Inhibitors of the NAD+-dependent protein desuccinylase and demalonylase sirt5

Article abstract of DOI:10.1021/ml3002709

NAD⁺-dependent histone deacetylases (sirtuins) play important roles in epigenetic regulation but also through nonhistone substrates for other key cellular events and have been linked to the pathogenesis of cancer, neurodegeneration, and metabol

Full text of DOI:10.1021/ml3002709

Letter
pubs.acs.org/acsmedchemlett
Inhibitors of the NAD+‑Dependent Protein Desuccinylase and
Demalonylase Sirt5
Benjamin Maurer,† Tobias Rumpf,† Michael Scharfe,‡ Diana A. Stolfa,† Martin L. Schmitt,†
Wenjuan He,§ Eric Verdin,§ Wolfgang Sippl,‡,∥ and Manfred Jung*,†,∥
†Institute of Pharmaceutical Sciences, Albert-Ludwigs Universitat, Albertstrasse 25, 79104 Freiburg, Germany
̈
‡Department of Pharmaceutical Chemistry, Martin-Luther Universitat Halle-Wittenberg, Universitatsplatz 7, 06120 Halle/Saale,
̈
̈
Germany
§Gladstone Institute of Virology and Immunology, 1650 Owens Street, San Francisco, California 94158, United States
S
* Supporting Information
ABSTRACT: NAD+-dependent histone deacetylases (sir-
tuins) play important roles in epigenetic regulation but also
through nonhistone substrates for other key cellular events and
have been linked to the pathogenesis of cancer, neuro-
degeneration, and metabolic diseases. The subtype Sirt5 has
been shown recently to act as a desuccinylating and
demalonylating enzyme. We have established an assay for
biochemical testing of Sirt5 using a small labeled succinylated
lysine derivative. We present a comparative study on the
profiling of several established sirtuin inhibitors on Sirt1−3 as
well as Sirt5 and also present initial results on a screening for
new compounds that block Sirt5. Thiobarbiturates were
identified as new Sirt5 inhibitors in the low micromolar range, which are selective over Sirt3 that can be found in the same
cell compartment as Sirt5.
KEYWORDS: sirtuin, Sirt5, desuccinylation, sirtinol, AMC assay
irtuins are nicotinamide adenine dinucleotide (NAD+)-
Various assay formats have been used for the detection of
Sdependent protein deacylases that are involved in many Sirt5 activity and its inhibition in vitro. This involves the use
of a small molecule lysine derivative,14 mass spectrometry,7
fundamental cellular processes. For some of them, adenosine
HPLC/absorbance,9 and 32P-NAD-thin-layer chromatography.7
diphosphate (ADP)-ribosyltransferase activity has also been
shown.1−4 Inhibitors and activators of sirtuins have been
We have synthesized an analogue of a known small molecule
substrate, namely, Z-Lys(succ)-aminomethyl coumarin (AMC)
{here termed 3-[(5-{[(benzyloxy)carbonyl]amino}-5-[(4-
methyl-2-oxo-2H-chromen-7-yl)carbamoyl]pentyl)carbamoyl]-
propanoic acid (ZK(s)A)} (synthesis scheme and spectral data
for intermediate compounds can be found in the Supporting
Information).
We used a Z (CbZ) protecting group rather than a Boc
derivative that has already been used by another group in a
succinyl lysine derivative,14 allowing us a better comparison of
the inhibition values of Sirt1−3, which have been obtained with
the same inhibitors using an acetylated analogue as the substrate.
