Synthesis, pharmacological evaluation, and σ1 receptor interaction analysis of hydroxyethyl substituted piperazines

Article abstract of DOI:10.1021/jm401707t

Starting from (S)- or (R)-aspartate, three synthetic strategies were explored to prepare hydroxyethyl substituted piperazines with different substituents at the N-atoms. σ receptor affinity was recorded using receptor material from both animal and human o

Full text of DOI:10.1021/jm401707t

Subscriber access provided by Universitaetsbibliothek | Erlangen-Nuernberg
Article
Synthesis, pharmacological evaluation and #1 receptor
interaction analysis of hydroxyethyl substituted piperazines
Frauke Weber, Steffi Brune, Katharina Korpis, Patrick J Bednarski, Erik Laurini,
Valentina Dal Col, Sabrina Pricl, Dirk Schepmann, and Bernhard Wünsch
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm401707t • Publication Date (Web): 11 Mar 2014
Downloaded from http://pubs.acs.org on March 18, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of Medicinal Chemistry
1
1
2
3
4
5
6
7
Synthesis, pharmacological evaluation and σ1 receptor interaction analysis of
hydroxyethyl substituted piperazines
8
9
Frauke Weber,a Stefanie Brune,a Katharina Korpis,b Patrick J. Bednarski,b Erik
Laurini,c Valentina Dal Col,c Sabrina Pricl,c,d Dirk Schepmann,a Bernhard Wünscha,*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a Institute of Pharmaceutical and Medicinal Chemistry, University of Münster,
Corrensstr. 48, Dꢀ48149 Münster, Germany
Tel.: +49ꢀ251ꢀ83ꢀ33311; Fax: +49ꢀ251ꢀ83ꢀ32144; Eꢀmail: wuensch@uniꢀmuenster.de
b Institute of Pharmacy, Department of Pharmaceutical and Medicinal Chemistry,
University of Greifswald, F.ꢀL.ꢀJahnꢀStraße 17, 17487 Greifswald, Germany
c Molecular Simulations Engineering (MOSE) Laboratory, Department of Engineering
and Architecture (DEA), University of Trieste, Via Valerio 6, 34127 Trieste, Italy
d National Interuniversity Consortium for Material Science and Technology (INSTM),
Research Unit MOSEꢀDEA, University of Trieste, Via Valerio 6, 32127 Trieste, Italy
Abstract
Starting from (S)ꢀ or (R)ꢀaspartate three synthetic strategies were explored to prepare
hydroxyethyl substituted piperazines with different substituents at the Nꢀatoms. σ
receptor affinity was recorded using receptor material from both animal and human
origin. σ1 affinities determined with guinea pig brain and human RPMI 8226 tumor
cell lines differed slightly but showed the same tendency. (S)ꢀ2ꢀ[4ꢀ
(Cyclohexylmethyl)ꢀ1ꢀ(naphthaleneꢀ2ꢀylmethyl)piperazinꢀ2ꢀyl]ethanol (7c) revealed
the highest affinity at human σ1 receptors (Ki = 6.8 nM). The potent σ1 receptor ligand
7c was able to inhibit selectively the growth of three human tumor cell lines with IC50
values in the low micromolar range. The reduced growth of the RPMIꢀ8226 cell line
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 36
2
1
2
3
4
5
6
7
8
9
was caused by apoptosis. The interaction of 7c with the σ1 receptor was analyzed in
detail using the 3D homology model of the σ1 receptor. The calculated free binding
energies of all hydroxyethylpiperazines nicely correlate with their recorded affinities
towards the human σ1 receptor.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords
hydroxyethyl piperazines, σ1 ligands, structure affinity relationships (SARs), cytotoxic
activity; apoptosis; 3D homology model; docking
Introduction
σ Receptors became generally recognized as a unique receptor class when it was
unequivocally demonstrated that their pharmacological profile differs from that of
opioid receptors, which they were formerly confused with.1,2 Two subtypes of σ
receptors have been identified so far, classified as σ1 and σ2 receptors. Both
subtypes differ in their distribution and in their ligand binding profile.3,4 The σ1
receptor has been cloned from different tissues like guinea pig liver5 and human
placental choriocarcinoma cells.6 Much less is known about the σ2 receptor, but new
findings hypothesize an association of the σ2 receptor with the PGRMCꢀ1 protein.7
This protein was found in liver microsomes first and could be cloned from human and
mammalian tissues. Both protein types show related intracellular distribution, similar
molecular weight and the same high affinity for [3H]ꢀprogesterone. Knockout of the
PGRMCꢀ1 gene led to decreased affinity of σ2 ligands.7,8 σ1 Receptors were shown
to be involved in the modulation of neurotransmitter systems and in the regulation of
ion channels.9ꢀ11 In addition, both subtypes are reported to regulate cancer cell
growth. Various tumor cell lines express large numbers of σ receptors, and, σ2
ACS Paragon Plus Environment

