Novel mouse GABA uptake inhibitors with enhanced inhibitory activity toward mGAT3/4 and their effect on pain threshold in mice

Article abstract of DOI:10.1016/j.ejmech.2019.111920

γ-Aminobutyric acid (GABA) uptake transporters are membrane transport proteins that are involved in the pathophysiology of a number of neurological disorders. Some types of chronic pain appear to result from the dysfunction of the GABAergic system. The deficiency of mouse GAT1 transporter (mGAT1) abolishes the nociceptive response, which means that mGAT1 inhibition is an appropriate medical approach to achieve analgesia. The mGAT4 transporter is the second most abundant GAT subtype in the brain; however, its physiological role has not yet been fully understood in the central nervous system. In this study, we examined whether the combination of mGAT1 and mGAT3/mGAT4 inhibition in a single molecule might lead to potentially synergistic effects improving analgesic activity to relieve neuropathic pain. To study this hypothesis, new GABA uptake inhibitors were designed, synthesized, and evaluated in terms of their activity and subtype selectivity for mGAT1-4. Among new functionalized amino acid derivatives of serine and GABA analogs, compounds with preferential mGAT3/4 inhibitory activity were discovered. Two selected hits (19b and 31c) were subjected to in vivo tests. We found a statistically significant antiallodynic activity in the von Frey test in diabetic and oxaliplatin-induced neuropathic pain model. The novel compounds (4-hydroxybutanoic, 4-hydroxypentanoic, and 4-aminobutanoic acid derivatives and serine analogs) provide new insights into the structure–activity relationship of mGAT3/mGAT4 inhibitors and indicate a new direction in the search for potential treatment of neuropathic pain of various origin.

Full text of DOI:10.1016/j.ejmech.2019.111920

Journal Pre-proof
Novel mouse GABA uptake inhibitors with enhanced inhibitory activity toward
mGAT3/4 and their effect on pain threshold in mice
Paula Zaręba, Beata Gryzło, Katarzyna Malawska, Kinga Sałat, Georg C. Höfner,
Alicja Nowaczyk, Łukasz Fijałkowski, Anna Rapacz, Adrian Podkowa, Anna Furgała,
Paweł Żmudzki, Klaus T. Wanner, Barbara Malawska, Katarzyna Kulig
PII:
S0223-5234(19)31072-4
DOI:
https://doi.org/10.1016/j.ejmech.2019.111920
EJMECH 111920
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 7 September 2019
Revised Date: 16 November 2019
Accepted Date: 27 November 2019
Please cite this article as: P. Zaręba, B. Gryzło, K. Malawska, K. Sałat, G.C. Höfner, A. Nowaczyk, Ł.
Fijałkowski, A. Rapacz, A. Podkowa, A. Furgała, Paweł. Żmudzki, K.T. Wanner, B. Malawska, K. Kulig,
Novel mouse GABA uptake inhibitors with enhanced inhibitory activity toward mGAT3/4 and their effect
on pain threshold in mice, European Journal of Medicinal Chemistry (2020), doi: https://doi.org/10.1016/
j.ejmech.2019.111920.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2019 Published by Elsevier Masson SAS.

Title: Novel mouse GABA uptake inhibitors with enhanced inhibitory activity toward mGAT3/4
and their effect on pain threshold in mice
1
*
1
1
1
2
3
Paula Zar
ęba , Beata Gryzło , Katarzyna Malawska , Kinga Sałat , Georg C. Höfner , Alicja Nowaczyk ,
3
1
1
1
1
Łukasz Fijałkowski , Anna Rapacz , Adrian Podkowa , Anna Furgała , Paweł
Ż
mudzki , Klaus T.
2
1
1
Wanner , Barbara Malawska , Katarzyna Kulig (†).
1
Faculty of Pharmacy, Jagiellonian University Medical College, 9 Medyczna St, 30–688 Kraków, Poland,
2
Department of Pharmacy, Center for Drug Research, Ludwig-Maximilians-Universität München,
3
Butenandtstr, 5-13, 81377 Munich, Germany, Faculty of Pharmacy, Collegium Medicum in Bydgoszcz,
Nicolaus Copernicus University, 2 dr. A. Jurasza St, 85-094, Bydgoszcz, Poland
Correspondence:
Paula Zaręba
9
Medyczna Street, 30–688 Kraków, Poland
E-mail: paula.zareba@uj.edu.pl
Tel.: +4812 6205464; fax: +4812 6205450
Acknowledgments
This study was financially supported by the National Science Center in Poland as part of grants
2
014/15/B/NZ7/00930 and 2015/16/T/NZ7/00032 and the Jagiellonian University Medical College grant
K/DSC/001419.
Abstract
γ-Aminobutyric acid (GABA) uptake transporters are membrane transport proteins that are involved in the
pathophysiology of a number of neurological disorders. Some types of chronic pain appear to result from
the dysfunction of the GABAergic system. The deficiency of mouse GAT1 transporter (mGAT1)
abolishes the nociceptive response, which means that mGAT1 inhibition is an appropriate medical
approach to achieve analgesia. The mGAT4 transporter is the second most abundant GAT subtype in the
brain; however, its physiological role has not yet been fully understood in the central nervous system. In
this study, we examined whether the combination of mGAT1 and mGAT3/mGAT4 inhibition in a single
molecule might lead to potentially synergistic effects improving analgesic activity to relieve neuropathic
pain. To study this hypothesis, new GABA uptake inhibitors were designed, synthesized, and evaluated in
terms of their activity and subtype selectivity for mGAT1-4. Among new functionalized amino acid
1

derivatives of serine and GABA analogs, compounds with preferential mGAT3/4 inhibitory activity were
discovered. Two selected hits (19b and 31c) were subjected to in vivo tests. We found a statistically
significant antiallodynic activity in the von Frey test in diabetic and oxaliplatin-induced neuropathic pain
model. The novel compounds (4-hydroxybutanoic, 4-hydroxypentanoic, and 4-aminobutanoic acid
derivatives and serine analogs) provide new insights into the structure–activity relationship of
mGAT3/mGAT4 inhibitors and indicate a new direction in the search for potential treatment of
neuropathic pain of various origin.
Keywords: GABA transporter, mGAT1, mGAT3, mGAT4, GAT inhibitors, N-benzylamides, biological
evaluation, molecular docking, neuropathic pain, analgesic activity, nociception
1
Introduction
The γ-aminobutyric acid (GABA) uptake transporters, which belong to the solute carrier (SLC6) group of
membrane transport proteins, are promising targets for the treatment and investigation of central nervous
system (CNS) pathophysiology. The decrease in the inhibitory post-synaptic response of the GABAergic
1–4
system in the CNS is the underlying cause of many neurological disorders such as epilepsy and
5
,6
neuropathic pain (NP). According to 2019 data from WHO, more than 50 million people in the world
7
suffer from various types of epilepsy. It is known that dysregulation of GABA neurotransmission is the
cause of this disease, and current therapies that focus on GABA uptake modulation show promising
results. Another example of a disease associated with GABAergic dysfunction is NP. Its consequences
8
include lowering of quality of life, general mood, and loss of occupational functioning. According to the
9
current data, it is estimated that between 3% and 17% of the world’s population is suffering from NP. It is
postulated that the development of NP is associated with the dynamic imbalance between the inhibitory
1
0–12
and excitatory neurotransmitters in the CNS, both in the brain and the spinal cord.
However, because
of the complex mechanism of neuropathic pain development, currently used therapies can only reduce
13–15
symptoms in approximately 30–40% of patients suffering from this chronic pain type.
In the CNS, GABA (1) is the most widely distributed inhibitory neurotransmitter that regulates neuronal
excitability and the nociceptive threshold. Enhancement of GABA action and the prolongation of its
activity can be achieved, for example, through the inhibition of its uptake. To date, four subtypes of
membrane-bound GABA transporters have been identified and the nomenclature commonly used is based
on the International Union of Basic and Clinical Pharmacology (IUPHAR) and include GAT1 (SLC6A1),
1
6
GAT2 (SLC6A13), GAT3 (SLC6A11), and BGT1 (SLC6A12). Human and rat GAT1, BGT1, GAT2,
1
and GAT3 correspond to mouse mGAT1, mGAT2, mGAT3, and mGAT4, respectively. For the purpose
2

of this manuscript, the mouse nomenclature (mGAT1–mGAT4) is used due to mouse GABA transporters
were used in in vitro assays.
As presented in Figure 1, GAT inhibitors representing a large degree of structural variety. Studies aiming
to describe GAT inhibitors' subtype selectivity relationships showed that N-substituted nipecotic acid
17,18
19
20
analogs (e.g. tiagabine (2) ), guvacine analogs (e.g. DDPM-2571 (5) and NO-711 (6) ) and muscimol
21
analogs (EF-1502 (7) ) display mGAT1 selectivity. Replacing the 4,4-diphenyl-3-butenyl (DPB) residue
of the lipophilic compounds with the 2-(tris(4-methoxyphenyl)methoxy)ethyl has led to the discovery of
22
22
(
S)-SNAP-5114 (3)), and its chemically more stable analogue, DDPM-859 (4) which showed
preference toward mGAT3/4 and mGAT4, respectively. mGAT4 inhibitor without an amino acid structure
23
element (10) has been identified in a group of 1H-indole-2,3-dione derivatives. The obtained compound
1
0 was approximately 170-fold more active against mGAT4 than mGAT1. It is worth highlighting that the
above group of compounds is characterized by non-competitive GABA uptake activity. Some studies have
suggested that the GABA uptake inhibitors, not belonging to structural GABA analogs, may have a
different binding site to the mGAT4 transporter than its natural substrate GABA (1). Other studies have
demonstrated that shortening of the aliphatic chain of the
γ-aminobutanoic acid (1) and the introduction of
the hydroxyl group at position C2 results in increased inhibitory activity of isoserine (12) toward the
24,25
transporters of subtypes 3 and 4 in comparison to the activity exhibited by GABA (1).
On the basis of
the presented structures, a general conclusion may be drawn indicating that unsubstituted ligands with an
2
3
26
25
aliphatic main chain shorter than that in GABA (1), such as β-alanine (11), isoserine (12), or 2,3-
24
diaminopropanoic acid (DAPA, 13) are mGAT inhibitors with noticeable preference toward subtype 3
and/or 4 (Figure 1).
It should be emphasized that GABA uptake inhibitors such as tiagabine (2), (S)-SNAP-5114 (3), or NO-
27–29
7
11 (6), initially designed as anticonvulsants, show analgesic activity in animal models.
Currently,
30–36
mGAT1 inhibitors are known to possess anticonvulsant and analgesic activities.
In addition, it is
postulated that mGAT4 also is involved in nociceptive transmission. Its inhibition may modulate
29
37
nociceptive threshold, and can reduce sleep disorders in patients with NP. One example is (S)-SNAP-
29
5
114 (3) which was demonstrated its antinociceptive activity in the acute and NP models. Thus mGAT1
and mGAT4 inhibitors are considered to be an important pharmacological tool to study the mechanisms
38–42
of NP.
Finally, the expression of mGAT2 and mGAT3 is very limited in the brain. They are found in
43
multiple peripheral organs, and their physiological role has not yet been fully understood.
3

