Potent and selective isophthalamide S2 hydroxyethylamine inhibitors of BACE1

Article abstract of DOI:10.1016/j.bmcl.2007.03.096

The design and synthesis of a novel series of potent BACE1 hydroxyethylamine inhibitors. These inhibitors feature hydrogen bonding substituents at the C-5 position of the isophthalamide ring with improved selectivity over cathepsin D.

Full text of DOI:10.1016/j.bmcl.2007.03.096

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3378–3383
Potent and selective isophthalamide S2
hydroxyethylamine inhibitors of BACE1
Steven W. Kortum,a,* Timothy E. Benson,a Michael J. Bienkowski,a Thomas L. Emmons,a
D. Bryan Prince,a, Donna J. Paddock,b Alfredo G. Tomasselli,a Joseph B. Moon,a
Alice LaBordea,à and Ruth E. TenBrinka
aPfizer Global Research and Development, Pfizer Inc., St. Louis Laboratories, 700 Chesterfield Parkway West, Chesterfield,
MO 63017, USA
bPfizer Global Research and Development, Pfizer Inc., 2800 Plymouth Road, Ann Arbor, MI 48105, USA
Received 27 February 2007; revised 23 March 2007; accepted 29 March 2007
Available online 3 April 2007
Abstract—The design and synthesis of a novel series of potent BACE1 hydroxyethylamine inhibitors. These inhibitors feature
hydrogen bonding substituents at the C-5 position of the isophthalamide ring with improved selectivity over cathepsin D.
Ó 2007 Elsevier Ltd. All rights reserved.
Alzheimer’s disease (AD) is a debilitating and ultimately
fatal form of dementia generally affecting people aged 60
and older. The disease progresses from mild cognitive
impairment through profound dementia, loss of motor
functions and finally death.1 In addition to the devastat-
ing impact AD has on individuals and families there is
also a societal impact. The financial cost to society,
brought about by the prolonged nature of AD, is esti-
mated at over $100 billion a year in the US alone.2,3
Inhibition of BACE1 appears to be a viable therapeutic
target for the reduction of Ab(1–40) and Ab(1–42). It
has been shown that BACE1 knockout mice, transgenic
for human APP, do not have amyloid plaque build-up in
the brain.6 This data helps validate BACE1 as a thera-
peutic target for AD.
Hydroxyethylamine (HEA), 1 (Fig. 1) was found to be a
potent inhibitor of BACE1 containing the optimized
transition state insert (TSI) 3,5-difluoro Phe and C-ter-
minal m-ethyl benzyl.8 Although highly potent, this
compound suffered from poor metabolic stability.
One of the major factors contributing to the onset of
AD is the build-up of amyloid plaques in the brain.
The primary component of the amyloid plaque is
aggregated Ab-peptide produced from the cleavage of
Amyloid Precursor Protein (APP) by beta- and
gamma-secretase.4–6 b-Secretase (BACE1) cleaves APP,
releasing the soluble portion of APP (sAPP) and leaving
behind an APP fragment still anchored to the
membrane. Gamma-secretase cleaves the anchored
APP fragment a second time producing two forms of
peptide, Ab(1–40) and Ab(1–42), with Ab(1–42) being
the major component in amyloid plaques.7
Transition
state
isostere
C-Terminus
N-terminus
S2
S1'
OH
H
H
N
N
N
I
Keywords: BACE1; Isophthalamide; HEA; Alzheimer’s disease;
Hydroxyethylamine; Inhibitor; S2 pocket; b-Secretase.
Corresponding author. Tel.: +1 636 247 3592; e-mail: steve.
kortum@pfizer.com
O
O
S2'
1
*
S3
Present address: Department of Chemistry, University of Oklahoma,
Norman, OK 73019, USA.
Retired.
S1
Figure 1. Isophthalamide hydroxyethyl amine.
