Synthesis and biological activity of splitomicin analogs targeted at human NAD+-dependent histone deacetylases (sirtuins)

Article abstract of DOI:10.1016/j.bmc.2011.01.026

Small molecules interfering with posttranslational modification of histones are of interest as tools to study epigenetic regulation of gene transcription. Specifically, drugs that interfere with histone deacetylation could be useful to induce differentiation, growth arrest as well as apoptotic cell death in tumor cells. One class of histone deacetylases is known as sirtuins some of which (Saccharomyces cerevisiae Sir2) are for example inhibited by the lactone splitomicin leading to telomeric silencing in yeast. However, splitomicin is only a micromolar inhibitor of yeast Sir2 and does not inhibit human subtypes and the lactone is prone to hydrolytic ring opening. In preliminary SAR-studies, splitomicin analogs lacking this hydrolytically labile ring were described as inactive while the naphthalene moiety could successfully be replaced by smaller aromatic rings in a fragment-like dihydrocoumarin. Here we report the synthesis and biological activity of a series of hydrolytically stable analogs with activity against human SIRT1 and 2. These comparatively small compounds characterized by high ligand efficiency are used as a starting point toward the development of specific inhibitors of histone deacetylases from the class of sirtuins.

Full text of DOI:10.1016/j.bmc.2011.01.026

Bioorganic & Medicinal Chemistry 19 (2011) 3669–3677
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and biological activity of splitomicin analogs targeted at human
NAD+-dependent histone deacetylases (sirtuins)
Marcus Freitag a, Jörg Schemies b, Tim Larsen a, Khattab El Gaghlab a, Felix Schulz a, Tobias Rumpf b,
Manfred Jung b, Andreas Link a,
⇑
a Institute of Pharmacy, Ernst-Moritz-Arndt-University Greifswald, Friedrich-Ludwig-Jahn-Str. 17, 17489 Greifswald, Germany
b Institute for Pharmaceutical Sciences, Albert-Ludwigs-University Freiburg, Albertstr. 25, 79104 Freiburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Small molecules interfering with posttranslational modification of histones are of interest as tools to
study epigenetic regulation of gene transcription. Specifically, drugs that interfere with histone deacety-
lation could be useful to induce differentiation, growth arrest as well as apoptotic cell death in tumor
cells. One class of histone deacetylases is known as sirtuins some of which (Saccharomyces cerevisiae
Sir2) are for example inhibited by the lactone splitomicin leading to telomeric silencing in yeast. How-
ever, splitomicin is only a micromolar inhibitor of yeast Sir2 and does not inhibit human subtypes and
the lactone is prone to hydrolytic ring opening. In preliminary SAR-studies, splitomicin analogs lacking
this hydrolytically labile ring were described as inactive while the naphthalene moiety could successfully
be replaced by smaller aromatic rings in a fragment-like dihydrocoumarin. Here we report the synthesis
and biological activity of a series of hydrolytically stable analogs with activity against human SIRT1 and 2.
These comparatively small compounds characterized by high ligand efficiency are used as a starting point
toward the development of specific inhibitors of histone deacetylases from the class of sirtuins.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 28 September 2010
Revised 7 January 2011
Accepted 13 January 2011
Available online 22 January 2011
Keywords:
Benzimidazole
Histone deacetylases
Ligand efficiency
Sirtuins
Splitomicin
1. Introduction
ciated with the pathogenesis of cancer,5 HIV,6 and metabolic,7
and neurological8 diseases. Within this context, sirtuin inhibition
The lactone splitomicin (1) was purified from a mixture of com-
pounds (a test sample named NSC-112546) containing cambinol as
major constituent and reported to be a micromolar inhibitor of
might offer an important novel starting point for therapeutic inter-
vention. A first example of a potent and selective SIRT1 inhibitor
(IC50 = 98 nM), EX-527 (19) has entered phase I clinical trials in
20105 as a drug for the treatment of Huntington’s disease. The sir-
tuin family is thought to contain key players in the control of lon-
gevity in response to caloric restriction in organisms such as yeast,9
worms, flies, and possibly mammals, thus both inhibitors10 as well
as activators are of potential interest as future drugs. The question
whether sirtuins should be inhibited or activated is far from being
settled, and novel sirtuin inhibitors with varying selectivity are of
research interest in order to establish a set of well-defined phar-
macological tools for further studies in this direction. Optimally,
research could lead to the discovery of arrays of highly selective
sirtuin inhibitors that could be classified as molecular probes.
While 1 is commercially available as a tool for experimental
studies, it does not inhibit human sirtuin subtypes such as SIRT1
and is prone to hydrolysis11 in aqueous media. The resulting ring
open form 2 is characterized by a profound loss of activity. The
straightforward replacement of the labile lactone functionality in
1 by a lactam function in analog 3 as depicted in Figure 1 led to
a significant loss of activity11 in the available assays. In preliminary
structure–activity relationship (SAR) studies on a limited number
of analogs, these results were interpreted in such a way that the
yeast Sir2p1 (IC50 = 60 M), an NAD+-dependent deacetylase re-
l
quired for chromatin-dependent silencing in yeast. This cell-per-
meable molecule is comparatively small in terms of the
1
number of rings and a molar mass (i.e., below 200) and thus repre-
sents an attractive lead for novel compounds intended as modula-
tors of histone deacetylase (HDAC) activity. Three of the four
classes of human HDACs known to date consist of zinc-dependent
amidohydrolases that cleave off acetyl groups from acetyl-lysine
residues in histones2 and various nonhistone proteins. The seven
members of the non-zinc dependent class III HDACs named sirtuins
for their homology to yeast silent information regulator Sir2p,1 uti-
lize nicotinamide adenine dinucleotide (NAD+) for their catalytic
activity. Recently, small molecule inhibitors of sirtuins have be-
come valuable tools3 for the investigation of sirtuin function4 and
have provided insight into their role in many physiological and
pathological processes. Specifically, sirtuin activity has been asso-
⇑
Corresponding author. Tel.: +49 03834 86 4893; fax: +49 03834 86 4895.
E-mail addresses: link@pharmazie.uni-greifswald.de, link@uni-greifswald.de
(A. Link).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.01.026

3670
M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
SH
O
O
S
SH
NH
COOH
OH
HN
HN
N
O
NH
N
NH
1
2
3
6a
6b
6c
Figure 1. Splitomicin (1), inactive hydrolyzed form 2 and inactivestabilized analog 3.
lactone ring was an essential part of the pharmacophoric structure.
In order to verify this notion and with the aim to identify a novel
lead structure with good stability and activity against human sir-
tuin subtypes, it appeared rewarding to synthesize and test addi-
tional analogs of
1 with other stable heterocyclic rings as
replacements for the dihydropyran or dihydropyridine ring.
2. Chemistry
With the intention to identify stable analogs of lead compound
1, the 2-oxo-substituted dihydropyran ring was formally replaced
by a 2-oxo substituted dihydroimidazole ring. The selection of
dihydroimidazole was based on the fact that this structural motif
is present in many biologically active compounds and can be called
a privileged structure in medicinal chemistry. In this way, the lac-
tone ring was replaced by hydrolytically stable cyclic urea func-
tionality resulting in compound 412 that had already been
prepared by Diels in 1922. Therefore, naphthalene-1,2-diamine
was treated with triphosgene as described. In addition, as de-
scribed before, compound 4 could easily be transformed regiose-
lectively into the nitro derivative 513 by nitration with nitric acid
in glacial acetic acid (Scheme 1) in good yield.
The regioselectivity of the above reaction enabled the straight-
forward purification of a single product. The structure of the
known single major product 5 could be verified by HSQC- and
HMBC-NMR experiments. Cyclization of naphthalene-1,2-diamine
by means of CS2 yielded related thiourea 614 in a mixture of tau-
tomers (6a–c, Fig. 2).
Figure 2. Known 1,3-dihydrobenzo[e]benzimidazole-2-thione (green) as thioenol
tautomer 6a as overlay on splitomicin (1, gray).
