Development of an N-Acyl Amino Acid That Selectively Inhibits the Glycine Transporter 2 to Produce Analgesia in a Rat Model of Chronic Pain

Article abstract of DOI:10.1021/acs.jmedchem.8b01775

Inhibitors that target the glycine transporter 2, GlyT2, show promise as analgesics, but may be limited by their toxicity through complete or irreversible binding. Acyl-glycine inhibitors, however, are selective for GlyT2 and have been shown to provide analgesia in animal models of pain with minimal side effects, but are comparatively weak GlyT2 inhibitors. Here, we modify the simple acyl-glycine by synthesizing lipid analogues with a range of amino acid head groups in both l- and d-configurations, to produce nanomolar affinity, selective GlyT2 inhibitors. The potent inhibitor oleoyl-d-lysine (33) is also resistant to degradation in both human and rat plasma and liver microsomes, and is rapidly absorbed following an intraperitoneal injection to rats and readily crosses the blood-brain barrier. We demonstrate that 33 provides greater analgesia at lower doses, and does not possess the severe side effects of the very slowly reversible GlyT2 inhibitor, ORG25543 (2).

Full text of DOI:10.1021/acs.jmedchem.8b01775

Subscriber access provided by Iowa State University | Library
Article
Development of an N-acyl amino acid that selectively inhibits the
glycine transporter 2 to produce analgesia in a rat model of chronic pain
Shannon N Mostyn, Tristan Rawling, Sarasa Mohammadi, Susan Shimmon,
Zachary J Frangos, Subhodeep Sarker, Arsalan Yousuf, Irina Vetter,
Renae M. Ryan, MacDonald J. Christie, and Robert J. Vandenberg
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01775 • Publication Date (Web): 04 Feb 2019
Downloaded from http://pubs.acs.org on February 5, 2019
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 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 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.
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 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
Development of an N-acyl amino acid that selectively inhibits the glycine
transporter 2 to produce analgesia in a rat model of chronic pain
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
Shannon N. Mostyn,1,§ Tristan Rawling,2, §,*, Sarasa Mohammadi1,, Susan Shimmon2, Zachary J.
Frangos1, Subhodeep Sarker1, Arsalan Yousuf1, Irina Vetter3, Renae M. Ryan1, Macdonald J.
Christie1, and Robert J. Vandenberg1,*
1. Discipline of Pharmacology, School of Medical Sciences, University of Sydney, Sydney,
NSW, 2006, Australia
2. School of Mathematical and Physical Sciences, Faculty of Science, The University of
Technology Sydney, Sydney, NSW, 2007, Australia
3. Institute for Molecular Bioscience & School of Pharmacy, The University of Queensland,
Brisbane Qld 4072 Australia
§ The first two authors contributed equally.
Corresponding authors email: robert.vandenberg@sydney.edu.au
tristan.rawling@uts.edu.au
Key words: Glycine Transport Inhibition, N-acylamino acid, analgesia, GlyT2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 57
1
2
3
4
Abstract
5
6
7
8
9
Inhibitors that target the glycine transporter 2, GlyT2, show promise as analgesics but may be limited
by their toxicity through complete or irreversible binding. Acyl-glycine inhibitors, however, are
selective for GlyT2 and have been shown to provide analgesia in animal models of pain with minimal
side effects, but are comparatively weak GlyT2 inhibitors. Here, we modify the simple acyl-glycine
by synthesising lipid analogues with a range of amino acid head groups in both L- and D-
configurations, to produce nanomolar affinity, selective GlyT2 inhibitors. The potent inhibitor oleoyl-
D-lysine (33) is also resistant to degradation in both human and rat plasma and liver microsomes, and
is rapidly absorbed following an intraperitoneal injection to rats and readily crosses the blood brain
barrier. We demonstrate that 33 provides greater analgesia at lower doses, and does not possess the
severe side effects of the very slowly reversible GlyT2 inhibitor, ORG25543 (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 3 of 57
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
9
Disease or injury to the somatosensory nervous system can cause long-term neuropathic pain,
characterised by the associated hyperalgesia (amplified pain sensation) and allodynia (pain response
to innocuous stimuli), which produce distinctive symptoms of numbness, burning, shooting, and
electric-like pain in sufferers.1 Changes within the central nervous system (CNS) produce disordered
synaptic pathways and central sensitisation which allow for symptoms to persist long after the site of
injury has healed.2-3 There is a great need to develop new drugs to treat chronic neuropathic pain as a
large number of cases are refractory to conventional analgesics4, perhaps due to these adaptations
within the CNS. Additionally, many current analgesics, such as opioids, have an unacceptable side
effect profile and potential for abuse.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
Recently, it has been shown in rodents suffering from neuropathic pain that there are reduced
inhibitory glycinergic inputs within the spinal cord dorsal horn,6 where primary sensory information is
relayed before transmission to the brain for cognitive processing.7 A reduction in inhibitory control
may allow for hyperexcitability and disinhibition of pain signalling.3,8 Inhibition of glycine
transporters to increase glycine at inhibitory synapses may therefore restore this control and reduce
spontaneous nerve activity and associated pain symptoms.
There are two subtypes of glycine transporters within the CNS, GlyT1 and GlyT2.9 GlyT1 is
expressed by astrocytes surrounding both inhibitory glycinergic and excitatory glutamatergic synapses
throughout the brain and spinal cord, whereas GlyT2 is located on presynaptic inhibitory glycinergic
neurons predominantly in the spinal cord, including within the dorsal horn.10 Conflicting reports have
been made concerning the effectiveness of GlyT1 inhibitors, with one study showing an initial
increase in pain due to NMDA receptor stimulation before analgesics effects are apparent11, whilst
other studies have shown analgesic effects.12 On the other hand, selective inhibition of GlyT2 could
selectively increase glycinergic neurotransmission in the ascending pain pathway and provide
effective relief from pain symptoms.3 This has been demonstrated using partial knockdown of GlyT2
via targeted siRNA, which ameliorates pain symptoms in a sciatic nerve injury model of neuropathic
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 57
1
2
3
4
5
6
7
8
pain.11 The role of glycine in the development of pain and its utility in producing analgesia has driven
the development of a number of glycine transport inhibitors (Figure 1) for the treatment of
neuropathic pain.
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 1: Chemical structures of selected GlyT2 inhibitors.
ALX139312 (1) and ORG2554314 (2) were among the first potent inhibitors of GlyT2 to be developed
(Figure 1). 1 inhibits GlyT2 with an IC50 of 100 nM15 and has been shown to be anti-allodynic when
intrathecally or intravenously injected into rodents suffering neuropathic pain.11-13,16 However, 1 does
not readily cross the blood brain barrier (BBB) and following IV administration the ratio of unbound
drug in brain and plasma (Kp,uu) is only 0.05. For this reason, 1 is generally delivered directly into
the CNS to produce analgesia in rodent models of pain,17 although analgesia has been reported
following IV delivery.11
2 is a potent inhibitor of GlyT2 (IC50 16 nM), and unlike 1, has no appreciable activity at GlyT1.14
2
ameliorates hyperalgesia and allodynia,11,15 but has also been shown to produce tremors, seizures, and
ACS Paragon Plus Environment

