You are on page 1of 11

European Journal of Medicinal Chemistry 92 (2015) 145e155

Contents lists available at ScienceDirect

European Journal of Medicinal Chemistry


journal homepage: http://www.elsevier.com/locate/ejmech

Original article

Functionalized tetrahydro-1H-pyrido[4,3-b]indoles: A novel


chemotype with Sirtuin 2 inhibitory activity
Tianming Yang a, Xiao Chen a, Hai-xiao Jin b, Gautam Sethi c, Mei-Lin Go a, *
a

Department of Pharmacy, National University of Singapore, 18 Science Drive 4, 117543, Republic of Singapore
Key Laboratory of Applied Marine Biotechnology, Ministry of Education, School of Marine Sciences, Ningbo University, Fenghua Road 818, Jiangbei District,
Ningbo, Zhejiang 315211, People's Republic of China
c
Department of Pharmacology, Yong Loo Lin School of Medicine, National University of Singapore, 10 Medical Drive, 117597, Republic of Singapore
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 22 August 2014
Received in revised form
1 December 2014
Accepted 17 December 2014
Available online 18 December 2014

Sirtuins are protein deacylases with regulatory roles in metabolism and stress response. Functionalized
tetrahydro-1H-pyrido[4,3-b]indoles were identied as preferential sirtuin 2 inhibitors, with in vitro
inhibitory potencies in the low micromolar concentrations (IC50 3e4 mM) for the more promising candidates. The functional relevance of sirtuin inhibition was corroborated in western blots that showed
hyperacetylation of p53 and a-tubulin in treated HepG2 and MDA-MB-231 cells. Molecular docking
showed that the tetrahydropyridoindole scaffold was positioned in the NAD pocket and the acetylated
substrate channel of the sirtuin 2 protein by van der Waals/hydrophobic, H bonding and stacking interactions. Functionalized tetrahydropyridoindoles represent a novel class of sirtuin 2 inhibitors that
could be further explored for its therapeutic potential.
2014 Elsevier Masson SAS. All rights reserved.

Keywords:
Tetrahydropyridoindoles
Sirtuin 2 inhibition
Molecular docking
Hyperacetylation of p53 and a-tubulin
Apoptosis
Growth inhibition

1. Introduction
Sirtuins are a class of evolutionary conserved nicotinamide
adenine dinucleotide (NAD) -dependent protein lysine deacylases
with important roles in diverse and interrelated cellular processes
such as stress response, gene expression, DNA damage repair and
metabolism [1]. They are implicated in age-related diseases such as
cancer, neurodegeneration and metabolic disorders and hence are
considered as prospective therapeutic targets for these conditions.
However, perplexing gaps remain in our understanding of their
modes of action. In carcinogenesis, sirtuins have bifurcated roles
and may function as tumor suppressors and promoters depending
on contextual variables such as the stage of the malignancy and the
tumor microenvironment [2]. High-throughput and in silico
screening have identied novel chemotypes with sirtuin inhibitory
activities such as thieno [3,2-d]pyrimidine-6-carboxamides [3],
macrocyclic peptides [4], chroman-4-ones [5], 2-hydroxy-1naphthalenes [6,7] and functionalized indoles [8e11]. Interestingly, certain scaffolds were observed to be more widely associated

* Corresponding author.
E-mail address: phagoml@nus.edu.sg (M.-L. Go).
http://dx.doi.org/10.1016/j.ejmech.2014.12.027
0223-5234/ 2014 Elsevier Masson SAS. All rights reserved.

with sirtuin inhibition than others. The indole ring is a case in point
- it is embedded in the bisindolylmaleimide derivative Ro 31-8220
[8], the oxindole GW 5074 [8,9], the tetrahydrocarbazole EX527
[10] and the carprofen analog 1 (2-(6-chloro-9H-carbazol-2-yl)
propanamide) [11] (Fig. 1A). We noted that EX527 and the caprofen
analog 1 bore a striking structural resemblance in that both compounds were ring-fused indoles with primary amide functionalities. They were also selective Sirt1 inhibitors but differed in their
cell-based activities. EX527 did not abolish cell growth and proliferation [12,13] and had varied effects on the acetylation status of
p53, a substrate of Sirt1 and Sirt2, depending on experimental
conditions [11e13]. In contrast, the carprofen analog 1 induced
apoptosis in leukemic cells and its sirtuin inhibitory activity was
conrmed in functional assays [11].
To further explore the potential of the indole scaffold, we evaluated a series of 2,3,4,5-tetrahydro-1H-pyrido[4,3-b]indoles for
sirtuin inhibition (Fig. 1B). The indole ring of this scaffold is fused to
a basic piperidine ring and not benzene or cyclohexane as in 1 and
EX527. Our results showed that tetrahydropyridoindole is a novel
chemotype associated with preferential Sirt2 inhibitory activity. A
representative member 18 inhibited sirtuin-mediated deacetylation of physiological substrates p53 and a-tubulin in liver (HepG2)
and breast (MDA-MB-231) cancer cells at low micromolar

146

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

Fig. 1. (A) Structures of known indole-based sirtuin inhibitors (B) Tetrahydropyridoindole scaffold with functionalization at positions 2 (R1), 5 (R2) and 8 (R3). R1,R2, R3 of lead
compound 8 are indicated.

concentrations (2e 4 mM). Molecular docking showed that the


tetrahydropyridoindole ring was positioned at the NAD pocket
and substrate channel of the Sirt2 protein, with binding afnity
modulated by the side chains attached to the scaffold.

2. Results and discussion


2.1. Design and synthesis of target compounds
The lead compound for this series was 8 (Fig. 1B) which was
serendipitously found to occupy the NAD binding pocket of human Sirt2 (PDB 3ZGV) [12]. As shown in Fig. 2A, the tetrahydropyridoindole ring of 8 straddled the nicotinamide-ribose site B
(Arg 97, Phe 96), substrate (acetyl lysine) channel (Phe 119, His 187,
Val 233, Phe 235) and nicotinamide site C (lle169). Several stacking
interactions (pp, p-cation) were identied, namely Phe 96, His 187
and the aromatic ring of the scaffold (pp, pH respectively, with
distances of 3.81 and 3.74 ), Arg 97 and m-tolyl ring (pH, 4.01 )
and Phe 235 and piperidine NH (pH, 3.60 ). The n-octyl side
chain of 8 occupied a hydrophobic channel that extended from site
C (Fig. 2B).
When 8 was screened for inhibition of Sirt1 and Sirt2 (as
described later), it was found to inhibit Sirt2 (43%, 10 mM) to a
greater extent than Sirt1 (5%, 10 mM). With this lead on hand, we
modied 8 at positions 2 (R1), 5 (R2) and 8 (R3) with the intent of
establishing structure-activity relationships (SAR) and optimizing
inhibition. To that end, thirty analogs that were structurally related
to 8 were synthesized. Having observed that the n-octyl side chain
of 8 was positioned in a lipophilic region of the NAD pocket, we
reasoned that it was an essential structural feature. However, its
presence would invariably impart considerable lipophilicity to the
resulting compounds. The lipophilicity of 8, as seen from its logP
(7.55) and logDpH 7.4 (5.93) values, were high. Hence to avoid
exacerbating the greasiness of the target compounds, a deliberate
attempt was made to introduce polar residues at R1 and R3. To that
end, several polar and sterically limited entities were introduced at
R1, such as hydroxyethyl, aminocarbonyl, methylsulfonyl, methyl,
ethyl and hydrogen. In the case of R3, an arbitrary selection of 30 tolyl, 40 -methylsulfonylphenyl, 20 -aminopyrimidin-50 -yl, 20 -uoropyridin-40 -yl and uoro was made, prompted in part by the
commercial availability of reagents. These R3 groups formed the
basis of classifying the synthesized compounds to their respective
series A-E (Table 1). As shown later, series D and E which had uoropyridinyl and uoro as R3 respectively, were weak sirtuin inhibitors, further emphasizing the need to deploy small or polar
groups at R1. Analogs with shorter (n-butyl, isoprenyl) or no R2 side
chains were also synthesized to conrm the necessity of retaining
the n-octyl side chain.
Most of the compounds (except series E) were derived from a
common intermediate ethyl 8-bromo-3,4-dihydro-1H-pyrido[4,3b]indole-2(5H)-carboxylate (2) which was obtained from a

