Second generation of BACE-1 inhibitors. Part 1: The need for improved pharmacokinetics

Article abstract of DOI:10.1016/j.bmcl.2009.03.165

Inhibition of the aspartyl protease BACE-1 has the potential to deliver a disease-modifying therapy for Alzheimer's disease. We have recently disclosed a series of transition-state mimetic BACE-1 inhibitors showing nanomolar potency in cell-based assays.

Full text of DOI:10.1016/j.bmcl.2009.03.165

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3664–3668
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Second generation of BACE-1 inhibitors. Part 1: The need for
improved pharmacokinetics
*
Nicolas Charrier, Brian Clarke, Leanne Cutler, Emmanuel Demont , Colin Dingwall, Rachel Dunsdon,
Julie Hawkins, Colin Howes, Julia Hubbard, Ishrut Hussain, Graham Maile, Rosalie Matico, Julie Mosley,
Alan Naylor, Alistair O’Brien, Sally Redshaw, Paul Rowland, Virginie Soleil, Kathrine J. Smith,
Sharon Sweitzer, Pam Theobald, David Vesey, Daryl S. Walter, 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
a r t i c l e i n f o
a b s t r a c t
Article history:
Inhibition of the aspartyl protease BACE-1 has the potential to deliver a disease-modifying therapy for
Alzheimer’s disease. We have recently disclosed a series of transition-state mimetic BACE-1 inhibitors
showing nanomolar potency in cell-based assays. Amongst them, GSK188909 (compound 2) had favor-
able pharmacokinetics and was the first orally bioavailable inhibitor reported to demonstrate brain amy-
loid lowering in an animal model. In this Letter, we describe the reasons that led us to favor a second
generation of inhibitors for further in vivo studies.
Received 25 February 2009
Revised 30 March 2009
Accepted 30 March 2009
Available online 17 April 2009
Keywords:
Alzheimer
Ó 2009 Elsevier Ltd. All rights reserved.
BACE-1
Aspartic protease
Hydroxyethylamine
Alzheimer’s disease (AD) is a devastating neurodegenerative
disorder for which no disease-modifying treatment is currently
available.1 It is characterized pathologically by the presence of
extracellular senile plaques and intracellular neurofibrillary tan-
gles.2 These plaques are mainly comprised of Amyloid-b (Ab) pep-
tides, predominantly of 40 or 42 amino acids in length (Ab40,
Ab42), generated by the proteolytic processing of a larger mem-
brane-bound precursor protein, known as the amyloid precursor
protein (APP). The identification of BACE-1 (b-site APP cleaving en-
zyme) as the elusive b-secretase, a key enzyme in the production of
Ab peptides was a major advance in the field of AD.3 Due to the po-
tential for disease modification, many pharmaceutical companies
and academic institutions have been actively developing BACE-1
inhibitors for the treatment of AD.4
We have recently reported the discovery of a series of novel
hydroxyethylamine BACE-1 inhibitors from a micromolar hit 1
(Fig. 1).5 Exemplars from this advanced series of inhibitors demon-
strated nanomolar potency in cell-based assays and GSK188909
(compound 2), an example with high potency and favorable
in vivo pharmacokinetics was successfully used to demonstrate
that inhibiting BACE-1 lowered brain amyloid levels in transgenic
mice.6 A second generation of inhibitors, typified by compound 3,
demonstrated improved pharmacokinetics whilst retaining the
same high level of potency.4c Herein, and in the following papers,
we will present the data that led us to the discovery of these inhib-
itors and the advantages they offered over earlier compounds.
Inhibitors 4 and 5 (Fig. 2) are further representative of our first
generation of inhibitors. As shown in Table 1, the intrinsic proper-
ties of these molecules are typical of those reported for marketed
renin or HIV protease inhibitors (PIs): high molecular weight
(>550), high polar surface area and a high molecular volume (as de-
scribed by their calculated molar refractivity (CMR) values). These
properties are fairly typical of inhibitors of aspartyl proteases that
need to make a relatively high number of hydrogen bonding and
lipophilic interactions to achieve significant levels of inhibition.
Compound 2 shows high microsomal clearance in all species
and also inhibits P450 enzymes at relatively low concentrations
(mainly the 3A4 isoforms—as is the case for most of the com-
pounds in this series, Table 1). However, it was possible to mini-
mize this inhibition by lowering the lipophilicity of the inhibitors
(compare 5–2 and 4) but to the detriment of their permeability.