S
X
O
HN
HN
R
NH
NH
O
N
COOH
7
8
9a-c
Figure 3. Active dihydrocoumarin 7 and general structure of analogs of type 8 and
9 (9a X = O, R = Cl, 9b X = S, R = Cl, 9c X = S, R = OCH3).
R2
X
It was evident from preliminary SAR investigations that the
annelated naphthalene moiety in 1 could be replaced by a benzene
ring11 leading to active dihydrocoumarin 7, thus biological evalua-
tion of smaller splitomicin analogs 8 and 9a–c (Fig. 3) were envi-
sioned as well.
N
NH
R1
The known compounds 815 and 9a–c16,17 were prepared accord-
ing to procedures described in the literature.
10a-h
X
R1
R2
Another known aspect of SAR within the splitomicin class of
compounds is the fact that substituents such as aryl rings on both
of the methylene groups of the lactone ring can contribute to im-
proved activity6,18 in selected assays. Therefore an additional series
of N-substituted fragment-like compounds 10a–h ( Fig. 4) were
prepared. The introduction of residues R2 via nucleophilic aromatic
substitution was aided by electron withdrawing substituents R1.
The seven novel compounds 10a–g were prepared from San-
ger’s reagent (11, X = F) with five aliphatic amines19–21 yielding
anilines 12a–e. Subsequent regioselective reduction of the nitro
O
O
S
S
S
O
NO2
NO2
NO2
NO2
NO2
NO2
2-propyl
cyclohexyl
a
b
c
d
e
f
2-aminoethyl HCl
3-(Boc-amino)propyl
3-aminopropyl HCl
2,4-dimethoxybenzyl
2,4-dimethoxybenzyl
1,3-benzodioxol-5-yl
S
O
NO2
CF3
g
h
Figure 4. Structure of N-substituted analogs 10a–h.
O
O
group adjacent to the newly introduced substituent of these inter-
mediates gave the o-dianilines 13a–e as outlined in Scheme 2. The
necessary regioselective reduction can be achieved by using ba-
ker’s yeast22 and many other reducing agents. In our hands, the
established method with NaHCO3/Na2S23 worked best. The boiling
methanolic solutions of 12a–e were treated with an aqueous equi-
molar solution of NaHCO3 and Na2S and refluxed for several hours.
The resulting 4-nitro substituted 1,2-dianilines 13a, b, and e could
be cyclized by means of triphosgene to form the cyclic ureas 10a, b,
HN
HN
NH
NH
a
NO2
4
5
Scheme 1. Regioselective nitration of compound 4 to nitro-derivative 5. (a) HNO3/
AcOH, 50 °C, 76% yield.

M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
3671
R
lated compounds.6,18 The annelated naphthalene moiety in 1 could
be omitted to yield active derivative 7 as mentioned before and
thus we anticipated that an additional gain of affinity might be best
X
NH
NO2
NO2
a
obtained by substituents in
position.
a different hitherto unexplored
NO2
NO2
X = F or Cl
R
11
12a-e
3. Results
The seven known splitomicin analogs 4–6 and 8–9c were eval-
uated for sirtuin inhibition together with the ten novel N-substi-
tuted derivatives 10a–h. The biological activity was determined
by using a homogeneous microplate reader-based nonisotopic his-
tone deacetylase activity assay24 as described before. The results of
the sirtuin assay with human recombinant proteins produced in
Escherichia coli are shown in Table 1. All compounds were tested
2-propyl
cyclohexyl
2-(Boc-amino)ethyl
3-(Boc-amino)propyl
a
b
c
d
e
2,4-dimethoxybenzyl
b
at concentrations of 25 and 50 lM. For those that showed a signif-
icant inhibition, IC50 values were determined.
R
NH
c
Neugebauer et al. reported the synthesis and biological evalua-
tion of a series of b-arylsplitomicins18 like 17 (Fig. 5). An investiga-
tion of the structure–activity relationships led to the discovery of
SIRT2 inhibitors with single digit micromolar potency, and a selec-
tivity for SIRT2 over SIRT1. In order to establish the binding mode
NH2
10a, b, f
d
10c, d, e, g
NO2
13a-e
Table 1
Scheme 2. Synthesis of N-substituted analogs 10a–g. Reagents and conditions: (a)
R-NH2, EtOH or acetone/H2O/NaHCO3 125 °C or reflux; (b) MeOH, reflux, NaHCO3,
Na2SÁ9H2O, 5 h; (c) CH2Cl2, NaOH, H2O, triphosgene, 12 h; (d) CS2, Et3N, EtOH,
reflux, 4–5 h.
Comparison of the biological activity of splitomicin analogs (IC50 SE [lM] or
inhibition @ 50
lM).
Entry
SIRT1
SIRT2
SIRT3
Ro 31-8220
4
5
6
8
9a
9b
9c
10a
10b
10c
10d
10e
10f
10g
10h
12e
13e
15
3.5 0.4
10.6%
n.i.
9.6 1.0
29.3 2.9
n.i.
>100
21.4 1.0
n.i.
0.8 0.2
46.7%
12.0 2.1
10.0 0.7
38.3 3.7
n.i.
26.9 6.4
16.3 2.0
n.i.
34.4 11.0
5.0 1.1
n.i.
n.i.
n.i.
43.1%
16.5%
14.8%
n.i.
41.4%
3.7 0.2
24.6%
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i
n.i.
n.i.
n.i.
n.i.
n.i.
and f. The thioureas 10d and g were obtained from the correspond-
ing 4-nitro substituted 1,2-dianilines 13c–e after cyclization with
CS2. Compounds 10c and e were prepared via their Boc-protected
precursors. The Boc-protected precursor 10d of compound 10e
was included in the biological evaluation.
Compound 10h bearing a trifluoromethyl group in position 5
was obtained from 1-fluoro-2-nitro-4-(trifluoromethyl)benzene
(14). In this case, an aromatic amine, 1,3-benzodioxol-5-ylamine
was selected for the introduction of an aromatic substituent in
the iminodibenzyl derivative 15. The absence of a second nitro
group in position 4 of this intermediate allowed for the reduction
of the nitro group in 15 to form commercially available 1,2-diani-
line 16 via straightforward catalytic hydrogenation (Scheme 3).
The six different amines were selected to introduce flexible al-
kyl chains with or without polar substituents as well as space fill-
ing aliphatic and aromatic rings into the molecular scaffold in
order to generate additional interaction motifs. The resulting sub-
stitution pattern of compounds 10a–h meets the hypothesis that
substituents on one of the two N-atoms might occupy the same
space as the beneficial substituents on the methylene groups of re-
n.i.
7.7 1.2
10.0%
n.i.
n.i.
14.4%
n.i.
n.i.
14.1%
18.8%
12.0%
n.i.
Reference inhibitor Ro 31-8220 was tested in the same system.3,25
n.i.: no inhibiton [<10% @ 50 lM].
H3 C
O
O
O
O
O
O
O
O
F
NH
NH
NO2
NO2
NH2
H3 C
Br
a
b
17
19
18
H3 C
CF3
CF3
CF3
O
14
15
16
NH
c
H3 C
10h
Scheme 3. Synthesis of N-substituted derivative 10h. Reagents and conditions: (a)
3,4-(methylenedioxy)aniline, neat, 80 °C; (b) Pd/C 10%, H2, EtOH; (c) CH2Cl2, NaOH,
H2O, triphosgene, 12 h.