à
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.03.096

S. W. Kortum et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3378–3383
3379
X-ray crystal structures of 1 bound to BACE1, along
with modeling, suggested a hydrogen bonding substitu-
ent at the C-5 position of the isophthalamide ring (S2
pocket of BACE1) might result in improved binding
with BACE1. Subsequent to our work on the isophthal-
amide HEA series, other groups have described work on
related HEA series of inhibitors.9
in order to increase affinity for BACE1 and selectivity
over Cat D.
The synthetic route used to prepare the C-3 amide and
nitrile isophthalamide rings is outlined in Scheme 1.
Mono methyl 5-nitroisophthalate was converted to
isophthalamide 3a–b using standard CDI amide cou-
pling conditions. Nitro 3a–b was reduced to aniline
4a–b under catalytic hydrogenation conditions with pal-
ladium on carbon. Diazotization of anilines 4a–b and
displacement with either copper (II) bromide or potas-
sium iodide gave halides 5a–d. Under palladium ace-
tate/dppp catalyzed carbonylation conditions halides
5a–d gave amides 6a–d.15 Cyanide displacement of
halides 5a–d using copper cyanide gave nitriles 8a–b.
The esters of amides 6a–d and 8a–b were hydrolyzed
using lithium hydroxide to give acids 7a–f.
Although 1 exhibits high affinity for BACE1 (2 nM) and
reasonable selectivity over cathepsin D (Cat D) (75-
fold), it was rapidly metabolized by rat and human
microsomes. The main pathways for metabolism are
N-debenzylation at the C-terminus and N-depropylation
at the N-terminus (data not shown). Unfortunately, re-
moval of the dipropyl amide results in a significant loss
of affinity for BACE1.11 Due to this loss we needed to
improve BACE1 affinity before addressing the metabolic
stability issues surrounding the HEA template.
The S2 pocket of Cat D is large and contains a lipophilic
valine residue whereas the S2 pocket of BACE1 contains
an arganine residue.12 Substituting the C-5 position of
the isophthalamide ring with hydrogen bonding substit-
uents may create an additional binding interaction with
hydrogen bonding substituents in BACE1 and improve
selectivity over Cat D. Towards these goals, we have
prepared a series of C-5 isophthalamide HEA analogs
The C-3 acetonitrile was prepared using the synthesis
outlined in Scheme 2.
Acid 9 was reduced to alcohol 10 using borane dimethyl
sulfide complex. Selective hydrolysis of one ester of alco-
hol 10 was accomplished using 0.9 equiv of lithium
hydroxide to give acid 11. Acid 11 was coupled with
dipropylamine using EDC/HOBt to give amide 12.
NH2
NO2
NO2
R1
N
R1
b
a
O
HO
O
N
O
O
O
O
O
O
O
2
4a R1 = H
4b R1 = Pr
3a R1 = H
3b R1 = Pr
c
R
R
O
N
O
N
X
R
O
R
R1
N
R1
R1
N
d
e
O
O
N
OH
O
O
O
O
O
5a x=Br, R1 = H
5b x=Br, R1 = Pr
5c x=I, R1 = H
5d x=I, R1 = Pr
7a R = H, R1 = H
6a R = H, R1 = H
6b R = Me, R1 = H
6c R = H, R1 = Pr
7b R = H, R1 = Pr
7c R = Me, R1 = H
7d R = Me, R1 = Pr
6d R = Me, R1 = Pr
f
N
N
R1
R1
N
e
N
O
OH
O
O
O
O
7e R1 = H
7f R1 = Pr
8a R1 = H
8b R1 = Pr
Scheme 1. Synthesis of 3-amide-5-[(dipropylamino)carbonyl]benzoic acids 7a–7f. Reagents and conditions: (a) CDI/CH2Cl2/dipropyl or
propylamine/rt/5 h (75%); (b) 5%Pd on carbon/50PSI H2/18 h (70%); (c) butyl nitrite/CuBr2/CH3CN/rt/3 h (5a,b) (75%); H2SO4/sodium nitrite/KI;
0 °C/1 h (5c,d) (80%); (d) Pd(OAc)2/dppp/CO/DIEA/dimethylamine or hexamethyldisilazane/NMP/rt/18 h (70%); (e) LiOH/methanol/rt/24 h then
HCl (70–80%); (f) CuCN/NMP/160 °C/6 h (70–80%).