Page 3 of 36
Journal of Medicinal Chemistry
3
1
2
3
4
5
6
7
8
9
receptors are overexpressed especially in proliferating tumor cells.12 Several studies
reported on the cytotoxic effects of σ1 antagonists and σ2 agonists.13,14 However,
their impact on the cell cycle or mechanism of cell death is not clearly understood.
Compounds I – III bearing a hydroxyalkyl substituted piperazine ring represent well
established σ1 receptor ligands. For example, recent studies have shown the high σ
receptor binding of 2ꢀhydroxyethyl substituted piperazines II in contrast to the
homologous hydroxypropyl derivatives III.15 With respect to selectivity towards the σ2
subtype, the hydroxyethyl derivatives II are superior to the hydroxymethyl ligands I
showing the same σ1 affinity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Piperazine derivatives I ꢀ III with different lengths of the hydroxyalkyl side
chain.
Herein we present different synthetic strategies to afford hydroxyethylpiperazines of
general formula II with varying substitution patterns. The σ1 and σ2 affinities of all
synthesized compounds were evaluated using receptor material from animal and
human origin. The interaction of the piperazine derivatives with the σ1 receptor was
analyzed in detail performing molecular dynamics simulations of the compounds in
complex with our recently validated σ1 3D homology model. The cytotoxic activity of
the most potent σ1 ligand was investigated in a panel of human tumor cell lines. The
ability of this compound to induce apoptosis was also studied.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 36
4
1
2
3
4
5
Synthesis
6
7
8
9
The chloroacetamide 3 was prepared from (S)ꢀaspartic acid (1) in four steps. This
reaction sequence includes esterification of diacid 1 to obtain the diester HCl 2.HCl,16
followed by reductive alkylation with benzaldehyde and subsequent acylation with
chloroacetyl chloride. The resulting chloroacetamide 3 was reacted with different
primary amines to obtain the dioxopiperazines 4aꢀe in a domino reaction, consisting
of a SN2 reaction followed by intramolecular aminolysis. Reduction with LiAlH4 in
refluxing THF led to the piperazine ethanols 5aꢀe (Scheme 2). The (R)ꢀenantiomer of
5e was synthesized in the same manner using (R)ꢀaspartic acid as starting material.
All derivatives shown are single enantiomers with (S)ꢀconfiguration, unless otherwise
stated. The enantiomers are characterized with the prefix ent. The high σ1 and σ2
affinity of the cyclohexylmethyl derivatives 5e and entꢀ5e prompted us to investigate
the σ affinity of piperazines 7 with different residues at Nꢀ1. For this purpose, the
benzyl group of 5e was cleaved off by transfer hydrogenolysis and the secondary
amine 6 was reductively alkylated with bulky biphenylꢀ4ꢀcarbaldehyde or
naphthaleneꢀ1ꢀcarbaldehyde in presence of NaBH(OAc)3 to obtain 7a and 7b in 7 – 8
% overall yields (Scheme 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 5 of 36
Journal of Medicinal Chemistry
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4, 5
a
b
c
d
R1
CH3
CH2CH2CH2CH3
CH2CH2CH(CH3)2
(CH2)2C6H5
CH2C6H11
e
7
a
b
R2
4ꢀPhꢀPh
1ꢀNaph
Scheme 2. Reagents and reaction conditions: (a) (H3C)3SiCl, H3COH, rt, 16 h; (b) 1.
PHꢀCH=O, NEt3, CH2Cl2, rt, 16 h; 2. NaBH4, H3COH, 0 °C, 40 min; 3. ClCH2COCl,
NEt3, CH2Cl2, rt, 2.5 h; (c) R1ꢀNH2, NEt3, CH3CN, rt, 16 h – 3 d; (d) LiAlH4, THF,
reflux, 16 h. The enantiomer entꢀ5e was prepared in the same manner starting with
(R)ꢀaspartic acid entꢀ1. (e) HCO2NH4, Pd/C, H3COH, 3.5 h, reflux; (f) RꢀCHO,
NaBH(OAc)3, CH3CN.
Since the bulky Nꢀ1ꢀresidues appeared to be the preferred substituents for high σ1
receptor binding, derivatives of 5a and 5b were synthesized by combining methyl and
butyl residues at Nꢀ4 with naphthylmethylꢀ and biphenylylmethyl residues at Nꢀ1. In
order to obtain piperazines 11 with different substituents on both Nꢀatoms via a
shorter route, the diester.HCl 2.HCl was acylated with chloroacetamide to afford the
secondary amide 8 (Scheme 3). Ring closure of 8 with methylamine and butylamine
led to the dioxopiperazines 9a and 9c, which were reduced by LiAlH4 to give the Nꢀ1
unsubstituted piperazines 10a and 10c, respectively. The substituted piperazines
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 36
6
1
2
3
4
5
11aꢀd were prepared by reductive alkylation of 10a and 10c with naphthaleneꢀ1ꢀ
6
7
carbaldehyde, biphenylꢀ4ꢀcarbaldehyde, 4ꢀmethoxybenzaldehyde and NaBH(OAc)3.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9ꢀ11
R1
CH3
R2
a
b
c
d
1ꢀNaph
4ꢀPhꢀPh
4ꢀPhꢀPh
CH3
CH2CH2CH2CH3
CH2CH2CH2CH3 4ꢀMeOꢀPh
Scheme 3. Reagents and reaction conditions: (a) ClCH2COCl, NaHCO3, CH2Cl2, 16
h, rt; (16) (b) R1ꢀNH2, NEt3, CH3CN, 16 h, rt; (c) LiAlH4, THF, reflux, 16 h; R2ꢀCHO,
NaBH(OAc)3, CH3CN, 7 d.
The reductive alkylation of Nꢀ1 unsubstituted piperazines 6 and 10 generated low
yields of products 7 and 11, respectively. This may be due to the hydroxyethyl side
chain that sterically shields the adjacent secondary amine. We assume that the
aldehydes could react with the γꢀaminoalcohols 6 and 10 to yield 1,3ꢀoxazinanes
(N/Oꢀacetals), which are less reactive leading to undesired side products and
reduced yields of 7 and 11.
Furthermore, it was not possible to perform the ring closure of chloroacetamide 8 with
cyclohexylmethylamine to yield piperazines with the preferred Nꢀ4ꢀresidue for σ1
affinity. We assume that the chloroacetamide 8 without a substituent on Nꢀ1 adopt
another conformation which is disfavored for the cyclization with large amines like
cyclohexylmethylamine.
ACS Paragon Plus Environment

