2-(N-Benzyl-N-phenylsulfonamido)alkyl amide derivatives as γ-secretase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2012.09.057

A series of (N-benzyl-N-phenylsulfonamido)alkyl amides were developed from classic and parallel synthesis strategies. Compounds with good in vitro and in vivo γ-secretase activity were identified and described.

Full text of DOI:10.1016/j.bmcl.2012.09.057

Bioorganic & Medicinal Chemistry Letters 22 (2012) 6828–6831
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
2-(N-Benzyl-N-phenylsulfonamido)alkyl amide derivatives as
c-secretase inhibitors
⇑
Michael F. Parker , Donna M. Barten, Carl P. Bergstrom, Joanne J. Bronson, Jason A. Corsa, Michael F. Dee,
Yonghua Gai, Valerie L. Guss, Mendi A. Higgins, Daniel J. Keavy, Alice Loo, Robert A. Mate, Larry R. Marcin,
Katharine E. McElhone, Craig T. Polson, Susan B. Roberts, John E. Macor
Molecular Sciences and Candidate Optimization and Neuroscience Discovery Biology, Bristol-Myers Squibb R&D, 5 Research Parkway, Wallingford, CT 06492, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 July 2012
Revised 7 September 2012
Accepted 17 September 2012
Available online 24 September 2012
A series of (N-benzyl-N-phenylsulfonamido)alkyl amides were developed from classic and parallel syn-
thesis strategies. Compounds with good in vitro and in vivo
described.
c-secretase activity were identified and
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Alzheimer’s
Gamma secretase inhitors
Phenyl sulfonamides
Alzheimer’s disease (AD) is a progressive, neurodegenerative
disorder characterized by memory impairment and cognitive dys-
function.1 Evidence suggests that soluble oligomers of b-amyloid
peptides (b-AP) may be responsible for the neuronal toxicity that
is associated with AD.2 b-AP in turn is generated by the sequential
proteolytic cleavage of a 695–770 amino acid called the amyloid
the caprolactam leads were not optimal due to their limited phar-
macokinetic properties and their potent cytochrome P450 (Cyp)
inhibition. As part of SAR elucidation and optimization of the cap-
rolactam series, we investigated ring-opened acyclic amide ana-
logues. This led to a new and promising lead series, exemplified
by the acyclic sulfonamide 2, which was significantly more potent
than 1.4 Based on the impressive improvement in potency achieved
with this simple structural modification, SAR studies were em-
precursor protein (APP) by the action of both b and c-secretases.
We have been interested in identifying small molecules that inhi-
bit the production of b-amyloid peptide (b-AP) from b-amyloid
ployed to optimize the acyclic amides for potency as c-secretase
precursor protein (b-APP) via
agents may be useful for the treatment or prevention of Alzhei-
mer’s disease. A caprolactam series of -secretase inhibitors, exem-
c
-secretase inhibition, since such
inhibitors. One of the most advanced leads from this series,
BMS-708163 (avagecestat), currently a clinical candidate, has been
recently disclosed.5 Herein we describe the initial SAR and optimi-
zation of the acyclic sulfonamide chemotype that led to the discov-
ery of avagecestat.
c
plified by compound 1, was previously reported from our
laboratories.3 While having good
c-secretase inhibitory potency,
H
O
CF3
N
O
S
O
O
O
N
SO2
H2N
Cl
H2N
S
Cl
N
N
O
O
O
O
N
F
Cl
O
2
1
N
(BMS-708163)
IC50 = 4.1
IC50 = 0.19 nM
IC50 = 0.30 nM
⇑
Corresponding author.
E-mail address: parkerm@bms.com (M.F. Parker).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.09.057

M. F. Parker et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6828–6831
6829
O
H2N
Method 1
c or d
b
NH
SO2
AA
O
H2N
O
R
H2N
a
N
5
X
Boc-AA-OH
NH2
AA
SO2
AA
4
Method 2
6
X
O
e
b
H2N
R
N
H
AA
7
Scheme 1. General synthesis of N-alkyl-N-phenylsulfonamide derivatives. Reagents and conditions: AA = readily available
D or L amino acid side chains (a) EDC, HOBt, DIEA,
NH4Cl, DMF; then TFA, CH2Cl2; (b) X-ArSO2Cl, CH2Cl2, Et3N; (c) Cs2CO3, MeCN, R-Br; (d) DEAD, TPP, THF; (e) R-CHO, NaCNBH4, ZnCl2, MeOH.