We compared the ability of trypsin to cleave ZK(s)A and its
desuccinylated metabolite benzyl N-{5-amino-1-[(4-methyl-2-
oxo-2H-chromen-7-yl)carbamoyl]pentyl}carbamate (ZKA) and
discovered that ZK(s)A was very stable even under a high trypsin
concentration (1.5 mg/mL for over 20 min). In contrast, ZKA as
proposed as potential new drugs for the treatment of a variety of
diseases. Sirtuins have mainly been characterized as deacetylases,
but recent examples show that selected subtypes are able to
cleave off other acyl groups. This has been shown on small
molecule substrates for the phenacetyl group5 but also in vivo for
long chain aliphatic acids for the Plasmodium falciparum Sir2A.6
Especially interesting is the fact that Sirt5 recently has been
shown to have demalonylase and desuccinylase activity in vitro
and in vivo.7,8 Initial inhibitor studies have shown that
thiosuccinylated peptides act as inhibitors9 similar to thioacety-
lated congeners for other sirtuins.10−12 Also, the standard sirtuin
inhibitor suramin seems to possess some activity against Sirt5,
which was shown in assays using acetylated, respectively,
succinylated substrates.9,13 Because of the involvement of Sirt5
in primary metabolism, development of selective inhibitors of
that isotype is an attractive area for characterizing the physio-
logical function and therapeutical potential. In this paper, we
present a modified in vitro assay that has been optimized for Sirt5
and results for profiling of a set of reference sirtuin inhibitors as
well as initial results of a screening campaign.
Received: September 4, 2012
Accepted: October 6, 2012
Published: October 6, 2012
© 2012 American Chemical Society
1050
dx.doi.org/10.1021/ml3002709 | ACS Med. Chem. Lett. 2012, 3, 1050−1053

ACS Medicinal Chemistry Letters
Letter
shown before can easily be cleaved by trypsin, and the amount
of released AMC can be measured because of the shift in
fluorescence wavelength. It has been shown before that εN-
modifications, like acetyl or succinyl on this kind of lysine
substrates, prevent tryptic cleavage15,16 (see Scheme 1).
Chart 1. Standard Sirtuin Inhibitors
Scheme 1. Sirt5 Activity Assay Using a Small Molecule
Substrate
We used a concentration of 200 μM ZK(s)A and an enzyme
concentration leading to a conversion of 5−10% in 1 h.
Converting more than 5% of ZK(s)A to AMC led to a very
stable and strong fluorescence signal, which could be measured
at λEx = 390 nm and λEm = 460 nm in a microtiter plate reader.
The kinetic constants were measured by end-point determina-
tion using the same assay with ZK(s)A concentrations ranging
from 1 to 1000 μM. The KM value for ZK(s)A was 17.4
1.8 μM, and the Vmax was 5.3 × 10−3 μM. A published
succinylated H3K9 peptide is reported to have a KM value of
5.8 2.7 μM for Sirt5,7 whereas other AMC derivatives like
H410−12-K(succ)-AMC or Ac-Lys(succ)-AMC showed slightly
lower values with 33 1.8 and 84 22 μM, respectively.17
We then tested a variety of reference sirtuin inhibitors for
Sirt5 inhibition and compared the results to the inhibition
values for Sirt1−3 using the reported acetylated substrate. The
reference inhibitors are Nicotinamide, Sirtinol,18 Cambinol,19
Suramin,20 EX-527 (Selisistat),21 a splitomicin derivative (HR
103),22 and Ro31-822015 (see Chart 1).
Table 1. IC50 Values of Standard Sirtuin Inhibitors (μM):
Inhibitory Values for Sirt1−3 from Ref 23 and Sirt5 from
This Study
compd
Sirt1
Sirt2
Sirt3
Sirt5
nicotinamide
sirtinol
50−100
37−131
1.2−100 30
46.6 3.0
48.9 6.3
38−58
24% at
50 μM
cambinol
56
59
NI
42.5 1.5
suramin
0.3−2.6
0.1−1
NI
1.1−20
20−33
1−5
NI
49
46.6 4.8
EX-527
NI
NI
NI
HR 103 (5a in lit.22
Ro 31-8220
)
NI
3.7
3.5
0.8
Table 2. IC50 Values of Thiobarbiturate Compounds
compd
Sirt1
Sirt2
Sirt3
Sirt5
As shown in Table 1, the relative potency of the reference
compounds on Sirt5 varies. While Cambinol shows similar
activity on Sirt5 as on Sirt1 and -2, the splitomicin derivative
HR 103 seems to be selective on its major target Sirt2. Suramin
is weaker on Sirt5 (in an assay using a nonfluorogenic peptide
as a substrate, an IC50 of 25 μM was measured9) and Ro31-8220
that is potent on Sirt1−3 did not show an effect on Sirt5.
Among all of the reference inhibitors, only Ro31-8220 shows
significant potency against Sirt3.
We then screened our internal library of thiourea inhibitors,
which mainly contains thiobarbiturates. The results are
presented in Table 2, and structures can be found in Chart 2.