Indeed, although compound 5 was permeable enough to show
good activity in a cellular assay,7 its oral bioavailability in rats,
even at high doses (100 mg/kg) and using an optimized vehicle
[suspension in 50% (w/v) Vit E TPGS (D-alpha-tocopheryl polyethyl-
ene glycol 1000 succinate), 40% (v/v) PEG 400 and 10% (v/v) etha-
nol], was poor (0.2%) despite having moderate blood clearance
(64 mL/min/kg) suggesting poor absorption. The opposite was
observed for the more lipophilic inhibitor 4 (Table 2): this inhibitor
* Corresponding author. Tel.: +44 0 1438764319; fax: +44 0 1438768232.
E-mail address: emmanuel.h.demont@gsk.com (E. Demont).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.03.165

N. Charrier et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3664–3668
3665
O
O
S
N
O
F
O
O
O
CF3
N
N
S
N
N
N
O
O
O
2
N
1
O
O
N
O
S
N
N
O
N
3
Figure 1. From micromolar hit to nanomolar orally available BACE-1 inhibitors.
O
O
F
O
S
N
O
O
CF3
N
N
N
N
N
N
O
O
N
N
N
4
5
Figure 2. Representative examples of BACE-1 inhibitors.
Table 1
In vitro profile of representative first generation BACE-1 inhibitors
Ab40a IC50 (nM)
CYP IC50
(lM)
Clic (m, r, h)
Log D @ pH 7.4
MW
PSA
CMR
b
Compd
BACE IC50
(lM)
2
4
5
0.002
0.009
0.010
5, 30
40, 519
78, 3633
>100, 9, 10, 6.2, 0.4, 1.6
75, 60, 15, 11, 75, 6.3
All >89 (3A4 DEF)
37, 8.7, 15
26, 7.6, 14
5.4, 1.7, 4.2
3.29
2.98
0.53
637
587
569
111
94
129
16.2
15.3
15.8
a
In SHSY5Y wild type and Swedish cells, respectively. See Ref. 6 for details.
1A2, 2C9, 2C19, 2D6, 3A4 (DEF, PPR).
Microsomal clearance in mouse, rat and human, respectively (in ml/min/g liver).
b
c
showed similar bioavailability (ꢀ35%) following oral administra-
tion (3 mg/kg po) or injection via the hepatic portal vein, suggest-
ing complete absorption.
Log D @ pH 7.4). As shown in Figure 3, prediction of absorption for
4 and 5 correlated with in vivo data. Compound 2 had some degree
These differences in absorption could be predicted in this series
using an in house model based on CMR and lipophilicity (measured
4
Cmpd 2
3.5
Table 2
In vivo pharmacokinetics for compounds 2 and 4
3
Dose route
Parameter
4
2
2.5
Intravenous (n = 3/4)
CLba (mL/min/kg)
Vss (L/kg)
51 9b
4.5 0.7
1.9 0.3
38 11
83 5d
5.5 0.6
1.5 0.3
Nd
Cmpd 4
2
t
(iv) (h)
½
Fipv (%)
1.5
Oral @ 3 mg/kg (n = 4)
Fpo (%)
t½ (po) (h)
AUC/dose (min g/L)
35 15c
2.1, nd, 1.8, nd
6.9 2.9
Nd
Nd
Nd
1
Good
Oral @ 10 mg/kg (n = 3/4)
Fpo (%)
67 45c
7
2e
t½ (po) (h)
AUC/dose (min kg/L)
1.8, 3.1, nd, nd
17.3 14.8
Nd
0.86 0.24
0.5
Good borderline
Poor borderline
a
0
CLb: in vivo clearance; Vss: volume of distribution; t : half-life; F: bioavail-
½
Cmpd 5
Poor
ability (following administration in portal vein:ipv or oral administration: po).
b
Dissolve in 2% (v/v) DMSO then added to saline containing 10% (w/v) kleptose.
Solution in 5% (v/v) ethanol, Capmul MCM C8 and solutol HS 15 (20:80).
0.9% (w/v) saline containing 10% (w/v) kleptose.
Dissolved in 1% (v/v) Tween 80 and 1% (w/v) methylcellulose aqueous. Nd = not
c
14.5 15 15.5 16 16.5 17 17.5 18 18.5
d
CMR
e
determined.
Figure 3. Prediction of absorption for the first generation of inhibitors.