Figure 5. Labile lactone-based racemic splitomicin (1) analogs 17 and 18 (HR73)
and stable lactam 19 reported by Neugebauer et al.18 and Pagans et al.6

3672
M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
of these compounds and to rationalize the SAR of this series of
SIRT2 inhibitors, docking simulations were performed.4 The
docked b-arylsplitomicin 17 revealed a binding mode that includes
a hydrogen bond to a water network around Gln167, similar to the
binding mode26,27 of the major constituent of NSC-112546, camb-
inol. The b-phenyl substituent of the (R)-enantiomer of 17 fits into
a hydrophobic channel and is sandwiched between Phe119 and
His187. The (R)-enantiomer of 17 was shown to be a potent SIRT2
inhibitor with an IC50 of 1.0 lM but the (S)-enantiomer displayed
only weak SIRT2 inhibition at high concentrations. While the
stereochemistry at the substituted b-carbon atom is of great
importance, the shift of the substituent from the b- to the a-carbon
as well led to the identification of similarly active sirtuin inhibitor
named HR736 (18). However, the notorious limitation of the insta-
bility of the lactone ring could be overcome neither by a- nor by b-
arylsplitomicins, but within the series of analogous compounds, a
stable lactam (19) could be shown to be equally active.18 Docking
studies of these compounds have given insight into the unknown
binding mode of, for example, compound 6 in its three tautomeric
forms in the nicotinamide binding pocket of human SIRT2, as sug-
gested in Figure 6.
While tautomers b and c of compound 6 bind in virtually iden-
tical position (the magenta colored skeleton of 6c is almost com-
pletely overlapped by light blue molecule 6b, only the thiol
hydrogen atom protrudes from underneath the thion sulfur of
6b), a preferred docking solution for 6a suggests that this enethiol
tautomer with a hydrogen atom on N1 is shifted significantly to-
wards Leu134 and Phe119.
The replacement of the six-membered tetrahydropyran ring of
splitomicins 1, 17, and 18 by the five-membered imidazole leads
to a conformationally closely related smaller ring system with cyc-
lic urea structure. While the resulting ureas are less prone to
hydrolysis, the N-substituents might interact with the molecular
target similar to splitomicins 17 and 18 so that substituents on
the N1 atom of the imidazole ring possibly can fit into the same
gap between Phe119 and His187. The cyclic ureas or thioureas
and their tautomers also could interact by a hydrogen bond net-
work with conserved water molecules in the region of Gln167
and form hydrogen bonds with His187. The rationale for the syn-
thesis of compounds 10a–h was the fact, that modifications (a–d)
could be combined in molecules of this type. Various substituents
could easily be introduced into position 1 of the ring system by
high yielding nucleophilic aromatic substitution. In contrast to
Figure 7. Predicted binding mode and molecular interactions for novel lead
structure 10c in the thione form.
X-ray structure of sirtuins with the compounds 5, 6 or 10a–h exists
to date, the proposed binding mode and molecular interactions de-
picted in figure 7 remain speculative. According to docking simula-
tions the binding mode of 1, 6, and 10c could be similar in terms of
the orientation of the carbocyclic part of the scaffold in the channel
between Phe96 and Phe119. The water network between Gln167,
Asn168 and the lactone or thiourea moiety of 1 or 6, respectively,
could be strengthened by the polar side chain of 10c as suggested
below. However, multiple binding modes are possible and 10c
could as well be flipped with the core aromatic ring system and
the N1 substituents changing positions as recently reported for a
biphenyl substituted thiobarbiturate.28
molecular framework of
a- or b-substituted derivatives 17 and
18 of 1, no stereocenters are formed by this operation which is
an advantage with respect to the ease of preparation. Because no
Due to the fact that the N1 atom is blocked by a substituent,
only two tautomeric forms of 10c are possible.
4. Discussion
Starting from the cell-permeable lead structure 1 that is inac-
tive against human sirtuins and prone to hydrolytic lactone ring
opening leading to a short half-life particularly at physiological
pH, stable analogs with inhibitory activity in the selected enzyme
were discovered in the present study. While the first analog 4 of
similar size as 1 was practically inactive towards SIRT1 and only
showed moderate inhibition of SIRT2, the introduction of a nitro
group in position 5 of the 1,3-dihydrobenzo[e]benzimidazol-2-
one ring of 4 lead to a more potent derivative 5 with an IC50 value
of 12.0 lM for SIRT2. An even more pronounced effect could be ob-
served after formally exchanging the oxygen atom in 4 by a sulfur
atom in compound 6. Interestingly, the SIRT2 selectivity in 5 is
completely lost in 6. The IC50 value of this spitomicin analog was
Figure 6. Predicted binding mode of the three tautomers of 6a (green), 6b (light
blue), and 6c (magenta) to human SIRT2 (1 J8F.pdb) in analogy to published docking
simulations for 17.
10.0 lM for SIRT2 and 9.6 lM for SIRT1. The question whether this

M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
3673
increase in SIRT1 inhibition may be attributed to the fact that the
H3 C
H
NH2
O
thiourea 6 is able to form enethiol tautomers thus altering hydro-
gen bond network architecture or by size or acidity remains to be
investigated. Both compounds 5 and 6 already show considerable
improvement over compound 1. As well (a) the 5-nitro-substitu-
tion as (b) the exchange of the urea moiety by a thiourea function
on the one hand appear to be beneficial modifications of 4 but lead
to an increase of molecular mass and potentially a decrease of
ligand efficiency and cell permeability on the other hand. As could
already be shown for compound 7, (c) size reduction of the
aromatic ring system of 1 is a feasible way to further improve
the ligand efficiency of the lead structure. It was known that the
introduction of substituents on either of both methylene carbon
O
NH
O
H3 C
Cl
17
20
NO2
Cl–
H3+ N
O
S
N
atoms in
a- or b-position of the splitomicin carboxyl function
NH
NH
can lead to enhanced activity, thus (d) the introduction of addi-
tional substituents on one of the two urea or thiourea nitrogen
atoms was envisioned. The combination of these four modifica-
tions (a–d) appeared to be especially attractive, because the
replacement of the naphthalene moiety by smaller aromatic rings
such as benzene and pyridine would compensate for the increase
in molar mass associated with the introduction of substituents
on N1 and in position 5 and the oxygen sulfur exchange. The fact
that only compounds 13e and 4 but not compounds of type 9
and 10 exhibited very moderate SIRT3 inhibition show, that these
scaffolds might be useful starting points to identify sirtuin inhibi-
tors with an advantageous selectivity profile.
Cl
Cl
NO2
21
10c
Figure 8. Known classes of sirtuin inhibitors (17, EX-527 (20), 21) and the novel
sirtuin inhibitors of type 10.
salts are well suited for crystal soaking experiments in aqueous
buffers, and this might aid in the determination of the binding
mode of 10c via X-ray crystallography. The structures reported
here consist of a unique sirtuin inhibitor scaffold, thus the novel
compounds 10 bridge the gap in chemical space between the labile
lead 1, sirtuin inhibiting indoles9 like EX-527 (20) and recently dis-
covered class of 3-arylidene-indolin-2-one sirtuin inhibitors30 21 (
Fig. 8). Analogs of 10c with a single defined binding mode would
represent interesting tools for the investigation of molecular inter-
actions and sirtuin functions.
5. Conclusion
The formal replacement of the lactone ring in sirtuin inhibitors
of the splitomicin type by a stable imidazole-derived heterocyclic
ring system enabled the synthesis of novel inhibitors with activity
within the low micromolar concentration range. While this activity
is superior to the lead compound splitomicin (1) it is still not useful
as an experimental tool as a chemical probe29 in the stricter sense
of this term. However, the successful identification of a unique
inhibitor scaffold adds to the knowledge in the field and comple-
ments nicely to the hits identified by a high throughput screening
approach4 performed by the group of Marmorstein and co-work-
ers. Since the ligand efficiency (i.e., potency/size) of the novel com-
pounds is good, decoration of the novel lead structure 6 in position
1 with aryl rings, and positions 5 or 7 with nitro groups and halo-
gen atoms, respectively, seems rewarding. Substitution of N1 with
various aromatic substituents would appear to be a logical choice
for two reasons. Firstly, SAR investigations from compound class
17 clearly show the beneficial effect upon binding affinity and sec-
ondly, the introduction of substituent will diminish the possibility
of the formation of tautomers like 6a. The only tautomers left to
form a dynamic equilibrium of types 6b and c were proposed to
bind in identical orientations and thus facilitate interpretation of
results of biological experiments. From the perspective of a medic-
inal chemist, compounds with aromatic nitro groups often give rise
to toxicity concerns and these metabolically unstable groups
should be replaced by more benign or less problematic functional
groups with similar electronic and steric properties. Here this is
less important, because compounds like 10c at this stage of inves-
tigation will only be regarded as a step towards pharmacological
tools or molecular probes and not leads for, for example, anticancer
drugs. While lipophilic substitutions are usually the key to in-
creased affinity via non-polar interactions, too many lipophilic
substituents often lead to disadvantageous physicochemical char-
acteristics in biological systems. Thus, additional analogs of 6 or
compounds from the series 10 with salt-forming basic nitrogen
atoms, such as the most active and highly soluble compound 10c
should be prepared and tested. Water soluble hydrochloric acid
A combination of SAR data on these related compound classes
based on careful exploration of binding modes of these molecules,
will facilitate the synthesis of more potent, selective and water sol-
uble sirtuin modulators for pharmacological research.