Several thiobarbiturates showed inhibition of Sirt5 in the
micromolar region, usually with similar potency than for Sirt1
and -2. These inhibitors were generally less potent on Sirt3, but
some of them showed an inhibition of Sirt3, which was rarely
seen among the reference inhibitors. Inhibitor 6 shows the best
selectivity for Sirt5 and might be used as a lead structure for
further selectivity optimizations.
1
2
3.4 0.3
5.3 0.7
10.4 3.0
9.7 1.6
20.3 1.5
30% at 50 μM
41% at 50 μM
14% at 50 μM
3.6 0.2
2.3 0.2
3
5.9 0.3
46.5 8.5
27.0 4.0
55.9 6.2
16.6 0.4
12.6 0.4
17.8 1.2
12.9 0.3
67.3 5.9
6.2 0.8
4
42.4 7.2
89.3 11.3
28.4 1.44
10.5 0.4
56.5 3.4
53.2 1.6
36.8 2.3
9.9 0.7
20% at 50 μM 17% at 50 μM
10% at 50 μM NI at 50 μM
20% at 50 μM 116.3 10.1
5
6
7
9.8 0.8
29.3 3.2
8
10.0 1.3
14.4 2.3
NI at 50 μM
3.4 0.8
22% at 50 μM
25% at 50 μM
13% at 50 μM
30.3 2.2
9
10
11
12
13
14
15
6.7 1.1
7.5 0.9
46.4 4.6
12.4 5.1
30.0 4.1
39.4 12.2
22.4 3.8
50.5 9.1
12.4 0.5
6.0 0.6
8.7 0.7
40.3 3.7
14.7 2.1
11.7 2.5
13% at 50 μM
43.3 5.0
ring is mimicking the succinyl group of the substrate and shows
hydrogen bonds with Tyr102, Arg105, and Gln140 (Figure SI1
in the Supporting Information). The interaction is stabilized by
strong electrostatic interactions between acidic thiobarbiturate
and basic guanidinium group of Arg105 (exemplarily shown for
To rationalize our findings for further optimization studies,
docking studies were carried out using the available X-ray
structures of Sirt5.7 The docking showed that the thiobarbiturate
1051
dx.doi.org/10.1021/ml3002709 | ACS Med. Chem. Lett. 2012, 3, 1050−1053

ACS Medicinal Chemistry Letters
Letter
We also tested a library of Sirt inhibitors with a thiobarbiturate
scaffold on Sirt5 and discovered several potent hits with activity
in the low micromolar region, and some of them inhibit Sirt5
better than the other subtypes. With N-alkylation, we have
identified a structural feature that favors Sirt5 inhibition. As
Sirt3 and Sirt5 are present in the same compartment of the cell
(mitochondria), the observed difference in affinity among these
is especially of interest. These results in view of the docking
analyses may guide further development of optimized inhibitors
of Sirt5.
Chart 2. Sirtuin Inhibitors with a Thiobarbiturate Scaffold
EXPERIMENTAL PROCEDURES
■
The Supporting Information contains the procedures for the
expression and purification of recombinant sirtuins and the synthesis
of the precursors of ZK(s)A. Compounds 1, 3−5, 8−11, and 13−15
were synthesized in our lab according to the Supporting Information.
Compounds 2, 6, 7, and 12 were ordered from Chembridge (San
Diego, CA) (Chembridge and lab-internal numbers can be found in
Table SI1 in the Supporting Information). High analytical grade
chemicals and solvents were purchased from Acros, Aldrich, Sigma, or
Fluka. When necessary, solvents were dried by standard techniques
and distilled. After extraction from aqueous layers, the organic solvents
were dried over anhydrous sodium sulfate.
Thin-layer chromatography (TLC) was performed on aluminum
sheets precoated with silica gel 60 F254 (0.2 mm) (E. Merck).
