Identification of Diketopiperazine-Containing 2-Anilinobenzamides as Potent Sirtuin 2 (SIRT2)-Selective Inhibitors Targeting the "selectivity Pocket", Substrate-Binding Site, and NAD+-Binding Site

Article abstract of DOI:10.1021/acs.jmedchem.9b00255

The NAD⁺-dependent deacetylase SIRT2 represents an attractive target for drug development. Here, we designed and synthesized drug-like SIRT2-selective inhibitors based on an analysis of the putative binding modes of recently reported SIRT2-selective inhibitors and evaluated their SIRT2-inhibitory activity. This led us to develop a more drug-like diketopiperazine structure as a "hydrogen bond (H-bond) hunter" to target the substrate-binding site of SIRT2. Thioamide 53, a conjugate of diketopiperazine and 2-anilinobenzamide which is expected to occupy the "selectivity pocket" of SIRT2, exhibited potent SIRT2-selective inhibition. Inhibition of SIRT2 by 53 was mediated by the formation of a 53-ADP-ribose conjugate, suggesting that 53 is a mechanism-based inhibitor targeting the "selectivity pocket", substrate-binding site, and NAD⁺-binding site. Furthermore, 53 showed potent antiproliferative activity toward breast cancer cells and promoted neurite outgrowth of Neuro-2a cells. These findings should pave the way for the discovery of novel therapeutic agents for cancer and neurological disorders.

Full text of DOI:10.1021/acs.jmedchem.9b00255

Subscriber access provided by BOSTON UNIV
Article
Identification of Diketopiperazine-Containing 2-Anilinobenzamides
as Potent Sirtuin 2 (SIRT2)-Selective Inhibitors Targeting the
+
“Selectivity Pocket”, Substrate-Binding Site, and NAD-Binding Site
Paolo Mellini, Yukihiro Itoh, Elghareeb E. Elboray, Hiroki Tsumoto, Ying Li, Miki
Suzuki, Yukari Takahashi, Toshifumi Tojo, Takashi Kurohara, Yuka Miyake,
Yuri Miura, Yuki Kitao, Masayuki Kotoku, Tetsuya Iida, and Takayoshi Suzuki
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00255 • Publication Date (Web): 30 May 2019
Downloaded from http://pubs.acs.org on May 30, 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 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Identification of Diketopiperazine-Containing 2-
Anilinobenzamides as Potent Sirtuin 2 (SIRT2)-
Selective Inhibitors Targeting the “Selectivity
Pocket”, Substrate-Binding Site, and NAD+-
Binding 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
Paolo Mellini,† Yukihiro Itoh,† Elghareeb E. Elboray,†,‡ Hiroki Tsumoto,# Ying Li,† Miki
Suzuki,† Yukari Takahashi,† Toshifumi Tojo,† Takashi Kurohara,† Yuka Miyake,† Yuri Miura,#
Yuki Kitao,† Masayuki Kotoku,†Tetsuya Iida,† and Takayoshi Suzuki*,†,§
†Graduate School of Medical Science, Kyoto Prefectural University of Medicine, 1-5
Shimogamohangi-cho, Sakyo-ku, Kyoto, 606-0823, Japan
‡Chemistry Department, Faculty of Science, South Valley University, Qena, 83523, Egypt
#Research Team for Mechanism of Aging, Tokyo Metropolitan Institute of Gerontology, 35-2
Sakae-cho, Itabashi-ku, Tokyo 173-0015, Japan
§CREST, Japan Science and Technology Agency (JST), 4-1-8 Honcho Kawaguchi, Saitama
332-0012, Japan
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 69
1
2
3
4
ABSTRACT
5
6
7
8
9
The NAD+-dependent deacetylase SIRT2 represents an attractive target for drug
development. Here, we designed and synthesized drug-like SIRT2-selective inhibitors based
on an analysis of the putative binding modes of recently reported SIRT2-selective inhibitors,
and evaluated their SIRT2-inhibitory activity. This led us to develop a more drug-like
diketopiperazine structure as a “hydrogen bond (H-bond) hunter” to target the substrate-
binding site of SIRT2. Thioamide 53, a conjugate of diketopiperazine and 2-
anilinobenzamide which is expected to occupy the “selectivity pocket” of SIRT2, exhibited
potent SIRT2-selective inhibition. Inhibition of SIRT2 by 53 was mediated by the formation
of a 53-ADP-ribose conjugate, suggesting that 53 is a mechanism-based inhibitor targeting
the “selectivity pocket”, substrate-binding site, and NAD+-binding site. Furthermore, 53
showed potent antiproliferative activity towards breast cancer cells and promoted neurite
outgrowth of Neuro-2a cells. These findings should pave the way for the discovery of novel
therapeutic agents for cancer and neurological disorders.
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
2
ACS Paragon Plus Environment