S
S
N
N
O
N
N
OH
OH
N
O
O
S
N
OH
S
N
Cl
OH
O
O
O
Cl
Tiagabine (2)
pIC50 mGAT1: 6.88 mGAT3: 64 %(100uM)
DDPM-2571 (5)
NO-711 (6)
EF1502 (7)
pIC50 mGAT1: 8.27
mGAT3: 4.35
mGAT4: 4.07
pIC50 mGAT1: 6.83 mGAT3: 3.62
pIC50 mGAT1: 5.15 mGAT3: <3.52
mGAT2: 4.59 mGAT4: <3.52
mGAT2: 50 %(100uM) mGAT4: 73 %(100uM)
mGAT2: 4.31
mGAT2: 3.20 mGAT4: 3.07
mGAT1 inhibitors
mGAT3 inhibitors
mGAT2 inhibitor
OH
O
OH
O
Cl
H2N
O
Cl
O
N
N
H2N
N
N
H
H
GABA (1)
OH
8
NH2
pIC50 mGAT1: 4.52 mGAT3: 5.15
mGAT2: n.d. mGAT4: 4.48
9
-alanine (11)
pIC50 mGAT1: 4.89 mGAT3: 5.00
pIC50 mGAT1: 4.72 mGAT3: 5.14
pIC50 mGAT1: 2.53 mGAT3: 4.18
mGAT2: 5.19 mGAT4: 5.04
mGAT2: 4.90 mGAT4: 5.00
mGAT2: 2.96 mGAT4: 3.96
mGAT3/4 inhibitors
mGAT4 inhibitors
O
O
O
O
OH
OH
O
N
O
S
H2N
O
H N
O
N
O
2
O
O
O
OH
N
H
NH2
OH
OH
(S)-SNAP-5114 (3)
O
O
Isoserine (12)
DAPA (13)
DDPM-859 (4)
10
pIC50 mGAT1: 4.07
mGAT2: 56 % mGAT4: 5.71
100uM)
mGAT3: 5.29
pIC50 mGAT1: 2.60 mGAT3: 5.40 pIC50 hGAT-1: < 3 hGAT-2: 4.4
mGAT2: 3.64 mGAT4: 5.30
hBGT-1: 3.2 hGAT-3: 4.5
pIC50 mGAT1: 4.19 mGAT3: 4.85
mGAT2: 4.12 mGAT4: 5.78
pIC50 mGAT1: < 3
mGAT3: 3.74
mGAT2: 3.21 mGAT4: 5.20
(
Figure 1. Structures of GAT inhibitors and their inhibitory activity toward GABA uptake transporters
2
6
19
22
22
19
(
(
pIC ): GABA (1), tiagabine (2), (S)-SNAP-5114 (3), DDPM-859 (4), DDPM-2571 (5), NO-711
50
18
21
44
45
23
26
25
24
6), EF1502 (7), 8, 9, 10, 11, 12, and 13. n.d. – not described.
Considering that the pharmacological importance of mGAT3/4 has not yet been fully elucidated, it is
important to design new molecules with such an activity profile. Recently, we reported new GAT
inhibitors among 4-hydroxy- and 4-aminobutanoic N-benzylamides bearing an aromatic 4,4-
44,45
diphenylbutan-1-amine domain.
Our previous studies have led to the identification of compound 8
with a chlorine substituent at the ortho position of the benzyl fragment of the molecule showed elevated
44
potency against mGAT2 (pIC50 = 5.19 ± 0.01), whereas its structural analog, compound 9, exhibited
45
higher potency against mGAT3 (pIC = 5.14 ± 0.10) (Figure 1). As presented in Figure 2, in the design
50
of the novel compounds we selected 8 and 9 as parent compounds. We were interested in an isosteric
replacement of the phenyl rings with thiophene. Thus, based on the finding that small amino acids such as
20,46,47
isoserine (12) or 2,3-diaminopropionic acid (13) are GABA uptake inhibitors,
we focused on the
modification in the carbon atom length in the core of 8 and 9. It was assumed that the lipophilic
substituent, the 4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl, at the 2-position of serine, 4-hydroxy- or 4-
amino-butyric and pentanoic acids should be the core structure of new compounds, while the N-substituent
4

of the amide group should be the variable fragment of the molecules. Various substituted N-benzylamines
were used to obtain appropriate amides, which was based on our previous experience that the benzyl
49–52
moiety is key for GABA uptake inhibitors.
Moreover, other groups such as 2-phenylethyl, 1-
phenylethan-1-amine, or morpholine were introduced to compare the importance of the N-benzylamine
substituent for GAT inhibitory activity. It is worth mentioning that the 4,4-bis(3-methylthiophen-2-yl)but-
3
-en-1-yl substituent is a characteristic group for the tiagabine structure (2), which has been shown to be
potent in some mouse models of NP. The main objective of this study was to determine whether the
inhibition of GABA uptake could lead to the attenuation of pain sensitivity. In our previous research we
focused on the selective GAT1 inhibition as an effective method to relieve NP symptoms and we showed
38
that a selective GAT1 inhibitor, tiagabine (2), has antiallodynic properties in mouse models of NP. In the
present study we investigated whether the combination of mGAT1 and mGAT3/mGAT4 inhibition in a
single molecule can lead to improving the analgesic effect in NP conditions. It was believed that the
designed pharmacological experiments would allow to determine the potential relationship between the in
vitro inhibitory potency for mGAT1-4 and the effect observed in vivo. All final compounds were
investigated for their inhibitory potency at four mouse GABA transport proteins (mGAT1-4) in in vitro
tests. Then, two selected compounds were subjected to preliminary in vivo tests to evaluate their
antinociceptive properties in rodent models of NP. Here, their activity is discussed with the reference to
3
8
48
the previously obtained data for tiagabine, and pregabalin. The metabolic (diabetic) painful peripheral
neuropathy induced by streptozotocin (STZ) and the toxic, chemotherapy-induced peripheral neuropathy
caused by an antitumor drug oxaliplatin were selected as models of NP in mice. The influence of tested
compounds on mechanical and thermal (heat or cold) nociceptive threshold was assessed. Additionally,
considering the potential involvement of mGAT1 and mGAT3/mGAT4 in NP etiology, we performed in
49
silico study using molecular docking techniques. Finally, the structure–activity relationships of the new
compounds obtained were determined.
Figure 2. The modification of parent compounds (8 and 9) and a general structure of target compounds.
5

2
2
Results and discussion
Chemistry
.1
2
.1.1
Synthesis of the 4-hydroxybutanoic acid derivatives (19a–f, 20–22) and the 2-substituted
derivatives of 4-hydroxypentanoic N-benzylamides (23a–e)
To prepare designed 4-hydroxybutanoic acid derivatives 19a–f and 20–22 and 4-hydroxypentanoic N-
benzylamides 23a–e, the nucleophilic substitution reaction between N-methyl-4,4-bis(3-methylthiophen-
5
0–53
2
-yl)but-3-en-1-amine
(14) and 3-bromodihydrofuran-2(3H)-one (15) or 3-bromo-5-
45
methyldihydrofuran-2(3H)-one (16) was first used to obtain compound 17 or 18 (Scheme 1). Then, the
27,45,54,55
final compounds were obtained in accordance with the previously described synthetic procedures.
Even though a simple modification was developed. Thus, the aminolysis of compound 17 or 18 was
performed with N-benzylamine and its derivatives (Scheme 1). The reaction was carried out under argon
atmosphere in refluxing dry THF. The crude product was purified by column chromatography to yield
1
9a–f, 20–22, and 23a–e as a yellow oils. Compounds 23a–e were obtained as a mixture of racemic
diastereomers 2RS,4RS-23 and 2RS,4SR-23 which were isolated in pure form. The diastereoselectivity (ds)
amounted to approximately 7:3 (see the experimental section).
Scheme 1. Synthesis of 2-substituted derivatives of 4-hydroxybutanamide (19a–f and 20–22) and
4
-hydroxypentanoic N-benzylamides (23a-e). Reagents and conditions: (a) TBAB, K CO , CH CN, 15
2 3 3
min at 0 °C and 16 h at rt, (b) argon, dry THF, reflux, 48 h.
2
.1.2
Synthesis of the 4-aminobutanoic acid derivatives (27a–e)
4
-Aminobutanoic acid derivatives (27a–e) were prepared according to the synthetic route shown in
Scheme 2. Compound 24 was obtained in accordance with the previously described synthetic
44,56
procedures.
The amide bond formation step to generate N-benzyl-2-bromo-4-(1,3-dioxoisoindolin-2-
44
yl)butanamide derivatives (25a–e) was performed in the presence of n-propanephosphonic acid
6

57
anhydride (T P), triethylamine (TEA) and appropriate N-benzylamine. Then, the N-alkylation of 14 with
3
2
5a–e yield corresponding amides library 26a–e. Finally, the hydrazinolysis of 26a–e led to obtaining 4-
aminobutanoic acid derivatives 27a–e as a yellow oil. Additionally, it was decided to study the effect of
substitution of the primary, aliphatic amino group at the 4-position of butanoic acid (27b) with an acetic
acid residue. In order to obtain 2-substituted 4-acetamide derivative of N-(2-chlorobenzyl)butanamide
(
28), the acetic acid was converted into a reactive acylation agent with dicyclohexylcarbodiimide (DCC)
58
which had been nucleophilically substituted with an amine residue of 27b (Scheme 2). The product was
purified by column chromatography to yield 28 as a yellow oil.
Scheme 2. Synthesis of 4-aminobutanoic acid derivatives (27a–e and 28). Reagents and conditions: (a)
T P, DCM, 2 h at 0 °C and 12 h at rt, (b) KI, K CO , MeCN, reflux, 24 h, (c) NH NH (50%–60%), EtOH,
3
2
3
2
2
2
h at 60 °C and 5 h at rt, (d) DMAP, DCC, CH COOH, DCM, 10 min at 0 °C and 20 h at rt.
3
2
.1.3
Synthesis of the serine analogs 31a–f
59
To obtain serine analogs 31a–f, a simple and efficient synthetic pathway was used (Scheme 3). Desired
amides 30a–f were obtained after the activation of acid 29 by n-propanephosphonic acid anhydride (T P).
3
The preparation of 2-substituted derivatives of serine 31a–f was accomplished by the alkylation of 14 by
2
-bromo-3-hydroxypropanoic acid N-benzylamides 30a–f in dry DMF with N,N-diisopropylethylamine
DIPEA) and tetra-n-butylammonium bromide (TBAB). Finally, the designed compounds 31a–f were
purified by column chromatography and obtained as a colorless oil.
(
7

Scheme 3. Synthesis of 2-substituted derivatives of serine (31a–f). Reagents and conditions: (a) T P,
3
TEA, N-benzylamine derivative, dry DCM, −17 °C at 30 min and 1 h at rt, argon; (b) DIPEA, TBAB,
DMF , reflux, 12 h.
dry
2
.2
In vitro evaluation and structure–activity relationship
The inhibitory potencies of all the obtained final compounds (19a–f, 23a–e, 27a–e, 28, and 31a–f,) were
determined for the four mouse GABA transporter subtypes (mGAT1 to mGAT4). The determination was
3
performed through [ H]GABA uptake using human embryonic kidney cells (HEK-293) stably expressing
1
8
the mouse GABA transporters according to the literature. The specific affinity to mGAT1 was
determined by a competitive MS Binding Assay with NO-711 (6) as a nonlabeled marker quantified by
60
LC–ESI–MS–MS. The compounds that could reduce the GABA uptake or NO-711 binding at least by
5
i
0% at an inhibitor concentration of 100 µM were considered as active and the pIC50 or pK values were
3
determined in [ H]GABA uptake or MS Binding Assays in triplicate in three independent experiments. If
3
the test compounds at a screening concentration of 100 µM could not reduce [ H]GABA uptake or NO-
3
7
11 binding below 50% (pIC50 = 4.00), the percentage of the remaining [ H]GABA uptake or NO-711
binding was given (in the presence of 100 µM of the inhibitor).
Newly synthesized 4-hydroxybutanoic acid derivatives (19a–f) shows a comparable inhibitory potency
with their parent compounds 4-hydroxy and 4-aminobutanoic acid derivatives having the 4,4-diphenylbut-
44,45
3
-en-1-yl group (8 and 9),
in the favor of mGAT3/4 (Table 1). Importantly, in the group of 4-
hydroxybutanoic acid (GHB) derivatives (19a-f), the replacement of the N-benzylamide moiety with a
phenylethylamide (20) or N-(1-phenylethyl)amide unit (21) resulted in a decrease in biological activity. In
addition, it was found that a morpholine moiety (22) cannot be tolerated, leading to losing the inhibitory
potency toward mGAT1 and mGAT2 but showed an improved selectivity toward mGAT3 and mGAT4
(
Table 1, Figure 3). Thus, the obtained data clearly emphasizes that the N-benzylamide moiety is an
important pharmacophore element of GAT inhibitors.
The replacement of the hydroxyl functionality with the amino group in the 4-position of the butanoic acid
(
27a–e) did not change the profile of inhibitory potency against mGAT1–4 (Table 1, Figure 3).
Interestingly, it was found that a chlorine substituent in the ortho position in the N-benzylamide moiety of
GHB and GABA, compounds 19b and 27b, slightly increased the inhibitory potency toward mGAT3
(
pIC = 5.31 ± 0.04 and 5.44 ± 0.08, respectively). Thus, compound 27b was selected for acylation of the
50
amino group in the 4-position. As a result, the obtained compound 28 showed a 1.5-fold higher
8