6. Experimental
6.1. General
NMR spectra were recorded on a Bruker BioSpin UltraShield 400
magnet with Avance III console, a DPX 200 MHz or a JEOL ECLIPSE+
500 spectrometer, using tetramethylsilane as internal standard.
The purity of novel compounds was deduced from NMR data as
well as evaluated by HPLC, using a VWR LaChrom or a Shimadzu
20A-Prominence HPLC-system with SPD-M20A photo diode array
and ESD-LT II evaporative light scattering detector using CC 125/
4 Nucleodur 100-5 C18 ec columns, supplied by Macherey-Nagel.
HRMS data were obtained offline on a Micromass Autospec (ESI,
methanol (1:1, v/v) infusion at 10 lL/min with polyethylene glycol
as reference) or online using a Shimadzu LC–MS-IT-TOF instru-
ment. TLC reaction control was performed on Macherey-Nagel
Polygram Sil G/UV254 precoated microplates, spots were visual-
ized under UV-illumination at 254 nm. IR-spectra were recorded
on a ThermoFisher Nicolet IR200 FT-IR Spectrometer. Refractive
indices were determined using a thermostated Abbé refractometer
(Carl Zeiss). Hydrogenations were performed in a HY 1000 appara-
tus (Hyscho, Bonn). Microwave reactions were executed using a
Discover reactor (CEM, Kamp-Lintfort). Docking simulations were
generated with MOE (The Molecular Operating Environment)
Version 2009.10 from the Chemical Computing Group Inc. with
London dG rescoring function and MMFF94Â forcefield refinement
at default settings.

3674
M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
6.2. General synthetic procedures
complete. Crystals are collected by filtration and washed with
THF and dried under reduced pressure. Yield: 0.275 g (0.62 mmol),
(cmÀ1) = 3428 cm; 2879 cm;
~
For the cyclization of urea derivatives, a solution of the corre-
sponding o-phenylene diamine in CH2Cl2 is combined with a solu-
tion of NaOH in H2O and stirred. A solution of triphosgene in
CH2Cl2 is added dropwise slowly into the mixture which is stirred
for 12 h. Subsequently the mixture is acidified with HCl (18%). The
solvents are evaporated under reduced pressure. The resulting
solid is purified as described.
For the cyclization of thiourea derivatives the ethanolic solution
of the corresponding o-phenylene diamine, triethyl amine and CS2
are heated to reflux. After the reaction has proceeded for a certain
time (as described) the mixture is allowed to cool. The collected
crystals are washed as described and dried under reduced
pressure.
83%, black crystals, mp: 229 °C. IR:
m
1614 cm. 1H NMR: (200 MHz, DMSO-d6) d (ppm) = 13.54 (s, 1H,
NH); 8.28 (s, 3H, NH3); 8.18 (dd, 1H, J = 8.9 Hz, H arom.); 7.96 (d,
1H, J = 2.1 Hz, H arom.); 7.83 (d, 1H, J = 8.9 Hz, H arom.); 4.57 (t,
2H, J = 6.1 Hz, 2CH2); 3.24 (m, 2H, 1CH2). 13C NMR: (50 MHz,
DMSO-d6) d (ppm) = 171.69; 143.15; 137.39; 130.76; 118.77;
109.18; 105.04; 41.00; 36.74. HRMS: [m/z]: calcd [M]+: 238.0524;
found: 238.0505.
6.2.4. tert-Butyl-[3-(5-nitro-2-thioxo-1,3-dihydro-1H-
benzimidazol-1-yl)propyl]carbamate (10d)
Compound 10d was prepared from 0.9 g (2.9 mmol) tert-butyl-
{3-[(2-amino-4-nitrophenyl)amino]propyl}carbamate, 0.5 g CS2,
and 0.6 g triethyl amine in 30 ml of boiling ethanol. After 4 h the
reaction was stopped. The collected crystals were washed with
ethanol. Yield: 0.85 g (2.4 mmol), 83%, yellow needles, mp:
~
6.2.1. 1-Isopropyl-5-nitro-1,3-dihydro-2H-benzimidazol-2-one
(10a)
Compound 10a was prepared from 0.4 g (2 mmol) N1-isopropyl-
4-nitrobenzene-1,2-diamine in 5 ml CH2Cl2, 0.75 g NaOH in 10 ml
H2O, and 0.25 g triphosgene in 5 ml CH2Cl2. The resulted solid was
dissolved in 50 ml ethyl acetate and washed with brine
(2 Â 50 ml) an H2O (1 Â 50 ml), dried with Na2SO4, evaporated un-
der reduced pressure. The resulted crystals were washed with meth-
anol and dried under reduced pressure. Yield: 0.34 g (1.5 mmol),
~
212 °C. IR:
m
(cmÀ1) = 3194; 2972; 1680; 1536; 1514; 1472;
1436; 1368; 1339; 1323; 1273; 1141; 752; 729; 470. 1H NMR:
(400 MHz, DMSO-d6) d (ppm) = 13.30 (s, 1H, NH); 8.13 (dd, 1H,
3Jo = 8.8 Hz, 4Jm = 2.0 Hz, Ar-H); 7.89 (d, 1H, 4Jm = 2.0 Hz, Ar-H);
7.60 (d, 1H, 3J = 8.8 Hz, Ar-H); 6.91 (s, 1H, NH-Boc); 4.27 (t, 2H,
3J = 7.2 Hz, CH2); 2.98 (m, 2H, CH2); 1.84 (m, CH2); 1.36 (s, 9H,
(CH3)3). 13C NMR: (100 MHz, DMSO-d6) d (ppm) = 171.5 (C@S);
155.6 (C@O); 142.9 (Ar); 137.2 (Ar); 130.7 (Ar); 118.6 (Ar); 109.3
(Ar); 104.8 (Ar); 77.6 (C(CH3)3); 41.5 (CH2); 37.5 (CH2); 28.2
(CH3); 27.8 (CH2). HRMS: [m/z]: calcd [C15H20N4O3S+H]+:
353.1278; found: 353.1289.
75%, yellow crystals, mp: 260–261 °C. IR:
m
(cmÀ1) = 2981; 2835;
2781; 1693; 1626; 1514; 1486; 1366; 1335; 1304; 1185; 1142;
1102; 1069; 1051; 947; 867; 833; 791; 755; 727; 704; 573; 550.
1H NMR: (400 MHz, DMSO-d6) d (ppm) = 11.42 (s, 1H, NH); 7.96
(dd, 1H, 3Jo = 8.8 Hz, 4Jm = 2.4 Hz, Ar-H); 7.73 (d, 1H, 4Jm = 2.4 Hz,
Ar-H); 7.47 (d, 1H, 3Jo = 8.8 Hz, Ar-H); 4.63 (sp, 1H, 3J = 6.8 Hz,
CH(CH3)2); 1.45 (d, 6H, 3J = 6.8 Hz, 2 Â CH3). 13C NMR: (100 MHz,
DMSO-d6) d (ppm) = 154.0 (C@S); 141.1 (Ar); 134.7 (Ar); 128.4
(Ar); 117.4 (Ar); 108.2 (Ar); 103.8 (Ar); 44.6 (CH(CH3)2); 19.7
(CH3). HRMS: [m/z]: calcd [C10H11N3O3ÀH]À: 220.0728; found:
220.0729.
