A bicyclic pentapeptide-based highly potent and selective pan-SIRT1/2/3 inhibitor harboring Nε-thioacetyl-lysine

Article abstract of DOI:10.1016/j.bmc.2020.115356

Past few years have seen an active pursuit of the inhibitors for the deacylation catalyzed by the seven human sirtuins (i.e. SIRT1-7) as valuable chemical biological/pharmacological probes of this enzymatic deacylation and lead compounds for developing novel therapeutics for human diseases. In the current study, we prepared eight monocyclic and one bicyclic analogs of a linear pentapeptide-based potent (sub-μM IC₅₀’s) pan-SIRT1/2/3 inhibitor Zheng laboratory discovered recently that harbors the catalytic mechanism-based SIRT1/2/3 inhibitory warhead N^ε-thioacetyl-lysine at its central position. We found that the bicyclic analog exhibited largely comparable SIRT1/2/3 inhibitory potencies to those of the parent linear pentapeptide, however, the former is proteolytically much more stable than the latter. Moreover, the bicyclic analog displayed very weak inhibition against SIRT5/6/7, was cell permeable, and exhibited an anti-proliferative effect on the human SK-MEL-2 melanoma cells. This bicyclic analog could be a lead for the future development of more potent and still selective pan-SIRT1/2/3 inhibitors whose use in studies on human sirtuin biology, pharmacology, and medicinal chemistry could complement with the use of the potent inhibitors selective for a single human sirtuin.

Full text of DOI:10.1016/j.bmc.2020.115356

Journal Pre-proofs
A bicyclic pentapeptide-based highly potent and selective pan-SIRT1/2/3 in-
ε
hibitor harboring N -thioacetyl-lysine
Renwu Li, Lingling Yan, Xun Sun, Weiping Zheng
PII:
S0968-0896(20)30147-4
https://doi.org/10.1016/j.bmc.2020.115356
BMC 115356
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
13 November 2019
29 January 2020
31 January 2020
Please cite this article as: R. Li, L. Yan, X. Sun, W. Zheng, A bicyclic pentapeptide-based highly potent and
ε
selective pan-SIRT1/2/3 inhibitor harboring N -thioacetyl-lysine, Bioorganic & Medicinal Chemistry (2020),
doi: https://doi.org/10.1016/j.bmc.2020.115356
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Ltd.

A bicyclic pentapeptide-based highly potent and selective pan-SIRT1/2/3

inhibitor harboring N -thioacetyl-lysine
Renwu Li , Lingling Yan , Xun Sun *, Weiping Zheng b,¶,*
a,§
b
a,c,
aDepartment of Natural Products Chemistry, School of Pharmacy, Fudan University, 826
Zhangheng Road, Shanghai 201203, P. R. China
bSchool of Pharmacy, Jiangsu University, 301 Xuefu Road, Zhenjiang 212013, Jiangsu
Province, P. R. China
cShanghai Key Laboratory of Clinical Geriatric Medicine, Shanghai 201203, P. R. China
*
Correspondence: wzheng@ujs.edu.cn (W. Zheng), sunxunf@shmu.edu.cn (X. Sun)
A visiting doctoral student at Zheng laboratory at Jiangsu University.
§
¶
lead contact.
1

Abstract
Past few years have seen an active pursuit of the inhibitors for the deacylation catalyzed
by the seven human sirtuins (i.e. SIRT1-7) as valuable chemical biological/pharmacological
probes of this enzymatic deacylation and lead compounds for developing novel therapeutics
for human diseases. In the current study, we prepared eight monocyclic and one bicyclic
analogs of a linear pentapeptide-based potent (sub-M IC ’s) pan-SIRT1/2/3 inhibitor Zheng
5
0
laboratory discovered recently that harbors the catalytic mechanism-based SIRT1/2/3

inhibitory warhead N -thioacetyl-lysine at its central position. We found that the bicyclic
analog exhibited largely comparable SIRT1/2/3 inhibitory potencies to those of the parent
linear pentapeptide, however, the former is proteolytically much more stable than the latter.
Moreover, the bicyclic analog displayed very weak inhibition against SIRT5/6/7, was cell
permeable, and exhibited an anti-proliferative effect on the human SK-MEL-2 melanoma
cells. This bicyclic analog could be a lead for the future development of more potent and still
selective pan-SIRT1/2/3 inhibitors whose use in studies on human sirtuin biology,
pharmacology, and medicinal chemistry could complement with the use of the potent
inhibitors selective for a single human sirtuin.

Keywords: N -thioacetyl-lysine; cyclic peptide; sirtuin; inhibitor; structure-activity
relationship.
2

1
. Introduction
Sirtuins refer to a family of intracellular proteins able to catalyze the -nicotinamide
+

adenine dinucleotide (-NAD )-dependent side chain deacylation of the N -acyl-lysine
1
-9

residues on their substrate proteins (Fig. 1) . As also indicated in Fig. 1, the N -acyl groups
can be acetyl and those bulkier ones1,2,4,5,7,8. The lysine side chain N -acylation on proteins
can be accomplished by enzymatic or non-enzymatic reactions1,6,10-16. The sirtuin proteins are
present in organisms from all the three evolutionary kingdoms of life, i.e. bacteria, archaea,
and eukarya; and seven sirtuins (i.e. SIRT1-7) have been found thus far in mammals including

1
7
humans . The sirtuin-catalyzed deacylation has been increasingly demonstrated to play an
important role in regulating a variety of crucial cellular processes such as transcription,
metabolism, and DNA damage repair2,5,18-28. Also, this enzymatic deacylation has been
regarded as a contemporary therapeutic target for developing novel therapeutics for human
diseases such as cancer, neurodegeneration, and metabolic diseases29-32. As such, past
O
R
HN
Acetyl: human SIRT1/2/3/7
Malonyl: human SIRT5
H N
2
O
Succinyl: human SIRT5/7
Glutaryl: human SIRT4/5
Myristoyl: human SIRT1/2/3/6/7
R
H
H
O
O
HN
HN
3-Methylglutaryl: human SIRT4
Sirtuins
Protein
Protein
O
NH2
O
H2O
NH2
O
1' OH
N
O
'
1' N
ADP
ADP
2'
O
R
3
HO
'
2'
Nicotinamide
O
3
HO
OH
-NAD+
2'-O-AADPR

+

Fig. 1. The sirtuin-catalyzed -NAD -dependent deacylation of N -acyl-lysine protein
substrates. Example R-C(=O)- removable by different human sirtuins are also indicated.
Note: The SIRT7-catalyzed deacetylation and demyristoylation can be promoted by
dsDNA, rRNA, or tRNA.
3

few years have seen an active pursuit of the chemical modulators (inhibitors and activators)
for the sirtuin-catalyzed deacylation as potential therapeutics and also as chemical probes for
sirtuin biochemistry, biology, and pharmacology3,4,8,29,31-34
.
As for the human sirtuin inhibitor development, while potent and selective inhibitors are
the major type of molecular entities that have been pursued and have been thus far indentified
for SIRT1, SIRT2, SIRT5, and SIRT64,29,31-34, those potent pan-sirtuin inhibitors such as the
3
5
potent and selective dual SIRT1/2 inhibitors Zheng laboratory recently identified may also
have their unique appeal in that their use in the studies on human sirtuin biology,
pharmacology, and medicinal chemistry could complement with the use of the potent
inhibitors selective for a single human sirtuin. In the current study, we discovered a

