Rapid P1 SAR of brain penetrant tertiary carbinamine derived BACE inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.01.005

This Letter describes the one pot synthesis of tertiary carbinamine 3 and related analogs of brain penetrant BACE-1 inhibitors via the alkylation of the Schiff base intermediate 2. The methodology developed for this study provided a convenient and rapid means to explore the P1 region of these types of inhibitors, where the P1 group is installed in the final step using a one-pot two-step protocol. Further SAR studies led to the identification of 10 which is twofold more potent in vitro as compared to the lead compound. This inhibitor was characterized in a cisterna magna ported rhesus monkey model, where significant lowering of CSF Aβ40 was observed.

Full text of DOI:10.1016/j.bmcl.2010.01.005

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1779–1782
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Rapid P1 SAR of brain penetrant tertiary carbinamine derived BACE inhibitors
Hong Zhu a, Mary B. Young a, Philippe G. Nantermet a, Samuel L. Graham a, Dennis Colussi b, Ming-Tain Lai b,
Beth Pietrak b, Eric A. Price b, Sethu Sankaranarayanan b, Xiao-ping Shi b, Katherine Tugusheva b,
Marie A. Holahan c, Maria S. Michener c, Jacquelynn J. Cook c, Adam Simon b, Daria J. Hazuda b,
Joseph P. Vacca a, Hemaka A. Rajapakse a,
*
a Department of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
b Department of Alzheimer’s Research, Merck Research Laboratories, West Point, PA 19486, USA
c Department of Imaging Research, Merck Research Laboratories, West Point, PA 19486, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
This Letter describes the one pot synthesis of tertiary carbinamine 3 and related analogs of brain pene-
trant BACE-1 inhibitors via the alkylation of the Schiff base intermediate 2. The methodology developed
for this study provided a convenient and rapid means to explore the P1 region of these types of inhibitors,
where the P1 group is installed in the final step using a one-pot two-step protocol. Further SAR studies
led to the identification of 10 which is twofold more potent in vitro as compared to the lead compound.
This inhibitor was characterized in a cisterna magna ported rhesus monkey model, where significant low-
ering of CSF Ab40 was observed.
Received 23 November 2009
Revised 30 December 2009
Accepted 4 January 2010
Available online 11 January 2010
Keywords:
BACE
AD
Ó 2010 Elsevier Ltd. All rights reserved.
Tertiary carbinamine
Alzheimer’s disease (AD) is a progressive neurodegenerative
disorder initially manifested by memory loss. Progression of the
disease leads to behavior and personality changes and ultimately
death as no true disease modifying therapy currently exists. The
implications of AD in terms of the financial and emotional burden
of caring for affected patients is immense.1 According to the amy-
loid cascade hypothesis, the deposition of amyloid b-peptide (Ab)
in the brain is one of the characteristics of AD pathogenesis.2 Ab
is formed by sequential processing of amyloid precursor protein
(APP) by two aspartyl proteases, b-secretase (BACE-1, b-site APP
ities were associated with 1, such as high clearance and poor oral
bioavailability in multiple species due to extensive first pass
P2
O O
S
N
Cl
N
Cleaving Enzyme) followed by c-secretase. BACE-1 knockout mice
N
N
show a complete absence of Ab production but are otherwise sim-
ilar to wild type animals. BACE-1 is therefore hypothesized to be an
attractive therapeutic target for the treatment of AD.3
N
O
NH2
OMe
P3
The development of brain penetrant BACE-1 inhibitors has been
extremely challenging.4 Recently, the discovery of a tertiary car-
binamine derived BACE-1 inhibitor was disclosed.5 This compound,
represented by structure 1 (Fig. 1), is a brain penetrant BACE-1
inhibitor with excellent in vitro (IP6 = 5 nM at pH 6.5)7 and cell-
based potencies (IC50 = 40 nM). Significant pharmacokinetic liabil-
P1
1
BACE-1 IP (pH 4.5) = 0.4 nM
BACE-1 IP (pH 6.5) = 5 nM
sAPP_NF IC50 = 46 22 nM
PGP ratio: (h) = 1.9, (m) = 2.3, Papp = 22
Brain/plasma ratio (i.p. 30 mpk): 9%
Log P > 3.2
Abbreviations: IP, inflection point; sAPP_NF, cell based assay utilizing NF
mutation for sAPP. See Ref. 15; PGP, P-glycoprotein; Papp, apparent permeability
value; PB, protein binding; IP, dosing: intraperitoneal dosing; Mpk, milligrams of
compound per kilogram of animal; BLQ, below limit of quantitation.