Fischer indole reaction between 4-bromophenylhydrazine hydrochloride and 1-carbethoxy-4-piperidone (Scheme 1). The indole
nitrogen of 2 was N-alkylated with 1-bromooctane in the presence
of sodium hydride, followed by alkaline hydrolysis and decarboxylation of the carbamate moiety at position 2 to give 4a. The latter
was reacted with formaldehyde, acetaldehyde or acetone in the
presence of sodium triacetoxyborohydride to give the N-methyl
(5a), N-ethyl (5b) or N-isopropyl (5c) intermediate. 4a was also
reacted with methanesulfonyl chloride, 2-bromoethanol or potassium cyanate to give 5d, 5e or 6 respectively. In the next step,
palladium catalyzed Suzuki coupling was carried out to introduce
substituted aromatic moieties at position 8. This sequence was
broadly followed throughout, except for analogs with 30 -tolyl at
position 8 (9, 10, 11, and 12) where higher yields were obtained
when the 30 -tolyl ring was introduced rst (Scheme 1). The 8-uoro
substituted analogs of series E (35, 36) were obtained by the Fischer
indole reaction between 4-uorophenylhydrazine hydrochloride
and 1-carbethoxy-4-piperidone, followed by the usual reactions
(Supplementary information, Section 9). In the same way, the
indole nitrogen of 2 was reacted with 1-bromobutane to give nbutyl analogs (16, 22, and 32). Alternatively, the indole nitrogen of 2
was left in its unsubstituted state and functionalized at positions 2
and 8 as described earlier to give 14, 15, 21 and 31 (Supplementary
information, Section 6).

2.2. Sirtuin inhibitory activity of test compounds


The tetrahydropyridoindoles were evaluated against recombinant human Sirt1 and Sirt2 in an assay that monitored the rate at
which the enzyme deacetylated an N-acetyl lysine residue in a
peptide substrate bearing the amino acid residues 317e320 of p53
and conjugated to aminoluciferin. The deacetylation reaction triggered a protease mediated hydrolysis which released aminoluciferin. The latter was oxidized by luciferase to oxyluciferin which
is luminescent. Inhibition of sirtuin resulted in less oxyluciferin
being generated and hence a reduction in the luminescent signal.
The compounds were rst evaluated at a xed concentration of
10 mM on Sirt1 and Sirt2 (Table 1). Of the 30 compounds evaluated,
8 analogs (9, 12, 13, 18, 19, 24, 26, 28) inhibited Sirt2 to a greater
extent (50% inhibition) than Sirt1 (50% inhibition) while 2
compounds (10, 20) inhibited both Sirt1 and Sirt2 by more than
50%. The remaining compounds were weak inhibitors (<50% inhibition) of both enzymes.
To ensure that the reduction in luminescence was not due to
aberrant inhibition of the protease or luciferase enzymes, the assay
was repeated using deacetylated substrate in the absence of sirtuin
for selected compounds (9, 10, 12, 13, 18, 19, 20, 22, 24, 26) at 10 mM
(Supplementary information, Section 12). We found luminescence
readings to be comparable to that of control readings (no test
compound) indicating the absence of non-specic inhibition of the
protease and luciferase enzymes.

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

147

Fig. 2. Docking pose of 8 in the NAD binding pocket of Sirt2 (PDB 3ZGV). (A) Ligand interaction map depicting amino acid residues in the NAD pocket which are in contact with
8. Polar and lipophilic residues are depicted in magenta and green respectively. (B) Orientation of 8 in the NAD binding pocket. The n-octyl side chain is anked by non-polar
residues Phe 119, Phe 131, lle232, lle169 which were found in site C and the hydrophobic channel extending from site C. (For interpretation of the references to colour in this gure
legend, the reader is referred to the web version of this article.)

We also explored the possible involvement of aggregate formation as a source of non-specic inhibition [13,14]. Some compounds are known to form submicrometer aggregates capable of
sequestering enzymes onto their surfaces, resulting in an artefactual reduction of enzyme activity. Such compounds (aggregators)
are generally characterized by poor solubility, high lipophilicity and
extended conjugation in their structures [13]. As some of these
features were present in our compounds, the involvement of
aggregate formation in the observed inhibition was investigated.
Aggregators normally form particles of 30e100 nm diameter that
were capable of scattering light. To that end, we monitored the light
scattering properties of selected compounds (9, 10, 12, 13, 18, 19, 20,
24, 26 and 28) at 1 mM and 10 mM in phosphate buffer pH 7.4
(Supplementary information, Section 14). Negligible light scattering was observed at 1 mM but higher levels were observed at
10 mM of 9, 12 and 13. This raised the question as to whether aggregation could have contributed to the inhibitory activities of
these compounds (Table 1). The likelihood was deemed to be slight
because 9, 12 and 13 (at a xed concentration of 10 mM) were
paradoxically weak inhibitors of Sirt1 although reasonably strong
inhibitors of Sirt2. We reasoned that if aggregate formation
contributed to artefactual inhibition of sirtuins, strong inhibition
should be observed for both and not just one enzyme.
We then proceeded to determine the IC50 (concentration
required to reduce basal oxyluciferin luminescence by 50%) of those
compounds that inhibited Sirt2/Sirt1 by more than 50% at 10 mM.
Some less inhibitory compounds were included for comparison.
The results are presented in Table 1.
The test compounds were found to be more potent inhibitors of
Sirt2 than Sirt1, with a preference for Sirt2 inhibition that varied
from 2-fold (8) to more than 40 fold (19) based on IC50 values. The
Sirt2 selective inhibitor AGK2 was used as a positive control and it
was gratifying to note that some of the Series A and B analogs were
more inhibitory than AGK2 on this assay.
Based on the inhibitory data at 10 mM and IC50 values, a
structure-activity relationship (SAR) for Sirt2 inhibition is proposed. First, it is important to maintain an alkyl substituted indole
nitrogen at R2. For compounds with the same R1 and R3, inhibition

was most pronounced when R2 was n-octyl. n-Octyl analogs were


consistently more inhibitory than their n-butyl or unsubstituted
(NH) counterparts in Series A (8 > 14, 16), B (18 > 22) and C
(24 > 32). There were however some n-octyl analogs with weak
Sirt2 inhibition (at 10 mM). For example, 11 and 17 had similar
groups at R1 (40 -hydroxybenzyl) and R3 (30 -tolyl) but different R2
groups (isoprenyl in 17 and n-octyl in 11). Stronger Sirt2 inhibition
was anticipated for the n-octyl bearing 11 but as seen from Table 1,
both compounds weakly inhibited Sirt2 (20e40% inhibition at
10 mM). Evidently, one or both substituents at R1 and R3 of 11 were
not optimal for activity. Thus, R2 notwithstanding, groups at R1 and
R3 should not be ignored as they may have modulatory roles.
That the groups at R1 and R3 were important was further supported by examining the actives (compounds that inhibited Sirt2
by 50% or more at 10 mM). By this measure, small polar groups were
preferred at R1 as seen from the recurring presence of
R1 hydrogen, methyl and aminocarbonyl among these compounds. On the other hand, compounds bearing ethoxycarbonyl
(polar, bulky) and 40 -hydroxybenzyl (bulky, less polar) side chains
were noticeably absent. In the case of R3, a key observation was that
actives were only found in series A (R3 30 -tolyl), B (R3 40 methylsulfonylphenyl) and C (R3 20 -aminopyrimidin-50 -yl), and
not series D (R3 30 -uoropyridin-4-yl) and E (R3 F). A comparison of IC50 values of compounds across Series A, B and C with
the same R1 and R2 (n-octyl) groups revealed the following trends:
R1 H: 18 (series B, 3.8 mM) > 8 (series A, 10.1 mM) z 24 (series C,
11.8 mM); R1 CH3: 19 (series B, 6.0 mM) z 10 (series A,
6.2 mM) > 26 (series C, 17.9 mM); R1 CONH2: 20 (series B,
4.0 mM) > 13 (series A, 7.8 mM) > 25 (series C, 14.0 mM). Taken
together, the SAR suggests that small (uoro in series E) and heteroaromatic azines (series C aminopyrimidinyls, series D uoropyridinyls) were not favored at R3.
There is concern that most of the identied actives were found
in series A or B which had more lipophilic substituted-phenyl rings
at R3, as compared to the polar heteroaromatic azines at R3 of series
C. Excessive lipophilicity would promote non-specic binding and
spurious cytotoxicity to normal tissues. Thus, we evaluated the
cytotoxic IC50 of 18 and 20, two of the more potent Sirt2 inhibitors,