Page 5 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
death in rodents,15 likely due to inhibition of vesicular refilling and reduced recycling of glycine into
the synapse, proposed to be via irreversible, or very slowly reversible, binding to GlyT2. 2 analogues
have since been synthesised to improve reversibility.15 More recently, Hayashi and colleagues
synthesised a series of phenoxymethylbenzamides which are potent inhibitors of GlyT2.18 One
derivative, GT-0198 (3) (Figure 1), is a potent and reversible inhibitor of GlyT2 (IC50 105 nM),
permeates the BBB following oral administration, and is analgesic in a partial nerve ligation model of
pain.19
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
Endogenous lipids have also been identified as novel agents capable of inhibiting glycine transporters.
N-arachidonyl glycine (NAGly) (5) is a partial, reversible, non-competitive inhibitor of GlyT2 (IC50
~9 µM) with little activity at GlyT1.20 Other endogenous N-acyl amino acids also inhibit GlyT2; with
an order of potency of oleoyl L-carnitine (340 nM) > N-oleoyl glycine (NOGly) (4) (500 nM) >
palmitoyl L-carnitine (600 nM) >> 5 ~ N-arachidonyl L-alanine (9 µM) > N-arachidonyl GABA (12
µM).20-21 These lipid inhibitors of GlyT2 are amphipathic; comprised of a long flexible lipid tail
conjugated to a charged or polar head group (Figure 1).
5 is very similar in structure to the endogenous cannabinoid anandamide but has no activity on the
cannabinoid receptors. It has been suggested that the analgesic effects of 5 are due to inhibition of
Fatty Acid Amide Hydrolase (FAAH), which would elevate levels of anandamide to stimulate
cannabinoid receptors.22 While the analgesic activity of 5 may in part be due to inhibition of FAAH, it
has also been shown that 5 directly increases glycinergic synaptic transmission within the dorsal horn
through inhibition of GlyT2, providing a cellular mechanism for its analgesic actions. 5 enhances the
actions of both exogenously applied and endogenous glycine at inhibitory glycinergic synapses of
lamina II of the superficial dorsal horn.23 5 also prolongs the decay phase of evoked inhibitory post-
synaptic currents, but has no effect on evoked N-methyl-D-aspartate mediated excitatory postsynaptic
currents; suggesting 5 exerts its effects by selectively blocking GlyT2 and increasing glycine
exclusively at inhibitory synapses. Application of 5 to rodents suffering from either inflammatory or
neuropathic pain reduces mechanical allodynia and thermal hyperalgesia.24-25 5 is also analgesic in a
formalin-induced pain model, where the second phase of pain thought to depict afferent input and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 57
1
2
3
4
5
6
7
8
9
central sensitisation in the dorsal horn is reduced.26 In contrast to 1 and 2, no adverse effects are
produced following intrathecal injection of 5.24 The lack of apparent side effects may be a
consequence of partial and reversible inhibition of GlyT2 by 5, which would allow for sufficient
glycine re-uptake into presynaptic terminals for loading of vesicles, but also slow the removal of
glycine to increase transient concentrations at inhibitory glycinergic synapses while minimising spill-
over into excitatory glutamatergic synapses.
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 in vivo efficacy and lack of side effects of 5 make it a promising lead compound for analgesic
drug development. To improve GlyT2 inhibitory potency we prepared a series of acyl-glycine
analogues incorporating lipid tails of varied chains lengths and double bond positions.26 We showed
that the lipid tail constituent is critical in their activity as inhibitors of GlyT2, and that the optimal
chain is 18 carbons in length with a cis-double bond in the ω-8, -9 or -10 positions. Thus, 4 was
identified as among most potent analogues in the series, and partially inhibited (max. 67%) GlyT2
with an IC50 of 500 nM and had no effect on GlyT1 at concentrations up to 10 µM.
In this study we report on the development of a second series of N-acyl amino acids as GlyT2
inhibitors. The optimal oleoyl tail (C18 ω9) was retained and conjugated to amino acids and amino
acid derivatives in both the L- and D- configurations. Inhibitory activity at GlyT1 and GlyT2 was
assessed, and select compounds were also screened for metabolic stability, with the most potent,
stable compound, oleoyl-D-Lys (33) tested for BBB permeability, selectivity against a range of
targets, and ability to reduce allodynia in a nerve ligation rat model of neuropathic pain.
Results and Discussion
Library design and Synthesis
We have previously shown that oleic acid tails impart optimal GlyT2 inhibitory activity to N-acyl
amino acids containing glycine head groups.27 To further investigate the structure activity relationship
for N-acyl amino acids as GlyT2 inhibitors, we synthesised a series of analogues in which an oleoyl
tail was conjugated to a number of amino acid head groups. A range of aliphatic, aromatic, polar,
ACS Paragon Plus Environment

Page 7 of 57
Journal of Medicinal Chemistry
1
2
3
4
negatively charged, and positively charged amino acids were used, in both the L- and D-
5
6
configurations.
7
8
9
N-acyl amino acids 6 – 31(Table 1), which lack a primary amine on the head group, were synthesised
in two steps (Scheme 1a). In the first step ester protected and enantiomerically pure D- and L-amino
acids were conjugated via amide bonds to oleic acid using the peptide coupling agent EDCI. The
resulting intermediates 6a – 31a were readily purified by silica gel chromatography and isolated in
good yields. The ester groups in 6a – 31a were then hydrolysed under basic conditions, and
acidification of the reaction mixtures precipitated 6 – 31, which were isolated in high yields by
filtration.
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
Compounds 32 – 38, which possess primary amine groups on the head group, were synthesised using
ester and BOC protected head groups (Scheme 1b). BOC protection was used to prevent amide bond
formation between the side arm amine and oleic acid. EDCI couplings to oleic acid afforded the BOC-
protected intermediates 32a – 37a. De-esterification by alkaline hydrolysis afforded 32b – 37b,
followed by removal of the BOC-protecting groups using HCl yielded the final products 32 – 37 as
hydrochloride salts. The ester 38 was prepared by removal of the BOC group in 32a using HCl in
ether.
Scheme 1. Synthesis of N-acyl amino acids without (a) or with (b) primary amine-substituted head
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 57
1
2
3
4
groups. Reagents and conditions: (i) EDCI, HOBt, NEt3, rt, 18h; (ii) NaOH, 40 oC, 3 h, then HCl; (iii)
5
6
HCl, rt, 4 h.
7
8
9
Inhibitory Activity at GlyT2 and GlyT1.
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
Application of glycine to oocytes expressing human GlyT1b or GlyT2a (herein referred to as GlyT1
and GlyT2) generates inward transport currents, which are reduced by co-application of the
synthesised N-acyl amino acids 6 – 38 (Figure 2A, B). Application of increasing concentrations of N-
acyl amino acid generated cumulative inhibition responses (representative trace shown, Figure 2C).
Concentration responses curves for selected compounds are shown in Figure 3 and the % max
inhibition and apparent IC50 values for each N-acyl amino acid at GlyT1 and GlyT2 shown in Table 1.
The specificity of inhibition of glycine transport was also measured using [3H]-glycine uptake by
oocytes expressing GlyT2 and GlyT1 in the presence of 100 nM and 1 µM 32 (Figure S2), confirming
the observations with electrophysiological measurements.
ACS Paragon Plus Environment

