Synthesis and biological evaluation of Esaprazole analogues showing σ1 binding and neuroprotective properties in vitro

Article abstract of DOI:10.1016/j.bmc.2013.02.058

Esaprazole, a molecule previously acknowledged to protect against stomach and intestinal ulcers was surprisingly discovered to have neuroprotective activities and σ₁ binding in vitro. A highly diverse set of Esaprazole analogues 2-5 was prepared in order to increase blood-brain barrier penetration. The analogues showed a structure-activity relationship at the σ₁ receptor closely matching already published pharmacophores. Many of the analogues were shown to have neuroprotective properties in two assays using primary cultures of cortical neurons exposed to glutamate and hydrogen peroxide. However, no apparent SAR for these two assays could be developed. Metabolic stability of the analogues were also investigated and the structure of R¹ had a significant bearing on the ADME properties of the compound resulting in two series of compounds. Compounds in which R ¹ was a H or acyl group had good metabolic stability in RLM but poor BBB penetration, whereas compounds where R¹ was a cyclo- or bicyclo-alkyl group had poor metabolic stability but good BBB penetration.

Full text of DOI:10.1016/j.bmc.2013.02.058

Bioorganic & Medicinal Chemistry 21 (2013) 3334–3347
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and biological evaluation of Esaprazole analogues showing
r1 binding and neuroprotective properties in vitro
⇑
Nicholas M. Kelly , Anja Wellejus, Heidi Elbrønd-Bek, Morten Sloth Weidner, Signe Humle Jørgensen
Nensius Research A/S, Symbion Science Park, Fruebjergvej 3, DK-2100 Copenhagen, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Esaprazole, a molecule previously acknowledged to protect against stomach and intestinal ulcers was
surprisingly discovered to have neuroprotective activities and r1 binding in vitro. A highly diverse set
of Esaprazole analogues 2–5 was prepared in order to increase blood–brain barrier penetration. The ana-
logues showed a structure–activity relationship at the r1 receptor closely matching already published
pharmacophores. Many of the analogues were shown to have neuroprotective properties in two assays
using primary cultures of cortical neurons exposed to glutamate and hydrogen peroxide. However, no
apparent SAR for these two assays could be developed. Metabolic stability of the analogues were also
investigated and the structure of R1 had a significant bearing on the ADME properties of the compound
resulting in two series of compounds. Compounds in which R1 was a H or acyl group had good metabolic
stability in RLM but poor BBB penetration, whereas compounds where R1 was a cyclo- or bicyclo-alkyl
group had poor metabolic stability but good BBB penetration.
Received 16 December 2012
Revised 12 February 2013
Accepted 16 February 2013
Available online 2 April 2013
Keywords:
Neuroprotection
Esaprazole
Oxidative stress
Sigma r1
Primary culture of cortical neurons
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
abilities comes from the discovery of mutations in the r1 receptor
leading to motor neuron degeneration in amyotrophic lateral scle-
rosis patients.5 Therefore, we tested whether Esaprazole would
show neuroprotective properties in experiments using primary
culture of cortical neurons. Compounds with such properties
would have a great potential in the treatment of a wide variety
of diseases such as cerebral ischemia, multiple sclerosis or Alzhei-
mer’s disease. Calculations indicated that Esaprazole is a relatively
polar compound (clogD at pH 7.4 is À1.256) and it was expected
that the ability to penetrate the BBB would not be ideal (logBB
À0.037). Thus analogues of Esaprazole were prepared8 with the
intention of creating a SAR around the initial hit and to develop
analogues with improved logBB values.
Loss of neurons due to neurodegenerative diseases or brain
ischemia is significant causes of death or long-term disability in
populations throughout the world. Although significant effort has
been made in finding treatments for neurodegenerative diseases
and to improve patient prospects after an ischemic attack, the re-
sult of most clinical trials has been disappointing. Herein we pres-
ent in vitro work which potentially could be the first step in finding
a compound with neuroprotective effects, high blood–brain barrier
(BBB) penetration and metabolic stability.
Esaprazole 1 (Fig. 1), also known as hexaprazole, was developed
in the 1980s as a drug for the treatment of gastric and duodenal ul-
cers.1 It was shown to have antiulcer, antisecretary and cytoprotec-
tive activities.2 Esaprazole completed phase II clinical trials with
only few minor side effects being reported, but was shown to be
less effective than Cimetidine and Ranitidine at healing ulcers.1,3
The cytoprotective effect of Esaprazole was shown by the inhi-
bition of the necrotizing action of gastric mucosa damaging agents2
although no specific mechanism of action was disclosed. We spec-
ulated whether Esaprazole had more general cytoprotective effects
and discovered through testing in receptor binding assays that
2. Chemistry
A range of Esaprazole like analogues were prepared based
around structure 2 (Table 1) along with the bicyclic, fused and
spiro piperazine analogues 3, 4 and 5, respectively (Fig. 2). Reaction
of acyl halide 6 with an amine (2.1 equiv) at ca. 15 °C in acetonitrile
or DCM gave after workup amide 7 in 75–96% yield and generally
>95% purity. In cases where the amine was expensive, equal equiv-
alents of amine, acyl halide and Cs2CO3 were used in acetonitrile at
5 °C to RT to give amide 7 in 60–99% yield.9 Amide 7 was then re-
acted with a variety of amines to give a broad range of Esaprazole
analogues 2–5. For example, reaction of amide 7 with N-alkyl or
acyl (homo) piperazines in ethanol at reflux gave the desired
aminoalkamide 2–3 (R1 – H) in generally >70% yield, although
Esaprazole was a weak
r1 ligand. High affinity r1 binders has pre-
viously been shown to have neuroprotective capabilities in animal
studies.4 Further evidence for r1 mediated neuroprotective
⇑
Corresponding author. Tel.: +45 30775824.
E-mail address: nk.nensius@gmail.com (N.M. Kelly).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.02.058

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3335
yields were lower in cases of steric hinderance (e.g., R3 – H or R4/
3. Results and discussion
5 – H). If R4 = R5 = alkyl then the yield was only 14–22%, and mod-
ification of the reaction conditions resulted in the preparation of
the Favorskii-like rearranged product.10
3.1. Binding properties of Esaprazole
In order to minimise dialkylation of the piperazine moiety, the
reaction of amide 7 with piperazine or homopiperazine was carried
out in the presence of aqueous HCl11 to afford aminoalkamide 2
(R1 = H) or 5 in 38–94% yield. Bicyclic aminoalkamides 4 were
formed from the reaction of amide 7 with Boc-(1S,4S)-(+)-2,5-
diazabicyclo[2.2.1]heptane under basic conditions in 30–75% fol-
lowed by deprotection with HCl to give amine 6. In both cases
the deprotection step proceeded in very poor yield.
Whilst it is known that Esaprazole 1 does not bind to the hista-
mine H2-receptor,2 to our knowledge no other receptor binding
data has been published and therefore Esaprazole was screened
for activity in a wide range of assays, including both receptor-,
enzyme- and ion channel assays (a list of assays is given in Supple-
mentary data). Esaprazole was active at only three targets; the r1
receptor (IC50 27
lM) and the muscarinic M3 and M5 receptors
(IC50 87 and 71 M, respectively) in radioligand binding studies.
l
A representative selection of nine analogues was tested at the
muscarinic M1, M2, M3, M4, and M5 and the sigma r1 and r2
receptors for radioligand binding. Six compounds did not show
HN
O
binding at any muscarinic receptor below 100
ers only showed activity at two receptors between 50 and 80
(Table 2). In comparison, most of the analogues tested showed
binding below 100 M at the r1 receptor and eight of the com-
lM whilst two oth-
N
N
H
lM
1
l
pounds bound with a 10-fold higher affinity to the r1 receptor
Figure 1. Structure of Esaprazole 1.
Table 1
Esaprazole-like aminoalkamide analogues 2
R1
R2
R3
R4
N
O
N
m
R6
N
n
R5 R7
2
Compound structure
R1
R2
R3
R4
R5
R6
R7
m
n
1
2a
2b
H
H
H
H
H
H
Me
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Et
H
H
H
H
H
H
n-Bu
H
H
H
H
Me
Me
H
H
H
Me
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
Me
H
Cyclohexyl
i-Pr
Cyclohexyl
Cyclohexyl
Cyclopentyl
–CH2CH(Me)Et
–CH(Et)2
n-Bu
Cyclohexyl
n-Bu
Cyclohexyl
–CH(Et)2
Cyclohexyl
Cyclopentyl
Cyclohexyl
H
H
H
H
Me
H
H
n-Pr
H
n-Pr
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
1
1
2
1
1
1
2
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
2
2
2
2
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
Me
Cyclopentyl
Et
EtC(O)
H
H
Cyclopropyl-C(O)
H
H
EtC(O)
EtC(O)
H
Cyclopentyl
H
( )-2c
( )-2d
( )-2e
2f
trans-2g
2h
2i
( )-2j
2k
2l
2m
2n
( )-2o
( )-2p
2q
2r
2s
2t
2u
2v
( )-2w
2x
2y
2z
–(CH2)2-cyclooctane
Cyclopentyl
–(CH2)2-tBu
exo-Norbornan-2-yl
(1S,2S,3S,5R)-Pinan-3-yl
i-Pr
H
H
H
H
Cyclohexyl
Cyclohexyl-C(O)
Cyclohexyl
EtC(O)
i-Pr
Me
CH2CH2tBu
(1S,2S,3S,5R)-Pinan-3-yl
Cyclopentyl
(R)-3,3-Dimethylbutan-2-yl
–(CH2)3-cyclopentane
n-Pentyl
Cycloheptyl
Cyclohexyl
–(CH2)2-cyclooctane
H
Ac
H
H
Me
Me
2aa
2ab
( )-2ac
Cyclopentyl
Cyclohexyl-C(O)
H
m
N
O
HN
O
N
HN
O
N
N
N
N
H
N
n
H
Figure 2. Novel bicyclic analogues 3, 4 and 5.