148

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

Table 1
Structures of synthesized compounds and their in vitro inhibitory activities on Sirt1 and Sirt2.
R1
N

R3
N
R2

Cpd

R1

R2

Sirt1

Sirt 2

% Inhibition (10 mM)a

IC50 (mM)a

% Inhibition (10 mM)a

IC50 (mM)a

Series A: R3

7
8
9

eCOeOeC2H5
H

C8H17n
C8H17n
C8H17n

Nilb
5.1 1.0
<5

Not done
20 11
>100

11.2 3.9
43.1 3.2
57.5 3.6

Not done
10.1 0.8
8.2 0.1

10
11

eCH3

C8H17n
C8H17n

66.4 2.5
Nilb

21 1
Not done

64.2 2.4
19.7 13.8

6.2 0.2
Not done

12
13
14
15
16
17

eCH(CH3)2
eCONH2
H
eCH3
H

C8H17n
C8H17n
H
H
C4H9n

2.4 1.9
Nilb
Not done
Not done
Nilb
Nilb

26 4
>100
Not done
Not done
Not done
Not done

56.3 4.3
49.1 9.5
Nilb
Nilb
Nilb
38.1 11.4

11.2 0.7
7.8 0.7
100 11
83 4
96 1
Not done

18
19
20
21
22
Series C: R3

H
eCH3
eCONH2
eCH3
H

C8H17n
C8H17n
C8H17n
H
C4H9n

48.5 2.4
12.8 3.9
51.8 11.4
Not done
Nilb

16.7 0.4
>250
28 3
Not done
Not done

82.5 3.3
79.1 7.4
76.5 4.4
Nilb
Nilb

3.8 0.3
6.0 0.4
4.0 0.2
>250
20 1

23
24
25
26
27
28
29
30
31
32
Series D: R3

eCOeOeC2H5
H
eCONH2
eCH3
eC2H5
eCH(CH3)2
eSO2CH3
eC2H4OH
eCH3
H

C8H17n
C8H17n
C8H17n
C8H17n
C8H17n
C8H17n
C8H17n
C8H17n
H
C4H9n

Nilb
2.4 1.1
Nilb
Nilb
Nilb
15.3 7.6
Nilb
Nilb
Not done
Nilb

Not done
38 2
63 4
>100
Not done
56 2
Not done
Not done
Not done
Not done

23.0
70.0
34.0
61.9
29.1
62.3
35.0
31.1
Nilb
Nilb

Not done
11.8 1.8
14.0 0.8
17.9 2.2
Not done
20 6
Not done
Not done
>250
>100

C8H17n
C8H17n

1.1 0.8
Nilb

Not done
Not done

27.7 3.3
18.5 8.4

Not done
Not done

C8H17n
C8H17n

Nilb
Nilb

Not done
Not done

17.4 4.9
Nilb

Not done
Not done

Series B: R3

33
H
34
eCONH2
Series E: R3 F
35
eCH3
36

AGK2c
EX527d
Nicotinamidee
a
b
c
d
e

Mean SD for n 3 separate determinations.


No inhibition was observed at 10 mM test compound.
Selective Sirt2 inhibitor [15].
Selective Sirt1 inhibitor [10].
Sirtuin 1 and 2 inhibitor [10].

Not done
0.33 0.03
149 12

3.3
8.8
1.5
11.4
1.6
7.9
3.5
2.4

13.9 1.0
Not done
6.3 1.4

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

149

Scheme 1. Reagents and conditions: (a) Ethanol, reux, 12 h; (b) NaH (60%), 1-bromooctane, DMF, 0  C to rt; (c) m-Tolylboronic acid or (2-aminopyrimidin-5-yl)boronic acid pinacol
ester, Pd(PPh3)4, K2CO3 (aqueous solution), 1, 4-dioxane, N2 atmosphere, reux, 10 h; (d) KOH (aqueous solution), ethanol, reux, 16 h; (e) For 5a, 5b and 5c: appropriate aldehyde or
ketone, NaBH(OAc)3, acetic acid, 1, 2-dichloroethane, rt, overnight; (f) For 5d: CH3SO2Cl, Et3N, rt; (g) For 5e: 2-BrC2H4OH, K2CO3, 80  C; (h) Appropriate boronic acid or boronic acid
pinacol ester, Pd(PPh3)4, potassium carbonate (K2CO3, aqueous solution), 1, 4-dioxane, microwave, 110  C, 15e30 min; (i) potassium cyanate (KNCO), 4 M HCl, ethanol, microwave
80  C, 60 min.

on non-malignant human lung broblast IMR90 cells. 18 was found


to be more cytotoxic (IC50 2.72mM 0.32) than 20 (IC50
14.5mM 1.5). The estimated log P of 18 was 6.1 as compared to 5.5
for 20 (ACD/Labs Version 12.0). However, 18 is a secondary cyclic
amine and its lipophilicity would be better represented by log
DpH7.4 which takes into account its ionization state. Log D (pH7.4)
was found to be 3.4 for 18 whereas the corresponding value for 20,
which has no ionizable group, was higher at 5.5 (ACD/Labs Version
12.0). Thus, the cytotoxicities of 18 and 20 on IMR90 cannot be
solely attributed to their (estimated) lipophilicities. Follow up investigations on other non-malignant cell lines should be carried out
for conrmation.
Besides cytotoxicity, excessive lipophilicity would adversely
affect drug-like character, in particular solubility. To that end, we
assessed the solubilities of selected compounds (10, 12, 18, 19, 20,
24) from Series A, B and C by a turbimetric method [16]. As seen
from Table 2, 24 from series C (R3 2-aminopyridimidin-5-yl)
which had moderate Sirt2 inhibitory activity (IC50 11.8 mM), had a

Table 2
Solubilities of selected compounds as determined by turbimetry at pH 7.4, 25  C.
Series

Compound

Solubilitya

Series

Compound

Solubilitya

A
A
B

10
12
18

3.1e6.3 mM
1.6e3.1 mM
3.1e6.3 mM

B
B
C

19
20
24

3.1e6.3 mM
1.6e3.1 mM
6.3e12.5 mM

From at least n 3 separate determinations.