Page 3 of 69
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
Since the discovery of Sir2 in S. cerevisae in 1979 by Klar and co-workers,1 the biological
function of sirtuins (SIRT1–7)2 has been widely investigated. Sirtuins were originally
identified as NAD+-dependent lysine deacetylases of histone and non-histone proteins,3,4 but
recent studies have revealed that their catalytic activity is not limited to deacetylation.
Sirtuins also efficiently catalyze the hydrolysis of N-acyl lysines such as long-chain fatty
acyl (SIRT1–3, 6, and 7),5–7 4-oxononanonyl (SIRT2),8 succinyl (SIRT5), and benzoyl
lysines (SIRT2).9–11 These findings, together with the different subcellular
compartmentalizations of the isotypes,12 highlight that sirtuins have key roles in regulating
multiple biological processes. To date, the activities of SIRT1, 3, and 6 have been linked to
metabolic functions such as mitochondrial biogenesis, fatty acid oxidation, cholesterol efflux,
insulin secretion, and glucose homeostasis.13,14 SIRT1 inhibits lipid accumulation through
inhibition of PPARγ activity15 and participates in the control of glucose metabolism by
deacetylation of PGC1-α16,17 and repression of UCP218. SIRT3 is highly expressed in
metabolically active tissues such as skeletal muscle, brain, liver, and heart.19,20 It is involved
in energy production and exerts antioxidant effects via deacetylation of AceCS2,21,22
SOD2,23,24 and IDH2.25 SIRT6 regulates glucose homeostasis via GLUT1 inhibition and
promotes expression of glycolytic genes, functioning as a HIF1-α co-repressor.26–28 Among
sirtuins, SIRT2 is most highly expressed in brain.29,30 Its genetic or pharmacological
inhibition is associated with neuroprotection in models of both Parkinson’s and Huntington
diseases;31–36 especially, treatment of the models of Parkinson's and Huntington’s diseases
with small-molecular SIRT2 inhibitors was found to be effective in reducing α-synuclein-
mediated toxicity and polyglutamine inclusions.33–36 Furthermore, a potential link between
inhibition of SIRT2 activity and antidepressant-like action has recently been revealed by
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
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 69
1
2
3
4
5
6
Tordera and co-workers.37 Thus, SIRT2 is interesting as a therapeutic target for neurological
disorders.
7
8
9
The role of SIRT1 and 2 in cancer progression is controversial.38–42 In normal cells,
SIRT1 promotes genomic stability. In some cancer cells, SIRT1 has been reported to promote
cell survival through deacetylation of p53, followed by silencing of tumor suppressor
HIC1.43,44 In addition, SIRT1-suppression by miR34a induces apoptosis of cancer cells.45 In
contrast, SIRT1 acts as a tumor suppressor via HIF1-α,46 leading to inhibition of tumor
growth and vascular formation. Accordingly, it is still unclear whether activation or inhibition
of SIRT1 is beneficial. The functions of SIRT2 are thought to be highly dependent on cancer
type. For example, a study on SIRT2-deficient mice suggests that SIRT2 acts as a tumor
suppressor in liver.47 On the other hand, other studies on patient samples, biochemical
pathway analysis, and small molecular SIRT2 inhibitors suggest that SIRT2 acts as a tumor
promoter in acute myeloid leukemia,48,49 prostate cancer,50 and neuroblastoma.51 Accordingly,
pharmacological inhibition of SIRT2 is effective on certain specific cancers.
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
Many SIRT inhibitors have been reported.52–54 SIRT1 and 2 inhibitors such as salermide,55
tenovin-6,56,57 and AC-9352358 (Chart1) showed interesting anticancer properties when tested
against a panel of cancer cells, including cancer stem cells, but their low inhibitory potency
combined with non-optimal isotype selectivity made it difficult to predict their biological
activity.
Currently,
chroman-4-ones,59,60
NCO-90,61,62
1,2,4-oxadiazoles,63
thienopyrimidinone,64 RK-9123016,65 NPD11033,66 SirReals,67–72 and TM73 have been
reported as SIRT2-selective inhibitors (Chart 1). Among them, SirReals and TM have been
studied well. SirReal2 inhibits the SIRT2-downstream target PEPCK1, which is associated
with mitochondrial metabolism and the RAS/ERK/JNK/MMP-9 pathway, and decreases
migration as well as invasion of human gastric cancer cells.74 TM shows not only
4
ACS Paragon Plus Environment

Page 5 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
antiproliferative activity towards a large panel of cancer cells including leukemia, non-small
cell lung cancer, colon, prostate, and breast cancer cells, but also the ability to discriminate
between normal breast cells and cancer cells: the anticancer effects are strictly dependent on
SIRT2 inhibition, which leads to time-dependent c-Myc degradation.73 These findings have
put the spotlight back on SIRT2-selective inhibitors as valuable tools to investigate the
druggability of SIRT2 in cancers.
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
We recently developed low-micromolar active, SIRT2-selective inhibitor KPM-175 (Chart
1 and Supporting Information, Figure S1a) by molecular fusion of NCO-90 and a
pseudopeptidic substrate-mimetic compound. The substitution of the KPM-1 acyl oxygen
with sulfur led to KPM-275 (Chart 1), a nanomolar SIRT2 mechanism-based inhibitor
(Supporting Information Figure S1b) with antiproliferative activity towards breast cancer
cells and potent neurite outgrowth activity. In this study, aiming to improve the non drug-like
peptide structure of KPM-1 and KPM-2, we explored in detail targeting of the SIRT2
“selectivity pocket”, substrate-binding site, and NAD+-binding site by means of structure-
activity relationship studies of NCO-90 bearing a 2-anilinobenzamide moiety. Herein, we
report the rational design, synthesis, and biological activity of SIRT2-selective inhibitors
based on the 2-anilinobenzamide scaffold.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 69
1
2
3
4
5
6
7
8
H
N
S
N
H
O
O
N
O
N
O
S
O
Cl
N
N
N
N
H
H
N
OH
Br
9
salermide
NH2
tenovin-6
chroman-4-one
AC-93253
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
O2N
Br
O
H
O
N
N
HN
N
N
N
O
NH
N
N
O
N
O
O
S
N
N
H
S
Br
O
NCO-90
1,2,4-oxadiazoles
thienopyrimidinone
RK-9123016
O
NH2
OH
HN
H
S
NH
N
R
HN
O
N
O
N
N
X
HN
O
N
NH
S
N
O
S
O
O
H
NH
N
O
O
N
H
O
O
SirReal1 (R = Ph)
SirReal2 (R = 1-naphtyl)
NPD11033
O
TM
KPM-1 (X = O)
KPM-2 (X = S)
Chart 1. Representative SIRT1 and SIRT2 inhibitors.
RESULTS AND DISCUSSION
Chemistry. Compounds 2, 3, 8, 12, 18–43, and 53 were synthesized as shown in Schemes
1–4. The preparation of 2, 3, and 8 is illustrated in Scheme 1. Alcohol 161 was etherified with
2-(2-bromoethyl)naphthalene76 or esterified with 2-[(4,6-dimethylpyrimidin-2-yl)thio]acetic
acid77 in the presence of coupling reagent 1-[(1-(cyano-2-ethoxy-2-oxoethylideneaminooxy)
dimethylaminomorpholino)]uronium hexafluorophosphate (COMU)78 to obtain 2 or 3,
respectively. Compound 8 was synthesized by Buchwald-Hartwig coupling between 4 and 5,
followed by reduction of the nitro group and amidation with 2-[(4,6-dimethylpyrimidin-2-
yl)thio]acetic acid.
6
ACS Paragon Plus Environment

