Adenosine mimetics as inhibitors of NAD+-dependent histone deacetylases, from kinase to sirtuin inhibition

Article abstract of DOI:10.1021/jm060118b

NAD⁺-dependent histone deacetylases, sirtuins, cleave acetyl groups from lysines of histones and other proteins to regulate their activity. Identification of potent selective inhibitors would help to elucidate sirtuin biology and could lead to

Full text of DOI:10.1021/jm060118b

J. Med. Chem. 2006, 49, 7307-7316
7307
Adenosine Mimetics as Inhibitors of NAD+-Dependent Histone Deacetylases, from Kinase to
Sirtuin Inhibition
Johannes Trapp,† Anne Jochum,‡ Rene Meier,§ Laura Saunders,| Brett Marshall,| Conrad Kunick, Eric Verdin,| Peter Goekjian,‡
Wolfgang Sippl,§ and Manfred Jung†,*
Institute of Pharmaceutical Sciences, Albert-Ludwigs-UniVersita¨t Freiburg, Albertstrasse 25, 79104 Freiburg, Germany, LCO2-Glycochimie,
UMR 5181 Me´thodologie de Synthe`se et Mole´cules BioactiVes, UniVersite´ Claude Bernard Lyon 1, 43 Bd du 11 NoVembre 1918, 69622
Villeurbanne Cedex, France, Department of Pharmaceutical Chemistry, Martin-Luther UniVersita¨t Halle-Wittenberg,
Wolfgang-Langenbeckstrasse 4, 06120 Halle/Saale, Germany, Gladstone Institute of Virology and Immunology, UniVersity of California, 1650
Owens Street, San Francisco, California 94158, and Institut fu¨r Pharmazeutische Chemie, Technische UniVersita¨t Braunschweig,
BeethoVenstrasse 55, 38106 Braunschweig, Germany
ReceiVed February 3, 2006
NAD+-dependent histone deacetylases, sirtuins, cleave acetyl groups from lysines of histones and other
proteins to regulate their activity. Identification of potent selective inhibitors would help to elucidate sirtuin
biology and could lead to useful therapeutic agents. NAD+ has an adenosine moiety that is also present in
the kinase cofactor ATP. Kinase inhibitors based upon adenosine mimesis may thus also target NAD+-
dependent enzymes. We present a systematic approach using adenosine mimics from one cofactor class
(kinase inhibitors) as a viable method to generate new lead structures in another cofactor class (sirtuin
inhibitors). Our findings have broad implications for medicinal chemistry and specifically for sirtuin inhibitor
design. Our results also raise a question as to whether selectivity profiling for kinase inhibitors should be
limited to ATP-dependent targets.
Chart 1. Known Inhibitors for Sirtuins
Introduction
Histone deacetylases (HDACs)a are enzymes that deacetylate
histones and certain nonhistone proteins, thereby altering their
conformational state or activity.1 Three classes of histone
deacetylases have been recognized in humans: class I and II
are zinc-dependent amidohydrolases of which 11 subtypes have
been discovered (HDAC1-11). Class III enzymes depend on
NAD+ for catalysis, and produce O-acetyl ADP ribose and
nicotinamide (1) as a consequence of the acetyl transfer. Due
to homology with the yeast histone deacetylase Sir2p, the
NAD+-dependent deacetylases are also termed sirtuins, and
seven members (SIRT1-7) are known in humans.2
In the past few years a considerable amount of knowledge
has accumulated on the biological activities of sirtuins.2 They
are linked to aging, and overexpression leads to an increased
lifespan in yeast.3 On the other hand, there are indications that
sirtuins play a role in the pathogenesis of viral diseases4 and
cancer.5-7
was sirtinol (2)11 (Chart 1), and its hydrolysis product, 2-hy-
droxynaphthaldehyde, also shows some activity. At least in
certain cases, precipitation of the sirtuin by sirtinol contributes
to enzyme inhibition.12 Structure-activity relationships of
sirtinol analogues have been reported.13 Another inhibitor,
splitomicin (3a), was discovered as an inhibitor of yeast
sirtuins14 and is basically inactive on human subtypes. HR73
(3b) was derived from splitomicin by our group as the first
selective (20-fold for SIRT1 over SIRT2) and potent inhibitor
(IC50 < 5 µM) of human sirtuins4 (See Chart 1). Two other
inhibitors of SIRT2 were discovered using a virtual screening
approach.15 Sirtuin inhibitors with half-inhibitory concentrations
as low as 98 nM have been reported recently.16
While class I and II HDAC inhibitors are already investigated
as new anticancer agents in clinical studies8 much less is known
about inhibitors of class III histone deacetylases. Only a few
sirtuin inhibitors are available, and several of them do not inhibit
human subtypes (Chart 1).9,10 Nicotinamide (1) is the physi-
ological sirtuin inhibitor. The first synthetic inhibitor discovered
Suramin and several related adenosine receptor antagonists
inhibit sirtuins as well.17 This prompted us to start a systematic
investigation of sirtuin inhibition by drugs that target enzymes
or receptors that bind adenosine-containing cofactors or ligands
to identify lead structures for sirtuin inhibitors. Among these
enzymes are kinases (using ATP) and dehydrogenases (using
NAD+). Certain kinase inhibitors, namely members of the
paullone cyclin-dependent kinase (CDK) inhibitor series (e.g.,
kenpaullone), also inhibit NAD+-dependent mitochondrial
malate dehydrogenase (mMDH).18 Thus, we tested a com-
* To whom correspondence should be addressed. E-mail: manfred.jung@
pharmazie.uni-freiburg.de; Tel: +49-761-203-4896; Fax: +49-761-203-
6321.
† Albert-Ludwigs-Universita¨t Freiburg.
‡ Universite´ Claude Bernard Lyon 1.
§ Martin-Luther Universita¨t Halle-Wittenberg.
| University of California.
Technische Universita¨t Braunschweig.
a Abbreviations: HDAC: histone deacetylase; mMDH: mitochondrial
malate dehydrogenase; CDK: cyclin-dependent kinase; BIM: bisindolyl-
maleimide; PKC: protein kinase C.
10.1021/jm060118b CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/10/2006

7308 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25
Trapp et al.
