BACE-1 inhibitors Part 1: Identification of novel hydroxy ethylamines (HEAs)

Article abstract of DOI:10.1016/j.bmcl.2007.12.017

Inhibition of the aspartyl protease BACE-1 has the potential to deliver a disease-modifying therapy for Alzheimer's disease. Herein, is described the lead generation effort which resulted, with the support of X-ray crystallography, in the discovery of pot

Full text of DOI:10.1016/j.bmcl.2007.12.017

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1011–1016
BACE-1 inhibitors Part 1: Identification of novel
hydroxy ethylamines (HEAs)
Brian Clarke, Emmanuel Demont,* Colin Dingwall, Rachel Dunsdon, Andrew Faller,
Julie Hawkins, Ishrut Hussain, David MacPherson, Graham Maile, Rosalie Matico,
Peter Milner, Julie Mosley, Alan Naylor, Alistair O’Brien, Sally Redshaw,
David Riddell, Paul Rowland, Virginie Soleil, Kathrine J. Smith, Steven Stanway,
Geoffrey Stemp, Sharon Sweitzer, Pam Theobald,
David Vesey, Daryl S. Walter, John Ward and Gareth Wayne
Neurology and Gastrointestinal Centre of Excellence for Drug Discovery, GlaxoSmithKline R&D, New Frontiers Science Park,
Third Avenue, Harlow, Essex CM19 5AW, United Kingdom
Received 30 October 2007; revised 7 December 2007; accepted 11 December 2007
Available online 15 December 2007
Abstract—Inhibition of the aspartyl protease BACE-1 has the potential to deliver a disease-modifying therapy for Alzheimer’s dis-
ease. Herein, is described the lead generation effort which resulted, with the support of X-ray crystallography, in the discovery of
potent inhibitors based on a hydroxy ethylamine (HEA) transition-state mimetic. These inhibitors were capable of lowering amyloid
production in a cell-based assay.
Ó 2007 Elsevier Ltd. All rights reserved.
Alzheimer’s disease is a devastating neurodegenerative
disorder characterized by the progressive formation of
insoluble amyloid plaques and neurofibrillary tangles
in the brain.1 These plaques are mainly comprised of a
R3
R4
N
R1
small 4 kDa amyloid-b (Ab) peptide, generated by the
proteolytic processing of a larger membrane bound pre-
cursor protein, known as the amyloid precursor protein
(APP). Cleavage of APP by BACE-1 (for b-site APP
cleaving enzyme, also known as b-secretase, Memap-
sin-2 or Asp-2) releases an extracellular soluble APP
fragment. This is concomitant with the generation of a
membrane-tethered C-terminal fragment that is subse-
quently processed by c-secretase to generate Ab pep-
tides, predominantly of 40 or 42 amino acids in length
(Ab40, Ab42).2 In an alternative non-amyloidogenic
pathway, cleavage of APP by a-secretase within the
amyloid-b region of APP precludes the release of intact
Ab.
R2
R5
+
N
H
BOCNH
O
O
1 R1 = iBu (S, S)
2 R1 = Bn (S, S)
3 R1 = Bn (S, R)
R3
R4
R1
N
R5
H2N
N
a, b
OH R2
O
4-6
R3
R4
N
O
R1
R5
R6
N
N
c
H
OH R2
O
Keywords: Alzheimer; BACE-1; Aspartic protease; Hydroxy
ethylamine.
Figure 1. Reagents and conditions: (a) EtOH, reflux; (b) TFA,
CH2Cl2, 25 °C or HCl, CH3CN or dioxan, 25 °C; (c) R6COOH,
EDACÆHCl, HOBT, CH2Cl2 or DMF, 25 °C.