Page 7 of 69
Journal of Medicinal Chemistry
1
2
3
4
Scheme 1. Synthesis of compounds 2, 3, and 8.a
5
6
7
8
9
O
O
R1
compd
NH2
NH
NH2
a or b
NH
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
O
O
OH
R1
O
O
1
2, 3, 8
S
N
N
3
8
O
N
N
e
NH2
NH2
O
O
NH2
NH2
S
N
H
4
c
d
NH
NH
+
Br
NO2
NH2
6
7
NO2
5
aReagents and conditions: (a) 2-(2-bromoethyl)naphthalene, K2CO3, acetone, reflux 18 h (for
2); (b) 2-[(4,6-dimethylpyrimidin-2-yl)thio]acetic acid, COMU®, Et3N, DMF, 0°C to rt, 18 h
(for 3); (c) Pd2dba3, 2-dicyclohexylphosphino-2’,4’,6’-triisopropyl biphenyl (XPhos), K2CO3,
t-BuOH, reflux, 18–20 h; (d) Pd/C, H2, MeOH, 2 h; (e) 2-[(4,6-dimethylpyrimidin-2-
yl)thio]acetic acid, EDCI·HCl, HOBt, Et3N, DMF, 0°C to rt, 17 h.
Scheme 2 shows the synthesis of 12. Condensation of acid 9 with ammonia afforded
amide 10. Reduction of nitro group of 10 in the presence of Pd/C catalyst yielded amine 11.
Compound 11 was converted to 12 by Buchwald-Hartwig coupling with 1-bromo-3-
phenethoxybenzene.61
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 69
1
2
3
4
Scheme 2. Synthesis of compound 12.a
NH2
5
6
7
8
O
OH
NH2
NH2
NH
a
b
c
O
O
O
NO2
NO2
NH2
9
10
11
9
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
12
aReagents and conditions: (a) (COCl)2, DMF, CH2Cl2, 0°C to rt, 180 min, then NH3, aqueous
solution, THF; (b) Pd/C, H2, MeOH, 5 h; (c) 1-bromo-3-phenethoxybenzene, Pd2dba3,
XPhos, K2CO3, t-BuOH, reflux, 8 h.
Scheme 3 shows the preparation of 19–43. Amine 17, which was used for the preparation
of 43, was synthesized from 13 in four steps: coupling reaction with glycine ethyl ester, Boc
deprotection under acidic conditions, diketopiperazine formation, and Cbz deprotection in the
presence of Pd/C catalyst. Compounds 19–31 and 33–39 were synthesized by coupling
reactions using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDCI·HCl)
and 1-hydroxybenzotriazole (HOBt) from the corresponding amines and acid 18, which was
prepared as described previously.61,75 Hydroxamic acid 32 was prepared by hydrolysis of the
tetrahydropyranyl acetal group of 29. Hydrolysis of ester 39 afforded acid 40, which was
reacted with appropriate amines to afford 41–43.
8
ACS Paragon Plus Environment

Page 9 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
Scheme 3. Synthesis of compounds 19–43.a
O
O
O
O
O
O
O
a
b
OEt
OEt
c
d
CbzHN
OH
NHBoc
CbzHN
N
CbzHN
N
CbzHN
NH
H2N
NH
4
4
4
4
4
NH2
H
H
NHBoc
O
O
HN
HN
7
13
14
15
16
17
8
9
O
O
O
O
OH
O
OH
NH
R2
NH
N
H
R3
N
H
NH
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
e
f
e
NH
O
O
O
O
O
18
19-39
40
39
(from
)
41-43
R2
R3
compd
R2
compd
R2
compd
compd
O
OCH3
H
N
N
H
19
26
N
H
33
N
H
41
N
H
O
O
N
H
N
H
N
20
21
22
23
24
25
27
28
29
30
31
32
N
H
34
35
36
37
38
39
42
43
O
N
H
N
H
N
O
O
H
N
OH
N
H
N
H
N
H
NH
HN
H
N
O
O
O
O
N
H
N
H
N
H
O
N
H
N
N
N
H
N
H
O
NH2
N
H
N
H
O
O
O
OCH3
OH
N
H
N
H
N
H
aReagents and conditions: (a) glycine ethyl ester hydrochloride, COMU®, Et3N, DMF, 0°C to
rt, 24 h; (b) CF3COOH, CH2Cl2, rt, overnight; (c) CH3COOH, N-methylmorpholine, 1-
butanol, reflux, 220 min; (d) H2, Pd/C, MeOH, rt, 3 h; (e) corresponding amine, EDCI·HCl,
HOBt, Et3N, CH2Cl2, 0°C to rt, 15–17 h; (f) LiOH, THF, MeOH, H2O, rt, 150 min.
Compound 53 was prepared from 13 (Scheme 4). Methyl esterification and deprotection
of the Cbz group of 13 afforded amine 45. Coupling reaction between 45 and N-Fmoc glycine
gave 46, and reaction using Lawesson’s reagent79 efficiently led to 47. Then, hydrolysis of 47
in the presence of trimethyltin hydroxide80 and condensation with glycine ethyl ester afforded
49. Compound 49 was treated with trifluoroacetic acid to furnish 50. Then, diketopiperazine
formation and deprotection of Fmoc group using piperidine were conducted to obtain 52.
Finally, condensation between 18 and 52 furnished the desired compound 53.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 69
1
2
3
4
5
6
7
8
Scheme 4. Synthesis of compound 53.a
O
O
O
X
O
b
a
c
FmocHN
CbzHN
OMe
NHBoc
44
CbzHN
OH
NHBoc
13
H2N
OMe
NHBoc
45
4
N
OMe
4
4
4
H
NHBoc
46
47
: X = O
: X = S
d
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
e
S
O
S
O
S
O
g
f
FmocHN
OEt
FmocHN
OEt
FmocHN
N
H
N
H
N
N
4
4
NHBoc
N
H
OH
NHBoc
H
H
4
NH2
O
O
50
49
48
h
S
O
O
S
O
O
FmocHN
H2N
i
N
NH
N
NH
4
HN
4
HN
H
H
O
O
O
H
N
51
52
N
NH
H
S
HN
j
NH
+
O
OH
NH
O
53
O
18
aReagents and conditions: (a) MeI, K2CO3, acetone, reflux, 26 h; (b) Pd/C, H2, MeOH, rt, 3 h;
(c) N-Fmoc-glycine, COMU®, i-Pr2NEt, DMF, CH2Cl2, 0°C to rt, 4 h; (d) Lawesson’s
reagent, toluene, 60C, 3 h; (e) Me3SnOH, ClCH2CH2Cl, 60C, 10 h; (f) glycine ethyl ester
hydrochloride, COMU®, i-Pr2NEt, DMF, CH2Cl2, 0°C to rt, 4 h; (g) CF3COOH, CH2Cl2, 0°C
to rt, 2 h; (h) CH3COOH, NMM, 1-butanol, reflux, 220 min; (i) piperidine, CH2Cl2, rt, 30
min; (j) COMU®, i-Pr2NEt, DMF, CH2Cl2, 0°C to rt, 3 h.
Design of NCO-90 derivatives and in vitro evaluation of inhibitory activity towards
SIRT1 and SIRT2. We recently solved the X-ray crystal structure of SIRT2 complexed with
NCO-90, which revealed that the inhibitor binds to the “selectivity pocket”67 where the
phenoxyethylphenyl moiety of NCO-90 establishes π-π and CH-π interactions, and the 2-
anilinobenzamide core protrudes towards the acetyl-lysine substrate-binding site (Figure
1a,b).75 Based on this X-ray structure and another structure of SIRT2 complexed with a
10
ACS Paragon Plus Environment