3
[
H]GABA inhibitory activity (pIC50 = 5.61 ± 0.07) toward mGAT3 than its analog 27b. Finally,
compound 28 was identified to exhibit the highest inhibitory potency against mGAT3 in the group of
compounds obtained, even when compared to the reference compounds (S)-SNAP-5114 (3) (pIC50 = 5.29
±
0.04, mGAT3) and DDPM-859 (4) (pIC = 4.85 ± 0.04, mGAT3, Table 1, Figure 3).
50
3
i
Table 1. Inhibitory potencies (pIC50 ± SEM) toward mGAT1–4 obtained from [ H]GABA uptake experiments and binding affinities (pK ± SEM)
toward mGAT1 from MS Binding Assays of obtained compounds.
pIC50 *± SEM
mGAT2
i
pK * ± SEM
mGAT1
60%
1
2
3
Compound
R
R
R
mGAT1
4.80 ± 0.08
4.86 ± 0.10
5.03 ± 0.09
4.81 ± 0.06
4.89 ± 0.01
4.91 ± 0.04
4.75 ± 0.07
4.52 ± 0.03
57%
4.88
4.63
51%
47%
63%
4.24
4.18
4.07
mGAT3
5.00 ± 0.04
5.31 ± 0.04
5.16 ± 0.11
5.05 ± 0.06
5.18 ± 0.07
5.20 ± 0.09
4.92 ± 0.04
5.04 ± 0.07
4.49 ± 0.10
5.23 ± 0.11
5.07 ± 0.12
5.44 ± 0.10
5.37 ± 0.09
4.80
mGAT4
4.99 ± 0.05
5.24 ± 0.05
5.13 ± 0.03
4.96 ± 0.03
5.15 ± 0.08
5.06 ± 0.04
4.79 ± 0.05
4.86 ± 0.05
4.48 ± 0.08
5.41 ± 0.07
4.73
5.02 ± 0.09
5.02 ± 0.04
4.54
4.88
4.75
1
1
9a
9b
H
2-Cl
4-Cl
4-F
4-Me
2,4-F
I
II
III
H
H
2-Cl
2-Cl
4-Cl
4-Cl
4-F
4-F
4-Me
4-Me
H
2-Cl
4-Cl
4-F
4-Me
2-Cl
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
4.69 ± 0.07
4.98 ± 0.05
5.14 ± 0.09
5.25 ± 0.12
5.01 ± 0.09
5.00 ± 0.10
4.76 ± 0.08
4.88 ± 0.10
76%
5.05
4.82
4.98 ± 0.09
5.03 ± 0.02
4.98
77%
81%
70%
58%
71%
73%
68%
84%
82%
59%
85%
88%
90%
91%
78%
70%
1
9c
9d
9e
9f
1
1
1
2
2
2
0
1
2
(
(
(
(
(
(
(
(
2RS,4RS)-23a
2RS,4SR)-23a
2RS,4RS)-23b
2RS,4SR)-23b
2RS,4RS)-23c
2RS,4SR)-23c
2RS,4RS)-23d
2RS,4SR)-23d
2RS,4RS)-23e
2RS,4SR)-23e
4.72
4.79
4.58
4.45
5.04 ± 0.06
4.47
4.53
(
(
4.91
4.60
4.77
4.64
5.24 ± 0.09
5.17 ± 0.03
5.20 ± 0.09
5.44 ± 0.08
5.10 ± 0.06
5.17 ± 0.09
5.10 ± 0.06
5.61 ± 0.07
5.26 ± 0.02
5.15 ± 0.06
5.18 ± 0.03
5.25 ± 0.04
5.09 ± 0.06
5.12 ± 0.05
5.16 ± 0.08
5.21 ± 0.05
79%
72%
2
2
7a
7b
NH
NH
NH
NH
NH
2
4.94 ± 0.07
5.12 ± 0.08
5.03 ± 0.05
4.93 ± 0.06
4.94 ± 0.01
5.06 ± 0.02
5.26 ± 0.10
5.14 ± 0.07
5.03 ± 0.09
5.10 ± 0.03
5.11 ± 0.10
5.35 ± 0.08
4.61 ± 0.08
4.93 ± 0.05
4.70 ± 0.06
4.66 ± 0.05
4.74 ± 0.08
68%
2
2
2
2
H
H
H
H
2
7c
7d
7e
2
2
2
8
NH(CO)CH
3
H
pIC50* ± SEM
i
pK * ± SEM
mGAT1
46%
96%
66%
4.29 ± 0.09
4.41 ± 0.06
92%
1
Compound
R
mGAT1
4.45 ± 0.09
61%
4.69 ± 0.06
4.82 ± 0.10
4.70 ± 0.04
4.87
mGAT2
4.51 ±0.04
65%
4.80 ± 0.01
4.65 ± 0.09
4.62 ± 0.10
4.88
mGAT3
4.67 ± 0.04
4.16 ± 0.01
4.83 ± 0.06
4.80 ± 0.08
5.02 ± 0.09
4.72
mGAT4
4.68 ± 0.09
4.11 ± 0.03
5.02 ± 0.03
4.77 ± 0.05
5.02 ± 0.05
4.77
3
3
1a
1b
H
2-Cl
4-Cl
4-F
4-Me
4-CF
3
3
1c
1d
1e
1f
3
3
3
1
8
61
Tiagabine (2)
6.88 ± 0.12
4.07 ± 0.09
52%
64%
73%
5.81 ± 0.10
7.43 ± 0.11
–
1
8
(
S)-SNAP-5114 (3)
56%
5.29 ± 0.04
2
2
–
DDPM-859 (4)
4.19 ± 0.07
4.12 ± 0.08
4.85 ± 0.04
5.78 ± 0.03
±
1
9
DDPM-2571 (5)
8.27 ± 0.03
6.83 ± 0.06
5.15
4.89 ± 0.10
4.72 ± 0.07
4.31
3.20 ± 0.09
4.59
5.19 ± 0.01
4.90 ± 0.13
4.35
3.62 ± 0.04
<3.52
5.00 ± 0.04
5.14 ± 0.10
4.07
3.07 ± 0.05
<3.52
5.04 ± 0.06
5.00 ± 0.04
3
8.29 ± 0.02
**
n.d.
4.96 ± 0.09
54%
1
8
NO-711 (6)
21
EF1502 (7)
4
4
4
5
8
9
*
Data are given as mean ± SEM of three independent experiments that were performed in triplicate. The percent result represents the [ H] GABA uptake or
NO-711 binding in the presence of 100 ꢀM inhibitor. Data without SEM implies that only one experiment was performed in triplicate. n.d. – not described,
9

6
2
d
**K = 26.5 ± 4.6 nM
6
5
5
5
5
5
4
4
4
4
4
.0
.8
.6
.4
.2
.0
.8
.6
.4
.2
.0
pIC50 (mGAT1)
pIC50 (mGAT2)
pIC50 (mGAT3)
pIC50 (mGAT4)
Figure 3. Bar chart showing pIC50 values of compounds 19a–f, 20–22, 27a–e, and 28 compared to the
values of reference compounds, (S)-SNAP-5114 (3) and DDPM-859 (4). Columns without error mean that
only one experiment was performed in triplicate. For the purpose of clarity, the activity of 22 and (S)-
SNAP-5114 (3) against mGAT1-2 and mGAT2 (pIC50 ≤ 4.00), respectively, was not presented on the bar
chart.
In the group of serine derivatives 31a–f, the activity started at comparatively low values (Table 1). The
dominant inhibitory potency was observed for 31c and 31e against mGAT3/4 (Table 1, Figure 4).
Unfortunately, 31c and 31e showed a significant decline in the biological activity as compared to their
homologs, 4-hydroxybutanamides 19c and 19e (Table 1, Figure 3 and 4). Surprisingly, it was found that a
derivative having chlorine in the 2-position of the N-benzylamide moiety showed a negative effect on the
activity. Compound (31b) exhibited a significant decrease in inhibitory potency toward mGAT3 and
mGAT4 and a total loss of inhibitory activity toward mGAT1 and mGAT2. Moreover, the same
unfavorable effect on the inhibitory activity was observed for the morpholine derivative in a group of 4-
hydroxybutanamide (22). It could be concluded that the reduction in the chain length of the molecule
leading to serine derivatives (31a–f) was not beneficial for GABA uptake inhibitory activity, as compared
to the butane analogs (19a–f). On the other hand, the elongation of the main carbon atom chain length to
five carbon atoms (23a–e) was found to be more tolerated than that of serine derivatives but did not
improve the activity as compared to the GABA analogs (27a–e; Table 1, Figure 4). Importantly,
(
2RS,4RS)-23b and (2RS,4SR)-23b showed an interesting subtype selectivity in the favor of mGAT3
1
0

transporter. (2RS,4RS)-23b showed 2.88 times higher preference to inhibit mGAT3 than mGAT2 and 2.66
times higher preference to inhibit mGAT3 and mGAT4. Furthermore, (2RS,4RS)-23b showed the same
pIC50 value as its structural analog 27b with a chlorine atom in the ortho position and primary amino
group in the 4-position (pIC50 = 5.44 ± 0.10 and 5.44 ± 0.08, respectively), while their analog 19b
possessing the hydroxyl group in the 4-position found to be slightly weaker mGAT3 inhibitor (pIC =
5
0
5
.31 ± 0.04). Taken together, the compounds discussed above presented a different profile of activity as
compared to tiagabine (2). Nevertheless, it was again confirmed that the 4,4-bis(3-methylthiophen-2-
yl)but-3-en-1-yl)(methyl)amine moiety is the source of increased potency of the GAT inhibitors.
6.0
5.8
5.6
5.4
5.2
5.0
4.8
4.6
4.4
4.2
4.0
pIC50 (mGAT1)
pIC50 (mGAT2)
pIC50 (mGAT3)
pIC50 (mGAT4)
Figure 4. Bar chart showing pIC50 values of the 2-substituted derivatives of serine (31a–f) and 2-
substituted derivatives of 4-hydroxypentanoic N-benzylamides ((2RS,4RS)-23a–e and ((2RS,4SR)-23a–e)
compared to the values of reference compounds, (S)-SNAP-5114 (3) and DDPM-859 (4). Columns
without error imply that only one experiment was performed in triplicate. For the purpose of clarity, if
compounds showed pIC50 ≤ 4.00 toward individual GAT transporters subtype, the activity was not
presented on the bar chart.
2
.3
In vivo evaluation
Two compounds 19b and 31c have been selected for in vivo studies based on their inhibitory activities
against mGATs. Compounds 19b and 31c have displayed balanced activity as inhibitors of mGAT1
(
pIC50 = 4.86 and 4.69, respectively) and mGAT3/4 (pIC50 = 5.31/5.24 and 4.83/5.02, respectively). Thus,
they showed the most interesting and appropriate in vitro profile for the verification of the postulated
hypothesis.
1
1