Chart 2. Structures of the Kinase Inhibitors Discovered in the Sirtuin Screening Showing SIRT2 Inhibitory Activity >60% at 12.5 µM
mercially available library of kinase and phosphatase inhibitors
(84 compounds), containing inhibitors of various structural
classes as well as CDK inhibitors (paullones), for their inhibitory
potency toward SIRT2. We identified new lead structures for
sirtuin inhibitors in several classes. Initial structure-activity
relationships were obtained for the bisindolylmaleimide group
of inhibitors (BIMs). The structural requirements for the binding
of ATP-mimics to the NAD+-binding site of sirtuins were
investigated by means of molecular docking and analyzing
favorable interaction fields using the program GRID.19 Selected
inhibitors from our screening were also investigated for selectiv-
ity toward SIRT1. Finally, hit validation was performed by
investigating cellular hyperacetylation of R-tubulin, a SIRT2
substrate.
The active compounds can be grouped into potentially redox-
active compounds 4, 5, 7, 8, and 10, Michael acceptors 6, 8, 9,
and 11, and indoles 6, 11, 12h, and 12j (see Charts 2 and 3).
Because the first two groups are not ideal lead structures due
to potential general bioreactivity, we focused on the indoles 12h
and 12j for further investigations. The BIMs are a class of
biologically active compounds originally identified in the 1980s
as ATP-competitive protein kinase C (PKC) inhibitors.23 Later
work identified the macrocyclic BIMs, such as LY333531
(ruboxistaurin) as potent, isoform-selective inhibitors of PKC-
beta, currently in phase III clinical trials for the treatment of
diabetic complications.24 We tested a set of BIMs 12a-k, 12m,
13a-c, and 14a,b (see Charts 3-5) related to the lead structures
12h and 12j to obtain structure-activity relationships. We found
that BIMs substituted on a single indole nitrogen (e.g., 12e-i)
were potent inhibitors of sirtuins and that these compounds were
significantly more potent than either the macrocyclic BIMs (12e
vs 13b, 12g vs 13a) or the indolocarbazoles (staurosporine 14a;
Go6976 14b). In contrast, the structural features of 13 and 14
led to substantially increased activity against protein kinases.25
Furthermore, BIMs bearing an alkyl group on the imide nitrogen,
which are inactive against many protein kinases,26-29 inhibit
SIRT2 (12b-d): methyl (12b) and benzyl substitution (12c)
decreased SIRT2 inhibition in comparison to the parent com-
pound 12a, but the IC50-values of the most active N-alkyl
compounds 12b and 12c (33 and 40 µM, respectively) were
still in the range of the physiological inhibitor nicotinamide (1),
determined to be 32.3 µM for SIRT2. Substituting the N-methyl
BIM 12b with simple alkyl or ω-hydroxyalkyl groups on both
indole nitrogens was detrimental to activity (12k, 12m).
However, substitution of the BIM 12a with a cyclohexylidene-
Results and Discussion
Enzyme Inhibition. To identify sirtuin inhibitors among
adenosine mimics, we screened a commercially available kinase
and phosphatase inhibitor library containing 84 compounds for
in vitro enzyme inhibition. We used a homogeneous fluorescent
assay,20 developed in our group, that utilizes a fluorescent
substrate for deacetylation12 and a tryptic digestion step21,22 to
quantitate enzymatic conversion. We identified 10 compounds
that had greater than 60% inhibition of SIRT2 at 12.5 µM,
including the stilbene piceatannol (4), a benzoic acid derivative
(5), the indolinone GW 5074 (6), rottlerin (7), an erbstatin
analogue (8), the naphthylketone ZM 449829 (9), a biphe-
nylpolyphenol (10), indirubin-3′-monooxime (11), and the two
BIMs GF 109203X (12h) and Ro 31-8220 (12j) (see Charts 2
and 3). Nicotinamide (1) and sirtinol (2) were tested as reference
compounds (Chart 1).

Adenosine Mimetics
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25 7309
Chart 4. Macrocyclic Bis(indolyl)maleimides
Chart 3. Bis(indolyl)maleimides
Chart 5. Indolocarbazoles and Paullones
bearing alcohol chain as in 12f and 12g produced inhibitors
with IC50 values of 2.8 and 2.5 µM, respectively, suggesting
that potent inhibitors can be designed based on nonpolar motifs.
The most active inhibitor was compound 12j,23 a disubstituted
BIM bearing an isothiourea structure in the side chain at one
indole nitrogen and a methyl substitution at the other, with an
IC50 of 0.8 µM. This is one of the most potent sirtuin inhibitors
described to date.
SIRT2 inhibition (see Table 1). Again, as in the BIM series,
alkylation of the lactam nitrogen strongly reduces kinase
inhibition,30 but here increased activity against SIRT2.
Selected inhibitors 12a, 12b, 12e-j, 13a-c, and 15b were
also tested for SIRT1 inhibition (see Table 1). All the BIMs
showed a marked selectivity toward SIRT2, while the paullone
15b did not.
Competition analysis revealed that the best inhibitor 12j is
competitive with respect to NAD+ (see Supporting Information).
Competition with the fluorescent substrate can only be analyzed
in a rather narrow concentration range due to substrate solubility
problems, but the concentration window that we investigated
shows no competition with the acetyl lysine derivative (data
Because of their known affinity for the NAD+-dependent
mMDH and the presence of an indole moiety, we tested a set
of three paullones for inhibition of SIRT2, including kenpaullone
(15a), a potent inhibitor of CDKs30 (see Chart 5). The prototype
CDK inhibitor kenpaullone displayed a very low inhibition of
Sirt2 at 100 µM, but benzylation of the lactam nitrogen (15b)
or the introduction of a hydroxyamidine structure (15c) increased

7310 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25
Table 1. Sirtuin Inhibition
Trapp et al.
% inhibition at concentration;
IC50 greater than concentration or IC50 + SE [µM]a
no.
SIRT1
SIRT2 [µM]
1
2
10.6% @ 50 µM
NTb
52.7 % @ 50 µM
>50
NT
NT
29.9% @ 50 µM
77.5% @ 50 µM
71.9% @ 50 µM
> 50
58.2% @ 50 µM
3.5 ( 0,4
NT
32.3 ( 6.6
53.0 ( 15.8
4.7 ( 1.1
33.0 ( 5.2
40.0 ( 3.0
78.3 ( 18.2
9.5 ( 1.5
2.8 ( 1.2
2.5 ( 0.6
7.3 ( 2.4
8.3 ( 0.5
0.8 ( 0.2
>100
12a
12b
12c
12d
12e
12f
12g
12h
12i
12j
12k
12m
13a
13b
13c
14a
14b
15a
15b
15c
NT
>100
55.7% @ 50 µM
NIc @ 50 µM
NI @ 50 µM
NT
NT
NT
20.8 ( 10.1
>50
>50
NI @ 50 µM
>100
>100
49.4% @ 50 µM
NT
42.8 ( 17.9
43.1% @ 50 µM
a Values are means ( SE of duplicate experiments. b Not tested. c No
inhibition.
