Discovery and initial optimization of 5,5′-disubstituted aminohydantoins as potent β-secretase (BACE1) inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.11.052

8,8-Diphenyl-2,3,4,8-tetrahydroimidazo[1,5-a]pyrimidin-6-amine (1) was identified through HTS, as a weak (micromolar) inhibitor of BACE1. X-Ray crystallographic studies indicate the 2-aminoimidazole ring forms key H-bonding interactions with Asp32 and Asp228 in the catalytic site of BACE1. Lead optimization using structure-based focused libraries led to the identification of low nanomolar BACE1 inhibitors such as 20b with substituents which extend from the S₁ to the S₃ pocket.

Full text of DOI:10.1016/j.bmcl.2009.11.052

Bioorganic & Medicinal Chemistry Letters 20 (2010) 632–635
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and initial optimization of 5,50-disubstituted aminohydantoins
as potent b-secretase (BACE1) inhibitors
a
a
b
a,
Pawel Nowak a, , Derek C. Cole , Ann Aulabaugh , Jonathan Bard , Rajiv Chopra , Rebecca Cowling a,
Kristi Y. Fan c, Baihua Hu c, Steve Jacobsen b, Minakshi Jani a, Guixan Jin a, Mei-Chu Lo a, Michael S. Malamas c,
Eric S. Manas a,à, Rani Narasimhan a, Peter Reinhart b, Albert J. Robichaud c, Joseph R. Stock a, Joan Subrath a,
Kristine Svenson d, Jim Turner b, Erik Wagner b, Ping Zhou c, John W. Ellingboe a
*
a Chemical Sciences, Wyeth Research, 401 N. Middletown Road, Pearl River, NY 10965, United States
b Discovery Neuroscience, Wyeth Research, Princeton, NJ 08852, United States
c Chemical Sciences, Wyeth Research, 865 Ridge Road, Monmouth Junction, NJ 08852, United States
d Chemical Sciences, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02139, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
8,8-Diphenyl-2,3,4,8-tetrahydroimidazo[1,5-a]pyrimidin-6-amine (1) was identified through HTS, as a
weak (micromolar) inhibitor of BACE1. X-Ray crystallographic studies indicate the 2-aminoimidazole ring
forms key H-bonding interactions with Asp32 and Asp228 in the catalytic site of BACE1. Lead optimiza-
tion using structure-based focused libraries led to the identification of low nanomolar BACE1 inhibitors
such as 20b with substituents which extend from the S1 to the S3 pocket.
Received 15 October 2009
Revised 10 November 2009
Accepted 16 November 2009
Available online 20 November 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
BACE1
Aminohydantoin
Alzheimer’s disease
Alzheimer’s disease (AD) is a progressive neurodegenerative dis-
ease that is the leading cause of long-term decline in cognitive func-
tion in the elderly. It is estimated that more than 27 million people
worldwide have AD and the number of cases is likely to increase in
the future.1 Current AD therapy is based on symptomatic treatment
with Aricept™ (Pfizer), Exelon™ (Novartis) and Reminyl™ (Janssen).
These acetylcholine esterase inhibitors are effective in patients only
for a limited period of time and do not treat the underlying cause of
the disease. Thus, there is an extremelyurgent need for development
of disease-modifying treatment for AD (Fig. 1).
8,8-Diphenyl-2,3,4,8-tetrahydroimidazo[1,5-a]pyrimidin-6-
amine (1)6 was identified as weak (IC50 = 38
M) inhibitor of
l
BACE1 using a fluorescence energy transfer (FRET)-based HTS
screen. An X-ray structure of 1 bound in the BACE1 active site6
indicates that the guanidine moiety forms key H-bonding interac-
tions with Asp32 and Asp228 in the catalytic site of BACE1. The
0
two phenyl groups occupy the S1 and S2 pockets and are conve-
Asp228
O
Overwhelming evidence suggests that the formation of b-amy-
loid plaques is the fundamental cause of AD.2,3 These plaques are
formed by aggregation of amyloid fibrils, which in turn are formed
from the amyloid b-peptide (b-amyloid, Ab), which is generated by
sequential proteolytic cleavage of amyloid precursor protein (APP)
Sovent
O
H
H
N
O
O
N
Asp32
by b-secretase (BACE1) and c
-secretase.4 Thus agents that inhibit
N
N
H
BACE1 may decrease levels of Ab, and have therapeutic benefit in
AD.5
S2'
S1
* Corresponding author. Tel.: +1 845 602 4303; fax: +1 845 602 5561.