3666
N. Charrier et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3664–3668
of oral bioavailability despite having high in vivo clearance (Table
2), again suggesting relatively good absorption, in agreement with
the prediction.
As can be seen, most of the compounds made in this series
were predicted to be poorly absorbed. The compounds which
were predicted to be well absorbed were in general too lipo-
philic to be metabolically stable. Overall, the chemical space
available for combining good potency (which tended to correlate
with high CMR), good absorption and drug-like lipophilicity was
small.
non linear kinetics (increased AUC/D between the 10 and 3 mg/kg
oral doses) suggest saturation of metabolism at the higher dose.
The oral bioavailability of compound 2 was lower than that of
compound 4 but its in vitro activity (especially in cell assays) made
it better suited for use as a tool molecule in animal models. Indeed,
it proved efficacious in lowering level of amyloid in the brain,
albeit at high doses (250 mg/kg po). This molecule, the result of
several rounds of optimization, was unlikely to be further
optimized in terms of in vitro activity and thus in order to identify
a molecule which was efficacious at lower oral doses, the poor
pharmacokinetics of our inhibitors had to be optimized further.
Our strategy towards optimization of the PK properties focused
on improving metabolic stability as much as possible in
The in vivo profile of inhibitor 4 is similar to what has been
reported for HIV PIs: first pass elimination may explain the sub-
optimal bioavailability observed after ipv administration and the
Table 3
Influence of C-4 substitution to enzyme activity and cell activity
O
O
R1
O
CF3
N
N
S
N
O
O
N
R2
N
A
B
Compd
R1
R2
BACE-1a IC50
(l
M)
BACE-2 IC50
(lM)
Cat-D IC50 (lM)
6
7
8
9
10
11
A
A
A
B
B
B
H
CH3
OCH3
H
CH3
OCH3
0.049
0.063
0.110
0.005
0.023
0.018
4.485
9.120
8.130
1.160
7.080
4.170
6.580
4.270
3.720
3.570
0.980
0.600
a
In all tables, IC50s reported are means of the values of three different experiments. Each IC50 is within threefold of the mean value.
O
O
O
O
O
O
O
N
N
N
N
N
N
I
O
O
O
c, d
e, f
a, b
O
O
N
B(OH)2
F3C
12
I
X
O
O
O
O
O2N
N
N
O
O
O2N
O2N
N
k
g, h
i, j
O
O
N
13
N
N
N
NO2
N
O
O
O
O2N
N
N
N
k, m-o
O
O
l
14
N
N
O
N
O
N
O
O
O
N
O2N
O2N
r, s
g, p, q
O
15
N
N
NO2
N
N
Scheme 1. Synthesis of key intermediates for the synthesis of biaryl non prime side substituent. Conditions: (a) pyrrolidinone, CuI, K3PO4, TMEDA, toluene, reflux 16 h, 28%;
(b) X, Pd(PPh3)4, K2CO3, DME/H2O 3:1, reflux 6 h, 60%; (c) (CF3CO)2O, pyridine, CH2Cl2, 25 °C, 61%; (d) allyl bromide, K2CO3, acetone, reflux, 3 h, 85%; (e) Grubbs II 10% w/w,
CH2Cl2, 25 °C, 16 h, 70%; (f) K2CO3, EtOH/H2O 5:1, 25 °C, 2 h then AcOEt, air, 16 h, 79%; (g) MeOH, H2SO4, reflux; (h) Pd/C, MeOH/c-C6H12, reflux, 80% (2 steps); (i) EtI, K2CO3,
DMF, 60 °C, 8 h, 50%; (j) HCOOH, 100 °C, 45 min, 79%; (k) NH4COOH, Pd/C, MeOH/H2O, 70 °C, 30 min, 81%; (l) NaNO2, AcOH, 60 °C, 1 h, 78%; (m) Cl(CH2)3COCl, NEt3, CH2Cl2,
25 °C; (n) NaH, THF, 25 °C; (o) NaH, DMF, 25 °C then EtI, 2 h, 61%; (p) Pd/C, MeOH/c-C6H12, reflux, 84% (2 steps); (q) NaNO2, HCl, H2O, 0 °C then urea, H2SO4, 50 °C, 15 min, 22%;
(r) EtI, K2CO3, DMF, 25–45 °C, 30 min, 35%; (s) NH4COOH, Pd/C, MeOH/H2O, 60 °C, 30 min, 80%;

N. Charrier et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3664–3668
3667
compounds which were predicted, using the model outlined above,
to be well absorbed. Qualitative in vitro studies with liver micro-
somes8 suggested that the main sites of metabolism for compound
2 were oxidation at the benzylic position on the prime side and de-
alkylation of the non prime side aniline substituent.