* Corresponding author.
PB > 98%
E-mail address: hemaka_rajapakse@merck.com (H.A. Rajapakse).
Figure 1. Profile of BACE-1 inhibitor 1.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.005

1780
H. Zhu et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1779–1782
metabolism. Nevertheless, upon IP dosing in transgenic mice at
high doses, reduction of brain Ab levels was observed. Upon co-
dosing with ritonavir, a CYP3A4 inhibitor widely used to improve
the pharmacokinetics of HIV protease inhibitors,8 twice daily dos-
ing with 1 (15 mpk) reduced CSF Ab levels by ꢀ40% in cisterna
magna ported (CMP) rhesus monkeys. This data provided, to the
best of our knowledge, the first example of Ab lowering upon oral
dosing of a BACE-1 inhibitor in non-human primates.
O O
S
O O
S
N
N
a, b, c
d
N
N
H
N
N
CO2H
N
NH
O
NHBoc
O
OMe
OMe
4
5
In a continuation of our efforts to discover an improved BACE-1
inhibitor, we sought to improve upon the profile of 1. The co-crys-
tal structure of 1 with the human BACE-1 enzyme was successfully
obtained,5 revealing that the inhibitor occupied S1-S3 sites as had
been observed for other inhibitors derived from this series. The
benzyl group occupied the S1 pocket, and we wished to explore
the feasibility of improving compound potency via optimization
the P1 substituent.
O O
S
O O
S
N
N
Cl
N
N
e, f
g, h
N
N
N
N
N
N
O
O
NH2
TFA
NHBoc
OMe
OMe
6
The oxadiazole moiety of 1 was derived from the bridging of
two carboxylic acids with hydrazine, followed by dehydration. In
particular, the right hand tertiary carbinamine fragment is derived
7
O O
S
O O
S
from an
blocks are not readily available from commercial sources, necessi-
tating custom synthesis. While accessing -disubstituted amino
a,a-disubstituted amino acid. These types of building
N
N
Cl
Cl
N
N
a,a
i
N
N
N
N
acids via Schiff base alkylation of
a-amino esters is well prece-
N
N
dented,9 the rapid synthesis of P1 variants was precluded due to
its multistep nature. Given that oxadiazoles are ester mimetics,
O
O
NH2
N
OMe
OMe
R
Ph
we felt that the alkylation of an
a-amino oxadiazole Schiff base,
Ph
2
3
though unprecedented, could provide a viable late stage P1 varia-
tion strategy (Scheme 1). Given the convenience offered by this ap-
proach, we pursued the synthesis of key Schiff base intermediate 2
as shown in Scheme 2.
Scheme 2. Synthesis of tertiary carbinamine derived BACE inhibitors. Reagents: (a)
Boc-hydrazine, EDC, HOAt, DMF, 72%; (b) HCl, EtOAc, 100%; (c) Boc-DL-Ala-OH, EDC,
HOAt, Hunig’s base, DMF, 81%; (d) Ph3P, CBr4, imidazole, DCM, 74%; (e) TFA, DCM,
100%; (f) NCS, DCM; (g) HCl, DCM, 58% (2 steps); (h) benzophenone imine, DCM,
84%; (i) LiHMDS, RCH2Br or RCH2Cl, THF, then aq 6 N HCl.