Chromatographic spots were visualized by UV light. Purification of
crude compounds was carried out by flash column chromatography on
Merck Silica Gel 60 (Kieselgel, 0.040−0.063 mm, E. Merck). 1H NMR
spectra were recorded in DMSO-d6 on a Bruker Avance DRX 400
MHz spectrometer and 13C NMR on a Varian 100 MHz. Chemical
shifts (δ scale) are reported in parts per million (ppm) relative to the
central peak of the solvent. Coupling constant (J values) are expressed
in Hertz (Hz). Spin multiplicities are given as s (singlet), b s (broad
singlet), d (doublet), t (triplet), dd (double doublet), dt (double
triplet), or m (multiplet). EI- and CI-mass spectra were measured with
a TSQ700 mass spectrometer (Thermoelectron). ESI- and PCI-mass
spectra were recorded with a LCQ-Advantage mass spectrometer. In
all cases, spectroscopic data are in agreement with known compounds
and assigned structures.
Compound ZKA (see SI4 in the Supporting Information) (300 mg,
0.69 mmol) was solved at 0 °C in 10 mL of dry THF, together with
diisopropylethylamine (DIPEA) (249 μL, 1.37 mmol). Succinic
anhydride (82 mg, 0.82 mmol) was added portionwise, and the
reaction was stirred for 1 h at the same temperature, then left overnight
at room temperature. When TLC indicated complete consumption of
the starting material, the reaction mixture was evaporated, and the
residue was solved in dichloromethane (DCM) and extracted by a 5%
aqueous solution of NaHCO3. The aqueous phase was collected and
acidified (pH ∼ 2), then poured into a separating funnel containing
DCM, and extracted. The organic layer was dried over MgSO4, filtered,
and evaporated to give a crude reaction mixture, which was purified by
flash column chromatography on silica gel with 5% methanol in DCM
as the eluent to furnish the titled compound ZK(s)A (Scheme 1) as a
white solid (280 mg, 76%): mp 148 °C. 1H NMR (DMSO-d6) δ: 12.02
(bs, 1H), 10.52 (s, 1H), 7.83 (t, J = 5.4 Hz, 1H), 7.78 (d, J = 1.6 Hz,
1H), 7.73 (d, J = 8.7 Hz, 1H), 7.65 (d, J = 7.5 Hz, 1H), 7.51 (dd, J =
1.6, 8.7 Hz, 1H), 7.40−7.15 (m, 5H), 5.04 (s, 2H), 4.19−4.08 (m, 1H),
3.05−2.98 (m, 2H), 2.44−2.33 (m, 5H), 2.27 (t, J = 7.5 Hz, 2H),
1.75−1.56 (m, 2H), 1.47−1.26 (m, 4H). 13C NMR (DMSO-d6) δ:
174.3, 172.4, 171.2, 160.5, 156.6, 154.1, 153.5, 142.7, 137.4, 128.8,
128.3, 128.2, 126.3, 115.7, 115.5, 112.7, 106.1, 65.9, 56.0, 38.7, 31.7,
30.5, 29.7, 29.2, 23.4, 18.4. LRMS (ESI) m/z 538 [M + H]+.
Figure 1. Docking result for 2 at Sirt5. Hydrogen bonds are shown as
dashed orange lines.
the most active thiobarbiturate 2 in Figure 1). This binding
mode was observed for all docked thiobarbiturates (see Figure
SI2 in the Supporting Information). Thiobarbiturates containing
a flexible side chain and an aromatic ring interact with aromatic
residues located in the lysine substrate channel of Sirt5 (His158
and Phe223), thus resulting in high inhibitory activity. The
docking results rationalized the finding that substituting the
thiobarbituric −NH− function by methyl, ethyl, or allyl groups is
tolerated. These can be accommodated by the binding pocket
but result in slightly different orientation of the inhibitor (see
Figure SI2 in the Supporting Information). Still, similar inhibi-
tion of Sirt2 is observed while this N-alkylation leads to an
increased potency as compared to Sirt3 and partially also Sirt1
(compare 7 vs 8 and 9).
With ZK(s)A, we have synthesized a small substrate that
works well for Sirt5 and allows for high-throughput screening
to discover Sirt5 inhibitors. Among sirtuin reference inhibitors,
Suramin, Sirtinol, and Cambinol did not show selectivity
toward the different sirtuins, while EX-527 is Sirt1 selective
also in comparison to Sirt5, and HR 103 works best on Sirt2.