Page 7 of 36
Journal of Medicinal Chemistry
7
1
2
3
4
5
6
7
CO2CH3
CO2CH3
CO2CH3
(b)
(a)
2.HCl
CO2CH3
R2
N
R2
N
H
H
H
8
9
15aꢀd
16aꢀd
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cl
CO2CH3
CO2CH3
(c)
(d)
N
N
O
O
N
H
CO2CH3
R2
O
H
R2
17aꢀd
18aꢀd
7, 15ꢀ18
R2
a
b
c
d
4ꢀPhꢀPh
1ꢀNaph
2ꢀNaph
(e)
N
N
OH
H
6ꢀMeOꢀ2ꢀNaph
R2
7aꢀd
Scheme 4. Reagents and reaction conditions: (a) R2ꢀCH=O, NEt3, CH2Cl2, rt, 16 h;
(b) NaBH4, H3COH, 0 °C, 40 min; (c) ClCH2COCl, NEt3, CH2Cl2, rt, 2.5 h; (d)
C6H11CH2ꢀNH2, NEt3, CH3CN, rt, 16 h; (e) LiAlH4, THF, reflux, 16 h. The enantiomers
entꢀ7a and entꢀ7b were prepared in the same manner starting with entꢀ2.HCl.
Therefore, an alternate synthetic route was explored to generate Nꢀ4ꢀ
cyclohexylmethyl substituted derivatives 5e, entꢀ5e, 7a, 7b, 7c and 7d with high σ1
affinity. The large hydrophobic aryl residue at Nꢀ1 was introduced at the second step
to avoid the reductive alkylation on the stage of piperazine ethanol 6 or 10. The
diester 2.HCl was reacted with different aldehydes and the resulting imines 15 were
reduced with NaBH4 to provide the secondary amines 16 (Scheme 4). Acetylation of
16 with chloroacetyl chloride led to the chloroacetamides 17, which were reacted with
cyclohexylmethylamine to obtain the piperazinediones 18. Reduction of 18 led to the
piperazinylethanols 7aꢀd in high yields. Thus, compounds 7aꢀd could be more
efficiently prepared according to the route outlined in Scheme 4.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 36
8
1
2
3
4
5
Pharmacological evaluation
6
7
8
Receptor binding studies
9
The σ1 and σ2 affinity of the test compounds was determined in competition
experiments with radioligands. All compounds were tested at σ1 and σ2 receptors of
animal origin prepared from guinea pig brain and rat liver, respectively. To investigate
the correlation of receptor binding, potent ligands were also assayed with human σ1
and σ2 receptors obtained from human tumor cell lines RPMI 8226 and RTꢀ4,
respectively. Membrane preparations containing the membrane bound animal and
human σ1 and σ2 receptors were obtained by homogenization, centrifugation and
washing of the respective tissue or cell line. The implementation of the receptor
binding assays using human tumor cell line membrane preparations was previously
described.17ꢀ20 Both σ1 assays were performed with [3H]ꢀ(+)ꢀpentazocine as
radioligand. [3H]ꢀDTG was used as radioligand in the σ2 assays. Compounds with
high affinity were tested three times. For compounds with low σ affinity, only the
inhibition (in %) of the radioligand binding at a concentration of 1 ꢁM is given.
The σ receptor affinities of the different piperazines are presented in Table 1.
Whereas the compounds 5a, 11a and 11b with a small methyl residue on Nꢀ4 do not
show strong affinity to the guinea pig σ1 receptor, the introduction of larger
substituents leads to increased σ1 affinity in the order methyl < butyl < isopentyl <
cyclohexylmethyl. The cyclohexylmethyl derivatives 5e, entꢀ5e, 7a, 7b, entꢀ7a, entꢀ7b
and 7c have Ki values in the low nanomolar range but show low selectivity against
the σ2 receptor. The introduction of a methoxy group into the hydrophobic aryl
residue of 7c leads to a decrease in both σ1 and σ2 receptor affinity as it is
demonstrated by the pairs 5b/11d and 7c/7d. The absolute configuration has a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 9 of 36
Journal of Medicinal Chemistry
9
1
2
3
4
5
6
7
negligible effect on the receptor affinity, since (R)ꢀ and (S)ꢀconfigured enantiomers
bind with similar IC50 values (e.g. 5e and entꢀ5e).
8
9
Table 1 σ1 and σ2 receptor affinity of hydroxyethyl substituted piperazines
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
σ1
σ2
σ1 (hum)c)
Ki [nM]
± SEM
σ2
(gp)a)
Ki [nM]
± SEM
28 %e)
(rat)b)
Ki [nM]
± SEM
27 %e)
(hum)d)
Ki [nM]
± SEM
n.d.
R1
R2
5a
5b
5c
5d
5e
entꢀ5e
7a
entꢀ7a
7b
entꢀ7b
7c
7d
11a
11b
11c
11d
(+)pentazocine
haloperidol
Me
Bu
Ph
Ph
Ph
Ph
Ph
n.d.
n.d.
279f)
46 ± 1
10 %e)
877f)
n.d.
n.d.
CH2CH2C(CH3)2
(CH2)2C6H5
CH2C6H11
CH2C6H11
CH2C6H11
CH2C6H11
CH2C6H11
CH2C6H11
CH2C6H11
CH2C6H11
Me
107 ± 13
41 ± 13
21 ± 4
23 ± 8
34 ± 8
29 ± 7
29 ± 10
7.5 ± 2.7
6.8 ± 2.3
27 ± 4
n.d.
31 ± 3
779f)
n.d.
4.2 ± 1.1
1.9 ± 0.4
3.5 ± 0.5
7.8 ± 2.4
1.9 ± 0.6
0.9 ± 0.2
4.7 ± 1.8
43 ± 10
3 %e)
116f)
168f)
Ph
60 ± 11
73 ± 45
38 ± 1
26 ± 12
14 ± 4
69 ± 24
253e)
251f)
4ꢀPhꢀPh
4ꢀPhꢀPh
1ꢀNaph
1ꢀNaph
2ꢀNaph
6ꢀMeOꢀ2ꢀNaph
1ꢀNaph
4ꢀPhꢀPh
4ꢀPhꢀPh
4ꢀMeOꢀPh
43 ± 13
39 ± 11
16 ± 4
37 ± 21
31 ± 13
n.d.
0 %e)
3 %e)
230f)
n.d.
n.d.
n.d.
n.d.
Me
Bu
Bu
17 %e)
711f)
n.d.
465f)
491f)
16 %e)
ꢀ
n.d.
5.4 ± 0.5
6.6 ± 0.9 78 ± 2.3
71 ± 8 58 ± 18
36 ± 5.2
40 ± 5.3
208 ± 26
ꢀ
200 ± 33
20 ± 5.8
diꢀoꢀtolylguanidine
a) guinea pig brain, b) rat liver c) RPMI 8226 cells, d) RTꢀ4 cells, e) Inhibition of
radioligand binding at 1ꢁM concentration of test compound, f) result from one
measurement, n.d.: not determined
The affinity of the synthesized compounds towards human σ1 receptors (RPMI 8226
cell line) is reduced compared to the affinity recorded in the guinea pig assay. The
same trend was observed for the reference compounds (+)ꢀpentazocine, haloperidol
and diꢀoꢀtolylguanidine. However, the data show the same tendency indicating that
the results obtained with receptor material from guinea pig brain and human tumor
cell line RPMI 8226 are well comparable. We assume that the higher Ki values
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 36
10
1
2
3
4
5
6
7
obtained in the human tumor cell line based σ1 assay originate from different test
materials and assay conditions.
8
9
For the discussion of the σ1/σ2 selectivity, the data recorded in the traditional guinea
pig brain (σ1) and rat liver (σ2) assays are used. These data indicate more than
tenfold selectivity for most of the piperazine derivatives for the σ1 over the σ2
subtype. Due to the higher Ki(σ1)ꢀvalues obtained with human tumor RPMI 8226 cell
line the σ1/σ2 selectivity is considerably reduced, when these values are compared
with the Ki(σ2) values obtained with rat liver membrane preparations. Moreover, the
Ki(σ2) values recorded with rat liver and human RTꢀ4 cell line preparations are very
similar indicating reduced σ1/σ2 selectivity when compared with the Ki(σ1)ꢀvalues
obtained with human RPMI 8226 cell line preparations. The very potent σ1 ligands
7a, entꢀ7a and 7b interacting in the low nanomolar range with the guinea pig σ1
receptors (Ki(σ1) = 1.9ꢀ3.5 nM) reveal high selectivity over the rat σ2 receptor (14 to
31ꢀfold), but can be regarded as nonꢀselective, when considering the Ki(σ1) values
obtained with the human tumor cell line RPMI 8226. These variations of the Ki values
are attributed to species differences and different preparations of receptor material. In
this work these species differences were observed for the first time and should be
considered in further σ1/σ2 selectivity discussions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Within this series of compounds, the cyclohexylmethyl derivative 7c represents the
most potent σ1 ligand with the lowest Ki value in the human σ1 assay (Ki = 6.8 nM)
and a promising σ1/σ2 selectivity for both the animal (15ꢀfold) and the human tumor
cell based assays (5ꢀfold). Therefore 7c was selected for further characterization in
the cytotoxicity assay with the aim to understand better the pharmacological
properties of potent σ1 ligands.
ACS Paragon Plus Environment