3336
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
Table 2
structurally different compounds (e.g., 2n, (R)-3a and 5) were all
highly active indicated that a large, flexible hydrophobic pocket
is available for binding.
Radioligand binding results of muscarinic and sigma receptors for Esaprazole and
selected analogues
Introducing small alkyl groups at R4 and/or R5 in structure 2,
such as dimethyl, ethyl or n-butyl, did not significantly alter the
binding of the analogues regardless of whether R1 was a proton
(such as for or compounds ( )-2c, 2n and 2ab) or a cyclopentyl
group (compare 1 with ( )-2j). However, the size of the groups
R6 and R7 in structure 2 did have a role in how well the analogues
Compound
IC50
(
l
M)
Ratio
r2 M1
IC50
(
l
M)
r1
r2
r1
:
M2
M3
M4
M5
71
63
Esaprazole 1 27.2
2a
2b
2k
2l
2n
2t
2ab
(R)-3a
(S)-3a
460
17:1
—
>100 >100
>100 >100
87 >100
83 >100
>100 >1000
3.25 41
>100 1130
>100 >1000
0.43
8.41
0.50
0.39
2.25
13:1
—
—
3.4:1
13:1
66:1
28:1
23:1
102 >100 >100 >100 >100
>100 >100 >100 >100 >100
>100 >100 >100 >100 >100
>100 >100 >100 >100 >100
>100 >100 >100 >100 >100
>100 >100 >100 >100 >100
bound to the
r
1 receptor. In general, if R6 was smaller than a cyclo-
1.45
104
33
11
51
hexyl group and R7 was a also very small (e.g., compounds 2a or
2m), then all activity was lost. However, if R6 was a slightly larger
group, such as n-pentyl (2z) or 3,3-dimethylbutyl (2q), then similar
activity to the cyclohexyl group was obtained. Bridged groups,
such as pinanyl (2s) or bicyclo[2.2.1]heptyl (2r) either did not give
any benefit or made the binding worse, respectively. Larger and/or
longer groups, such as cyclooctylethyl or cyclopentylpropyl, did
not significantly improve binding (compare 2ac–2u and 2y–2b).
56 >100 >100 >100
50 93 57 73
74
26
compared to the
r2 receptor. Due to this and the well documented
connection between r1 binding and neuroprotective abilities,12
only the binding to the r1 receptor was investigated further.
3.3. r1 Receptor pharmacophore
3.2. Binding to the r1 receptor
Glennon,15 Laggner,16 Pricl17 and Wünsch14 have all published
A great difference in binding affinities to the r1 receptor was
observed among the analogues and therefore we investigated
whether a correlation between binding to the r1 receptor and
the structure of the Esaprazole analogues existed. About 80 com-
pounds were prepared using the chemistry described in Scheme 1,
and a representative sample is presented in Table 3. Examination
of the data showed improved binding to the r1 receptor when R1
was an alkyl or cycloalkyl group and that increasing the size and
therefore lipophilicity generally increased the binding. For exam-
ple, increasing the size of R1 from proton to Me to cyclopentyl in-
creased the binding by 10- and 60-fold, respectively (compare
compounds 1, 2b and 2n, respectively). However, compounds of
structure 2 in which R1 was an acyl group, such as acetyl or propi-
onyl (e.g., ( )-2e, 2k, 2l and ( )-2w), generally did not bind to the
pharmacophore models for the
r1 receptor. Since both the Laggner
and Wünsch have been compared to the Glennon model and were
found to be similar, we decided to evaluate our results based on
the Glennon pharmacophore. This pharmacophore for the r1
receptor has an amine binding site flanked by a primary hydropho-
bic region, 6–10 Å away, and a secondary hydrophobic region 2.5–
3.9 Å away (Fig. 3). To obtain affinity in the sub nanomolar range,
the optimal distance between the amine binding site and the pri-
mary hydrophobic region was about five carbons whereas it was
about three carbons to the secondary hydrophobic region. Apart
from the chain length having a large effect on compound potencies,
Glennon also showed that the
r1 affinity decreased when the chain
linking the hydrophobic region to the amine site was functional-
ised and not a hydrocarbon linker. Since the compounds presented
here have an amide functionality in the linker, this may explain
why the majority of compounds shown in Table 3 are significantly
less active.
r1 receptor at less than 100 lM. If the length of the linker n be-
tween the amide and amine group was two though (e.g., com-
pounds 2h and 2aa) or the acyl group was large, such as a
cyclohexylcarbonyl group (e.g., 2u and ( )-2ac), then activity was
restored back to the level of Esaprazole 1.
However, even though the activities of our ligands and those
presented by Glennon are not of the same order of magnitude,
our results still fit the general pharmacophore model proposed
by Glennon. In order for our compounds to show good binding,
R6 needs to be a large hydrophobic group so it fits into one hydro-
phobic region, whilst R1 needs to be an alkyl or cycloalkyl group to
fit into the other hydrophobic region. All of the ligands 1–5 pre-
pared contain a piperazine group and provided that R1 is not an
acyl group, either of the two nitrogen atoms N-1 or N-4 could bind
to the amine site. This means that binding can theoretically occur
in three different binding modes. If the amine binding site binds to
N-1, then R6 will fit into the primary hyrophobic region and R1 will
fit into the secondary hydrophobic region (binding mode 1).
Increasing the size of R1 and/or R6 will improve the binding in
the respective hydrophobic regions and thus will result in more ac-
tive compounds. Alternatively the amine site can bind to N-4 and
either R6 or R1 will fit into the primary hyrophobic region (binding
Compounds based around structure 3, in which the piperazine
moeity was replaced by a bicyclic moeity, bound at the
r1 receptor
at around 0.3 M, ca. 100 times better than Esaprazole 1. Interest-
l
ingly, whilst the size of the bicyclic ring in 3 did not significantly
alter the binding, the (R)-enantiomer of 3a was about 5-fold more
active than the (S)-enantiomer. Since the bicyclic analogue was sig-
nificantly more active, we decided to investigate whether other
substituents on the piperazine moeity would also be beneficial.
Introducing a small substituent, such as 2-methyl or a trans-2,5-di-
methyl substitution onto the piperazine ring (compounds ( )-2d
and trans-2g, respectively), did not significantly improve binding.
However, it was found that spirocycle 5 was one of the most active
compounds prepared, binding to the r1 receptor at 0.2 lM. Other
spirocyclic r1 ligands are known but they are structurally
dissimilar to the ligands presented here.13,14 The fact that three
O
O
R4
R4
X
X
R6
i or ii
iii, iv or v
2a-2ac, 3a-b, 4a-b, 5
X
N
n
n
R5
6, X=Br, Cl
R5 R7
7, X=Br, Cl
Scheme 1. Reagents and conditions: (i) Amine, MeCN or DCM, 15 °C, 1–2 h; (ii) amine, Cs2CO3, MeCN, 5 °C to rt, 1 h; (iii) N-alkyl or acyl (homo)piperazine, Na2CO3, EtOH,
reflux, 2–18 h; (iv) (homo)piperazine, HCl, EtOH, reflux, 3–18 h; (v) Boc-(1S,4S)-(+)-2,5-diazabicyclo[2.2.1]heptane, Na2CO3, EtOH, reflux, 4 h, then HCl, EtOH.

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3337
Table 3
Percent reduction in cell death in primary culture of cortical neurons after glutamate or peroxide insult
Compound number
Structure
r1 IC50
(
lM)
Percent reduction in cell deatha
Glutamate intoxication
Concentration of test compound
Peroxide intoxication
Concentration of test compound
1
l
M
10 100
l
M
lM
1
lM
10
lM
100 lM
HN
O
1
N
27.2
>100
3.25
6
10⁄
27⁄⁄⁄
25⁄⁄⁄
19⁄⁄⁄
23⁄⁄
14⁄⁄⁄
3
33⁄⁄⁄
17⁄⁄⁄
13⁄
44⁄⁄⁄
23⁄⁄⁄
21⁄⁄⁄
N
H
HN
O
2a
2b
N
11⁄
16⁄⁄
12⁄⁄
N
H
O
N
8⁄
N
N
H
N
O
( )-2c
( )-2d
0.59
1.2
28⁄⁄⁄
39⁄⁄⁄
39⁄⁄⁄
3
6
12⁄
20⁄⁄⁄
N
N
H
N
O
9⁄⁄⁄
14⁄⁄⁄
19⁄
b
b
N
N
O
N
O
2e
2f
>100
15.2
14
11
9
21⁄⁄
2
0
11
12
28⁄⁄
N
N
H
O
HN
HN
15
21⁄⁄
33⁄⁄⁄
N
N
H
O
N
N
trans-2g
7.60
8.83
16.3
11.5
>100
11⁄⁄⁄
15⁄⁄⁄
21⁄⁄⁄
8
4
11
41⁄⁄⁄
O
N
N
H
2h
2
8
21⁄⁄⁄
12⁄
27⁄⁄⁄
N
O
HN
HN
O
N
20⁄⁄⁄
2
9
25⁄⁄⁄
N
2i
À2
3
O
N
N
H
c
c
c
c
c
c
( )-2j
O
N
O
2k
2
7
25⁄⁄⁄
0
4
17⁄
N
N
H
(continued on next page)

3338
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
Table 3 (continued)
Compound number
Structure
r1 IC50
(lM)
Percent reduction in cell deatha
Glutamate intoxication
Concentration of test compound
Peroxide intoxication
Concentration of test compound
1
l
M
10
lM
100
14
lM
1
3
lM
10
3
lM
100 lM
O
N
O
2l
>100
>100
0.43
3
3
24⁄⁄⁄
16⁄
10
N
N
H
HN
HN
O
2m
2n
À1
7
4
29⁄⁄⁄
À3
À1
N
N
N
O
3
9
9
11
N
O
N
H
b
b
( )-2o
( )-2p
7.24
NA
6
4
1
9
14⁄⁄
N
N
H
O
HN
HN
À1
9⁄⁄
5
7
12
N
N
H
O
O
c
c
c
N
N
2q
2r
18.8
86.8
3
2
4
2
8
2
N
H
HN
HN
c
c
c
c
c
c
N
H
O
2s
2t
15.8
8.41
2
2
0
À9
N
N
H
N
O
7⁄
9⁄⁄
6
10
18
N
N
H
O
N
O
2u
21.1
0.18
>100
4
8
3
5
13⁄⁄
5
5
2
N
N
H
N
N
O
O
b
c
c
c
2v
À3
4
N
N
N
H
O
( )-2w
8⁄⁄⁄
1
3
12
N
H
O
N
c
c
c
c
c
c
2x
2y
11.8
0.93
0
2
1
2
5
0
N
N
H
O
N
N
N
H

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3339
Table 3 (continued)
Compound number
Structure
r1 IC50
(lM)
Percent reduction in cell deatha
Glutamate intoxication Peroxide intoxication
Concentration of test compound Concentration of test compound
1
l
M
10
lM
100
lM
1
lM
10
lM
100
lM
O
HN
c
c
c
2z
15.2
18
1
5
16
N
N
O
2aa
0
6
12⁄⁄
3
16⁄⁄
35⁄⁄⁄
N
N
H
O
N
N
O
c
c
c
c
c
c
2ab
0.50
11.6
N
N
H
O
N
O
c
c
c
c
c
c
( )-2ac
N
N
H
H
N
N
O
O
c
c
c
c
c
c
(R)-3a
(S)-3a
0.39
2.25
N
N
H
H
N
c
c
c
c
c
c
N
H
N
O
( )-3b
0.27
1
3
16⁄⁄⁄
8
10
27⁄⁄⁄
N
N
H
HN
O
4a
4b
N
N
NA
NA
2
5
2
6
9
7
6
13⁄
13
30⁄⁄⁄
N
H
O
10⁄
45⁄⁄⁄
N
H
HN
HN
b
b
5
0.21
4
6
3
À1
O
N
N
Threshold for statistical significance: ⁄P <0.05, ⁄⁄P <0.01, ⁄⁄⁄P <0.001.
a
b
c
Compound cytotoxic at 100
nd – Not determined.
lM.
modes 2 and 3, respectively). However, for binding modes 2 and 3
this forces either an amine or an amide group into the secondary
hydrophobic region which would be electronically disfavoured. It
is therefore clear that binding mode 1 best fits the data and this
conclusion also fits the data presented by Glennon who showed,
by removing either the N-1 or N-4 nitrogen of a piperazine ana-
logue, that it was the N-1 nitrogen which interacted with the
amine binding site.
Since the amine group is in the middle of the molecule, there are
two possible binding modes such that R6 can be either located to-
wards the primary hydrophobic region or towards the secondary
hydrophobic region. In both cases though, an amide group is forced
into the secondary hydrophobic region and R6 (for binding mode 2)
or R1 (for binding mode 3) has to be quite large to be able to reach
the primary hydrophobic region. Since having an amide functional-
ity in a hydrophobic region is unlikely to be favourable, it is there-
fore not surprising that compounds where R1 is an acyl group
either bind poorly or not at all.
In the cases when R1 is an acyl group, there is only 1 basic nitro-
gen, N-4, which is able to interact with the amine binding site.