*
Corresponding author. Tel.: +44 (0) 1279 643468; fax: +44 (0) 1279
622727; e-mail: emmanuel.h.demont@gsk.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.12.017

1012
B. Clarke et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1011–1016
Table 1. BACE-1 inhibition for compounds 7–10
S1
'
S1
'
'
P1
P1
S2
R1
S2
O
O
O
H
N
S
'
N
N
R2
P2
H
OH
O
P2
Compound
R1
CH(OH)
R2
BACE-1
a
IC50 (lM)
7
CH2Ph
R
R
S
H
5.4
8
CH2CH(CH3)2
CH2Ph
CH2Ph
H
>500
>500
>500
9
10
H
CH3
R
a For assay protocol, see Ref. 7. IC50 value is the mean of three experiments.
Table 2. SAR at the non-prime side of BACE-1 HEA inhibitors
O
H
R1
R2
N
N
N
H
H
OH
O
Compound
R1
R2
H
BACE-1a IC50 (lM)
BACE-2 IC50 (lM)
Cat-D IC50 (lM)
7
CH3SO2–
5.4 (3)
42
7.7
O
S
O
11
12
13
14
H
1.9 (4)
5.3 (4)
62 (1)
1.8 (1)
>21
79
10
N
O
H
26
N
H
CH3CONH–
>100
>100
n-C5H11SO2–
H
3.6
0.6
a In all tables, IC50s reported are means of the values of n different experiments, n being reported in bracket and identical for BACE-1, BACE-2 and
Cat-D. Each IC50 is within threefold of the mean value.
Mice genetically deficient in BACE-1 show no Ab pro-
duction3a,b and are healthy, viable and fertile,3a–c sug-
gesting that BACE-1 inhibition is unlikely to be
associated with mechanism-based toxicity. For these
reasons, inhibition of this enzyme is considered to be
an attractive therapeutic target. Following the classifica-
tion of BACE-1 as an aspartyl protease,4 extensive ef-
forts have resulted in the discovery of potent and
selective inhibitors.5 In most cases, the strategy for the
design of inhibitors has been based on the transition-
state mimetic concept, an approach that has been used
successfully to design inhibitors of other aspartyl prote-
ases, most notably HIV protease.6 This approach typi-
cally relies on replacement of the scissile amide bond
of an appropriate substrate with a stable mimetic of
the putative transition state.
GSK188909 was recently reported to be the first BACE-
1 inhibitor capable of lowering brain Ab in APP trans-
genic mice following oral administration.7 Herein, and
in subsequent letters, the studies which led to the discov-
ery of this orally active inhibitor are described.
Bearing in mind the need to deliver inhibitors with drug-
like properties, we preferred not to start with large pep-
tidic compounds based on the EVKM-DAE (APP Wild
Type, WT) or EVNL-DAE (APP Swedish Mutation,
SWE) sequences of the natural substrates. Our strategy
instead focused on the preparation of arrays of hydroxy
ethylamines (HEAs) starting from the known8 Boc
epoxides 1, 2 and 3, and limiting ourselves to molecules
with MW < 700 and containing no more than 2 amide
bonds (Fig. 1).

B. Clarke et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1011–1016
1013
Figure 2. Key interactions between inhibitor 14 and BACE-1.
With this in mind, a number of a-amino amides were se-
lected to add to epoxides 1–3. Removal of the Boc pro-
tecting group provided amines 4–6 which were then
coupled to a small set of carboxylic acids.
Interestingly, the co-crystal structure showed that the
sulfone on the non-prime side of the inhibitor makes a
hydrogen bond with the enzyme, allowing access to
the S3 pocket from the other meta-position (Fig. 3). It
seemed reasonable that filling this narrow pocket with
an appropriate substituent might increase the potency
of our inhibitors.
From an initial array, 7 was one of the first compounds
to show micromolar activity against BACE-1 (Table 1).
The analogue 8, with Leu in the P1 position, proved
much less active. This was somewhat surprising since
the Swedish-mutant substrate has Leu in the P1 position
and is turned over more rapidly than the wild-type sub-
strate. The requirement for R-stereochemistry at the hy-
droxyl group was confirmed by compound 9 derived
from the S,R-epoxide 3. N-Methylation of the hydroxy
ethylamine isostere was not tolerated (compound 10).