Table 2. Inhibition of SIRT1 and -2 in Different Deacetylase Assays
fluorimetric assay,
scintillation assay,
IC50 [µM]
IC50 [µM]
compound
SIRT1
SIRT2
SIRT1
SIRT2
12b
12h
12j
>50
>50
3.5
∼50
33.0
7.3
0.8
15.7
23.0
5.1
24.1
11.7
1.1
Figure 1. Sir2-Af2-NAD+ complex. Interaction of NAD+ (carbon
atoms in green) with corresponding amino acid residues at the binding
pocket. (a) The GRID interaction field obtained with the aromatic probe
(contour level -2.5 kcal/mol, colored orange) is in perfect agreement
with the aromatic adenine ring system. Hydrogens bonds between
adenosine and SIRT2 are shown in orange. (b) The GRID field obtained
with the carbonyl probe (contour level -4.5 kcal/mol, colored magenta)
indicates the favorable hydrogen bond acceptor regions within the
binding pocket.
15b
42.8
8.0
10.0
Scheme 1. Synthesis of Bisalkylated BIMs
not shown). Thus, competition with the protein substrate cannot
be ruled out completely, but its apparent lack agrees with the
docking results (see below).
Because fluorescence-based deacetylase assays led to some
false positives during a screen for sirtuin activators,17,31,32 we
tested selected inhibitors with another biochemical assay relying
on scintillation counting after release of [3H] acetic acid from
the histone H4 N-terminal peptide (amino acids 1-25). This
radioactive assay confirmed that BIMs and paullone 15b are
potent inhibitors of human sirtuins (Table 2). The results of the
fluorescent and the radioactive assays are comparable. The
scintillation assay identified the paullone 15b as being more
active on SIRT1 and -2 than in the fluorescent assay (8-10
µM vs 42.8 µM, respectively). Also 12h was found to be more
active in the radioactive assay. 12b and 12h did not show
selectivity in the radioactive assay. The nature of the substrate
(peptide vs small molecule) may influence the results. But, as
the natural substrate is a histone in the context of a nucleosome
or a nonhistone protein and not an isolated peptide, it remains
to be determined which assay has the better predictive power
Figure 2. SIRT2-Ligand Complexes. Interaction of the adenosine
part of NAD+ (carbon atoms in green) and BIM 12j (magenta) with
the corresponding amino acid residues at the human SIRT2 binding
pocket. Hydrogen bonds are shown in orange.
for cellular effects. 12j is a potent inhibitor of SIRT2 in both
assays with a four- to fivefold selectivity over SIRT1.
The conformational flexibility of the BIM moiety has allowed
exquisite selectivity in terms of kinase inhibition, by constraining

Adenosine Mimetics
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25 7311
Figure 4. Docking results for BIM 12a. Comparison of the two
favorable docking solutions obtained for inhibitor 12a.
THP group in the case of 12m (Scheme 1). The monosubstituted
imides 12e-g were prepared by adapting the method of Faul
et al.33 (Scheme 2). Alkylation of indole acetamide with the
tosylate 16,34 followed by condensation of 17 with methyl
4-fluoroindole-3-glyoxylate and acidic workup, led to a mixture
of the triol 12e and the cyclohexylidene 12g. The nonfluorinated
compound 12f was prepared by an analogous route using methyl
indole-3-glyoxylate.35
Examination of the Putative Inhibitor Binding Site. All
Sir2 structures contain a conserved catalytic domain of 270
amino acids with variable N- and C-termini. Consistent with
the high sequence similarity in the catalytic domain, the available
structural data on the Sir2 proteins also show conservation in
the tertiary structure.36,37 The structure of the catalytic domain
consists of a large classical Rossmann-fold and a small zinc-
binding domain. The acetylated peptide binds in the cleft
between the two domains and forms an enzyme-substrate
â-sheet with two flanking strands from the enzyme. The acetyl-
lysine residue inserts into a conserved hydrophobic pocket, and
NAD+ binds nearby.37 The interaction of NAD+ at the binding
pocket can be examined in several sirtuin X-ray structures in
which the cofactor has been cocrystallized.37,38 Structural
comparison of available NAD+-sirtuin complexes from the
Protein Data Bank revealed a highly conserved and rigid NAD+
binding site.39
Figure 3. Docking results for BIMs. (a) Superimposition of docked
inhibitors 12a (orange), 12h (green-blue) and 12j (magenta). (b)
Superimposition of docked inhibitors 12a (orange) and paullone 15b
(purple). Hydrogen bonds are shown in orange.
the conformation. Our results demonstrate a similar potential
for sirtuin inhibition. The BIMs are important lead compounds
for the design of new, possibly subtype selective, sirtuin
inhibitors.
Chemistry. The BIMs 12d and 12k,l were prepared by
bisalkylation of the parent compound 12b with alkyl bromides
after deprotonation with NaH, followed by deprotection of the
First, by using the program GRID19 the NAD+ putative
binding site was examined to determine possible favorable
interactions. Since the NAD+ molecule in human SIRT2 was
Scheme 2. Synthesis of Monoalkylated BIMs

7312 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25
Trapp et al.