Page 9 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 2. Representative current traces from oocytes expressing glycine transporters, clamped at -60
mV. A. The EC50 concentration of glycine (30 µM) was applied to an oocyte expressing GlyT2 (open
bar) to produce an inward current, which was reduced by application of 0.06 µM 32 (grey bar). B.
GlyT1 mediated glycine currents were not reduced by 32, even at 3µM. C. Cumulative concentration
response curves were generated by applying increasing concentrations of N-acyl amino acid (13 in
this example, grey segmented bar) to reduce glycine induced currents at GlyT2.
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
Prolonged application of >10 µM N-acyl amino acids to oocytes expressing transporters or uninjected
oocytes did not produce any currents, with the exception of acyl-serine (15 and 16) and positively
charged N-acyl amino acid compounds. Concentration response curves for these inhibitors were
therefore limited to 3 µM to circumvent non-specific membrane effects.
Figure 3. N-acyl amino acids inhibit glycine transport currents of GlyT2 expressing Xenopus
laevis oocytes. 30 μM glycine transport currents were measured in the presence of N-acyl amino acids
to generate concentration inhibition curves A. Representative N-acyl amino acids from aliphatic (8),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 57
1
2
3
4
5
6
7
8
acidic (17), aromatic (26), and positively charged (32) groups. B. Positively charged oleoyl L-amino
acids (28 and 30) and the backbone modified oleoyl N-acyl compounds (37 and 38). C. Positively
charged oleoyl D-amino acids (29, 31, and 33). D. “Lysine-like” oleoyl L-amino acids, where the
9
+
pendant NH3 is linked to the amino acid backbone with 3 (34), 2 (35), or 1 (36) carbons. Only 1 data
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
point is presented for 36 at the maximal concentration (3 μM) as the data could not be reliably fit. The
responses are normalized mean values ± SEM (n ≥ 3 cells) fit using least-squares analysis. L-Val
(8);
L-Asp (17); L-Trp (26) L-His (28); D-His (29); L-Arg (30); D-Arg (31); L-Lys
(32); D-Lys (33); L-C3-Lys (34); L-C2-Lys (35); L-C1-Lys (36);
Lys OMe (38).
C5-NH3 (37);
L-
N-acyl amino acid with head groups in the L-configuration all reduced GlyT2 transport currents, with
IC50 values ranging from 25.5 nM – 4.35 µM (Table 1). For compounds containing small aliphatic
side chains (L-Ala (6) and L-Val (8)) weak inhibition of GlyT2 was observed with IC50 values above 1
µM, whereas the bulkier iso-butyl group in the L-Leu analogue 10 produced a large improvement in
potency (IC50 143 nM). The iso-butyl branched side chain in 10 appears to be beneficial to activity, as
the straight chain n-butyl analogue 14 was 2-fold less potent than 10. Interestingly, although
methionine is often grouped with aliphatic amino acids, the L-Met analogue 12, which is structurally
analogous to 14 but contains a side chain sulphur, is a potent, near complete inhibitor (29.2 nM,
91.2%). This ~9-fold improvement in activity indicates the sulfur is providing additional interactions
with GlyT2. The methionine sulfur atom may also behave as an electrophile; the σ-hole of the S
possibly interacting with an oxygen atom, or with aromatic π electrons in the binding site.28-27
For acidic amino acids, where their side chains would be negatively-charged under the assay
conditions, conjugation to the oleoyl tail produces compounds which are all relatively low potency
inhibitors of GlyT2 (865 nM (17) and 4.35 µM (19)). Interestingly, the IC50 of the amide derivative of
17, oleoyl L-Asn (21), is not largely changed (613 nM) compared to 17, which suggests that the
negative charge is not providing any significant binding or repulsive interactions with GlyT2.
ACS Paragon Plus Environment

Page 11 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The N-acyl amino acids with aromatic side chains (22, 24, and 26) are all relatively potent inhibitors
of GlyT2. The L-Phe analogue 19 inhibited GlyT2 with an IC50 of 214 nM, and lengthening the alkyl
chain between the aromatic ring and amino acid back bone marginally improved potency (24; IC50
151 nM). Extending the side chain, as in the L-Trp analogue 26 further improved potency (IC50 54.6
nM).
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 most potent N-acyl amino acids in the series were those with positively ionisable groups on the
side chain. Thus, 28 (L-His), 30 (L-Arg) and 32 (L-Lys) inhibited GlyT2 with IC50s of 130, 47.9 and
25.5 nM, respectively. Encouraged by the activity of 32, we prepared a series of derivatives of 32 to
gain additional structure-activity insights. Inhibitory activity decreased as the butyl side chain in 32
was shortened to propyl (34), ethyl (35) and methyl (36). These findings are consistent with aromatic
N-acyl amino acids, where extension of the side chain improves potency. The importance of the
carboxylic acid group to activity was examined through two lysine derivatives; 37, where the
carboxylic acid is removed, and 38 which instead contains a methyl ester. In both cases, the
compounds were less active than 32 (IC50 560 nM and 91.3 nM respectively), with complete removal
of the carboxylic acid reducing potency by ~22-fold.
Pairwise comparison of D- and L- enantiomers showed that with the exception of serine analogues 15
and 16, the L-enantiomers were equal, and in most cases, superior GlyT2 inhibitors. The IC50 values
obtained for L-analogues with less bulky or flexible head group side chains were between 2-9 fold
lower than their corresponding the D-enantiomers. For example, 33 (D-Lys) inhibits GlyT2 with an
IC50 of 48.3 nM compared with 25.5 nM for 32. Interestingly, the decrease in activity of D-
enantiomers was much greater when the head group side chain was bulkier and/or less flexible; the D-
enantiomers 11 (D-Leu), 23 (D-Phe), 25 (D-C2-Phe), and 27 (D-Trp) were inactive at the highest assay
concentration of 10 µM, compared with their L-enantiomers that are all nanomolar inhibitors (IC50s
from 54.6 – 214 nM).
Given the wide range of head groups that are tolerated, the binding region must be large and/or
flexible to accommodate a number of different chemical groups and conformations. The chemical
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 57
1
2
3
4
5
6
7
8
9
nature of the binding site can also be inferred through structure activity data. We propose that the
head group binding site should be comprised of residues that strongly coordinate a positively charged
or aromatic group, but do not repel a negatively charged group. It is therefore unlikely that ionic
interactions mediate binding, but rather aromatic residues that would generate similar energies for
cation-π and π-π interactions for the most potent N-acyl amino acids.
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 1: Inhibitory activity of N-acyl amino acids 2, 4, 6 – 38 at GlyT2 and GlyT1*
R
O
GlyT1
IC50
Max. GlyT2
inhibition
(%)
GlyT2
Head
Compound
R
group
IC50 (nM)
(µM)
3.8
2
(ORG25543)
n.d
93.00 ± 0.04
66.8 ± 2.9
(2.4-6.0)
HO2C
4 (NOGly)26
Gly
500
>30
NH
1230
HO2C
HO2C
a
>30
6
7
L-Ala
D-Ala
89.8 ± 5.8
81.7 ± 6.2
NH
NH
(649 – 2340)
2330
>30
>30
(1.26 – 4.34)
1170
HO2C
8
L-Val
90.1 ± 5.9
NH
(643 – 2120)
ACS Paragon Plus Environment

Page 13 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
1000
HO2C
9
D-Val
>30
56.7 ± 5.1
NH
(448 – 2570)
7
8
9
143
HO2C
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
10
11
12
13
14
15
16
17
18
L-Leu
D-Leu
L-Met
D-Met
>10
>30
>3
72.2 ± 3.8
35.6 ± 2.6
91.2 ± 2.7
68.6 ± 8.1
64.5 ± 6.9
73.1 ± 5.1
73.6 ± 4.4
76.7 ± 5.3
92.5 ± 6.9
NH
(70.7 – 287)
HO2C
a
>30 µM
NH
29.2
(18.6 – 45.7)
259
HO2C
S
S
NH
HO2C
HO2C
>3
NH
(79.6 – 845)
259
L-
>10
>3
NH
NorLeu
(166 – 372)
496
HO2C
OH
OH
L-Ser
D-Ser
L-Asp
D-Asp
NH
NH
(308 – 799)
64.8
HO2C
>3
(31.1 – 135)
865
HO2C
OH
>10
>10
NH
O
O
(468 – 1600)
1810
HO2C
OH
NH
(1050 – 3140)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 57
1
2
3
4
5
6
7
8
9
O
4350
HO2C
OH
OH
19
20
L-Glu
D-Glu
>10
>10
100 ± 14.0
91.6 ± 7.9
NH
NH
(2350 – 8070)
O
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
3750
HO2C
HO2C
HO2C
HO2C
(2240 – 6260)
613
NH2
21
22
23
L-Asn
L-Phe
D-Phe
>10
>10
>10
92.7 ± 7.2
54.7 ± 3.8
42.3 ± 2.8
NH
O
(265 – 1420)
214
NH
NH
(86.4 – 530)
a
>10 µM
151
L-C2-
Phe
HO2C
24
25
26
27
>10
>10
>10
>10
75.1 ± 2.9
10.2 ± 1.2
86.2 ± 3.8
14.2 ± 3.5
NH
(88.8 – 257)
D-C2-
Phe
HO2C
a
>10 µM
NH
54.6
HO2C
L-Trp
D-Trp
NH
NH
N
H
(31.4 – 95.8)
HO2C
a
>10 µM
N
H
ACS Paragon Plus Environment