As shown in Scheme 1, readily accessible (D)- or (L)-Boc-amino
Table 1
acids such as D-norvaline were subjected to carbodiimide activa-
tion and primary amide formation with ammonium chloride, fol-
lowed by Boc deprotection6 to provide the synthetically useful
chiral amide/amine 4. Sulfonylation of the primary amine afforded
intermediates 5. The resulting secondary sulfonamides were then
alkylated on the sulfonamide nitrogen by treatment with an array
of alkyl or benzyl halides in acetonitrile in the presence of Cs2CO3
to afford the sulfonamides 6. Alternatively, side-chains were intro-
duced by treatment of 5 with an alcohol under Mitsunobu condi-
tions in cases where the alcohols were more readily available or
where reactivity considerations favored the Mitsunobu coupling
approach. The R-group and sulfonyl substituents could also be
introduced in reverse order (Scheme 1, Method 2). In this se-
quence, amine 4 was first subjected to reductive amination by
treatment with an aldehyde and sodium cyanoborohydride in
methanol to provide intermediates 7. The resulting secondary
amines were treated with a variety of sulfonyl chlorides in the
presence of triethyl amine to produce the desired analogs 6.
Binding activity4 for amide replacement derivatives
O
S
O
Cl
R
N
X
Compd
R
X
Ab42 IC50 (nM)10
O
9
4-MeO
4-MeO
1100
N
O
10
380
N
H
O
O
HO
11
12
13
14
3-F,4-MeO
4-MeO
4-MeO
4-CF3
19
N
H
Our initial investigation focused on whether c-secretase activ-
ity was related to a particular enantiomer of the ring-opened struc-
ture. Sulfonamide 8, the (S) enantiomer of 2, was prepared and
found to be weakly active (IC50 = 315 nM), indicating the impor-
tance of the (R)-stereochemistry. Similar results were seen for
other enantiomer pairs, where the (R)-isomer was always more po-
tent than the (S)-isomer (data not shown). Not surprisingly this
preference for the (R)-enantiomer was also seen with the parent
caprolactams.3
O
160
53
N
O
H2N
N
H
O
H2N
110
N
H
N
(S)
O
15
16
17
4-MeO
2200
1200
220
N
O
S
O
H2N
Cl
N
S
N
4-MeO
O
N
O
O
3-F,4-MeO
8
HO
IC50 = 315 nM
even though this secondary amide more closely resembles the ini-
tial caprolactam lead 1. This is likely due to the preferred (E)-geom-
etry of the acyclic amide compared with the enforced (Z)-geometry
of the lactam.
With the preferred stereochemistry established, our attention
turned to modification of the primary amide. The results from test-
ing of several amide isosteres and replacements are shown in Table
1. The N-hydroxyamide 11 was the only functional group that con-
ferred reasonable potency although it was 100 times less potent
than 2.7 Most amide replacements suffered a greater than 1000-fold
loss of potency. Surprisingly, the N–Me amide 10 was far less active,
As had been found with the caprolactam series,3 the sulfon-
amide moiety was preferred over a non-sulfonamide group as
shown with the examples in Table 2. Thus, replacement of the sul-
fonamide by an amide (18) a 2-phenylacetamide (not shown), or
urea (19) resulted in a dramatic loss of binding activity. For the

6830
M. F. Parker et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6828–6831
Table 2
Binding activity4 for sulfonamide replacement derivatives
H2N
R
N
O
X
Compd
R0
X
Ab42IC50 (nM)10
18
19
20
21
22
23
24
25
26
27
28
4-Chlorobenzamide
1-Phenylurea
Benzenesulfonamide
4-Chlorobenzene sulfonamide
3-Chlorobenzene sulfonamide
4-Fluorobenzene sulfonamide
4-Fluorobenzene sulfonamide
4-Methylbenzene sulfonamide
4-Methylbenzene sulfonamide
3-Methylbenzene sulfonamide
3-F,4-MeO
3-F,4-MeO
3-F,4-MeO
3-F,4-MeO
3-F,4-MeO