We investigated their antiallodynic and antihyperalgesic properties 1 h after intraperitoneal administration
and we used two mouse models of chronic pain: NP induced by STZ and NP induced by oxaliplatin
administration. To induce type I diabetes and diabetic NP, mice were first administered intraperitoneally
with STZ. Then, 3 weeks later, i.e. during the advanced stage of diabetes-induced neuropathy, the effect of
the test compounds on tactile allodynia in diabetic neuropathic mice was evaluated. For this purpose, the
6
3
von Frey test was used. Moreover, the effect of 19b and 31c on heat hyperalgesia in diabetic neuropathic
63
mice was assessed using the hot plate test.
To induce chemotherapy-induced peripheral neuropathy (CIPN), oxaliplatin was administered to mice.
This model of pain allowed to determine the effect of test compounds on tactile allodynia in oxaliplatin-
treated neuropathic mice in von Frey test, as well as the effect on cold allodynia in oxaliplatin-treated
64
mice in the cold plate test.
2
.3.1
STZ-induced diabetic neuropathy: influence on tactile allodynia (von Frey test)
In the von Frey test, an overall effect of the treatment with compound 19b was observed (F[6,63]=33.54;
p < 0.0001). Compared to vehicle-treated mice, STZ significantly (p < 0.0001) reduced mechanical
nociceptive threshold in diabetic mice, and treatment with 19b at doses 10 and 30 mg/kg significantly
(
p < 0.01 and p < 0.0001, respectively) attenuated tactile allodynia. The dose 5 mg/kg was not effective in
this assay (Figure 5A). An overall effect of treatment was also noted for compound 31c in STZ-treated
mice (F[6,63]=49.14; p < 0.0001). STZ significantly (p < 0.0001 vs. vehicle-treated normoglycemic mice)
reduced mechanical nociceptive threshold in diabetic mice, and treatment with 31c at the dose 30 mg/kg
significantly (p < 0.0001) attenuated this effect. The doses 5 and 10 mg/kg were not effective in this assay
(
Figure 5B).
1
2

Figure 5. Effect of the test compounds 19b (Figure 5A) and 31c (Figure 5B) on tactile allodynia measured
in the von Frey test in diabetic, STZ-treated neuropathic mice (n=10). Results are shown as the mean force
(
± SEM) applied to elicit paw withdrawal. Statistical analysis: one-way ANOVA followed by Tukey’s
####
post hoc test.
p < 0.0001 (vs. paw withdrawal force of normoglycemic control) and
*
*p < 0.01, and ****p < 0.0001 (vs. pre-drug paw withdrawal force of STZ-treated mice).
Based on the results obtained here and in the previous studies, we could compare the antiallodynic activity
of compounds 19b and 31c to that of two reference drugs, i.e. tiagabine and pregabalin. These previous
experiments were performed under the same laboratory conditions and a detailed description of the results
38
obtained is shown in assessed antiallodynic properties of tiagabine in the von Frey test and the results
obtained previously for pregabalin, an α δ calcium channel ligand approved for the treatment of diabetic
2
NP, revealed that pregabalin at doses 10 and 30 mg/kg induced a significant (p < 0.001) elevation of pain
6
3
thresholds in STZ-treated mice. In the STZ-induced NP model, none of the test compounds was more
effective than tiagabine, but 19c reduced tactile allodynia at doses equal to those of pregabalin.
1
3

2
.3.2
STZ-induced diabetic neuropathy: influence on heat hyperalgesia (hot plate test)
The hot plate test used to determine the effect of compounds 19b and 31c (used at doses reducing tactile
allodynia) on heat hyperalgesia revealed a significant effect of treatment with STZ and 19b or 31c on heat
nociceptive threshold (F[4,43]=4.5, p < 0.01 and F[4,45] = 3.405, p < 0.05, respectively). However, the
comparison of pre-drug and post-drug latencies to pain reaction measured in diabetic mice did not show
any significant effect of 19b or 31c on heat hyperalgesia (Figure 6A, Figure 6B, respectively), while
38
63
reference drugs: tiagabine and pregabalin were effective.
Figure 6. Effect of the test compounds 19b (Figure 6A) and 31c (Figure 6B) on heat hyperalgesia
measured in the hot plate test in diabetic, STZ-treated neuropathic mice (n=9-10). Results are shown as
the mean latency to pain reaction (± SEM). Statistical analysis: one-way ANOVA followed by Tukey’s
#
##
post hoc test. p < 0.05 and p < 0.01 (vs. latency of normoglycemic control).
1
4

2
.3.3
Oxaliplatin-induced peripheral neuropathy: effect on tactile allodynia (von Frey test)
Compounds 19b and 31c at doses active in a diabetic NP model were also tested in a mouse model of NP
induced by oxaliplatin. For compound 19b, an overall effect of the treatment was demonstrated in the von
Frey test (F[8,69]=32.82, p < 0.0001). The administration of oxaliplatin significantly (p < 0.0001 vs.
vehicle-treated non-neuropathic mice) decreased the paw withdrawal force, i.e., it increased pain
sensitivity of mice both in the acute and late phases of neuropathy (Figure 7A). Intraperitoneal 19b, at
doses 10 and 30 mg/kg, effectively reduced acute-phase tactile allodynia (significant at p < 0.0001 vs. pre-
drug paw withdrawal force of oxaliplatin-treated mice). In contrast, only the dose 30 mg/kg was effective
in
the
late
phase
(
p < 0.001 vs. pre-drug paw withdrawal force). For compound 31c, an overall effect of the treatment was
determined by the von Frey test (F[4,35]=66.87, p < 0.0001). The administration of oxaliplatin
significantly (p < 0.0001 vs. non-neuropathic control) decreased the paw withdrawal force, i.e., it
increased pain sensitivity of mice both in the acute and in the late phases of neuropathy (Figure 7B).
Intraperitoneal 31c, at 30 mg/kg, effectively reduced acute-phase and late-phase tactile allodynia in
oxaliplatin-treated mice (significant at p < 0.01 vs. pre-drug paw withdrawal force of oxaliplatin-treated
mice).
1
5

Figure 7. Effect of the test compounds 19b (Figure 7A) and 31c (Figure 7B) on tactile allodynia
determined by the von Frey test in oxaliplatin-induced NP in mice (n=7-10). Results are shown as mean
force (± SEM) applied to elicit paw withdrawal. For statistical analysis, one-way ANOVA followed by
#
###
Tukey’s post hoc test was performed.
p < 0.0001 (vs. paw withdrawal force of non-neuropathic
control) and **p < 0.01, ***p < 0.001, and ****p < 0.0001 (vs. pre-drug paw withdrawal force of
oxaliplatin-treated mice).
The comparison of results obtained in the present experiment with the previous ones conducted under the
3
8
same laboratory conditions demonstrates a higher efficacy of tiagabine and, to a lesser degree,
48
pregabalin in reducing tactile allodynia in oxaliplatin-treated mice.
2
.3.4
Oxaliplatin-induced peripheral neuropathy: effect on cold allodynia (cold plate test)
In the cold plate test, an overall effect of the treatment was revealed (F[8,69]=6.802, p < 0.0001) for 19b.
Post hoc analysis revealed that the administration of oxaliplatin significantly (p < 0.001) decreased the
latency to pain reaction (i.e., it increased the pain sensitivity of mice) both in the acute and late phases of
neuropathy. As in the diabetic NP model, intraperitoneal 19b, at 10 and 30 mg/kg doses, could not reduce
cold allodynia, either in the early or the late phase of neuropathy caused by oxaliplatin (Figure 8A). For
compound 31c, an overall effect of the treatment was revealed in this test (F[4,33]=2.55, p < 0.0001). The
administration of oxaliplatin significantly (p < 0.0001) reduced the latency to pain reaction in mice in both
phases of neuropathy. Intraperitoneal 31c, at the dose 30 mg/kg, did not attenuate cold allodynia induced
38
by oxaliplatin (Figure 8B). Also, the previously tested reference drugs: tiagabine and pregabalin, were
4
8
not able to increase the cold sensitivity threshold of oxaliplatin-treated mice.
1
6

Figure 8. Effect of the test compounds 19b (Figure 8A) and 31c (Figure 8B) on cold allodynia determined
by the cold plate test in oxaliplatin-induced NP model in mice (n=7-10). Results are shown as mean
latency to pain reaction (± SEM). For statistical analysis, one-way ANOVA followed by Tukey’s post hoc
##
###
####
test was performed. p < 0.01 p < 0.001
p < 0.0001 (vs. latency of non-neuropathic control).
The results obtained from the two mouse models of NP showed that both compounds 19b and 31c could
reduce tactile allodynia, but they could not attenuate thermal (heat/cold) allodynia in neuropathic mice.
The comparison of doses effectively reducing tactile allodynia revealed a higher efficacy of 19b than that
of 31c. This difference in in vivo efficacy might result from a distinct profile of mGAT affinity for these
two compounds, with compound 19b being a stronger inhibitor of mGAT3 and mGAT4 than 31c. Of note,
the test compounds were less active as the previously tested in the same experimental conditions reference
drugs: tiagabine and pregabalin.
1
7

In the present study, the mouse STZ model was implemented, with tests performed 21 days after STZ
administration. At this time point, diabetic mice exhibited a significant decrease in mechanical nociceptive
threshold, which results in the development of tactile allodynia, heat hyper-, or hypoalgesia, with the latter
6
5
more often observed at advanced stages of diabetes (i.e., ≥12 weeks after STZ treatment). In the protocol
used in the study, STZ-treated diabetic mice exhibited significantly increased blood glucose levels, urine
output, and decreased body weight gain as compared to their normoglycemic littermates. Under these
conditions, 19b and 31c exhibited a significant tactile allodynia reduction in diabetic neuropathic mice. In
fact, in the von Frey test, these compounds at the dose of 30 mg/kg almost returned the mechanical
nociceptive threshold of the mice to levels similar to those of control (normoglycemic) mice. Taken
together, our study showed that although test compounds can reduce some aspects of diabetic neuropathy
in mice (i.e. tactile allodynia), this does not extend to other aspects, notably thermal hyperalgesia. A lack
of influence on thermal nociceptive threshold could limit their potential use as a treatment for diabetic NP.
Test compounds were also active in another model of NP, i.e., the CIPN model. Here, a single dose of
oxaliplatin induced painful peripheral neuropathy accompanied by mechanical (tactile) and cold allodynia
66
(
but not heat hyperalgesia). The test compounds at the dose of 30 mg/kg attenuated tactile allodynia in
both the acute-phase and late-phase neuropathy. Their efficacy in the von Frey test was, however, lower
than the reference drugs. Similarly to the reference drugs, compounds 19b and 31c failed to modify
responses in the cold plate test.
To summarize the in vivo part of this study, mGAT3/4 seem to be targets for analgesic drugs used for the
relief of NP symptoms. However, the comparison of the efficacy of the test compounds and reference
drugs indicates that mGAT1 and calcium channels play a key role as targets for analgesics used to
attenuate tactile allodynia in the course of NP.
2
.4
Molecular docking
To confirm the role of mGAT1 in NP and to establish whether mGAT1 inhibition might be the potential
mechanism of the analgesic action we selected compounds that were used for in silico studies. For this,
three highly and two weakly active compounds toward mGAT1 were used as reference sets. The former
consisted of 5, 6, and (R,S)-7, while the later of 3, 4 (Figure 1). Selected reference compounds have
analgesic activity adequately related to affinity to mGAT1 (highly active toward mGAT1) and mGAT3/4
(
weakly active toward mGAT1). The test set contained the two compounds (R/S-19b and R/S-31c)
selected also for the in vivo study. It is worth mentioning that all compounds investigated in the study,
both reference and test sets, showed the differences between the calculated pK and pIC50 value within
empirically reasonable range (pK ≈ pIC50, pK - pIC50 < 1, see Table 1 and 2, Figure 1, 3 and 4).
i
i
i
1
8