E-mail address: nowakp@wyeth.com (P. Nowak).
1
Present address: Novartis Institute for Biomedical Research, 250 Massachusetts
Ave, Cambridge, MA 02139, United States.
Present address: Department of Computational and Structural Chemistry,
Molecular Discovery Research, GlaxoSmithKline, 1250 South Collegeville Road,
Collegeville, 19426 PA, United States.
Figure 1. The structure of 8,8-diphenyl-2,3,4,8-tetrahydroimidazo[1,5-a]pyrimi-
din-6-amine (1) showing hydrogen bonding interactions between the guanidine
moiety and Asp32 and Asp228 of BACE1, and the two phenyl groups projecting into
à
0
S1 and S2 pockets (Model developed based on the X-ray structure PDB ID code:
3IGB6).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.11.052

P. Nowak et al. / Bioorg. Med. Chem. Lett. 20 (2010) 632–635
633
nient handles for SAR optimization, while the six-membered ring
O
OH
OH
of 1 extends toward solvent and does not interact with the enzyme.
On the basis of the X-ray structure it was hypothesized that re-
moval of tetrahydropyrimidine ring of 1, while lowering the
molecular weight, should have little effect on the binding affinity
since the resulting compound 2-amino-3-methyl-5,5-diphenylhy-
dantoin (2a) or its partially or fully saturated analogs (2b, 2c)
should retain all the key interactions with protein.
OH
OMe
OMe
b,c
a
O
O
O
10
9
8
O
O
R
Cl
d
e
O
O
12
11
H2N
H2N
H2N
H2N
N
N
N
N
Scheme 2. Reagents and conditions: (a) H2 (50 psi), PtO2, AcOH, 16 h (99%);
(b) (COCl)2, DMSO, ꢀ78 °C, 1 h then 9, 1 h, then Et3N, 1 h (85%); (c) NaOH, THF–
MeOH–H2O, 5 h (75%); (d) (COCl)2, DMF (cat), DCM, 3 h (100%); (e) RMgBr, CuBr
(1 equiv), LiBr (2 equiv), THF, ꢀ78 °C, 0.5 h, then 11, 0.5 h.
N
N
N
N
N
O
O
O
1
2c
2b
2a
NHBoc
NHBoc
The aminohydantoins are prepared by reaction of diketones
with substituted aminoguanidines via a benzil like rearrangement
(Scheme 1).
Under mild base conditions, such as sodium carbonate, the reac-
tion produces almost exclusively the endo-substituted kinetic prod-
uct 6, while strong bases, for example, sodium hydroxide, lead
preferentially to exo-substituted, thermodynamically favored prod-
uct 7.7,8 Compounds of structure 7 were found to be much less active
in the BACE1 FRET assay.
a
+
Br
15
14
13
O
O
c, d
b
O
O
The required substituted aminoguanidines were produced by
the reaction of primary amines with 1H-pyrazole-1-carboxamidine
hydrochloride.9
The synthesis of aminohydantoin libraries required access to
non-commercial, unsymmetrical, cyclohexyl substituted diketones
for example, 12 and 17, so two routes outlined in Schemes 2 and 3
were devised.10
NHCOR
NHBoc
17
16
Scheme 3. Reagents and conditions: (a) Pd(PPh3)4, CuI (cat), DIPEA, toluene, 60 °C,
16 h, 99%; (b) KMnO4, acetone, water, 3 h, 65%; (c) DCM–TFA, rt, 1 h, 100%; (d)
RCOOH, EDC, DMAP, DCM, rt, 16 h.