3-F,4-MeO
4-CN
3-F,4-MeO
4-CN
3-F,4-MeO
3-F,4-MeO
Inactive
1100
1.7
0.15
7.0
0.26
2.7
0.45
2.8
9.2
0.27
4-Trifluoromethyl benzenesulfonamide
Table 3
Binding activity4 for N-alkyl side chain replacement derivatives
Table 4
Binding activity4 for substituted sulfonamide- N-benzyl derivatives
O
S
O
O
O
H2N
Cl
H2N
S
Cl
N
N
O
R
O
R
Compd
R
Ab42 IC50 (nM)10
Compd
R
Ab42 IC50 (nM)10
29
CH3CH2–
2.3
37
38
39
2
40
41
42
43
44
45
46
47
Benzyl
0.35
23
2-Methoxybenzyl
3-Methoxybenzyl
4-Methoxybenzyl
4-Trifluoromethyl benzyl
4-Chloromethyl benzyl
4-Trifluoromethoxybenzyl
4-Cyanobenzyl
4-Fluorobenzyl
4-Isopropylbenzyl
4-t-Butylbenzyl
4-Dimethylaminobenzyl
N
30
31
17
0.59
0.19
0.39
0.35
0.69
0.67
0.43
0.58
0.39
0.27
4.2
S
O
O
32
33
34
9.9
3.5
4.9
N
(CH2)4
(CH2)6
N
N
N
N
48
7.7
(CH2)6
3-Phenylpropyl
Cyclohexylmethyl
49
21
50
3,4-Difluorobenzyl
3-Fluoro-4-methoxybenzyl
3,5-Difluorobenzyl
0.44
0.15
0.62
35
36
0.41
0.79
substitution on the aryl group on the sulfonamide, removal of the
4-chloro substituent (as with the phenyl sulfonamide 20) or shift-
ing the substituent to the 3-position (22, 27) resulted in a decrease
in potency compared with the para-chlorosulfonamide (21). 4-Flu-
oro, 4-methyl and 4-trifluoromethyl groups were found to be
acceptable chloro replacements in vitro. However futher in vivo
evaluation elliminated these chloro alternatives.8
Analogs with non-benzylic side chains R on the sulfonamide
nitrogen were also examined (Table 3). The simple ethyl analog
29 had reasonable potency, although this compound was 10-fold
less potent than 21. A variety of substituents were tolerated on
non-benzylic side-chains as seen with compounds 30–34, although
derivatives with polar or bulky groups were consistently less po-
tent than the benchmark 21. The most potent of the non-benzylic
R analogs were the phenylpropyl derivative 35 (IC50 = 0.41 nM)
and the cyclohexyl analog 36 (IC50 = 0.79 nM). As has been more
reciently shown by this data and others9 the superior potency of
21, 35, and 36 indicate a preference for a lipophilic group on the
R side-chain, with slight favoring for a benzylic substituent.
Analogs shown in Table 4 were focused on examination of sub-
stitution on the N-benzyl moiety. Ortho-substitution was clearly
disfavored as seen with the 2-methoxybenzyl derivative 38. By
comparison, both meta- and para- substituted analogs were gener-
ally equipotent, regardless of whether the substituent was elec-
tron-donating or electron-withdrawing (compounds 38–47).
Combination of a para- and a meta-substituent was acceptable, as
seen with analogues 21, 49, and 50.
With SAR for key regions of the molecule established, modifica-
tion of the amino acid derived alkyl side chain was investigated
next (Table 5). It was apparent that a chain length of 3 atoms or
more was preferred (compare 52, 53, 54, and 56), with branching

M. F. Parker et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6828–6831
6831
Table 5
In summary, potent
c
-secretase inhibitors emerged from SAR
Binding activity4 for alpha-alkyl side chain replacement derivatives
studies focused on the initial optimization of our sulfonamide lead.
The most potent compounds were the 3-fluoro-4-methoxybenzyl
compound 55 and the 4-cyanobenzyl compound 58, both of which
are derived from readily available
showed promising in vivo effects on brain Ab40 reduction in early
screening studies. Efforts to further optimize this sulfonamide ser-
ies of c-secretase inhibitors as well as their SAR versus other pro-
teins cleaved by b-secretase (i.e. Notch) will be reported in due
course.