Moreover, in silico results (Table 2) are in agreement with the data obtained from MS binding studies
showing that, in most cases, the R-isomer of given active compound has higher affinity than respective S-
isomer toward mGAT1. The objective of this part of the project was to perform the in silico study on the
series of twelve inhibitors (Table 2). The measurement procedures applied in the study are typical for
38–40
docking studies.
The results of the docking experiments, such as pK was calculated from the
i
AutoDock4 estimated inhibition constant K , which was reported in the AutoDock4 output, the complex
i
binding energy and the specific hydrogen bond components (energies, lengths, and angles), are shown in
Table 2. The binding modes between hGAT1’s and studied compounds are illustrated in Figure 2S-13S at
Supplementary materials.
Table 2. The summary of results of docking experiments.
E
B
H
B
Angle
L
HB
E
HB
Amino acid
residues
Complex
pK
i
pIC50
(
kcal/mol
)
Donor
Acc
(θ)
(Å
)
(kcal/mol)
LEU460
#CONH
%COO
156.74
1.78
-5.54
(
R
)-SNAP-5114
(
3
)
4.24
4.22
-
-5.78
(
S)-SNAP-5114
(
3
)
4.07
-5.4
TRP68
#ind_NH
#CONH
%COO
156.76
161.66
1.80
2.21
-5.61
-3.73
SER459
%COO1
(
R
)-DDPM-859
(
4
)
4.17
-5.68
LEU460
MET458
TRP68
#CONH
#CONH
#ind_NH
#ind_NH
#CONH
#CONH
#CONH
#CONH
%OH
%COO2
%COO
%O
168.64
149.90
140.73
146.74
165.68
148.14
160.01
138.50
125.02
127.63
159.27
144.69
1.93
1.87
2.17
1.69
2.08
1.93
1.75
2.18
1.81
2.08
1.74
2.06
-6.64
-4.39
-1.84
-1.69
-5.09
-3.99
-5.92
-1.75
0.40
4
.19*
(
S)-DDPM-859
(4
)
4.03
8.61
6.63
-5.49
-11.75
-9.05
DDPM-2571
(
5)
8.27
6.83
3
9
TRP68
%O
NO-711
(6)
MET458
SER459
MET458
SER459
ASP451
SER295
TYR139
TRP68
%N_oxasole
%N_oxasole
%N oxasole
%N_oxasole
#CONH
(
R
)-EF1502
)-
39
5.27
4.87
5.15
-
-7.19
-6.65
((R 7)
(
S)-EF1502
((S)-7)
39
-
6.26
6.02
6.45
5.55
(
R)-19b
4.59
4.42
4.73
4.06
4
.86*
.69*
-
%OH
#CONH
-0.69
-0.40
-2.81
(
(
S
)-19b
-
#PhOH
#ind_NH
%NHCO
R
)-31c
4
-
%OH
(
S)-31c
B
H —hydrogen bond. acc—hydrogen bond acceptor. hydrogen bond components: from the ligand % and from the protein #. and E
B
—complex
energy binding. p
K
i
was calculated from the AutoDock4 estimated inhibition constant . which is reported in the AutoDock4 output. pIC50 toward
K
i
22
22
19
18
21
GAT1: ( )-SNAP-5114 (3) DDPM-859 (4) DDPM-2571 (5) NO-711 (6) EF1502 (7) , * data for racemic mixture
S
The preliminary analysis of the docking study indicated that all studied compounds might interact with the
active site of hGAT1. On the basis of predicted binding affinity, the highest affinity was observed with the
R-isomer in all cases of studied compounds. The strongest binding affinity was formed by 5, (calculated
pK ≥ 8 ) and 6 (calculated pK ≥ 6 ) together with the strongest chemical interaction (see binding energy
i
i
Table 2). Visual of the docking poses of 5 and 6 (Figure 2S-3S) revealed almost identical interaction with
hGAT1. There are two main likenesses between the complexes of 5 and 6 and hGAT1. The first one
concerns the formation of H-bonds design bonded interaction to exactly the same residues, i.e., indole NH
1
9

moieties of Trp68 donates one normal H-bond to oxygen atoms of (amino)oxy)ethyl guvacine moieties of
and 6. Second concerns the distances associated with the H-bonds lie in the range 2.17 ÷ 1.69 Å. The H-
5
bonds energy ranges 1.84 ÷ -1.69 kcal/mol. Interestingly in cases of (R)-7 and (S)-7 compounds it was
found that they interact with hGAT1 with binding energies of -7.19 for (R)-7 and -6.65 for (S)-7 kcal/mol,
respectively. Both of these molecule form their complexes with hGAT1 via bifurcated H-bonds (Table 2S,
Figure 4S and 5S in Supplementary materials). The nitrogen atom of oxazole ring of (R and S)-7 isomers
acts as donor in H-bond with Met458 and Ser459. As it is depicted in Figure 4S and 5S this bifurcated
system is highly asymmetric, as the major H-bond component involves the Met45 and the minor the
Ser459. The major H-bond component have moderate energy > -5 kcal/mol, and average length about 2 Å,
and non-linear geometry (165 ÷ 138°). Whereas the minor H-bond component is weak get long (2.20 Å)
and are non-linear interaction (Table 2).
In the case of reference set containing compounds weakly active toward mGAT1 (R/S-3, R/S-4), they are
highly active toward mGAT3/4. Our calculations showed low inhibitory potency of reference set
compounds toward GAT1, corresponding to the respective biological data (pK ≈ pIC , pK - pIC < 1,
i
50
i
50
see Table 1 and 2, Figure 1, 3 and 4). The interactions with the active site of the protein showed binding
energies of R/S-3 comparable to these of R/S-4 (about −5.5 kcal/mol) which suggested the stability of the
complex with hGAT1 about 2-fold lower than of that with 5 and 6. On the contrary, the strongest, among
all calculated H-bonds, interaction was found (EHB ∈ (6.64, 4.39) kcal/mol) in the case of 3 and 4 R/S-
enantiomer. The molecular docking study revealed that R/S-3 and S-4 isomers interact via single normal
hydrogen bonds. R-4-hGAT1 complex is formed via one bifurcated H-bonds.
In the case of test set including two compounds R/S-19b and R/S-31c docking studies showed that these
i
compounds interact with hGAT1 with low binding affinities (pK ∈ (4.06, 4.73), Table 2). Comparing the
data obtained for R and S enantiomers led to a conclusion that binding affinity between (R)-31c and
hGAT1 is higher than this between (S)-31c and hGAT1 (difference about 0.7, Table 2). The same situation
was found in case of (R)-19b and (S)-19b where the binding affinity differences were about 0.17 (Table
2
). Our docking results implied that the stability of the complexes of hGAT1 with R/S-19b and (R)-31c are
comparable while that with (S)-31c is lower (E difference about 1 kcal/mol, Table 2). All these
B
complexes form one key interaction with residues of hGAT1. The hydroxyl group of R/S-19b donates one
single normal H-bond to Asp450 ((R)-19b, Figure 10S at Supplementary materials) and Ser295 ((S)-19b,
Figure 11S at Supplementary materials). While the CONH moieties of (R)-31c and OH moieties of (S)-
3
1c act as an acceptors of single H-bond involving Tyr139 ((R)-31c, Figure 12S at Supplementary
materials) and Trp68 ((S)-31c, Figure 13S at Supplementary materials). All calculated molecular
configurations indicated that the single hydrogen bond energy between R/S-19b, R/S-31c and hGAT1 was
2
0

relatively small (Table 2). The distances of hydrogen bonds range from 1.74 to 2.08 Å and has non-linear
geometry (θ ∈ (125°, 160°), Table 2).
To sum up, based on all the information obtained from the docking tests, we can state that the test set
compounds are characterized by low binding affinities toward hGAT1 and low stability of the resulting
complexes, which support the hypothesis that observed analgesic effect in in vivo tests is more likely
related to higher binding affinity between the compounds and mGAT 3/4 than mGAT1.
The validation was carried out by docking of few molecules with no affinity to hGAT1 (as a control
group). The control group of neurotransmitters active in CNS was chosen: leucine, isoleucine and valine.
This choice was made based on the structural similarity to GABA molecule and as basic substrates for the
67,68
LeuT model but from the set of compounds inactive in GAT transport.
The control group included also
69
typical analgesic inactive in GAT transport like aspirin (acetylsalicylic acid, ASA). The results of the
validation experiments, such as: the complex binding energies, the specific hydrogen bond components,
and detailed data of the hydrogen bonds features (energies, lengths and angles) are gathered in Table 3.
Table 3. The summary of validation experiments results
E
B
amino
residues
acid
H
B
Angle
L
HB
E
HB
kcal/mol
-5,276
complex
pK
i
kcal/mol
donor
#CONH
acc
%COO1
θ
Å
MET458
153,536
1,79
SER459
LEU460
ALA61
TYR140
TRP68
#CONH
#CONH
%NH3+
#PhOH
%COO1
%COO2
#CONH
%COO
%COO
#COO1
#COO2
%COO
#COO1
#COO2
163,074
178,059
120,86
1,78
-6,639
-5,469
-0,121
0,653
ASA
3.31
3.04
2.97
-4.52
2,09
2,101
2,158
1,838
1,894
1,854
1,829
1,945
1,821
Leu
Ile
-4.15
-4.05
173,846
166,41
#ind_NH
%NH1+
%NH2+
#ind_NH
%NH1+
%NH2+
-7,066
-0,118
-0,093
-5,945
-0,137
-0,201
ASP451
ASP451
TRP68
134,064
128,069
158,302
131,827
132,557
ASP451
ASP451
Val
2.86
-3.9
Abbreviations in table H
protein #, E – complex energy binding,
calculated from the AutoDock4 estimated inhibition constant
B
– hydrogen bond, acc – hydrogen bond acceptor, Hydrogen bond components: from the ABX % and from the
- hydrogen bond angle, LHB -hydrogen bond length, EHB - hydrogen bond energy, p was
, which is reported in the AutoDock4 output.
B
θ
K
i
K
i
The binding modes between hGAT1 and control compounds are illustrated in Figure 14S-17S in
supplementary file. As it can be seen from the data all control complexes have calculated pK ≤ 4 which is
i
commonly used as threshold and can be considered to be an inactive ligand for that protein. In addition all
2
1

control complexes have higher binding energy values and low hydrogen bond energy than GABA-hGAT1
38
complex. These results confirm the validity of the approach taken in the docking experiments.
2
.5 Physicochemical properties: Lipophilicity
Physicochemical properties of molecules have a significant role in their pharmacokinetic and
pharmacodynamic activities; lipophilicity, in particular, is a physicochemical property of principal
importance in drug discovery and development. Considering the abovementioned properties, the
lipophilicity of final compounds was estimated by computational methods (ChemBioDraw and
SwissADME). Then, the obtained values were verified by the experimental method using Ultra
70
Performance Liquid Chromatography coupled with Mass Spectrometry (UPLC/MS). As shown in Table
, the calibration and reliability of the UPLC/MS method were confirmed using a set of well-known
4
compounds with established logP values. In addition, the analysis of the results obtained indicates that the
lipophilicity parameters determined using both the computational methods (ChemBioDraw and
SwissADME) and the experimental (UPLC/MS) are comparable (Table 4). Figure 9 shows that the
2
calculated correlation coefficients (R ) between log P obtained by UPLC/MS and in silico methods are
0
.8762 and 0.8575 (ChemBioDraw and SwissADME, respectively). The obtained logP showed that the
value of lipophilicity is dependent on the type of substituents in the N-benzyl fragment of the molecules
and the length of the carbon atom linker in the core of the structures, as expected (Table 4). Most
lipophilic properties are due to the presence of 4-hydroxybutanoic and pentanoic amides with a chlorine
atom in the N-benzyl fragment of molecules (19b–c, 28, and 23b–c, respectively). The unsubstituted
derivative of 4-aminobutyric acid (27a) or the one possessing a fluorine atom at the para position of N-
benzylamide moiety (27d) showed the lower value of logP as compared to their appropriate serine analogs
(
31a and 31d). Unfortunately, the obtained compounds possessed still a surprisingly high lipophilicity
index.
Table 4. Log P values obtained for selected compounds.
logP
Compound
ChemBioDraw
5.41
SwissADME
5.09
UPLC/MS
5.71
1
9a
1
9b
5.97
5.58
6.23
1
1
1
1
9c
9d
9e
9f
5.97
5.57
5.90
5.73
5.34
5.42
5.46
5.66
6.38
5.87
6.20
6.37
2
2
2
0
1
2
5.85
5.65
3.52
5.67
5.25
3.84
6.21
6.10
4.75
2
2
7a
7b
5.03
5.59
5.04
5.52
5.64
6.22
2
2
2
7c
7d
7e
5.59
5.19
5.52
5.57
5.35
5.35
5.95
5.67
5.99
2
2