Inhibition of recombinant Sirt1/2/3 was determined using a homo-
geneous fluorescence deacetylase assay.16 Sirt1/2/3 was mixed with the
fluorescent substrate ZMAL (assay concentration, 10.5 μM), NAD+
(500 μM), the inhibitor, DMSO [5% (v/v)], and assay buffer (25 mM
Tris·HCl, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, pH 8.0) to a
volume of 60 μL, incubated for 4 h (37 °C, 180 rpm), then treated with
1052
dx.doi.org/10.1021/ml3002709 | ACS Med. Chem. Lett. 2012, 3, 1050−1053

ACS Medicinal Chemistry Letters
Letter
60 μL of trypsin solution [50 mM Tris·HCl, 100 mM NaCl, 8 mM
nicotinamide, 5.5 U/μL trypsin, 6.7% DMSO (v/v)] and incubated
(37 °C, 180 rpm, 20 min). The fluorescence intensity was then
measured with a microplate reader (BMG Polarstar, λEx = 390 nm,
λEm = 460 nm). To ensure initial state conditions, the conversion of
ZMAL was adjusted to 10−30% substrate conversion without inhibitor.
A mixture with only DMSO was used as the control. Inhibition rates
were calculated in reference to the DMSO control. All inhibition
experiments were run at least three times independently. IC50 values
were determined with Graphpad Prism 4.0 software (La Jolla, CA).
Sirt5 was mixed with the fluorescent substrate ZK(s)A (assay
concentration, 200 μM), NAD+ (500 μM), the inhibitor, DMSO [<2%
(v/v)], and assay buffer (25 mM Tris·HCl, 130 mM NaCl, 3 mM KCl,
1 mM MgCl2, pH 8.0, and 0.1% PEG8000) to a volume of 51 μL,
incubated for 1 h (37 °C, 750 rpm), then treated with 10 μL of trypsin
solution (50 mM Tris·HCl, 130 mM NaCl, and 6 mg/mL trypsin)
and incubated (37 °C, 1250 rpm, 2 min). The fluorescence intensity
was then measured with a microplate reader (BMG Polarstar, λEx =
390 nm, λEm = 460 nm). To ensure initial state conditions, the
conversion of ZK(s)A was adjusted to 4−10% substrate conversion
without inhibitor. A mixture with only DMSO was used as the control.
Inhibition rates were calculated in reference to the DMSO control.
All inhibition experiments were run at least two times independently.
IC50 values were determined with Graphpad Prism 4.0 software.
(4) Van Meter, M.; et al. Repairing split ends: SIRT6, mono-ADP
ribosylation and DNA repair. Aging (Albany NY) 2011, 3 (9), 829−
835.
(5) Heltweg, B.; et al. Subtype selective substrates for histone
deacetylases. J. Med. Chem. 2004, 47 (21), 5235−5243.
(6) Zhu, A. Y.; et al. Plasmodium falciparum Sir2A Preferentially
Hydrolyzes Medium and Long Chain Fatty Acyl Lysine. ACS Chem.
Biol. 2012, 7 (1), 155−159.
(7) Du, J.; et al. Sirt5 is a NAD-dependent protein lysine
demalonylase and desuccinylase. Science 2011, 334 (6057), 806−809.
(8) Hirschey, M. D. Old enzymes, new tricks: sirtuins are NAD(+)-
dependent de-acylases. Cell Meta. 2011, 14 (6), 718−719.
(9) He, B.; Du, J.; Lin, H. Thiosuccinyl peptides as sirt5-specific
inhibitors. J. Am. Chem. Soc. 2012, 134 (4), 1922−1925.
(10) Fatkins, D. G.; Monnot, A. D.; Zheng, W. Nepsilon-thioacetyl-
lysine: a multi-facet functional probe for enzymatic protein lysine
Nepsilon-deacetylation. Bioorg. Med. Chem. Lett. 2006, 16 (14), 3651−
3656.
(11) Smith, B. C.; Denu, J. M. Mechanism-based inhibition of Sir2
deacetylases by thioacetyl-lysine peptide. Biochemistry 2007, 46 (50),
14478−14486.
(12) Hawse, W. F.; et al. Structural insights into intermediate steps in
the Sir2 deacetylation reaction. Structure 2008, 16 (9), 1368−1377.
(13) Schuetz, A.; et al. Structural basis of inhibition of the human
NAD+-dependent deacetylase SIRT5 by suramin. Structure 2007, 15
(3), 377−389.