more favorable solubility prole than compounds from series A and


B which were more potent Sirt2 inhibitors. At this stage of lead
optimization, moderately active but less lipophilic compounds like
those in series C should not be entirely ignored as they may yet
yield more potent and drug-like analogs upon structural
modications.
2.3. Kinetics of Sirt2 inhibition by 18
The kinetics of Sirt2 inhibition was investigated for a representative compound 18. The activity of Sirt2 as a function of the
acetylated substrate (Fig. 3A) or NAD (Fig. 3B) was monitored in
the presence of varying concentrations of 18. Both plots showed a
dose-dependent decrease in the maximal enzyme activity with
increasing 18 that was not overcome at high substrate concentrations. Such a prole was indicative of non-competitive inhibition
which meant that 18 binds to both the enzyme and the enzymeesubstrate complex (ES) or subsequent species [17]. To determine the comparative afnity of 18 for the enzyme or the ES
complex, the double reciprocal plots of 18 for the two substrates
were examined. As seen from Fig. 3C and D, both plots displayed a
nest of lines that converged below the x-axis, indicating that 18 had
greater afnity for the ES complex than the free enzyme [17]. To
obtain the Ki of 18, the slopes of the double reciprocal plots obtained in presence of different concentrations of 18 were re-plotted
against inhibitor concentration (Supplementary information,
Section 13) [18]. Apparent Ki of 18 was found to be 2.28mM 0.42

150

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

Fig. 3. Kinetics of inhibition of SIRT2 by 18 at concentrations of 10 mM ( ), 3.3 mM ( ), 1.1 mM ( ) and 0 mM ( ). Panel (A) shows the rate (luminescence per min) of catalysis as a
function of acetylated substrate at a xed NAD concentration (5 mM). Panel (B) depicts the rate of catalysis as function of NAD at a xed concentration (35 mM) of the acetylated
substrate. The double reciprocal plot of 1/rate versus 1/[acetylated substrate] and 1/[NAD] at different concentrations of 18 are depicted in Panels C and D respectively.

(for acetylated substrate) and 3.70mM 0.35 (for NAD).


2.4. Molecular docking of series AeC compounds in the Sirt2
binding pocket
In our preliminary investigations, we examined the docking
pose of lead compound 8 in the Sirt2 pocket (PDB 3ZGV) and found
that its tetrahydropyridoindole ring straddled several sites in the
NAD pocket of Sirt2, namely site B, site C and the substrate
channel (Fig. 2). We found that 18, 19, 20 of Series B and 24, 25, 26,
28 of Series C were similarly positioned in the NAD binding
pocket as 10. Fig. 4 depicts the docking poses of 18 (Series B) and 24
(Series C) in the Sirt2 pocket. These poses were representative of
other compounds in the respective series. Signicant interactions
were the positioning of the n-octyl side chain along a hydrophobic
channel which extended from site C, and pp/p-cation interactions
between the scaffold and Phe 96 (site B), His 187 (substrate channel) and Phe 235 (substrate channel). The p-cation (pH) interaction
with Phe235 was observed only when the piperidine nitrogen was
protonated and was absent in 20 and 25 which had an aminocarbonyl substituent at R1. We noted that the aromatic rings at
position 8 of 18 and 24 were involved in H bonding but this was at
the expense of p-cation stacking interactions that were characteristic of the series A analog 8 (Fig. 2). As shown in Fig. 4, the sulfonyl
moiety of 18 was H bonded to Ser 263 and the amino group of 24
was H bonded to Gly 261. These residues were located in the polar
region that accommodates the phosphoryl residues of the cocrystalized ligand ADP-ribose. The docking pose of ADP-ribose
showed that its phosphate residues were H bonded to Ser 263.
The involvement of Ser 263 in H bonding with the sulfonyl moiety
of the series B analogs 18, 19, 20 may account for the stronger Sirt2
inhibition observed for these compounds.
Although 18 exhibited non-competitive inhibition of Sirt2,
molecular docking showed that it occupied the NAD and substrate
binding sites in the enzyme. While this may seem paradoxical, we
noted that EX-527 which inhibited several sirtuins non-

competitively or uncompetitively, was also reported to occupy


the nicotinamide binding site of the enzyme [10,19].
2.5. 18 increased levels of acetylated p53 and alpha-tubulin in
cancer cell lines
Having shown that 18 inhibited the Sirt2 mediated deacetylation of a luminogenic peptide, we proceeded to determine if it
inhibited deacetylation of physiological substrates of Sirt2. To that
end, HepG2 and MDA-MB-231 cells were incubated with varying
concentrations of 18 (2e8 mM) after which acetylated p53 (acetylated at Lys 382) and acetylated a-tubulin were probed by immunoblotting (Fig. 5A, B). p53 is a substrate of both Sirt1 and Sirt2
whereas a-tubulin is a Sirt2 substrate [2]. 18 had a growth inhibitory IC50 of 2.1 mM on HepG2 and MDA-MB-231 (Supplementary
information, Section 15). As seen in Fig. 5A, B, 18 increased the
levels of acetylated p53 and a-tubulin in these cells at 2e4 mM.
Acetylation of p53 increased its stability by limiting ubiquitination of key lysine residues and subsequent proteasomal degradation [20,21]. The stabilized and hence activated p53 is thus able
to increase transcription of genes involved in cell-cycle arrest, DNA
repair and apoptosis. Having shown that 18 increased levels of
acetylated p53, we proceeded to determine if this would translate
to apoptotic cell death. Hence, two apoptotic marker proteins
(caspase 3 and PARP) were monitored by immunoblotting. In the
event of apoptosis, elevated levels of cleaved caspase 3 and cleaved
PARP would be observed and this were duly noted in 18- treated
MDA-MB-231 cells (Fig. 5C).
3. Conclusion
We have shown that functionalization of the tetrahydropyridoindole scaffold resulted in analogs with low micromolar Sirt2
inhibitory activity. There was a modest preference for Sirt2 inhibition (2e40 fold) among the potent members, unlike the structurally related EX527 and caprofen analog 1 which were Sirt1

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

151

Fig. 4. Docking poses of 18 (A,B) and 24 (C,D) in the NAD binding pocket of Sirt2 (PDB 3ZGV). (A) and (C) are ligand interaction maps depicting amino acid residues in the
NAD pocket in contact with 18 and 24. (B) and (D) depict the key residues anking 18 and 24 in the binding pocket.

Fig. 5. Compound 18 induced hyperacetylation of p53 (Ac p53) and a-tubulin (Ac a-tubulin) in (A) HepG2 and (B) MDA-MB-231 cells. Incubation times were 6 h and 24 h for HepG2
and MDA-MB-231 cells respectively. For the detection of p53, cell lysate protein was loaded at 20 mg (HepG2) and 100 mg (MDA-MB-231). For the detection of a-tubulin, cell lysate
protein was loaded at 1.5 mg (HepG2) and 2.0 mg (MDA-MB-231). (C) 18 induced apoptosis in MDA-MB-231 cells as seen from the increases in apoptotic marker proteins cleaved
caspase 3 and cleaved PARP (89 kDa) after 24 h incubation. Cell lysate protein was loaded at 100 mg. b-actin was the loading control.

selective inhibitors. SAR analysis revealed structure specic requirements for inhibition, of which substitution at the indole nitrogen (R2) and position 8 (R3) of the scaffold were deemed
important. Of the groups explored at the indole nitrogen, Sirt2 inhibition decreased in the order n-octyl > n-butyl > H, implicating
lipophilicity and steric bulk as essential features for optimal

activity. In the case of substitution at position 8 (R3), there was an


apparent preference for Series B which had a 40 -methylsulfonylphenyl ring at the position 8. Inhibitory potency was
affected to a lesser degree by R1 as inferred from the laxity in terms
of size and polarity permitted at this site. Nonetheless, small groups
were preferred at R1 to avoid exacerbating lipophilicity of the

152

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

resulting compounds. The more promising Sirt2 inhibitors were


found in Series B, notably 18 and 20 (IC50 3e4 mM) which were
more inhibitory than the Sirt2 selective inhibitor AGK2 (14 mM).
However, the moderately active members in series C (24,25 IC50
11e14 mM) should also be pursued for lead optimization as they
have more favorable solubility proles than series B. Kinetic analysis of Sirt2 inhibition by 18 revealed non-competitive inhibition
against both NAD and the acetylated substrate. Molecular docking
showed the functionalized scaffold of 18 (and other analogs)
positioned at both the NAD binding pocket (sites B, C) and the
acetylated substrate channel by stacking, van der Waals/hydrophobic and H bonding interactions. The functional relevance of
Sirt2 inhibition by 18 was corroborated by Western blot analyses
which showed that 18 increased levels of acetylated p53 and atubulin in MDA-MB-231 and HepG2 cell lysates. Levels of apoptotic
marker proteins (cleaved PARP, cleaved caspase 3) were also
elevated in MDA-MB-231 cell lysates, implicating induction of
apoptosis by 18. Further exploration of this scaffold may yield
useful compounds that could be used to interrogate the biology of
Sirt2.