2
8
6.53
5.31
5.87
5.47
5.73
5.73
6.29
6.29
6.29
5.89
6.22
6.22
6.08
4.80
5.34
5.09
5.39
5.39
5.93
5.93
5.93
5.72
5.68
5.68
6.77
5.51
6.15
5.50
6.11
5.85
6.45
6.61
6.46
6.26
6.47
6.45
3
3
1a
1c
3
1d
(
(
2
2
RS,4RS)-23a
RS,4SR)-23a
RS,4RS)-23b
RS,4RS)-23c
RS,4SR)-23c
2RS,4RS)-23d
RS,4RS)-23e
RS,4SR)-23e
(
2
(
(
(
(
(
2
2
2
2
Selected reference compounds
Loratadine
ChemBioDraw
DrugBank
5.20
UPLC/MS
6.06
4.13
2.14
1.39
3.68
4.63
Phenytoin
Benzocaine
Promazine
Amitriptyline
2.47
1.86
4.55
4.92
2.93
1.95
3.83
4.27
Considering the above results, we found that the UPLC/MS method is an appropriate and useful tool to
determine the lipophilicity properties. However, the obtained results showed that it is important to note
that the lipophilicity of studied compounds should be reduced, in regard to the Lipinski’ rule of five, and
this will be the next stage of our research in future.
7
6
5
4
3
7
6
5
4
3
R² = 0.8762
R² = 0.8575
3
4
5
6
7
3
4
5
6
7
log P (ChemBioDraw)
log P (SwissADME)
Figure 9. Correlation between the experimental method (UPLC/MS) and in silico studies (ChemBioDraw
and Swiss ADME).
3
Conclusion
The performed studies included synthesis, in vitro and in vivo evaluation of pharmacological properties,
molecular docking, and definition of the relationships between the structures and physicochemical
properties. According to the data obtained, the preferred inhibitory activity toward mGAT3 and mGAT4
was observed. N-benzylamides with a chlorine substituent in the ortho position in both GHB and GABA
2
3

series, and similarly in the series of 4-hydroxypentanoic acid derivatives showed the highest inhibitory
activity toward mGAT3 (19b, pIC = 5.31 ± 0.04; 27b, pIC = 5.44 ± 0.08, (2RS,4RS)-23b (pIC = 5.44
5
0
50
50
±
0.10). Thus, the structure–activity relationship studies clearly revealed the importance of the chlorine
substituent in the ortho position of the N-benzylamide moiety. Subsequent substitution of primary and
aliphatic amino groups at the 4-position of GABA analogs with acetic acid leads to obtaining the non-
selective GAT inhibitor (28) but with the highest inhibitory potency toward mGAT3 (IC = 2.5 ꢀM; pIC
50
50
=
5.61 ± 0.07). Finally, compound 28 was found to be more effective mGAT3 inhibitor than the reference
compounds (S)-SNAP-5114 (3, pIC50 = 5.29 ± 0.04, mGAT3) and DDPM-1457 (4, pIC50 = 4.85 ± 0.04,
mGAT3). Furthermore, in the in vivo part of the performed study, by using two mouse models of chronic
pain, namely NP induced by STZ and oxaliplatin, it was shown that both compounds 19b and 31c could
reduce tactile allodynia in neuropathic mice without inducing any acute adverse effects or toxicity during
the experiment but they were not able to reduce increased pain sensitivity to thermal stimulation. The
proven effect of analgesia (in in vivo test) indicated the correlation of the observed analgesic effect (in
vivo tests) and inhibitory activity toward mGAT3/4 of compounds in a test set (in vitro test). However,
further research on the structure optimization with a focus on enhancing drug-like properties is needed.
4
Experimental section
4
.1 Chemistry
Commercially available reagents were purchased from Merck, Aldrich, Acros, or ChemPur and were used without
further purification. Solvents for reactions carried out under inert gas (argon), such as THF and methylene chloride
before use, were dried, distilled, and collected under argon atmosphere. THF was distilled from the mixture of
sodium and benzophenone, while methylene chloride was distilled from calcium hydride. Thionyl chloride and TEA
were distilled under vacuum before use. Reactions under microwave irradiation were carried out using the Discover
LabMate (CEM Corporation, USA). Purification of chemical compounds by column chromatography was carried out
using silica gel (mesh: 0.063–0.200 mm, Sigma-Aldrich) as a stationary phase. The reactions were monitored by
thin-layer chromatography (aluminum sheets precoated with silica gel 60 F254 (Merck). Compounds were visualized
with UV light (254 nm). In addition, the chromatogram was stained in a 0.5% solution of ninhydrin in n-propanol or
in a solution of 5% (NH ) Mo O and 0.2% Ce(SO ) in 5% H SO . The retardation factor R was defined using the
4
6
7
24
4 2
2
4
f
following solvent systems: S1 (chloroform/acetone, 1:1 v/v), S2 hexane/ethanol/TEA (7:2:1, v/v/v), S3
(
DCM/acetone, 7:3 v/v), S4 (DCM/acetone, 3:2 v/v), S8 (DCM/methanol/AcOH, 8:2:0.5, v/v/v), and S9
1
13
1
(
DCM/acetone 9:1, v/v). H NMR and C NMR spectra were recorded using Varian Mercury-VX 300, with H at
1
3
3
00.08 MHz and C at 75.46 MHz. The chemical shifts (δ) are reported in ppm and were calculated in relation to the
frequency of the deuteron field stabilization signal. The coupling constant J is reported in Hz. High-resolution mass
spectrometry was performed using Synapt G2-S HDMS mass spectrometer (Waters Inc.) equipped with an
electrospray ion source and q-Time-of-Flight type mass analyzer. The instrument was controlled and recorded data
2
4

were processed using MassLynx V4.1 software package (Waters Inc). The purities and lipophilicity of the final
compounds were recorded using Acquity™ UPLC (Waters, Milford, MA, USA) coupled to a tandem quadrupole
(
TQD, Waters) mass spectrometer (electrospray ionization mode ESI-tandem quadrupole). Chromatographic
separations were carried out using the Acquity UPLC BEH (bridged ethyl hybrid) C18 column; 2.1 × 100 mm, and
.7 µm particle size, equipped with Acquity UPLC BEH C18 VanGuard pre-column, 2.1 × 5 mm, and 1.7 µm
1
particle size. All of the final compounds showed purities of >95%. Elemental analyses (C, H, N, and S) were carried
out within 0.4% of the theoretical values and were performed on Vario Elementar EL III (Elementar
Analysensysteme, Hanau, Germany). Melting points (mp) were determined in open glass capillaries on a Melting
Point Büchi 353 apparatus and are uncorrected.
4
.1.1
3-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}oxolan-2-one (17)
Anhydrous K CO (1.80 mmol, 0.25 g, 1 equiv) was added to the solution of N-methyl-4,4-bis(3-methylthiophen-2-
2
3
yl)but-3-en-1-amine (1.80 mmol, 0.5 g, 1 equiv) and tetrabutylammonium bromide (TBAB, 0.18 mmol, 0.06 g, 0.01
equiv) in acetonitrile (7 ml), and the mixture was stirred at 0 °C for 15 min. Then, a solution of 3-
bromodihydrofuran-2(3H)-one (15) (1.80 mmol, 0.30 g, 1 equiv) was added dropwise, and stirring was continued for
1
6 h at room temperature. After the reaction was completed, the precipitate was filtered off and the filtrate was
concentrated under vacuum. The obtained crude product was purified by column chromatography over silica gel
1
(
CH Cl /acetone = 9:1) to yield 17 (468 mg, 72 %, R = 0.69 (S )) as a yellow oil: H NMR (CDCl ) δ ppm 2.01 (s, 3
2
2
f
9
3
H (Me)), 2.04 (s, 3 H (Me)), 2.18 - 2.30 (m, 2 H (NCHCH )), 2.31 - 2.37 (m, 5 H (Me; =CHCH )), 2.67 - 2.79 (m, 2
2
2
H (CH N)), 3.63 (t, J = 9.62 Hz, 1 H (NCH)), 4.10 - 4.21 (m, 1 H (OCH )), 4.29 - 4.39 (m, 1 H (OCH )), 6.06 (t, J =
2
2
2
7
.31 Hz, 1 H (C=CH)), 6.76 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d, J = 5.13 Hz, 1 H (SCHCH)), 7.05 (d, J = 5.13
+
Hz, 1 H (SCH)), 7.21 (d, J = 5.13 Hz, 1 H (SCH)). Formula: C19
.1.2
Anhydrous K CO (2.24 mmol, 0.31 g, 1 equiv) was added to the solution of N-methyl-4,4-bis(3-methylthiophen-2-
H23NO
2
S
2
, MS: m/z 262 (M+H ).
4
3-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-5-methyloxolan-2-one (18)
2
3
yl)but-3-en-1-amine (14) (2.30 mmol, 0.64 g, 1 equiv) and TBAB (0.68 mmol, 0.22 g, 0.3 equiv) in the acetonitrile
(
(
8 ml), and the mixture was stirred at 0 °C for 15 min. Then, a solution of 3-bromo-5-methyldihydrofuran-2(3H)-one
16) (2.29 mmol, 0.41 g, 1 equiv) was added dropwise and stirring was continued for 20 h at room temperature. After
the reaction was completed, the precipitate was filtered and the filtrate was concentrated under vacuum. The obtained
crude product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield compound 18
2
2
1
(
(
(
(
(
740 mg, 85 %, R = 0.78 and 0.72 (S )) as yellow oil: H NMR (300 MHz, CHLOROFORM-d) σ ppm 1.32 - 1.44
f 9
m, 3 H (Me)) 1.78 (td, J=12.25, 10.39 Hz, 2 H (NCHCH
2
)) 1.98 - 2.02 (m, 3 H (Me)) 2.04 (s, 3 H (Me)) 2.27 - 2.40
m, 5 H (Me; =CHCH )) 2.66 - 2.78 (m, 2 H (CH N)) 3.71 (dd, J=12.18, 8.34 Hz, 1 H (NCH)) 4.34 - 4.47 (m, 1 H
2
2
OCH )) 6.00 - 6.11 (m, 1 H (C=CH)) 6.75 (d, J=5.13 Hz, 1 H (SCHCH)) 6.84 (d, J=5.13 Hz, 1 H (SCHCH)) 7.04
2
+
d, J=5.13 Hz, 1 H (SCH)) 7.20 (d, J=5.13 Hz, 1 H (SCH)). Formula: C H NO S , MS: m/z 376 (M+H ).
20
25
2 2
4
.1.3
General procedure for the synthesis of N-benzyl-2-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-
yl)(methyl)amino)-4-hydroxybutanamide derivatives 19a–f and 20–22.
2
5