Page 11 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pseudopeptide substrate (Figure 1c,d), we previously identified KPM-1 and KPM-2 as
SIRT2-selective inhibitors, which were designed by conjugating NCO-90 and a
pseudopeptide substrate structure (Figure 1e and Supporting Information Figure S1a,b).75,81
We also found that introduction of small glycinamide handles into NCO-90 did not affect the
SIRT2-inhibitory activity. We assume that glycinamide dislodges a water molecule (HOH4)
and forms an H-bonding interaction with Val233 (Figure 1b,d,e) and that potent SIRT2
inhibition was achieved by introduction of the pseudopeptide scaffold, which mimics a lysine
residue substrate. We considered that the peptide bonds could interact with Gly236, Glu237,
and Gln267 through H-bond formation (Figure 1d,e). In order to provide additional insights
into drug-like fragments targeting the binding site of acetyl-lysine substrate, we prepared and
screened a new small library of NCO-90 analogues, using SIRT1 and SIRT2 inhibition
assays. The naphthyl (2) and pyrimidine derivatives (3 and 8) were designed to probe the
versatility of NCO-90 as a SIRT2 inhibitor scaffold (Figure 1f, part A). The acetamide (12),
carboxylic acid (18) and amides with small/bulky hydrocarbon groups (19–27) were tested to
elucidate the structure-activity relationship of the benzamide moiety, and amides with polar
groups (28–40) were expected to interact directly with Val233 (Figure 1f, part B).
Furthermore, compounds 41–43 were prepared as “H-bonds hunters” to target Gly236,
Glu237, and Gln267 in SIRT2 (Figure 1f, part C). All the compounds prepared in this study
were screened in in vitro assays using SIRT1 and SIRT2, and NCO-90 was used as a
reference compound (Table 1). The SIRT1-inhibitory activities of the tested compounds were
relatively low even at 50 M (Table 1). Therefore, we focus mainly on the SIRT2-inhibitory
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
activities at 10 M in the following section.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 69
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. (a) X-ray structure of SIRT2 in the complex with NCO-90 (black) (PDB code:
5Y5N). (b) Schematic diagram of Figure 1a. (c) X-ray structure of SIRT2 in the complex
with myristoylated substrate peptide (black, purple, and green) (PDB code: 4Y6O). (d)
Schematic diagram of Figure 1c. (e) Plausible binding mode of KPM-1 on SIRT2. NCO-90
motif, glycinamide handle, and pseudopeptide structure of KPM-1 are presented in black,
purple, and green, respectively. The π-π and CH-π interactions are represented by black
dashed lines, and H-bonds are represented by purple or green dashed lines. (f) Drug design of
SIRT2-selective inhibitors.
12
ACS Paragon Plus Environment

Page 13 of 69
Journal of Medicinal Chemistry
1
2
3
4
Table 1. SIRT1 and SIRT2 inhibitory activities of compounds 2, 3, 8, 12, 18–43, 53 and 54.
5
6
7
Cmpd.
% SIRT1 inhibition at 50 µMa
% SIRT2 inhibition at 10 µMa
8
9
NCO-90
12 ± 0.36
7 ± 2.2
90 ± 1.5
75 ± 0.2
2
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7 ± 0.07
26 ± 2.7
93 ± 0.13
47 ± 0.30 (at 1 M)
68 ± 4.2
8
16 ± 0.03
12
18
19
20
21
19 ± 1.0
2.5 ± 2.5
12 ± 1.3
8 ± 0.6
12 ± 3.3
77 ± 0.01
58 ± 0.22
2 ± 2.0
57 ± 0.3
45 ± 1
70 ± 1.1 (at 50 M)
50 ± 0.33
22
8 ± 0.06
23
24
25
26
27
28
29
30
31
32
33
9 ± 2.1
15 ± 2.3
9 ± 3.7
28 ± 0.67
44 ± 4.8
12 ± 0.6
5 ± 0.83
8 ± 1
70 ± 0.92
46 ± 0.09
56 ± 0.23
54 ± 1.1
8 ± 0.54
11 ± 0.41
4 ± 0.81
13 ± 2.9
6 ± 0.06
61 ± 0.95
71 ± 0.81
68 ± 1.9
63 ± 2.7
59 ± 1.7
74 ± 0.52 (at 50 M)
76 ± 0.4
34
18 ± 1.6
35
36
37
38
39
40
41
42
12 ± 1.1
11 ± 0.41
4 ± 0.06
15 ± 1
77 ± 5.6
57 ± 2.1
81 ± 0.9
72 ± 0.45
50 ± 1.3
83± 0.25
17 ± 0.73
8 ± 2.6
23 ± 0.15
14 ± 0.23
83 ± 1.0
92 ± 0.47
59 ± 0.91 (at 1 M)
73 ± 6.8 (at 1 µM)
43
9 ± 3.9
53
54
41 ± 2.1
10 ± 1.9b
85 ± 0.47b
aFluor de Lys assay, values are means  SD of at least two experiments. aData from ref. 75.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 69
1
2
3
4
5
6
7
8
9
Initially, we tested 2, 3, and 8 (Scheme 1), which were designed by modifying the phenyl
group of NCO-90. Because the phenyl group of NCO-90 is located at a hydrophobic pocket
formed by Phe96, Phe143, and Phe190,75 we anticipated that more hydrophobic moiety such
as a naphthyl group might improve the SIRT2-inhibitory activity. However, the replacement
of the phenyl group of NCO-90 with a naphthyl group (2) resulted in decreased SIRT2-
inhibitory activity (Table 1). We next focused on the superimposition of SIRT2 crystal
structures in the complexes with NCO-90 (PDB code: 5Y5N) and SirReal1 (PDB code:
4RMI).67 The phenyl group and the thiazole ring of SirReal1 are partly superimposable on the
2-anilinobenzamide core (Figure 2). Based on this, we designed 3, in which the pyrimidine
handle of SirReal1 is connected with 2-anilinobenzamide of NCO-90. However, 3 showed
only very weak SIRT2 inhibition at 10 μM concentration. We speculated that the ester group
of 3 favored an elongated conformation that does not optimally fit in the selectivity pocket.
As previously reported by Rumpf et al.,67 a peculiarity of SirReal2 is the formation of an
intramolecular H-bond (Figure 2) that confers conformational rigidity. Such a molecular
feature was also expected in NCO-90. Thus, we designed 8, in which the pyrimidine moiety
of SirReal1 is connected with the 2-anilinobenzamide of NCO-90 via amide bond to obtain a
new scaffold with two intramolecular H-bonds. One is formed between the amide (-C=O) and
aromatic amine (-NH), and the other is formed between the secondary amide (-NH) and
nitrogen of the pyrimidine ring (Figure 2). As we expected, compound 8 exhibited similar
SIRT2 inhibitory potency to NCO-90 (Table 1).
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
14
ACS Paragon Plus Environment