All subsequent SAR was therefore developed using the
central core of compound 7.
This idea was explored further after first replacing the sul-
fone H-bond acceptor with a pyrrolidinone. Incorpora-
tion of this group gave inhibitors with similar potency
(compare 7 and 12, Table 2), but it was felt that the lower
polar surface area would increase the chance of obtaining
compounds with a degree of brain penetration. The
substituted benzoic acids required to complete this study
0
Initially, the effect of modifications of the P10 and the P2
substituents on BACE-1 inhibition was explored, but
despite extensive effort, no improvement of potency
could be found compared to the hit 7. The contribution
of the meta-substituted benzamide non-prime side was
then investigated (Table 2), choosing BACE-2 and
Cat-D,9 both structurally related to BACE-1, as repre-
sentative aspartyl protease for selectivity screens. A
large range of hydrogen bond acceptors (HBAs) were
tolerated in this position (see compounds 11 and 12 as
representative examples) whilst inhibitors with other
substitution patterns were less potent (see compound
13 for example).10 More lipophilic meta-HBAs led to a
modest increase in potency (compound 14) which was
sufficient to obtain a co-crystal structure of the inhibitor
with a human BACE-1 construct.11
The key H-bond interactions between inhibitor 14 and
BACE-1 are depicted below (Fig. 2) and help with
understanding some of the results presented above.
Figure 3. Non-prime side of inhibitor 14 bound to BACE-1 and
potential access to the S3 pocket.

1014
B. Clarke et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1011–1016
O
O
O
I
48%
a, b
OH
N
O
j, g
O
Br
O
Br
N
OH
O
O
O
H2N
OH
N
65%
X
O
R1
k, g
a,c, d, e
NO2
NH2
l, g
O
O
O
O
O
N
72%
H2N
23%
O
O
O
O
c, d, e
f, g, h, i
O
OH
OBn
H
N
H2N
N
H
OH
O
15
Figure 4. Reagents and conditions: (a) SOCl2, MeOH, reflux; (b) pyrrolidinone, Pd2(dba)3, xantphos, Cs2CO3, dioxane, 55 °C; (c) Cl(CH2)3COCl,
NEt3, CH2Cl2, 25 °C; (d) NaH, THF, 25 °C; (e) H2, Pd/C, MeOH, 25 °C; (f) BnBr, K2CO3, acetone, reflux; (g) NaOH, THF/H2O, 25 °C; (h) DPPA,
toluene, 80 °C then (CH3)3Si(CH2)2OH; (i) TBAF, THF, 25 °C; (j) BINAP, alkyl amine, toluene, K2CO3, Pd(OAc)2, 100 °C; (k) aldehyde or ketone,
(CH2Cl)2, NaHB(OAc)3, 25 °C; (l) R1Br or R1I, acetone, K2CO3, reflux.
Table 3. SAR at the S3 pocket using a pyrrolidinone as HBA
O
X
N
H
N
N
N
H
H
OH
O
R1
BACE-1 IC50 (nM) BACE-2 IC50 (nM) Cat-D IC50 (nM) Ab40a,b IC50 (lM) Ab42a,b IC50 (lM)
Compound
X
R1
12
16
17
18
19
20
21
CO
CO
CO
CO
CO
H
5270 (4)
745 (2)
59 (1)
78,530
28,600
5130
25,630
30,015
7940
—
—
—
—
–OCH3
–OC2H5
0.97
1.13
—
1.42
1.18
—
–OCH(CH3)2 120 (1)
32 (1)
4790
400
12,300
430
6170
–OC5H11
CH2 –OC2H5
CO –OC3H7
1450 (1)
605 (2)
12,880
8235
—
1.69
—
1.38
47,080
a See Ref. 7 for protocol.
b In Tables 3 and 4, IC50 values are means of at least two separate experiments. Each IC50 is within threefold of the mean value.
were prepared as depicted in Figure 4. The acids were cou-
pled to amine 15 to complete the synthesis of the inhibi-
tors. Access to the S3 pocket could be achieved via a
nitrogen, oxygen, or carbon linker.