Figure 5. Favorable interaction fields (GRID) obtained with: (a) carbonyl probe (colored red at -4.5 kcal/mol), (b) aromatic probe (colored green
at -3.0 kcal/mol), (c) amine probe (colored red at -8.0 kcal/mol), and (d) amidine probe (colored green at -10.0 kcal/mol). The docked inhibitors
12a (colored green) and 12j (colored magenta) are shown for comparison. The molecular surface of the binding pocket is calculated using the
MOLCAD program colored according the cavity depth (blue ) exposed, orange ) buried).
not resolved, we analyzed the archaeal homologue Sir2-Af2.37
The calculated GRID contour maps were superimposed on the
crystal structure of Sir2-Af2 and compared with the positions
of the cocrystallized NAD+ molecule and the amino acids of
the active site (Figure 1A and 1B). On the basis of GRID probes,
the carbonyl group had an interaction energy of -5 kcal/mol
near the backbone NH of Cys324 and Asp325. The amide NH
probe yielded an interaction energy of -6 kcal/mol near the
carboxyl groups of Asp325 and Glu323 (data not shown). The
aromatic probe yielded a favorable interaction field between
the aliphatic portions of Thr89, Lys287, and Glu323 in perfect
agreement with the position of the aromatic adenine ring of
NAD+ as taken from the Sir2-Af2 X-ray structure (Figure 1A
and 1B).37 Because the NAD+ binding site is highly conserved,
similar GRID results were obtained for the human SIRT2
structure (data not shown).
Second, we tested whether the docking program GOLD40 can
reproduce the experimentally observed interaction of NAD+ at
Sir2-Af2. The NAD+ molecule and the less flexible adenosine
moiety were docked in the putative binding pocket near Cys324.
The observed low heavy atom rmsd value between top-ranked
docking solution and X-ray structure (0.75 Å for the adenosine
fragment and 1.06 Å for the whole NAD+ molecule) showed
that GOLD can correctly predict the bioactive conformation of
the adenosine moiety as well as for the complete NAD+ (see
Figure 6. In vivo SIRT2 inhibition. A549 cells were treated with the
indicated amounts of 12j, 15b, and TSA for 6 h and analyzed by western
blotting with antibodies against acetylated R-tubulin and R-tubulin.
Supporting Information). Similar results were obtained when
the adenosine fragment was docked in the corresponding pocket
of human SIRT2 (see Supporting Information). In the human
sirtuin, the adenine ring of adenosine is surrounded by several
residues (Gly86, Asn286, Lys287, Glu288, Glu323, Cys324,
and Asp325, see Figure 1) similar to those observed in the X-ray
structure of Sir2-Af2.37
In a subsequent step, the inhibitors were docked into this well-
defined binding pocket. The docking of the most potent inhibitor
12j revealed a second putative binding site for the polar
isothiourea group near Ile93 and Asp95 (Figure 2). The docking
solutions of four other potent inhibitors are shown within the
binding pocket in Figures 3A and 3B. As a relatively rigid
compound, 12a led to two reasonable docked conformations

Adenosine Mimetics
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25 7313
Conclusion
which are in agreement with the GRID maps. The two docking
solutions differ only in the orientation of one of the indole rings.
Figure 4 shows 12a bound in a folded and in a C2-symmetrical
conformation. This observation has also been made for BIMs
bound to kinases where different orientations of the indole rings
can be observed.41
The planar ring systems of the docked inhibitors fit perfectly
in the adenine-binding pocket and form hydrogen bonds with
Cys324 and Asp325. The maleimide ring of the potent BIMs
lies in a central position, in close proximity to Cys324. Due to
the rigidity of staurosporine (14a), it was not possible to obtain
a favorable docking solution within the adenine-binding pocket.
Thus, a certain flexibility within the BIMs is needed so that
the two indole rings can adopt the appropriate conformation.
This observation is in agreement with the observed inactivity
of compounds 14a and 14b against SIRT2.
The docking results showed that the active inhibitors can
interact in similar ways with the adenine-binding pocket.
However, on the basis of the docking results (i.e., Goldscore),
no discrimination could be derived between more and less active
inhibitors. This is often observed when dealing with docking
scores.42 Therefore, we focused on a qualitative analysis of key
interactions necessary for high inhibitory activity.
A common feature of all docked inhibitors is the strong
interaction of the aromatic ring moiety with the hydrophobic
portions of residues Thr89, Lys287, Glu323, and Cys324. This
type of interaction can also be observed for the adenine base of
NAD+. Figure 2 shows the superimposition of 12j and the
docked adenosine fragment. The indole ring of 12j occupies
the same region in space as the adenine ring of NAD+. In
addition, all docked active inhibitors show a strong hydrogen
bond of the carbonyl group and the backbone NH of Cys324.
The carbonyl groups are located in the same area as the
favorable GRID interaction field of the carbonyl probe (Figure
5A), suggesting the importance of an interaction between the
inhibitor’s hydrogen bond acceptor group and the backbone NHs
of Cys324.
The most potent BIMs bear an additional polar or basic side
chain attached to one of the indole rings. This polar or basic
group interacts with a second sub-pocket nearby Ile93 and
Thr89. Favorable interactions within this subpocket were
detected using the GRID amidine and amine probe (Figure 5A-
D). The hydroxyl probe led to similar results as the amine probe
(data not shown). The potent BIM derivatives interact with this
second polar binding site by making electrostatic interactions
with Asp95 (12h, 12i) or by making hydrogen bonds to the
backbone NH of Ile93 (12j, Figure 3). Due to the high flexibility
of the two side chains of the inactive BIMs, 12k and 12m, no
suitable docking results were obtained for these molecules.
Validation of Sirtuin Inhibition in Cell Culture. To
evaluate the in ViVo inhibition of SIRT2, compounds 12j and
15b were tested for their abilities to induce hyperacetylation of
the SIRT2 target tubulin.43 A549 human lung adenocarcinoma
cells were treated with compounds 12j and 15b for 6 h. Western
blot analysis using an antibody to the acetylated form of tubulin
showed that compound 12j and to a lesser extent 15b induced
hyperacetylation of tubulin (Figure 6). Tubulin was also
hyperacetylated by 0.1 µM TSA, which inhibits class I and II
HDACs, including the tubulin deacetylase HDAC6. Certain
combinations of compound 12j and 15b with TSA did not
induce greater levels of tubulin acetylation (Figure 6). More
detailed studies would be necessary to prove or rule out additive
or synergistic effects. Thus, compounds 12j and 15b inhibit
SIRT2 in ViVo and induce hyperacetylation of tubulin.
Random screening and target-fishing approaches had previ-
ously identified adenosine mimics, such as suramin, and kinase
inhibitors, namely the paullones, as compounds that also target
NAD+-dependent enzymes. We developed a systematic ap-
proach to establish structure-activity relationships between
selective inhibitors and different classes of enzymes that use
adenosine-containing cofactors. We show that an overlap
between ATP and NAD+ mimesis by adenosine mimetics is
more general than previously suspected. As several of the
micromolar sirtuin inhibitors are more potent as kinase inhibi-
tors, selectivity is most likely not an issue in those instances.