Page 11 of 36
Journal of Medicinal Chemistry
11
1
2
3
4
5
Cytotoxicity of 7c in human cancer cell lines
6
7
8
9
The antiproliferative effect of 7c was investigated in seven human tumor cell lines,
namely 5637 and RTꢀ4 (bladder cancer), A427 (small cell lung cancer), LCLCꢀ103H
(large cell lung cancer), MCFꢀ7 (breast cancer), DANꢀG (pancreas cancer) and RPMI
8226 (multiple myeloma). In case of the suspensionꢀcell line RPMI 8226, 7c was
tested in a MTT assay.21 The optical density of the resulting formazane product was
used to determine the cytotoxic activity of the compound. For the adherent cell lines,
crystal violet assay21 was used. The results are given as part of adherent cells in
relation to an untreated control. Five serially diluted stock solutions were used to
determine the IC50 values.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2 Average IC50 values (± SD) for the growth inhibition of 7c employing different
tumor cell lines. Results are from at least 3 independent experiments.
cell line
5637
RTꢀ4
IC50 [ꢁM]
2.1 ± 1.9
˃20
Aꢀ427
2.5 ± 2.6
˃20
˃20
15 ± 2
7.3 ± 3.9
LCLCꢀ103H
DANꢀG
MCFꢀ7
RPMI 8226
7c inhibited the growth of the cell lines 5637, Aꢀ427 and RPMI 8226, with IC50 values
in the low micromolar range, whereas lower activity was observed for the MCFꢀ7 cell
line. The growth inhibition of LCLCꢀ103H, RTꢀ4, and DANꢀG cell lines was not
reduced at a concentration of 20 ꢁM of 7c. These results indicate a selective
mechanism of cytotoxicity for 7c.
Induction of apoptosis in RPMIꢀ8226 cell lines
For these studies, RPMIꢀ8226 cells were used because of their high density of σ1
receptors.18 After incubation of RPMIꢀ8226 cells with 7c at the IC50ꢀ and IC90ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 36
12
1
2
3
4
5
6
7
8
9
concentrations for 24 or 48 h, the cells were harvested and doubleꢀstained with
annexinꢀV and propidium iodide (PI). The relative numbers of early (i.e. those stained
with annexinꢀVꢀFITC) and late apoptotic cells (i.e. those stained with both annexinꢀVꢀ
FITC and PI) were quantified by flow cytometry.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Population of annexinꢀV positive, PI negative RPMIꢀ8226 cells after 24 h
and 48 h treatment with 7c.
Table 3 Apoptotic activity in untreated controls and in 7c treated RPMIꢀ8226 cells
24 h incubation period
48 h incubation period
Concentration earlyꢀapoptotic lateꢀapoptotic earlyꢀapoptotic lateꢀapoptotic
(%)
0
7.3
13.2
cells (%)a
12.8 ± 2.59
8.60 ± 1.50
9.94 ± 0.76
cells (%)b
8.11 ± 1.44
9.71 ± 2.41
15.67 ± 7.55
cells (%)a
14.6 ± 3.09
22.55 ± 5.46
25.82 ± 5.89
cells (%)b
6.43 ± 0.77
10.59 ± 4.16
17.74 ± 6.86
a) annexinꢀV positive, PI negative; b) annexinꢀV positive, PI positive
The results presented in Figure 1 and Table 3 show a significant increase in the
amount of earlyꢀapoptotic cells (annexinꢀV positive, PI negative cells) in relation to
the untreated control after 48 h. Shorter incubations (i.e. 24 h) did not result in an
increased number of apoptotic cells, neither at the IC50 nor at the IC90 concentration.
ACS Paragon Plus Environment

Page 13 of 36
Journal of Medicinal Chemistry
13
1
2
3
4
5
6
7
Further studies will be necessary to elucidate the mechanism of apoptosis; e.g.,
whether it is a caspase dependent or independent event.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molecular Modeling
To support the SAR analysis developed above with a molecularꢀbased rationale for
the binding mode of these series of new, potent piperazine σ1 ligands, the model
structure of the best derivative 7c was docked into the putative binding site of the 3D
σ1 receptor homology model22 and the drug/protein free energy of binding (
ꢀGbind)
was estimated using a validated molecular dynamics (MD) procedure23 based on
MM/PBSA calculations24 (see Supporting Information for details). The analysis of the
corresponding MD trajectories reveals that four major types of interactions are
involved in the binding mode of 7c to the
σ
1 receptor, as highlighted in Figure 2: 1)
π
interactions between residues Arg119 (
π
ꢀcation) and Tyr120 (Tꢀstacking π−π),
respectively, with the naphthalene ring of 7c; 2) a hydrogen bond (HB) between the
side chain –OH donor of Thr181 and the acceptor oxygen atom of the hydroxyethyl
chain of 7c; 3) a salt bridge between the –NH+ of the Nꢀalkyl moiety of 7c and the
protein Asp126 –COOꢀ and 4) a robust network of stabilizing van der Waals
interactions between the high hydrophobic cavity leaned by the side chains of Ile128,
Phe133, Tyr173, and Leu186 and the cyclohexylmethyl portion of 7c.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 36
14
1
2
3
4
5
a)
b)
6
7
8
173Tyr
133Phe
D126
L186
T181
Ile128
186Leu
9
I128
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
R119
Hydrophobic interaction
F133
Y120
126Asp
HO
Y173
salt bridge
Thr181
N
OH
O
N
OH
H
interaction
NH
H2N
OH
HN
Tyr120
Arg119
Figure 2. a) 2D schematic representation of the identified interactions between the
3D homology model σ1 receptor and 7c. The lines/arrows indicate key interactions
between the receptor and its ligand. b) Equilibrated MD snapshot of the complex of
the σ1 receptor with 7c showing the key interactions depicted in part a. The main
protein residues involved in these interactions are Arg119 and Tyr120 (πꢀinteractions,
cyan), Asp126 (salt bridge, red), Ile128, Phe133, Tyr173 and Leu186 (hydrophobic
interaction, magenta) and Thr181 (hydrogen bond, green). Compound 7c is shown in
atomꢀcolored sticksꢀandꢀballs (C, gray; N, blue; and O, red). H atoms are omitted, but
the salt bridge and the Hꢀbond are indicated as black lines or black dotted lines,
respectively. Na+ and Clꢀ ions are represented as dark gray and light gray sphere,
respectively. Water molecules are not shown for clarity. The 3D model of the σ1
receptor was obtained 22 by homology modeling techniques starting from the
following model templates: 3ClA.pdb, 1l24.pdb, 2Z2Z.pdb, and 2Q8l.pdb.
All the interactions described above are quantified by a calculated ꢀGbind of 7c for σ1
equal to ꢀ10.85 ± 0.36 kcal/mol, corresponding to an estimated affinity σ1Ki,calcd value
of 11.2 nM (Table 4), in stringent agreement with the experimental quantity σ1Ki,exp of
6.8 nM (Table 1).
ACS Paragon Plus Environment

Page 15 of 36
Journal of Medicinal Chemistry
15
1
2
3
4
5
6
7
8
9
Table 4 Major components and total binding free energy (kcal/mol) for compound 7c
in complex with the σ1 receptor. The calculated σ1Ki (σ1Ki,calcd, nM) is also reported
for comparison with the corresponding experimental value
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Components
ꢀEVDW
ꢀEELE
7c
ꢀ54.40 ± 0.05
ꢀ90.84 ± 0.09
ꢀ145.24 ± 0.10
126.11 ± 0.14
ꢀ4.31 ± 0.03
121.80 ± 0.14
ꢀ23.44 ± 0.18
12.59 ± 0.31
ꢀ10.85 ± 0.36
11.2
ꢀEMM
ꢀGPB
ꢀGNP
ꢀGSOL
ꢀHbind
ꢀTꢀSbind
ꢀGbind
a
σ1Ki,calcd
a The σ1Ki,calc values were obtained from the corresponding ꢀGbind values using the
relationship ꢀGbind = ꢀRT ln(1/σ1Ki,calcd).
Similarly to other σ1/ligand complexes,22,24,25 the nonbonded mechanical energy
components of ꢀGbind, ꢀEVDW and ꢀEELE, afford a substantial, favorable contribution
to binding. On the other hand, the desolvation penalty paid by 7c upon binding
(ꢀGPB) is also quite substantial, so that the net, resulting electrostatic contribution to
the affinity of 7c for the σ1 receptor is markedly unfavorable. Specifically, the mean
value of the electrostatic energy (ꢀEELE + ꢀGPB) is 35.27 kcal/mol, whilst the
corresponding value of the van der Waals and hydrophobic interaction energies
(ꢀEVDW + ꢀGNP) is ꢀ58.71 kcal/mol. Accordingly, it follows that the association
between the σ1 receptor and 7c is mainly driven by more favorable nonpolar
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 36
16
1
2
3
4
5
6
7
interactions in the complex than in the solution, in harmony with a proposed general
scheme for noncovalent association.
8
9
To investigate in details the binding mode of 7c to the σ1 receptor, a deconvolution of
the enthalpic component (ꢀHbind) of the binding free energy into contributions from
each protein residue was carried out, as illustrated in Figure 3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Per residue binding enthalpy decomposition for the σ1/7c complex. Only σ1
amino acids from position 100 to 200 are shown, as for all remaining residues the
contribution to ligand binding is irrelevant.
As seen in Figure 4, substantial van der Waals and electrostatic interactions are
contributed by residues Arg119 (ꢀ0.75 kcal/mol) and Y120 (ꢀ1.64 kcal/mol) ꢀ through
the aforementioned πꢀcation and Tꢀstacking π−π interaction, respectively ꢀ and by the
residues belonging to the hydrophobic pocket Ile128 (ꢀ1.12 kcal/mol), Phe133 (ꢀ0.97
kcal/mol), Tyr173 (ꢀ0.88 kcal/mol) and Leu186 (ꢀ0.83 kcal/mol) (Figure 2) The stable
salt bridge (average dynamic length (ADL) = 3.91 ± 0.07 Å) involving Asp126 and the
permanent hydrogen bond (ADL = 1.99 ± 0.05 Å) featured by Thr181are responsible
for stabilizing contributions of ꢀ2.67 kcal/mol and ꢀ2.33 kcal/mol, respectively. All
other receptor residues were characterized by negligible interaction enthalpy values
(|ꢀHbind| < 0.30 kcal/mol).
ACS Paragon Plus Environment