N -thioacetyl-lysine-containing bicyclic pentapeptide (i.e. compound 10 in Fig. 2) as a potent
(
sub-M IC ’s) pan-inhibitor selective for human SIRT1/2/3 (versus human SIRT5/6/7).
5
0
2
. Materials and Methods
2
.1. General
The following were purchased for the compound preparation and were used as received
from commercial vendors without further treatment. Sigma-Aldrich China:
N-methylmorpholine (NMM), trifluoroacetic acid (TFA), N,N-dimethylformamide (DMF),
anhydrous tetrahydrofuran (THF); TCI Shanghai: 4-bromobutyric acid, 6-bromohexanoic acid,
1
3
,8-diazabicyclo[5.4.0]undec-7-ene (DBU), 1,1,1,3,3,3-hexafluoroisopropanol (HFIP),
-bromopropylamine hydrobromide; Enamine Ltd.: 7-bromoheptylamine hydrobromide; Alfa
Aesar China: 11-bromoundecanoic acid, sodium iodide (NaI), sodium methoxide (MeONa),
4

phenol, thioanisole, ethanedithiol, N-hydroxybenzotriazole (HOBt),
2
2
-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethylaminium hexafluorophosphate (HBTU),
-(7-aza-1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HATU),
N,N-diisopropylethylamine (DIPEA); J&K Scientific Ltd.: 8-bromooctanoic acid;
Honeywell China: acetonitrile, dichloromethane (DCM), methanol (MeOH), hexanes;
Sinopharm Chemical Reagent Co., Ltd.: piperidine, diethyl ether, acetic anhydride; Shanghai
Plus Bio-Sci & Tech Co., Ltd.: Rink amide 4-methylbenzhydrylamine (MBHA) resin; GL
Biochem (Shanghai) Ltd.: 2-chlorotrityl chloride resin.


N -9-Fluorenylmethoxycarbonyl (Fmoc)-N -thioacetyl-lysine was synthesized from
N -Fmoc-lysine hydrochloride (purchased from TCI Shanghai) and ethyl dithioacetate
purchased from Sigma-Aldrich China) according to the procedure Zheng laboratory

(
3
6

previously reported and was used in the current study for incorporating N -thioacetyl-lysine

into compounds 2-10. All the other N -Fmoc-protected amino acids were purchased from the
following commercial vendors: Sigma-Aldrich China, Alfa Aesar China, TCI Shanghai, GL
Biochem (Shanghai) Ltd., or Chem-Impex International Inc.
Routine unit-resolution mass spectrometry (MS) was performed with a Thermo LXQ
LC-ion trap mass spectrometer at Jiangsu University. High-resolution mass spectrometry
(HRMS) was performed with an AB 5600+ Q TOF high-resolution mass spectrometer at the
Pharmacy School of Fudan University.
The following were purchased for the sirtuin inhibition assay and the pronase digestion
assay, and were used as received from commercial vendors without further treatment.
+
Sigma-Aldrich China: the active human recombinant His -SIRT1, Trizma, Hepes, -NAD , a
6
5

1
.0 M solution of MgCl (molecular-biology grade), the pronase from Streptomyces griseus;
2
Cayman Chemical: the active human recombinant GST-SIRT1, the active human recombinant
His -SIRT2, the active human recombinant His -SIRT3, the active human recombinant
6
6
GST-SIRT5, the active human recombinant His -SIRT6, the active human recombinant
6
His -SIRT7; Thermo Fisher Scientific China: yeast tRNA (catalog #: AM7119),
6
SUPERase-In RNase inhibitor (catalog #: AM2694); TCI Shanghai: DL-dithiothreitol (DTT);
Alfa Aesar China: NaCl, KCl.
The following are the substrate peptides prepared and used in the sirtuin inhibition assay:

the SIRT1/2/3 substrate H N-HK-[N -acetyl-lysine]-LM-COOH that corresponds to amino
2
acids 380-384 of the K382-acetylated human tumor suppressor protein p53; the SIRT5

substrate CH CONH-AR-[N -succinyl-lysine]-ST-CONH that corresponds to amino acids
3
2
7
-11 of the K9-succinylated human histone protein H3; the SIRT6 substrate

H N-EALPK-[N -myristoyl-lysine]-TGGPQ-CONH that corresponds to amino acids 15-25
2
2
of the K20-myristoylated human tumor necrosis factor  (TNF); the SIRT7 substrate

H N-GGKAPR-[N -acetyl-lysine]-QLATKA-CONH that corresponds to amino acids 12-24
2
2
of the K18-acetylated human histone protein H3.
2
2
.2. Compound preparation
.2.1. Preparation of monocyclic pentapeptides 2-5 (Scheme 1)
These compounds were synthesized by the Fmoc chemistry-based manual solid phase
peptide synthesis (SPPS) on the Rink amide MBHA resin. For each amino acid coupling

reaction, 4 (or 2) equiv. of a N -Fmoc-protected amino acid, 3.8 (or 2) equiv. of the coupling
6

S
S
Trt
Trt
H3C
HN
NH
O
H3C
HN
NH
O
N
COOH
N
CH3
CH3
S
S
N
(CH2)
N
O
n
O
O
H
SPPS
O
O
Fmoc
N
H2N
H
N
H
Br
N
H
H
HFIP/DCM (1/1)
N
H
N
N
N
N
H
N
N
H
CH3
Rink Amide
MBHA Resin
H
HATU,
0.4N NMM/DMF
(CH2)n
H
O
O
Br
O
O
2
-Cl-Trt
CH3
O
O
H3C
H3C
2-Cl-Trt
S
S
S
Trt
Trt
H3C
HN
NH
H3C
HN
NH
H3C
HN
NH
O
N
HN
N
CH3
CH3
CH3
S
S
S
N
N
N
O
O
O
O
O
O
O
O
O
O
O
N
H
H
NaI, 1% DBU/DMF
N
H
H
Reagent K
N
H
H
N
H
N
N
H
N
N
NH2
N
N
H
N
N
N
H
N
(CH2)n
H
(CH2)n
O
H
(CH2)n
H
O
H
Br
O
O
O
O
O
CH3
CH3
CH3
H3C
HO
H3C
O
H3C
2: n=1; 3: n=3; 4: n=5; 5: n=8.
Scheme 1. The synthetic scheme of the monocyclic analogs 2-5. Reagent K, a
TFA-containing peptide cleavage cocktail.
reagent HBTU and the additive HOBt were used in the presence of 0.4 N NMM/DMF, and
the coupling reaction was allowed to proceed at room temperature for 1 h (or 2 h). A 2% (v/v)
3
7
DBU/DMF solution was used for Fmoc removal (2 x 5 min at room temperature) . Following
the on-resin assembly and the N-terminal Fmoc removal of the linear pentapeptide composed

of His(Trt), Tyr(2-Cl-Trt), N -thioacetyl-lysine, Leu, and Met, the N-term -amino group was
acylated with a bromoalkanoic acid (4-bromobutyric acid (n=1), 6-bromohexanoic acid (n=3),
8
-bromooctanoic acid (n=5), or 11-bromoundecanoic acid (n=8)) (4 equiv.) in the presence of
HATU (3.8 equiv.) and 0.4 N NMM/DMF (2 x 1 h at room temperature). The side chain
-Cl-Trt protecting group of Tyr(2-Cl-Trt) was then selectively removed with 50% (v/v)
2
HFIP/DCM (10 min at room temperature), and the exposed free side chain phenolic group
was subsequently allowed to react intramolecularly with the N-term bromide in the presence
of NaI (2 equiv.) and 1% (v/v) DBU/DMF (24 h at room temperature). The resulting
peptidyl-resin was then treated with reagent K (83.6% (v/v) TFA, 5.9% (v/v) phenol, 4.2%
(
v/v) ddH O, 4.2% (v/v) thioanisole, 2.1% (v/v) ethanedithiol) at room temperature for 4 h,
2
followed by the filtration of the reaction mixture and the removal of the volatiles in the filtrate
7