Exhaustive hydrogenation of methyl mandelate 8 affords the
hydroxyester 9, which is oxidized to the corresponding keto ester
under Swern oxidation conditions and hydrolyzed to the 2-cyclo-
hexyl-2-oxoacetic acid 10. The acid is activated as an acid chloride
and reacted with a cuprate reagent generated in situ from the
appropriate Grignard reagent leading to diketones 12.
Alternatively, aryl-amido functionalized diketones for example,
16 used for synthesis of aminohydantoin array 20a–h are prepared
by a Sonogashira coupling of Boc-protected bromoaniline 14 with
cyclohexylacetylene 13 (Scheme 3). The disubstituted acetylene
15 is oxidized to form the diketone 16. Removal of the protecting
group and amide formation affords 17.
Analogs 2a–c were tested in the BACE1 and BACE2 FRET assays
to determine potency and selectivity. Cathepsin D, which plays an
important role in cellular protein catabolism,11 was used as an
unrelated protease (39% similarity at active site with BACE1) to
predict pan-aspartyl protease inhibition.
BACE2 is the closest related human aspartyl protease to BACE1
(79% identity in the active site).12 Despite high similarity of BACE2
to BACE1, and the ability of BACE2 to cleave APP within the Ab do-
main, its biological function, although not well understood, ap-
pears to be independent. BACE2 is likely not responsible for
pathogenesis of Alzheimer’s disease and it was even postulated
that potentiation of this protease in the elderly may protect against
AD pathogenesis.13 Comparative analysis of crystal structures of
these two proteases indicates residue differences around S3, S2
and S2 pockets which may allow design of potent inhibitors.14
0
To our surprise, removal of tetrahydropyrimidine ring of 1 had a
very pronounced effect, with 10-fold improvement of BACE1 en-
zyme and cellular activity. This coupled with the reduction in
MW and clogP, results in aminohydantoin 2a, with clogP, ligand
efficiency (LE),15 LELP16 and LipE17 all significantly improved, and
in the range of an attractive lead.16 The improvement in calculated
properties also correlates with improved pharmaceutical proper-
ties, including permeability (Parallel Artificial Membrane Perme-
ability Assay (PAMPA) and BBB-PAMPA),18,19 aqueous solubility,
brain and plasma exposure and brain/plasma (B/P) ratio (Table 1).
Saturation of one of the phenyl substituents (compound 2b)
gave further improvement of the BACE1 potency while the BACE2
activity was almost unaffected. Full saturation (compound 2c) did
not improve potency; however it had a detrimental effect on brain
and plasma concentrations as well as the B/P ratio. We could not
find a simple relationship between the B/P ratio and clogP or total
polar surface area (TPSA). In fact the B/P ratio was counter intui-
tively highest when clogP was lowest. In all cases (1–2c) the
Cathepsin D activity was well separated from BACE activity.
H2N
R3
O
N
R2
R1
N
R1
O
R2
O
4
NH
NH2
6
R3 NH2
R3
+
N
a
b
H
3
R3HN
5
NH
N
R1
O
R2
Compound 2b with its low lM activity in enzyme and cellular
7
assay, good calculated and pharmaceutical properties and brain
exposure, was selected for further development. We envisaged that
the potency of our series could be improved by introducing substit-
Scheme 1. Reagents and conditions: (a) 1H-pyrazole-1-carboxamidine.HCl, DIPEA,
DMF, 40 °C, 16 h; (b) Na2CO3, EtOH/H2O (5:1), reflux, 3 h.