3340
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
did not show any significant activity. Investigating this further re-
2.5-3.9 Å
R1
6-10 Å
O
vealed that most of the other compounds which had a substituent
at R4 and/or R5 also reduced cell death in a weak, but significant
manor (( )-2p, 2t, ( )-2w) at 100 lM in the glutamate assay. In
N
R6
addition, most compounds in which both R6 and R7 were an alkyl
substituent were also significantly active (( )-2d, trans-2g, 2i,
2m). However, having a substituent at R4, R5 and/or R7 was not a
requirement for a compound to have activity since there were
compounds lacking both structural features which were also sig-
nificantly activity (2b, 2k). No single class of substituent at R1 or
R6 was required for activity. For example, R1 could be a proton, acyl
or alkyl group resulting in both active (1, 2b, 2k) and inactive com-
pounds (2q, 2x). Likewise R6 could be a cyclohexyl or an acyclic
group again resulting in both active (1, 2a) and inactive analogues
(2x, 2y). One structural feature which did not reduce cell death was
a bridged ring system at R6 (2r, 2s, 2v).
1
N
Binding mode 1
Binding mode 2
Binding mode 3
N
R7
4
Secondary
Primary
hydrophobic
region
hydrophobic
region
O
N
R6
R1
1
N
4
N
R7
O
R6
N
R1
1
N
N
R7
4
As for the glutamate assay, it was not possible to generate a SAR
from the peroxide assay data (Table 3). However, it was possible to
find structural features which tended to prevent cell death in the
assay. The three most active compounds in the peroxide assay (1,
2f, 4b) were those compounds where R6 was cyclohexyl-like and
all other R groups were a proton. The chain length, n, and ring size,
m, did not have any effect on the protective effects. Constraining
the piperazine ring as a diazabicyclo[2.2.1]heptyl moiety highly
significantly reduced cell death, with 4a and 4b reducing cell death
Amine
binding
site
Figure 3.
ligands.
r
1 Binding model proposed by Glennon15 when applied to Esaprazole like
3.4. Activity in primary culture of cortical neurons
by 30% and 45% at 100 lM, respectively. Whilst compounds in
which R6 was cyclohexyl and all other R groups were a proton
showed significant neuroprotective effects, these structural fea-
tures were not a prerequisite for significant activity. For example,
significant activity could be obtained if R1 was either an acyl group
(2h, 2aa) or some type of alkyl group (2b, ( )-2d, ( )-3b). However,
the vast majority of analogues tested in which R1 was a proton
showed some neuroprotective effects in the peroxide assay. Other
structural features appear to be less important for imparting activ-
ity, and as such there were analogues showing neuroprotective ef-
fects when R6 was a cyclic or non-cyclic alkyl group and/or when
R6 was a proton or an alkyl group (1, 2f, 2i).
Prior to in vitro testing in primary culture of cortical neurons
the cytotoxicity of compounds 1–5 were evaluated. It was found
that none of the compounds showed any cytotoxicity at 10
and only 4 out of the 38 compounds tested were cytotoxic even
at 100 M. The compounds were then tested for their ability to
lM
l
prevent cell death in primary cultures of cortical neurons when ex-
posed to glutamate or hydrogen peroxide. In these experiments,
primary cultures of cortical neurons consisting of neuronal and
glial cells were grown for 10 days. The test compound was added
and after 30 min glutamate or peroxide concentrations known to
induce cell death in approximately 50% of the cultured cells was
added. After the insult, the cells were washed and further incu-
bated with the test compound alone. After 24 h cell death was
measured by analyzing the cellular release of LDH.
Esaprazole 1 prevented cell death in a dose dependent manner
against both glutamate and peroxide induced cell death, with a
27% and 44% decrease in cell death, respectively (Table 3). These
results were confirmed by showing a corresponding increase in
viable cell counts against both glutamate and peroxide induced in-
sults (data not shown). Many of the analogues prepared also pre-
vented cell death in the glutamate or peroxide assay. Overall
As previously mentioned, defining a SAR for both the glutamate
and peroxide assays was not possible. This is probably in part due
to assay variability although it may also be that different com-
pounds possess different neuroprotective mechanisms of action.
For example, both 4a and 4b were significantly active in the perox-
ide assay and yet showed very little activity in the glutamate assay.
It is therefore possible that these two compounds possess abilities
directly aimed at the peroxide assay (e.g., antioxidant or scaveng-
ing effects) compared to the glutamate assay and that
r1 binding is
not the only important neuroprotective mode of action for this
class of compound.
there was no clear correlation between
r1 binding and prevention
of cell death in either the glutamate or the peroxide assay. Whilst
3.5. Metabolic stability and blood–brain barrier penetration
the most active analogue in the glutamate assay (( )-2c) was active
at the r1 receptor at around 1
activity in the glutamate assay (2a, 2k or 2m) did not have any sig-
nificant 1 binding. Conversely there were also compounds which
did not protect against cell death in the glutamate assay but either
bound (2n) or did not bind (2r) to the 1 receptor. For the peroxide
assay, weak r1 binding analogues (1, 2f, trans-2g and 2aa) tended
to show the prevention of cell death. However, the weak r1 bind-
l
M, other analogues with similar
Calculations indicated that Esaprazole is a relatively polar com-
pound and it was expected that the ability to penetrate the blood–
brain barrier would not be ideal (Table 4). Therefore the compound
library was designed such that the majority of compounds pre-
pared would be more lipophilic (higher clogD) and would have
better blood–brain barrier penetration ability (higher clogBB).
Two different calculation models were used indicating that the
compounds being prepared had a clogD value of À2.05 to 6.06
and clogBB value of À0.06 to 1.67.7
r
r
ing analogue 2u was not neuroprotective and both strong
r1 bind-
ing analogue and non-binding analogues were active (( )-2c, 2a)
and inactive (2n, ( )-2w).
The results of the peroxide and glutamate assays were also
examined to find possible correlations with compound structure
(Table 3). The most active compound in the glutamate assay was
A number of analogues were also tested in vitro for metabolic
stability, blood–brain barrier penetration and whether they were
P-gp substrates (Table 4). In general the type of substituent present
on R1 dominated the blood–brain barrier potential and efflux ratio,
whilst other substitutions had a smaller effect. All of the analogues
tested in which R1 was a proton, regardless of any other structural
( )-2c with 39% more neurons surviving at both 10 and 100
lM.
The analogue of ( )-2c lacking the ethyl group in the linker (2n)

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3341
Table 4
Blood–brain barrier penetration and metabolic stability in rat liver microsomes
Compound number
PappA–B
PappB–A
Efflux ratio
Half-life in RLMa (min)
clogBB PAb
clogBB ACBc
clogD PAb
clogD ACDc
1
2a
2b
( )-2c
( )-2d
2e
2f
trans-2g
2h
2i
( )-2j
2k
2l
2m
2n
( )-2o
( )-2p
2q
2r
2s
0.61
0.38
3.1
5.0
0.45
28
8.2
1.2
8.8
8.8
>60
>60
22
À0.03
À0.06
0.25
1
0.4
0.1
0.29
0.02
0.39
1.14
0.24
0.47
0.36
0.83
0.16
0.41
1.14
0.45
0.49
0.12
0.69
1.45
0.32
0.36
0.18
0.87
0.78
1.2
À1.25
À1.89
À0.56
2.34
0.53
0.49
À2.05
0.43
2.22
0
0.85
0.71
1.16
À1.89
1.93
1.1
À0.02
À1.19
0.08
3.14
0.74
1.81
À0.67
1.16
1.08
0.27
2.12
1.81
1.93
À0.81
2.12
2.77
À0.79
0.22
À0.16
1.36
2.4
3.69
4.07
1.71
0.86
1.12
À0.73
1.46
3.01
6
5.7
50
7.0
2.6
39
15
21
d
d
d
d
d
d
d
d
d
d
d
d
0.05
0.28
0.61
0.2
0.51
0.13
0.23
À0.03
0.8
0.57
0.07
0.01
À0.02
0.4
0.59
0.74
1.34
0.13
0.55
0.6
0.08
0.53
0.95
1.38
0.39
0.39
0.56
0.03
0.1
0.69
49
71
>60
d
d
d
d
d
d
d
d
2.8
6.7
0.28
43
46
3.8
16
6.8
14
>60
>60
>60
16
44
2.7
9.3
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
À1.89
À0.88
À0.75
0.6
2t
2u
2v
( )-2w
2x
8.0
12
56
47
54
50
41
45
8.1
31
7.0
3.9
7.1
16
24
8.9
40
4.9
21
1.44
2.75
3.19
0.57
0.02
0.75
À1.1
1.97
2.42
4.52
1.06
1.06
1.74
À1.45
À0.31
1.72
7.6
3.1
1.7
5.1
0.2
0.2
3.1
>60
3.2
1.8
3.4
>60
1.48
0.4
0.82
0.85
0.19
0.35
1.03
1.67
0.33
0.33
0.56
0.18
À0.02
0.68
2y
2z
2aa
2ab
( )-2ac
(R)-3a
(S)-3a
( )-3b
4a
137
17
44
2.5
1.7
d
d
d
d
12
13
35
43
3.0
3.3
8.2
4.2
1.1
1.1
1.69
À0.8
À1.65
0.94
14
40
2.9
18
d
d
d
d
4b
5
0.31
1.1
0.41
57
1.3
52
45
17
0.46
a
b
c
Half life in rat liver microsomes.
Using the algorithm by Pharma Algorithms.
Using the algorithm by Advanced Chemical Development.
nd – Not determined.
d
feature, had a low brain penetration potential (PappA–B from 0.2 to
1.1, e.g., 1, 2z, 4b, 5). In addition, the three analogues containing
the longer linker (i.e., n = 2, 2h, 2aa, 4b) also had PappA–B from
0.2 to 0.7. However, for acyl piperazines 2h and 2aa this was be-
cause they were strong P-gp substrates, with efflux ratios of 71
and 137, respectively, rather than poor penetration. It therefore ap-
pears that the structural combination of when R1 is an acyl group
and n = 2 results in a compound which interacts very strongly with
the P-gp receptor resulting in the high efflux ratios.
Alkyl or acyl piperazines generally had moderate brain penetra-
tion potential (PappA–B from 2.0 to 12, e.g., 2b, 2l, 2u) and were
moderate P-gp substrates (efflux ratio 3.9–24). Whilst the substitu-
tion of R1 was the dominating feature, the substituent at R4 and R5
also had some effect. If R5 was a cyclohexyl group it had a higher
PappA–B and lower efflux ratio than if R5 was a cyclopentyl (com-
pare 2l vs ( )-2w) or acyclic substituent (compare 2k vs 2l). How-
ever, the compounds with the best brain penetration potential
were those where R1 was a cycloalkyl group (entries 2n, 2t) or
especially when R1 cyclised with the piperazine moiety to form a
pyrido- or pyrrolo[1,2-a]pyrazine. These latter compounds had a
PappA–B around 12 and an efflux of around three (entries (R)-3a,
(S)-3a and ( )-3b). It is interesting to note that whilst there was
a difference in the r1 binding for the two enantiomers (R)-3a
and (S)-3a, the PappA–B and efflux ratio were effectively the same.
We compared the clogBB from both prediction algorithms to
the measured PappA–B results (the prediction program does not
take into account efflux mechanisms when calculating clogBB)
and found that there was a rough correlation between the pre-
dicted and measured results (R2 = 0.42 for both algorithms). Whilst
the program accepts additional training data to improve the pre-
diction, we did not investigate this feature.
The metabolic stability of the analogues was tested using rat li-
ver microsomes. Again, the type of substituent present on R1
tended to dominate the metabolic stability of the analogue,
although not to the same extent as for the blood–brain barrier pen-
etration. Ventura et al.18,19 have shown in human volunteers that
about 85% of Esaprazole is excreted in urine unchanged after
24 h and that the metabolites detected in rat, dog and human are
N-formulation, N-actylation and 4-hydroxylation of the cyclohexyl
ring. Therefore it could be expected that structurally similar ana-
logues will also have a long half life in rat liver microsomes.
This proved to be the case, with the majority of analogues in
which R1 was either a proton or an acyl group having long half lives
(T1/2 >60 min) in rat liver microsomes. Interestingly this set of ana-
logues also included a number of structural features which may
have been expected to be hydroxylated, such as when R4 or R6
was a short alkyl chain (2k, ( )-2w), or N-demethylated, such as
when R7 was a methyl group (2m). However, if R6 was a longer al-
kyl chain, then the compound had a short half life (2z).
All of the analogues when R1 was an alkyl, cycloalkyl or bicy-
cloalkyl substituent had poorer half lives than when R1 was either
a proton or an acyl group. If R1 was a cyclopentyl or cyclohexyl
group (2v, 2ab and ( )-2c) then the half life was below 10 min. If
R1 contained an alkyl chain and R6 was a cyclohexyl group then