with a stream of nitrogen gas in a well-ventilated fuming hood. To the resulting residue was
added cold diethyl ether to precipitate out the crude pentapeptides 2-5 which were each
purified with semi-preparative reversed-phase high performance liquid chromatography
(
RP-HPLC) on a C18 column (1 x 25 cm, 5 m) eluted with a gradient of ddH O containing
2
0
.05% (v/v) of TFA and acetonitrile containing 0.05% (v/v) of TFA at 4.5 mL/min and
monitored at 214 nm. The desired RP-HPLC fractions were pooled and concentrated under
reduced pressure on a rotary evaporator to remove volatiles and the remaining aqueous
solution was lyophilized to afford the purified pentapeptides 2-5 all as puffy white solids. Of
note, for a synthetic scale of 0.1 mmol, the purified 2-5 were obtained in overall yields of
3
.1%, 12.8%, 9.2%, and 6.6%, respectively. The purified 2-5 were all >95% pure per
RP-HPLC analysis on an analytical C18 column (0.46 x 25 cm, 5 m), and their exact masses
were all confirmed with HRMS analysis (Table 1).
Table 1. The HRMS analysis of compounds 2-10a
Compound
Ionic Formula
Calculated m/z
816.3895
844.4208
872.4521
914.4991
867.3980
895.4293
923.4606
965.5076
956.4133
Observed m/z
816.3875
844.4200
872.4513
914.4981
867.3989
895.4300
923.4599
965.5070
956.4164
+
+
+
+
2
3
4
5
6
7
8
9
[C H N O S ]
3
8
58
9
7 2
[C H N O S ]
4
0
62
9
7 2
[C H N O S ]
4
2
66
9
7 2
[C H N O S ]
4
5
72
9
7 2
+
+
+
+
[C H N O S Na]
3
9
60 10
7 2
[C H N O S Na]
4
1
64 10
7 2
[C H N O S Na]
4
3
68 10
7 2
[C H N O S Na]
4
6
74 10
7 2
+
1
0
[C H N O S Na]
46 63 9 8 2
a
All compounds were measured with the positive ion mode of electrospray ionization (ESI).
8

2
.2.2. Preparation of monocyclic pentapeptides 6-9 (Scheme 2)
These compounds each resulted from the condensation between a cyclic dipeptidic
amine and a side chain fully protected linear tripeptidic acid. (a) Preparation of a cyclic
dipeptidic amine: This started with the loading of Fmoc-Met onto the 2-chlorotrityl chloride
o
resin in the presence of DIPEA (2 equiv.) and DCM initially at 0-2 C for 15 min and then at
room temperature for 1 h, which was followed by treating the reaction mixture with
DIPEA/MeOH (v/v, 1/9) at room temperature for 15 min. The reaction mixture was then
filtered and washed with DMF, isopropanol, and DCM, affording the 2-chlorotrityl resin
loaded with Fmoc-Met. Following the Fmoc removal on Fmoc-Met with 20% (v/v)
piperidine/DMF (2 x 10 min at room temperature), the exposed free amino group was reacted
at room temperature for 2 h with Boc-Tyr(OH)-OH (4 equiv.) in the presence of HBTU (3.8
equiv.), HOBt (3.8 equiv.), and 0.4N NMM/DMF. The subsequent treatment with 50% (v/v)
HFIP/DCM at room temperature for 20 min liberated from the resin the dipeptidic acid which
was then reacted with a bromoalkylamine hydrobromide (n=1,5: commercially available;
n=3,8: synthesized as below described according to Scheme 3) (1.2 equiv.) at room
temperature for 24 h in the presence of HATU (1.2 equiv.), DIPEA (3.0 equiv.), and THF,
affording the bromide intermediate with a free phenolic group which was extracted into ethyl
acetate from the reaction mixture that had been concentrated under reduced pressure on a
rotary evaporator, dried over anhydrous Na SO , and concentrated under reduced pressure on
2
4
a rotary evaporator. To the obtained crude intermediate was added MeONa (1.5 equiv.), NaI
0.1 equiv.), and MeOH (freshly degassed with a stream of nitrogen gas) and the reaction
(
mixture was stirred at room temperature for 48-72 h, affording the intermediate presumably
9

H3C
H3C
O
S
S
Cl
HO
Cl
O
Cl
O
Fmoc-Met-OH
1. 20% piperidine/DMF
H
Fmoc
O
N
O
N
H
Boc
N
H
2. Boc-Tyr(OH)-OH
HBTU, HOBt
2
-Chlorotrityl
0.4N NMM/DMF
chloride resin
HO
H3C
S
H3C
H3C
H3C
O
NH2 HBr
CH2)n
S
S
S
(
O
H
O
MeONa/MeOH
NaI(cat.)
O
Br
H
H
50%TFA/DCM
HFIP/DCM (1/1)
Boc
N
O
Boc
N
O
2
H N
O
N
N
H
OH
N
H
Boc
HATU, DIPEA
N
N
H
H
HN
O
HN
HN
(CH2)n
Br
HO
(CH2)n
(CH2)n
HO
(n=1,3,5,8)
HATU
DIPEA
O
O
S
S
S
Trt
N
S
Trt
HN
CH3
OH
HN
CH3
H3C
NH
H3C
NH
N
N
N
HFIP/DCM
(1/1)
O
OH
O
Cl
H
Fmoc
H
Cl
N
H
Cl
O
SPPS
N
O
N
O
HN
H3C
O
N
HO
Fmoc
O
HN
H3C
N
H
N
H
O
O
H
O
O
O
2-Chlorotrityl
chloride resin
NH
Boc
NH
Boc
S
S
Trt
N
HN
CH3
H
N
HN
CH3
CH3
S
CH3
HN
S
N
O
N
O
O
H
O
O
O
H
N
TFA/DCM/ddH O
H
H
N
2
N
O
N
O
H3C
N
H
N
N
H3C
N
N
N
H
H
H
H
H
O
O
HN
O
O
(CH2)n
(CH2)n
NH
H2N
Boc
O
O
6: n=1; 7: n=3; 8: n=5; 9: n=8
Scheme 2. The synthetic scheme of the monocyclic analogs 6-9. Note: While the
bromoalkylamine hydrobromide (highlighted in red) with n=1 (i.e. 3-bromopropylamine
hydrobromide) and n=5 (i.e. 7-bromoheptylamine hydrobromide) were purchased from
commercial vendors, the bromoalkylamine hydrobromide with n=3 (i.e. 5-bromopentylamine
hydrobromide) and n=8 (i.e. 10-bromodecylamine hydrobromide) were synthesized according
to Scheme 3.
with the depicted cyclic structure which was extracted into ethyl acetate from the reaction
mixture that had been concentrated under reduced pressure on a rotary evaporator, and dried
over anhydrous Na SO . To the obtained crude intermediate was added 50% (v/v) TFA/DCM
2
4
and the reaction mixture was stirred at room temperature for 8 h, affording the crude cyclic
dipeptidic amine which was used directly for the later condensation with the side chain fully
protected linear tripeptidic acid. (b) Preparation of the side chain fully protected linear

tripeptidic acid: This started with the loading of Fmoc-N -thioacetyl-lysine onto the
o
2
-chlorotrityl chloride resin in the presence of DIPEA (2 equiv.) and DCM initially at 0-2 C
1
0

for 15 min and then at room temperature for 1 h, which was followed by treating the reaction
mixture with DIPEA/MeOH (v/v, 1/9) at room temperature for 15 min. The reaction mixture
was then filtered and washed with DMF, isopropanol, and DCM, affording the 2-chlorotrityl

resin loaded with Fmoc-N -thioacetyl-lysine. The sequential incorporation of Lys(Boc) and
His(Trt) onto the charged 2-chlorotrityl resin was then performed based on the Fmoc
chemistry-based SPPS with each amino acid coupling reaction being performed with an