Page 17 of 36
Journal of Medicinal Chemistry
17
1
2
3
4
5
6
7
8
9
To corroborate the experimental structureꢀactivity relationship formulated for this
series of σ1 ligands, the same computational recipe was applied to all remaining 15
compounds. Accordingly, the corresponding receptor/ligand complexes were built
and optimized by a combined molecular mechanics/molecular dynamics procedure,
and the relevant values of ꢀGbind/Ki,calcd evaluated applying the MMꢀPBSA
computational ansatz. These results are listed in Table 5.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5 Binding free energies ꢀGbind (kcal/mol) and σ1Ki,calcd values (nM) for the entire
set of the 12 bioactive ligands in complex with the σ1 receptor. Errors are given in
parenthesis as standard errors of the mean. The experimental σ1Ki values (nM) and
the calculated σ1Ki (nM), as estimated from the corresponding ꢀGbind values (ꢀGbind
= ꢀRT ln(1/σ1Ki,calcd)), are also reported for comparison.
ꢀꢀH
(kcal/mol)
ꢀ17.44
(0.15)
ꢀ20.13
(0.18)
ꢀ22.02
(0.19)
ꢀ22.43
(0.17)
ꢀ22.19
(0.16)
ꢀ22.25
(0.16)
ꢀ23.31
(0.18)
ꢀ23.27
(0.19)
ꢀ23.34
(0.17)
ꢀ23.11
(0.17)
ꢀ23.44
(0.18)
ꢀ23.26
(0.19)
ꢀ18.87
(0.19)
ꢀ18.73
(0.16)
ꢀ20.80
ꢀTꢀS
(kcal/mol)
ꢀ10.40
(0.31)
ꢀ11.62
(0.32)
ꢀ12.22
(0.29)
ꢀ12.11
(0.29)
ꢀ12.01
(0.28)
ꢀ12.00
(0.30)
ꢀ13.02
(0.29)
ꢀ12.91
(0.31)
ꢀ12.62
(0.27)
ꢀ12.41
(0.30)
ꢀ12.59
(0.31)
ꢀ12.78
(0.31)
ꢀ11.72
(0.29)
ꢀ11.63
(0.28)
ꢀ11.74
ꢀGbind
(kcal/mol)
ꢀ7.04
(0.34)
ꢀ8.51
σ1Ki,calcd
σ1Ki
Compound
5a
[nM]
[nM] ± SEM
6900
580
66
28%
279
5b
(0.37)
ꢀ9.80
5c
107 ± 13
41 ± 13
21 ± 4
23 ± 8
34 ± 8
29 ± 7
29 ± 10
7.5 ± 2.7
6.8 ± 2.3
27 ± 4
3%
(0.35)
ꢀ10.32
(0.35)
ꢀ10.18
(0.32)
ꢀ10.25
(0.34)
ꢀ10.29
(0.34)
ꢀ10.36
(0.36)
ꢀ10.72
(0.32)
ꢀ10.70
(0.34)
ꢀ10.85
(0.36)
ꢀ10.48
(0.36)
ꢀ7.15
5d
27
5e
35
entꢀ5e
7a
31
29
entꢀ7a
7b
26
14
entꢀ7b
7c
15
11
7d
21
11a
5800
(0.35)
ꢀ7.10
(0.32)
11b
11c
6300
230
17%
465
ꢀ9.06
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 36
18
1
2
3
4
5
6
7
(0.15)
(0.30)
(0.34)
ꢀ20.02
(0.18)
ꢀ11.53
(0.30)
ꢀ8.49
(0.35)
11d
600
491
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
It is interesting to observe how, notwithstanding all approximation involved, the
applied methodology was able to rank correctly the entire set of compounds with
respect to their affinity towards the human σ1 receptor protein. Specifically, only those
compounds carrying a bulky substituent on Nꢀ1 (5cꢀe and 7aꢀd) are endowed with
the most favorable free energy of binding (ꢀGbind < 10 kcal/mol). The presence of a
butyl chain (5b and 11cꢀd) reflects in a weak but still appreciable ligand binding
capacity while the introduction of the small methyl moiety on Nꢀ1 (5a and 11aꢀb)
leads to a dramatic loss of protein/ligand affinity (3ꢀ4 kcal/mol, Table 5). On the other
hand, an aromatic substituent in R2 and, even more intriguingly, the absolute
configuration of the enantiomeric compounds appear to play a negligible role in
receptor binding (e.g., 5e and entꢀ5e, 7a and entꢀ7a, and 7b and entꢀ7b, Table 5).
Lastly, the dissection of the binding enthalpy into per residue contribution applied to
7c has been extended to the entire molecular series of σ1 ligands. To facilitate a
direct comparison with the lead compound 7c, the values of ꢀHbind for the critical σ1
residues were clustered according to a specific underlying interaction type, as follow:
Arg119 and Tyr120 were considered for π interactions (π), Asp126 for salt bridge
(SB), Thr181 for hydrogen bond (HB) and Ile128, Phe133, Tyr173, and Leu186 for
hydrophobic interactions (HY). Moreover, compounds 5a, 7a, and 11c were selected
for a meaningful comparison with 7c. These results are reported in Figure 4.
ACS Paragon Plus Environment