634
P. Nowak et al. / Bioorg. Med. Chem. Lett. 20 (2010) 632–635
Table 1
Table 2
BACE1, BACE2, Cathepsin D activities, permeability, brain and plasma concentrations
BACE1 and BACE2 activity of compounds 20a–l
data and calculated properties for compounds 1–2c
H2N
N
Compound
BACE1 IC50
Cellular Ab ELISA EC50
BACE2 IC50 M)
1
2a
2b
2c
N
O
(
l
M)
38
16
38
3.40
4.17
0.77
(34)
3.0
11.1
258
384
0.7
0.99
0.81
0.51
1.08
1.03
0.51
H
(lM)
N
R
(l
Cathepsin D IC50 (or% Inhibition @ 100
l
M) (2)
24 (38) (33)
O
PAMPA (Pe * 10ꢀ6 cm/s)
1.0
8.6
21
115
0.2
55
290
3.0
7.2
12.8
99
208
0.5
2.9
7.7
17
143
0.1
>100
277
3.6
56
0.41
8.8
2.4
20
BBB-PAMPA (Pe * 10ꢀ6 cm/s)
Brain concentration Cmax (ng/g)a
Plasma concentration Cmax(ng/g)a
B/P ratio
Compound
R
BACE1 IC50
0.10
(
lM)
BACE2 IC50
0.02
(lM)
O
O
20a
Solubility @ pH 7.4 (
l
g/mL)
>100 >100
MW
clogP
TPSA
LEb
265
1.8
56
271
2.8
56
0.41
6.8
3.2
20b
20c
0.04
0.86
0.03
0.18
53
0.22 0.38
16.6 4.7
1.4
LELPc
LipEd
N
3.7
a
b
c
Dose 3 mg/kg oral.
20d
20e
2.85
0.21
0.49
0.04
LE = ꢀRT(Ln(IC50)/HA), HA—non-H atoms.
LELP = clogP/LE.
O
d
LipE = ꢀlog(IC50) – clogP.
O
20f
7.86
1.36
1.85
0.49
uents extending from the S1 to the S3 pocket. This strategy was al-
ready shown to be successful on the acylguanidine series of BACE1
inhibitors, exemplified by comparing 18 and 19.20
20g
O
NH2
O
20h
0.38
1.15
HN NH
O
O
R
N
Introduction of the N-methyl-2-pyrrole group (20c), however
was not beneficial. The small n-butyric amide (20d) led to loss of
potency. Introduction of an oxygen atom into the alkyl chain of
20d lead to fivefold potency improvement (20e), while extending
the chain by one methylene (20f) resulted in significant loss of
activity. With the exception of compound 20h, in the amide series
the BACE2 potency was equal or higher than BACE1. Subsequent
modeling studies indicated that increased potency of compounds
20a, 20b and 20e may be due to additional hydrogen bonding
interactions between the oxygen atom in the furan or ethoxy group
and Ser229 through conserved water observed in the X-ray struc-
ture of similar lead in this series.21
Compound 20b is a 40 nM BACE1 enzyme inhibitor with an EC50
of 180 nM in the Ab cellular ELISA assay and hence it was more
fully characterized (Table 3). While the BACE1 enzymatic activity
was improved 25-fold relative to lead 2b, the improvement of cel-
lular activity was only fivefold, perhaps reflecting the lower per-
18
19
R = H
IC50 = 3.7 µM
R = m-Br-Ph-CONH IC50 = 0.6 µM
Overlaying compound 2b with 18 in the BACE1 binding site
(Fig. 2), indicated that the meta-position of the phenyl ring of 2b
is ‘equivalent’ to para-position of phenyl ring of aminoguanidine
18 and it should allow for introduction of substituents extending
into the deeper S3 pocket.
The possibility of reaching further into the S1–S3 pocket was ex-
plored with a set of amides 20 (Table 2). 2-Furoic and 3-methyl-2-
furoic amides (20a and b) gave compounds with 10- and 25-fold
improved potency over the original aminohydantoin 2a.
Table 3
BACE1, BACE2, Cathepsin D activities, permeability, brain and plasma concentrations
data and calculated properties for compounds 20b
Property
Value
BACE1 IC50
(
l
M)
0.04
0.18
0.03
17.1
1.4
Cellular Ab ELISA EC50
BACE2 IC50 M)
Cathepsin D ( M)
PAMPA (Pe * 10ꢀ6 cm/s)
BBB-PAMPA (Pe * 10ꢀ6 cm/s)
(lM)
(l
l
13.7
41
Solubility @ pH 7.4 (lg/mL)
MW
clogP
TPSA
LE
LELP
LipE
395
2.7
100
0.35
7.7
Figure 2. The overlay of compound 2b (magenta) with 18 (cyan) (PDB ID code:
2QU2) in the BACE1 binding site. The structure alignment is based on protein
sequences of both X-ray structures. For clarity, only BACE1 site in complex with
compound 2b is shown in white ribbon.