N -Fmoc-protected amino acid (i.e. Fmoc-Lys(Boc)-OH or Fmoc-His(Trt)-OH) (4 equiv.),
HBTU (3.8 equiv.), HOBt (3.8 equiv.) in 0.4 N NMM/DMF at room temperature for 1 h.
Following the removal of the N-term Fmoc with 20% (v/v) piperidine/DMF (2 x 10 min at
room temperature) and the subsequent on-resin N-term acetylation with acetic anhydride in
0
5
.4 N NMM/DMF (1 h at room temperature), the resulting tripeptidyl resin was treated with
0% (v/v) HFIP/DCM (10 min at room temperature), affording the crude side chain fully
protected linear tripeptidic acid. (c) Preparation of monocyclic pentapeptides 6-9: The
obtained crude cyclic dipeptidic amine and the crude side chain fully protected linear
tripeptidic acid were subsequently condensed in THF with HATU (1.2 equiv.) and DIPEA
(1.5 equiv.) at room temperature for 24 h. The reaction mixture was then concentrated under
reduced pressure on a rotary evaporator and partitioned in ddH O and ethyl acetate. The
2
organic layer was isolated, washed once with 5% (w/v) aqueous tartaric acid solution and
once with saturated aqueous Na CO solution, dried over anhydrous Na SO , and
2
3
2
4
concentrated under reduced pressure on a rotary evaporator. The obtained crude intermediate
was then treated with TFA/DCM/ddH O (90%/5%/5%, v/v/v) at room temperature for 4 h,
2
followed by the filtration of the reaction mixture and the removal of the volatiles in the filtrate
1
1

with a stream of nitrogen gas in a well-ventilated fuming hood. To the resulting residue was
added cold diethyl ether to precipitate out the crude pentapeptides 6-9 which were each
purified with semi-preparative RP-HPLC on a C18 column (1 x 25 cm, 5 m) eluted with a
gradient of ddH O containing 0.05% (v/v) of TFA and acetonitrile containing 0.05% (v/v) of
2
TFA at 4.5 mL/min and monitored at 214 nm. The pooled desired RP-HPLC fractions were
concentrated under reduced pressure on a rotary evaporator and the remaining aqueous
solution was lyophilized to afford the purified pentapeptides 6-9 all as puffy white solids. Of
note, for a synthetic scale of 0.1 mmol, the purified 6-9 were obtained in overall yields of
2
.4%, 6.9%, 6.7%, and 8.5%, respectively. The purified 6-9 were all >95% pure per
RP-HPLC analysis on an analytical C18 column (0.46 x 25 cm, 5 m), and their exact masses
were all confirmed with HRMS analysis (Table 1).
2
1
.2.3. Preparation of 5-bromopentylamine hydrobromide (n=3) and
0-bromodecylamine hydrobromide (n=8) (Scheme 3)
(a) A solution of 5-hydroxylpentylamine (or 10-hydroxyldecylamine) (1.0 mmol) and
(
Boc) O (1.2 mmol) in dry DCM (10 mL) was stirred at room temperature for 2 h before the
2
reaction mixture was partitioned in ddH O (20 mL) and DCM (20 mL), the DCM layer was
2
isolated and the ddH O layer was extracted with DCM for three times. The combined
2
organics were then sequentially washed with saturated aqueous NaHCO solution and brine,
3
dried over anhydrous Na SO , filtered, and concentrated under reduced pressure on a rotary
2
4
evaporator. (b) To the obtained colorless oily residue was added CBr (1.2 mmol) and dry
4
o
DCM (10 mL), and the resulting solution was stirred at 0 C while to which a solution of PPh
3
1
2

Boc
NH
(CH )
Boc
NH
(CH )
NH2
CH2)n
NH2 HBr
(CH )
PPh , CBr
4
1. 50% TFA/DCM
. aq. HBr
(
Boc) O
3
2
(
2 n
2 n
2 n
2
(n=3,8)
HO
HO
Br
Br
Scheme 3. The synthetic scheme of 5-bromopentylamine hydrobromide (n=3) and
0-bromodecylamine hydrobromide (n=8).
1
(1.2 mmol) in dry DCM (5 mL) was added. After the addition was complete, the reaction
mixture was stirred at room temperature overnight before being concentrated under reduced
pressure on a rotary evaporator. To the residue was then added hexanes (30 mL) while stirring,
the resulting precipitate was filtered off and the filtrate was concentrated under reduced
pressure again, the resulting white crystalline precipitate was filtered off again and the filtrate
was then concentrated under reduced pressure. (c) To the residue (a white solid) was added 50%
(v/v) TFA/DCM (6 mL) and the resulting solution was stirred at room temperature for 8 h
before being concentrated. The resulting residue was then dissolved in cold diethyl ether (6
mL), and to the stirred solution was added hydrobromic acid (≧40%) (~1.0 mmol). The
resulting white precipitate was collected via filtration and washed with cold diethyl ether,
affording the crude material that predominantly contained 5-bromopentylamine hydrobromide
1
(
or 10-bromodecylamine hydrobromide). 5-bromopentylamine hydrobromide: H NMR (400
MHz, CD OD) δ 3.50 (t, J = 8.0 Hz, 2H), 2.96 (t, J = 8.0 Hz, 2H), 1.97-1.89 (m, 2H),
3
+
1
1
.75-1.67 (m, 2H), 1.61-1.53 (m, 2H); ESI-MS calcd for C H BrN ([M+H] ) 166.02; found:
5
13
1
66.03. 10-bromodecylamine hydrobromide: H NMR (400 MHz, CD OD) δ 3.46 (t, J = 8.0
3
Hz, 2H), 2.93 (t, J = 8.0 Hz, 2H), 1.89-1.82 (m, 2H), 1.71-1.64 (m, 2H), 1.49-1.37 (m, 12H);
+
ESI-MS calcd for C H BrN ([M+H] ) 236.10; found: 236.23. The two crude preparations
1
0
23
were then used directly for the synthesis depicted in Scheme 2.
1
3

2
.2.4. Preparation of bicyclic pentapeptide 10 (Scheme 4)
This compound resulted from the condensation between the cyclic dipeptidic amine and
the side chain fully protected cyclic tripeptidic acid. (a) Preparation of the cyclic dipeptidic
amine: This particular cyclic dipeptidic amine (b=3) was synthesized as above-described
according to Scheme 2. The obtained crude material was used directly for the condensation
with the crude side chain fully protected cyclic tripeptidic acid whose preparation is detailed
below. (b) Preparation of the side chain fully protected cyclic tripeptidic acid: This

started with the loading of Fmoc-N -thioacetyl-lysine onto the 2-chlorotrityl chloride resin in
o
the presence of DIPEA (2 equiv.) and DCM initially at 0-2 C for 15 min and then at room
temperature for 1 h, which was followed by treating the reaction mixture with DIPEA/MeOH
(v/v, 1/9) at room temperature for 15 min. The reaction mixture was then filtered and washed
with DMF, isopropanol, and DCM, affording the 2-chlorotrityl resin loaded with