3342
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
the half life was around 20 min, but this dropped to only a couple
of minutes when R6 was an alkyl-like chain (compare entries 2b,
2y). The bicyclic analogues 3a and 3b showed a moderate to poor
half life, with the [6.5] ring system (3b) being slightly more stable
than the [5.5] ring system and the (R)-enantiomer of 3a being
slightly more stable than the (S)-enantiomer.
solvent signal (d 7.26 for CDCl3 and d 3.31 for CD3OD) was used
as reference.
Gas–liquid chromatography (GC) was carried out on an Agilent
6890N instrument with FID detection at 250 °C, H2 flow 25 ml/min,
air flow 450 ml/min. The column used was
a Varian VF5
30 m  0.32 mm  0.25 m. The initial temperature was set at
l
We also investigated how much compound penetrated the
brain using an in vivo rat model. A single dose of 1600 mg/kg po
of 1 or 2l was given to rats (n = 2) and blood samples were taken
just prior to dosing, and then after 30, 60 and 120 min at which
point the rats were sacrificed and the brains perfused. For 1 the ra-
tio of compound in brain vs. in plasma at 120 min was 0.25 ( 0.04)
whereas for 2l the ratio was 0.33 ( 0.01). As expected from the
in vitro data, 2l had a higher brain-plasma ratio than 1, although
the difference was not as large as expected. No side effects from
this single high dose of 1 or 2l were observed.
40 °C for 2 min, and the temperature was ramped at 15 °C/min to
250 °C, where it was held for 4 min, for a total run time of 22 min.
Compound purification was carried out using a TeledyneIsco
Combiflash Companion. The crude sample was dissolved in DCM,
celite 545 (ca. 10 g per gram of crude sample) was added and the sol-
vent removed in vacuo to give a dry powder. The dry powder was
loaded into an empty reservoir and gently compacted. The silica col-
umn was wetted with A buffer (Et2O), then the solvent flow was
passed through the load and silica column using an increasing ratio
of B buffer (0.1 M NH3 in 1:1 MeOH/Et2O) to elute off the product.
Esaprazole 1 was purchased from Chess GmbH, Mannheim
(cat# 2455) and N-isopropyl-1-piperazineacetamide 2a was pur-
chased from Acros Organics (cat# 205610250).
4. Conclusion
We were interested in developing an orally available drug can-
didate with good blood–brain barrier penetration as a neuropro-
tective compound for the treatment of neurodegenerative
diseases. A structurally broad range of analogues to Esaperazole
1 were prepared and were tested for activity at the r1 receptor.
A SAR of the analogues was found and which matched previously
published r1 receptor pharmacophores. However, there did not
appear to be any SAR when examining the compounds ability to
prevent cell death in a primary culture of cortical neurons against
a glutamate or peroxide insult.
Examination of the ADME properties of the analogues showed
that there were two series of compounds. One series in which R1
was H or acyl, resulted in compounds with good metabolic stability
(T1/2 >60 min) but with poor blood–brain barrier penetration (low
PappA–B and/or moderate to high P-gp efflux). The other series
was where R1 was a cyclo- or bicyclo-alkyl group giving com-
pounds with poor metabolic stability (T1/2 <20 min) but with good
blood–brain barrier penetration (high PappA–B and low to moder-
ate P-gp efflux). In conclusion we have prepared a broad series of
Esaprazole analogues with many compounds showing neuropro-
tective properties in two primary culture of cortical neuron assays.
Esaprazole 1 has since been sent for in vivo testing.
5.1.2. General procedure 1 (GP1)
The amine (1.0 equiv) in ethanol (1 mL per mmol amine) was
added to the alkyl halide (1.1 equiv) and sodium carbonate
(3 equiv) in ethanol (2 mL per mmol amine). The reaction was
heated to reflux for 2–3 h then water (5 mL per mmol amine)
and aqueous sodium carbonate (1 M, 5 mL per mmol amine) were
added. The product was extracted with ethyl acetate (3 times,
20 mL per mmol amine) and the organic layers combined.
DowexÒ 50WX2 hydrogen form 100–200 mesh (4 g per mmol
amine) was washed with methanol (10 mL per gram resin). The
reaction mixture was loaded onto the ion exchange resin using
gravity filtration and the resin was washed with methanol
(10 mL per gram resin). The product was eluted off the resin with
ca. 1 M NH3 in methanol (10 mL per gram resin) and the solution
was concentrate to give the product.
5.1.3. General procedure 2 (GP2)
The amine (2.0 equiv) was added to aqueous hydrochloric acid
(2.0 equiv, 1.2 M) then the alkyl halide (1.0 equiv) in ethanol
(2 mL per mmol amine) was added and the reaction was heated
to reflux. The reaction was cooled to room temperature and aque-
ous sodium carbonate (1 M, 10 mL per mmol alkyl halide) was
added. The product was extracted with ethyl acetate (3 times,
20 mL per mmol alkyl halide) and the combined organic layers
were dried over MgSO4, filtered and concentrated. The crude mate-
rial was purified using the Combiflash (A buffer Et2O, B buffer
0.1 M NH3 in 1:1 MeOH/Et2O) and the relevant fractions combined
and concentrated to give the product.
5. Experimental section
5.1. Chemistry
5.1.1. General methods
Purity was determined by HPLC and is stated for each example.
HPLC was performed on a Waters Alliance e2695 Separations Mod-
ule with detection using a Waters 2998 photodiode array detector
with the wavelength set to 220 nm. The column used was a Waters
5.1.4. Example A.6.1: general procedure 3 (GP3)
The amine (1.0 equiv) in ethanol (1 mL per mmol amine) was
added to the alkyl halide (1.1 equiv) and sodium carbonate (3 equiv)
in ethanol (2 mL per mmol amine). The reaction was heated to reflux
then water (5 mL per mmol amine) and aqueous sodium carbonate
(1 M, 5 mL per mmol amine) were added. The product was extracted
with ethyl acetate (3 times, 10 mL per mmol amine) and the com-
bined organic layers were dried over MgSO4, filtered and concen-
trated. The crude material was purified using the Combiflash (A
buffer Et2O, B buffer 0.1 M NH3 in 1:1 MeOH/Et2O) and the relevant
fractions combined and concentrated to give the product.
XBridge C18, 100 Â 4.6 mm, 3.5
lm, 135 Å which was kept in a
heater unit at 30 °C. Buffer A was 0.10% trifluoroacetic acid in
95% water, 5% acetonitrile. Buffer B was 0.08% trifluoroacetic acid
in 10% water, 90% acetonitrile. The flow was 1.0 mL/min. Following
an initial hold of 0.5 min at 0% buffer B, the compounds were
eluted off using a gradient of 0–40% buffer B over 10.0 min, a gra-
dient of 40–100% buffer B over 2.5 min followed by a hold at 100%
buffer B for 2.5 min, and the column was re-equilibrated at 0% buf-
fer B for a total run time of 20 min.
High resolution mass spectra (HRMS) were performed on a
Micromass Q-TOF 1.5 mass spectrometer, using polyethylene gly-
col as an external standard. 1H NMR spectra were run on a Varian
Mercury 300 MHz, or a JEOL ECX300 (MHz) or ECX400 (MHz) spec-
trometer using either CDCl3 or CD3OD as solvent. The residual
5.1.5. General procedure 4 (GP4)
A halide intermediate (1.0 equiv), amine (0.9–10 equiv) and
optionally a base (2.7 equiv) and halogenating agent (0.5 equiv)
were mixed together in industrial methylated spirits (15 mL per
g halide) and heated in a microwave at 150 °C for 1 h, repeating