Page 15 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
H
N
130
HO2C
28
L-His
>3
78.4 ± 3.4
N+
H
NH
NH
(84.4 - 199)
H
N
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
313
HO2C
29
30
31
D-His
L-Arg
D-Arg
>3
>3
>3
82.1 ± 3.3
87.2 ± 3.7
70.9 ± 2.2
N+
H
(229 – 889)
NH2
47.9
HO2C
NH NH+2
NH
NH
(27.2 – 84.2)
NH2
69.2
HO2C
NH NH+2
(48.0 – 99.9)
HO2C
HO2C
NH+3
25.5
32
33
34
35
L-Lys
D-Lys
>3
>3
>3
>3
86.8 ± 2.4
91.0 ± 1.9
86.0 ± 2.2
82.7 ± 3.3
NH
(17.2 – 37.7)
NH+3
48.3
(37.2 – 62.7)
31.6
NH
HO2C
NH+3
L-C3-
Lys
NH
(23.1 – 43.3)
152
HO2C
NH+3
L-C2-
Lys
NH
(100 - 231)
b
182
HO2C
NH+3
L-C1-
Lys
36
>3
58.2 ± 7.6
NH
(27.3 – 1210)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 57
1
2
3
4
5
6
7
8
9
NH+3
560
37
38
C5-NH3
>3
>3
86.7 ± 7.2
78.8 ± 5.8
NH
(312 – 1010)
91.3
MeO2C
NH+3
L-Lys-
OMe
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
NH
(37.5 - 222)
* Compounds were tested for inhibition of 30 µM glycine transport by GlyT2 and GlyT1, expressed
in Xenopus laevis oocytes. n ≥ 3 with measurements taken from at least 2 batches of oocytes. Data
presented are mean and 95 % confidence intervals or mean ± SEM.
a. Where significant inhibition was not reached, IC50 are presented as greater than the maximum
concentration of each compound applied.
b. Curve could not reliably be fit from the data
The GlyT2 inhibitor 2 produces adverse side effects that have been attributed to its full and slowly
reversible inhibition of GlyT2.14 We therefore assessed the maximum level of GlyT2 inhibition for all
of our N-acyl amino acids. Positively charged N-acyl amino acids all inhibited glycine transport in the
range 71 – 95% (Table 1). Similarly, all negatively charged N-acyl amino acids produced a high %
max inhibition (77 – 100%). For aliphatic and aromatic N-acyl amino acids, however, the level of
inhibition was influenced by the size and configuration of the side chain. For these compounds, the L-
isomer possessed a higher % max inhibition compared to the D- for all cases (for example 8 and 9
inhibited 90.1 ± 5.9 and 56.7 ± 5.1, respectively). This was particularly apparent for aromatic N-acyl
amino acids, where all of the D- compounds displayed low level or no inhibition at GlyT2. The
differences in level of inhibition by each compound may be a reflection on the stability of GlyT2 in its
inhibited conformation.
Mechanism of binding
GlyT2 possesses the smallest substrate (S1) binding site of the NSS family and is only able to
transport the smallest amino acid, glycine,30,31 which suggests much larger bioactive lipids are
unlikely to bind in the S1 central binding site. To determine the mechanism of inhibition for synthetic
ACS Paragon Plus Environment

Page 17 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N-acyl amino acids, glycine concentration responses were measured in the presence of increasing
concentrations of 32. The Imax of transport was significantly reduced for all concentrations of 32 (%
Imax [10 nM] = 82.8 ± 4.0%*; [30 nM] = 62.4 ± 5.8%***; [1 µM] = 30.2 ± 4.4%***) (Figure 4).
Conversely the EC50 values for glycine were unchanged compared to glycine transport in the absence
of any inhibitor. This indicates that 32 acts as a non-competitive inhibitor of GlyT2, however an
Eadie-Hofstee plot of the data suggests that the mechanism may not be purely non-competitive
(Figure 3S). Lipid inhibitors may therefore display a mixed mechanism of action where binding at an
allosteric site influences glycine binding at the substrate site.
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. 32 reduces concentration-dependent glycine currents. Concentration-dependent glycine
transport currents with increasing concentrations of 32. EC50 (absence of inhibitor) = 15.8 ± 1.3 µM ;
EC50 [10 nM 32] = 14.0 ± 2.7 µM; [30 nM 32] = 23.5 ± 8.0 µM; [1 µM 32] = 12.8 ± 7.6 µM)
ANOVA tests were performed for EC50 and Imax values. gly,
10 nM 32, 30 nM 32, 1 µM
32.
In vitro metabolism
32 emerged as a potent GlyT2 inhibitor that causes a maximum inhibition of 91% and is suitable for
further development. N-acyl amino acids may be metabolised through several pathways that can limit
their in vivo half-lives and therapeutic potential. Degradation occurs primarily through amide bond
hydrolysis carried out by fatty acid amide hydrolase (FAAH) and other hydrolytic enzymes, as well as
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 57
1
2
3
4
oxidations of the lipid chains and head groups.33-34 Thus, the stability of 32 and its D-enantiomer 33
5
6
were assessed in human and rat plasma and liver microsomes.
7
8
9
Table 2. Degradation half-lives of N-acyl amino acids in human and rat plasma.
Degradation half-life (t1/2; min)
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
Compound
Head group
Human plasma
> 1022a
Rat plasma
27
4
Gly
32
33
L-Lys
D-Lys
> 1022a
86
> 1022a
> 1022a
a. Minimal degradation (˂15%) was observed over the course of the 240 minute incubation period.
All compounds were resistant to degradation in human plasma (see Table 2 and Figure S1 in
Supporting Information). In rat plasma 4 was rapidly metabolised with a degradation half-life of 27
minutes. The degradation half-life of 32 improved ~ 3-fold relative to 4, and interestingly 33, the D-
enantiomer of 32, did not exhibit measurable degradation in either species. A similar pattern of
stability was observed in liver microsomes (Table 3). Thus, 32 and 33 were both stable in human liver
microsomes, and in rat liver microsomes 32 was rapidly metabolised (t1/2 = 54 minutes) while 33 was
minimally degraded over the course of the assay.
The improved stability of 32 relative to 4 may arise from the lysine side chain of 32. The side chain is
adjacent to the amide bond and can act as a steric shield to protect the amide from hydrolysis. This
phenomenon is well-known in medicinal chemistry and has been used to develop stabilised analogues
of anandamide, a structurally related fatty acid amide.35-36 The remarkable stability of 33 relative to 32
is likely to result of the unnatural head group orientation in 33, which hinders its ability to fit within
the active site of hydrolytic enzymes. Indeed, incorporation of D-amino acids into synthetic peptides is
employed to produce peptide bonds that are resistant to the actions of proteases and other endogenous
enzymes.37
ACS Paragon Plus Environment