Fmoc-N -thioacetyl-lysine. The subsequent sequential incorporation of Tyr(OH) and His(Trt)
onto the charged 2-chlorotrityl resin was performed based on the Fmoc chemistry-based SPPS.
Specifically, while Tyr(OH) was introduced with 2 equiv. of both Fmoc-Tyr(OH)-OH and
HATU in 0.4 N NMM/DMF at room temperature for 2 h, the introduction of His(Trt) was
realized with 4 equiv. of Fmoc-His(Trt)-OH and 3.8 equiv. of both HBTU and HOBt in 0.4 N

NMM/DMF at room temperature for 1 h. The N-term N -Fmoc removal with 20% (v/v)
piperidine/DMF (2 x 10 min at room temperature) afforded the peptidyl resin with the free

N-term N -amino group and the free side chain phenolic group of Tyr(OH). The N-term

-amino group was then selectively acylated with 4-bromobutyric acid (a=1) (4 equiv.) in the
presence of HATU (3.8 equiv.) and 0.4 N NMM/DMF (2 x 1 h at room temperature). The free
1
4

S
S
S
Trt
HN
CH3
O
H3C
NH
H3C
NH
N
COOH
CH2)a
(
N
Fmoc
OH
O
Cl
HO
N
H
Cl
O
SPPS
Cl
O
H
O
Br
Fmoc
O
N
H2N
N
N
HATU,
H
H
O
0.4N NMM/DMF
2-Chlorotrityl
chloride resin
HO
HN
S
S
S
Trt
CH3
Trt
HN
CH3
OH
Trt
HN
CH3
N
N
N
N
N
N
HFIP/DCM (1/1)
r.t.; 10 min)
O
O
Cl
O
O
H
Cl
O
H
H
(
NaI, 1% DBU/DMF
N
N
N
N
H
O
HN
N
O
HN
N
HN
O
O
2
H
H
O
O
O
O
O
(CH2)a
(H C)
a(H2C)
a
O
O
Br
HO
S
H3C
O
S
S
HN
CH3
HN
CH3
H
Trt
CH3
HN
H2N
O
N
N
H
N
CH3
S
S
HN
N
O
O
N
H
H
N
O
O
(CH2)b
H
H
N
HFIP/DCM (1/1)
(r.t.; 4 h)
N
O
N
O
HN
N
N
H
O
HN
N
N
H
H
O
O
O
H
O
O
O
HATU, DIPEA
HN
a(H2C)
(CH2)b
a(H2C)
(CH2)b
O
O
O
O
10: a=1, b=3
Scheme 4. The synthetic scheme of the bicyclic analog 10. Note: The cyclic dipeptidic amine
b=3) whose chemical structure is highlighted in red was synthesized as depicted in Scheme
(
2
.
side chain phenolic group of Tyr(OH) was subsequently allowed to react intramolecularly
with the N-term bromide in the presence of NaI (2 equiv.) and 1% (v/v) DBU/DMF (2 x 36 h
at room temperature). The final treatment with 50% (v/v) HFIP/DCM at room temperature for
1
0 min afforded the crude side chain fully protected cyclic tripeptidic acid. (c) Preparation of
bicyclic pentapeptide 10: The obtained crude cyclic dipeptidic amine and side chain fully
protected cyclic tripeptidic acid were subsequently condensed in THF with HATU (1.2 equiv.)
and DIPEA (1.5 equiv.) at room temperature for 24 h. The reaction mixture was then
concentrated under reduced pressure on a rotary evaporator and the obtained residue was
taken into ethyl acetate. The organic layer was washed once with 5% (w/v) aqueous tartaric
acid solution and once with saturated aqueous Na CO solution, and dried over anhydrous
2
3
1
5

Na SO . After filtration, the filtrate was concentrated under reduced pressure on a rotary
2
4
evaporator, and the obtained residue was then treated with 50% (v/v) HFIP/DCM at room
temperature for 4 h followed by the removal of the volatiles in the reaction mixture under
reduced pressure on a rotary evaporator. To the resulting residue was added cold diethyl ether
to precipitate out the crude pentapeptide 10 which was purified with semi-preparative
RP-HPLC on a C18 column (1 x 25 cm, 5 m) eluted with a gradient of ddH O containing
2
0
.05% (v/v) of TFA and acetonitrile containing 0.05% (v/v) of TFA at 4.5 mL/min and
monitored at 214 nm. The pooled desired RP-HPLC fractions were concentrated under
reduced pressure on a rotary evaporator and the remaining aqueous solution was lyophilized
to afford the purified pentapeptide 10 as a puffy white solid. Of note, for a synthetic scale of
0
.1 mmol, the purified 10 was obtained in an overall yield of 1.4%. The purified 10 was >95%
pure per RP-HPLC analysis on an analytical C18 column (0.46 x 25 cm, 5 m), and its exact
mass was confirmed with HRMS analysis (Table 1).
2
.3. In vitro sirtuin inhibition assay with compounds 2-10
The HPLC-based sirtuin inhibition assay was employed in the current study and was
performed as described previously38-40. An assay solution (50 L) contained the following
components: 50 mM Hepes (pH 8.0), 137 mM NaCl (150 mM for the SIRT7 assay), 2.7 mM
KCl and 1 mM MgCl (both omitted for the SIRT7 assay), 1 mM DTT, yeast tRNA (3-fold
2
molar excess relative to SIRT7) and SUPERase-In RNase inhibitor (to a final concentration of
+
~
1 U/L, to inhibit tRNA degradation) (both for the SIRT7 assay only), -NAD (0.5 mM for
the SIRT1 and SIRT2 assays, 3.5 mM for the SIRT3 assay, 0.8 mM for the SIRT5 assay, 0.2
1
6

mM for the SIRT6 assay, or 3 mM for the SIRT7 assay), the peptide substrate (0.3 mM of the
above-mentioned SIRT1/2/3 acetyl-lysine substrate for the SIRT1 assay, 0.39 mM of the
above-mentioned SIRT1/2/3 acetyl-lysine substrate for the SIRT2 assay, 0.105 mM of the
above-mentioned SIRT1/2/3 acetyl-lysine substrate for the SIRT3 assay, 0.88 mM of the
above-mentioned SIRT5 succinyl-lysine substrate, 0.02 mM of the above-mentioned SIRT6
myristoyl-lysine substrate, or 1.8 mM of the above-mentioned SIRT7 acetyl-lysine substrate),
one test compound (purified 2-10) with varied concentrations including 0, and a sirtuin
(
His -SIRT1 or GST-SIRT1, 320 nM; His -SIRT2, 309 nM; His -SIRT3, 320 nM;
6 6 6
GST-SIRT5, 370 nM; His -SIRT6, 313 nM; or His -SIRT7, 570 nM). The same [S]/Km
6
6
+
ratios for both substrates (~3.2 for the peptide substrates and ~5.6 for -NAD ) were
employed for the inhibition assays with all the six human sirtuins (SIRT1/2/3/5/6/7) employed
in the current study. An enzymatic reaction was initiated by the addition of a sirtuin at 37 °C
and was allowed to be incubated at 37 °C for 10 min (for the SIRT1 assay), 12 min (for the
SIRT2 assay), 10 min (for the SIRT3 assay), 5 min (for the SIRT5 assay), 12 min (for the
SIRT6 assay), or 30 min (for the SIRT7 assay) until quenched with the following stop
solution: 100 mM HCl and 0.16 M acetic acid. The quenched assay solutions were directly
injected into a reversed-phase C18 column (0.46 x 25 cm, 5 m) eluted with a gradient of
ddH O containing 0.05% (v/v) TFA and acetonitrile containing 0.05% (v/v) TFA at 1 mL/min
2
and monitored at 214 nm. Turnover of the limiting substrate was maintained at < 10%. Stock
solutions of the test compounds were all prepared in ddH O. IC values were estimated from
2
50
4
1
the Dixon plots (1/v vs.[inhibitor]) as an indication of the inhibitory potency.
0
1
7