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3343
if neccesary. Upon completion of the reaction, the mixture was
passed through a phase seperation cartridge, concentrated and
purified.
methanol (40 mL). The reaction mixture was loaded onto the ion
exchange resin using gravity filtration and the resin was washed
with methanol (40 mL). The product was eluted off the resin with
ca. 1 M NH3 in methanol (40 mL) and the solution was concentrate
to give 2f (184 mg, 81%) as a colourless oil; 1H NMR (300 MHz,
CD3OD) d 3.69 (tt, J = 9.0, 4.9 Hz, 1H), 3.19 (s, 2H), 2.92 (m, 4H),
2.77 (m, 4H), 1.81 (m, 2H), 1.58 (m, 2H), 1.44 (m, 2H), 0.91 (t,
J = 7.4 Hz, 6H); HPLC Rt 5.67 min (92.8% pure); HRMS (ESI-TOF)
calcd for C12H25N3NaO (M+Na) 250.1895, found 250.1877.
5.1.6. N-Cyclohexyl-2-(4-methyl-1,4-diazepan-1-yl)acetamide
(2b)
1-Methyl-1,4-diazepane (30 mmol, 3.43 g), 2-chloro-N-cyclo-
hexylacetamide (33 mmol, 5.81 g) and sodium carbonate
(90 mmol, 9.60 g) were used and the reaction was heated to
50 °C for 15 h according to GP3. The crude material purified using
a 80 g silica column eluting 0% buffer B for 2 min, then 0–100% buf-
fer B over 21 min and 100% buffer B for 4 min to give 2b (1.59 g,
21%) as a colourless oil; 1H NMR (300 MHz, CD3OD) d 3.69 (m,
1H), 3.12 (s, 2H) 2.73 (m, 8H), 2.39 (s, 3H), 1.81 (m, 6H), 1.66 (m,
1H), 1.33 (m, 5H); HPLC Rt 6.68 min (99.8% pure); HRMS (ESI-
TOF) calcd for C14H28N3O (M+H) 254.2232, found 254.2235.
5.1.11. N-Butyl-2-(trans-2,5-dimethylpiperazin-1-yl)-N-
propylacetamide (trans-2g)
trans-2,5-Dimethylpiperazine (6 mmol, 686 mg), hydrochloric
acid (6 mmol, 5 mL) and N-butyl-2-chloro-N-propylacetamide
(2.0 mmol, 385 mg) were used according to GP2. The crude mate-
rial purified using a 12 g silica column eluting 0% buffer B for
1 min, then 0–100% buffer B over 11 min and 100% buffer B for
3 min to give trans-2g (135 mg, 50%) as a yellow oil; 1H NMR
(300 MHz, CD3OD) d 3.74 (dd, J = 14.7, 1.3 Hz, 1H), 3.5–3.2 (m,
4H), 2.99 (m, 3H), 2.86 (dd, J = 11.8, 2.9 Hz, 1H), 2.58 (m, 2H),
2.15 (ddd, J = 11.9, 10.6, 1.4 Hz, 1H), 1.60 (m, 4H), 1.36 (m, 2H),
1.09 (t, J = 6.2 Hz, 6H), 0.95 (m, 6H); HPLC Rt 9.04 min (99.7% pure);
HRMS (ESI-TOF) calcd for C15H32N3O (M+H) 270.2545, found
270.2525.
5.1.7. N-Cyclohexyl-2-(4-cyclopentylpiperazin-1-yl)butanamide
(( )-2c)
1-Cyclopentylpiperazine (4 mmol, 619 mg), 2-bromo-N-cyclo-
hexylbutanamide (4.4 mmol, 1.10 g) and sodium carbonate
(12 mmol, 1.33 g) were used and the reaction was heated to reflux
for 18 h according to GP3. The crude material purified using a 12 g
silica column eluting 0% buffer B for 1 min, then 0–100% buffer B
over 11 min and 100% buffer B for 3 min to give ( )-2c (953 mg,
74%) as a white powder; 1H NMR (300 MHz, CD3OD) d 3.68 (tt,
J = 10.5, 3.9 Hz, 1H), 2.76 (dd, J = 8.2, 6.0 Hz, 1H), 2.53 (m, 8H),
2.53 (m, 1H), 1.87 (m, 4H), 1.68 (m, 9H), 1.30 (m, 7H), 0.89 (t,
J = 7.4 Hz, 3H); HPLC Rt 9.63 min (94.4% pure); HRMS (ESI-TOF)
calcd for C19H35N3NaO (M+Na) 344.2678, found 344.2686.
5.1.12. N-Cyclohexyl-3-(4-(cyclopropanecarbonyl)piperazin-1-
yl)propanamide (2h)
Cyclopropyl(piperazin-1-yl)methanone hydrochloride (3 mmol,
573 mg), 3-chloro-N-cyclohexylpropanamide (3.3 mmol, 638 mg)
and sodium carbonate (12 mmol, 1.27 g) were used and the reac-
tion was heated to reflux for 18 h according to GP3. The crude
material purified using a 12 g silica column eluting 0% buffer B
for 1 min, then 0–50% buffer B over 11 min and 50% buffer B for
10 min to give 2h (132 mg, 14%) as a white solid; 1H NMR
(300 MHz, CD3OD) d 3.77 (m, 2H), 3.68 (m, 1H), 3.62 (m, 2H),
2.69 (t, J = 7.1 Hz, 2H), 2.55 (m, 2H), 2.47 (m, 2H), 2.38 (t,
J = 7.1 Hz, 2H), 1.96 (m, 1H), 1.85 (m, 2H), 1.76 (m, 2H), 1.64 (m,
1H), 1.30 (m, 5H), 0.85 (m, 4H); HPLC Rt 8.63 min (99.7%); HRMS
(ESI-TOF) calcd for C17H30N3O2 (M+H) 308.2338, found 208.2331.
5.1.8. N-Cyclopentyl-2-(4-ethyl-3-methylpiperazin-1-yl)-N-
methylacetamide (( )-2d)
1-Ethyl-2-methylpiperazine (1 mmol, 128 mg), 2-chloro-N-
cyclopentyl-N-methylacetamide (1.1 mmol, 193 mg) and sodium
carbonate (3 mmol, 317 mg) were used according to GP1 to give
( )-2d as a yellow oil; 1H NMR (300 MHz, CD3OD, mixture of two
rotamers in ca. 2:3 ratio) d 4.84 (m, 2/5 H), 4.50 (m, 3/5 H), 3.25
(s, 6/5H), 3.19 (s, 4/5H), 2.94 (s, 6/5H), 2.88 (m, 4H), 2.79 (s, 9/
5H), 2.53 (m, 1H), 2.38 (m, 3H), 2.03 (m, 1H), 1.88 (m, 1H), 1.75
(m, 3H), 1.63 (m, 4H), 1.07 (t, J = 7.2 Hz, 3H), 1.07 (d, J = 6.3 Hz,
3H); HPLC Rt 6.68 min (98.6% pure); HRMS (ESI-TOF) calcd for
C15H29N3NaO (M+Na) 290.2208, found 290.2205.
5.1.13. N-Butyl-2-(piperazin-1-yl)-N-propylacetamide (2i)
Piperazine (4.5 mmol, 388 mg), hydrochloric acid (4.5 mmol,
3.75 mL) and N-butyl-2-chloro-N-propylacetamide (2.25 mmol,
431 mg) were used and the reaction was heated to reflux for 5 h
according to GP2. The crude material purified using a 12 g silica
column eluting 0% buffer B for 1 min, then 0–100% buffer B over
11 min and 100% buffer B for 5 min to give 2i (480 mg, 88%) as a
pale yellow oil; 1H NMR (300 MHz, CD3OD) d 3.40–3.20 (m, 6H),
2.97 (m, 4H), 2.58 (m, 4H), 1.59 (m, 4H), 1.35 (m, 2H), 0.99 (t,
J = 7.3 Hz, 3H), 0.95 (t, J = 7.2 Hz, 3H); HPLC Rt 8.48 min (98.9%
pure); HRMS (ESI-TOF) calcd for C13H28N3O (M+H) 242.2232, found
242.2239.
5.1.9. N-(2-Methylbutyl)-2-(4-propionylpiperazin-1-
yl)acetamide (( )-2e)
1-(Piperazin-1-yl)propan-1-one (1 mmol, 144 mg), 2-chloro-N-
(2-methylbutyl)acetamide (1.1 mmol, 182 mg) and sodium
carbonate (3 mmol, 316 mg) were used according to GP1 to give
( )-2d (258 mg, 96%) as a colourless oil; 1H NMR (300 MHz,
CD3OD) d 3.61 (m, 4H), 3.18 (dd, J = 13.2, 6.2 Hz, 1H), 3.05 (dd,
J = 13.1, 7.4 Hz, 1H), 3.05 (s, 2H), 2.52 (m, 4H), 2.41 (q, J = 7.5 Hz,
2H), 1.60 (td, J = 13.4, 6.7 Hz, 1H), 1.42 (m, 1H), 1.20 (m, 1H),
1.11 (dd, J = 8.6, 6.3 Hz, 3H), 0.94 (t, J = 7.3 Hz, 3H), 0.91 (d,
J = 6.7 Hz, 3H); HPLC Rt 7.91 min (99.6% pure); HRMS (ESI-TOF)
calcd for C14H27N3NaO2 (M+Na) 292.1995, found 292.1900.
5.1.14. N-Cyclohexyl-2-(piperazin-1-yl)hexanamide (( )-2j)
Piperazine (12 mmol, 1.04 g), hydrochloric acid (12 mmol,
10 mL) and 2-bromo-N-cyclohexylhexanamide (4 mmol, 1.10 g)
were used and the reaction was heated to reflux for 2 days accord-
ing to GP2. Piperazine (6 mmol, 500 mg) was added and the reac-
tion was heated to reflux for one additional day. The crude
material purified using a 12 g silica column eluting 0% buffer B
for 1 min, then 0–100% buffer B over 11 min and 100% buffer B
for 7 min to give ( )-2j (550 mg, 49%) as a white powder; 1H
NMR (300 MHz, CD3OD) d 3.67 (m, 1H), 2.86 (m, 4H), 2.82 (m,
1H), 2.59 (m, 4H), 1.97–1.50 (m, 7H), 1.50–1.08 (m, 9H), 0.92 (t,
J = 7.1 Hz, 3H) HPLC Rt 10.42 min (94.7% pure); HRMS (ESI-TOF)
calcd for C16H32N3O (M+H) 282.2545, found 282.2566.
5.1.10. 2-(1,4-Diazepan-1-yl)-N-(pentan-3-yl)acetamide (2f)
1,4-Diazepane (2 mmol, 199 mg) was added to aqueous hydro-
chloric acid (2 mmol, 1.67 mL, 1.2 M) and stirred for ca. 5 min.
2-Chloro-N-(pentan-3-yl)acetamide (1.0 mmol, 169 mg) in ethanol
(2 mL) was added and the reaction was heated to reflux for 3 h. The
reaction was cooled to room temperature and aqueous sodium car-
bonate (1 M, 10 mL) was added. The product was extracted with
ethyl acetate (3 Â 20 mL) and the organic layers combined. Dow-
exÒ 50WX2 hydrogen form 100–200 mesh (4 g) was washed with