Page 15 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Fragment switching design applied to NCO-90 with compounds 3 and 8 using the
pyrimidine moiety of SirReal1. These compounds were designed based on superimposition of
the SIRT2/SirReal2 (PDB code: 4RMI) and SIRT2/NCO-90 (PDB code: 5Y5N) crystal
structures.
Subsequently, we evaluated compounds 12 (Scheme 2) and 18–40 (Scheme 3), aiming to
optimize the benzamide part of NCO-90. The switch from aromatic amide of NCO-90 to
acetamide (12) did not lead to potent inhibition. Amide functionalization with aliphatic
amines (19–23) gave compounds showing 45–77% SIRT2 inhibition. Modeling studies with
21 and 23 (Figure 3a,b) showed that the aliphatic moiety displaces water molecule HOH4,
originally located near His187 and Val233 (Figure 1a), and the only interactions were with
the water molecules HOH3 and HOH11, which are thought to be weakly bound, because they
appear to be displaced upon NAD+ binding. The cyclohexyl amide derivative (24) showed the
weakest SIRT2 inhibition (28% inhibition) among the amide compounds. Docking
simulations indicate that steric hindrance near the anilinobenzamide core plus ring flexibility
might induce a suboptimal binding mode (Figure 3c), which would explain the low inhibitory
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 69
1
2
3
4
5
6
7
8
9
activity. When a phenyl ring was introduced instead of the cyclohexyl ring, a slight
improvement of SIRT2 inhibition was observed (25 vs 24). Introduction of a methylene
spacer between the aromatic ring and amide function in benzylamide derivative (26) provided
greater activity (26 vs 25), which might be ascribable to sandwiched π-π and CH-π
interaction with His187 and Phe235 (Figure 3d). However, the SIRT2-inhibitory activity was
decreased when a longer linker with a phenethylamide was used (27 vs 26). Bulky
compounds 28 and 29 showed moderate SIRT2 inhibition ≥ 54%, and N,N-diethylamide (30)
and morpholine amide (31) retained moderate SIRT2-inhibitory activity (61% and 71%
inhibition, respectively). The shift from the polar carboxylic acid precursor (18) to
hydroxamic and methylhydroxamic derivatives (32 and 33) increased SIRT2 inhibition to a
moderate level. The introduction of a nitrogen atom into the isopentylamide 22 (compound
34) did not significantly affect SIRT2 inhibition (22: 45% at 10 M, 70% at 50 M; 34: 59%
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
at 10 M, 74% at 50 M). However, the activity was improved when ethanolamide (35) was
used as an H-bond donor (35 vs 22 vs 20). Docking simulations suggested that 35 forms an
H-bond with Val233 (Figure 3e), a similar interaction pattern to that seen for 36 (Figure 3f).
Further elongation of the Boc-diaminoethane linker with three carbon atoms (37) reduced the
SIRT2 inhibition activity (36: 77% vs 37: 57%), indicating that a two-carbon spacer is
preferred for the interaction with Val233. We also tested compound 38 with an alaninamide
fragment, which is expected to interact directly with Val233 (Figure 3g). As expected, 38
showed potent SIRT2-inhibitory activity (81% inhibition). Furthermore, we screened the
glycinamide precursors 39 and 40 to evaluate the effect of the glycine derivatives without any
H-bond donor. The SIRT2-inhibitory activity of methyl ester derivative 39 was maintained,
but that of the acid derivative 40 was not. As in the case of the previously reported
glycinamide derivatives,75 the above-mentioned studies suggest that glycinamide 5475 (Figure
3h, Table 1) is an optimal fragment to mimic acetyl-lysine in the substrate binding site, but
16
ACS Paragon Plus Environment

Page 17 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
the interaction with Val233 alone did not allow full binding stabilization when longer
fragments were applied, and consequently only mild inhibition was obtained.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Docking poses of (a) 21, (b) 23, (c) 24, (d) 26, (e) 35, (f) 36, and (g) 38 in the
SIRT2/NCO-90 crystal structure. Crystallized ligand NCO-90 and docking poses are colored
in green and gray, respectively. His187 and Val233 are colored in red and delimited by green
spheres. The phenylalanines composing the “selectivity pocket” are colored in yellow. H-
bonds are represented by dashed lines. (h) Structure of glycinamide 54.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 69
1
2
3
4
5
6
7
8
9
A possible strategy to obtain potent SIRT2 inhibitors is the application of pseudopeptidic
extensions in a similar manner to previously reported KPM-1, targeting Gly236, Glu237, and
Gln267 of SIRT2 as potential H-bond sources (Figures 1e,f and 4). However, lysine
derivatives with N-, C-terminal extensions have unfavorably high molecular weight or low
stability/membrane permeability, with a nonoptimal drug-like profile. To overcome this
dilemma, we designed compounds 41–43 (Scheme 3), in which NCO-90 is functionalized
with a glycinamide linked with an alkyl chain spacer and terminal Boc or diketopiperazine
moieties carrying H-bond acceptor/donor groups. Compounds 41 and 42 exhibited over 80%
inhibition of SIRT2 at 10 M, and 43 bearing diketopiperazine, a useful fragment to target
key substrate amino acid residues, showed 92% inhibition. These results support the
hypothesis that the diketopiperazine structure might establish H-bonds with Val233, Gly236,
Glu237, and Gln267 of SIRT2. Furthermore, we designed and synthesized 53 as a thioamide
analogue of 43 (Scheme 4 and Figure 4). We expected that the sulfur of 53 would
nucleophilically attack NAD+ at the active site of SIRT2 to enable stable conjugation with
ADP-ribose (Supporting Information, Figure S1c), leading to potent SIRT2 inhibition in a
similar manner to KPM-2 (Supporting Information, Figure S1b). As we expected, 53 potently
inhibited SIRT2 even at 1 M, with relatively low inhibitory activity toward SIRT1. Indeed,
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
53 was a more potent SIRT2 inhibitor than 8 and 43 at 1 M (53: 73%; 8: 47 %; 43: 59%)
(Table 1).
18
ACS Paragon Plus Environment