Nevertheless, the first kinase inhibitors have already obtained
approval as drugs, and as many more are in clinical trials, the
investigation of selectivity should not be limited too readily to
kinases. This is especially true for a potential inhibition of
sirtuins due to their general involvement in aging.44 Additionally,
the IC50 values obtained for kinase inhibition are dependent on
the ATP concentration used in the in Vitro assays. Often these
ATP concentrations are considerably lower than in the cell, and
thus, such assays tend to overestimate kinase inhibitory po-
tency.29
Our results have implications for sirtuin inhibitor design. We
identified new lead structures in the low- or even sub-
micromolar range. 12j is one of the most potent sirtuin inhibitors
reported to date. This supports the use of our nonradioactive
screening assay as a tool for sirtuin inhibitor discovery. Among
a series of related PKC inhibitors, only 12j led to the induction
of apoptosis at 10 µM,45 and in this concentration range we
observe cellular tubulin hyperacetylation. Although it is tempting
to tie these proapoptotic properties to sirtuin inhibition, further
studies are needed to substantiate these assumptions.
Additionally, with N-alkylation and in particular N-benzy-
lation, we identified a promising structural descriptor for
achieving sirtuin over kinase inhibition. Cell-permeable inhibi-
tors, such as 12b (bisindolylmaleimide V) and 12c, that are
inactive as inhibitors for most kinases,29 may be useful for
cellular studies aimed at determining the biological roles of
sirtuins. These results provide the outline of a synthetic program
aimed at combining features of sirtuin inhibition potency and
selectivity for improved sirtuin inhibitor design. The modeled
protein-inhibitor complexes will be a valuable tool toward this
end.
Experimental Section
Materials. All chemicals were purchased from Sigma, Aldrich,
or Fluka and used without further purification. The kinase and
phosphatase inhibitor library was purchased from BIOMOL Inter-
national.
Inhibitors. Paullones were synthesized as described (15a;46
15b,c30). BIM 12c47 and macrocycles 13a, 13b, 13c were synthe-
sized as described.34 12a,b, 12h-j, and 14a,b were purchased from
Calbiochem.
General Procedure for the Synthesis of Bisindolylmaleimides
12d, 12k, 12l. To a solution of the bisindolylmaleimide (1 equiv)
in DMF (400 µL/0.14 mmol) were added NaH (2.2 eq, 60%
dispersion in mineral oil) and the alkyl halide (2 equiv). The mixture
was stirred at room temperature for 2 h. The mixture was quenched
with saturated aqueous NH4Cl and extracted with CH2Cl2. The
organic layer was washed with brine, dried over Na2SO4, filtered,
and concentrated to dryness. Purification (SiO2: 0-50% gradient
of EtOAc in CH2Cl2) afforded the compound as a red solid (For
NMR and purity data of all compounds, see Supporting Informa-
tion).
12d: Yield 69%; MS (CI, isobutane) m/z 522 +; HRMS (CI,
isobutane) calcd for C35H28N3O2 522.2182, found m/z 522.2183.

7314 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25
Trapp et al.
12k: Yield 30%; MS (CI, isobutane) m/z 538 [MH]+; HRMS
(CI, isobutane) calcd for C35H44N3O2 538.3433, found m/z 538.3434.
12l: Yield 27%; MS (FAB, NBA) m/z 709 [M]+; HRMS (CI,
isobutane) calcd for C43H55N3O6 709.4091, found m/z 709.4094.
Bisindolylmaleimide 12m. p-Toluenesulfonic acid (1 mg) was
added to a solution of 12l (27 mg, 0.04 mmol) in MeOH (1 mL).
The mixture was stirred at room temperature for 2 h. The mixture
was quenched with saturated aqueous NH4Cl and extracted with
EtOAC. The organic layer was washed with brine, dried over Na2-
SO4, filtered, and concentrated to dryness. Purification (SiO2:
0-50% gradient of EtOAc in CH2Cl2) afforded 12m (6 mg, 28%)
as a red solid. MS (FAB, glycerol) m/z 541 [M]+, 542 [M + H]+;
HRMS (FAB, glycerol) calcd for C33H39N3O4 541.2941, found m/z
541.2942.
Indole Acetamide Derivative 17. To a solution of NaH (39 mg,
1.6 mmol, 60% dispersion in mineral oil) in DMF (300 µL) at 0
°C was added indole-3-acetamide (224 mg, 1.3 mmol), and the
mixture was allowed to warm to room temperature and stir for 30
min. Tosylate 1634 (500 mg, 0.76 mmol) in DMF (300 µL) was
added, and the reaction mixture was stirred at 70 °C for 18 h. The
reaction was cooled to room temperature, quenched with saturated
aqueous NH4Cl, and diluted with EtOAc. The organic layer was
washed with brine, dried over Na2SO4, filtered, and concentrated
to dryness. Purification (SiO2: EtOAc) led to the isolation of the
indole acetamide derivative 17 (387 mg, 77%) as a yellow oil. MS
(LSIMS, NBA) m/z 658 [M]+, 681 [M + Na]+; HRMS (LSIMS,
NBA) calcd for C42H46N2O5 658.3407, found m/z 658.3406.
Bisindolylmaleimides 12e and 12g. To a solution of 17 (400
mg, 0.61 mmol) and methyl 4-fluoroindole-3-glyoxylate (376 mg,
1.70 mmol)33 in a mixture of THF:DMF (2:1, 5.7 mL) was added
potassium tert-butoxide (3.60 mL, 1 M in THF, 3.6 mmol), and
the reaction mixture was stirred at 70 °C for 5 h. The mixture was
quenched with saturated aqueous NH4Cl, and diluted with EtOAc.
The organic layer was washed with H2O, dried over Na2SO4,
filtered, and concentrated to dryness to afford the alcohol (500 mg).
The alcohol (161 mg, 0.19 mmol) was dissolved in acetic acid
(3 mL), and the reaction mixture was stirred at 56 °C for 6 h. The
mixture was quenched with H2O and diluted with EtOAc. The
organic layer was washed with brine and saturated aqueous
NaHCO3, dried over Na2SO4, filtered, and concentrated to dryness.
Purification (SiO2: 100% EtOAc) afforded the cyclohexylidene 12g
(45 mg, 39%) as a red oil and the diol 12e (6 mg, 8%) as a red oil.