Page 19 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Table 3. Degradation half-lives of 32 and 33 in human and rat liver microsomes.
Degradation half-life (t1/2; min)
Compound
Head group
9
Human
> 255a
Rat
54
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
32
33
L-Lys
D-Lys
> 255 a
> 255 a
a. Minimal degradation (˂15%) was observed over the course the 60 minute incubation period.
Taken together, these studies show that incorporation of unnatural head group configuration can
improve stability to N-acyl amino acids, and that 33 is a metabolically stable candidate for in vivo
evaluation.
Selectivity of 33
As shown in Table 1, all N-acyl amino acids were selective for GlyT2 over the closely related GlyT1.
For 33, GlyT2 was potently inhibited (IC50 48.3 nM), while little inhibition of GlyT1 was seen at
maximal concentrations (3 µM). We further explored the selectively of 33 at 1 and 10 µM against
other targets, including transporters, GPCRs, and ion-channels (full panel results in Tables 2S, 3S,
4S). Three additional targets were identified as potential binding sites for 33. 10 µM 33 caused greater
than 50% displacement of radioactive ligands for the Adrenergic α2A receptor, norepinephrine
transporter, µ opioid receptor (µOR). As µORs are a well established target for analgesics, we further
examined the mechanism of interaction by 33 in AtT20 cells expressing µORs (see Supporting
Information). We found 33 had no activity as a µOR agonist or antagonist at 1 µM, which suggests 33
is unlikely to produce analgesia via this mechanism.
We also tested 33 at Nav, Cav channels, and α7 nicotinic acetylcholine receptors (Table 3S) which are
known pain targets37 to rule out analgesia through this mechanism, and no significant inhibition was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 57
1
2
3
4
5
6
7
8
9
observed. The cannabinoid receptors, CB1 and CB2, as well as the sphingosine-1-phosphate receptor
are targeted by similar lipid based ligands, such as anandamide, and so these receptors were also
screened for 33 activity (Table 4S), but no activity as either an agonist or antagonist was observed for
all three receptors.
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
In vivo pharmacokinetics
GlyT2 is predominantly expressed in the dorsal horn of the spinal cord and for GlyT2 inhibitors to
show in vivo efficacy they must cross the BBB. Structurally related fatty acid amides such as
methanandamide39 and DHA-dopamine (NMI 8739)40 are both BBB permeable, and we anticipated
that 33 would act similarly. We tested this directly by studying the pharmacokinetics and brain
penetration of 33 in vivo.
Figure 5. Plasma and brain concentrations of 33 in male Sprague Dawley rats following ip
administration at 27.5 mg/kg.
ACS Paragon Plus Environment

Page 21 of 57
Journal of Medicinal Chemistry
1
2
3
4
Table 4. Pharmacokinetic parameters of 33 after ip administration of a single 27.5 mg/kg dose in
5
6
Sprague Dawley rats.
7
8
9
Parameter
Cmax (ng/mL, ng/g)
Tmax (h)
Plasma
16690
1
Brain
1854
24
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
t1/2 (h)
10
c.n.c.
38479
AUC0-24 (ng*h/mL, ng*h/g)
Time averaged B/P ratio
122473
0.31
c.n.c – Could not calculate as the flat profile precluded an estimation of half life
Male Sprague Dawley rats received a single intraperitoneal dose of 33 (27.5 mg/kg) and plasma and
brain concentrations were measured over a 24-hour period (Figure 5). The Cmax, Tmax, terminal half-
life (t1/2), AUC0-24h, and total brain-to-plasma ratio (B/P) calculated from the data are presented in
Table 4. Following administration 33 was rapidly adsorbed with maximum plasma concentrations
observed at 1 hour. This was followed by a clearly defined terminal phase where 33 was slowly
cleared from plasma with a terminal half-life (t1/2) of 10 hours. 33 was also rapidly taken up into the
brain, however unlike in plasma, there was no clear terminal phase and brain concentrations of 33
remained relatively constant over the 24 hour sampling period. Accordingly, the B/P ratio increased
over the sampling period from 0.08 at 1 hour to 1.1 at 24 hours, and a time averaged B/P ratio of 0.31
was calculated from the AUC0-24h values.
Encouragingly, Cmax for 33 in brain was 1854 ng/g (approximately 4.5 µM), which although well
above the IC50 for 28 at GlyT2, is comprised of both bound and unbound fractions of 33. Given that
33 it is likely to be highly bound within the CNS, we assessed drug binding in vitro. Rat brain
homogenate was spiked with 33 and incubated for 4.2 hours, with ultracentrifugation used to separate
bound and unbound fractions. From these studies 33 was estimated to be 99.98% bound in rat brain.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 57
1
2
3
4
5
6
7
8
9
Thus, at Cmax concentrations of 33 in brain, the concentration of unbound drug (퐶unbound) is
approximately 90 nM (cf. IC50 = 48.3 nM). Together, these studies indicate that systemic
administration of 33 can produce unbound concentrations of 33 within the CNS required for
analgesia. The high degree of binding may also explain the slow rate of elimination of 33 from the
brain. It is likely that 33 is mostly sequestered to brain tissue once it crosses the BBB, and therefore
unable to rapidly leave the CNS.
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
Analgesia
Based on the pharmacokinetic findings, the analgesic efficacy of 33 was assessed in the same species.
Male Sprague Dawley rats received a single intraperitoneal bolus of 33 and the efficacy in reversing
allodynia in a neuropathic pain model was assessed over 6 h. A significant main interaction effect of
treatment over time was observed (F (18, 138) = 1.765, p < 0.05), and the 30 mg/kg dose of 33
produced significant analgesia over the first 90 min after injection compared to vehicle (p < 0.001 at
15 and 30 min, p < 0.01 at 60 and 90 min; Bonferroni’s multiple comparisons test; Figure 6 A). A
lower dose of 3 mg/kg alleviated allodynia in a manner similar to the positive control (2, 30 mg/kg).
Mild side effects were observed in one 33-treated (30 mg/kg) animal up to 15 min post injection,
where localised pain (abdominal contraction) was evident at the ip site of drug delivery (p > 0.05
compared to vehicle treated group). In contrast, more severe side effects were observed in 4 out of the
6 animals treated with 2 (30 mg/kg; p < 0.01 compared to vehicle). 2-treated animals remained
recumbent and exhibited acute abdominal constriction for up to 60 min, with two animals exhibiting
writhing behaviour. No side effects were seen in the vehicle treated animals so these behaviours are
likely due to on/off-target effects of the administered drugs (Figure 6 B). Thus, at the equivalent dose,
33 was able to offer significantly greater analgesia, with milder side effects, than the GlyT2 inhibitor
2.
ACS Paragon Plus Environment