B. Clarke et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1011–1016
1015
O
O
R1
N
N
N
22
H
H
OH
NH
Non peptidic RHS
Figure 6. Towards drug-like inhibitors.
that allows access and additive binding in the S3 pocket
by a group in the other meta-position.
The latter hypothesis appeared to be corroborated by a
co-crystal structure of compound 21 with the enzyme
(Fig. 5) in which an appropriate H-bond was observed.
The structure did not, however, make it entirely clear as
to why an n-propyl chain led to a 5- to 20-fold loss in
potency when compared with shorter or longer alkyl
chains in the same position.12
Figure 5. Key interactions between the non-prime side of inhibitor 21
and the enzyme.
Extensive exploration of the SAR at the S3 pocket was
initiated, with the aim of increasing activity and selectiv-
ity whilst keeping the polar surface area as low as possi-
ble in order to improve the likelihood of obtaining brain
penetrant inhibitors. It appeared that a linear three-
atom chain gave the best compromise between size, po-
tency and selectivity (Table 4, compare 22 and 23–25),
with the NH-linked derivatives being the most potent
and selective (compare 22 and 17, 26).
An additional substituent of sufficient bulk did appear
to improve binding and inhibitors 17–19 (Table 3), for
example, were more potent than the unsubstituted ana-
logue 12 or derivative 16 with the smaller methoxy sub-
stituent. On examining the length of the chain in the S3
pocket, it was found that a three-atom chain length led
to a more substantial increase in potency (100-fold,
compare 17, 18 and 12, 16), whilst compounds with a
longer chain in this position proved no more active
and were generally less selective versus BACE-2 and
Cat-D (compound 19). The inhibition of BACE-1 by
these compounds was also demonstrated to result in a
reduction in amyloid-b (Ab40 and Ab42) production
in cells expressing APP WT.
At this stage, efforts were shifted towards further simpli-
0
0
fication of the prime-side by replacement of the P1 –P2
amide, as it was felt that this strategy was more likely to
lead to the identification of drug-like inhibitors of
BACE-1 with the potential for good oral bioavailability
and CNS penetration (Fig. 6).
Replacement of the pyrrolidinone by a pyrrolidine ring
(compound 20) resulted in a 25-fold reduction in po-
tency, suggesting that a hydrogen bonding interaction
with the enzyme in this position was needed in order
to lock the benzamide non-prime side in a conformation
In summary, in this first round of optimisation of hydro-
xy ethylamine BACE-1 inhibitors, P2 and P3 substitu-
ents which gave good potency and selectivity over
Table 4. SAR at the S3 pocket: influence of linker and nature of the substituent
O
O
H
N
N
N
N
H
H
OH
O
X
R1
Compound
X
R1
BACE-1
IC50 (nM)
BACE-2
IC50 (nM)
Cat-D IC50
(nM)
Ab40 IC50
(lM)
Ab42
IC50 (lM)
22
23
24
NH
NH
NH
C2H5
13 (2)
39 (6)
180 (1)
1810
830
2695
5413
3090
0.31
0.33
0.36
0.28
CH(CH3)2
CH2CH
(CH3)2
650
25
26
Pyrrolidine
CH2
860 (3)
72 (1)
8793
1580
9277
4570
C2H5
0.99
0.69

1016
B. Clarke et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1011–1016
amel, M.-C.; Ellis, J.; Lai, M.-T.; Espeseth, A. S.; Shi, X.-
P.; Jin, L.; Colussi, D.; Pietrak, B.; Huang, Q.; Xu, M.;
Simon, A. J.; Graham, S. L.; Vacca, J. P.; Selnick, H. J.