Page 19 of 36
Journal of Medicinal Chemistry
19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Comparison of per residue binding enthalpy decomposition for compounds
5a, 7a, 7c, and 11c in complex with the σ1. Critical receptor residues are clustered
according to the specific underlying interactions: π (π interactions); SB (salt bridge);
HB (hydrogen bond); HY (hydrophobic interactions) (see legend and main text for
more details).
Figure 4 clearly shows that the lead compound 7c and the derivative 7a exhibit a
similar interaction profile except for negligible quantitative differences. Indeed, the
naphthalene ring (7c) and the biphenyl moiety (7a) are both able to establish the best
interactions with residues Arg119 and Tyr120; furthermore, the similar scaffold of
these two compounds allows for the adoption of an optimal conformational fit of the
molecule within the entire binding pocket. In contrast, the absence of a large,
hydrophobic substituent on Nꢀ1 as in compounds 5a and 11c leads to a drastic loss
not only in HY stabilization (ꢀH(HY,5a) = ꢀ0.75 kcal/mol, ꢀH(HY,11c) = ꢀ1.57 kcal/mol
vs. ꢀH(HY,7c) = ꢀ3.80 kcal/mol) but also in all other main interactions, by virtue of a
global rearrangement of the receptor/complex binding site. Accordingly, the resultant
affinity for the σ1 protein of 7a is similar to that of 7c, whilst the binding of 5a and 11a
to the same receptor is substantially weaker.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 36
20
1
2
3
4
5
Conclusion
6
7
8
9
Hydroxyethylpiperazines with a small methyl group at an Nꢀatom do not reveal high
σ1 receptor affinity, whereas analogs with larger lipophilic residues at both Nꢀatoms
show high σ1 affinity. The cyclohexylmethyl group proved to be the preferred
substituent at the Nꢀ4 position, as demonstrated for the compounds 5e, entꢀ5e, 7aꢀd
and entꢀ7a,b. The configuration in 2ꢀposition of the piperazine ring does not exert a
strong impact on σ1 receptor affinity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The Ki values recorded with the human tumor cell line RPMI 8226 are slightly higher,
but show the same tendency as the Ki values obtained with receptor material from
guinea pig brain. Since the cyclohexylmethyl substituted derivative 7c showed the
highest affinity towards the human σ1 receptor, the growth inhibition of seven tumor
cell lines was investigated. It was shown that 7c inhibited selectively the growth of
three tumor cell lines indicating a relationship between σ1 affinity and growth
inhibition of these tumor cell lines. The growth inhibition by 7c is mediated by
inducing apoptosis.
The interaction of 7c with the human σ1 receptor was analyzed in detail by docking
studies using the recently reported 3D homology model of the σ1 receptor. Four types
of interactions were found to be crucial for high σ1 affinity: π interactions between
Arg119/Tyr120 and the naphthalene ring, a hydrogen bond between Thr181 and the
Oꢀatom of the hydroxyethyl moiety, an ionic interaction between Asp126 and a
protonated piperazine Nꢀatom, and finally lipophilic van der Walls interactions
between the side chains Ile128, Phe133, Tyr173, and Leu186 with the
cyclohexylmethyl substituent. The contribution of the single amino acids to the free
energy resulting from binding of 7c with the human σ1 receptor was analyzed.
ACS Paragon Plus Environment

Page 21 of 36
Journal of Medicinal Chemistry
21
1
2
3
4
5
6
7
8
9
The interaction of the whole set of hydroxyethylpiperazines towards the human σ1
receptor was analyzed in detail and it was found that the calculated free binding
energies correlate nicely with the recorded Ki values. It should be emphasized that
this is the first study using the human tumor cell line RPMI 8226 for recording σ1
affinity of a set on novel ligands. Therefore this is the first study, which allows the
correlation of the calculated free binding energy with recorded Ki values.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Part
Chemistry, general
Thin layer chromatography: Silica gel 60 F254 plates (Merck). Flash chromatography
(fc): Silica gel 60, 40–43 µm (Merck); parentheses include: diameter of the column,
eluent, Rf value. Melting point: Melting point apparatus SMP 3 (Stuart Scientific),
1
uncorrected. H NMR (400 MHz), 13C NMR (100 MHz): Unity Mercury Plus AS 400
NMR spectrometer (Varian); δ in ppm related to tetramethylsilane; coupling constants
1
are given with 0.5 Hz resolution; the assignments of 13C and H NMR signals were
supported by 2D NMR techniques. The purity of all compounds was determined by
HPLC analysis. HPLC (method ACN): Merck Hitachi Equipment; UV detector: Lꢀ
7400; autosampler:Lꢀ7200; pump: Lꢀ7100; degasser: Lꢀ7614; column: LiChrospher®
60 RPꢀselect B (5 µm); LiCroCART® 250ꢀ4 mm cartridge; flow rate: 1.0 mL/min;
injection volume: 5.0 µL; detection at λ = 210 nm; solvent A: demineralized water with
0.05% (v/v) trifluoroacetic acid; solvent B: acetonitrile with 0.05% (v/v) trifluoroacetic
acid: gradient elution (% A): 0ꢀ4 min: 90.0 %; 4ꢀ29 min: gradient from 90 % to 0 %;
29ꢀ31 min: 0 %; 31ꢀ31.5 min: gradient from 0 % to 90.0 %; 31.5ꢀ40 min: 90 %.
According to HPLC analysis the purity of all test compounds is > 95 %.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 36
22
1
2
3
4
5
6
7
8
9
General Procedure A: Under N2, 1 equivalent of the monocyclic piperazinedione
was dissolved in THF abs. and the mixture was cooled down to 0 °C. At this
temperature, lithium aluminium hydride (6 equivalents) was added. The reaction
mixture was stirred at 0 °C for 10 min and then heated to reflux for 16 h. Finally water
was added under iceꢀcooling until H2 liberation was finished. The mixture was stirred
at 0 °C for 10 min and then heated to reflux for 30 min. After cooling to room
temperature, the mixture was filtered and the solvent was removed in vacuo. The
residue was purified by fc.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Dimethyl (
Under ice cooling, chlorotrimethylsilane (33.2 mL, 263.0 mmol) was added dropwise
to a suspension of ( )ꢀaspartate (1) (10 g, 75.1 mmol) in methanol abs. (150 mL) and
S
)ꢀ2ꢀaminobutanedioate hydrochloride (2.HCl)16
S
the reaction mixture was stirred for 16 h at room temperature. The solvent and
volatile byproducts were evaporated in vacuo. The residue was suspended in
methanol (1 x 30 mL) and the solvent was removed in vacuo. Then diethyl ether was
added (30 mL) and the solvent was removed under reduced pressure. This
procedure was repeated three times. The product was dried in highꢀvacuum.
1
Colorless solid, mp 112 – 115 °C, yield 15 g (99.9 %). C6H12ClNO4, Mr = 197.6. H
NMR (CD3OD): δ [ppm] = 3.07 (d, J = 5.3 Hz, 2H, CHC
H
2CO2CH3), 3.75 (s, 3H,
OC
H3), 3.84 (s, 3H, OCH3), 4.40 (t, J = 5.4 Hz, 1H, CHCH2CO2CH3). The signal for
+
the protons of the NH3 ꢀgroup is not seen.
Dimethyl (S)ꢀ2ꢀ[2ꢀchloroꢀNꢀ(naphthalenꢀ2ꢀylmethyl)acetylamino]butanedioate
(17c)
2.HCl (1 g, 5.06 mmol) was dissolved in CH2Cl2 abs. (15 mL) and triethylamine (0.74
mL, 5.06 mmol) and 2ꢀnaphthaldehyde (0.79 g, 5.06 mmol) were added slowly. Then
Na2SO4 was added and the mixture was stirred for 16 h at room temperature. For
ACS Paragon Plus Environment