Our initial attempts to improve metabolic stability in this series
focused on trying to hinder the non prime side aniline based on an
early finding that a substitution ortho to the hydrogen bond acceptor
(HBA) was not detrimental to the activity (compare 6 and 9 with 7, 8
and 10, 11, respectively, Table 3). Moreover, substitution in this
position was not detrimental to the selectivity observed over
BACE-2 and Cat-D (two aspartic proteases used as our selectivity
panel). We hoped to prevent the observed metabolic dealkylation
by tying the aniline nitrogen back to the R2 substituent in a ring.
The syntheses of the most challenging bicyclic non-prime side
chemical intermediates 12–15 which emerged from this endeav-
our are shown in Scheme 1. The routes described also provided
the opportunity to introduce different HBAs at the meta position
(i.e., R1):
Figure 4. Superimposition on inhibitors 17 (green) and 22 (yellow) bound to
BACE-1.
As shown in Table 4, some 5,6 biaryls inhibitors were slightly
more potent than the parent compound 6 (compounds 16–20)
whilst others or 6,6-biaryls were less well tolerated (compounds
23 and 22, respectively). The latter result (compound 22) was
not easily rationalized and X-ray crystallography failed to provide
any significant insights. Superimposition of the structures of com-
pounds 17 and 22 bound to BACE-19 did not reveal any major
movements of enzyme residues that could explain the differences
in activity observed (Fig. 4).
unfortunately without demonstrating a significant improvement
in the in vivo pharmacokinetics. Compound 28 (Fig. 5), for exam-
ple, proved to be potent in vitro (BACE-1 IC50 = 35 nM) but rapidly
cleared in vivo (Rat CLb (1 mg/kg iv) = 83 5 mL/min/kg). This
result may be explained by either invoking alternative in vivo met-
abolic pathways compared to those observed for compound 2 in
microsomes, or by concluding that the changes made to the mole-
cule may have introduced other metabolic liabilities.
As shown with representative examples in Table 5, it was also
possible to maintain a good level of activity and at the same time
potentially minimize the risk of oxidation of the benzylic prime
side position either by substitution of this position, by cyclisation
or by lowering the electron density of the adjacent ring (com-
pounds 25–27, respectively).
A further set of compounds which combined the optimal prime
and non prime side substituents predicted to afford the appropri-
ate activity and improved metabolic stability were prepared but
Overall, finding a molecule with adequate in vitro potency, good
absorption and metabolic stability proved difficult to achieve in
this series. GSK188909 (compound 2) appeared to represent one
of the best compromises in terms of potency and pharmacokinet-
ics, hence its successful use for target validation, albeit at high oral
doses (250 mg/kg po). The discovery of an inhibitor with similar
efficacy at acceptable oral doses for clinical studies required the
discovery of a novel template which will be the subject of the fol-
lowing publications.
Table 4
Activity and selectivity of selected BACE-1 inhibitors bearing bicyclic non prime side substituents
R1
N
N
O
N
N N
N
O
CF3
N
16
17
19
N
N
18a: R = H
18b: R = CH3
O
X
N
N
O
N
N
N
N
O
20
21
23
22
Compd
BACE-1 IC50
(lM)
BACE-2 IC50
(lM)
Cat-D IC50 (lM)
6
0.049
0.019
0.023
0.026
0.028
0.034
0.04
0.087
0.21
0.42
4.485
5.75
8.32
5.75
6.31
8.91
5.13
12.59
13.18
23.44
6.580
53.7
13.49
4.07
14.13
11.22
3.72
15.14
14.39
23.99
16
17
18a
19
20
18b
21
22
23a
a
Using m-MeO benzylic prime side substituent instead of m-CF3.

3668
N. Charrier et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3664–3668
Table 5
in vitro profile of inhibitors with potential improved metabolic stability
CF3
CF3
O
NH
NH
O
S
N
O
24
N
NHR
25
N
O
N
N
NH
NH
27
O
26
Compd
BACE-1 IC50
(lM)
BACE-2 IC50
(lM)
Cat-D IC50 (lM)
24
25
26a
27
0.003
0.006
0.018
0.040
1.41
0.35
2.95
7.76
3.89
0.60
1.07
35.48
a
Obtained as a 1:1 mixture of isomers.