Page 19 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Rational design applied to NCO-90 targeting Val233, Gly236, Glu237, and Gln267
of SIRT2, which are involved in substrate binding stabilization. Conformation of NCO-90
bound to SIRT2 (PDB code: 5Y5N) is shown with key H-bonds indicated by dashed lines,
together with the chemical structures of 43 and 53.
Determination of IC50 values and evaluation of SIRT2-selectivity of selected inhibitors.
Compounds 8, 43, and 53 were selected for further studies. First, the IC50 values for SIRT2
inhibition were determined (Table 2 and Figure 5). To assess the SIRT2-selectivity, we also
evaluated SIRT1-, SIRT3-, and SIRT5-inhibitory activity (Table 2). Compound 8 strongly
inhibited SIRT2 with an IC50 of 1.18 μM (Table 2 and Figure 5), which is in the same range
as NCO-90. Notably, diketopiperazine derivative 43 and its thioamide analog 53 were potent
and selective SIRT2 inhibitors (Table 2 and Figure 5). The SIRT2-inhibitory activity of 53
was greater than that of 43 (43: IC50 = 0.62 μM, 53: IC50 = 0.31 μM), 6 times more potent
than NCO-90, and similar to that of KPM-1 and SirReal2. In addition, 43 and 53 did not
strongly inhibit SIRT1, SIRT3, or SIRT5: indeed, the SIRT2-inhibitory activity of 53 was
over 220 times higher than the activities towards SIRT1 and SIRT3 (SIRT1 IC50/SIRT2 IC50
= 250; SIRT3 IC50/SIRT2 IC50 = 223). Accordingly, 43 and 53 are highly potent and selective
SIRT2-inhibitors.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 69
1
2
3
4
5
6
7
8
9
Table 2. IC50 values for compounds 8, 43, and 53.
IC50 ± SD [µM]a or % inhibitionb
Cmpd.
SIRT1
17% at 100 µM
51% at 100 µMd
-
SIRT2
SIRT3
17% at 100 µM
61% at 100 µMd
-
SIRT5
8% at 100 µM
6% at 100 µMd
-
NCO-90
KPM-1
1.74 ± 0.26c
0.37 ± 0.03 c
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
c
SirReal2
0.30 ± 0.06
80% at 100 µM
63% at 50 µM
17% at 100 µM
69.2 ± 2.6
28% at 100 µM
16% at 100 µM
8
1.18 ± 0.02
13% at 100 µM
2% at 100 µM
0% at 100 µM
43
53
0.60 ± 0.10
0.31 ± 0.12
77.4 ± 9.5
aFluor de Lys assay, values are calculated from three independent determinations giving
altogether at least 21 data points. b Values represent the mean ± standard deviation of at least
c
two experiments. Data from ref 75. d Not fully soluble at tested concentration.
Figure 5. IC50 curves for inhibition of SIRT2 by compounds 8, 43, and 53.
We also confirmed that compound 53 is not a pan assay interference compound (PAIN)
using other enzymes and assay systems. The inhibitory activities of 53 against histone
deacetylase 1 (HDAC1) and lysine demethylase 5A (KDM5A) were evaluated in
fluorescence assay and AlphaScreen assay, respectively.82,83 As a result, we found that 53 did
20
ACS Paragon Plus Environment

Page 21 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
not inhibit HDAC1 and KDM5A at 2.5–100 M (Supporting Information, Figure S2) and
concluded that 53 is not a PAIN.
7
8
9
Inhibitory mechanism. We next investigated whether the inhibitory activity of 43 and 53
is competitive with respect to an acetylated lysine SIRT2 substrate. The Michaelis–Menten
plot (Figure 6a,c) showed an increase of the apparent Km with increasing 43 and 53
concentration at saturating NAD+ concentration, and the Ki values were 0.47 ± 0.032 μM and
0.068 ± 0.0034 μM, respectively. A double-reciprocal plot of 1/V versus 1/[S] showed a
series of regression lines that intersect on the 1/V axis, which is characteristic of competitive
inhibition (Figure 6b,d).
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
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 69
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Competition analysis of compounds 43 and 53 with acetylated lysine substrate. (a)
Michaelis-Menten plot showing acetylated Fluor de Lys SIRT2 substrate (μM) competition
analysis at 0, 0.1, 0.4, 0.6, and 1 μM 43. (b) Lineweaver-Burk plot 1/V versus reciprocal
SIRT2 substrate concentration in the presence of 0, 0.1, 0.4, 0.6, and 1 μM 43. (c) Michaelis-
Menten plot showing Fluor de Lys SIRT2 substrate (μM) competition analysis at 0, 0.02,
0.06, 0.1, and 0.4 μM 53. (d) Lineweaver-Burk plot 1/V versus reciprocal SIRT2 substrate
concentration in the presence of 0, 0.02, 0.06, 0.1, and 0.4 μM 53.
Next, we performed mass spectrometric analysis of an incubation mixture of SIRT2 with
53 to examine whether it inhibits SIRT2 by reacting with NAD+ to afford 53-ADP-ribose
conjugate (Supporting Information, Figure S1). As depicted in Figure 7a, a significant peak
was observed at m/z 1113.3. The peak corresponds to the predicted molecular weight of the
53-ADP-ribose conjugate (Figure 7e). As this peak was not detected in the absence of SIRT2,
NAD+, or 53 (Figure 7b,c,d), these results suggest that the 53-ADP-ribose conjugate was
22
ACS Paragon Plus Environment