Page 23 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
33
33
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
33
33
Figure 6. In vivo 33 reverses allodynia with fewer side effects than 2 in a neuropathic pain model
in male Sprague Dawley rats. A. Significant reversal of allodynia was observed after ip
administration of 33 (30 mg/kg) compared to vehicle. Non-significant reversal of allodynia was
observed for 33 at 3 mg/kg and 2 at 30 mg/kg. B. Side effects for 33 were less severe than for 2 at
equal doses. Side effects were scored as described in the methods.
Reversibility of 2 and 33
The toxicity produced by 2 has previously been suggested to be due to irreversible binding at GlyT2.14
As 33 displayed considerably less side effects, we measured the relative reversibility of 2 and 33 in an
electrophysiology washout assay (Figure 5S). The reversibility of inhibition was investigated by first
applying an IC50 concentration of inhibitor and once a stable level of inhibition was observed,
washing the recording chamber with buffer. Glycine was then reapplied at 5 minute intervals
throughout the wash time course to determine transport recovery, which is indicative of the
reversibility of an inhibitor. For 2, we observed 35% recovery of transport currents after 30 minutes
which suggests 2 is a slowly reversible, but not irreversible, inhibitor. Conversely, transport following
inhibition by 33 was completely restored after 25 minutes.
Conclusions
By conjugating a broad range of amino acids to the oleoyl tail, a large pool of inhibitors of GlyT2
have been identified, most of which have similar or improved potency compared with 4 and other
acyl-glycine compounds. It has been determined that acyl-amino acids are active with the order of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 57
1
2
3
4
5
6
7
8
9
potency: positively charged > aromatic >> aliphatic > negatively charged, and that L- isomers are
more active than D-isomers in most cases. While the most potent compound identified, 32 (IC50 25.5
nM), is ~353-fold more potent than 4, it is limited by its metabolic instability. The D-enantiomer 33,
however, is also a potent GlyT2 inhibitor (IC50 48.3 nM), is metabolically stable, and is rapidly
absorbed into the CNS following ip administration. It has also been demonstrated that ip injection of
33 reduces allodynia in the chronic nerve constriction model of neuropathic pain at concentrations
that have minimal overt side effects. In contrast, the very slowly reversible GlyT2 inhibitor, 2, was
less effective as an analgesic and caused more substantial side effects. At this stage it is not possible
to establish whether short-term analgesic effects of 33 without overt side effects is a consequence of
partial occupancy of the transporter or full occupancy with partial inhibition. However, with high
protein binding it is unlikely that free concentrations of the lipids in vivo are likely to saturate the
binding site on GlyT2 and so this question may become mute. Further investigations into the
availability of free lipids over a longer time frame and their potential for toxic side effects will be
needed to assess the efficacy of this class of compounds for long term treatment of chronic pain.
These novel bioactive lipids provide a promising lead in the development of a new class of analgesics.
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
1. General Chemistry. Ester protected valine, alanine and phenylalanine in L- and D- configurations
were purchased from Sigma Aldrich (Castle Hill, NSW, Australia). All other protected amino acids
were purchased from Fluorochem (Derbyshire, United Kingdom). General reagents and anhydrous
solvents were purchased from Sigma Aldrich and Astatech (Bristol, USA). Reactions were monitored
by thin-layer chromatography (TLC) using silica gel 60 F254 plates. TLC plates were visualised with
UV light and potassium permanganate TLC stain. Reaction products were purified by dry column
vacuum chromatography on silica gel using gradient elutions. 1H and 13C NMR spectra were recorded
1
on an Agilent 500 MHz NMR. Spectra were referenced internally to residual solvent (CDCl3; H 
1
7.26, 13C  77.10. DMSO-d6; H  2.49, 13C  39.52). High resolution mass spectra (HRMS) were
recorded on an Agilent Technologies 6510 Q-TOF LCMS. The purity of all test compounds was
confirmed to be >95% by absolute quantitative NMR (see Supporting Information for further details).
ACS Paragon Plus Environment

Page 25 of 57
Journal of Medicinal Chemistry
1
2
3
4
2. Synthesis
5
6
7
8
9
General procedure A for EDCI coupling. To a solution of oleic acid (2.0 mmol) in anhydrous DMF
(10 mL) was added hydroxybenzotriazole hydrate (2.40 mmol), and EDCI (2.80 mmol). The mixture
was stirred at room temperature for 1 h, then the amino acid or amino acid derivative (1.0 mmol) and
triethylamine (6.0 mmol) were added. The reaction mixture was stirred for 18 h, then diluted with
water (50 mL). The crude product was extrated with ethyl acetate (3 x 25 mL) and purified on silica
gel by stepwise gradient elution with chloroform/isopropanol (100:0 to 90:10).
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
General procedure B for ester hydrolysis. To a solution of the ester (0.50 mmol) in ethanol (30 mL)
was added 1M NaOH (10 mL), and the resulting solution was stirred at 40oC for 3 h. The reaction
volume was reduced ethanol was removed under reduced pressure, and the aqueous residue was
adjusted to pH 2 with 0.5M HCl. The resulting suspension was filtered and the solid product washed
with water (10 mL) and ethanol (5 mL) and dried in vacuo.
General procedure C for BOC removal. The BOC-protected intermediate was dissolved in
dichloromethane (2 mL) and 2M HCl in diethyl ether (2 mL) was added. The solution was stirred at
room temperature for 4 hours and then concentrated in vacuo. The resulting solid was triturated with
diethyl ether and dried in vacuo.
(2S)-2-[[(Z)-octadec-9-enoyl]amino]propanoic acid (6). General procedure B. White solid, yield =
51%. Mp = 48 – 50 oC. 1H NMR (500 MHz, DMSO-d6): δ 8.00 (d, J=8.0Hz, 1H), 5.32-5.30 (m, 2H),
4.16-4.13 (m, 1H) 2.06 (t, J=8.0Hz, 2H), 1.98-1.95 (m, 4H), 1.45 (p, J=7.0Hz, 2H), 1.28-1.23 (m,
23H), 0.84 (t, J=7.0Hz, 3H). 13C NMR (125 MHz, DMSO-d6): δ 174.31, 171.94, 129.63, 129.62,
47.30, 34.98, 31.28, 29.12, 29.10, 28.84(2C), 28.69, 28.61, 28.59, 28.57, 26.61, 26.57, 25.18, 22.09,
17.21, 13.93. [α]23D -11.1 (c = 2.1, MeOH). HRMS (ESI) m/z [M+H]+ calcd for C21H39NO3, 354.3003;
found, 354.3009.
Methyl (2S)-2-[[(Z)-octadec-9-enoyl]amino]propanoate (6a). General procedure A. White solid,
o
1
yield = 70%. Mp = 28 – 30 C. H NMR (500 MHz, DMSO-d6): δ 8.17 (d, J=7.0Hz, 1H), 5.32-5.30
(m, 2H), 4.24-4.21 (m, 1H) 3.59 (s, 3H), 2.07 (t, J=7.5Hz, 2H), 1.99-1.96 (m, 4H), 1.45 (p, J=7.5Hz,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 57
1
2
3
4
13
2H), 1.30-1.23 (m, 23H), 0.84 (t, J=7.0Hz, 3H). C NMR (125 MHz, DMSO-d6): δ 172.23, 172.08,
5
6
7
8
129.61(2C), 51.70, 47.39, 34.86, 31.27, 29.11, 29.09, 28.82, 28.68, 28.66, 28.58, 28.55, 28.54, 26.60,
26.56, 25.13, 22.08, 16.94, 13.93. HRMS (ESI) m/z [M+H]+ calcd for C22H41NO3, 368.3159; found,
368.3154.
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
(2R)-2-[[(Z)-octadec-9-enyl] amino]propanoic acid (7). General procedure B. White solid, yield =
51%. Mp = 44 – 46 oC. 1H NMR (500 MHz, DMSO-d6): δ 8.02 (d, J=7.5Hz, 1H), 5.32-5.30 (m, 2H),
4.18-4.15 (m, 1H) 2.06 (t, J=7.5Hz, 2H), 1.99-1.95 (m, 4H), 1.45 (p, J=7.5Hz, 2H), 1.28-1.23 (m,
23H), 0.84 (t, J=7.0Hz, 3H). 13C NMR (125 MHz, DMSO-d6): δ 174.28, 171.93, 129.61, 129.60,
47.30, 34.98, 31.28, 29.13, 29.11, 28.84(2C), 28.69, 28.61, 28.59, 28.57, 26.61, 26.56, 25.18, 22.09,
17.21, 13.93. [α]23 +14.3 (c = 2.2, MeOH). HRMS (ESI) m/z [M+H]+ calcd for C21H39NO3,
D
354.3003; found, 354.3003.
Methyl (2S)-2-[[(Z)-octadec-9-enoyl]amino]propanoate (7a). General procedure A. White solid,
o
1
yield = 70%. Mp = 26 – 28 C. H NMR (500 MHz, DMSO-d6): δ 8.11 (d, J=6.5Hz, 1H), 5.32-5.30
(m, 2H), 4.21-4.17 (m, 1H), 3.60 (s, 3H), 2.07 (t, J=7.5Hz, 2H), 1.99-1.96 (m, 4H), 1.45 (p, J=7.0Hz,
2H), 1.30-1.23 (m, 23H), 0.84 (t, J=7.0Hz, 3H). 13C NMR (125 MHz, DMSO-d6): δ 174.26, 172.10,
129.63(2C), 51.71, 47.40, 34.86, 31.27, 29.11, 29.10, 28.83, 28.68, 28.66, 28.58, 28.55(2C), 26.60,
26.56, 25.13, 22.08, 16.94, 13.93. HRMS (ESI) m/z [M+H]+ calcd for C22H41NO3, 368.3159; found,
368.3154.
(2S)-3-methyl-2-[[(Z)-octadec-9-enoyl]amino]butanoic acid (8). General procedure B. White solid,
o
1
yield = 33%. Mp = 58 – 60 C. H NMR (500 MHz DMSO-d6): δ 7.85 (d, J=8.5Hz, 1H), 5.32-5.30
(m, 2H), 4.50-4.47 (m, 1H) 2.18-2.09 (m 2H), 2.04-1.95 (m, 5H), 1.48-1.44 (m, 2H), 1.28-1.23 (m,
13
20H), 0.84 (t, J=7.0Hz, 9H). C NMR (125 MHz, DMSO-d6): δ 172.25, 172.36, 129.59(2C), 57.07,
34.98, 31.28, 29.78, 30.66, 29.83, 29.10, 28.82(2C), 28.69(2C), 28.65, 28.59, 26.61, 26.56, 25.37,
22.08, 19.17, 18.02, 13.92. [α]23 -2.0 (c = 2.1, MeOH). HRMS (ESI) m/z [M+H]+ calcd for
D
C23H43NO3, 382.3316; found, 382.3319.
ACS Paragon Plus Environment