12g: MS (LSIMS, thioglycerol) m/z 587 [M]+; HRMS (LSIMS,
thioglycerol) calcd for C33H34FN3O6 587.2431, found m/z 587.2434.
12e: MS (LSIMS, thioglycerol) m/z 507 [M]+, 508 [MH]+;
HRMS (LSIMS, thioglycerol) calcd for C27H26FN3O6 507.1806,
found m/z 507.1805.
force field and the GB/SA continuum50 solvent model for water
was carried out. During the minimization, a tethering constant of
100 kcal/(mol Å) on the backbone atoms was applied, after a
stepwise reduction of the tethering to 1 kcal/(mol Å).
Interaction possibilities were analyzed using the GRID program
(Molecular Discovery Inc.). GRID is an approach to predict
noncovalent interactions between a molecule of known three-
dimensional structure (i.e., a sirtuin) and a small group as a probe
(representing chemical features of a ligand).19 The calculations were
performed using version 22 of the GRID program and the crystal
structures mentioned above. The calculations were performed on a
cube (20 × 20 × 20 Å, spacing 1 Å), including the NAD+ binding
pocket, to search for binding sites complementary to the functional
groups of the inhibitors. The following probes were used for
calculation: carbonyl, aromatic, amidine, and amine probes. The
calculated GRID contour maps were then viewed superimposed
on the crystal structure of the sirtuins with the MOE program.
Docking of the cocrystallized NAD+ and adenosine, as well as
the tested inhibitors, was carried out using program GOLD 2.3.40
All torsion angles in each compound were allowed to rotate freely.
Goldscore was chosen as the fitness function. For each molecule,
30 docking runs were performed. The resulting solutions were
clustered on the basis of the heavy atom rmsd values (0.75 Å).
The top-ranked poses for each ligand were retained and viewed
graphically within MOE. The docking solution that showed the best
agreement with the calculated GRID maps was manually selected
for each inhibitor.
Recombinant Proteins. Human SIRT2 (N-terminally tagged
with 6 His) was prepared as described51 with minor modifications.
In brief, the plasmid pEV1440, containing the full-length human
SIRT2 cDNA was transformed in Escherichia coli strain BL21 for
expression. The culture was grown in LB medium to an optical
density of 0.6 (A600) at 37 °C, induced with 0.1 mM IPTG for 2 h,
and pelleted. Lysis was performed with a French press. The soluble
overexpressed recombinant protein was purified using Ni-NTA
resin. The identity of the produced His-SIRT2 was verified by SDS
electrophoresis. Deacetylase activity of the produced SIRT2 was
dependent on NAD+ and could be inhibited with sirtinol and
nicotinamide. The human sirtuin hSIRT1 was purchased from
Biomol (activity: 3.5 U/µL).
Fluorescent Deacetylase Assay. All compounds were evaluated
for their abilities to inhibit recombinant sirtuins in a homogeneous
fluorescent deacetylase assay.20 Stock solutions of inhibitors were
prepared in DMSO, and 3 µL or less of a suited DMSO inhibitor
solution was added to the incubation mixture. The assay was carried
out in 96-well plates: 60 µL reaction volume contained the
fluorescent histone deacetylase substrate ZMAL (10.5 µM), NAD+
(500 µM), and Sirt2 (3 µL) or Sirt1 (1.2 µL). After 4 h of incubation
at 37 °C, the deacetylation reaction was stopped, and the metabolite
ZML (deacetylated form of ZMAL) was developed using a tryptic
digest to form a fluorophore. Fluorescence was measured in a plate
reader (BMG Polarstar) with a coumarin filter (excitation 390 nm
and emission 460 nm). The amount of remaining substrate in the
positive control with inhibitor versus negative control without
inhibitor was employed to calculate inhibition. All determinations
were carried out at least in duplicate. IC50 data were analyzed using
GraphPad Prism Software.
Bisindolylmaleimide 12f. The nonfluorinated bisindolylmale-
imide 12f was prepared by the same procedure. 12f (Yield: 17%):
MS (LSIMS, glycerol) m/z 569 [M]+; HRMS (LSIMS, glycerol)
calcd for C33H35N3O6 569.2526, found m/z 569.2524.
Molecular Modeling. All calculations were performed on a
Pentium IV 1.8 GHz based Linux cluster. The molecular structures
of the inhibitors were generated using the MOE modeling package
(Chemical Computing Group).48 The structures were energy
minimized using the MMFF94s force field and the conjugate
gradient method, until the default derivative convergence criterion
of 0.01 kcal/(mol Å) was met. The crystal structures of human
SIRT2 (pdb code 1J8F)36 and archaeal Sir2-Af2(pdb code 1YC2)37
were taken from the Protein Data Bank. SIRT2 is a monomer in
solution, and therefore only chain B was chosen from the trimeric
SIRT2 structure of 1J8F. Monomer B was selected for SIRT2 and
monomer C for Sir2-Af2, as they showed the best stereochemical
quality examined with the program PROCHECK.49 In addition to
the uncomplexed form of human SIRT2, the archaeal Sir2-Af2
crystal structure was used for the current investigation to inspect
the NAD+-enzyme interaction. Since the X-ray structures might
contain residual energetic tensions from the crystallization process,
both structures were first energy minimized. After removal of the
cocrystallized water molecules and addition of hydrogen atoms to
the protein structure, a descent minimization using the MMF94s
Radioactive Deacetylase Assay. The deacetylase assay with
recombinant SIRT1 and 2 was performed as described. [3H]-
Acetylated H4 histone N-terminal peptide (residues 1-25) was used
as peptide substrate.51
Analysis of Tubulin Acetylation. A549 human lung carcinoma
cells were cultured in Dulbecco’s modified Eagle’s medium
(DMEM; Mediatech, Herndon, VA) supplemented with 10% fetal
bovine serum (Gemini Bio-products, Woodland, CA), 1% penicil-
lin-streptomycin, and 2 mM l-glutamine (GIBCO Invitrogen
Corporation). A549 cells (50K/well) were plated in 12-well dishes
1 day before treatment with 12j, 15b, and Trichostatin A (TSA;
Wako Chemicals USA, Richmond, VA). After 6 h of treatment,
cells were washed with PBS and lysed in 50 mM Tris-HCl, pH
7.5, 0.5 mM EDTA, 150 mM NaCl, 0.5% NP-40, 400 nM TSA,

Adenosine Mimetics
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25 7315
(20) Heltweg, B.; Trapp, J.; Jung, M. In vitro assays for the determination
of histone deacetylase activity. Methods 2005, 36, 332-337.