Med. Chem. 2007, 50, 3431; (d) 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; (e) Kortum, S. W.; Benson,
T. E.; Bienkowski, M. J.; Emmons, T. L.; Prince, D. B.;
Paddock, D. J.; Tomasselli, A. G.; Moon, J. B.; LaBorde,
A.; TenBrick, R. E. Bioorg. Med. Chem. Lett. 2007, 17,
3378; (f) Hanessian, S.; Yun, H.; Hou, Y.; Yang, G.;
Bayrakdarian, M.; Therrien, E.; Moitessier, N.; Roggo, S.;
Veenstra, S.; Tintelnot-Blomley, M.; Rondeau, J.-M.;
Ostermeier, C.; Strauss, A.; Ramage, P.; Paganetti, P.;
Neumann, U.; Betschart, C. J. Med. Chem. 2005, 48, 5175,
and references cited herein.
other aspartyl proteases were identified. Efforts to re-
duce the peptidic nature of these compounds and in-
crease the drug-like properties of this series will form
the subject of a subsequent publication.
References and notes
1. Glenner, G. G.; Wong, C. W. Biochem. Biophys. Res.
Commun. 1984, 120, 885.
2. Selkoe, D. J. Annu. Rev. Cell Biol. 1994, 10, 373.
3. (a) Luo, Y.; Bolon, B.; Kahn, S.; Bennett, B. D.; Babu-
Khan, S.; Denis, P.; Fan, W.; Kha, H.; Zhang, J.; Gong,
Y.; Martin, L.; Louis, J.-C.; Yan, Q.; Richards, W. G.;
Citron, M.; Vassar, R. Nat. Neurosci. 2001, 4, 231; (b) Cai,
H.; Wang, Y.; McCarthy, D.; Wen, H.; Borchelt, D. R.;
Price, D. L.; Wong, P. C. Nat. Neurosci. 2001, 4, 233; (c)
Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.;
Branstetter, D. G.; Chen, K. S.; Freedman, S. B.; Frigon,
N. L.; Games, D.; Hu, K.; Johnson-Wood, K.; Kappen-
man, K. E.; Kawabe, T. T.; Kola, I.; Kuehn, R.; Lee, M.;
Liu, W.; Motter, R.; Nichols, N. F.; Power, M.; Robert-
son, D. W.; Schenk, D.; Schoor, M.; Shopp, G. M.; Shuck,
M. E.; Sinha, S.; Svensson, K. A.; Tatsuno, G.; Tintrup,
H.; Wijsman, J.; Wright, S.; McConlogue, L. Hum. Mol.
Genet. 2001, 10, 1317.
4. (a) Hussain, I.; Powell, D.; Howlett, D. R.; Tew, D. G.;
Meek, T. D.; Chapman, C.; Gloger, I. S.; Murphy, K. E.;
Southan, C. D.; Ryan, D. M.; Smith, T. S.; Simmons, D. L.;
Walsh, F. S.; Dingwall, C.; Christie, G. Mol. Cell. Neurosci.
1999, 14, 419; (b) Sinha, S.; Anderson, J. P.; Barbour, R.;
Basi, G. S.; Caccavello, R.; Davis, D.; Doan, M.; Dovey, H.
F.; Frigon, N.; Hong, J.; Jacobson-Croack, K.; Jewett, N.;
Keim, P.; Knops, J.; Lieberburg, I.; Power, M.; Tan, H.;
Tatsuno, G.; Tung, J.; Schenk, D.; Seubert, P.; Suomensa-
ari, S. M.; Wang, S.; Walker, D.; Zhao, J.; McConlogue, L.;
John, V. Nature 1999, 402, 537; (c) 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; (d) Yan, R.; Bienkowski, M. J.; Shuck, M. E.; Miao,
H.; Tory, M. C.; Pauley, A. M.; Brashler, J. R.; Stratman,
N. C.; Mathews, W. R.; Buhl, A. E.; Carter, D. B.;
Tomasselli, A. G.; Parodi, L. A.; Heinrikson, R. L.;
Gurney, M. E. Nature 1999, 402, 533; (e) Lin, X.; Koelsch,
G.; Wu, S.; Downs, D.; Dashti, A.; Tang, J. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 1456.