R
O
S
O
H2N
Cl
D-leucine. These compounds
N
O
R'
Compd
R
R0
Ab42 IC50
(nM)10
51
52
H
3-Fluoro-4-
methoxybenzyl
3-Fluoro-4-
2000
5.3
CH3–
Supplementary data
methoxybenzyl
4-Cyanobenzyl
4-Cyanobenzyl
3-Fluoro-4-
53
54
55
CH3CH2–
CH3CH2CH2–
CH3CH2CH2–
4.4
2.6
0.52
Supplementary data (experimental details for synthetic proce-
dures and associated chemical data for compounds 2, and 9–17
are available) associated with this article can be found, in the on-
line version, at http://dx.doi.org/10.1016/j.bmcl.2012.09.057.
methoxybenzyl
4-Cyanobenzyl
3-Fluoro-4-
56
57
Butyl–
(CH3)2CH–
1.0
1.4
methoxybenzyl
3-Fluoro-4-
methoxybenzyl
21
(CH3)2CHCH2–
0.15
References and notes
58
59
60
Cyclopropylmethyl 4-Cyanobenzyl
Cyclopropylmethyl 4-Trifluoromethyl
0.94
0.68
0.35
1. Beier, M. T. Pharmacotherapy 2007, 27, 399.
2. Haass, C.; Selkoe, D. J. Nat. Rev. Mol. Cell Biol. 2007, 8, 101.
3. Parker, M. F.; Bronson, J. J.; Barten, D. M.; Corsa, J. A.; Du, W.; Felsenstein, K. M.;
Guss, V. L.; Izzarelli, D.; Loo, A.; McElhone, K. E.; Marcin, L. R.; Padmanabha, R.;
Pak, R.; Polson, C. T.; Toyn, J. H.; Varma, S.; Wang, J.; Wong, V.; Zheng, M.;
Roberts, S. B. Bioorg. Med. Chem. Lett. 2007, 17, 5790.
CH2@CHCH2–
3-Fluoro-4-
methoxybenzyl
4-Methoxybenzyl
4-Cyanobenzyl
4-Cyanobenzyl
4-Cyanobenzyl
4-Cyanobenzyl
4-Cyanobenzyl
61
62
63
64
65
66
Ph-CH2–
390
19
1800
4300
48
CH3S(CH2)2–
CH3SO(CH2)2–
CH3SO2(CH2)2–
CH3OCH2–
4. The cellular GS H4-8 SW1-42 assay was performed as described (a) Truong, A.
P.; Aubele, D. L.; Probst, G. D.; Neitzel, M. L.; Semko, C. M.; Bowers, S.; Dressen,
D.; Hom, R. K.; Konradi, A. W.; Sham, H. L.; Garofalo, A. W.; Keim, P. S.; Wu, J.;
Dappen, M. S.; Wong, K.; Goldbach, E.; Quinn, K. P.; Sauer, J.-M.; Brigham, E. F.;
Wallace, W.; Nguyen, L.; Hemphill, S. S.; Bova, M. P.; Basi, G. Bioorg. Med. Chem.
Lett. 2009, 19, 4920; (b) Parker, M. F.; Barten, D. M.; Bergstrom, C. P.; Bronson, J.
J.; Corsa, J. A.; Deshpande, M. S.; Felsenstein, K. M.; Guss, V. L.; Hansel, S. B.;
Johnson, G.; Keavy, D. J.; Lau, W. Y.; Mock, J.; Prasad, C. V. C.; Polson, C. T.; Sloan,
C. P.; Smith, D. W.; Wallace, O. B.; Wang, H. H.; Williams, A.; Varma, S.; Zheng,
M. Bioorg. Med. Chem. Lett. 2007, 17, 4432.
5. Gillman, K. W.; Starrett, J. E.; Parker, M. F.; Xie, K.; Bronson, J. J.; Marcin, L. R.;
McElhone, K. E.; Bergstrom, C. P.; Mate, R. A.; Williams, R.; Meredith, J. E.;
Burton, C. R.; Barten, D. M.; Toyn, J. H.; Roberts, S. B.; Lentz, K. A.; Houston, J. G.;
Zaczek, R.; Albright, C. F.; Decicco, C. P.; Macor, J. E.; Olson, R. E. Med. Chem. Lett.
2010, 1, 12.
(CH3)3COCH2–
54
as in analogues 2, 57–59 giving the best potency. Removal of the
side-chain altogether (51) resulted in a significant loss of potency.
The presence of unsaturation as in a terminal alkene 60 was well-
tolerated. Aromatic substituents (61) were much less preferred as
was the addition of polar substituents (62–66) on the side chain.