(14) Peng, C.; et al. The first identification of lysine malonylation
substrates and its regulatory enzyme. Mol Cell Proteomics 2011, 10
(12), M111012658.
(15) Trapp, J.; et al. Adenosine mimetics as inhibitors of NAD
+-dependent histone deacetylases, from kinase to sirtuin inhibition. J.
Med. Chem. 2006, 49 (25), 7307−7316.
(16) Heltweg, B.; Trapp, J.; Jung, M. In vitro assays for the
determination of histone deacetylase activity. Methods 2005, 36 (4),
332−337.
(17) Madsen, A. S.; Olsen, C. A. Substrates for Efficient Fluorometric
Screening Employing the NAD-Dependent Sirtuin 5 Lysine Deacylase
(KDAC) Enzyme. J. Med. Chem. 2012, 55 (11), 5582−5590.
(18) Grozinger, C. M.; et al. Identification of a class of small molecule
inhibitors of the sirtuin family of NAD-dependent deacetylases by
phenotypic screening. J. Biol. Chem. 2001, 276 (42), 38837−38843.
(19) Heltweg, B.; et al. Antitumor activity of a small-molecule
inhibitor of human silent information regulator 2 enzymes. Cancer Res.
2006, 66 (8), 4368−4377.
(20) Howitz, K. T.; et al. Small molecule activators of sirtuins extend
Saccharomyces cerevisiae lifespan. Nature 2003, 425 (6954), 191−196.
(21) Napper, A. D.; et al. Discovery of indoles as potent and selective
inhibitors of the deacetylase SIRT1. J. Med. Chem. 2005, 48 (25),
8045−8054.
ASSOCIATED CONTENT
* Supporting Information
■
S
Description of computational methods and docking results,
recombinant sirtuin expression and purification, additional
synthetic procedures and spectral data, and HPLC purity of
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: manfred.jung@pharmazie.uni-freiburg.de.
■
Present Address
∥Freiburg Institute of Advanced Studies (FRIAS), Universitat
̈
Freiburg, Germany.
Funding
We thank the Deutsche Forschungsgemeinschaft (Ju295/8-1,
Si868/6-1) and the EU (Nr. 241865-FP7 Health) for funding
and A. Walter for supporting inhibitor synthesis.
Notes
The authors declare no competing financial interest.
(22) Neugebauer, R. C.; et al. Structure-activity studies on
splitomicin derivatives as sirtuin inhibitors and computational
prediction of binding mode. J. Med. Chem. 2008, 51 (5), 1203−1213.
(23) Lawson, M.; et al. Inhibitors to understand molecular
mechanisms of NAD(+)-dependent deacetylases (sirtuins). Biochim.
Biophys. Acta 2010, 1799 (10−12), 726−739.
ABBREVIATIONS
■
ADP, adenosine diphosphate; AMC, aminomethyl coumarin;
NAD+, nicotinamide adenine dinucleotide; ZKA, benzyl N-{5-
amino-1-[(4-methyl-2-oxo-2H-chromen-7-yl)carbamoyl]-
pentyl}carbamate; ZK(s)A, 3-[(5-{[(benzyloxy)carbonyl]-
amino}-5-[(4-methyl-2-oxo-2H-chromen-7-yl)carbamoyl]-
pentyl)carbamoyl]propanoic acid; TLC, thin-layer chromatog-
raphy; DCM, dichloromethane; DIPEA, diisopropylethylamine
REFERENCES
■
(1) Tanny, J. C.; et al. An enzymatic activity in the yeast Sir2 protein
that is essential for gene silencing. Cell 1999, 99 (7), 735−745.
(2) Frye, R. A. Characterization of five human cDNAs with
homology to the yeast SIR2 gene: Sir2-like proteins (sirtuins)
metabolize NAD and may have protein ADP-ribosyltransferase
activity. Biochem. Biophys. Res. Commun. 1999, 260 (1), 273−279.
(3) Haigis, M. C.; et al. SIRT4 inhibits glutamate dehydrogenase and
opposes the effects of calorie restriction in pancreatic beta cells. Cell
2006, 126 (5), 941−954.
1053
dx.doi.org/10.1021/ml3002709 | ACS Med. Chem. Lett. 2012, 3, 1050−1053