3344
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
5.1.15. N-(Pentan-3-yl)-2-(4-propionylpiperazin-1-yl)acetamide
(2k)
5.1.20. N-Cyclopentyl-2-(1,4-diazepan-1-yl)propanamide (( )-
2p)
1-(Piperazin-1-yl)propan-1-one (4 mmol, 566 mg), 2-chloro-N-
(pentan-3-yl)acetamide (4.4 mmol, 726 mg) and sodium carbonate
(12 mmol, 1.27 g) were used and the reaction was heated to reflux
for 18 h according to GP3. The crude material purified using a 12 g
silica column eluting 0% buffer B for 1 min, then 0–100% buffer B
over 11 min and 100% buffer B for 3 min to give 2k (1.04 g, 100%)
as a colourless oil; 1H NMR (300 MHz, CD3OD) d 3.70 (m, 1H),
3.61 (m, 4H), 3.07 (s, 2H) 2.53 (m, 4H), 2.41 (q, J = 7.5 Hz, 2H),
1.59 (m, 2H), 1.43 (m, 2H), 1.12 (t, J = 7.5 Hz, 3H), 0.91 (t,
J = 7.4 Hz, 6H); HPLC Rt 7.49 min (98.8% pure); HRMS (ESI-TOF)
calcd for C14H28N3O2 (M+H) 270.2182, found 270.2198.
2-Chloro-N-cyclopentylpropanamide (5.7 mmol, 1 g) and homo-
piperazine (57 mmol, 5.7 g) were used according to GP4, purifying
by Biotage SP4 (2 cycles) to give ( )-2p (668 mg, 49%) as a yellow
oil; 1H NMR (300 MHz, CDCl3) d 7.28 (m, 1H), 4.18 (dp, J = 7.9,
6.7 Hz, 1H), 3.23 (q, J = 6.9 Hz, 1H), 2.90 (m, 4H), 2.64 (m, 4H), 1.94
(m, 2H), 1.67 (m, 4H), 1.36 (m, 2H), 1.20 (d, J = 7.0 Hz, 3H); HPLC Rt
5.75 min (91.7% pure); GC: Rt 12.93 min; HRMS (ESI-TOF) calcd for
C13H25N3NaO (M+Na) 262.1890, found 262.1853.
5.1.21. N-(3,3-Dimethylbutyl)-2-(piperazin-1-yl)acetamide (2q)
2-Chloro-N-(3,3-dimethylbutyl)acetamide (5.63 mmol, 1 g) and
piperazine (56.3 mmol, 4.85 g) were used according to GP4, purify-
ing by Biotage SP4 to give 2q (1.09 g 94%) as a pale yellow solid; 1H
NMR (300 MHz, CDCl3) d 7.05 (s, 1H), 3.26 (m, 2H), 2.95 (s, 2H),
2.90 (m, 4H), 2.56 (s, 2H), 2.49 (dd, J = 6.2, 3.5 Hz, 4H), 1.39 (m,
2H), 0.91 (s, 9H); HPLC Rt 8.31 min (99.4% pure); GC: Rt
11.69 min; HRMS (ESI-TOF) calcd for C12H26N3O (M+H) 228.2070,
found 228.2045.
5.1.16. N-Cyclohexyl-2-(4-propionylpiperazin-1-yl)acetamide
(2l)
1-(Piperazin-1-yl)propan-1-one (4 mmol, 569 mg), 2-chloro-N-
cyclohexylacetamide (4.4 mmol, 777 mg) and sodium carbonate
(12 mmol, 1.29 g) were used and the reaction was heated to reflux
for 18 h according to GP3. The crude material purified using a 12 g
silica column eluting 0% buffer B for 1 min, then 0–100% buffer B
over 11 min and 100% buffer B for 3 min to giv 2l (1.08 g, 96%) as
a colourless oil; 1H NMR (300 MHz, CD3OD) d 3.70 (m, 1H), 3.61
(m, 4H), 3.03 (s, 2H) 2.51 (m, 4H), 2.41 (q, J = 7.5 Hz, 2H), 1.81
(m, 4H), 1.65 (m, 1H), 1.31 (m, 5H), 1.12 (t, J = 7.5 Hz, 3H), HPLC
Rt 8.00 min (99.7%); HRMS (ESI-TOF) calcd for C15H28N3O2 (M+H)
282.2182, found 282.2218.
5.1.22. N-((1S,2S,4R)-rel-Bicyclo[2.2.1]heptan-2-yl)-2-
(piperazin-1-yl)acetamide (2r)
N-((1S,2S,4R)-rel-Bicyclo[2.2.1]heptan-2-yl)-2-chloroacetamide
(5.35 mmol, 1 g) and piperazine (53.5 mmol, 4.6 g) were used
according to GP4, purifying by Biotage SP4 to give 2r (991 mg,
78%) as a yellow oil which solidified on standing; 1H NMR
(300 MHz, CDCl3) d 7.01 (s, 1H), 3.73 (td, J = 8.0, 3.5 Hz, 1H), 2.90
(m, 6H), 2.48 (m, 4H), 2.28 (m, 1H), 2.17 (m, 1H), 1.79 (ddd,
J = 13.1, 8.0, 2.2 Hz, 1H), 1.47 (m, 2H), 1.21 (m, 5H); HPLC Rt
7.28 min (97.0% pure); GC: Rt 13.20 min; HRMS (ESI-TOF) calcd
for C13H24N3O (M+H) 238.1919, found 238.1909.
5.1.17. N-Cyclopentyl-N-methyl-2-(piperazin-1-yl)acetamide
(2m)
Piperazine (12 mmol, 1.04 g), hydrochloric acid (12 mmol,
10 mL) and 2-chloro-N-cyclopentyl-N-methylacetamide (4 mmol,
701 mg) were used and the reaction was heated to reflux for
18 h according to GP2. The crude material purified using a 12 g sil-
ica column eluting 0% buffer B for 1 min, then 0–100% buffer B over
11 min and 100% buffer B for 7 min to give 2m (601 mg, 67%) as a
pale yellow oil; 1H NMR (300 MHz, CD3OD, mixture of two rota-
mers in ca. 1:1 ratio) d 4.83 (m, 1/2H), 4.50 (m, 1/2H), 3.28 (s,
1H), 3.23 (s, 1H), 2.94 (s, 3/2H), 2.92 (m, 4H), 2.79 (s, 3/2H), 2.55
(m, 4H), 1.91 (m, 1H), 1.77 (m, 3H), 1.63 (m, 4H); HPLC Rt
5.92 min (97.5% pure); HRMS (ESI-TOF) calcd for C12H24N3O
(M+H) 226.1919, found 226.1913.
5.1.23. 2-(Piperazin-1-yl)-N-((1S,2S,3S,5R)-2,6,6-
trimethylbicyclo[3.1.1]heptan-3-yl)acetamide (2s)
2-Chloro-N-((1S,2S,3S,5R)-2,6,6-trimethylbicyclo[3.1.1]heptan-
3-yl)acetamide (4.35 mmol, 1 g) and piperazine (43.5 mmol,
3.75 g) were used according to GP4, purifying by Biotage SP4 to
give 2s (1.04 g, 86%) as a viscous yellow oil; 1H NMR (300 MHz,
CDCl3) d 7.01 (s, 1H), 4.25 (tt, J = 9.6, 6.4 Hz, 1H), 2.97 (s, 2H),
2.91 (m, 4H), 2.48 (m, 6H), 2.04 (m, 1H), 1.95 (m, 1H), 1.80 (m,
1H), 1.75 (m, 1H), 1.48 (ddd, J = 13.9, 6.0, 2.4 Hz, 1H), 1.22 (s,
3H), 1.10 (d, J = 7.1 Hz, 3H), 1.04 (s, 3H), 0.86 (d, J = 9.8 Hz, 1H);
HPLC Rt 10.95 min (98.0% pure); GC: Rt 14.28 min; HRMS (ESI-
TOF) calcd for C16H30N3O (M+H) 280.2389, found 280.2390.
5.1.18. N-Cyclohexyl-2-(4-cyclopentylpiperazin-1-yl)acetamide
(2n)
1-Cyclopentylpiperazine (4 mmol, 622 mg), 2-chloro-N-cyclo-
hexylacetamide (4.4 mmol, 771 mg) and sodium carbonate
(12 mmol, 1.26 g) were used and the reaction was heated to reflux
for 4 h according to GP3. The crude material purified using a 12 g
silica column eluting 0% buffer B for 1 min, then 0–100% buffer B
over 11 min and 100% buffer B for 3 min to give 2n (1.01 g, 86%)
as a white solid; 1H NMR (300 MHz, CD3OD) d 3.68 (m, 1H), 2.99
(s, 2H) 2.57 (m, 8H), 2.51 (m, 1H), 1.91 (m, 4H), 1.66 (m, 6H),
1.31 (m, 7H); HPLC Rt 8.84 min (99.1% pure); HRMS (ESI-TOF) calcd
for C17H32N3O (M+H) 294.2545, found 294.2548.
5.1.24. 2-(4-Cyclohexylpiperazin-1-yl)-N-isopropyl-2-
methylpropanamide (2t)
Propan-2-amine (210 mmol, 18 ml) was dissolved in DCM
(21 mL) and added drop wise to a stirred solution of 2-bromo-2-
methylpropanoyl bromide (100 mmol, 23.0 g) in DCM (100 mL)
at ca. 10 °C. The reaction was stirred at room temperature for
30 min then concentrated HCl (10 mL) and water (40 mL) were
added. The layers were separated and the organic phase was
washed with saturated aqueous sodium hydrogen carbonate
(30 mL), dried over magnesium sulphate, filtered and concentrated
in vacuo to give 2-bromo-N-isopropyl-2-methylpropanamide
(20.2 g, 97%) as a white solid; 1H NMR (300 MHz, CD3OD) d 6.50
(s, 1H), 4.01 (dhept, J 7.8, 6.6 Hz, 1H), 1.95 (s, 6H), 1.20 (d,
J = 6.5 Hz, 6H); HPLC Rt 12.06 min.
1-Cyclohexylpiperazine (20 mmol, 3.37 g), 2-bromo-N-isopro-
pyl-2-methylpropanamide (48 mmol, 9.92 g) and sodium carbonate
(60 mmol, 6.35 g) were used and the reaction was heated to reflux
for 16 h according to GP3. The crude material purified using a 80 g
silica column eluting 0% buffer B for 2 min, then 0–100% buffer B
over 23 min and 100% buffer B for 10 min. The product containing
5.1.19. N-(2-Cyclooctylethyl)-2-(piperazin-1-yl)propanamide
(( )-2o)
2-Chloro-N-(2-cyclooctylethyl)propanamide (4.07 mmol, 1 g),
piperazine (40.7 mmol, 3.5 g) and NaI (4.07 mmol, 610 mg) were
used according to GP4, purifying by Biotage SP4, to give ( )-2o
(1.13 g, 94%) as a pale yellow oil; 1H NMR (300 MHz, CDCl3) d 7.16
(t, J = 6.1 Hz, 1H), 3.24 (dtd, J = 11.0, 7.0, 6.5, 3.8 Hz, 2H), 2.93 (m,
5H), 2.46 (m, 5H), 1.51 (m, 15H), 1.27 (m, 2H), 1.20 (d, J = 7.0 Hz,
3H); HPLC Rt 12.36 min (96.6% pure); GC: Rt 15.95 min; HRMS
(ESI-TOF) calcd for C17H34N3O (M+H) 296.2702, found 296.2726.