3380
S. W. Kortum et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3378–3383
OH
OH
O
OH
a
b
HO
O
O
O
O
O
O
O
O
O
O
O
11
10
9
c
OH
Br
N
O
d
e
N
O
N
O
N
O
O
O
O
O
12
13
O
14
b
N
N
OH
O
O
7g
Scheme 2. Synthesis of 3-(cyanomethyl)-5-[(dipropylamino)carbonyl]benzoic acid 7g. Reagents and conditions: (a) (CH3)2S BH3/THF/rt/18 h/
MeOH (71%); (b) LiOH–H2O/H2O/EtOH/rt/30 min (75%); (c) EDC/HOBt/CH2Cl2/dipropylamine/rt/18 h (50%); (d) PBr3/50 °C/4 h (85%); (e)
NaCN/DMSO/rt/3.5 h (95%).
The alcohol of amide 12 was converted to bromide 13
using PBr3 at 50 °C. Bromide 13 was displaced with
sodium cyanide to give nitrile 14. Hydrolysis of the ester
of nitrile 14 gave acid 7g.
decrease in Cat D affinity. The dimethyl amide 19m
exhibited an 8-fold decrease in Cat D affinity when
compared with the parent compound. Increasing the
alkyl chain length of the C-5 amide to propyl as in 19i
resulted in little change in Cat D affinity. BACE1 affinity
was reduced by a factor of 2.5 as compared with the
parent compound. Acid 19l was 3-fold more potent
and 4-fold more selective than amide 19c while ester
19n lost selectivity 4-fold compared to 19c.
The P1 ꢀ P02 fragments synthesis and final coupling are
outlined in Scheme 3.
Epoxides 15a–c were refluxed with benzyl amines 16a–b
to form P1 ꢀ P02 fragments 17a–c. BOC deprotection of
P1 ꢀ P02 fragments 17a–c using TFA gave amines 18a–
c. HATU coupling of amines 18a–c to acids 7a–g gave
HEA analogs 19a–k and 19m–n. The lower yields asso-
ciated with this coupling reaction was due to unwanted
coupling of the acid with the benzyl amine portion of
18a–c.
Compounds 19d and 19j illustrate the contributions the
difluoro Phe and m-ethyl benzyl amine moieties make
toward BACE1 affinity compared to the amide 19c.
Replacing the P1 Phe with 3,5 difluoro Phe increased
BACE1 affinity 10-fold whereas Cat D affinity was un-
changed compared to 19c. Replacing the P02 m-methoxy
19d with m-ethyl 19j results in a two-fold increase in
BACE1 affinity.
The ester and acid HEA analogs 19l and 19n, respec-
tively were made from acid 9 utilizing standard EDC/
HOBt coupling to form the desired amide and selective
hydrolysis as described above gave the desired isoph-
thalamide. The acid of 19l was formed by saponification
of ester 19n using lithium hydroxide.10
The X-ray crystal structure of 19d in the BACE1 enzyme
shows a stacking interaction between the amide car-
bonyl group and Arg 235 as shown by Figure 2.13 The
distance between the amide carbonyl group and Arg
˚
235 is between 2.8 and 3.1 A. In Cat D the amide car-
BACE1 and Cat D IC50 values are contained in Table
1.14 Compounds 19a and 19b are included for
comparison.8
bonyl likely interacts with a more lipophilic Val 238,
resulting in a less desirable interaction which is reflected
in the affinities for Cat D. This same interaction may be
responsible for the results obtained for 19e.