6. Rodriguez-Barrios, F.; Gago, F. Curr. Top. Med. Chem.
2004, 4, 991.
7. Hussain, I.; Hawkins, J.; Harrison, D.; Hille, C.; Wayne,
G.; Cutler, L.; Buck, T.; Walter, D.; Demont, E.; Howes,
C.; Naylor, A.; Jeffrey, P.; Gonzalez, M. I.; Dingwall, C.;
Michel, A.; Redshaw, S.; Davis, J. B. J. Neurochem. 2007,
100, 802.
8. Luly, J. R.; Dellaria, J. F.; Plattner, J. J.; Soderquist, J. L.;
Yi, N. J. Org. Chem. 1987, 52, 1487.
9. Demont, E.; Faller, A.; Macpherson, D.; Milner, P.;
Naylor, A.; Redshaw, S.; Stanway, S.; Vesey, D.; Walter,
D. PCT Int. Appl., 2004, WO 2004050619 and references
cited herein.
10. Substitution ortho to the amide functionality was in
general poorly tolerated (data not shown).
11. Asp2N153, 172, 223, 354Q/Fc was expressed in CHO
secretion system with thrombin cleavage sites engineered
after residues 45 and 460 (pro/thr/Asp2QQQQ/thr/Fc).
Pro/thr/Asp2QQQQ/thr/Fc was captured by ProSep-A
High Capacity resin (Bioprocessing Limited) and neutral-
ized eluate dialysed into 25 mM Hepes, 0.25 M NaCl, pH
7.4. Soluble Asp2QQQQ was released by cleavage with
bovine alpha thrombin (Haematologic Technologies Inc.).
ProSep A unbound of the thrombin digest was further
truncated by pepsin at pH 4.6. V61-E452 form of
Asp2QQQQ was purified from the neutralized pepsin
digest by MonoQ (GE Healthcare) at pH 7.4. Crystals of
apo Asp2 were grown at 20 °C using the hanging drop
vapour diffusion method combined with streak seeding.
Protein at 10 mg/mL was mixed in a 1:1 ratio with the
reservoir solution consisting of 10% PEG8000 and 0.1 M
glycine, pH 3.2. Crystals were soaked overnight in the
mother liquor to which approximately 0.1 mg of solid
compound was added. The crystals were cryoprotected by
serial transfer into a solution of mother liquor with 30%
glycerol and then flash-frozen in liquid nitrogen prior to
data collection.
5. Review (a) Durham, T. B.; Shepherd, T. A. Curr. Opin.
Drug Discovery Dev. 2006, 9, 776; See for examples (b)
Baxter, E. W.; Conway, K. A.; Kennis, L.; Bischoff, F.;
Mercken, M. H.; De Winter, H. L.; Reynolds, C. H.;
Tounge, B. A.; Luo, C.; Scott, M. K.; Huang, Y.; Braeken,
M.; Pieters, S. M. A.; Berthelot, D. J. C.; Masure, S.;
Bruinzeel, W. D.; Jordan, A. D.; Parker, M. H.; Boyd, R.
E.; Qu, J.; Alexander, R. S.; Brenneman, D. E.; Reitz, A.
B. J. Med. Chem. 2007, 50, 4261; (c) Stanton, M. G.;
Stauffer, S. R.; Gregro, A. R.; Steinbeiser, M.; Nantermet,
P.; Sankaranarayanan, S.; Price, E. A.; Wu, G.; Crouth-
The X-ray was collected at ESRF. The PDB deposition
codes and refinement details for the BACE-1 complex
˚
crystal structures are: 14 (2viy, 1.8 A resolution,
˚
R = 0.181, Rfree = 0.216); 21 (2viz, 1.6 A resolution,
R = 0.198, Rfree = 0.227).
12. This effect was also observed with HBAs other than
pyrrolidinone.