Page 23 of 36
Journal of Medicinal Chemistry
23
1
2
3
4
5
6
7
8
9
workꢀup, the suspension was filtered, the precipitate was washed with CH2Cl2 and
the filtrate was concentrated in vacuo. Diethyl ether was added to the resulting
residue and the mixture was filtered again. Evaporation of the filtrate gave the imine
15c (1.24 g, 4.14 mmol) as a pale yellow solid, which was dissolved in methanol (15
mL) and converted directly by addition of NaBH4 (260.4 mg, 7.0 mmol) over 40 min
under ice cooling. The mixture was stirred at room temperature for another 5 min,
before the solvent was evaporated in vacuo. Water (15 mL) was added to the residue
and the aqueous layer was extracted with diethyl ether (5 x 10 mL). The combined
organic layers were dried (K2CO3), filtered and the solvent was removed under
reduced pressure. The oily residue (16c, 1.03 g, 3.42 mmol) was dissolved in CH2Cl2
abs. (15 mL). Triethylamine (0.38 mL, 2.73 mmol) and chloroacetyl chloride (0.44 mL,
5.47 mmol) were added dropwise under N2ꢀatmosphere and ice cooling. The mixture
was stirred for 2.5 h at room temperature. The solvent was removed in vacuo and
diethyl ether (50 mL) was added to the resulting residue. The suspension was
filtered, washed and the filtrate was concentrated in vacuo. Purification of the residue
was performed by fc (∅ 3.5 cm, h = 16 cm, v = 30 mL, cyclohexane/ethyl acetate =
4/1, Rf = 0.12). Pale yellow oil, yield 291.6 mg (15 % over 3 steps). C19H20ClNO5, Mr =
377.8. 1H NMR (CDCl3): δ [ppm] = 2.69 (dd, J = 17.4 / 7.9 Hz, 0.2H,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CHC
H
2CO2CH3NR), 2.80 (dd, J = 17.1 / 6.9 Hz, 0.8H, CHC
2CO2CH3NR), 3.30 (dd, J = 17.1 / 6.3 Hz, 0.8H,
2CO2CH3HR), 3.40 (s, 3 x 0.2H, OCH3NR), 3.57 (s, 3 x 0.2H, OCH3NR), 3.60 (3 x
0.8H, OCH3HR), 3.68 (s, 3 x 0.8H, OCH3HR), 4.11 (d, J = 12.7 Hz, 0.8H,
2ClHR), 4.17 (d, J = 12.6 Hz, 0.8H, O=CC 2ClHR), 4.37 (d, J = 15.8 Hz, 0.2H,
2ArNR), 4.51 (s, broad, 2 x 0.2H, O=CC 2ClNR), 4.59 (t, J = 6.6 Hz, 0.8H,
CH2CO2CH3HR), 4.84 (d, J = 16.7 Hz, 0.8H, NC 2ArHR), 4.91 (d, J = 16.8 Hz,
H
2CO2CH3HR), 3.01 (dd, J
= 17.3 / 6.4 Hz, 0.2H, CHC
H
CHC
H
O=CC
NC
H
H
H
H
CH
H
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 36
24
1
2
3
4
5
6
7
8
9
0.8H, NC
J = 8.5 Hz, 0.2H, ArꢀHNR), 7.43 – 7.45 (m, 0.2H, ArꢀHNR + 0.8H, ArꢀHHR), 7.50 – 7.54
(m, 2 x 0.8H, ArꢀHHR), 7.59 (s, broad, 0.2H, NC 2ArNR), 7.77 (d, J = 8.6 Hz, 3 x 0.2H,
2ArNR), 7.78 – 7.88 (m, 0.2H, ArꢀHNR + 4 x 0.8H, ArꢀHHR). Ratio of rotamers is
H H H
2ArHR), 5.05 – 5.15 (m, 2 x 0.2H, NC 2ArNR + C CH2CO2CH3NR), 7.33 (d,
H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NC
H
80:20.
(S)ꢀMethyl 2ꢀ[4ꢀ(cyclohexylmethyl)ꢀ1ꢀ(naphthalenꢀ2ꢀylmethyl)ꢀ3,6ꢀdioxoꢀ
piperazinꢀ2ꢀyl]acetate (18c)
17c (200 mg, 0.53 mmol) was dissolved in dry acetonitrile (molecular sieves 3 Å, 10
mL) and triethylamine (0.1 mL, 0.64 mmol) and cyclohexylmethylamine (0.1 mL, 0.69
mmol) were added slowly. The reaction mixture was stirred at room temperature for
16 h. For workup, the solvent was evaporated in vacuo and the residue was
dissolved in ethyl acetate (10 mL). The organic layer was washed with 0.5 M HCl (2 x
4 mL), 0.5 M NaOH (1 x 4 mL) and brine (1 x 4 mL), dried (Na2SO4), filtered and
concentrated in vacuo. The residue was purified by fc (∅ 2 cm, h = 20 cm, v = 10 mL,
cyclohexane/ethyl acetate = 4/1, Rf = 0.12). Colorless solid, mp 119 ꢀ 123 °C, yield
173 mg (77 %). C25H30N2O4, Mr = 422.5. 1H NMR (CDCl3): δ [ppm] = 0.91 – 1.00 (m,
2H, NCH2C6
2.82 (dd, J = 17.2 / 5.1 Hz, 1H, CHC
CHC 2CO2CH3), 3.18 – 3.30 (m, 2H, NC
17.2 Hz, 1H, O=C 2N), 4.13 (t, J = 4.1 Hz, 1H, C
Hz, 1H, O=C 2N), 4.43 (d, J = 15.3 Hz, 1H, NC
NC 2Ar), 7.35 (dd, J = 8.4 / 1.8 Hz, 1H, Arꢀ ), 7.46 – 7.51 (m, 2H, Arꢀ
Arꢀ 1ꢀH), 7.77 – 7.83 (m, 3H, Arꢀ ).
H
11), 1.13 – 1.28 (m, 3H, NCH2C6
2CO2CH3), 3.02 (dd, J = 17.2 / 3.7 Hz, 1H,
2C6H11), 3.54 (s, 3H, OC 3), 3.97 (d, J =
CH2CO2CH3), 4.38 (d, J = 17.5
2Ar), 5.20 (d, J = 15.1 Hz, 1H,
), 7.69 (s, 1H,
H11), 1.65 – 1.74 (m, 6H, NCH2C6H11),
H
H
H
H
H
H
H
H
H
H
H
H
H
ACS Paragon Plus Environment