The synthesis was initiated by sequential EDC coupling of isoni-
cotinic acid 45 with Boc-hydrazine, deprotection, then coupling
with Boc-DL-Ala-OH to afford the diacylhydrazide 5. Cyclodehy-
dration with Ph3P, CBr4 and imidazole provided desired oxadiazole
6.10 Deprotection of the Boc group required mildly acidic TFA con-
ditions to prevent unraveling of the oxadiazole. Chlorination of the
amine trifluroacetate salt with NCS gave a 3:1 mixture of regioi-
somers favoring the desired product which was separable via pre-
parative HPLC. As an interesting aside, we consistently observed
that the rate of chlorination was faster when the hydrochloride salt
was utilized, at the cost of lower regioselectivity and diminished
yield due to oxadiazole hydrolysis. Conversely, the transformation
of 7 to the desired Schiff base 2 necessitated the utilization of the
HCl salt of 7 to achieve full conversion, presumably due to the
insoluble nature of ammonium chloride in dichloromethane driv-
ing the equilibrium toward the desired product.
tion of the reaction mixture via preparative HPLC gave the desired
product as a mixture of diastereomers at the
a-amino acid cen-
ter.11 The diastereomeric mixtures could be readily resolved via
chiral chromatography for analogs of interest.12 The abovemen-
tioned alkylation reaction proved to be robust, where desired ana-
logs were obtained with various substituted benzyl halides (Table
1) as well as heteroaryl substituted halo methane derivatives and
alkyl halides.
In general, SAR trends revealed that substitution at the para po-
sition was better tolerated than at either the ortho or meta posi-
tions. In fact, the incorporation of a 4-fluoro substituent gave
compound 10,13 with a twofold improvement in both our in vitro
and cell based assays as compared to 1. However, a substituent lar-
ger than fluorine at the para position lead to a large potency loss
compared to reference compound 1. A variety of heteroaryl groups
and cyclopropyl derived phenyl replacements were also screened,
but these transformations universally led to large losses of po-
tency. Unfortunately no improvement in the intrinsic metabolic
clearance was observed for the fluorinated compound 10 (Fig. 2).
As compound 10 represented a twofold improvement for
in vitro potency over our lead compound 1, we wondered how this
comparison would translate in vivo. In particular, we hoped that in
the established CMP rhesus monkey model,14 we would observe
enhanced reduction of CSF Ab levels at the same dose, or similar
reduction of the biomarker at a lower dose. The rhesus pharmaco-
kinetics of compound 10 when co-administered with ritonavir
were determined in a satellite experiment (n = 3 rhesus). Even
though the profile for compound 1 was determined previously,
the experiment was repeated, and the plasma levels at 4 h and
24 h are tabulated below (Table 2). Given the small sampling size,
we were faced with large standard deviations for the averages of
plasma levels at a given timepoint.
With intermediate 2 in hand, we wished to examine the feasi-
bility of the key alkylation reaction. The optimal conditions that
were discovered involved the deprotonation of 2 in DMF with
LiHMDS, followed by addition of the alkylating agent. Once the
reaction was complete, the addition of a few drops of 6 N HCl
achieved the removal of the Schiff base in the same pot. Purifica-
O O
S
O O
S
N
N
Cl
O
Cl
N
N
N
N
N
N
N
N
O
R
NH2
N
OMe
OMe
Ph
Ph
2
3
Scheme 1. P1 scan strategy.

H. Zhu et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1779–1782
1781
Table 1
O O
S
Summary of in vitro data for P1 analogs
N
Compound
R
BACE-1 IP (nM)
5
sAPP_NF15 IP (nM)
46 22
Cl
O
N
N
N
1
N
NH2
F
OMe
8
9
15*
17*
2*
195*
304*
F
F
F
10
BACE-1 IP = 2 nM
sAPP_NF IC50 = 15 6 nM
PGP ratio: (h) = 1.1, (m) = 1.9, Papp = 20
Prot. Bind 97.8%
10
15 6*
Cl
Figure 2. Profile of BACE-1 inhibitor 10.
11
12
13
340*
65*
4*
>6700*
4315*
320*
Table 2
Cl
Rhesus pharmacokinetics for compounds
10 mpk ritonavir (RTV)
1 and 10 when co-administered with
Dose (+10 mpk RTV)
(Plasma)4 h
(lM)
(Plasma)24 h (lM)
15 mpk 1
5 mpk 10
15 mpk 10
3.36 1.95
1.31 0.88
2.34 1.96
0.10 0.09
BLQ
0.03 0.02s
Cl
Br
14
15
>600,000*
29,000
Br
CSF A 40 Levels
β
Br
1,200
1,100
1,000
900
15 mpk 1
5 mpk 10
15 mpk 10
Vehicle
F
Br
O
16
17
3500
1500
800
O
700
600
18
180
4300
1100
O
500
400
-20
O
19
20
119
0
4
24
O
Time (h)
Figure 3. CSF Ab40 reduction in rhesus monkeys.