Compounds 55, 58 and 59 were among the more potent ana-
logues and were selected for in vivo screening to compare with
the caprolactam 1. Amyloid precursor protein Transgenic mice
(Tg2576 mice)7, 3–6 months of age, were treated by oral gavage
6. McMurray, J. S.; Wang, W. Tetrahedron Lett. 1999, 2501.
7. Kitas, E. A.; Galley, G.; Jakob-Roetne, R.; Flohr, A.; Wostl, W.; Mauser, H.; Alker,
A. M.; Czech, C.; Ozmen, L.; David-Pierson, P.; Reinhardt, D.; Jacobsen, H. Bioorg.
Med. Chem. Lett. 2008, 18, 304.
8. Compounds 21, 24, 26, and 28 were analyzed in our standard transgenic model
(described in the text) for their potential to reduce Ab40 in plasma and brain at
with test compounds at a single dose of 200 lmol/kg. The effect
of compound treatment on Ab40 levels6 was measured 180 min
after dosing in brain and plasma (Table 6), along with exposures
of the test compounds. All three compounds showed a marked
improvement in brain Ab40 lowering relative to 1, presumably
due in part to their greater intrinsic potency. Interestingly, only
58 and 59 showed higher reduction in plasma Ab40 levels. Com-
pound 58 had the best metabolic stability and showed correspond-
ingly better plasma exposures at the 180 min time-point, although
brain concentrations were not appreciably increased relative to 1.
As is with single time-point and single dose studies, these data
cannot be used to draw conclusions about correlations between
exposure and lowering of Ab40 levels. However, these data served
to demonstrate the in vivo potential of the acyclic sulfonamides,
and were the impetus for continued interest in and optimization
of this chemotype.
a dose of 200
advantage.
lmol/kg. The 4-chloro analog (21) was shown to have a clear
9. (a) Stepan, A. F.; Karki, K.; McDonald, W. S.; Dorff, P. H.; Dutra, J. K.; DiRico, K. J.;
Won, A.; Subramanyam, C.; Efremov, I. V.; O’Donnell, C. J.; Nolan, C. E.; Becker,
S. L.; Pustilnik, L. R.; Sneed, B.; Sun, H.; Lu, Y.; Robshaw, A. E.; Riddell, D.;
O’Sullivan, T. J.; Sibley, E.; Capetta, S.; Atchison, K.; Hallgren, A. J.; Miller, E.;
Wood, A.; Obach, R. S. J. Med. Chem. 2011, 54, 7772; (b) Stepan, A. F.;
Subramanyam, C.; Efremov, I. V.; Dutra, J. K.; O’Sullivan, T. J.; DiRico, K. J.;
McDonald, W. S.; Won, A.; Dorff, P. H.; Nolan, C. E.; Becker, S. L.; Pustilnik, L. R.;
Riddell, D.; Kauffman, G. W.; Kormos, B. L.; Zhang, L.; Lu, Y.; Capetta, S.; Green,
M. E.; Karki, K.; Sibley, E.; Atchison, K.; Hallgren, A. J.; Oborski, C. E.; Robshaw,
A. E.; Sneed, B.; O’Donnell, C. J. J. Med. Chem. 2012, 55, 3414.
10. IC50 values were measured using a GS H4-8 SW1-42 assay. Values are means of
P3 experiments, with 12 drug concentrations in each exp.; intra-assay
variance <10%.
11. The stability in liver microsomes was determined by a high throughput in-
house assay using substrate concentrations of 3 lM microsomal protein across
species. Incubations were performed at 37 °C in sodium phosphate buffer
(100 mM), pH 7.4, and quenched after 10 min. Samples were analyzed by
HPLC/MS.
Table 6
In vivo evaluation of selected sulfonamnide
c
-secretase inhibitors
Brain concn (nM)
Compd
Brain Ab40 Inh. (%)
Plasma Ab40 Inh. (%)
Plasma concn (nM)
Metabolic stability11 (nmol/min/mg)
1
55
58
59
49
72
75
78
2200 1200
380 110
2800 510
6300 1100
47
41
81
65
280 90
840 240
8600 260
3600 1300
0.23 (H)/0.22 (M)
0.30 (H)/0.30 (M)
0.16 (H)/0.17 (M)
0.23 (H)/0.27 (M)
Note: H = Human liver microsomes, M = Mouse liver microsomes.