Page 23 of 69
Journal of Medicinal Chemistry
1
2
3
4
generated as a result of SIRT2-catalyzed reaction of 53 with NAD+, and 53 inactivates SIRT2
5
6
through mechanism-based inhibition.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. MALDI-TOF mass spectrometric detection of the ADP-ribose conjugate formed
between 53 and NAD+ (a) in the presence of SIRT2 and in the absence of (b) SIRT2, (c)
NAD+, or (d) compound 53. (e) Chemical structure of the conjugate and the calculated m/z
value for the [M−H]− ion.
Because compound 53 was a mechanism-based inhibitor, compound 43 was assumed to be
hydrolyzed in a similar mechanism (Supporting Information, Figure S3a). Therefore, we also
tested if hydrolysis of 43 occurs. We analyzed a mixture of SIRT2, NAD+ and 43 by MS
analysis. However, peaks which correspond to hydrolysis products were not detected
(Supporting Information, Figure S3b–g), suggesting that 43 is not hydrolyzed by SIRT2. It
indicates that 43 inhibits SIRT2 by occupying both SIRT2-selectivety pocket and substrate-
binding site.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 69
1
2
3
4
5
6
7
8
9
Cellular Assay. It has been reported that SIRT2 inhibition leads to growth inhibition of
breast cancer cells.73,75 Therefore, we investigated the antiproliferative activity of 43 and 53
towards breast cancer MCF-7 cells. NCO-90 was used as a reference compound. As shown in
Figure 8a, 53 significantly reduced the proliferation of MCF-7 cells at 30 M, and its
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
antiproliferative activity was greater than that of 43 or NCO-90 at 30 M (53: 96% cell
growth inhibition; 43: 40% cell growth inhibition; NCO-90: 30% cell growth inhibition).
Moreover, we determined the half-maximal growth-inhibitory concentration (GI50) of 53.
Compound 53 dose-dependently inhibited cell growth, and the GI50 value was 11.5 µM. We
also compared the growth inhibitory activity of 53 with those of TM and KPM-2. As shown
in Figure 8b, the growth inhibitory activity of 53 is superior to TM and comparable to KPM-2
(reported KPM-2 GI50: 8.43 M75). These results suggest that 53 with its diketopiperazine
structure could be a good lead structure for the development of anticancer agents. We next
examined the cellular SIRT2-selectivity of 53 by means of western blotting. As shown in
Figure 9 and Supporting Information Figure S4, treatment of MCF-7 cells with 1–5 M 53
induced the accumulation of acetylated -tubulin, which is known to be a substrate of
SIRT2.84 On the other hand, 53 did not affect the level of acetylated lysine 9 in histone H3
(H3K9Ac), which is a SIRT1 substrate,85–87 in MCF-7 cells. These results suggest that 53
selectively inhibits SIRT2 over SIRT1 in cellulo.
24
ACS Paragon Plus Environment

Page 25 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. (a) Antiproliferative activity of NCO-90, 43, and 53 towards MCF-7 cells. (b)
Antiproliferative activity of TM, KPM-2, and 53 towards MCF-7 cells. Cells were exposed to
test compound for 72 h. Values were calculated from three independent determinations.
Figure 9. Western blot detection of acetylated -tubulin or H3K9 in MCF-7 cells after 24 h
treatment with 53 or vorinostat, which is a pan-HDAC inhibitor. The latter was used as a
positive control that induces acetylation of both -tubulin and H3K9. Values of Ac--tubulin
and H3K9Ac ratio determined by optical density measurement of the blots are shown.
As it was previously suggested that SIRT2 inhibitors might be beneficial in the context of
neurological disorders,31–36 we examined the effect of 43 and 53 on the neurite outgrowth of
Neuro-2a (N2a) cells (Figure 10 and Supporting Information, Figures S5 and S6). The
treatment of N2a cells with 43 (0.2 and 2 μM) for 24 or 48 h did not significantly induce
neurite outgrowth. On the other hand, treatment with 53 (2 M) induced neurite outgrowth
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 69
1
2
3
4
5
6
7
8
9
and significantly increased the percentage of differentiated cells relative to the control group
(53 vs. ctrl for 24 h: 4.54 ± 0.48% vs. 2.53 ± 0.76%, p < 0.0001; for 48 h: 12.27 ± 1.28% vs.
4.18 ± 1.67%, p < 0.001). The percentages of differentiated cells were higher than those of
NCO-90, TM, and KPM-2, which were used as a reference compound. As TM has a long
fatty acyl chain, its plasma proteins binding rate is estimated to be high. Therefore, TM may
easily bind to plasma proteins in medium and may not effectively inhibit SIRT2 in cells.
These data suggest that 53 bearing a diketopiperazine structure may be a promising candidate
or lead compound for therapeutic agents for neurological disorders.
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 10. Effect of 43, 53, NCO-90, TM, and KPM-2 on N2a differentiation after 24 h or 48
h treatment, represented as the % of differentiated cells relative to the total counted cells (at
least 100 cells for each condition). Bars represent the mean values ± SD from three
independent experiments; *p < 0.05, ***p < 0.001, ****p < 0.0001, compared to the control
group. For representative images showing N2a cells treated with 43, 53, NCO-90, TM, and
KPM-2, see Supporting Information Figures S5 and S6.
CONCLUSION
To identify novel SIRT2-selective inhibitors, we used structure-based drug design based on
the previously reported X-ray structure of SIRT2/NCO-90 complex. Using a simple fragment
replacement strategy between NCO-90 and SirRea1, we found that 2-anilinobenzamide also
26
ACS Paragon Plus Environment