Under argon atmosphere, 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)dihydrofuran-2(3H)-one
17) (1 equiv) was heated with relevant amine (2 equiv) in dry THF at reflux for 48 h. When the reaction was
completed, the mixture was ice-cooled, and 1 M solution of HCl (1.5 mL) was added to the mixture. The mixture
(
was then extracted with dichloromethane (3 × 10 mL). The combined organic fractions were dried over Na
evaporated under vacuum. The crude product was purified by column chromatography.
2 4
SO and
4
.1.3.1 N-benzyl-2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-4-hydroxybutanamide (19a)
Compound 19a was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
(3H)-one (17) (2.21 mmol, 0.80 g) and N-benzylamine (4.42 mmol, 0.47 g) in 10 ml dry THF. The obtained crude
2
product was purified by column chromatography over silica gel (CH
2
Cl
2
/acetone = 9:1) to yield 19a (384 mg, 39%,
1
R
f
= 0.60 (S
1
)) as yellow oil: H NMR (CDCl
3
) δ ppm 1.78 - 1.92 (m, 5 H (NCHCH
2
; Me)), 2.13 (s, 3 H (Me)), 2.00
(
s, 3 H (Me)), 2.21 - 2.34 (m, 2 H (=CHCH )), 2.48 - 2.62 (m, 2 H (CH N)), 3.24 (dd, J = 3.59, 8.98 Hz, 1 H
2 2
(
NCH )), 3.51 - 3.61 (m, 1 H (CH OH)), 3.79 - 3.88 (m, 1 H (CH OH)), 4.35 - 4.53 (m, 2 H (CH NH)), 5.99 (t, J =
2 2 2 2
7
.31 Hz, 1 H (C=CH)), 6.74 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d, J = 5.13 Hz, 1 H (SCHCH)), 7.02 (d, J = 5.13
1
3
Hz, 1 H (SCH)), 7.17 -7.33 (m, 6 H (Ar; SCH)), 7.82 (t, J = 5.90 Hz, 1 H (CONH)). C NMR (CDCl ) δ ppm 14.51
3
(
(
thiophene-Me), 14.71 (thiophene-Me), 27.10 (=CH-CH
2 2
-), 28.42 (-N-CH-CH -), 37.82 (N-Me), 43.31
CO(NH)CH -), 54.00 (-CH -N-), 61.50 (-CH -OH), 66.81 (-N-CH-), 122.71 (-C=CH-), 127.41 (Ar), 127.71 (2 C,
2
2
2
Ar), 128.70 (2 C, Ar), 129.71 (2 C, thiophene), 131.51 (2 C, thiophene), 132.50 (-S-CH-), 133.82 (-S-CH-), 134.81
(
(
Ar), 135.61 (-C-Methiophene), 138.31 (-C-Methiophene), 139.00 (-C=CH-), 174.10 (C=O). Anal. calcd for C H N O S
26 32 2 2
2
+
+
468.67): C: 66.63, H: 6.88, N: 5.98, S: 13.68; Found C: 66.71, H: 6.91, N: 5.98, S: 13.71. HRMS-ESI m/z [M+H]
calcd for C H N O S : 469.1983, found: 469.1996.
2
6
32
2
2 2
4
.1.3.2 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-N-[(2-chlorophenyl)methyl]-4-
hydroxybutanamide (19b)
Compound 19b was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
(3H)-one (17) (2 mmol, 0.72 g) and (2-chlorophenyl)methanamine (4 mmol, 0.57 g) in 8 ml dry THF. The obtained
2
2 2
crude product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield 19b (321 mg,
1
3
2%, R = 0.62 (S )) as yellow oil: H NMR (CDCl ) δ ppm 1.84 (td, J = 4.20, 8.27 Hz, 2 H (NCHCH )), 2.01 (s, 3
f
1
3
2
H (Me)), 1.90 (s, 3 H (Me)), 2.14 (s, 3 H (Me)), 2.22 - 2.36 (m, 2 H (=CHCH )), 2.56 (dt, J = 2.56, 7.05 Hz, 2 H
2
(
CH N)), 3.24 (dd, J = 3.98, 8.34 Hz, 1 H (NCH )), 3.50 - 3.60 (m, 1 H (CH OH)), 3.77 - 3.86 (m, 1 H (CH OH)),
2 2 2 2
4
.42 - 4.62 (m, 3 H (CH NH; OH)), 6.00 (t, J = 7.31 Hz, 1 H (C=CH)), 6.73 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d,
2
J = 5.13 Hz, 1 H (SCHCH)), 7.02 (d, J = 5.13 Hz, 1 H (SCH)), 7.16 - 7.25 (m, 3 H (Ar; SCH)), 7.29 - 7.36 (m, 2 H
1
3
(
Ar)), 7.91 (t, J = 6.03 Hz, 1 H (CONH)). C NMR (CDCl
3
) δ ppm 14.41 (thiophene-Me), 14.70 (thiophene-Me),
2
6
7.01 (=CH-CH -), 28.41 (-N-CH-CH -), 37.90 (N-Me), 41.42 (CO(NH)CH -), 54.02 (-CH -N-), 61.51 (-CH -OH),
2
2
2
2
2
6.70 (-N-CH-), 122.71 (-C=CH-), 124.54 (Ar), 127.00 (Ar), 128.91 (Ar), 129.10 (Ar), 129.52 (Ar), 130.02 (2 C,
thiophene), 131.42 (2 C, thiophene), 132.32 (-S-CH-), 133.52 (-S-CH-), 133.72 (Ar), 134.82 (-C-Methiophene), 135.54
-C-Methiophene), 139.00 (-C=CH-), 174.21 (C=O). Anal. calcd for C H ClN O S (503.12): C: 62.07, H: 6.21, N:
(
26
31
2
2 2
2
6

+
+
5
5
.57, S: 12.74; Found C: 62.17, H: 6.22, N: 5.61, S: 12.78. HRMS-ESI m/z [M+H] calcd for C H ClN O S :
26 31 2 2 2
03.1594, found: 503.1584.
4
.1.3.3 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-N-[(4-chlorophenyl)methyl]-4-
hydroxybutanamide (19c)
Compound 19c was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
(3H)-one (17) (2 mmol, 0.72 g) and (4-chlorophenyl)methanamine (4 mmol, 0.57 g) in 8 ml dry THF. The obtained
crude product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield 19c (450 mg,
2
2
2
1
4
5%, R = 0.56 (S )) as a yellow oil: H NMR (CDCl ) δ ppm 1.77 - 1.92 (m, 5 H (NCHCH ;Me)), 2.12 (s, 3 H
f 1 3 2
(
Me)), 2.01 (s, 3 H (Me)), 2.22 - 2.34 (m, 2 H (=CHCH )), 2.50 - 2.63 (m, 2 H (CH N)), 3.23 (dd, J = 3.33, 8.98 Hz,
2 2
1
=
H (NCH)), 3.50 - 3.61 (m, 1 H (CH OH)), 3.79 - 3.88 (m, 1 H (CH OH)), 4.34 - 4.46 (m, 2 H (CH NH)), 5.99 (t, J
2 2 2
7.44 Hz, 1 H (C=CH)), 6.75 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d, J = 5.13 Hz, 1 H (SCHCH)), 7.03 (d, J =
1
3
5
.13 Hz, 1 H (SCH)), 7.10 - 7.16 (m, 1 H (SCH)), 7.19 - 7.27 (m, 4 H (Ar)), 7.80 - 7.90 (m, 1 H (CONH)). C NMR
(
CDCl ) δ ppm 14.41 (thiophene-Me), 14.71 (thiophene-Me), 27.10 (=CH-CH -), 28.32 (-N-CH-CH -), 37.81 (N-
3
2
2
Me), 42.62 (CO(NH)CH -), 54.00 (-CH -N-), 61.43 (-CH -OH), 66.62 (-N-CH-), 122.61 (-C=CH-), 124.51 (Ar),
2
2
2
1
28.71 (Ar), 129.10 (Ar), 129.20 (Ar), 129.70 (Ar), 131.51 (2 C, thiophene), 132.51 (2 C, thiophene), 133.00 (-S-CH-
, 133.81 (-S-CH-), 134.81 (Ar), 135.62 (-C-Methiophene), 136.90 (-C-Methiophene), 138.80 (-C=CH-), 174.22 (C=O).
Anal. calcd for C26 (503.12): C: 62.07, H: 6.21, N: 5.57, S: 12.74; Found C: 62.15, H: 6.20, N: 5.60, S:
)
2 2 2
H31ClN O S
+
+
1
2.77. HRMS-ESI m/z [M+H] calcd for C26
2 2 2
H31ClN O S : 503.1594, found: 503.1584.
4
.1.3.4 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-N-[(4-fluorophenyl)methyl]-4-
hydroxybutanamide (19d)
Compound 19d was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
2
(3H)-one (17) (1.5 mmol, 0.54 g) and (4-fluorophenyl)methanamine (3 mmol, 0.38 g) in 7 ml dry THF. The
obtained crude product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield 19d
2
2
1
(
327 mg, 45%, R = 0.55 (S )) as a yellow oil: H NMR (CDCl ) δ ppm 1.79 - 1.92 (m, 5 H (NCHCH ;Me)), 2.12 (s,
f 1 3 2
3
H (Me)), 2.00 (s, 3 H (Me)), 2.21 - 2.34 (m, 2 H (=CHCH )), 2.51 - 2.61 (m, 2 H (CH N)), 3.23 (dd, J = 3.33, 8.72
2
2
Hz, 1 H (NCH
2
)), 3.51 - 3.61 (m, 1 H (CH
2
OH)), 3.78 - 3.89 (m, 1 H (CH
2
2
OH)), 4.33 - 4.44 (m, 2 H (CH NH)), 5.99
(
(
(
(
(
t, J = 7.31 Hz, 1 H (C=CH)), 6.74 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.91 - 7.05
1
3
m, 3 H (Ar; SCH)), 7.12 - 7.23 (m, 3 H (Ar)), 7.78 - 7.89 (m, 1 H (CONH)). C NMR (CDCl ) δ ppm 14.39
3
thiophene-Me), 14.67 (thiophene-Me), 26.95 (=CH-CH -), 28.23 (-N-CH-CH -), 37.74 (N-Me), 42.58
2
2
CO(NH)CH -), 53.95 (-CH -N-), 61.45 (-CH -OH), 66.74 (-N-CH-), 115.29 (-CH-C-F), 115.57 (-CH-C-F), 122.63
2
2
2
-C=CH-), 124.54 (Ar), 129.10 (Ar), 129.39 (thiophene), 129.49 (thiophene), 129.65 (thiophene), 131.51 (thiophene),
1
1
5
4
32.52 (-S-CH-), 133.83 (-S-CH-), 134.10 (Ar), 134.70 (-C-Methiophene), 135.59 (-C-Methiophene), 138.85 (-C=CH-),
1
62.03 (d, JC-F=245.25 Hz, C-F), 174.18 (C=O). Anal. calcd for C26
H31FN
2
O
2
S
2
(486.66): C: 64.17, H: 6.42, N:
+
+
.76, S: 13.18; Found C: 64.19, H: 6.52, N: 5.79, S: 13.28. HRMS-ESI m/z [M+H] calcd for C H FN O S :
26
31
2
2 2
87.1889, found: 487.1888.
2
7