6.2.5. 1-(3-Aminopropyl)-5-nitro-1,3-dihydro-2H-benzimid-
azole-2-thione hydrochloride (10e)
Through a solution of 0.54 g (1.5 mmol) 10d in 50 ml dry THF
HCl gas is bubbled until precipitation of a solid is complete. Crys-
tals are collected by filtration and washed with THF and dried un-
der reduced pressure. Yield: 0.3 g (1 mmol), 69%, yellow crystals,
~
mp: 235–239 °C. IR:
m
(cmÀ1) = 3043; 2357; 1619; 1514; 1470;
1433; 1336; 1321; 1203; 748; 727; 476. 1H NMR: (200 MHz,
DMSO-d6) d (ppm) = 13.50 (s, 1H, NH); 8.18 (dd, 1H, 3Jo = 8.8 Hz,
4Jm = 2.0 Hz, Ar-H); 8.07 (s, 3H, NH2ÁHCl); 7.97 (d, 1H, 4Jm = 2.0 Hz,
Ar-H); 7.76 (d, 1H, 3Jo = 8.8 Hz, Ar-H); 4.39 (t, 2H, 3J = 7.4 Hz, CH2);
2.89 (t, 2H, 3J = 7.4 Hz, CH2); 2.05 (qn, 2H, 3J = 7.4 Hz, CH2). 13C
NMR: (50 MHz, DMSO-d6) d (ppm) = 171.6 (C@S); 143.0 (Ar);
137.2 (Ar); 130.8 (Ar); 118.6 (Ar); 109.5 (Ar); 105.0 (Ar); 40.9
(CH2); 36.3 (CH2); 25.8 (CH2). HRMS: [m/z]: calcd
[C10H12N4O2SÀH]À: 251.0608; found: 251.0618.
6.2.2. 1-Cyclohexyl-5-nitro-1,3-dihydro-2H-benzimidazol-2-
one (10b)
Compound 10b was prepared from 0.39 g (1.7 mmol) N1-iso-
propyl-4-nitrobenzene-1,2-diamine in 4 ml CH2Cl2, 0.55 g NaOH
in 4 ml H2O, and 0.18 g triphosgene in 4 ml CH2Cl2. The resulting
solid was dissolved in 40 ml ethyl acetate and washed with brine
(2 Â 50 ml) and H2O (1 Â 50 ml), dried with Na2SO4, evaporated
under reduced pressure. Resulted crystals were washed with
methanol and dried under reduced pressure. Yield: 0.37 g
~
(1.4 mmol), 82%, orange crystals, mp: 282 °C. IR:
m
(cmÀ1) =
3017; 2935; 2858; 1599; 1621; 1514; 1487; 1443; 1373; 1336;
1299; 1140; 1107; 1069; 816; 748; 710; 657; 616; 550; 441. 1H
NMR: (200 MHz, DMSO-d6) d (ppm) = 11.45 (s, 1H, NH); 7.96
(dd, 1H, 3Jo = 8.8 Hz, 4Jm = 2.4 Hz, Ar-H); 7.75 (d, 1H, 4Jm = 2.4 Hz,
Ar-H); 7.53 (d, 1H, 3Jo = 8.8 Hz, Ar-H); 4.28–4.15 (m, 1H, CH);
2.14–1.32 (m, 10H, CH2 (cyclohexyl)). 13C NMR: (50 MHz,
DMSO-d6) d (ppm) = 154.0 (C@O); 141.0 (Ar); 134.9 (Ar); 128.4
(Ar); 117.3 (Ar); 108.3 (Ar); 103.7 (Ar); 52.3 (CH); 29.3
(CH2 (cyclohexyl)); 25.4 (CH2 (cyclohexyl); 24.6 (CH2 (cyclo-
hexyl)). HRMS: [m/z]: calcd [C13H15N3O3+H]+: 262.1186; found:
262.1196.
6.2.6. 1-(2,4-Dimethoxybenzyl)-5-nitro-1,3-dihydro-2H-benzi-
midazol-2-one (10f)
Compound 10f was prepared from 0.15 g (0.5 mmol) 13e in
2 ml CH2Cl2, 0.4 g NaOH in 2 ml H2O, and 0.22 g triphosgene in
2 ml CH2Cl2. After 5 h the reaction was stopped. The resulting solid
was washed with H2O (2 Â 15 ml) and crystallized from ethanol.
Yield: 0.124 g (0.38 mmol), 75%, colorless solid, mp: 224–226 °C.
~
IR:
m
(cmÀ1) = 3123; 3007; 1711; 1617. 1H NMR: (200 MHz,
DMSO-d6) d (ppm) = 7.44–7.49 (m, 1H, H arom.); 7.42 (t, 1H,
J = 2.5 Hz NH); 7.06 (d, 1H, J = 8.5 Hz, H arom.); 6.59 (d, 1H,
J = 2.2 Hz, H arom.); 6.47 (dd, 1H, J = 2.2/8.4 Hz, H arom.); 6.25–
6.37 (m, 2H, H arom.); 5.22 (s, 2H, NH2); 4.30 (d, 1H, J = 5.3 Hz,
CH2); 3.82 (s, 3H, CH3); 3.74 (s, 3H, CH3). 13C NMR: (126 MHz,
DMSO-d6) d (ppm) = 160.24; 157.72; 154.62; 141.31; 135.80;
129.37; 128.37; 117.54; 115.73; 107.70; 104.73; 103.81; 98.35;
55.42; 55.14; one signal covered by DMSO signal. HRMS: [m/z]:
calcd [M]+: 329.1012; found: 329.1019.
6.2.3. 1-(2-Aminoethyl)-5-nitro-1,3-dihydro-2H-benzimidazole-
2-thione hydrochloride (10c)
Through a solution of 0.25 g (0.74 mmol) tert-butyl [2-(5-nitro-
2-thioxo-2,3-dihydro-1H-benzimidazol-1-yl)ethyl]carbamate
in
10 ml dry THF HCl gas is bubbled until precipitation of a solid is

M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
3675
6.2.7. 1-(2,4-Dimethoxybenzyl)-5-nitro-1,3-dihydro-2H-benz-
imidazole-2-thione (10g)
with H2 (3 bar) in a hydrogenation apparatus until constant pres-
sure is reached. After filtration, the product solution 0.65 ml CS2
and a pellet of NaOH are added, the mixture is heated to reflux un-
til complete reaction (TLC control) is observed. Subsequently the
solvent is removed; the resulted solid is washed several times with
H2O. A brown solid is isolated. Yield: 0.251 g, (0.68 mmol) 68%,
(cmÀ1) = 3058; 2948; 1643. 1H NMR:
~
mp: 362–364 °C. IR: m
Compound 10g was prepared from 0.152 g (0.5 mmol) 13e,
0.6 g CS2, 0.1 g triethyl amine, 20 ml ethanol. After 4 h the reaction
was stopped. The mixture was evaporated under reduced pressure
and crystallized from ethanol. Yield: 0.137 g (0.39 mmol), 79%,
~
beige solid, mp: 271–274 °C. IR:
m
(cmÀ1) = 3063; 2962; 1617. 1H
NMR: (200 MHz, DMSO-d6) d (ppm) = 13.41 (s, 1H, NH); 8.10 (dd,
1H, J = 2.3/8.9 Hz, H arom.); 7.92 (d, 1H, J = 2.2 Hz, H arom.); 7.35
(d, 1H, J = 8.9 Hz, H arom.); 6.90 (d, 1H, J = 8.5 Hz, H arom.); 6.61
(d, 1H, J = 2.1 Hz, H arom.); 6.42 (dd, 1H, J = 2.2/8.4 Hz, H arom.);
5.39 (s, 2H, CH2), 3.86 (s, 3H, CH3); 3.72 (s, 3H, CH3). 13C NMR:
(126 MHz, DMSO-d6) d (ppm) = 172.13; 160.18; 157.59; 142.88;
137.18; 130.61; 128.89; 118.59; 114.98; 109.69; 104.77; 104.75;
98.32; 55.49; 55.13; 41.46. Elemental Anal. Calcd for C16H15N3O4S:
C, 55.65; H, 4.38; N, 12.17. Found: C, 55.70; H, 44.1; N, 11.67. HRM:
[m/z]: calcd [M]+: 345.0783; found: 345.0773.