O
9900
N
In order to compare the pharmacodynamics of the two analogs,
a four-way single dose crossover experiment was designed. Both
plasma and CSF were sampled at time points À20 (predose), 0, 4
and 24 h and analyzed for peptides related to the amyloid cascade
(Fig. 3), Gratifyingly, significant reductions of Ab40 (ꢀ30%) were
observed in the CSF at 4 h in all three treated groups. Surprisingly,
the reduction of this biomarker at a 5 mpk dose of 10 was compa-
rable to the 15 mpk dose 1. However, increasing the dose of 10 to
15 mpk did not result in enhanced Ab40 at the 4 h timepoint. Also
observed were decreases in Ab42 and sAPPb concentrations, which
are consistent with the inhibition of BACE-1.14c At the second and
final timepoint of 24 h, more robust Ab40 reduction was observed
when treated with 15 mpk of compound 1, which could be attrib-
uted to its superior pharmacokinetics when co-administered with
ritonavir. Overall, the pharmacodynamics of compound 10 were
not meaningfully different from 1 despite the in vitro potency
improvement.
21
1400
O
N
22
23
375
142
3186
1934
O
O
N
24
25
49,000
1300
O
>2200
*
Racemate was separated by chiral HPLC. Data showed is of the active enantiomer.

1782
H. Zhu et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1779–1782
A.; Rush, D.; Jones, K. L. G.; Zhu, H.; Rajapakse, H. A.; Graham, S. L.; Simon, A.;
In summary, we developed a one pot procedure for the facile
Holloway, M. K.; Allison, T. J.; Munshi, S. K.; Espeseth, A. S.; Zuck, P.; Colussi, D.;
Wolfe, A.; Pietrak, B. L.; Lai, M.-T.; Vacca, J. P. Bioorg. Med. Chem. Lett. 2009, 19,
2977.
synthesis of P1 analogs of active BACE-1 inhibitor 1 via late stage
Schiff base alkylation. A more potent inhibitor of the enzyme was
identified from this effort and its efficacy in the rhesus CSF Ab low-
ering was evaluated. Unfortunately, the modest improvement of
in vitro potency did not result in superior pharmacodynamic re-
sponse for compound 10.
8. Kempf, D. J.; Marsh, K. C.; Kumar, G.; Rodrigues, A. D.; Denissen, J. F.; McDonald,
E.; Kukulka, M. J.; Hsu, A.; Granneman, G. R.; Baroldi, P. A.; Sun, E.; Pizzuti, D.;
Plattner, J. J.; Norbeck, D. W.; Leonard, J. M. Antimicrob. Agents Chemother. 1997,
41(3), 654.
9. O’Donnell, M. J.; Bennett, W. D.; Wu, S. J. Am. Chem. Soc. 1989, 111, 2353.
10. Rajapakse, H. A.; Nantermet, P. G.; Selnick, H. G.; Munshi, S.; McGaughey, G. B.;
Lindsley, S. R.; Young, M. B.; Lai, M. T.; Espeseth, A. S.; Shi, X. P.; Colussi, D.;
Pietrak, B.; Crouthamel, M. C.; Tugusheva, K.; Huang, Q.; Xu, M.; Simon, A. J.;
Kuo, L.; Hazuda, D. J.; Graham, S.; Vacca, J. P. J. Med. Chem. 2006, 49, 7270.
11. The purities of the target compounds and the enantiomeric excess for
compounds that were separated by chiral chromatography were greater than
97%. For experimental details, see Supplementary data.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.01.005.
12. The diastereomers resulting from this alkylation could be separated on a
ChiralPack AD, gradient from 90/10 to 60/40 hexane/iPrOH (hexane modified
w/0.1% DEA).