Page 27 of 57
Journal of Medicinal Chemistry
1
2
3
4
Methyl (2S)-3-methyl-2-[[(Z)-octadec-9-enoyl]amino]butanoate (8a). General procedure A.
5
6
1
Colourless liquid, yield = 53%. H NMR (500 MHz, DMSO-d6): δ 8.03 (d, J=8.0Hz, 1H), 5.32-5.30
7
8
9
(m, 2H), 4.17-4.14 (m, 1H) 3.61 (s, 3H), 2.18-2.08 (m 2H), 2.02-1.95 (m, 5H), 1.48-1.44 (m, 2H)
1.28-1.23 (m, 20H), 0.87 (dt, J=7.0, 9.50 Hz, 9H). 13C NMR (125 MHz, DMSO-d6): δ 172.56, 172.23,
129.59(2C), 57.25, 51.51, 34.82, 31.28, 29.78, 29.10, 28.83, 28.68(2C), 28.62(2C), 28.59, 28.57,
28.55, 26.60, 26.56, 25.69, 22.08, 18.97, 18.16, 13.93. HRMS (ESI) m/z [M+H]+ calcd for
C24H45NO3, 396.3472; found, 396.3469.
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
(2R)-3-methyl-2-[[(Z)-octadec-9-enoyl]amino]butanoic acid (9). General procedure B. White solid,
o
1
yield = 56%. Mp = 61 – 63 C. H NMR (500 MHz, DMSO-d6): δ 7.85 (d, J=9.0Hz, 1H), 5.32-5.30
(m, 2H), 4.13-4.10 (m, 1H) 2.16-2.10 (m 2H), 2.00-1.95 (m, 5H), 1.48-1.44 (m, 2H) 1.28-1.23 (m,
13
20H), 0.84 (t, J=7.0Hz, 9H). C NMR (125 MHz, DMSO-d6): δ 172.96, 172.14, 129.38(2C), 56.80,
34.72, 31.28, 30.43, 29.53, 28.84, 28.83(2C), 28.56, 28.42, 28.38, 28.32, 28.31, 26.34, 26.30, 25.11,
21.83, 18.93, 17.79, 13.92. [α]23 +1.8 (c = 2.1, MeOH). HRMS (ESI) m/z [M+H]+ calcd for
D
C23H43NO3, 382.3316; found, 382.3318.
Methyl (2R)-3-methyl-2-[[(Z)-octadec-9-enoyl]amino]butanoate (9a). General procedure A.
1
Colourless liquid, yield = 65%. H NMR (500 MHz, DMSO-d6): δ 8.03 (d, J=8.0Hz, 1H), 5.32-5.30
(m, 2H), 4.16-4.13 (m, 1H) 3.60 (s, 3H), 2.17-2.07 (m 2H), 2.01-1.95 (m, 5H), 1.47- 1.44 (m, 2H)
1.28-1.23 (m, 20H), 0.87 (dt, J=7.0, 9.50 Hz, 9H). 13C NMR (125 MHz, DMSO-d6): δ 172.57, 172.25,
129.60(2C), 57.26, 51.51, 34.82, 31.28, 29.78, 29.10, 28.83, 28.68(2C), 28.62(2C), 28.59, 28.57,
28.55, 26.60, 26.56, 25.69, 22.09, 18.97, 18.16, 13.93. HRMS (ESI) m/z [M+H]+ calcd for
C24H45NO3, 396.3472; found, 396.3467.
(2S)-4-methyl-2-[[(Z)-octadec-9-enoyl]amino]pentanoic acid (10). General procedure B. White
solid, yield = 60%. Mp = 66 – 68 oC. 1H NMR (500 MHz, DMSO-d6): δ 7.98 (d, J=8.0Hz, 1H), 5.32-
5.30 (m, 2H), 4.20-4.18 (m, 1H), 2.09-2.06 (m, 2H), 1.99-1.97 (m, 4H), 1.62-1.58 (m, 1H), 1.50-1.43
(m, 4H), 1.30-1.17 (m, 20H), 0.83 (t, J=7.0Hz, 9H). 13C NMR (125 MHz, DMSO-d6): δ 174.32,
172.19, 129.62(2C), 50.00, 35.02, 31.26, 29.08, 28.81, 28.67, 28.62, 28.59(3C), 28.57, 28.49, 26.62,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 57
1
2
3
4
26.57, 25.25, 24.32, 22.89, 22.08, 21.17, 13.94. [α]23 -10.9 (c = 2.1, MeOH). HRMS (ESI) m/z
D
5
6
[M+H]+ calcd for C24H45NO3, 396.3472; found, 396.3472.
7
8
9
Methyl (2S)-4-methyl-2-[[(Z)-octadec-9-enoyl]amino]pentanoate (10a). General procedure 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
1
Colourless liquid, yield = 86%. H NMR (500 MHz, CDCl3): δ 5.80 (d, J=8.5Hz, 1H), 5.34-5.32 (m,
2H), 4.66-4.64 (m, 1H), 3.73 (s, 3H), 2.20 (t, J=7.5Hz, 2H), 2.02-1.99 (m, 4H), 1.66-1.62 (m, 5H),
1.53-1.50 (m, 1H), 1.30-1.26 (m, 20H), 0.95-0.93 (m, 6H), 0.87 (t, J=7.0Hz, 3H). 13C NMR (125
MHz, CDCL3): δ 173.12, 172.99, 130.13, 129.89, 52.19, 50.65, 41.95, 36.73, 32.04, 29.91, 29.84,
29.67, 29.46, 29.45, 29.37, 29.33, 29.28, 27.36, 27.31, 25.70, 25.03, 22.94, 22.82, 22.13,14.25.
HRMS (ESI) m/z [M+H]+ calcd for C25H47NO3, 410.3629; found, 410.3628.
(2R)-4-methyl-2-[[(Z)-octadec-9-enoyl]amino]pentanoic acid (11). General procedure B. White
solid, yield = 70%. Mp = 64 – 66 oC. 1H NMR (500 MHz, DMSO-d6): δ 7.98 (d, J=7.5Hz, 1H), 5.32-
5.30 (m, 2H), 4.21-4.17 (m, 1H), 2.10-2.06 (m, 2H), 1.98-1.96 (m, 4H), 1.63-1.58 (m, 1H), 1.52-1.42
(m, 4H), 1.27-1.20 (m, 20H), 0.84 (t, J=7.0Hz, 9H). 13C NMR (125 MHz, DMSO-d6): δ 174.31,
172.18, 129.61(2C), 49.99, 35.01, 31.27, 29.09, 28.82, 28.67, 28.63, 28.60(3C), 28.58, 28.50, 26.62,
26.56, 25.25, 24.31, 22.87, 22.08, 21.15, 13.93. [α]23 +11.5 (c = 2.0, MeOH). HRMS (ESI) m/z
D
[M+H]+ calcd for C24H45NO3, 396.3472; found, 396.3470.
Methyl (2R)-4-methyl-2-[[(Z)-octadec-9-enoyl]amino]pentanoate (11a). General procedure A.
1
Colourless liquid, yield = 80%. H NMR (500 MHz, CDCl3): δ 5.79 (d, J=8.5Hz, 1H), 5.35-5.32 (m,
2H), 4.66-4.64 (m, 1H), 3.73 (s, 3H), 2.20 (t, J=7.5Hz, 2H), 2.03-1.99 (m, 4H), 1.66-1.62 (m, 5H),
1.53-1.50 (m, 1H), 1.30-1.26(m, 20H), 0.95-0.93(m, 6H), 0.87 (t, J=7.0Hz, 3H). 13C NMR (125 MHz,
CDCl3): δ 173.12, 172.99, 130.13, 129.89, 52.19, 50.65, 41.95, 36.73, 32.04, 29.91, 29.84, 29.67,
29.46, 29.45, 29.37, 29.33, 29.28, 27.36, 27.31, 25.70, 25.03, 22.94, 22.82, 22.13,14.25. HRMS (ESI)
m/z [M+H]+ calcd for C25H47NO3, 410.3629; found, 410.3627.
(2S)-4-methylsulfanyl-2-[[(Z)-octadec-9-enoyl]amino]butanoic acid (12). General procedure B.
White solid, yield = 59%. Mp = 28 – 30 oC. 1H NMR (500 MHz, DMSO-d6): δ 8.04 (d, J=7.5Hz, 1H),
5.32-5.30 (m, 2H), 4.29-4.26 (m, 1H), 2.47-2.40 (m, 2H), 2.11-2.07 (m, 2H), 2.02 (s, 3H), 1.99-1.97
ACS Paragon Plus Environment