(200 MHz, DMSO-d6) d (ppm) = 13.42 (s, 1H, NH); 12.86 (s, 1H,
NH); 8.33 (d, 1H, J = 8.5 Hz, H arom.); 7.86 (d, 1H, J = 8.1 Hz, H
arom.); 7.38–7.73 (m, 4H, H arom.). 13C NMR: (50 MHz, DMSO-
d6) d (ppm) = 166.08; 129.25; 128.59; 128.36; 126.59; 126.56;
124.41; 123.06; 120.80; 119.00; 110.54.
6.2.12. 2-Thioxo-2,3-dihydro-1H-benzimidazole-5-carboxylic
acid (8)
A mixture of 0.912 g (6 mmol) 3,4-diaminobenzoic acid, 0.24 g
NaOH, 1 ml CS2 and 20 ml H2O is heated until reflux. Subsequently
the free acid is precipitated with glacial acetic acid. The precipitate
is washed several times with H2O and dried under reduced pres-
sure. A brown solid is isolated. Yield: 1.0 g (4.1 mmol), 68%, mp:
(cmÀ1) = 3065; 2965; 1683; 1629 cm. 1H NMR:
~
362–366 °C. IR: m
6.2.8. 1-(1,3-Benzodioxol-5-yl)-5-(trifluoromethyl)-1,3-dihy-
dro-2H-benzimidazol-2-one (10h)
Compound 10h was prepared from 0.1 g (0.33 mmol) N1-1,3-
benzodioxol-5-yl-4-(trifluoromethyl)benzen-1,2-diamine in 2 ml
CH2Cl2, 0.4 g NaOH in 2 ml H2O, 0.22 g triphosgene in 2 ml CH2Cl2.
The resulting solid was washed with H2O (2 Â 15 ml) and crystal-
lized from ethanol. Yield: 0.089 g (0.046 mmol), 14%, brown solid,
~
(200 MHz, DMSO-d6) d (ppm) = 12.80 (s, 3H, NH, COOH); 7.75
(dd, 1H, J = 8.3/1.4 Hz, H arom.); 7.65 (m, 1H, H arom.); 7.20 (d,
1H, J = 8.3 Hz,
H d
arom.). 13C NMR: (50 MHz, DMSO-d6)
(ppm) = 134.61; 133.94; 133.42; 129.39; 128.75; 117.43; 116.65;
111.39.
mp: 224–235 °C. IR:
m
(cmÀ1) = 3127; 3042; 2907; 1699; 1625. 1H
NMR: (200 MHz, DMSO-d6) d (ppm) = 11.47 (s, 1H, NH); 7.30–7.39
(m, 2H, H arom.); 6.95–7.12 (m, 4H, H arom.); 6.14 (s, 2H, CH2). 13
C
6.2.13. 5-Chloro-1,3-dihydro-2H-imidazo[4,5-b]pyridin-2-one
(9a)
NMR: (50 MHz, DMSO-d6)
d (ppm) = 153.47; 147.80; 146.86;
133.55; 128.45; 127.30; 124.59 (q, 1J(C,F); À272 Hz, CF3); 122.01
(q, 2J(C,F); 32 Hz, C5); 120.14; 118.17 (q, 3J(C,F); 4 Hz, C4);
108.47; 108.16; 107.78; 105.40 (q, 3J(C,F): 4 Hz, C6); 101.77.
HRMS: [m/z]: calcd [M]+: 322.0565; found: 322.0553.
Compound 9a was prepared from 0.143 g (1 mmol) 6-chloro-
pyridine-2,6-diamine in 2 ml CH2Cl2, 0.4 g NaOH in 2 ml H2O,
and 0.22 g triphosgene in 2 ml CH2Cl2 in analogy to described7 pro-
cedures. Colorless crystals are obtained from ethanol. Yield: 0.08 g
~
(0.47 mmol), 47%, mp: 374 °C. IR:
m
(cmÀ1) = 3094; 3002; 2817;
6.2.9. 1,3-Dihydro-2H-naphtho[1,2-d]imidazol-2-one (4)
A solution of 0.445 g (2 mmol) 4-chloro-2-nitro-1-naphthyl-
amine in 50 ml CH3OH and 0.06 g Pd/C (10%) are allowed to shake
with H2 (3 bar) in a hydrogenation apparatus until constant pres-
sure was reached. After filtration the product solution is evaporated
under reduced pressure. Subsequently 10 ml CH2Cl2 and 10 ml
NaOH (5 mol/l) and 0.2 g triphosgene is added and the mixture is al-
lowed to stir for 12 h. The mixture is acidified with HCl (18%). The li-
quid is removed by filtration, resulting in a brown solid. Yield:
~
1770. 1H NMR: (200 MHz, DMSO-d6) d (ppm) = 11.58 (s, 1H, NH);
11.03 (s, 1H, NH); 7.27 (d, 1H, J = 8.0 Hz, H arom.); 7.01 (d, 1H,
J = 8.1 Hz,
H
arom.). 13C NMR: (50 MHz, DMSO-d6)
d
(ppm) = 154.20; 144.34; 139.82; 122.81; 117.08; 115.63.
6.2.14. 5-Chloro-1,3-dihydro-2H-imidazo[4,5-b]pyridine-2-
thione (9b)
A mixture of 0.144 g (1 mmol) 6-chlorpyridine-2,3-diamine,
20 ml ethanol, one pellet of NaOH and 0.5 ml CS2 is heated to re-
flux7 as described. When the reaction is completed (TLC control)
the mixture is acidified with HCl (2 mol/l), the solvent is evapo-
rated under reduced pressure. The resulting solid is washed several
times with H2O. Beige solid, yield: 0.125 g (0.67 mmol), 67%, mp:
(cmÀ1) = 3130; 3063; 2945; 1604. 1H NMR:
~
m
0.251 g (1.36 mmol), 68%, mp >360 °C. IR:
m
(cmÀ1) = 3112; 1724;
1662. 1H NMR: (200 MHz, DMSO-d6) d (ppm) = 11.48 (s, 1H, NH);
10.90 (s, 1H, NH); 8.08 (d, 1H, J = 8.3 Hz, H arom.); 7.86 (d, 1H,
J = 8.4 Hz, H arom.); 7.26–7.57 (m, 4H, H arom.). 13C NMR:
(50 MHz, DMSO-d6) d (ppm) = 155.22; 128.53; 128.45; 125.61;
124.91; 123.08; 122.96; 120.53; 120.41; 118.99; 110.57.
290–295 °C. IR:
(200 MHz, DMSO-d6) d (ppm) = 13.31 (s, 1H, NH); 12.86 (s, 1H,
NH); 7.50 (d, 1H, J = 8.2 Hz, H arom.); 7.19 (d, 1H, J = 8.2 Hz, H
arom.). 13C NMR: (50 MHz, DMSO-d6) d (ppm) = 170.36; 145.80;
142.48; 124.73; 118.75; 117.43.
6.2.10. 5-Nitro-1,3-dihydro-2H-naphtho[1,2-d]imidazol-2-one
(5)
A mixture of 0.184 g 1,3-dihydro-2H-naphtho[1,2-d]imidazol-
2-one, 2 ml glacial acetic acid and 122
heated carefully to 50 °C. After 2.5 h the mixture is poured on ice
l
L nitric acid (60%) are
6.2.15. 5-Methoxy-1,3-dihydro-2H-imidazo[4,5-b]pyridine-2-
thione (9c)
and washed several times with H2O, a brown solid is isolated.