2
.4. Pronase digestion assay with compounds 1 and 10
This assay was also performed as described previously . Fifty (50) µL of a solution of a
3
9
test compound (1 or 10) in ddH O (160 µM) was mixed thoroughly with 50 µL of a pronase
2
solution in 100 mM Tris•HCl (pH 7.3) (8 ng/µL), and the resulting solution was incubated at
3
7 °C until quenched with a 1.0 M solution of acetic acid in ddH O at 0, 1, and 2 min for 1 or
2
at 0, 2.5, 5.0, 10, 20, and 30 min for 10. At each time point, 20 µL of a pronase digestion
mixture was taken and treated with 40 µL of the 1.0 M acetic acid solution, and the whole
mixture was vigorously vortexed, centrifuged, and the supernatant was injected into a
reversed-phase C18 analytical HPLC column (0.46 x 25 cm, 5 m). The C18 column was
eluted with a gradient of ddH O containing 0.05% (v/v) TFA and acetonitrile containing 0.05%
2
(v/v) TFA at 1 mL/min and monitored at 214 nm. The HPLC peak areas obtained for a given
test compound at different time points were used to estimate the percentage remaining for this
test compound with the digestion time. The graph of the percentage remaining versus time
was used to compare the proteolytic stability of different test compounds, as shown in Fig. 3.
2
.5. Western blotting analysis with compound 10
This assay was also performed as previously described by Zheng laboratory35,39. Briefly,
5
HCT116 human colon cancer cells (Procell, Wuhan, China) (5 x 10 ) were cultured in
McCoy5A culture medium containing 10% FBS and were treated for 8 h at 37 °C with
compound 10 (stock solution prepared in ddH O) at different final concentrations (0, 0.5, 2,
2
1
0, 50, and 100 µM) before cell lysis and SDS-PAGE. Proteins on a SDS-PAGE gel were
then transferred onto a PVDF membrane (Millipore). The transblotted membrane was blocked
1
8

for 2 h at room temperature and then incubated overnight at 4 °C with the following primary
antibodies: acetyl-p53 (K382) antibody (Cell Signaling Technology, United States) and p53
antibody (Proteintech-CN, Wuhan, China). After the treatment with the above primary
antibodies, the transblotted membrane was treated with the goat anti-rabbit IgG (H+L)
secondary antibody-horseradish peroxidase (HRP) conjugate (Boster (PRC), Wuhan, China)
for 2 h at 37 °C, and the immunoblots (for K382-acetylated p53 protein and total p53 protein)
were visualized by enhanced chemiluminescence.
2
.6. Cell proliferation assay with compound 10
This assay was performed as previously described by George et al. on the
4
2
anti-proliferative effect of 4’-bromo-resveratrol on melanoma cells . Of note,
4
’-bromo-resveratrol was found earlier by Nguyen et al. to be a potent inhibitor against the
4
3
deacetylase activity of human SIRT1 and SIRT3 under the assay condition of the study . The
human MCF-7 breast cancer cells (Procell, Wuhan, China) and the human SK-MEL-2
melanoma cells (Beijing zhongkezhijian Biotechnology Co. Ltd., Beijing, China) were
employed in our current study, and the anti-proliferative effect of compound 10 was assessed
with the 3‐(4,5-dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT)-based assay.
3
3
Briefly, MCF-7 cells (5 x 10 , 100 L cell suspension) or SK-MEL-2 cells (5 x 10 , 100 L
cell suspension) were cultured in 96-well plates in DMEM culture medium containing 10 %
FBS (for MCF-7 cells) or MEM culture medium containing 10% FBS (for SK-MEL-2 cells)
overnight and then were treated at 37 °C with compound 10 (stock solution prepared in
ddH O) at different final concentrations (0, 5, 12.5, 25, 50, or 100 µM) for 24 h, 48 h, or 72 h.
2
1
9

The cells in each treatment well were then washed with phosphate buffered saline (PBS) and
subsequently incubated with 10 L of a 5 mg/mL solution of MTT in PBS (to a final
concentration of 0.5 mg/mL) at 37 °C for 4 h. After the removal of the incubation buffer by
pipetting, to each treatment well was then added 150 L of DMSO, and the resulting mixture
was agitated for 10 min before the absorbance at 568 nm was measured for each treatment
well. The assay results are presented in Fig. 5.
3
. Results and Discussion
3
.1. Compound design
Fig. 2 depicts the monocyclic and bicyclic pentapeptides 2-10 we prepared in the
current study which are the analogs of the linear pentapeptide-based potent (sub-M IC ’s)
5
0
4
4
pan-SIRT1/2/3 inhibitor 1 Zheng laboratory discovered recently that harbors the catalytic

mechanism-based SIRT1/2/3 inhibitory warhead N -thioacetyl-lysine (discovered during
3
6
45
2
006-2007 by Zheng laboratory and that of Denu ) at its central position. Despite being a
potent pan-SIRT1/2/3 inhibitor, the linear peptide 1 could be less proteolytically stable and
cell permeable than its cyclic analogs because peptide chain cyclization is known to be able to
potentially enhance the proteolytic stability and the cell permeability of a parent linear
4
6
peptide . To pursue the cyclic analogs of 1, we initially designed and prepared four
monocyclic analogs with N-term macrocycles and four monocyclic analogs with C-term
macrocycles depicted in Fig. 2, hoping to find the N-term and the C-term macrocycles able to
confer stronger SIRT1/2/3 inhibition than 1 so that these macrocycles could be employed to
2
0

CH3
S
M384
HN
+2 (M384)
K386
-2 (H380)
N
CH3
S
AcK382
HN
O
O
H
H
N
K381
L383
F385
T387
N
O
HN
N
N
H
CH3
H
O
O
NH2
O
Bridging
Unit
Bridging
Unit
CH3
H380
-
1 (K381)
NH2
H3C
1 (L383)
+
1
Bridg Ui nn gi t
Peptide Chain
Cyclization
CH3
CH3
S
CH3
S
HN
S
HN
HN
CH3
NH2
CH3
CH3
S
N
S
N
S
N
HN
HN
HN
O
O
O
O
O
H
O
H
H
N
H
H
N
H
N
N
N
O
N
O
N
N
H
CH3
HN
N
N
HN
O
O
N
N
HN
H
H
H
H
H
O
O
O
O
O
HN
O
O
HN
O
O
O
CH3
H3C
(CH2)n
(CH2)a
O
(CH2)b
(CH2)n
O
NH2
2
4
: n=1; 3: n=3;
: n=5; 5: n=8.
6: n=1; 7: n=3;
8: n=5; 9: n=8.
10: a=1, b=3
Fig. 2. The design of the monocyclic analogs 2-9 and the bicyclic analog 10 of the linear

pentapeptide 1 via peptide chain cyclization. Note: (i) The central N -thioacetyl-lysine residue