4. Experimental
4.1. General conditions for organic synthesis
Reagents were purchased from SigmaeAldrich Chemical
(Singapore) or Alfa Aesar (Ward Hill, MA) and used without further
purication. Microwave reactions were carried out on the Biotage
Initiator Microwave Synthesizer. 1H NMR (300 MHz or 400 MHz)
and 13C NMR (75 MHz or 100 MHz) spectra were measured on a
Bruker Spectrospin 300 or 400 Ultrashield magnetic resonance
spectrometer. Chemical shifts (d) were reported in ppm and referenced to residual solvents: CDCl3 (d7.26), DMSO-d6 (d2.50), CD3OD
(d 3.31) (for 1H spectra) or CDCl3 (d 77.00), DMSO-d6 (d 39.43),
CD3OD (d 49.05) (for 13C spectra). Coupling constants (J) were reported in Hz. Reactions were monitored by TLC on Silica Gel 60
F254 (Merck). Column chromatography was carried out on Merck
Silica Gel 60 (0.04e0.06 mm). Mass spectra were recorded in positive ion mode using electro-spray ionization (ESI) (Applied Biosystem, Q-Trap 2000 LC/MS) or high-resolution LC-MS (IT TOF:
Waters-Micromass QTOF premier mass spectrometer). Synthetic
protocols and spectral data of synthesized intermediates/target
compounds are provided in Supplementary information (Sections
1e9). Purity of nal compounds were determined by reverse
phase HPLC and found to be  95% (Supplementary information,
Section 10).

4.2. Ethyl 8-bromo-3,4-dihydro-1H-pyrido[4,3-b]indole-2(5H)carboxylate (2)


The mixture of 4-bromophenylhydrazine hydrochloride (10.0 g,
44.7 mmol) and 1-carbethoxy-4-piperidone (7.66 g, 44.7 mmol) in
absolute EtOH (20 mL) was reuxed for 12 h. On cooling to room
temperature (25  C), the solid product was ltered, washed with
50% aqueous EtOH, and recrystallized from 95% EtOH to give 2 as an
off-white solid: 12.71 g, yield: 88.0%. 1H NMR (400 MHz, DMSO-d6)
d 11.23 (s, 1H), 7.60 (d, 1H, J 2.0 Hz), 7.26 (d, 1H, J 8.6 Hz),
7.14e7.11 (m, 1H), 4.54 (s, 2H), 4.08 (q, 2H, J 7.2 Hz), 3.72 (t, 2H,
J 5.6 Hz), 2.78 (t, 2H, J 5.6 Hz),1.20 (t, 3H, J 7.2 Hz). 13C NMR
(100 MHz, DMSO-d6) d 155.09, 134.48, 134.34, 126.83, 122.95,
119.59, 112.84, 111.06, 105.43, 60.85, 40.76, 23.03, 22.87, 14.59.

4.3. Ethyl 8-bromo-5-n-octyl-3,4-dihydro-1H-pyrido[4,3-b]indole2(5H)-carboxylate (3a)


To a mixture of 2 (2.39 g, 7.40 mmol) and sodium hydride (60%,
740 mg, 18.5 mmol, 2.5 equiv.) was added dry DMF (10 mL) under
N2 at 0  C. The reaction mixture was stirred at room temperature
for 15 min after which a solution of 1-bromooctane (1.72 g,
8.9 mmol, 1.2 equiv.) in dry DMF (5 mL) was added drop wise to the
stirred mixture at 0  C. The reaction mixture was stirred at room
temperature (z3 h), poured into an ice-water and extracted with
DCM (15  3 mL). The combined DCM layer was sequentially
washed with water (20  3 mL) and brine, dried (Na2SO4), ltered,
and concentrated under reduced pressure and the residue was
puried by silica gel column chromatography (hexane: EtOAc 3:1)
to give 3a as yellow oil (2.70 g, yield: 83.5%). 1H NMR (300 MHz,
CDCl3) d 7.57 (d, 1H, J 2.0 Hz), 7.25e7.22 (m, 1H), 7.15e7.12 (m,
1H), 4.65 (br, 2H), 4.20 (q, 2H, J 7.2 Hz), 3.97 (t, 2H, J 7.2 Hz), 3.87
(br, 2H), 2.80 (br, 2H), 1.74e1.67 (m, 2H), 1.33e1.25 (m, 13H), 0.87 (t,
3H, J 6.6 Hz). 13C NMR (75 MHz, CDCl3) d 155.96, 134.97, 128.98,
128.17, 126.71, 123.69, 120.23, 112.20, 110.53, 61.53, 53.39, 43.21,
41.08, 41.02, 31.71, 30.24, 29.22, 29.09, 26.95, 22.54, 14.70, 14.02.
4.4. 8-Bromo-5-n-octyl-2, 3, 4, 5-tetrahydro-1H-pyrido[4,3-b]
indole (4a)
To the solution of 3a (1.66 g, 3.82 mmol) in 15 mL ethanol was
added an aqueous solution of KOH (4.28 g,76.4 mmol, in 5 mL
water) and reuxed for 16 h. Ethanol was removed under reduced
pressure and the residual mixture was extracted with DCM
(15  3 mL). The combined DCM layer was sequentially washed
with water (15  3 mL) and brine, dried (Na2SO4), ltered, and
concentrated under reduced pressure to give a brown residue
which was puried by silica gel column chromatography (DCM:
methanol 25:1) to give 4a as a brown oil, 1.03 g, yield: 74.5%. 1H
NMR (400 MHz, CDCl3) d 7.49 (d, 1H, J 1.6 Hz), 7.21e7.18 (m, 1H),
7.11 (d, 1H, J 8.8 Hz), 4.00 (br, 2H), 3.93 (t, 2H, J 7.2 Hz), 3.24 (t,
2H, J 5.6 Hz), 2.73 (t, 2H, J 5.6 Hz), 2.68 (br, 1H), 1.68 (t, 2H,
J 7.2 Hz), 1.28e1.24 (m, 10H), 0.87 (t, 3H, J 6.8 Hz). 13C NMR
(100 MHz, CDCl3) d 134.69, 134.56, 127.12, 123.26, 120.15, 111.93,
110.33, 107.74, 43.17, 42.98, 41.91, 31.68, 30.19, 29.20, 29.06, 26.94,
23.19, 22.51, 13.98. MS (ESI), [MH], Calcd for C19H28BrN2, 363.1;
Found 363.3, 365.5.
4.5. 8-(40 -(Methylsulfonyl)phenyl)-5-n-octyl-2, 3, 4, 5-tetrahydro1H-pyrido[4,3-b]indole (18)
To the mixture of 4a (200 mg, 0.55 mmol), 4-(methanesulfonyl)
phenylboronic acid (132.6 mg, 0.66 mmol) and Pd(PPh3)4 (32.0 mg,
0.028 mmol) in 4 mL 1, 4-dioxane was added 0.5 mL aqueous solution of K2CO3 (229.0 mg, 1.66 mmol). The mixture was heated
(110  C, 0.5 h) in a microwave reactor with stirring. On cooling, the
solvent was evaporated and the resulting residue was extracted
with DCM (10 mL  3), the DCM layer was washed with brine, dried
(Na2SO4) and ltered. The residue obtained on removal of the solvent was puried by silica gel column chromatography (DCM:
methanol 20:1) to give the target compound. 18 was obtained as an
off-yellow oil (120 mg, yield: 49.6%). 1H NMR (400 MHz, CDCl3)
d 7.96 (d, 2H, J 8.4 Hz), 7.79 (d, 2H, J 8.4 Hz), 7.63 (br, 1H),
7.43e7.41 (m, 1H), 7.37e7.35 (m, 1H), 4.29 (br, 2H), 4.02 (t, 2H,
J 7.2 Hz), 3.44 (br, 2H), 3.09 (s, 3H), 2.96 (br, 2H), 1.76e1.71 (m,
2H), 1.31e1.25 (br, 10H), 0.87 (t, 3H, J 6.8 Hz). 13C NMR (100 MHz,
CDCl3) d 147.91, 137.92, 136.43, 133.45, 130.44, 127.81, 127.76, 127.70,
125.84, 120.93, 116.84, 109.82, 48.77, 48.13, 44.68, 43.38, 31.74,
30.32, 29.25, 29.12, 29.05, 27.04, 22.56, 14.03. MS (ESI), [MH],
calcd for C26H35N2O2S, 439.2, found, 439.2. HRMS (ESI), [MH],