Direct comparison of the parent compound 19b with the
corresponding carboxamide substituted analog 19c
showed comparable values for BACE1 but a 6-fold
In an effort to increase the interaction between the
amide carbonyl and Arg 235, a methylene spacer was

S. W. Kortum et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3378–3383
3381
O
O
H
O
H
OH
H
O
N
O
N
N
R3
R4
R5
R4
R5
+
a
NH2
R3
R4
R4
17a R3 = OMe, R4 = H, R5 = OBn
17b R3 = Et, R4 = F, R5 = H
17c R3 = Et, R4, R5 = H
15a R4 = H, R5 = OBn
15b R4 = F, R5 = H
15c R4, R5 = H
16a R3 = OMe
16b R3 = Et
b
OH
R2
H
H2N
N
R3
R1
R4
R5
N
OH
+
O
O
R4
7a-g
18a R3 = OMe, R4 = H, R5 = OBn
18b R3 = Et, R4 = F, R5 = H
18c R3 = Et, R4,5 = H
c
R2
R1
N
OH
H
H
N
N
R3
O
O
R4
R5
19a-n
R4
Scheme 3. Synthesis of (2R,3S)-3-amino-1-(benzylamino)-4-phenylbutan-2-ols 18a–c and final coupling. Reagents and conditions: (a) isopropanol/
reflux/5 h (70–80%); (b) TFA/CH2Cl2/rt/1 h/sodium bicarbonate (90%); (c) HATU/TEA/CH2Cl2;/rt/18 h (30–50%).
Table 1. Table of BACE1 and Cat D IC50 values
R2
R1
OH
H
N
H
N
N
R5
R3
R4
O
O
R3
Compound
R1
R2
R3
R4
R5
BACE IC50 (nM)
Cat D IC50 (nM)
19a
19b
19c
19d
19e
19f
19g
19h
19i
Pr
Pr
Pr
Pr
Pr
Pr
H
Me
Me
F
H
H
H
H
H
H
H
H
H
H
OH
H
H
H
Et
2
100
100
11
150
450
H
H
F
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
Et
CONH2
CONH2
CN
2800
2400
660
H
H
H
H
H
F
90
CH2CN
CN
360
770
1900
250
5
570
>20,000
>20,000
3600
1600
>20,000
3600
3800
866
H
CONH2
CON(Pr)2
CONH2
CONH2
COOH
CON(Me)2
COOEt
Pr
Pr
Pr
Pr
Pr
Pr
19j
19k
19l
H
H
H
H
OMe
OMe
OMe
OMe
300
30
19m
19n
70
132

3382
S. W. Kortum et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3378–3383
Acknowledgments
We would like to thank Benjamin Amore and Richard
Steenwick for their work in metabolic stability. We
would also like to acknowledge the use of the IMCA-
CAT beamline 17-ID at the Advanced Photon Source
which is supported by the companies of the Industrial
Macromolecular Crystallography Association through
a contract with the Center for Advanced Radiation
Sources at the University of Chicago. Use of the
Advanced Photon Source was supported by the US
Department of Energy, Office of Science, Office of Basic
Energy Sciences, under Contract No. W-31-109-Eng-38.
References and notes
˚
Figure 2. A 3D view of the 2.5 A structure of 19d bound to BACE
showing the stacking of the amide to Arg235 using van der Waals
surfaces (pink and blue). The carbonyl oxygen of Phe108 which is
˚
3.5 A away from the para position of the P1 phenyl is also shown.
Waters have been omitted for clarity.
1. (a) Celsis, P. Ann. Med. 2000, 32, 6; (b) Pendlbury, B.;
Solomon, P. Clin. Symp. 1999, 48, 1.
2. Stavrovskaya, Irina G.; Kristal, Bruce S. Free Radic. Biol.
Med. 2005, 38, 687.
3. Milligan, C. Nat. Med. 2000, 6, 385.
4. Thorsett, E.; Latimer, L. Curr. Opin. Chem. Biol. 2000, 4,
377.
placed between the isophthalamide ring and the nitrile
substituent to give 19f. Unfortunately this brought
about a decrease in BACE1 affinity with a slight increase
in the Cat D affinity. Attempts to convert 19f to the pri-
mary carboxamide were unsuccessful.
