active site was constructed using the in-
house X-ray structure of BACE 1 in complex with 3.
The program QXP/FLO9 was used in constructing ini-
tial model for inhibitor 9 and then further minimized
using MMFF94 force field10 implemented in SYBYL.11
The details regarding the interactions with the specific
pocket is illustrated in Figure 1. As postulated earlier,
the secondary alcohol of the phenyl statine moiety
formed hydrogen bonds with the two catalytic aspartic
acids (D32 and D228). The S1 is large hydrophobic
pocket and has very nice contact with the groups like
Acknowledgements
The authors acknowledge the members of the Wyeth
Discovery Analytical Chemistry group for analytical
and spectral determinations, Laura Lin, Mark Stahl,
and Richard Zollner for producing and purifying the
protein for functional and structural analyses.

3460
B. Hu et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3457–3460
5. For molecular modeling studies on BACE 1, see: (a)
References and notes
Tounge, B. A.; Reynolds, C. H. J. Med. Chem. 2003, 46,
2074; (b) Fan, K. Y.; Hu, B.; Cole, D.; Bridges, K.;
Chopra, R.; Manas, E. S.; Katz, A.; Alvarez, J. C.;
Lovering, F. E.; Zhou, P.; Jin, G.; Cowling, R.; Bard, J.
Abstracts of Papers, 226th ACS National Meeting, New
York, NY, United States, September 7–11, 2003, COMP-
134.
1. For recent reviews on BACE 1, see: (a) Varghese, J.; Beck,
J. P.; Bienkowski, M. J.; Sinha, S.; Heinrikson, R. L. J.
Med. Chem. 2003, 46, 4625; (b) Vassar, R. Adv. Drug
Deliv. Rev. 2002, 54, 1589; (c) Vassar, R. J. Mol. Neurosci.
2001, 17, 157; (d) Ghosk, A. K.; Hong, L. Curr. Med.
Chem. 2002, 9, 1135; (e) Roggo, S. Curr. Top. Med. Chem.
2002, 2, 359.
2. For recent papers on statine-containing BACE 1 inhibitors
see: (a) Tung, J. S.; Davis, D. L.; Anderson, J. P.; Walker,
D. E.; Mamo, S.; Jewett, N.; Hom, R. K.; Sinha, S.;
Thorsett, E. D.; John, V. J. Med. Chem. 2002, 45, 259; (b)
Hu, J.; Cwi, C. L.; Smiley, D. L.; Timm, D.; Erickson, J.
A.; McGee, J. E.; Yang, H.; Mendel, D.; May, C. P.;
Shapiro, M.; McCarthy, J. R. Bioorg. Med. Chem. Lett.
2003, 13, 4335; (c) 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; (d) Ghosh, A. K.; Bilcer, G.; Harwood, C.;
Kawahama, R.; Shin, D.; Hussain, K. A.; Hong, L.; Loy,
J. A.; Nguyen, C.; Koelsch, G.; Ermolieff, J.; Tang, J. J.
Med. Chem. 2001, 44, 2865.
6. For a preliminary account of SARs on bis-statine deriv-
atives, see: Hu, B.; Fan, K.; Bridges, K.; Cole, D.; Zhou,
P.; Ellingboe, J.; Jin, G.; Cowling, R.; Bard, J. Abstracts,
58th Northwest Regional Meeting of the American
Chemical Society, Bozeman, MT, United States, June
12–14, 2003, 171.
7. All final compounds and novel intermediates were char-
acterized by 1H NMR, mass spectrometry, and high-
resolution mass spectrometry or CHN analysis.
8. BACE 1 inhibition was measured using recombinant
enzyme produced from mammalian expression as de-
scribed similarly in Ref. 2a. Fluorogenic substrate Abz-
SEVNL-DAEFR-DPa was used with 10 nM human
BACE 1 in 0.05 M Na-acetate þ 0.05% CHAPS and 25%
PBS, pH 4.5 at room temperature. The excitation wave-
length was 354 nm, and the emission wavelength was
280 nm. Details of this assay will be described shortly.
9. Mcmartin, C.; Bohacek, R. S. J. Comp.-Aided Mol. Des.
1997, 11, 333.
3. John, V.; Tung, J.; Fang, L.; Mamo, S. S. WO 0077030,
2000; Chem. Abstr. 2000, 134, 56968.
4. Bridges, K.; Chopra, R.; Lin, L.; Svenson, K.; Tam, A.;
Jin, G.; Cowling, R.; Lovering, L.; Akopian, L. N.; Bard,
J.; Elizabeth, D.; Bethany, A.; Edward, R. L.; Tod, H. M.;
Richard, S. Z.; William, S. S.; Mark, L. S.; Ronald, W. K.,
unpublished results.
10. Halgren, T. A. J. Comp. Chem. 1996, 17, 616, and the
references cited therein.
11. Drug Discovery Software SYBYL 6.7 developed by Tripos,
St. Louis, USA.