Suomensaari, S. M.; Wang, S.; Walker, D.; Zhao, J.; McConlogue, L.; John, V.
Nature 1999, 402, 537.
O
O
4. Review: (a) Durham, T. B.; Shepherd, T. A. Curr. Opin. Drug Discovery Dev. 2006, 9,
776; (b) Sankaranarayanan, S.; Holahan, M. A.; Colussi, D.; Crouthamel, M.-C.;
Devanarayan, V.; Ellis, J.; Espeseth, A.; Gates, A. T.; Graham, S. L.; Gregro, R.;
Hazuda, D.; Hochman, J. H.; Holloway, K.; jin, L.; Kahana, J.; lai, M.-T.; Lineberger,
J.; McGaughey, G.; Moore, K. P.; Nantermet, P.; Pietrak, B.; Price, E. A.; Rajapakse,
H.; Stauffer, S.; Steinbeiser, M. A.; Seabrook, G.; Selnick, H. G.; Shi, X.-P.; Stanton,
M. G.; Swestock, J.; Tugusheva, K.; Tyler, K. X.; Vacca, J. P.; Wong, J.; Wu, G.; Xu,
M.; Cook, J. J.; Simon, A. J. J. Pharmacol. Exp. Ther. 2009, 328, 131; (c) Charrier, N.;
Clarke, B.; Cutler, L.; Demont, E.; Dingwall, C.; Dunsdon, R.; Hawkins, J.; Howes,
C.; Hussain, I.; Maile, G.; Matico, R.; Mosley, J.; Naylor, A.; O’Brien, A.; Redshaw,
S.; Rowland, P.; Soleil, V.; Smith, K. J.; Sweitzer, S.; Theobald, P.; Vesey, D.;
Walter, D. S.; Wayne, G. J. Med. Chem. 2008, 51, 3313. and references cited
herein; (d) Machauer, R.; Laumen, K.; Veenstra, S.; Rondeau, J.-M.; Tintelnot-
Blomley, M.; Betschart, C.; Jaton, A.-L.; Desrayaud, S.; Staufenbiel, M.; Rabe, S.;
Paganetti, P.; Neumann, U. Bioorg. Med. Chem. Lett. 2009, 19, 1366.
5. Beswick, P.; Charrier, N.; Clarke, B.; Demont, E.; Dingwall, C.; Dunsdon, R.; Faller,
A.; Gleave, R.; Hawkins, J.; Hussain, I.; Johnson, C. N.; MacPherson, D.; Maile, G.;
Milner, P.; Naylor, A.; O’Brien, A.; Redshaw, S.; Riddell, D.; Rowland, P.;
Skidmore, J.; Soleil, V.; Smith, K.; Stanway, S.; Stemp, G.; Stuart, A.; Theobald,
P.; Vesey, D.; Walter, D. S.; Ward, J.; Wayne, G. Bioorg. Med. Chem. Lett. 2008, 18,
1022.
N
O
N
N
O
N
Figure 5. Structure of inhibitor 28.
References and notes
1. Cummings, J. L. N. Eng. J. Med. 2004, 351, 56.
2. (a) Glenner, G. G.; Wong, C. W. Biochem. Biophys. Res. Commun. 1984, 120, 885;
(b) Goedert, M.; Wischik, C. M.; Crowther, R. A.; Walker, J. E.; Klug, A. Proc. Natl.
Acd. Sci. U.S.A. 1988, 85, 4051.
3. (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)
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; (c) Yan,
R.; Bienkowski, M. J.; Shuck, M. E.; Miao, H.; Tory, M. C.; Pauley, A. M.; Brashier, 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; (d) 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.;
6. Hussain, I.; Hawkins, J.; Harisson, 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.
7. For a discussion around the variability of the transposition of enzyme activity in
cellular assays, see Ref. (5).
8. Compound
2 was incubated in liver microsomes and the structures of
metabolites were determined via mass-spectroscopy.
9. The PDB deposition codes and refinement details for the BACE-1 complex crystal
structures are: 17 (2wez, 1.7 Å resolution, R = 0.200, Rfree = 0.223); 22 (2wf0,
1.6 Å resolution, R = 0.188, Rfree = 0.203).