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

calcd for C26H35N2O2S, 439.2419; found 439.2415.


4.6. In vitro assay for Sirt1 and Sirt2 activities
Sirt1 and Sirt2 activities were determined using the Sirt-Glo
Assay Kit (G6450, Promega, Madison, WI, USA) following manufacturer's instructions. Compounds were dissolved in DMSO and
tested over a 103 fold concentration range (nal DMSO concentration 2.5%). IC50 values were determined in triplicate and from 2 or
more independent experiments. Briey, assays were carried out on
a white wall 384-well plate with human recombinant Sirt1 (BMLSE239) and Sirt2 (BML-SE251) enzymes from Enzo Life Sciences
(NY, USA). The amount of enzyme used in the assay was determined
by measuring the signal-to-noise ratio of serially diluted enzyme in
Sirt-Glo Buffer solution. The concentration of enzyme corresponding to the mid-range linear portion of the signal to noise ratio
versus concentration plot was selected and was 0.1 unit Sirt1 per
well and 0.2 unit Sirt2 per well. Serially diluted inhibitor solutions
(10 mL in Sirt-Glo Buffer solution, with 1 mL of DMSO) were added
to each well followed by the sirtuin enzyme (10 mL in Sirt-Glo
Buffer solution). The plate was agitated (400 rpm, 30 min, 25  C,
Tecan Innite 200 plate reader) after which 20 mL of Sirt-Glo
Reagent solution was added per well, the contents mixed by
shaking (400 rpm) for another 30 min, 25  C, after which luminescence was read on the plate reader. Enzyme activity (%) was
measured by the following expression:

h
i 

LumCompound  Lum Blank
 
  100%
Enzyme Activity% 
Lum Control  Lum Blank
where Lum_Compound luminescence of wells containing
enzyme and test compound in vehicle (SIRT Glo Buffer solution),
Lum_Control luminescence of well containing enzyme only in
vehicle; and Lum_Blank luminescence of well containing vehicle
only. The IC50 of test compound was determined from the sigmoidal
curve obtained by plotting % enzyme activity versus logarithmic
concentration of test compound (GraphPad Prism, Version 5, San
Diego, USA). EX527 (Sigma LifeScience, MO, USA), AGK2 (Tocris
Bioscience, Bristol, UK) and nicotinamide (SIRT-Glo Assay Kit
provided, Promega, Madison, WI, USA) were used as positive controls. Representative dose response curves are presented in
Supplementary information (Section 11).
4.7. In vitro assay for evaluating inhibition of protease and
luciferase
Stock solutions (10 mM) of test compounds (9, 10, 12, 13, 18, 19,
20, 22, 24 and 26) were prepared in DMSO and serially diluted with
Sirt-Glo Buffer solution. 10 mL of the diluted solution was added to
each well in a whiteewall 384 well plate followed by 10 mL of SirtGlo Buffer solution and 20 mL of Sirt-Glo Reagent solution
containing 1 mM SIRT-Glo Control Substrate (non-acetylated
peptide, a gift from Promega, Madison, WI, USA). The nal concentration of test compound in the well was 10 mM and DMSO
content per well was 2.5% v/v. The contents of the wells were mixed
by shaking (400 rpm) for 30 min, 25  C, after which luminescence
was read on the plate reader. Luminescence (%) was measured by
the following expression:

i 

LumCompound  Lum Blank
 
  100%
Luminescence % 
Lum Control  Lum Blank
where Lum_Compound luminescence of wells containing non-

153

acetylated peptide, Sirt-Glo Reagent and test compound in


vehicle (SIRT Glo Buffer solution), Lum_Control luminescence
of wells containing non-acetylated peptide, Sirt-Glo Reagent in
vehicle Sirt-Glo Reagent and Lum_Blank luminescence of well
containing vehicle only. Results are given in Supplementary
information, Section 12.
4.8. Kinetic study of Sirt2 inhibition by 18
The kinetics of Sirt2 inhibition by 18 was investigated using a
deconstructed Sirt-Glo Kinetics Assay Kit which was a gift from
Promega (Madison, WI, USA). Human recombinant SIRT2 (BMLSE251, Enzo Life Sciences, NY, USA) (0.2 unit/well in 5 mL of assay
buffer) was aliquoted into each well of a 384-well plate which
contained test compound (18) at 0, 1.1, 3.3 or 10 mM in 5 mL of assay
buffer per well. The contents of the plate were incubated for 60 min
at 25  C with agitation (400 rpm). Final concentration of DMSO per
well was kept at 2.5%. At the same time, the Luciferin Detection
Reagent (LDR) solution was prepared by combining the acetylated
substrate (Z-QPK(Me)2K(Ac)-aminoluciferin) (4.12, 12.4, 37.0, 111.1
or 333.3 mM), NAD (1 mM) and developer reagent. The mixture
was incubated for 60 min at 37  C, after which aliquots (10 mL) were
dispensed to each well containing the 18-Sirt2 enzyme mixture.
Incubation was continued for another 30 min, 25  C with agitation
(400 rpm). Luminescence readings were then taken. The data from
these experiments were used to obtain the double-reciprocal plot
from which Km of Sirt2 for acetylated substrate was obtained. The
slope and intercept of the lines obtained in presence of 18 were
replotted against concentration of 18 according to Equation (1) to
give apparent Ki of 18 (Supplementary information, Section 13,
Fig. S2). The Km of Sirt2 for acetylated peptide was also obtained
from EadieeHofstee plot (Supplementary information, Section 13,
Fig. S3) which is based on Equation (2) and involved plotting reaction rate against rate/(substrate concentration).