(
highlighted in red) of 1 corresponds to the N -acetyl-lysine residue (AcK382) of the peptide
substrate that is depicted to be bound at the active site of archaeal Sir2-Af2 in the upper left
-dimensional rendering (adapted from ref. 47 with permission) of the X-ray
3
crystallographically determined structure of the complex of Sir2-Af2 and the depicted peptide
substrate (Protein Data Bank code: 1MA3). (ii) The side chains at the -2, -1, +1, and +2
positions of 1 respectively correspond to the H380, K381, L383, and M384 residues of the
bound peptide substrate depicted in the upper left 3-dimensional rendering. (iii) The red,
yellow, and green regions in the upper left 3-dimensional rendering refer to the degree of
amino acid conservation (conserved, similar, and variable, respectively) among the following
ten sirtuins based on sequence alignment: Sir2-Af1, Sir2-Af2, Sir2-Tm, Sir2, Hst1, Hst2,
CobB, SIRT1, SIRT2, and SIRT5. (iv) The two thick red curves in the upper left depiction
refer to the macrocycle bridging units present in cyclic analogs 2-10.
construct bicyclic analogs, which was pursued because the other half of the molecule of a
monocyclic analog is still a proteolytically labile linear peptide harboring two consecutive
standard amino acids.
The monocyclic analogs 2-9 were designed based on the following: (i) the substrate
recognition behavior at a sirtuin active site would be applicable to a

N -thioacetyl-lysine-containing catalytic mechanism-based sirtuin inhibitor since such an
2
1

inhibitor would be first recognized and processed as a sirtuin substrate, however with the
formation of a sirtuin inhibiting species latter on along the catalytic coordinate8,32,48; (ii)

N -acetyl-lysine and its four immediately flanking amino acid residues (two on each side) in a
substrate form a -strand in a sirtuin bound substrate and this -strand forms an anti-parallel

-sheet with two -strands from a sirtuin active site, as first observed in the X-ray co-crystal
4
7
structure of archaeal Sir2-Af2 complexed with the substrate peptide depicted in Fig. 2 ; (iii)
the tyrosine replacement for the native amino acid residues at -1 and +1 positions relative to

N -thioacetyl-lysine was previously demonstrated by Zheng laboratory to be tolerable
concerning the SIRT1 inhibitory potency with the sequence of

39
HK-(N -thioacetyl-lysine)-LM ; (iv) the N-term and the C-term macrocycles in the
monocyclic pentapeptides 2-9 are or closely resemble those that have been demonstrated
previously by Fairlie and co-workers to be effective local -strand inducers in a peptide
sequence49,50. In addition, according to the 3-dimensional rendering depicted in Fig. 2, the
N-term and C-term macrocycle units in pentapeptides 2-9 would have a favorable global fit at
sirtuin active site.
3
.2. Compound preparation
The monocyclic pentapeptides 2-5 with N-term macrocycle units were synthesized
according to Scheme 1. The linear pentapeptide composed of His(Trt), Tyr(2-Cl-Trt),

N -thioacetyl-lysine, Leu, and Met was initially assembled on the Rink amide MBHA resin

based on the Fmoc chemistry-based manual SPPS followed by the N-term N -Fmoc removal,
the N-term free -amino group was then acylated with a bromoalkanoic acid (4-bromobutyric
2
2

acid (n=1), 6-bromohexanoic acid (n=3), 8-bromooctanoic acid (n=5), or
1-bromoundecanoic acid (n=8)) in the presence of HATU and 0.4 N NMM/DMF. The side
1
chain 2-Cl-Trt protecting group of Tyr(2-Cl-Trt) was then selectively removed with 50% (v/v)
HFIP/DCM, and the exposed free side chain phenolic group was subsequently allowed to
react intra-molecularly with the N-term bromide in the presence of NaI and 1% (v/v)
DBU/DMF. The final treatment of the resulting peptidyl-resin with Reagent K (a
TFA-containing peptide cleavage cocktail) afforded the crude pentapeptides 2-5 which were
each purified by semi-preparative RP-HPLC. The exact masses of the purified 2-5 were
confirmed by HRMS analysis (Table 1). The purified 2-5 were each >95% pure based on a
RP-HPLC analysis.
The monocyclic pentapeptides 6-9 with C-term macrocycle units were synthesized
according to Scheme 2. As depicted, these compounds each resulted from the condensation
between a cyclic dipeptidic amine and a side chain fully protected linear tripeptidic acid. The
synthesis of a cyclic dipeptidic amine started with the loading of Fmoc-Met onto the
2
-chlorotrityl chloride resin, followed by Fmoc removal and the subsequent coupling with
Boc-Tyr-OH. Following treating the resulting dipeptidyl resin with 50% (v/v) HFIP/DCM,
the dipeptidic acid cleaved from the resin was reacted in solution with a bromoalkylamine
hydrobromide (3-bromopropylamine hydrobromide (n=1), 5-bromopentylamine
hydrobromide (n=3), 7-bromoheptylamine hydrobromide (n=5), or 10-bromodecylamine
hydrobromide (n=8)). The phenolic group and the bromide in the resulting intermediate were
then allowed to undergo an intra-molecular cyclization in the presence of MeONa and NaI.
The final treatment with 50% (v/v) TFA/DCM afforded the crude cyclic dipeptidic amine
2
3

which was used directly for the later condensation with the side chain fully protected linear
tripeptidic acid, whose synthesis is detailed below starting with the loading of

Fmoc-N -thioacetyl-lysine onto the 2-chlorotrityl chloride resin. The sequential incorporation
of Lys(Boc) and His(Trt) onto the charged 2-chlorotrityl resin based on the Fmoc

chemistry-based SPPS followed by the N-terminal N -acetylation with acetic anhydride
afforded the side chain fully protected linear tripeptidyl resin. The final treatment with 50%
(v/v) HFIP/DCM afforded the crude side chain fully protected linear tripeptidic acid which
was used directly for the condensation with the above-described crude cyclic dipeptidic amine
in solution, and the resulting intermediate was then treated with a TFA-containing peptide
cleavage cocktail, affording the crude pentapeptides 6-9 which were each also purified by
semi-preparative RP-HPLC. The exact masses of the purified 6-9 were also confirmed by
HRMS analysis (Table 1). The purified 6-9 were each also >95% pure based on a RP-HPLC
analysis.
The bicyclic pentapeptide 10 was synthesized according to Scheme 4. As depicted, this
compound resulted from the condensation between the cyclic dipeptidic amine and the side
chain fully protected cyclic tripeptidic acid. The synthesis of the side chain fully protected

cyclic tripeptidic acid started with the loading of Fmoc-N -thioacetyl-lysine onto the
2
-chlorotrityl chloride resin. The subsequent sequential incorporation of Tyr(OH) and His(Trt)

based on the Fmoc chemistry-based SPPS followed by the N-term N -Fmoc removal afforded

the peptidyl resin with the free N-term N -amino group and the free side chain phenolic group
of Tyr(OH). The N-term -amino group was then selectively acylated with 4-bromobutyric
acid (a=1) in the presence of HATU and 0.4 N NMM/DMF. The free side chain phenolic
2
4

group of Tyr(OH) was subsequently allowed to react intra-molecularly with the N-term
bromide in the presence of NaI and 1% (v/v) DBU/DMF. The final treatment with 50% (v/v)
HFIP/DCM afforded the crude side chain fully protected cyclic tripeptidic acid which was
subsequently used directly for the condensation in solution with the crude cyclic dipeptidic
amine (b=3) which was synthesized as above-described according to Scheme 2. The resulting
intermediate was then treated with 50% (v/v) HFIP/DCM (4 h at room temperature), affording
the crude bicyclic pentapeptide 10 which was also purified by semi-preparative RP-HPLC.
The exact mass of the purified 10 was also confirmed by HRMS analysis (Table 1). The
purified 10 was also >95% pure based on a RP-HPLC analysis.
3
.3. In vitro sirtuin inhibition assay
When the purified monocyclic pentapeptides 2-9 were subjected to a HPLC-based in
vitro sirtuin inhibition assay, we found that the SIRT1/2/3 inhibitory potencies of 2-5
especially 2) were much closer to those of the linear pentapeptide 1 than the SIRT1/2/3
(
inhibitory potencies of 6-9 (Table 2). Therefore, it is clear that, even though the tyrosine side

chains at both -1 and +1 positions relative to N -thioacetyl-lysine in the template sequence