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3345
fractions were repurified using a 80 g silica column eluting 0% buffer
B for 2 min, then 0–50% buffer B over 21 min to give 2t (1.31 g, 22%)
as a very pale yellow oil; 1H NMR (300 MHz, CD3OD) d 3.94 (hept,
J = 6.7 Hz, 1H), 2.66 (s, 4H), 2.55 (m, 4H), 2.25 (m, 1H), 1.97 (m,
2H), 1.83 (m, 2H), 1.66 (m, 1H), 1.26 (m, 5H), 1.17 (s, 6H), 1.15 (d,
J = 6.7 Hz, 6H); HPLC Rt 8.53 min (95.8% pure); HRMS (ESI-TOF) calcd
for C17H34N3O (M+H) 296.2702, found 296.2709.
(s, 2H), 2.79 (m, 7H), 2.50 (s, 3H), 1.94 (m, 2H), 1.74 (m, 3H),
1.54 (m, 6H), 1.32 (m, 2H), 1.05 (m, 2H); HPLC Rt 10.33 min
(88.4% pure); GC: Rt 14.71 min; HRMS (ESI-TOF) calcd for
C16H32N3O (M+H) 282.2545, found 282.2559.
5.1.30. 2-(1,4-Diazepan-1-yl)-N-methyl-N-pentylacetamide (2z)
Homopiperazine (12 mmol, 1.19 g), hydrochloric acid (12 mmol,
10 mL) and 2-chloro-N-methyl-N-pentylacetamide (4 mmol,
709 mg) were used and the reaction was heated to reflux for 3.5 h
according to GP2. The crude material purified using a 12 g silica col-
umn eluting 0% buffer B for 1 min, then 0-100% buffer B over 11 min
and 100% buffer B for 7 min to give 2z (367 mg, 38%) as a pale yellow
oil; 1H NMR (300 MHz, CD3OD, mixture of 2 rotamers in ca. 1:1 ratio)
d3.41 (s, 1H), 3.39 (s, 1H), 3.36 (m, 2H), 3.08 (s, 3/2H), 2.91 (m, 4H),
2.90 (s, 3/2H), 2.76 (dt, J = 6.8, 2.5 Hz, 4H), 1.82 (m, 2H), 1.60 (m, 2H),
1.34 (m, 4H), 0.95 (t, J = 7.1 Hz, 3/2 H), 0.92 (t, J = 7.1 Hz, 3/2 H); HPLC
Rt 7.53 min (99.7% pure); HRMS (ESI-TOF) calcd for C13H28N3O
(M+H) 242.2232, found 242.2218.
5.1.25. 2-(4-(Cyclohexanecarbonyl)piperazin-1-yl)-N-(3,3-
dimethylbutyl)acetamide (2u)
2-Chloro-N-(3,3-dimethylbutyl)acetamide (5.63 mmol, 1 g), 1-
(cyclohexylcarbonyl)-piperazine (5.12 mmol, 1 g) and Na2CO3
(15.2 mmol, 1.61 g) were used according to GP4, purifying by Bio-
tage SP4 to give 2u (1.56 g, 88%) as a pale yellow solid; 1H NMR
(300 MHz, CDCl3) d 6.97 (s, 1H), 3.55 (m, 4H), 3.30 (m, 3H), 3.01
(m, 2H), 2.45 (m, 5H), 1.79 (m, 2H), 1.69 (m, 4H), 1.46 (m, 5H),
1.35–1.25 (m, 3H), 0.94 (s, 9H); HPLC Rt 12.43 min (91.4% pure);
GC: Rt 18.14 min; HRMS (ESI-TOF) calcd for C19H35N3NaO2
(M+Na) 360.2621, found 360.2584.
5.1.31. 3-(4-Acetyl-1,4-diazepan-1-yl)-N-
5.1.26. 2-(4-Cyclohexylpiperazin-1-yl)-N-((1S,2S,3S,5R)-2,6,6-
trimethylbicyclo[3.1.1]heptan-3-yl)acetamide (2v)
cycloheptylpropanamide (2aa)
1-(1,4-Diazepan-1-yl)ethanone (3 mmol, 426 mg), 3-bromo-N-
cycloheptylpropanamide (4.5 mmol, 1.12 g), sodium carbonate
(6 mmol, 636 mg) and potassium iodide (3.3 mmol, 549 mg) were
used and the reaction was heated to reflux for 17 h according to
GP3. The crude material purified using a 12 g silica column eluting
0% buffer B for 1 min, then 0–100% buffer B over 11 min and 100%
2-Chloro-N-((1S,2S,3S,5R)-2,6,6-trimethylbicyclo[3.1.1]heptan-
3-yl)acetamide
(4.35 mmol,
1 g),
4-cyclohexylpiperazine
(3.96 mmol, 666 mg) and Na2CO3 (11.74 mmol, 1.25 g) were used
according to GP4, purifying by Biotage SP4 to give 2v (1.15 g, 87%)
as a yellow orange oil; 1H NMR (300 MHz, CDCl3) d 6.98 (s, 1H),
4.25 (tt, J = 9.6, 6.5 Hz, 1H), 3.48 (s, 1H), 2.98 (s, 2H), 2.59 (m, 8H),
2.42 (dtd, J = 9.7, 6.2, 2.3 Hz, 1H), 2.25 (m, 1H), 1.83 (m, 7H), 1.63
(dt, J = 13.2, 3.0 Hz, 1H), 1.48 (ddd, J = 14.0, 6.0, 2.5 Hz, 1H), 1.21
(m, 7H), 1.10 (d, J = 7.2 Hz, 3H), 1.04 (s, 3H), 0.86 (d, J = 9.7 Hz, 1H);
HPLC Rt 13.55 min (97.5% pure); GC: Rt 18.40 min; HRMS (ESI-TOF)
calcd for C22H40N3O (M+H) 362.3171, found 362.3187.
buffer B for 3 min to give 2aa (661 mg, 71%) as a colourless oil; 1
H
NMR (300 MHz, CD3OD) d3.83 (m, 1H), 3.59 (m, 4H), 2.81 (m, 3H),
2.69 (m, 3H), 2.33 (td, J = 6.9, 1.8 Hz, 2H), 2.11 (s, 3H), 1.86 (m, 4H),
1.56 (m, 10H), HPLC Rt 9.18 min (99.7% pure); HRMS (ESI-TOF)
calcd for C17H32N3O2 (M+H) 310.2495, found 310.2499.
5.1.32. N-Cyclohexyl-2-(4-cyclopentylpiperazin-1-yl)-2-
methylpropanamide (2ab)
5.1.27. N-Cyclopentyl-2-(4-propionylpiperazin-1-
yl)propanamide (( )-2w)
1-Cyclopentylpiperazine (40 mmol, 6.17 g), 2-bromo-N-cyclo-
hexyl-2-methylpropanamide (40 mmol, 9.93 g), sodium carbonate
(80 mmol, 7.3 g) and potassium iodide (44 mmol, 8.48 g) were
used and the reaction was heated to reflux for 2 days according
to GP3. The crude material purified using a 80 g silica column elut-
ing 0% buffer B for 2 min, then 0–100% buffer B over 21 min and
100% buffer B for 9 min to give 2ab (1.78 g, 14%) as a colourless
oil; 1H NMR (300 MHz, CD3OD) d 3.64 (m, 1H), 2.96–2.17 (m,
9H), 1.91 (m, 2H), 1.68 (m, 9H), 1.34 (m, 7H), 1.18 (s, 6H); HPLC
Rt 10.47 min (98.8% pure); HRMS (ESI-TOF) calcd for C19H36N3O
(M+H) 322.2858, found 322.2845.
2-Chloro-N-cyclopentylpropanamide (5.7 mmol, 1 g), N-propio-
nylpiperazine (5.2 mmol, 736 mg) Na2CO3 (15 mmol, 1.63 g) and
NaI (2.84 mmol, 430 mg) were used according to GP4, purifying
by biotage SP4 (2 cycles), to give ( )-2w (794 mg, 54%) as a pale
yellow solid; 1H NMR (300 MHz, CDCl3) d6.99 (s, 1H), 4.17 (m,
1H), 3.60 (m, 2H), 3.45 (q, J = 4.8 Hz, 2H), 3.01 (q, J = 7.0 Hz, 1H),
2.47 (m, 4H), 2.31 (q, J = 7.5 Hz, 2H), 1.93 (m, 2H), 1.61 (m, 4H),
1.35 (m, 2H), 1.19 (d, J = 7.0 Hz, 3H), 1.11 (t, J = 7.4 Hz, 3H); HPLC
Rt 7.17 min (97.0% pure); GC Rt 15.2 min; HRMS (ESI-TOF) calcd
for C17H27N3NaO2 (M+Na) 304.2001, found 304.1994.
5.1.28. (R)-N-(3,3-Dimethylbutan-2-yl)-2-(4-isopropyl-1,4-
diazepan-1-yl)acetamide (2x)
5.1.33. 2-(4-(Cyclohexanecarbonyl)piperazin-1-yl)-N-(2-
cyclooctylethyl)propanamide (( )-2ac)
(R)-2-Chloro-N-(3,3-dimethylbutan-2-yl)acetamide
2-Chloro-N-(2-cyclooctylethyl)propanamide (4.07 mmol, 1 g),
1-cylohexylcarbonyl piperazine (3.70 mmol, 727 mg), Na2CO3
(10.99 mmol, 1.16 g) and NaI (4.07 mmol, 610 mg) were used
according to GP4, purifying by Biotage SP4, to give ( )-2ac
(1.58 g, 96%) as a yellow oil; 1H NMR 400 MHz (CDCl3) d 7.03 (s,
1H), 3.62 (m, 2H), 3.49 (m, 2H), 3.26 (tt, J = 7.3, 5.9 Hz, 2H), 3.04
(q, J = 7.0 Hz, 1H), 2.46 (m, 5H), 1.79 (m, 2H), 1.55 (m, 19H), 1.23
(m, 8H); HPLC Rt 14.93 min (94.3% pure); HRMS (ESI-TOF) calcd
for C24H44N3O2 (M+H) 406.3434, found 406.3413.
(5.63 mmol, 1 g), N-isopropyl(1,4)-diazapane (5.12 mmol, 0.73 g)
and Na2CO3 (15.2 mmol, 1.61 g) were used according to GP4, puri-
fying by Biotage SP4 to give 2x (794 mg, 55%) as a yellow oil; 1H
NMR (300 MHz, CDCl3) d 7.32 (s, 1H), 3.83 (m, 1H), 3.10 (s, 2H),
2.89 (m, 1H), 2.69 (m, 8H), 1.77 (m, 2H), 1.04 (d, J = 6.9 Hz, 3H),
0.98 (d, J = 6.9 Hz, 6H), 0.89 (s, 9H); HPLC Rt 7.69 min (95.1% pure);
GC: Rt 12.74 min; HRMS (ESI-TOF) calcd for C16H34N3O (M+H)
284.2702, found 284.2706.
5.1.29. N-(3-Cyclopentylpropyl)-2-(4-methyl-1,4-diazepan-1-
yl)acetamide (2y)
5.1.34. (R)-N-Cyclohexyl-2-(hexahydropyrrolo[1,2-a]pyrazin-
2(1H)-yl)acetamide ((R)-3a)
2-Chloro-N-(3-cyclopentylpropyl)acetamide (4.9 mmol, 1 g),
1-methylhomopiperazine (4.47 mmol, 0.56 mL) and Na2CO3
(13.23 mmol, 1.4 g) were used according to GP4, purifying by Bio-
tage SP4 to give 2y (727 mg, 58%) as an orange oil; 1H NMR
(300 MHz, CDCl3) d 7.45 (s, 1H), 3.27 (td, J = 7.1, 5.9 Hz, 2H), 3.15
(R)-Octahydropyrrolo[1,2-a]pyrazine (20 mmol, 2.52 g), 2-
chloro-N-cyclohexylacetamide (22 mmol, 3.86 g) and sodium car-
bonate (60 mmol, 6.35 g) were used and the reaction was heated
to reflux for 16 h according to GP3. The crude material purified
using a 80 g silica column eluting 0% buffer B for 2 min, then