Slope Km =Vmax Km =Vmax I=Ki

(1)

Rate Km Rate=Subtrate Vmax

(2)

The above procedure was repeated at a xed concentration of


the acetylated substrate (38.5 mM) and varying amounts of NAD
(185, 556, 1667, 5000 mM). The resulting double reciprocal plot gave
the Km of Sirt2 for NAD and the secondary plots gave apparent Ki
of 18 with NAD was substrate. Plots and kinetic parameters Km, Ki,
Vmax were obtained with GraphPad Prism (Version 5, San Diego,
CA). From the double reciprocal plots, the Km values of Sirt2 for
acetylated substrate and NAD were found to be 38.0 2.0 mM and
1317 208 mM respectively. Km values from EadieeHofstee plots of
the same data were comparable at 29.9 1.6 mM (acetylated substrate) and 1298 218 mM (NAD) (Supplementary information,
Section 13, Fig. S3).
4.9. Molecular docking
The human Sirt2 enzyme was retrieved from the RCSB protein
data bank (PDB 3ZGV) [12]. Water molecules were removed and the
monomeric enzyme was processed for docking using LigX in Molecular Operating Environment (MOE, version 2011, Chemical
Computing Group, Montreal, Canada). The structures of the test
compounds were separately prepared for docking on MOE. Docking
was carried out on GOLD v 5.2 (Cambridge Crystallographic Data
Centre Software Ltd, Cambridge, UK) with default genetic algorithm
settings. The binding pocket was dened by the atoms within 10
radius of the co-crystallized ligand (ADP-ribose). Docking was
carried out without the reference ligand. GOLD uses a genetic

154

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

algorithm for docking exible ligands into the binding pocket to


explore the full range of ligand conformational exibility [22]. The
GOLD Score was used as the tness function for selection of the best
docked conformations of test compounds in the binding pocket. For
each molecule, the top 5 docked conformations were retained and
visualized on MOE.
4.10. Assessment of aggregation tendency by light scattering
Stock solutions (10 mM) of test compounds prepared in DMSO,
diluted to 1 mM with DMSO and then serially diluted with potassium phosphate buffer (5 mM, pH 7.4, preltered before use) to give
nal concentrations of 1 mM and 10 mM. Final concentration of
DMSO was 1% v/v. Measurements were carried out on the Malvern
Instrument Zetasizer Nano ZS system equipped with a 4 mW
HeeNe laser at 633 nm and detector angle of 90 . Three or more
determinations of derived count rates (kilocounts per second, kcps)
were obtained from each concentration of test compound, using
two separately prepared stock solutions. Data collection was carried out using the software supplied with the instrument. Results
are represented as mean standard deviation. The positive control
was benzyl benzoate which gave a count rate of 1148kcps 37
(250 mM) and 76 5 (25 mM). The vehicle (phosphate buffer, 1%
DMSO) gave a reading of 15.5 0.4. The results are presented in
Supplementary information, Section 14.
4.11. Determination of cell growth inhibition
Human breast cancer MDA-MB-231, hepatocellular carcinoma
HepG2 and human lung broblast IMR90 cells were purchased
from ATCC (Rockville, MD). All cells were grown in DMEM (Invitrogen, High Glucose.) at 37  C, 5% CO2. DMEM was supplemented
with 10% fetal bovine serum (Hyclone, heat treated at 65  C for
30 min before use), 50 units/L penicillin (Gibco) and 50 mg/mL
streptomycin (Gibco). Cells were subcultured at 85e90% conuency
and used within 15 passages. Cell viability was assessed using
CellTitre 96 Aqueous One Solution (Promega, Madison, WI) containing the tetrazolium salt 3-(4,5-dimethylthiazol-2-yl)-5-(3carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS).
Seeding densities were 2500 cells/well in the 96-well plates. Cells
were grown in media (10% FBS) in 96-well plates for 24 h, at 37  C,
5% CO2 after which aliquots of 18 (stock solution in DMSO, diluted
with media containing 10% FBS) were added to each well and the
plates were incubated for 72 h, at 37  C, 5% CO2. Final concentration
of DMSO per well was 0.5% v/v. At the end of the incubation period,
10 mL of the MTS solution was added to each well, the plates were
incubated for 4 h in the dark after which absorbance readings were
read at 490 nm (Tecan Innite M200 Microplate reader). Cell
viability was determined from the expression:

 
Absorbancecellscpd  Absorbancecpd
Cell viability %
Absorbancecellsvc  Absorbancevc
 100%
where Absorbancecellscpd absorbance of wells containing cells
and test compound in vehicle (media 0.5% DMSO),
Absorbancecellsvc absorbance of wells containing cells in vehicle
(vc) only; Absorbancevc absorbance of wells containing vehicle;
Absorbancecpd absorbance of wells containing test compound.
The % viability readings were plotted against log concentration on
GraphPad Prism (Version 5.0, San Diego, CA) to give a sigmoidal
curve from which IC50 (concentration required to reduce viability
by 50% compared to control/untreated cells) was obtained. The plot
was constrained to 0 and 100%. At least 3 separate

determinations of IC50 were determined. Representative dose


response curves are presented in Supplementary information,
Section 15.
4.12. Western blot
Rabbit monoclonal antibodies to PARP and caspase 3, mouse
monoclonal antibodies to p53, goat anti-rabbit-horse radish
peroxidase (HRP) conjugate and goat anti-mouse HRP were from
Santa Cruz Biotechnology (CA, USA). Rabbit monoclonal antibodies
to cleaved caspase 3 and acetylated p53 (K382) were purchased
from Cell Signaling (MA, USA). Mouse monoclonal antibodies to
acetylated a-tubulin and a-tubulin were from Sigma Aldrich (Sigma
Lifescience, MO, USA). Mouse monoclonal antibodies to b-actin
were purchased from Invitrogen Life Technologies (CA, USA).
HepG2 and MDA-MB-231 cells were seeded at a cell density of in a
10 cm petri dish at 50,000 cells/mL (total 10 mL in DMEM, high
glucose, Invitrogen) containing a known concentration of test
compound. Final concentration of DMSO in the ask was maintained at 0.5% v/v. Treated cells were incubated at 37  C in a humidied 5% CO2 atmosphere for 6 h (for HepG2 cells) or 24 h (for
MDA-MB-231 cells) after which they were transferred to a 15 mL
Falcon tube, pelleted at 500 g (5 min) and washed twice with icecold 1x PBS. The cells were transferred to a microcentrifuge tube
containing 70 mL lysis buffer (20 mM Tris pH 7.4, 250 mM NaCl,
2 mM EDTA pH 8.0, 0.1% Triton X-100, 0.01 mg/ml aprotinin,
0.005 mg/ml leupeptin, 0.4 mM PMSF, and 4 mM NaVO4) and
incubated at 0  C for at least 20 min. After the incubation, the lysates were spun at 13,300 g for 10 min to remove cell debris. An
aliquot (3 mL) was retained for protein determination (Bradford
Protein Assay Kit, Bio-Rad Laboratories Inc, CA, USA) while the
remaining supernatant was diluted with 4  SDS solution (0.2 M
Tris pH 6.8, 0.28 M SDS, 40% v/v glycerol, 0.59 M b-mercaptoethanol, 50 mM EDTA, 1.1 mM bromophenol blue) to give 1  SDS solution which was then deactivated at 100  C, 5 min and
subsequently stored for no more than 1 week at 80  C. Cell lysates
were separated on the SDS-PAGE Bio-Rad Mini-Protean II system
(Bio-Rad Laboratories Inc, CA, USA). Membranes were blocked with
blocking buffer for at least 60 min at 25  C and then incubated with
primary antibodies at appropriate dilutions in 2.5% BSA in TBST
(50 mM Tris, 150 mM NaCl, 0.1% Tween 20) solution overnight
(4  C). Membranes were than washed thrice with TBST (10 min per
wash), incubated for 1 h with secondary antibodies in TBST at 25  C
followed by washing in TBST (thrice, 10 min per wash). The
immunoreactive bands were detected by the ECL reagent (GE
Healthcare, Little Chalfont, UK) using Bio-Rad Universal Hood II Gel
Doc. If needed, the membranes were stripped with stripping buffer,
blocked with blocking buffer for 30 min 25  C, and re-probed with
other antibodies. Western blotting for each compound at stated
concentrations were repeated at least 3 times.
4.13. Determination of solubility
The solubilities of 10, 12, 18, 19, 20 and 24 were determined in
phosphate buffer pH 7.4 by a turbimetric method [16]. Briey, stock
solutions of test compound were prepared in DMSO and aliquots
were added to a 96-well plate containing phosphate buffer in each
well to give a range of concentrations (1e50 mM). DMSO concentration was kept at 1% v/v. After shaking the plate for 30 min at
25  C, absorbance readings were taken at 620 nm. Absorbance
readings were plotted against concentration. When the concentration exceeded compound solubility, the occurrence of precipitation led to an increase in absorbance. Solubility was then assigned
to the range of concentrations where a two-fold or greater increase
in absorbance was observed from the absorbance-concentration

T. Yang et al. / European Journal of Medicinal Chemistry 92 (2015) 145e155

plot. Results were obtained from 3 separate determinations.