4.7

P. Nowak et al. / Bioorg. Med. Chem. Lett. 20 (2010) 632–635
635
Table 4
ment over the initial hit (compound 1 vs 20b). The newly designed
compounds were selective over Cathepsin D. Future studies will fo-
cus on continued optimization of the series, optimizing BACE2
selectivity, and improving the physical and pharmaceutical
properties.
BACE1 activity of compounds 21a–f
R
H2N
N
N
O
References and notes
21
1. Brookmeyera, R.; Johnsona, E.; Ziegler-Grahamb, K.; Arrighic, M. H. Alzheimer’s
& Dementia 2007, 3, 186.
2. Blass, J. P.; Poirier, J. In Clinical Diagnosis and Management of Alzheimer’s Disease;
Gauthier, S., Ed.; Butterworth-Heinemann: Woburn, Mass, 1996; pp 17–31.
3. Hardy, J.; Selkoe, D. J. Science 2002, 297, 353.
Compound
R
BACE1 IC50
1.39
(l
M)
BACE2 IC50
0.27
(lM)
Me
4. Seloke, D. J. Physiol. Rev. 2001, 81, 741.
21a
(CH2)5
5. For an excellent review see: Stachel, S. J. Drug Dev. Res. 2009, 70, 101.
6. Malamas, M. S.; Erdei, J. J.; Gunawan, I. S.; Nowak, P.; Harrison, B. L. W.O. Patent
2006076284, 2006; Chem. Abstr. 2006, 145, 167278; Malamas, M. S.; Erdei, J.;
Gunawan, I.; Barnes, K.; Johnson, M.; Hui, Y.; Turner, J.; Hu, Y.; Wagner, E.; Fan,
K.; Olland, A.; Bard, J.; Robichaud, A. J. J. Med. Chem. 2009, 52, 6314.
7. Call, L. Monatsh. Chem. 1970, 101, 344.
COOH
(CH2)5
21b
0.95
0.31
COOH
8. Nishimura, T.; Kitajima, K. J. Org. Chem. 1979, 44, 818.
21c
21d
0.39
1.24
0.12
0.57
9. Bernatowicz, M. S.; Wu, Y.; Matsueda, G. R. J. Org. Chem. 1992, 57, 2497.
10. Other convenient ways to prepare unsymmetrical diketones were also
developed. For example see: Nowak, P. Synth. Commun., in press.
11. Diment, S.; Leech, M. S.; Stahl, P. D. J. Biol. Chem. 1998, 263, 6901.
12. Farzan, M.; Schnitzler, C. E.; Vasillieva, N.; Leung, D.; Choe, H. Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 9712.
F
21e
1.08
1.04
0.41
13. Sun, X.; He, G.; Song, W. FASEB J. 2006, 20, 1369.
14. Ostermann, N.; Eder, J.; Eidhoff, U.; Zink, F.; Hassiepen, U.; Worpenberg, S.;
Maibaum, J.; Simic, O.; Hommel, U.; Gerhartz, B. J. Mol. Biol. 2006, 335, 249.
15. Hopkins, A. L.; Groom, C. R.; Alex, A. Drug Discovery Today 2009, 9, 430.
16. Keseru, G. M.; Makara, G. M. Nat. Rev. Drug Disc. 2009, 8, 203.
17. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 86, 881.
18. Kerns, E. H.; Di, L. Drug Discovery Today 2003, 8, 316.
F
O
21f
0.17
1.65
OH
21g
56% @ 5 lM
19. Di, L.; Kerns, E. H.; Fan, K.; McConnell, O. J.; Carter, G. T. Eur. J. Med. Chem. 2003,
38, 223.