3346
N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
0–100% buffer B over 21 min and 100% buffer B for 4 min to give
(R)-3a (3.81 g, 72%) as a white wax; 1H NMR (300 MHz, CD3OD)
d 3.68 (m, 1H), 3.08 (m, 2H), 3.06 (s, 2H) 2.97 (m, 1H), 2.81 (m,
1H), 2.43 (m, 2H), 2.29 (m, 2H), 2.11 (m, 1H), 1.83 (m, 7H), 1.64
(m, 1H), 1.32 (m, 6H); HPLC Rt 7.82 min (99.6% pure); HRMS
(ESI-TOF) calcd for C15H28N3O (M+H) 266.2232, found 266.2230.
crude material purified using a 12 g silica column eluting 0% buffer
B for 1 min, then 0–100% buffer B over 8 min and 100% buffer B for
3 min to give Boc-4b (463 mg, 33%) as a colourless oil; 1H NMR
(300 MHz, CD3OD) d 4.28 (br s, 1H), 3.64 (tt, J = 9.5, 3.2 Hz, 1H),
3.55 (br s, 1H), 3.45 (dd, J = 10.4, 4.8 Hz, 1H), 3.18 (ddd, J = 10.3,
8.0, 2.2 Hz, 1H), 2.83 (m, 3H), 2.66 (t, J = 9.4 Hz, 1H), 2.31 (t,
J = 6.9 Hz, 2H), 1.84 (m, 3H), 1.73 (m, 3H), 1.64 (m, 1H), 1.46 (s,
9H), 1.28 (m, 5H); HPLC Rt 11.84 min.
Boc-4b (463 mg) was dissolved in ethanol (5 mL) then concen-
trated HCl (5 drops) was added. The reaction was heated to 60 °C
for 18 h. Aqueous Na2CO3 (1 M, 20 mL) was added, the product
was extracted with EtOAc (3 Â 30 mL) and the organic phase dried
over MgSO4, filtered and concentrated. The crude material purified
using a 4 g silica column eluting 0% buffer B for 1 min, then 0–100%
buffer B over 11 min and 100% buffer B for 8 min to give 4b (35 mg,
11%) as a pale yellow oil; 1H NMR (300 MHz, CD3OD) d 3.63 (m,
1H), 3.52 (br s, 1H), 3.45 (br s, 1H), 3.09 (dd, J = 10.6, 1.3 Hz, 1H),
2.81 (m, 4H), 2.48 (dd, J = 10.0, 1.3 z, 1H), 2.31 (t, J = 7.1 Hz, 2H),
1.78 (m, 5H), 1.62 (m, 2H), 1.32 (m, 5H); HPLC Rt 7.06 min (99.2%
pure).
5.1.35. (S)-N-Cyclohexyl-2-(hexahydropyrrolo[1,2-a]pyrazin-
2(1H)-yl)acetamide ((S)-3a)
(S)-Octahydropyrrolo[1,2-a]pyrazine
(20 mmol,
2.53 g),
2-chloro-N-cyclohexylacetamide (22 mmol, 3.93 g) and sodium
carbonate (60 mmol, 6.41 g) were used and the reaction was
heated to reflux for 16 h according to GP3. The crude material puri-
fied using a 80 g silica column eluting 0% buffer B for 2 min, then
0–100% buffer B over 21 min and 100% buffer B for 4 min to give
(S)-3a (3.82 g, 72%) as a white wax; 1H NMR (300 MHz, CD3OD) d
3.69 (m, 1H), 3.05 (s, 2H) 3.01 (m, 3H), 2.79 (m, 1H), 2.37 (m,
2H), 2.22 (m, 2H), 2.05 (m, 1H), 1.87 (m, 5H), 1.64 (m, 1H), 1.33
(m, 6H); HPLC Rt 7.82 min (99.9% pure); HRMS (ESI-TOF) calcd
for C15H28N3O (M+H) 266.2232, found 266.2241.
5.1.36. N-Cyclohexyl-2-(dihydro-1H-pyrido[1,2-a]pyrazin-
2(6H,7H,8H,9H,9aH)-yl)acetamide (3b)
5.1.39. N-Methyl-N-pentyl-2-(1,4-diazaspiro[5.5]undecan-4-
yl)acetamide (5)
Octahydro-1H-pyrido[1,2-a]pyrazine
(1 mmol,
141 mg),
1,4-Diazaspiro[5.5]undecane dihydrochloride (1 mmol, 229 mg)
in ethanol (1 mL) was added to 2-chloro-N-methyl-N-pentylaceta-
mide (1.1 mmol, 197 mg) and sodium carbonate (5 mmol, 530 mg)
in ethanol (2 mL) and the reaction was heated to reflux for 3 h. The
reaction was cooled to room temperature then water (5 mL) and
aqueous Na2CO3 (1 M, 5 mL) was added. The product was extracted
with EtOAc (3 Â 20 mL). DowexÒ 50WX2 hydrogen form 100–200
mesh (4 g) was washed with methanol (40 mL). The reaction mix-
ture was filtered then loaded onto the ion exchange resin using
gravity filtration. The resin was washed with methanol (40 mL).
The product was eluted off the resin with ca. 1 M NH3 in methanol
(20 mL) and the solution was concentrated. The crude material was
purified by Combiflash (A buffer Et2O, B buffer 0.1 M NH3 in 1:1
MeOH/Et2O) using a 4 g silica column eluting 0% buffer B for
1 min, then 0–100% buffer B over 11 min and 100% buffer B for
2-chloro-N-cyclohexylacetamide (1.1 mmol, 194 mg) and sodium
carbonate (3 mmol, 321 mg) were used according to GP1 to give
3b (274 mg, 98%) as a cream solid; 1H NMR (300 MHz, CD3OD) d
3.69 (m, 1H), 2.98 (s, 2H), 2.83 (m, 2H), 2.72 (m, 2H), 2.37 (m, 2H),
2.12 (m, 2H), 2.02 (dd, J = 19.6, 9.1 Hz, 2H), 1.85–1.7 (m, 5H), 1.65–
1.5 (m, 4H), 1.45–1.2 (m, 7H); HPLC Rt 8.16 min (99.5% pure); HRMS
(ESI-TOF) calcd for C16H30N3O (M+H) 280.2389, found 280.2381.
5.1.37. 2-((1S,4S)-2,5-Diazabicyclo[2.2.1]heptan-2-yl)-N-
cyclohexylacetamide (4a)
(1S,4S)-tert-Butyl 2,5-diazabicyclo[2.2.1]heptane-2-carboxylate
(4 mmol, 792 mg), 2-chloro-N-cyclohexylacetamide (4.4 mmol,
776 mg) and sodium carbonate (12 mmol, 1.27 g) were used and
the reaction was heated to reflux for 4 h according to GP3. The
crude material purified using a 12 g silica column eluting 0% buffer
B for 1 min, then 0–100% buffer B over 11 min and 100% buffer B
for 3 min to give Boc-4a (1.04 g, 77%) as a colourless oil; 1H NMR
(300 MHz, CD3OD) d 4.29 (1H, br s), 3.68 (m, 1H), 3.52 (br s, 1H),
3.45 (m, 1H), 3.26 (d, J = 16.0 Hz, 1H), 3.18 (d, J = 16.0 Hz, 1H),
2.93 (dd, J = 9.6, 2.1 Hz, 1H), 2.63 (m, 1H), 1.90 (m, 2H), 1.78 (m,
4H), 1.65 (m, 1H), 1.47 (s, 9H), 1.32 (m, 5H); HPLC Rt 11.52 min
Boc-4a (1.04 g) was dissolved in ethanol (9 mL) then HCl in
ether (4.6 mL of 2.0 M solution) was added. The reaction was stir-
red at room temperature for 8 days then at 60 °C for 18 h. Aqueous
Na2CO3 (1 M, 20 mL) was added, the product was extracted with
EtOAc (3 Â 30 mL) and the organic phase dried over MgSO4, filtered
and concentrated. The crude material purified using a 4 g silica col-
umn eluting 0% buffer B for 1 min, then 0–100% buffer B over
11 min and 100% buffer B for 6 min to give 4a (17 mg, 2%) as a
white precipitate; 1H NMR (300 MHz, CD3OD) d 3.69 (m, 1H),
3.62 (br s, 1H), 3.43 (br s, 1H), 3.24 (d, J = 16.0 Hz, 1H), 3.17 (d,
J = 16.0 Hz, 1H), 3.09 (dd, J = 10.5, 1.1 Hz, 1H), 2.90 (dd, J = 10.0,
2.4 Hz, 1H), 2.82 (dd, J = 10.5, 2.4 Hz, 1H), 2.58 (dd, J = 10.0,
1.1 Hz, 1H), 1.81 (m, 5H), 1.63 (m, 2H), 1.30 (m, 5H); HPLC Rt
6.62 min (94.7% pure).
2 min to give
5 (102 mg, 35%) as a
yellow oil; 1H NMR
(300 MHz,, CD3OD, mixture of two rotamers in ca. 1:1 ratio) 3.47
(m, 1/2H), 3.36 (m, 1/2H), 3.13 (s, 3H) 3.12 (s, 2/2H), 2.90 (s, 2/
2H), 2.87 (m, 2H), 2.37 (m, 4H), 1.66 (m, 3H), 1.56 (m, 5H), 1.35
(m, 11H), 0.95 (t, J = 7.1 Hz, 3/2 H), 0.94 (t, J = 7.1 Hz, 3/2 H); HPLC
Rt 10.26 min (92.5% pure); HRMS (ESI-TOF) calcd for C17H37N3O
(M+H) 296.2702, found 296.2680.
5.2. Biological assays
In vitro ADME properties in terms of blood–brain barrier pene-
tration potential and metabolic stability of the compounds in addi-
tion to the sigma and muscarinic bindings were performed as
previously described by Elbrønd-Bek et al.8 Likewise, the methods
of the in vitro neuronal damage assays (hydrogen peroxide and
glutamate) was outlined by Wellejus et al.20 All receptor-, enzyme-
and ion channel assays were carried out by Ricerca Biosciences,
Taipei, Taiwan and the pocedures are available on Ricerca Biosci-
ences homepage.
Acknowledgments
5.1.38. 3-((1S,4S)-2,5-Diazabicyclo[2.2.1]heptan-2-yl)-N-
cyclohexylpropanamide (4b)
(1S,4S)-tert-Butyl 2,5-diazabicyclo[2.2.1]heptane-2-carboxylate
(4 mmol, 794 mg), 3-chloro-N-cyclohexylpropanamide (4.4 mmol,
837 mg) and sodium carbonate (12 mmol, 1.27 g) were used and
the reaction was heated to reflux for 4 h according to GP3. The
The authors would like to thank Paul Blaney, Mark Ruston,
Laura Montgomery and Ieuan Roberts from Peakdale Molecular
Ltd, Chapel-en-le-Frith, England for preparing some of the com-
pounds; Nina Vartiainen from Cerebricon (now Charles River),
Kuopio, Finland for carrying out the primary culture of cortical
neuron assays; Absorption Systems, Exton, USA for the BBB

N. M. Kelly et al. / Bioorg. Med. Chem. 21 (2013) 3334–3347
3347
6. Calculated using Advanced Chemistry Development ACD/Labs ADME Suite 5.0
using Ionisation/LogD module.
penetration and metabolic stability testing; Ricerca Biosciences,
Taipei, Taiwan for performing the receptor binding and enzyme
inhibition assays and Anette Andersen at the Department of Chem-
istry, Copenhagen University, Denmark for running the HRMS.
7. Calculated using Advanced Chemistry Development ACD/Labs ADME Suite 5.0
using BBB module and Pharma Algorithms algorithm. clogBB is the calculated
blood–brain partition coefficient and is the ratio of the equilibrium
concentrations of the drug in the brain and the blood.
8. Elbrønd-Bek, H.; Wellejus, A.; Kelly, N. M.; Weidner, M. S.; Jørgensen, S. H.
Neurochem. Int. 2011, 59, 821.
Supplementary data
9. Kelly, N. M.; Wellejus, A.; Jørgensen, S. H.; Weidner, M. S. PCT Int. Appl., WO
2011/000945.
10. Kelly, N. M.; Wellejus, A.; Elbrønd-Bek, H.; Weidner, M. S.; Jørgensen, S. H.
Synth. Commun. 2013, 43, 1243.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2013.02.058.
11. Corvi-Mora, C. U.S. Patent 4,123,530, 1977; Corvi-Mora, C. U.S. Patent
4,278,796, 1981.
References and notes
12. Maurice, T.; Su, T. P. Pharmacol. Ther. 2009, 124, 195.
13. Oberdorf, C.; Schmidt, T. J.; Wünsch, B. Eur. J. Med. Chem. 2010, 45,
3116.
1. Dal Monte, P. R.; Gullini, S.; Saggioro, A.; Beinat, L.; Castiglioni, C. L. Drugs Exp.
Clin. Res. 1987, XIII, 305.
14. Jasper, A.; Schepmann, D.; Lehmkuhl, K.; Vela, J. M.; Buschmann, H.; Holenz, J.;
Wünsch, B. Eur. J. Med. Chem. 2009, 44, 4306.
2. Scuri, E.; Mondani, G.; Fantini, P. L.; Dalla Valle, V.; Valsecchi, B. Boll. Chim.
Farm. 1984, 123, 425.
15. Glennon, R. A. Mini-Rev. Med. Chem. 2005, 5, 927.
16. Laggner, C.; Schieferer, C.; Fiechtner, B.; Poles, G.; Hoffmass, R. D.; Glossmann,
H.; Langer, T.; Moebius, F. F. J. Med. Chem. 2005, 48, 4754.
17. Zampieri, D.; Mamolo, M. G.; Laurini, E.; Florio, C.; Zanette, C.; Fermeglia,
M.; Posocco, P.; Paneni, M. S.; Pricl, S.; Vio, L. J. Med. Chem. 2009, 52,
5380.
3. (a) Parente, F.; Ardizzone, S.; BianchiPorro, G.; Civardi, R. Clin. Trials J. 1989, 26,
376; (b) Luminari, M.; Giannelli, C.; Manini, G.; Fltilia, F.; Gandolfi, L.; Giaccari,
S.; Guilini, S.; Missale, G.; Moettini, A. Clin. Trials J. 1989, 26, 384.
4. (a) Allahtavakoli, M.; Jarrott, B. Brain Res. Bull. 2011, 85, 219; (b) Ajmo, C. T.;
Vernon, D. O. L.; Collier, L.; Pennypacker, K. R.; Cuevas, J. Curr. Neurovasc. Res.
2006, 3, 89.
18. Ventura, P.; Schiavi, M.; Risoli, G. Farmaco Prat. 1984, 39, 190.
19. Girometta, M. A.; Loschi, L.; Ventura, P. J. Chromatogr., B: Biomed. Sci. Appl. 1986,
383, 85.
20. Wellejus, A.; Elbrønd-Bek, H.; Kelly, N. M.; Weidner, M. S.; Jørgensen, S. H.
Restor. Neurol. Neurosci. 2012, 30, 21.
5. (a) Al-Saif, A.; Al-Mohanna, F.; Bohlega, S. Ann. Neurol. 2011, 70, 913; (b) Luty,
A. A.; Kwon, J. B. J.; Dobson-Stone, C.; Loy, C. T.; Coupland, K. G.; Karlström, H.;
Sobow, T.; Tchorzewska, J.; Maruszak, A.; Barcikowska, M.; Panegyres, P. K.;
Zekanowski, C.; Brooks, W. S.; Williams, K. L.; Blair, I. P.; Mather, K. A.; Sachdev,
P. S.; Halliday, G. M.; Schofield, P. R. Ann. Neurol. 2010, 68, 639.