(21) Wegener, D.; Wirsching, F.; Riester, D.; Schwienhorst, A. A
fluorogenic histone deacetylase assay well suited for high-throughput
activity screening. Chem. Biol. 2003, 10, 61-68.
(22) Wegener, D.; Hildmann, C.; Riester, D.; Schwienhorst, A. Improved
fluorogenic histone deacetylase assay for high-throughput screening
applications. Anal. Biochem. 2003, 321, 202-208.
(23) Davis, P. D.; Elliott, L. H.; Harris, W.; Hill, C. H.; Hurst, S. A. et
al. Inhibitors of protein kinase C. 2. Substituted bisindolylmaleimides
with improved potency and selectivity. J. Med. Chem. 1992, 35, 994-
1001.
(24) Ishii, H.; Jirousek, M. R.; Koya, D.; Takagi, C.; Xia, P. et al.
Amelioration of vascular dysfunctions in diabetic rats by an oral PKC
beta inhibitor. Science 1996, 272, 728-731.
10 mM nicotinamide, and 1x complete protease inhibitors (Roche,
Penzberg, Germany). Protein concentration was determined with
the DC Protein Assay (Bio-Rad). Protein samples were separated
by electrophoresis on 10% SDS-polyacrylamide gels and transferred
to a nitrocellulose membrane (Bio-Rad). Membranes were blocked
with 5% nonfat dry milk in TBS-Tween (10 mM Tris-HCl, pH
7.5, 150 mM NaCl, and 0.1% Tween-20) and probed with anti-
acetylated R-tubulin (6-11B-1; Sigma) and anti-R-tubulin (B-5-1-
2; Sigma) at 1:2000.
Acknowledgment. Funding by the European Commission
(Contract No LSHB-CT-2004-503467, to P.G. and C.K.) and
the Albert-Ludwigs-University of Freiburg (Wilhelm-Seiter-
Stiftung and Eugen-Graetz-Preis, to M.J.) is gratefully acknowl-
edged.
(25) Goekjian, P. G.; Jirousek, M. R. Protein kinase C in the treatment of
disease: signal transduction pathways, inhibitors, and agents in
development. Curr. Med. Chem. 1999, 6, 877-903.
(26) Marmy-Conus, N.; Hannan, K. M.; Pearson, R. B. Ro 31-6045.,
the inactive analogue of the protein kinase C inhibitor Ro 31-8220.,
blocks in vivo activation of p70(s6k)/p85(s6k): implications for the
analysis of S6K signalling. FEBS Lett. 2002, 519, 135-140.
(27) Toullec, D.; Pianetti, P.; Coste, H.; Bellevergue, P.; Grand-Perret,
T. et al. The bisindolylmaleimide GF 109203X is a potent and
selective inhibitor of protein kinase C. J. Biol. Chem. 1991, 266,
15771-15781.
(28) Davis, P. D.; Hill, C. H.; Lawton, G.; Nixon, J. S.; Wilkinson, S. E.
et al. Inhibitors of protein kinase C. 1. 2,3-Bisarylmaleimides. J. Med.
Chem. 1992, 35, 177-184.
(29) Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P. Specificity and
mechanism of action of some commonly used protein kinase
inhibitors. Biochem. J. 2000, 351, 95-105.
(30) Schultz, C.; Link, A.; Leost, M.; Zaharevitz, D. W.; Gussio, R. et al.
Paullones, a series of cyclin-dependent kinase inhibitors: synthesis,
evaluation of CDK1/cyclin B inhibition, and in vitro antitumor
activity. J. Med. Chem. 1999, 42, 2909-2919.
(31) Kaeberlein, M.; McDonagh, T.; Heltweg, B.; Hixon, J.; Westman,
E. A. et al. Substrate specific activation of sirtuins by resveratrol. J.
Biol. Chem. 2005, 280, 17038-17045.
(32) Borra, M. T.; Smith, B. C.; Denu, J. M. Mechanism of human SIRT1
activation by resveratrol. J. Biol. Chem. 2005, 280, 17187-17195.
(33) Faul, M. M.; Winneroski, L. L.; Krumrich, C. A. Synthesis of
Rebeccamycin and 11-Dechlororebeccamycin. J. Org. Chem. 1999,
64, 2465-2470.
(34) Liu, Z. Design and Synthesis of Conformationally Restricted Indole-
Fluorinated Macrocyclic Bis(indolyl)maleimides. M.Sc. Thesis, Mis-
sissippi State University, 2001.
(35) Goekjian, P. G.; Wu, G.-Z.; Chen, S.; Zhou, L.; Jirousek, M. R. et
al. The Synthesis of Fluorinated Macrocyclic Bis(indolyl)maleimides
as Potential 19F-NMR Probes for PKC. J. Org. Chem. 1999, 64,
4238-4246.
(36) Finnin, M. S.; Donigian, J. R.; Pavletich, N. P. Structure of the histone
deacetylase SIRT2. Nat. Struct. Biol. 2001, 8, 621-625.
(37) Avalos, J. L.; Bever, K. M.; Wolberger, C. Mechanism of sirtuin
inhibition by nicotinamide: altering the NAD(+) cosubstrate speci-
ficity of a Sir2 enzyme. Mol. Cell 2005, 17, 855-868.
(38) Zhao, K.; Harshaw, R.; Chai, X.; Marmorstein, R. Structural basis
for nicotinamide cleavage and ADP-ribose transfer by NAD(+)-
dependent Sir2 histone/protein deacetylases. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 8563-8568.
(39) Denu, J. M. The Sir2 family of protein deacetylases. Curr. Opin.
Chem. Biol. 2005.
(40) Jones, G.; Willet, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J. Mol. Biol. 1997, 267, 727-748.
(41) Komander, D.; Kular, G. S.; Schuttelkopf, A. W.; Deak, M.; Prakash,
K. R. et al. Interactions of LY333531 and other bisindolyl maleimide
inhibitors with PDK1. Structure 2004, 12, 215-226.
(42) Tame, J. R. Scoring functions-the first 100 years. J. Comput.-Aided
Mol. Des. 2005, 19, 445-451.