(CH2)3
20. Cole, D. C.; Stock, J. R.; Chopra, R.; Cowling, R.; Ellingboe, J. W.; Fan, K. Y.;
Harrison, B. L.; Hu, Y.; Jacobsen, S.; Jennings, L. D.; Jin, G.; Lohse, P. A.; Malamas,
M. S.; Manas, E. S.; Moore, W. J.; O’Donnell, M.-M.; Olland, A. M.; Robichaud, A.
J.; Svenson, K.; Wu, J.; Wagner, E.; Bard, J. Bioorg. Med. Chem. Lett. 2008, 18,
1063.
21. Malamas, M. S.; Barnes, K.; Johnson, M.; Hui, Y.; Zhou, P.; Turner, J.; Hu, Y.; Erik
Wagner, E.; Fan, K.; Olland, A.; Bard, J.; Jacobsen, S.; Ronald L. Magolda, R.;
Pangalos, M.; Robichaud, A. J. J. Med. Chem., in press.
22. 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.; Zhou, P.; Bard, J. J.
Med. Chem. 2006, 49, 6158.
23. Jennings, L. D.; Cole, D. C.; Stock, J. R.; Sukhdeo, M. N.; Ellingboe, J. W.; Cowling,
R.; Jin, G.; Manas, E. S.; Fan, K. Y.; Malamas, M. S.; Harrison, B. L.; Jacobsen,
Steve; Chopra, R.; Lohse, P. A.; Moore, W. J.; O’Donnell, M.-M.; Hu, Y.;
Robichaud, A. J.; Turner, M. J.; Wagner, E.; Bard, J. Bioorg. Med. Chem. Lett. 2008,
18, 767.
24. Fobare, W. F.; Solvibile, W. R.; Robichaud, A. J.; Malamas, M. S.; Manas, E.;
Turner, J.; Hu, Y.; Wagner, E.; Chopra, R.; Cowling, R.; Jin, G.; Bard, J. Bioorg.
Med. Chem. Lett. 2007, 17, 5353.
25. Stachel, S. J.; Coburn, C. A.; Steele, T. G.; Jones, K. G.; Loutzenhiser, H. A.; Gregro,
A. R.; Rajapakse, H. A.; Lai, M.-T.; Crouthamel, M.-C.; Xu, M.; Tugusheva, K.;
Lineberger, J. E.; Pietrak, B. L.; Espeseth, A. S.; Shi, X.-P.; Chen-Dodson, E.;
Holloway, M. K.; Munshi, S.; Simon, A. J.; Kuo, L.; Vacca, J. P. J. Med. Chem. 2004,
47, 6447.
meability of 20b versus 2b (1.4 vs 7.2 Pe * 10ꢀ6 cm/s). The increase
in potency resulted from additional functionality, so the MW and
TPSA were increased, but the clogP, LE, LELP and LipE remained
unchanged.
The N-methyl group in aminohydantoins of type 2 is oriented
0
towards the hydrophilic S1 pocket of BACE1 and suggests that
extending polar substituents into this pocket may result in addi-
tional binding interactions, with for example Arg235. Improved po-
0
tency by additional interactions in the S1 region has been achieved
in our laboratories on the acylguanidine series22–24 as well by
others.25,26
A set of compound 21a–f bearing polar and hydrophobic groups
was selected based on docking experiments. In general, the addi-
tional N-substituent had minimal impact on BACE1 potency, with
the exception of 21c in which small improvement of potency
was realized. The hydroxypropyl group (21g), previously shown
to be beneficial for the potency of acylguanidines,22,24 gave surpris-
ingly poor activity. Compounds of type 21 were, however uni-
formly more active in BACE2 than in BACE1 (Table 4).
26. Garino, C.; Tomita, T.; Pietranscosta, N.; Laras, Y.; Rosas, R.; Herbette, G.;
Maigret, B.; Quelever, G.; Iwatsubo, T.; Kraus, J.-L. J. Med. Chem. 2006, 49, 4275.
In general this hit-to-lead optimization guided by crystallo-
graphic and modeling studies led to a 1000-fold potency improve-