References and notes
13. The stereochemistry of the active diastereomer is known to always be R at the
tertiary carbinamine center based on earlier X-ray crystallographic analysis of
tertiary carbinamine inhibitors Refs. 5,9
1. Nussbaum, R. L.; Ellis, C. E. N. Eng. J. Med. 2003, 348, 1356.
2. Cummings, J. L. N. Eng. J. Med. 2004, 351, 56.
3. Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.; Branstetter, D. G.; Chen,
K. S.; Freedman, S.; Frigon, N. L.; Games, D.; Hu, K.; Johnson-Wood, K.;
Kappenman, K. E.; Kawabe, T.; Kola, I.; Kuehn, R.; Lee, M.; Liu, W.; Motter, R.;
Nichols, N. F.; Power, M.; Robertson, 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. Recent reviews on BACE-1 inhibitors (a) Durham, T. B.; Shepherd, T. A. Curr.
Opin. Drug Disc. Dev. 2006, 9, 776; (b) Guo, T.; Hobbs, D. W. Curr. Med. Chem.
2006, 13, 1811; (c) Beher, D.; Graham, S. L. Expert Opin. Investig. Drugs 2005, 14,
1385; (d) Gosh, A. K.; Kumaragurubaran, N.; Tang, J. Curr. Top. Med. Chem. 2005,
5, 1609; (e) Thompson, L. A.; Bronson, J. J.; Zusi, F. C. Curr. Pharm. Des. 2005, 11,
3383.
5. Nantermet, P. G.; Rajapakse, H. A.; Stanton, M. G.; Stauffer, S. R.; Barrow, J. C.;
Gregro, A. R.; Moore, K. P.; Steinbeiser, M. A.; Swestock, J.; Selnick, H. G.;
Graham, S. L.; McGaughey, G. B.; Colussi, D.; Lai, M. T.; Sankaranarayanan, S.;
Simon, A. J.; Munshi, S.; Cook, J. J.; Holahan, M. A.; Michener, M. S.; Vacca, J. P.
ChemMedChem 2009, 4, 37–40.
14. (a) Gilberto, D. B.; Zeoli, A. H.; Szczerba, P. J.; Gehret, J. R.; Holahan, M. H.; Sitko,
G. R.; Johnson, C. A.; Cook, J. J.; Motzel, S. L. Contemp. Top. Lab. Anim. Sci. 2003,
42, 53; (b) Cook, J. J.; Sankaranarayanan, S.; Holahan, M.; Michener, M.;
Crouthamel, M.; Colussi, D.; Wu, G.; Pietrak, B.; Miller, R.; Barbacci, D.;
Paweletz, C.; Lee. A.; Hendrickonn, R.; Devanarayan, V.; Shearman, M.; Simon,
A. Translational biomarker discovery in rhesus monkey csf and plasma, Vol. 4
presented at Alzheimer’s and Parkinson’s Diseases: Progress and new
prospectives 8th international conference, Salzburg, Austria, March 18–21,
2007, 144.; (c) Sankaranarayanan, S.; Holahan, M. A.; Colussi, D.; Crouthamel,
M.; Devanarayan, V.; Ellis, J.; Espeseth, A.; Gates, A. T.; Graham, S. L.; Gregro, A.
R.; Hazuda, D.; Hochman, J. H.; Holloway, K.; Jin, L.; Kahana, J.; Lai, M.;
Lineberger, J.; McGaughey, G.; Moore, K. P.; Nantermet, P. G.; 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(1), 131.
15. Shi, X.-P.; Tugusheva, K.; Bruce, J. E.; Lucka, A.; Chen-Dodson, E.; Hu, B.; Wu, G.-
X.; Price, E.; Register, R. B.; Lineberger, J.; Miller, R.; Tang, M.-J.; Espeseth, A.;
Kahana, J.; Wolfe, A.; Crouthamel, M.-C.; Sankaranarayanan, S.; Simon, A.;
Chen, L.; Lai, M.-T.; Pietrak, B.; DiMuzio, J.; Li, Y.; Xu, M.; Huang, Q.; Garsky, V.;
Sardana, M. K.; Hazuda, D. J. J. Alzheimer’s Dis. 2005, 7, 139.
6. See abbreviations.
7. It is more predictive of the measured cellular potency if the enzymatic BACE-1
assay is conducted at PH 6.5 rather than pH 4.5. The data from our study
suggests that APP cleavage is occurring in less acidic cellular components than
previously theorized. For additional information see: Stachel, S. J.; Coburn, C.