Acknowledgments
This work was supported by grants from the Biomedical
Research Council (10/1/21/19/664) and Ministry of Health Faculty
Research Grant (R148000171112) to M.L.G. We thank Promega for
providing the SIRT-Glo Kinetics Assay Kit and SIRT-Glo Control
Substrate, and Dr. Andrew Niles for helpful advice and discussion
on the assay details.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2014.12.027.
References
[1] C. Sebasti
an, F.K. Satterstrom, M.C. Haigis, R. Mostoslavsky, From sirtuin
biology to human diseases: an update, J. Biol. Chem. 287 (2012)
42444e42452.
[2] M. Roth, W.Y. Chen, Sorting out functions of sirtuins in cancer, Oncogene 33
(2014) 1609e1620.
[3] J.S. Disch, G. Evindar, C.H. Chiu, C.A. Blum, H. Dai, L. Jin, E. Schuman, K.E. Lind,
S.L. Belyanskaya, J. Deng, F. Coppo, L. Aquilani, T.L. Graybill, J.W. Cuozzo,
S. Lavu, C. Mao, G.P. Vlasuk, R.B. Perni, Discovery of thieno[3,2-d]pyrimidine6-carboxamides as potent inhibitors of SIRT1, SIRT2, and SIRT3, J. Med. Chem.
56 (2013) 3666e3679.
[4] K. Yamagata, Y. Goto, H. Nishimasu, J. Morimoto, R. Ishitani, N. Dohmae,
N. Takeda, R. Nagai, I. Komuro, H. Suga, O. Nureki, Structural basis for potent
inhibition of SIRT2 deacetylase by a macrocyclic peptide inducing dynamic
structural change, Structure 22 (2014) 345e352.
[5] M. Friden-Saxin, T. Seifert, M.R. Landergren, T. Suuronen, M. Lahtela-Kakkonen, E.M. Jarho, K. Luthman, Synthesis and evaluation of substituted chroman4-one and chromone derivatives as sirtuin 2-selective inhibitors, J. Med.
Chem. 55 (2012) 7104e7113.
[6] D. Rotili, D. Tarantino, A. Nebbioso, C. Paolini, C. Huidobro, E. Lara, P. Mellini,
A. Lenoci, R. Pezzi, G. Botta, M. Lahtela-Kakkonen, A. Poso, C. Steinkuhler,
P. Gallinari, R. De Maria, M. Fraga, M. Esteller, L. Altucci, A. Mai, Discovery of
salermide-related sirtuin inhibitors: binding mode studies and antiproliferative effects in cancer cells including cancer stem cells, J. Med. Chem.
55 (2012) 10937e10947.
[7] S.S. Mahajan, M. Scian, S. Sripathy, J. Posakony, U. Lao, T.K. Loe, V. Leko,
A. Thalhofer, A.D. Schuler, A. Bedalov, J.A. Simon, Development of pyrazolone
and isoxazol-5-one cambinol analogues as sirtuin inhibitors, J. Med. Chem. 57
(2014) 3283e3294.

155

[8] J. Trapp, A. Jochum, R. Meier, L. Saunders, B. Marshall, C. Kunick, E. Verdin,


P. Goekjian, W. Sippl, M. Jung, Adenosine mimetics as inhibitors of NADdependent histone deacetylases, from kinase to sirtuin inhibition, J. Med.
Chem. 49 (2006) 7307e7316.
[9] B. Suenkel, F. Fischer, C. Steegborn, Inhibition of the human deacylase sirtuin 5
by the indole GW5074, Bioorg. Med. Chem. Lett. 23 (2013) 143e146.
[10] A.D. Napper, J. Hixon, T. McDonagh, K. Keavey, J.F. Pons, J. Barker, W.T. Yau,
P. Amouzegh, A. Flegg, E. Hamelin, R.J. Thomas, M. Kates, S. Jones, M.A. Navia,
J.O. Saunders, P.S. DiStefano, R. Curtis, Discovery of indoles as potent and
selective inhibitors of the deacetylase SIRT1, J. Med. Chem. 48 (2005)
8045e8054.
[11] P. Mellini, V. Carafa, B. Di Rienzo, D. Rotili, D. De Vita, R. Cirilli, B. Gallinella,
D.P. Provvisiero, S. Di Maro, E. Novellino, L. Altucci, A. Mai, Carprofen analogues as sirtuin inhibitors: enzyme and cellular studies, ChemMedChem 7
(2012) 1905e1908.
[12] S. Moniot, M. Schutkowski, C. Steegborn, Crystal structure analysis of human
Sirt2 and its ADP-ribose complex, J. Struct. Biol. 182 (2013) 136e143.
[13] J. Seidler, S.L. McGovern, T.N. Doman, B.K. Shoichet, Identication and prediction of promiscuous aggregating inhibitors among known drugs, J. Med.
Chem. 46 (2003) 4477e4486.
[14] S.L. McGovern, B.T. Helfand, B. Feng, B.K. Shoichet, A specic mechanism for
nonspecic inhibition, J. Med. Chem. 46 (2003) 4265e4272.
[15] T.F. Outeiro, E. Kontopoulos, S.M. Altmann, I. Kufareva, K.E. Strathearn,
A.M. Amore, C.B. Volk, M.M. Maxwell, J.C. Rochet, P.J. McLean, A.B. Young,
R. Abagyan, M.B. Feany, B.T. Hyman, A.G. Kazantsev, Sirtuin 2 inhibitors rescue
alpha-synuclein-mediated toxicity in models of Parkinson's disease, Science
317 (2007) 516e519.
[16] J.H.M. Lange, Hein K. A.C. Coolen, M.A.W. van der Neut, A.J.M. Borst, B. Stork,
P.C. Verveer, C.G. Kruse, Design, synthesis, biological properties and molecular
modeling investigations of novel tacrine derivatives with a combination of
acetylcholinesterase inhibition and cannabinoid CB1 receptor antagonism,
J. Med. Chem. 53 (2010) 1338e1346.
[17] R.A. Copeland, Enzyme reaction mechanisms, in: Evaluation of Enzyme Inhibitors in Drug Discovery, Wiley-Interscience, New Jersey, 2013, pp. 48e81.
[18] X.Q. Wu, J. Wang, Z.R. Lu, H.M. Tang, D. Park, S.H. Oh, J. Bhak, L. Shi, Y.D. Park,
F. Zou, Alpha-glucosidase folding during urea denaturation: enzyme kinetics
and computational prediction, Appl. Biochem. Biotechnol. 160 (2010)
1341e1355.
[19] M. Gertz, F. Fischer, G.T.T. Nguyen, M. Lakshminarasimhan, M. Schutkowski,
M. Weyand, C. Steegborn, EX527 inhibits sirtuins by exploiting their unique
NAD dependent deacetylation mechanism, Proc. Nat. Acad. Sci. U S A 110
(2013) E2772eE2781.
[20] M. Li, J. Luo, C.L. Brooks, W. Gu, Acetylation of p53 inhibits its ubiquitination
by Mdm2, J. Biol. Chem. 277 (2002) 50607e50611.
[21] H. Yamaguchi, N.T. Woods, L.G. Piluso, H.H. Lee, J. Chen, K.N. Bhalla,
A. Monteiro, X. Liu, M.C. Hung, H.G. Wang, p53 acetylation is crucial for its
transcription-independent proapoptotic functions, J. Biol. Chem. 284 (2009)
11171e11183.
[22] M.L. Verdonk, J.C. Cole, M.J. Hartshorn, C.W. Murray, R.D. Taylor, Improved
protein-ligand docking using GOLD, Proteins 52 (2003) 609e623.

You might also like