Page 27 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
offers a versatile scaffold to probe the substrate-binding site and selectivity pocket.
Furthermore, focusing on the key roles of Val233, Gly236, Glu237, and Gln267 in SIRT2 in
stabilizing the binding of Ac-lysine-mimetic fragments, we developed a novel
diketopiperazine structure as an “H-bond hunter” to target these amino acid residues.
Compounds 43 and 53 with a diketopiperazine structure on a 2-anilinoamide scaffold were
potent and selective SIRT2 inhibitors with inhibition profiles comparable to that of KPM-1.
This finding opens up new perspectives in the design of previously unexplored drug-like
sirtuin mechanism-based inhibitors. In cell-based assays, 53 showed potent antiproliferative
activity in breast cancer cells and neurite outgrowth-inducing activity in N2a cells, suggesting
that diketopiperazine/2-anilinobenzamide-based SIRT2 inhibitors are promising candidates or
lead compounds for novel therapeutic agents for cancer and neurological disorders.
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 SECTION
Chemistry. The chemical reagents and solvents used in this study were commercial
products of the highest available purity. Reagents and solvents were purchased from Sigma
Aldrich, Wako Pure Chemical Industries, and TCI Tokyo Chemical Industry Co., Ltd.
Organic solvents were dried over anhydrous sodium sulfate. Compounds 1, 18, 39, and 40
were prepared according to Suzuki et al.61 or Mellini et al.75 (additional data given below).
Commercially available compounds 4, 5, 9, and 13 were used without purification. The
synthetic route to 17 was adapted from Kaur et al.88 NMR spectra were recorded on a Bruker
Avance 300 AV (Bruker Biospin, Swizerland) spectrometer operating at 300.1 MHz (1H) or
75.5 MHz (13C). The chemical shift values are reported as δ (ppm) relative to TMS
(tetramethylsilane) as an internal reference (δ = 0), and coupling constants are given in Hz.
Positive/negative LRMS ion mass spectra were recorded on a Bruker HCT-Plus. The purity
of all tested compounds was > 95%, except for 32 (92%), as determined by HPLC using a
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 69
1
2
3
4
5
6
7
8
9
Shimadzu UFLC (SPD-M20A UV detector, DGU-20A3R degassing unit, LC-20AD solvent
delivery unit and CBM-20A system) and a C18 column (Inert Sustain, 4.6*150, 5 µM), at a
flow rate of 1 mL/min, with UV detection (λ = 220 or 254 nm). HPLC conditions: eluent A:
H2O containing 0.1% TFA; eluent B: acetonitrile containing 0.1% TFA. Gradient: B: 0 to 20
min, 10–90%; 20 to 30 min, 90%; 30 to 40 min, 90–10%. Melting points were determined
using a Yanaco Micro Melting Point apparatus. High-resolution mass spectra (HRMS) were
recorded on a JEOL JMS-SX102A mass spectrometer or a Shimadzu LCMS-IT-TOF mass
spectrometer.
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
Synthesis of 2-({3-[2-(naphthalen-2-yl)ethoxy]phenyl}amino)benzamide (2). To a
solution of 161 (0.22 g, 0.96 mmol) in dry acetone (15.0 mL) were added K2CO3 (0.40 g, 2.9
mmol) and 2-(2-bromoethyl)naphthalene76 (0.45 g, 1.9 mmol). The mixture was heated at
reflux under an N2 atmosphere for 18 h, then insoluble materials were filtered off, and the
filtrate was concentrated in vacuo. The crude product was purified by column
chromatography on silica Kieselgel 60 (n-hexane:EtOAc = 6:4) to give a white solid (0.063 g,
1
0.16 mmol, 17%); Rf = 0.28 (n-hexane:EtOAc = 6:4); m.p. 152–153 ºC; H NMR (DMSO-
d6): δ 9.94 (s, 1H), 8.03 (s br, 1H), 7.89–7.82 (m overlap, 4H), 7.70–7.67 (m, 1H), 7.52–7.42
(m, 4H), 7.31–7.29 (m overlap, 2H), 7.20 (t, 1H, J = 8.3 Hz), 6.82–6.76 (m, 1H), 6.76–6.70
13
(m overlap, 2H), 6.58–6.54 (m, 1H), 4.27 (t, 2H, J = 6.8 Hz), 3.19 (t, 2H, J = 6.8 Hz); C
NMR (DMSO-d6): δ = 171.17, 159.45, 144.62, 142.83, 136.04, 133.06, 131.99, 131.74,
130.09, 129.26, 127.64 (2C), 127.58, 127.40, 127.28, 126.99, 125.94, 125.33, 117.91, 115.30,
111.85, 108.06, 105.58, 67.92, 35.03; ESI-MS (m/z): 383.2 [M + H]+; HRMS (EI) calcd for
C25H22N2O2, 382.16813, found, 382.16749; HPLC: purity 98% at 254 nm, tR: 21.9 min.
Synthesis of 3-[(2-carbamoylphenyl)amino]phenyl 2-[(4,6-dimethylpyrimidin-2-
yl)thio]acetate (3). To a solution of 2-[(4,6-dimethylpyrimidin-2-yl)thio]acetic acid76 (0.18
g, 0.91 mmol) in dry DMF, COMU® (0.47 g, 1.1 mmol) were added Et3N (0.38 mL, 2.7
28
ACS Paragon Plus Environment

Page 29 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mmol) and 1 (0.21 g, 0.91 mmol) with cooling in an ice bath. The mixture was allowed to
warm to room temperature. After 18 h, the resulting dark solution was diluted with brine (~30
mL) and extracted with EtOAc (4 x 40 mL). The combined organic phase was washed with
saturated aqueous NaHCO3 solution (twice) and brine, and dried over Na2SO4. Filtration,
evaporation, and purification of the residue by column chromatography on silica Kieselgel 60
(n-hexane:EtOAc = 1:1) gave a light pink solid (0.098 g, 0.24 mmol, 27%); Rf = 0.22 (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
1
hexane:EtOAc = 1:1); m.p. 71–72 ºC; H NMR (DMSO-d6): δ = 10.04 (s, 1H), 8.06 (s br,
1H), 7.71 (dd, 1H, J = 7.9, 1.1 Hz), 7.49 (s br, 1H), 7.33–7.25 (m overlap, 3H), 7.02–6.98 (m
overlap, 2H), 6.87–6.82 (m overlap, 2H), 6.65 (dd, 1H, J = 7.9, 2.1 Hz), 4.18 (s, 2H), 2.32 (s,
6H); 13C NMR (DMSO-d6): δ = 171.02, 168.61, 167.91, 167.02, 151.45, 143.88, 142.97,
131.99, 130.23, 129.32, 118.63, 118.61, 116.44, 116.19, 115.57, 114.10, 111.43, 33.05,
23.25; ESI-MS (m/z): 409.1 [M + H]+; HRMS (ESI) calcd for C21H20N4O3SNa+, 431.1148;
found, 431.1151; HPLC: purity 99% at 254 nm, tR: 17.1 min.
Synthesis
of
2-[(3-{2-[(4,6-dimethylpyrimidin-2-
yl)thio]acetamido}phenyl)amino]benzamide (8). Step 1: preparation of 2-[(3-
nitrophenyl)amino]benzamide (6). A solution of 4 (0.32 g, 2.4 mmol), 5 (0.4 g, 1.98
mmol), K2CO3 (0.38 g, 2.8 mmol), Pd2dba3 (0.16 g, 0.18 mmol) and XPhos (0.19 g, 0.39
mmol) in tert-BuOH (8 mL) was heated at reflux under an N2 flow for 20 h, then EtOAc (~20
mL) was added to the reaction mixture. Insoluble materials were removed by filtration, and
the filtrate was evaporated in vacuo. The residue was purified by column chromatography on
silica Kieselgel 60 (n-hexane:EtOAc = 6:4) to give an orange solid (0.25 g, 0.99 mmol, 50%);
Rf = 0.29 (n-hexane:EtOAc = 6:4); 1H NMR (DMSO-d6): δ = 10.05 (s, 1H), 8.09 (s br, 1H),
7.91–7.90 (m, 1H), 7.76–7.69 (m overlap, 2H), 7.58–7.52 (m overlap, 3H), 7.43–7.39 (m
overlap, 2H), 7.01–6.95 (m, 1H); 13C NMR (DMSO-d6): δ = 170.68, 148.71, 143.72, 142.49,
29
ACS Paragon Plus Environment