4
.1.3.5 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-4-hydroxy-N-[(4-
methylphenyl)methyl]butanamide (19e)
Compound 19e was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
2
(3H)-one (17) (2 mmol, 0.72 g) and p-tolylmethanamine (4 mmol, 0.48 g) in 8 ml dry THF. The obtained crude
2 2
product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield 19e (308 mg, 32%,
1
R = 0.57 (S )) as yellow oil: H NMR (CDCl ) δ ppm 1.77 - 1.94 (m, 5 H (NCHCH ;Me)), 2.00 (s, 3 H (Me)), 2.13
f
1
3
2
(
s, 3 H (Me)), 2.18 - 2.29 (m, 2 H (=CHCH )), 2.32 (s, 3 H), 2.49 - 2.62 (m, 2 H (CH N)), 3.23 (dd, J = 3.59, 8.98
2 2
Hz, 1 H (NCH )), 3.51 - 3.60 (m, 1 H (CH OH)), 3.79 - 3.89 (m, 1 H (CH OH)), 4.31 - 4.48 (m, 2 H (CH NH)), 5.98
2
2
2
2
(
t, J = 7.31 Hz, 1 H (C=CH)), 6.74 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.84 (d, J = 5.13 Hz, 1 H (SCHCH)), 7.03 (d, J =
5
.13 Hz, 1 H (SCH)), 7.05 - 7.17 (m, 4 H (Ar)), 7.20 (d, J = 5.13 Hz, 1 H (SCH)), 7.77 (t, J = 6.03 Hz, 1 H (CONH)).
1
3
C NMR (CDCl
3 2
) δ ppm 14.42 (thiophene-Me), 14.72 (thiophene-Me), 21.21 (Ar-Me), 27.10 (=CH-CH -), 28.42 (-
N-CH-CH -), 37.82 (N-Me), 43.13 (CO(NH)CH -), 53.92 (-CH -N-), 61.52 (-CH -OH), 66.81 (-N-CH-), 122.70 (-
2
2
2
2
C=CH-), 124.61 (Ar), 127.70 (2 C, Ar), 129.02 (Ar), 129.32 (2 C, thiophene), 129.72 (thiophene), 131.41
(
thiophene), 132.51 (-S-CH-), 133.71 (-S-CH-), 134.81 (Ar), 135.32 (-C-Methiophene), 135.52 (-C-Methiophene), 137.05
(
C -Me), 139.01 (-C=CH-), 174.00 (C=O). Anal. calcd for C H N O S (482.70): C: 67.18, H: 7.10, N: 5.80, S:
Ar
27 34
+
2
2 2
+
1
34 2 2 2
3.29; Found C: 67.28, H: 7.13, N: 5.83, S: 13.31. HRMS-ESI m/z [M+H] calcd for C27H N O S : 483.2140,
found: 483.2135.
4
.1.3.6 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-N-[(2,4-difluorophenyl)methyl]-4-
hydroxybutanamide (19f)
Compound 19f was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
2
(3H)-one (17) (1 mmol, 0.36 g) and (2,4-difluorophenyl)methanamine (2 mmol, 0.29 g) in 5 ml dry THF. The
obtained crude product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1) to yield 19f
2
2
1
(
(
(
393 mg, 78%, R = 0.57 (S )) as yellow oil: H NMR (CDCl ) δ ppm 1.77 - 1.87 (m, 2 H (NCHCH )), 1.89 (s, 3 H
f 1 3 2
Me)), 2.01 (s, 3 H (Me)), 2.11 (s, 3 H (Me)), 2.22 - 2.33 (m, 2 H (=CHCH )), 2.52 - 2.59 (m, 2 H (CH N)), 3.21
2
2
dd, J = 3.85, 8.72 Hz, 1 H (NCH
2
)), 3.51 - 3.59 (m, 1 H (CH
2
OH)), 3.77 - 3.85 (m, 1 H (CH
2
OH)), 4.38 - 4.51 (m, 2
2
H (CH NH)), 5.99 (t, J = 7.31 Hz, 1 H (C=CH)), 6.71 - 6.86 (m, 4 H (Ar)(SCHCH)), 7.02 (d, J = 5.13 Hz, 1 H
1
3
(
(
(
(
SCHCH)), 7.20 - 7.24 (m, 2 H (SCH)), 7.82 (t, J = 6.03 Hz, 1 H (CONH)). C NMR (CDCl ) δ ppm 14.32
3
thiophene-Me), 14.61 (thiophene-Me), 27.27 (=CH-CH -), 28.30 (-N-CH-CH -), 36.72 (N-Me), 37.62
2
2
CO(NH)CH ), 53.96 (-CH -N-), 61.15 (-CH -OH), 66.20 (-N-CH-), 103.39 (-CH-C-F), 104.06 (-CH-C-F), 111.41
2
2
2
Ar), 121.35 (Ar), 122.59 (-C=CH-), 124.48 (2 C, thiophene), 129.60 (2 C, thiophene), 131.40 (-S-CH-), 133.70 (-S-
CH-), 134.77 (-C-Methiophene), 135.50 (-C-Methiophene), 138.87 (-C=CH-), 159.83 (C-F para), 163.12 (C-Forto), 174.14
C=O). Anal. calcd for C26 (504.66): C: 61.88, H: 5.99, N: 5.55, S: 12.71; Found C: 61.93, H: 5.61, N:
(
30 2 2 2 2
H F N O S
+
+
5
.58, S: 12.75. HRMS-ESI m/z [M+H] calcd for C H F N O S : 505.1795, found: 505.1785.
26 30 2 2 2 2
4
.1.3.7 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl](methyl)amino}-N-[2-(4-fluorophenyl)ethyl]-4-
hydroxybutanamide (20)
2
8

Compound 20 was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
(3H)-one (17) (1.5 mmol, 0.54 g) and 2-(4-fluorophenyl)ethan-1-amine (3 mmol, 0.42 g) in 7 ml dry THF. The
2
obtained crude product was purified by column chromatography over silica gel (CH
2
Cl
2
/acetone = 9:1) to yield 20
1
(
568 mg, 76%, R
f
= 0.58 (S
1
)) as yellow oil: H NMR (CDCl
3
) δ ppm 1.69 - 1.82 (m, 2 H (CH
2
CH OH)), 1.88 (s, 3 H
2
(
Me)), 1.99 (s, 3 H (Me)), 2.02 (s, 3 H (Me)), 2.15 - 2.27 (m, 2 H (C=CHCH )), 2.43 - 2.55 (m, 2 H (NHCH CH )),
2 2 2
2
2
.64 - 2.77 (m, 2 H (NHCH )), 3.12 (dd, J = 4.23, 8.08 Hz, 1 H (NCH)), 3.30 - 3.42 (m, 1 H (CH N)), 3.44 - 3.59 (m,
2 2
H (CH OH)), 3.69 - 3.78 (m, 1 H (CH N)), 5.95 (t, J = 7.31 Hz, 1 H (C=CH)), 6.71 (d, J = 5.13 Hz, 1 H
2
2
(
(
(
SCHCH)), 6.80 (d, J = 5.13 Hz, 1 H (SCHCH)), 6.88 (t, J = 8.72 Hz, 2 H (Ar)), 7.03 (dt, J = 5.26, 9.30 Hz, 3 H
1
3
SCH)(Ar)), 7.17 (d, J = 5.13 Hz, 1 H (SCH)), 7.54 (t, J = 5.90 Hz, 1 H (CONH)). C NMR (CDCl
thiophene-Me), 14.64 (thiophene-Me), 27.03 (=CH-CH -), 28.26 (-N-CH-CH -), 34.80 (-CH -Ar), 37.70 (N-Me),
0.51 (CO(NH)CH ), 53.88 (-CH -N-), 61.45 (-CH -OH), 66.69 (-N-CH-), 115.10 (-CH-C-F), 115.37 (-CH-C-F),
3
) δ ppm 14.38
2
2
2
4
1
1
2
2
2
22.58 (-C=CH-), 129.02 (Ar), 129.62 (Ar), 130.21 (2 C, thiophene), 131.50 (2 C, thiophene), 132.51 (-S-CH-),
1
33.83 (-S-CH-), 134.42 (Ar), 134.74 (-C-Methiophene), 135.55 (-C-Methiophene), 138.91 (-C=CH-), 161.51 (d, JC-
F=244.49 Hz, C-F), 174.11 (C=O). Anal. calcd for C H F N O S (500.69): C: 64.77, H: 6.64, N: 5.59, S: 12.81;
27
33
2
2
2 2
+
+
Found C: 64.83, H: 6.71, N: 5.64, S: 12.86. HRMS-ESI m/z [M+H] calcd for C27
H
33
F
2
N
2
O
2
S
2
: 501.2046, found:
5
01.2047.
4
.1.3.8 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-4-hydroxy-N-(1-phenylethyl)butanamide
(
21)
Compound 21 was prepared using 3-((4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-dihydrofuran-
(3H)-one (17) (1 mmol, 0.36 g) and 1-phenylethan-1-amine (2 mmol, 0.24 g) in 5 ml dry THF. The obtained crude
product was purified by column chromatography over silica gel (CH Cl /acetone = 9:1 → CH Cl /ethyl acetate =
2
2
2
2
2
1
7
:3) to yield 21 (232 mg, 48%, R = 0.54 (S )) as yellow oil: H NMR (CDCl ) δ ppm 1.77 - 1.91 (m, 5 H (NCHCH ;
f
1
3
2
Me)), 2.03 (s, 3 H (Me)), 2.14 (s, 3 H (Me)), 2.19 - 2.32 (m, 2 H (=CHCH
2
)), 2.49 - 2.63 (m, 2 H (CH
2
N)), 3.25 (dd,
J = 3.59, 8.98 Hz, 1 H (NCH)), 3.47 (s, 3 H (Me)), 3.54 - 3.62 (m, 1 H (CH
.30 - 4.53 (m, 1 H (CHNH(CO))), 5.96 (t, J = 7.31 Hz, 1 H (C=CH)), 6.73 (d, J = 5.14 Hz, 1 H (SCHCH)), 6.85 (d,
J = 5.13 Hz, 1 H (SCHCH)), 7.05 (d, J = 5.10 Hz, 1 H (SCH)), 7.10 -7.40 (m, 6 H (Ar; SCH)), 7.81 (t, J = 5.91 Hz, 1
2 2
OH)), 3.79 - 3.88 (m, 1 H (CH OH)),
4
1
3
H (CONH)). C NMR (CDCl ) δ ppm 14.44 (thiophene-Me), 14.75 (thiophene-Me), 27.00 (CO(NH)CH-Me), 28.51
3
(
=CH-CH -), 28.62 (-N-CH-CH -), 35.01 (N-Me), 37.72 (CHNH(CO)), 51.32 (-CH -N-), 53.33 (-CH -OH), 60.90 (-
2 2 2 2
N-CH-), 122.60 (-C=CH-), 122.81 (Ar), 124.31 (Ar), 126.42 (Ar), 127.30 (Ar), 127.41 (Ar), 128.81 (2 C, thiophene),
29.60 (2 C, thiophene), 131.32 (-S-CH-), 133.23 (-S-CH-), 133.60 (Ar), 135.41 (-C-Methiophene), 137.41 (-C-
Methiophene), 139.31 (-C=CH-), 172.40 (C=O). Anal. calcd for C H N O S (482.70): C: 67.18, H: 7.10, N: 5.80, S:
1
27
34
2
2 2
+
+
1
4
3.29; Found 67.21, H: 7.11, N: 5.85, S: 13.31. HRMS-ESI m/z [M+H] calcd for C H N O S :483.2140, found:
27 34 2 2 2
83.2133.
4
.1.3.9 2-{[4,4-bis(3-methylthiophen-2-yl)but-3-en-1-yl)(methyl)amino)-4-hydroxy-1-morpholinobutan-1-one (22)
2
9