5. Yan, R.; Bienkowski, M. J.; Shuck, M. E.; Miao, H.;
Tory, M. C.; Pauley, A. M.; Brashier, J. R.; Stratman, N.
C.; Mathews, R. W.; Buhl, A. E.; Carter, D. B.;
Tomasselli, A. G.; Parodi, L. A.; Heinrikson, R. L.;
Gurney, M. Nature 1999, 402, 533.
6. Lin, X.; Koelsch, G.; Wu, S.; Downs, D.; Dashti, A.;
Tang, J. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,
145601460.
7. (a) Iwatsubo, T.; Odaka, A.; Suzuki, N.; Mizusawa, H.;
Nukina, N.; Ihara, Y. Neuron 1994, 13, 45; (b) Lippa, C.;
Nee, L.; Mori, H.; St George-Hyslop, P. Lancet 1998, 352,
1117.
Since a major metabolite of compounds such as 19a is
depropylation of the dipropyl amide on the N-terminus,
compounds such as 19g and 19h were prepared to elim-
inate this mode of metabolism and improve overall met-
abolic stability. Unfortunately the added interaction at
the C-5 position of the isophthalamide ring did not com-
pensate for the loss in affinity resulting from the removal
of one of the propyl groups in the dipropyl amide.
BACE1 affinity decreased ꢁ9 to 19-fold for the mono
propyl amide analogs.
8. (a) Maillaird, M.; Hom, R.; Gailunas, A.; Jagodzinska, B.;
Fang, L. Y.; Varhgese, J.; Freskos, J. N.; Pulley, S. R.;
Beck, J. P.; TenBrink, R. E. WO Patent 2002002512, 2002;
(b) Hom, R. K.; Maillard, M.; Mamo, S.; Dappen, M. A.;
Brogley, L.; Varghese, J.; Thorsett, E. D.; Abstracts of
Papers 232nd National Meeting of the American Chemical
Society, San Fransisco, CA, Sept. 10–14, 2006; (c)
Maillard, M. C.; Hom, R. K.; Benson, T. E.; Moon, J.
B.; Mamo, S.; Bienkowski, M. J.; Tomasselli, A. G.;
Woods, D. D.; Prince, D. B.; Paddock, D. J.; Emmons, T.
L.; Tucker, J.; Dappen, M. A.; Brogley, L.; Thorsett, E.
D.; Jewett, N.; Sinha, S.; John, V. J. Med. Chem. 2007, 50,
776.
9. (a) Stachel, S. J.; Coburn, C. A.; Steele, T. G.; Jones,
K. G.; Loutzenhiser, E. F.; Gregro, A. R.; Rajapaske,
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. 2005, 5, 1609; (b) Stachel, S. J.;
Coburn, C. A.; Steele, T. G.; Crouthamel, M.-C.;
Pietrak, B. L.; Lai, M.-T.; Holloway, M. K.; Munshi,
S.; Sanjeev, K.; Graham, S. L.; Vacca, J. P. Bioorg.
Med. Chem. Lett. 2006, 16, 641.
Due to the loss in affinity for BACE1 in compounds
such as 19g and 19h we needed to further improve upon
the affinity for BACE1. Based on the X-ray crystal
structure of 19d in the BACE1 enzyme, Phe-108, which
resides in the bottom of the S3 pocket of BACE1, may
offer an additional hydrogen bonding opportunity. In
an attempt to gain hydrogen bonding between the
BACE1 inhibitor and the carbonyl oxygen of Phe-108,
we replaced the difluoro phenyl portion of 19d with a
para-hydroxy group; however 19k showed an approxi-
mately 30-fold decrease in BACE1 affinity. As a result,
further analogs with the tyrosine HEA insert were not
pursued.
A carboxamide HEA with optimized P1 and P02 groups
19j resulted in a 5 nM BACE1 inhibitor with 320-fold
selectivity over Cat D compared with 75-fold selectivity
for 19a. The added interaction with BACE1 at S2 was
not enough to overcome the affinity loss associated with
the removal of one of the propyl substituents of the
N-terminal dipropyl amide. Several other HEAs with
various substituents in the C-5 position of the isophthal-
amide ring were prepared but did not improve upon 19d.