Supporting Information Available: Additional spectral data
for compounds 12d-m and 17 and information on compound
purity, kinetic analysis, and docking studies. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Grozinger, C. M.; Schreiber, S. L. Deacetylase enzymes: Biological
functions and the use of small-molecule inhibitors. Chem. Biol. 2002,
9, 3-16.
(2) North, B. J.; Verdin, E. Sirtuins: Sir2-related NAD-dependent protein
deacetylases. Genome Biol. 2004, 5, 224.
(3) Imai, S.; Armstrong, C. M.; Kaeberlein, M.; Guarente, L. Transcrip-
tional silencing and longevity protein Sir2 is an NAD- dependent
histone deacetylase. Nature 2000, 403, 795-800.
(4) Pagans, S.; Pedal, A.; North, B. J.; Kaehlcke, K.; Marshall, B. L. et
al. SIRT1 Regulates HIV Transcription via Tat Deacetylation. PLoS
Biol. 2005, 3, e41.
(5) Vaziri, H.; Dessain, S. K.; Ng, Eaton, E.; Imai, S. I.; Frye, R. A. et
al. hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase.
Cell 2001, 107, 149-159.
(6) Bereshchenko, O. R.; Gu, W.; Dalla-Favera, R. Acetylation inactivates
the transcriptional repressor BCL6. Nat. Genet. 2002, 32, 606-613.
(7) Ota, H.; Tokunaga, E.; Chang, K.; Hikasa, M.; Iijima, K. et al. Sirt1
inhibitor, Sirtinol, induces senescence-like growth arrest with at-
tenuated Ras-MAPK signaling in human cancer cells. Oncogene 2005.
(8) Johnstone, R. W. Histone-deacetylase inhibitors: novel drugs for the
treatment of cancer. Nat. ReV. Drug DiscoVery 2002, 1, 287-299.
(9) Scha¨fer, S.; Jung, M. Chromatin modifications as targets for new
anticancer drugs. Arch. Pharm. Chem. Life Sci. 2005, 338, 347-
357.
(10) Biel, M.; Wascholowski, V.; Giannis, A. Epigenetics - An epicenter
of gene regulation: Histones and histone-modifying enzymes. Angew.
Chem., Int. Ed. 2005, 44, 3186-3216.
(11) Grozinger, C. M.; Chao, E. D.; Blackwell, H. E.; Moazed, D.;
Schreiber, S. L. Identification of a class of small molecule inhibitors
of the sirtuin family of NAD-dependent deacetylases by phenotypic
screening. J. Biol. Chem. 2001, 276, 38837-38843.
(12) Heltweg, B.; Dequiedt, F.; Verdin, E.; Jung, M. A non isotopic
substrate for assaying both human zinc and NAD+-dependent histone
deacetylases. Anal. Biochem. 2003, 319, 42-48.
(13) Mai, A.; Massa, S.; Lavu, S.; Pezzi, R.; Simeoni, S. et al. Design,
synthesis and biological evaluation of sirtinol analogues as class III
histone/protein deacetylase (sirtuin) inhibitors. J. Med. Chem. 2005,
48, 7789-7795.
(14) Bedalov, A.; Gatbonton, T.; Irvine, W. P.; Gottschling, D. E.; Simon,
J. A. Identification of a small molecule inhibitor of Sir2p. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 15113-15118.
(15) Tervo, A. J.; Kyrylenko, S.; Niskanen, P.; Salminen, A.; Leppanen,
J. et al. An in silico approach to discovering novel inhibitors of human
sirtuin type 2. J. Med. Chem. 2004, 47, 6292-6298.
(16) Napper, A. D.; Hixon, J.; McDonagh, T.; Keavey, K.; Pons, J. F. et
al. Discovery of indoles as potent and selective inhibitors of the
deacetylase SIRT1. J. Med. Chem. 2005, 48, 8045-8054.
(17) Howitz, K. T.; Bitterman, K. J.; Cohen, H. Y.; Lamming, D. W.;
Lavu, S. et al. Small molecule activators of sirtuins extend Saccha-
romyces cerevisiae lifespan. Nature 2003, 425, 191-196.
(18) Knockaert, M.; Wieking, K.; Schmitt, S.; Leost, M.; Grant, K. M. et
al. Intracellular Targets of Paullones. Identification following affinity
purification on immobilized inhibitor. J. Biol. Chem. 2002, 277,
25493-25501.
(43) North, B. J.; Marshall, B. L.; Borra, M. T.; Denu, J. M.; Verdin, E.
The human Sir2 ortholog, SIRT2, is an NAD+-dependent tubulin
deacetylase. Mol. Cell 2003, 11, 437-444.
(44) Guarente, L. Sir2 links chromatin silencing, metabolism, and aging.
Genes DeV. 2000, 14, 1021-1026.
(45) Han, Z.; Pantazis, P.; Lange, T. S.; Wyche, J. H.; Hendrickson, E.
A. The staurosporine analog, Ro-31-8220., induces apoptosis
independently of its ability to inhibit protein kinase C. Cell Death
Differ. 2000, 7, 521-530.
(46) Kunick, C. Synthese von 7,12-Dihydro-indolo[3,2-d][1]benzazepin-
6(5H)-onen und 6,11-Dihydro-thieno-[3′,2′:2,3]azepino[4,5-b]indol-
5(4H)-on. Arch. Pharm. 1992, 325, 297-299.
(19) Goodford, P. J. A computational procedure for determining energeti-
cally favorable binding sites on biologically important macromol-
ecules. J. Med. Chem. 1985, 28, 849-857.

7316 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 25
Trapp et al.
(47) Joyce, R. P.; Gainor, J. A.; Weinreb, S. M. Synthesis of the aromatic
and monosaccharide moieties of staurosporine. J. Org. Chem. 1987,
52, 1177-1185.
(50) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
Semianalytical treatment of solvation for molecular mechanics and
dynamics. J. Am. Chem. Soc. 1990, 112, 6127-6129.
(51) North, B. J.; Schwer, B.; Ahuja, N.; Marshall, B.; Verdin, E.
Preparation of enzymatically active recombinant class III protein
deacetylases. Methods 2005, 36, 338-345.
(48) Molecular Operating EnVironment (MOE); Chemical Computing
Group Inc.: Montreal, Quebec, Canada, 2005.
(49) Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M.
PROCHECK: A program to check the stereochemical quality of
protein structures. J. Appl. Crystallogr. 1993, 26, 283-291.
JM060118B