Page 29 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(m, 4H), 1.93-1.90 (m, 1H), 1.83-1.80 (m, 1H), 1.46 (p, J=7.0Hz, 2H), 1.31-1.23 (m, 20H), 0.84 (t,
J=7.0Hz, 3H). 13C NMR (125 MHz, DMSO-d6): δ 173.48, 172.34, 129.60(2C), 50.75, 35.04, 31.28,
30.70, 29.74, 29.12, 29.10(2C), 28.84, 28.68, 28.67, 28.59, 28.56, 26.62, 26.57, 25.22, 22.09, 14.57,
13.92. [α]23D -7.4 (c = 2.1, MeOH). HRMS (ESI) m/z [M+H]+ calcd for C23H43NO3, 414.3036; found,
414.3038.
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
Methyl (2S)-4-methylsulfanyl-2-[[(Z)-octadec-9-enoyl] amino] butanoate (12a). General
1
procedure A. Colourless liquid, yield = 89%. H NMR (500 MHz, CDCl3): δ 6.14 (d, J=7.5Hz, 1H),
5.34-5.32 (m, 2H), 4.74-4.70 (m, 1H), 3.75 (s, 3H), 2.52-2.48 (m, 2H), 2.22 (t, J=7.5Hz, 2H). 2.19-
2.14 (m, 1H), 2.08 (s, 3H), 1.99-1.94 (m, 5H), 1.63 (p, J=7.0Hz, 2H), 1.32-1.24 (m, 20H), 0.87 (t,
J=7.0Hz, 3H). 13C NMR (125 MHz, CDCl3): δ 173.06, 172.77, 130.12, 129.86, 52.64, 51.53, 36.72,
32.02, 31.90, 30.11, 29.89, 29.83, 29.64, 29.44, 29.43, 29.36, 29.34, 29.26, 27.37, 27.30, 25.70, 22.80,
15.63, 14.24. HRMS (ESI) m/z [M+H]+ calcd for C24H45O3S, 428.3193; found, 428.3198.
(2R)-4-methylsulfanyl-2-[[(Z)-octadec-9-enoyl]amino]butanoic acid (13). General procedure B.
White solid, yield = 98%. Mp = 30 – 32 oC. 1H NMR (500 MHz, DMSO-d6): δ 8.04 (d, J=7.5Hz, 1H),
5.32-5.30 (m, 2H), 4.29-4.26 (m, 1H), 2.47-2.42 (m, 2H), 2.11-2.07 (m, 2H), 2.02 (s, 3H), 1.99-1.97
(m, 4H), 1.93-1.90 (m, 1H), 1.83-1.80 (m, 1H), 1.46 (p, J=7.0Hz, 2H), 1.31-1.23 (m, 20H), 0.84 (t,
J=7.0Hz, 3H). 13C NMR (125 MHz, DMSO-d6): δ 173.48, 172.34, 129.60(2C), 50.75, 35.04, 31.28,
30.70, 29.74, 29.12, 29.10(2C), 28.84, 28.68, 28.67, 28.59, 28.56, 26.62, 26.57, 25.22, 22.09, 14.57,
13.92. [α]23D +7.2 (c = 2.0, MeOH). HRMS (ESI) m/z [M+H]+ calcd for C23H43NO3, 414.3036; found,
414.3039.
Methyl (2R)-4-methylsulfanyl-2-[[(Z)-octadec-9-enoyl] amino] butanoate (13a). General
1
procedure A. Colourless liquid, yield = 56%. H NMR (500 MHz, CDCl3): δ 6.12 (d, J=7.5Hz, 1H),
5.34-5.32 (m, 2H), 4.74-4.70 (m, 1H), 3.75 (s, 3H), 2.52-2.48 (m, 2H), 2.22 (t, J=7.5Hz, 2H), 2.19-
2.14 (m, 1H), 2.08 (s, 3H), 1.99-1.94 (m, 5H), 1.63 (p, J=7.0Hz, 2H), 1.32-1.24 (m, 20H), 0.87 (t,
J=7.0Hz, 3H). 13C NMR (125 MHz, CDCl3): δ 173.06, 172.77, 130.12, 129.86, 52.64, 51.53, 36.72,
ACS Paragon Plus Environment