HK-(N -thioacetyl-lysine)-LM could be accommodated favorably at a sirtuin active site, as
3
9
previously demonstrated with SIRT1 , the engagement of the side chain phenolic group of
the tyrosine residue at the +1 position as in the C-term macrocycles of 6-9 led to dramatically
attenuated binding at SIRT1/2/3 active sites. While the above finding with monocyclic
pentapeptides 6-9 is disappointing, their macrocycles may still have a favorable global fit
(albeit with unfavorable binding details) at sirtuin active site according to the 3-dimensional
2
5

Table 2. SIRT1/2/3 inhibition by compounds 1-10a
b
IC (M)
5
0
Compound
SIRT1
SIRT2
SIRT3
1c
2
3
4
5
6
7
8
9
0.165 ± 0.048
0.171 ± 0.059
0.646 ± 0.204
0.673 ± 0.199
1.25 ± 0.57
34.3 ± 8.3
0.332 ± 0.062
0.68 ± 0.3
0.27 ± 0.024
0.75 ± 0.3
0.53 ± 0.16
0.55 ± 0.26
1.1 ± 0.54
90.9 ± 7.6
177.1 ± 47.4
217.3 ± 38.7
13.1 ± 3.0
0.72 ± 0.36
1.03 ± 0.29
1.33 ± 0.078
4.31 ± 2.74
60.8 ± 15.6
115.1 ± 41.6
25.5 ± 0.85
17.9 ± 0.42
0.253 ± 0.02
13.6 ± 1.9
22.6 ± 7.4
56.7 ± 7.4
1
0
0.54 ± 0.07
a
The SIRT1/2/3-catalyzed reaction monitored in the inhibition assay was deacetylation. See
b
“
Materials and Methods” for assay detail. IC , the inhibitor concentration at which an
5
0
c
enzymatic reaction velocity is reduced by 50%. Taken from ref. 44 in which compound 1
was evaluated under identical SIRT1/2/3 inhibition assay condition to that used in the current
study.
rendering depicted in Fig. 2, we were therefore interested in constructing a bicyclic
pentapeptide composed of the macrocycles selected from those in the monocyclic
pentapeptides 2-9, hoping the binding of the N-term macrocycle would incur a better binding
at sirtuin active site for the C-term macrocycle. Fig. 2 shows the chemical structure of the
bicyclic pentapeptide (i.e. 10) we prepared in the current study. Its N-term macrocycle
corresponds to that of 2 since 2 was the most potent SIRT1/2/3 inhibitor among 2-5; its
C-term macrocycle corresponds to that of 7 since 7 exhibited the strongest SIRT1 inhibition
among 6-9.
When the purified bicyclic pentapeptide 10 was subjected to the HPLC-based in vitro
sirtuin inhibition assay, this compound was found to exhibit a strong pan-SIRT1/2/3
2
6

inhibition. As shown in Table 2, the SIRT1/2/3 inhibitory potencies of the bicyclic
pentapeptide 10 are largely comparable to those of the linear pentapeptide 1 (IC values
5
0
within 3.3-fold of difference). We further found that the inhibition of 10 against the
SIRT5/6/7-catalyzed deacylation reactions (i.e. desuccinylation, demyristoylation, and the
tRNA-activated deacetylation, respectively) was very weak with respective IC values being
5
0
4
42.7 ± 178.9 M, 390.2 ± 23.3 M, and 78.9 ± 13.9 M, which ought to be consistent with
the fact that deacetylation is a less proficient catalytic activity of SIRT5/6/71,4,5,7,8,51
.
3
.4. Pronase digestion assay
Even though the bicyclic pentapeptide 10 was found to be a comparably strong
pan-SIRT1/2/3 inhibitor to the linear pentapeptide 1, the potential of peptide chain cyclization
4
6
to enhance the proteolytic stability of a parent linear peptide prompted us to subject the fully
cyclic compound 10 and the fully linear compound 1 to a proteolysis experiment to compare
their proteolytic stability. Pronase was used in the experiment as the proteolytic enzyme
preparation because pronase is a mixture of different types of proteases and peptidases and
5
2
thus exhibits a very broad substrate specificity . As indicated in Fig. 3, the bicyclic
pentapeptide 10 is proteolytically much more stable than the linear pentapeptide 1. Therefore,
compound 10 rather than compound 1 would be a more appropriate lead compound for further
development. As such, in the current study compound 10 was evaluated further with the
following preliminary studies.
2
7

3
.5. Western blotting analysis
As revealed by such an analysis and indicated in Fig. 4, the bicyclic pentapeptide 10 was
found to be cell permeable and able to inhibit the SIRT1-catalyzed deacetylation of the
5
3
K382-acetylated tumor suppresser protein p53 (a native SIRT1 substrate ) inside the human
HCT116 colon cancer cells, as judged by a concentration-dependent increase in the p53
acetylation at K382 following the treatment of cells with 10.
Fig. 3. The proteolysis profiles of compounds 10 (●) and 1 (▲). Each data point is the
mean of two measurements with standard deviation less than 15%. See “Materials and
Methods” for the proteolysis assay details.
Fig. 4. The Western blotting analysis of the tumor suppresser protein p53 K382-acetylation
level change in the human HCT116 colon cancer cells following the treatment with
compound 10. See “Materials and Methods” for the experimental details.
3
.6. Cell proliferation assay
The human MCF-7 breast cancer cells and the human SK-MEL-2 melanoma cells were
employed in this assay, and the anti-proliferative effect of compound 10 was assessed with
2
8

the MTT-based assay. The roughly comparable SIRT1/2 dual inhibition by compound 10
made us select the human MCF-7 cells for our current assay since a dual SIRT1/2 inhibition
was previously demonstrated with the human MCF-7 cells to be necessary for cell growth
5
4
inhibition . Moreover, the roughly comparable SIRT1/3 dual inhibition by compound 10
made us also select the human SK-MEL-2 cells for our current assay since this type of the
cells were also used previously to examine the anti-proliferative effect of
4
2
4
’-bromo-resveratrol which was found earlier to be a potent SIRT1/3 dual inhibitor under
4
3
the assay condition of the study , even though its inhibitory action on other human sirtuins
e.g. SIRT2) was not assessed in the study. As indicated in Fig. 5, while there is a general
(
trend of increased concentration-dependent cell growth inhibition as the treatment time of
compound 10 was increased (from 24 h to 48 h and to 72 h) with both types of the cells, the
overall inhibitory strength of compound 10 is much greater toward the human SK-MEL-2
cells than that toward the human MCF-7 cells. This finding suggests that a SIRT1/2/3 triple
inhibition seems to be more able to confer a growth inhibition against the human melanoma
cells than against the human breast cancer cells. It should be noted that, under our assay
condition, tamoxifen treatment at 100 M and 24h/48h/72h was found to be all able to confer
an over 80% growth inhibition against the human MCF-7 cells (data not shown), suggesting
that the observed weak inhibition of compound 10 treatment was not due to possible problems
with the human MCF-7 cells used in our assay.
4
. Conclusion
To summarize, we found in the current study that the bicyclic analog 10 of the
2
9