Compound 9c was obtained as described17 using 2,6-dichloro-
pyridine as the starting material in five steps, including methoxy-
lation, nitration, amination, reduction, and cyclization. The overall
yield of this sequence was up to 45.7%. The structure of the product
was identified by 1H NMR, MS, and IR. A solution of 0.5 g (3 mmol)
6-methoxy-3-nitropyridine-2-amine 100 ml CH3OH, 2 ml H2O, and
0.1 g Pd/C (10%) are allowed to shake with H2 (3 bar) in a hydroge-
nation apparatus until constant pressure. After filtration, the prod-
uct solution is heated with 1 ml CS2 and 0.225 g KOH until reflux.
When the reaction is completed (TLC control) the solvent is evap-
orated under reduced pressure. The resulted solid is recrystallized
Yield: 0.146 g, 76%. IR:
(cmÀ1) = 3338; 3011; 2960; 1722; 1596.
~
m
1H NMR: (200 MHz, DMSO-d6) d (ppm) = 12.28 (s, 1H, NH); 11.37
(s, 1H, NH); 8.61–8.69 (m, 1H, H arom.); 8.20–8.26 (m, 1H, H
arom.); 8.17 (s, 1H, H arom.); 7.63–7.72 (m, 2H, H arom.). 13C
NMR: (50 MHz, DMSO-d6)
d (ppm) = 155.34; 138.43; 130.34;
127.58; 126.77; 123.55; 123.13; 121.72; 121.48; 117.95; 110.06.
6.2.11. 1,3-Dihydro-2H-naphtho[1,2-d]imidazole-2-thione (6)
A solution of 0.223 g (1 mmol) 4-chloro-2-nitro-1-naphthyl-
amine in 50 ml CH3OH and 0.06 g Pd/C (10%) are allowed to shake

3676
M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
were purified as described31 with minor modifications. Purity
and identity of the produced enzymes were verified using SDS
gel electrophoresis. The deacetylase activity of the sirtuins was
dependent on the sirtuin cofactor NAD+ and could be inhibited
with the endogenous sirtuin inhibitor nicotinamide.
~
from ethanol. Yield: 0.187 g (1.1 mmol), 35%, mp: 223–225 °C. IR:
m
(cmÀ1) = 3262; 3051; 1624. 1H NMR: (200 MHz, DMSO-d6) d
(ppm) = 12.99 (s, 1H, NH); 12.54 (s, 1H, NH); 7.43–7.47 (d, 1H,
J = 8.5 Hz, H arom.); 6.55–6.60 (d, 1H, J = 8.5 Hz, H arom.); 3.83
(s, 3H, CH3). 13C NMR: (50 MHz, DMSO-d6) d (ppm) = 167.60;
160.12; 143.62; 120.17; 119.65; 103.90; 53.49.
6.4. Fluorescent deacetylase assay
6.2.16. N-(2,4-Dinitrophenyl)-1,3-benzodioxol-5-amine (12e)
Compound 12e was obtained by heating a mixture of 1.709 g
(7 mmol) 2,4-dimethoxybenzylamine and 5 mmol 2,4-dinitrofluo-
robenzene, 10 mmol NaHCO3, 50 ml acetone, and 5 ml H2O to re-
flux for 1–2 h. The solvent is evaporated under reduced pressure.
The resulted solid is washed with H2O and recrystallized from eth-
anol. Yield: 1.37 g (4.1 mmol), 82%, yellow needles, mp: 107–
~
Rates of inhibition of recombinant human SIRT1, SIRT2, and
SIRT3 were determined using a homogeneous fluorescent deacetyl-
ase assay32 carried out in 96-well plates. The assay mixture con-
tained the fluorescent deacetylase substrate ZMAL (10.5
NAD+ (500
M), the inhibitors at various concentrations and DMSO
(5–10% v/v). Total assay volume was 60 L. The amount of enzyme
lM),
l
l
109 °C. IR:
m
(cmÀ1) = 3352; 3112; 1624. 1H NMR: (200 MHz,
preparation was dependent on the activity of the preparation that
was used and varied from batch to batch. To assure initial state
conditions we usually adjusted to 10–30% substrate conversion
without inhibitor. A mixture solely with DMSO was used as a con-
DMSO-d6) d (ppm) = 9.18 (t, 1H, J = 6.0 Hz, NH); 8.85 (d, 1H,
J = 2.7 Hz, H arom.) 8.22 (dd, 1H, J = 2.7/9.6 Hz, H arom.); 7.06–
7.18 (m, 2H, H arom.); 6.61 (d, 1H, J = 2.3 Hz, H arom.); 6.47 (dd,
1H, J = 2.3/8.3 Hz, H arom.); 4.59 (d, 2H, J = 6.0 Hz, CH2); 3.85 (s,
3H, CH3); 3.73 (s, 3H, CH3). 13C NMR: (126 MHz, DMSO-d6) d
(ppm) = 160.17; 157.78; 147.95; 134.83; 129.88; 129.79; 128.75;
123.49; 116.39; 115.31; 104.65; 98.50; 55.49; 55.14; 41.44. HRMS:
[m/z]: calcd [M]+: 333.0951; found: 333.0951.
trol. After incubation (4 h, 37 °C) a stop solution (60 lL) was added
containing trypsin (1 mg  mlÀ1, from bovine or porcine pancreas,
10,000–200,000 BAEE units  mgÀ1) and nicotinamide (8 mM) to
cleave off the fluorescent aminocoumarin. Fluorescence intensity
was measured with
a microplate reader (BMG Polarstar,
kex = 390 nm, kem = 460 nm). Comparison of the fluorescence inten-
sity of the DMSO control with a probe containing a potential inhib-
itor was used to determine the rate of inhibition. All inhibition
determinations were carried out at least in duplicates. IC50 values
were determined using GraphPad Prism software.
6.2.17. N1-(2,4-Dimethoxybenzyl)-4-nitro-benzene-1,2-diamine
(13e)
To a boiling solution of 0.833 g (2.5 mmol) 12e in 90 ml meth-
anol a solution of 2.1 g NaHCO3 and 6.1 g Na2SÁ9H2O in 25 ml
H2O is dropped over 1 h. Subsequently the mixture is allowed to
boil at reflux for 4 h. When the reaction is finished, the mixture
is poured in 300 ml cold H2O. The resulted crystals are washed
with H2O and recrystallized from ethanol. Yield: 1.37 g (2.1 mmol),
~
Acknowledgment
The authors acknowledge financial support by DFG Grants
Li765/4-2 and Ju295/8-1.
82% red crystals, mp: 160–162 °C. IR:
m
(cmÀ1) = 3390; 3328; 1644.
1H NMR: (200 MHz, DMSO-d6) d (ppm) = 7.44–7.49 (m, 1H, H
arom.); 7.42 (t, 1H, J = 2.5 Hz, NH); 7.06 (d, 1H, J = 8.5 Hz, H arom.);
6.59 (d, 1H, J = 2.2 Hz, H arom.); 6.47 (dd, 1H, J = 2.2/8.4 Hz, H
arom.); 6.25–6.37 (m, 2H, H arom.); 5.22 (s, 2H, NH2); 4.30 (d,
1H, J = 5.3 Hz, CH2); 3.82 (s, 3H, CH3); 3.74 (s, 3H, CH3). 13C NMR:
(126 MHz, DMSO-d6) d (ppm) = 159.85; 157.90; 142.38; 136.65;
134.53; 128.79; 117.87; 115.63; 107.28; 107.24; 104.50; 98.37;
55.42; 55.15; 40.98. Elemental Anal. Calcd for C15H17N3O4: C,
59.40; H, 5.65; N, 13.85. Found: C, 59.41; H, 5.72; N, 13.50.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2011.01.026. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
References and notes
6.2.18. 2-Nitro-N-[4-(trifluoromethyl)-phenyl]-1,3-benzodi-
oxol-5-amine (15)
1. Bedalov, A.; Gatbonton, T.; Irvine, W. P.; Gottschling, D. E.; Simon, J. A. Proc.
Natl. Acad. Sci. U.S.A. 2001, 98, 15113.