10. Varhgese, J.; Jagodzinska, B.; Maillaird, M.; Beck, J. P.;
TenBrink, R. E., Getman, D. WO Patent 072535A2, 2003.
11. Freskos, J. N.; Fobian, Yvette M.; Benson, Timothy E.;
Bienkowski, Micheal J.; Brown, David L.; Emmons,
Thomas L.; Heintz, Robert; Laborde, Alice; McDonald,
Joseph J.; Mischke, Brent V.; Molyneaux, John M.; Moon,
Joseph B.; Mullins, Patrick B.; Prince, D. Bryan; Paddock,

S. W. Kortum et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3378–3383
3383
Donna J.; Tomasselli, Alfredo G.; Winterrowd, Gregory
Bioorg. Med. Chem. Lett. 2007, 17, 73.
12. Silva, F. P.; Ribeiro, F.; Katz, N.; Giovanni-De-Simone,
S. FEBS Lett. 2002, 514, 141.
13. Coordinates for the complex of BACE with inhibitor 36d
have been deposited in the Protein Data Bank
(www.rcsb.org) under PDB ID 2P83.
14. The BACE1 and cathepsin D assays were used to measure
enzymatic activities in a fluorescence polarization format
and in an endpoint mode. Both assays were performed in
non-binding 384-well plates and in a volume of 30 lL/well.
Enzyme, substrate, and stop solutions were added to the
assay plates with a LabSystems Multidrop 384. The
reactions in the BACE1 assay were run at 37 °C for 3 h
under the following conditions: 100 lM sodium acetate
(pH 4.5), 150 nM substrate 1, 0.1 nM BACE1, 2% DMSO,
and 0.001% (v/v) Tween-20. The cathepsin D assay
reactions were run at 37 °C for 90 min under the following
conditions: 100 lM sodium acetate (pH 4.5), 150 nM
substrate 2, 0.5 nM cathepsin D, 2% DMSO, and 0.001%
(v/v) Tween-20. Reactions were terminated by the addition
of stop solution (30 lL/well). Following a 15-min incuba-
tion at room temperature, sample fluorescence polariza-
(Sunnyvale, CA) using an excitation 485 nm filter, a
530 nm emission, and G factor settings of 0.872 and 0.864
for the BACE1 and the cathepsin D assays, respectively.
15. Carbonylation procedure for compound 6d: To a mixture
of methyl 3-[(dipropylamino)carbonyl]-5-iodobenzoate
(0.42 g, 1.08 mmol) in dry NMP (6.0 mL) bubbled with
carbon monoxide for 10 min was added diisopropyl
ethylamine (0.14 g, 1.08 mmol), palladium (II) acetate
(0.025 g, 0.108 mmol), 2 M dimethylamine in THF
(1.0 mL, 2.0 mmol) and diphenylphosphino propane
(0.067 g, 0.162 mmol). The mixture was stirred at room
temperature under a carbon monoxide atmosphere over-
night. The mixture was partitioned between water and
ethyl acetate. The layers were separated and the organic
layer washed twice with water, dried over anhydrous
magnesium sulfate and concentrated under reduced
pressure. The residue was purified via flash chromatog-
raphy on silica gel (100 mL) using 3% CH3OH in CH2Cl2
to give 0.249 g (69%) of the title compound: 1H NMR
(CDCl3, 400 MHz) d 0.745–0.765 (m, 3 H), 0.972–0.998
(m, 3H), 1.51–1.59 (m, 2H), 1.65–1.73 (m, 2H), 2.99 (s,
3H), 3.09–3.17 (m, 5H), 3.41–3.53 (m, 2H), 3.94 (s, 3H),
7.62 (d, J = 1.2 Hz, 1H), 8.07 (d, J = 1.2 Hz, 1H), 8.12 (d,
J = 1.2 Hz, 1H).
tion was measured on
a LJL Biosystems Acquest