Page 25 of 36
Journal of Medicinal Chemistry
25
1
2
3
4
5
(S
)ꢀ2ꢀ[4ꢀ(Cyclohexylmethyl)ꢀ1ꢀ(naphthalenꢀ2ꢀylmethyl)piperazinꢀ2ꢀyl]ethanol
6
7
(7c)
8
9
7c was synthesized according to General Procedure A: Reaction of 18c (150 mg,
0.36 mmol) with LiAlH4 solution (1 M in THF, 2.1 mL, 2.1 mmol) was performed in
THF abs. (12 mL). Fc (∅ 2 cm, h = 18 cm, v = 10 mL, cyclohexane/ethyl acetate =
3/2, Rf = 0.13). Colorless solid, mp 55 – 59 °C, yield 75.2 mg (57 %). C24H34N2O, Mr =
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
366.5. 1H NMR (CDCl3): δ [ppm] = 0.81 – 0.90 (m, 2H, NCH2C6
2H, NCH2C6 11), 1.45 – 1.52 (m, 1H, NCH2C6 11), 1.63 – 1.79 (m, 5H, NCH2C6H11
CHC 2CH2OH), 1.89 – 1.91 (m, 2H, NCH2C6 11, CHC 2CH2OH), 2.04 – 2.07 (m,
1H, NCH2C6 11 or CHC 2CH2OH), 2.11 (d, J = 7.2 Hz, 2H, NC 2C6H11), 2.32 – 2.42
2), 2.57 – 2.60 (m, 1H, NC 2), 2.86 – 2.89 (m, 1H, NC ), 2.92 – 2.98
2), 3.57 (d, J = 12.9 Hz, 1H, NC 2Ar), 3.77 – 3.83 (m, 1H,
2OH), 4.28 (d, J = 12.2 Hz, 1H,
), 7.72 (s, 1H, Arꢀ 1ꢀH), 7.79 – 7.83 (m, 3H, Arꢀ ).
H11), 1.13 – 1.23 (m,
H
H
,
H
H
H
H
H
H
(m, 4H, NC
H
H
H
(m, 1H, NC
H
H
CHCH2C
NC 2Ar), 7.42 – 1.51 (m, 3H, Arꢀ
H2OH), 3.87 – 3.92 (m, 1H, CHCH2CH
H
H
H
H
The signal for the proton of the OH group is not seen.
Molecular modeling
The model structures of compounds 5aꢀe, 7aꢀd, and 11aꢀd were sketched and
geometrically optimized using Discovery Studio (DS, version 2.5, Accelrys, San
Diego, CA). A conformational search was then carried out using a wellꢀvalidated, ad
hoc developed combined molecular mechanics/molecular dynamics simulated
annealing (MDSA) protocol22 ꢀ 24, 26 ꢀ 32 using Amber 1233 and the ff03 force field.34
Accordingly, the relaxed structures were subjected to five repeated temperature
cycles (from 300 to 1000 K and back) using constantꢀvolume/constantꢀtemperature
(NVT) molecular dynamics (MD) conditions. At the end of each annealing cycle, the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 36
26
1
2
3
4
5
6
7
8
9
structures were again energy minimized to converge below 10ꢀ4 kcal/mol Å, and only
the structures corresponding to the minimum energy were used for further modeling.
The atomic partial charges for the geometrically optimized compounds were obtained
using the RESP procedure,35ꢀ37 and the electrostatic potentials were produced by
singleꢀpoint quantum mechanical calculations at the Hartree – Fock level with a 6ꢀ
31G * basis set, using the Merz − Singh − Kollman van der Waals parameters.
Eventual ff03 missing force field parameter of Amber 12 were generated using the
general Amber force field (GAFF).38
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The compound optimized structures were then docked into the σ1 putative binding
pocket by applying a consolidated procedure performed with Autodock 4.3/Autodock
Tools 1.4.639 on a win64 platform. The resulting structures were visualized and
subjected to cluster analysis with a 1 Å tolerance for an allꢀatom rootꢀmeanꢀsquare
(rms) deviation from a lower energy structure representing each cluster family. Then,
for each compound only the molecular conformation satisfying the combined criteria
of having the lowest (i.e., more favorable) Autodock energy and belonging to a highly
populated cluster was selected to carry for further modeling.
Each ligand/receptor complex obtained from the docking procedure was further
refined in Amber 12 using the quenched molecular dynamics (QMD) method.40–49
According to QMD, 1 ns MD simulations at 300 K were employed to sample the
conformational space of each ligand/receptor complex in the GB/SA continuum
solvation environment.50,51 The integration step was equal to 1 fs. After each
picosecond, each system was cooled to 0 K, and the structure was extensively
minimized and stored. To prevent global conformational changes of the protein, the
backbone atoms of the protein binding site were constrained by a harmonic force
constant of 100 kcal/Å, whereas the amino acid side chains and the ligands were
ACS Paragon Plus Environment

Page 27 of 36
Journal of Medicinal Chemistry
27
1
2
3
4
5
6
7
8
9
allowed to move without any constraint. The best energy configuration of each
complex resulting from the previous step was subsequently solvated by a cubic box
of TIP3P52 water molecules extending at least 10 Å in each direction from the solute.
Each system was then neutralized and, furthermore, the solution ionic strength was
adjusted to the physiological value of 0.15 M by adding the required amounts of Na+
and Clꢀ ions. Each solvated system was relaxed by 500 steps of steepest descent
followed by 500 other conjugateꢀgradient minimization steps and then gradually
heated to a temperature of 300 K in intervals of 50 ps of NVT MD, using a Verlet
integration time step of 1.0 fs. The Langevin thermostat was used to control
temperature, with a collision frequency of 2.0 psꢀ1. The SHAKE method53 was used
to constrain all of the covalently bound hydrogen atoms, while longꢀrange nonbonded
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
van der Waals interactions were truncated by using dual cutoffs of 6 and 12 Å. The
particle mesh Ewald (PME) method54 was applied to treat longꢀrange electrostatic
interactions. The protein was restrained with a force constant of 2.0 kcal/(mol Å), and
all simulations were carried out with periodic boundary conditions.
The density of the system was subsequently equilibrated via MD runs in the
isothermal − isobaric (NPT) ensemble, with isotropic position scaling and a pressure
relaxation time of 1.0 ps, for 50 ps with a time step of 1 fs. Each system was further
equilibrated using NPT MD runs at 300 K, with a pressure relaxation time of 2.0 ps.
Five equilibration steps were performed, each 2 ns long and with a time step of 2.0
fs. To check the system stability, the fluctuations of the rmsd of the simulated position
of the backbone atoms of the σ1 receptor with respect to those of the initial protein
were monitored. All chemicoꢀphysical parameters and rmsd values showed very low
fluctuations at the end of the equilibration process, indicating that the systems
reached a true equilibrium condition.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 36
28
1
2
3
4
5
6
7
8
9
The equilibration phase was followed by a data production run consisting of 20 ns of
MD simulations in the canonical (NVT) ensemble. Only the last 10 ns of each
equilibrated MD trajectory were considered for statistical data collections.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
All simulations were carried out using the Sander and Pmemd modules of Amber 12,
running on the EURORAꢀCPU/GPU calculation cluster of the CINECA
supercomputer facility (Bologna, Italy). The entire MD simulation and data analysis
procedure was optimized by integrating Amber 12 in modeFRONTIER, a
multidisciplinary and multiobjective optimization and design environment.55
Supporting Information Available
Supporting Information is available free of charge via the Internet at
http://pubs.acs.org. and includes physical, spectroscopic and purity data of all
compounds, synthetic methods and description of the σ receptor binding assays,
cytotoxicity assay and of the molecular modeling methods.
Corresponding author
Bernhard Wünsch*
Institute of Pharmaceutical and Medicinal Chemistry, University of Münster,
Corrensstr. 48, Dꢀ48149 Münster, Germany
Tel.: +49ꢀ251ꢀ8333311; Fax: +49ꢀ251ꢀ8332144; Eꢀmail: wuensch@uniꢀmuenster.de
Acknowledgement
This work was supported by the Deutsche Forschungsgemeinschaft (DFG) and the
Cells-in-motion Cluster of Excellence (EXC, 1003ꢀCiM), University of Münster,
Germany.
ACS Paragon Plus Environment

Page 29 of 36
Journal of Medicinal Chemistry
29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Abbreviations
APCI, atmospheric pressure chemical ionization; DTG, diꢀ
oꢀtolylguanidine; EM, exact
mass; MM/PBSA, molecular mechanics/Poisson Boltzmann Surface Area; MTT , 3ꢀ
(4,5ꢀdimethylthiazolꢀ2ꢀyl)ꢀ2,5ꢀdiphenyltetrazolium bromide; PGRMCꢀ1, progesterone
receptor membrane component 1; PI, propidium iodide; SEM, standard error of the
mean.
ACS Paragon Plus Environment