2. Lin, S. J.; Defossez, P. A.; Guarente, L. Science 2000, 289, 2126.
3. Lawson, M.; Uciechowska, U.; Schemies, J.; Rumpf, T.; Jung, M.; Sippl, W.
Biochim. Biophys. Acta 2010, 1799, 726.
4. Sanders, B. D.; Jackson, B.; Brent, M.; Taylor, A. M.; Dang, W. W.; Berger, S. L.;
Schreiber, S. L.; Howitz, K.; Marmorstein, R. Bioorg. Med. Chem. 2009, 17, 7031.
5. Peck, B.; Chen, C. Y.; Ho, K. K.; Di Fruscia, P.; Myatt, S. S.; Coombes, R. C.;
Fuchter, M. J.; Hsiao, C. D.; Lam, E. W. F. Mol. Cancer Ther. 2010, 9, 844.
6. Pagans, S.; Pedal, A.; North, B. J.; Kaehlcke, K.; Marshall, B. L.; Dorr, A.; Hetzer-
Egger, C.; Henklein, P.; Frye, R.; McBurney, M. W.; Hruby, H.; Jung, M.; Verdin,
E.; Ott, M. PLoS Biol. 2005, 3, 210.
7. Guarente, L. Nature 2006, 444, 868.
8. Gao, J.; Wang, W. Y.; Mao, Y. W.; Gräff, J.; Guan, J. S.; Pan, L.; Mak, G.; Kim, D.;
Su, S. C.; Tsai, L. H. Nature 2010, 466, 1105.
9. Sinclair, D. A.; Lin, S. J.; Guarente, L. Science 2006, 312, 195.
10. Grozinger, C. M.; Schreiber, S. L. Chem. Biol. 2002, 9, 3.
11. Posakony, J.; Hirao, M.; Stevens, S.; Simon, J. A.; Bedalov, A. J. Med. Chem. 2004,
47, 2635.
A mixture of 1.646 g (12 mmol) 3,4-(methylenedioxy)aniline
and
10 mmol
1-fluoro-2-nitro-4-(trifluoromethyl)benzene,
20 mmol NaHCO3, 100 ml acetone, and 10 ml H2O is heated until
reflux for 1–2 h. The solvent is evaporated under reduced pressure.
The resulted solid is washed with H2O and recrystallized from eth-
anol. Yield: 1.625 g (0.67 mmol), 67%, red solid, mp: 172–175 °C.
~
IR:
m d
(cmÀ1) = 3350; 1583. 1H NMR: (500 MHz, DMSO-d6)
(ppm) = 10.07 (s, 1H, NH); 8.88 (d, 1H, J = 2.8 Hz, H arom.); 8.21
(dd, 1H, J = 2.8/9.6 Hz, H arom.); 6.85 (dd, 1H, J = 1.6/8.3 Hz, H
arom.); 7.00–7.10 (m, 2H, H arom.); 6.98 (d, 1H, J = 1.8 Hz, H
arom.); 6.85 (m, 1H, H arom.); 6.11 (s, 2H, CH2). 13C NMR:
(126 MHz, DMSO-d6) d (ppm) = 148.72; 148.01; 147.01; 136.53;
131.97; 131.40; 130.16; 123.79; 120.21; 117.54; 109.51; 108.36;
102.27. Elemental Anal. Calcd for C13H9N3O6: C, 51.49; H, 2.99;
N, 13.86. Found: C, 51.16; H, 3.23; N, 13.71.
12. Diels, O. Liebigs Ann. Chem. 1922, 429, 1.
13. Reich, S. H.; Fuhry, M. A. M.; Nguyen, D.; Pino, M. J.; Welsh, K. M.; Webber, S.;
Janson, C. A.; Jordan, S. R.; Matthews, D. A.; Smith, W. W.; Bartlett, C. A.; Booth,
C. L. J.; Herrmann, S. M.; Howland, E. F.; Morse, C. A.; Ward, R. W.; White, J. J.
Med. Chem. 1992, 35, 847.
6.3. Recombinant proteins
14. Ghosh, T. N. J. Indian Chem. Soc. 1933, 583.
15. Nakayama, T.; Taira, S.; Ikeda, M.; Ashizawa, H.; Oda, M.; Arakawa, K.; Fujii, S.
Chem. Pharm. Bull. 1993, 41, 117.
16. Bebenburg, W. V.; Steinmetz, G.; Thiele, K. Chem. Ztg. 1979, 103, 387.
17. Li, M.-L.; Cheng, J.; Cao, S.-T.; Liu, Z.-L. Yingyong Huaxue 2010, 27, 122.
Human Sirt1 was expressed in E. coli as an N-terminally GST-
tagged, human Sirt2 as an N-terminally His6-tagged and human
Sirt3 as an C-terminally His6-tagged fusion protein. The enzymes

M. Freitag et al. / Bioorg. Med. Chem. 19 (2011) 3669–3677
3677
18. Neugebauer, R. C.; Uchiechowska, U.; Meier, R.; Hruby, H.; Valkov, V.; Verdin,
E.; Sippl, W.; Jung, M. J. Med. Chem. 2008, 51, 1203.
19. Larsson, A.; Angbrant, J.; Ekeroth, J.; Mansson, P.; Liedberg, B. Sens. Actuators, B.
Chem. 2006, 113, 730.
20. Willitzer, H.; Brauniger, H.; Engelmann, D.; Krebs, D.; Ozegowski, W.; Tonew,
M. Pharmazie 1978, 33, 30.
21. Blanksma, J. J. Recl. Trav. Chim. Pays-Bas. J. Roy. Neth. Chem. Soc. 1948, 67, 1005.
22. Olguin, L. F.; Jimenez-Estrada, M.; Barzana, E.; Navarro-Ocana, A. Synlett 2005,
340.
26. Medda, F.; Russell, R. J. M.; Higgins, M.; McCarthy, A. R.; Campbell, J.; Slawin, A.
M. Z.; Lane, D. P.; Lain, S.; Westwood, N. J. J. Med. Chem. 2009, 52, 2673.
27. Heltweg, B.; Gatbonton, T.; Schuler, A. D.; Posakony, J.; Li, H. Z.; Goehle, S.;
Kollipara, R.; DePinho, R. A.; Gu, Y. S.; Simon, J. A.; Bedalov, A. Cancer Res. 2006,
66, 4368.
28. Uciechowska, U.; Schemies, J.; Neugebauer, R. C.; Huda, E. M.; Schmitt,
M. L.; Meier, R.; Verdin, E.; Jung, M.; Sippl, W. ChemMedChem 2008, 3,
1965.
29. Frye, S. V. Nat. Chem. Biol. 2010, 6, 159.
23. Heim, E. Ber. Dtsch. Chem. Ges. 1888, 21, 2301.
30. Huber, K.; Schemies, J.; Uciechowska, U.; Wagner, J. M.; Rumpf, T.; Lewrick, F.;
Suss, R.; Sippl, W.; Jung, M.; Bracher, F. J. Med. Chem. 2010, 53, 1383.
31. North, B. J.; Schwer, B.; Ahuja, N.; Marshall, B.; Verdin, E. Methods 2005, 36, 338.
32. Heltweg, B.; Trapp, J.; Jung, M. Methods 2005, 36, 332.
24. Heltweg, B.; Dequiedt, F.; Verdin, E.; Jung, M. Anal. Biochem. 2003, 319, 42.
25. Trapp, J.; Jochum, A.; Meier, R.; Saunders, L.; Marshall, B.; Verdin, E.; Goekjian,
P.; Kunick, C.; Sippl, W.; Jung, M. J. Med. Chem. 2006, 49, 737.