Discovery of thieno[3,2-d ]pyrimidine-6-carboxamides as potent inhibitors of SIRT1, SIRT2, and SIRT3

Article abstract of DOI:10.1021/jm400204k

The sirtuins SIRT1, SIRT2, and SIRT3 are NAD⁺ dependent deacetylases that are considered potential targets for metabolic, inflammatory, oncologic, and neurodegenerative disorders. Encoded library technology (ELT) was used to affinity screen a 1

Full text of DOI:10.1021/jm400204k

Subscriber access provided by BALL STATE UNIV
Article
Discovery of Thieno[3,2-d]pyrimidine-6-carboxamides
as potent inhibitors of SIRT1, SIRT2 and SIRT3.
Jeremy S. Disch, Ghotas Evindar, Cynthia H. Chui, Charles A. Blum, Han Dai, Lei Jin, Eli Schuman,
Kenneth E. Lind, Svetlana L. Belyanskaya, Jianghe Deng, Frank T. Coppo, Leah Aquilani, Todd
L. Graybill, John W. Cuozzo, Siva Lavu, Cheney Mao, George P. Vlasuk, and Robert Bruno Perni
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400204k • Publication Date (Web): 09 Apr 2013
Downloaded from http://pubs.acs.org on April 15, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
Discovery of Thieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamides
7
8
9
as potent inhibitors of SIRT1, SIRT2 and SIRT3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jeremy S. Disch*, Ghotas Evindar†*, Cynthia H. Chiu†, Charles A. Blum, Han Dai, Lei Jin, Eli
Schuman, Kenneth E. Lind†, Svetlana L. Belyanskaya†, Jianghe Deng‡, Frank Coppo‡, Leah Aquilani‡,
Todd L. Graybill‡, John W. Cuozzo†, Siva Lavu, Cheney Mao⊥, George P. Vlasuk, and Robert B. Perni
Sirtris a GSK Company, Cambridge, Massachusetts, United States
† ELT Boston, Platform Technology and Science, GlaxoSmithKline, Waltham, Massachusetts, United
States
‡ ELT UP, Platform Technology and Science, GlaxoSmithKline, Upper Providence, Pennsylvania,
United States
⊥ Viva Biotech Ltd., 334 Aidisheng Road, Zhangjiang Highꢀtech Park, Shanghai, China
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 45
ABSTRACT
1
2
3
4
5
6
7
8
9
The sirtuins SIRT1, SIRT2 and SIRT3 are NAD+ dependent deacetylases that are considered potential
targets for metabolic, inflammatory, oncologic and neurodegenerative disorders. Encoded Library
Technology (ELT) was used to affinity screen a 1.2 million heterocycle enriched library of DNA
encoded small molecules, which identified panꢀinhibitors of SIRT1/2/3 with nanomolar potency (e.g.
11c: IC50 = 3.6, 2.7 and 4.0 nM for SIRT1, SIRT2 and SIRT3 respectively). Subsequent SAR studies to
improve physiochemical properties identified the potent drug like analogs 28 and 31. Crystallographic
studies of 11c, 28 and 31 bound in the SIRT3 active site revealed that the common carboxamide binds
in the nicotinamide Cꢀpocket and the aliphatic portions of the inhibitors extend through the substrate
channel, explaining the observable SAR. These pan SIRT1/2/3 inhibitors, representing a novel
chemotype, are significantly more potent than currently available inhibitors which makes them valuable
tools for sirtuin research.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS
SIRT1, SIRT2, SIRT3, sirtuin inhibitor, pan SIRT1/2/3 inhibitor, thieno[3,2ꢀd]pyrimidineꢀ6ꢀ
carboxamides, Xꢀray structure, Encoded Library Technology, ELT.
ACS Paragon Plus Environment

Page 3 of 45
Journal of Medicinal Chemistry
INTRODUCTION
1
2
3
4
In recent years, the sirtuins, a series of seven Class III NAD+ dependent histone deacetylases, have
been implicated as important regulators of numerous cellular processes and play physiological roles
related to lifespan and aging.1ꢀ2 Modulating sirtuin activity3, either through overexpression (activation)
or knockdown (inhibition), has been proposed to be beneficial in numerous disease states including
those related to metabolism,4 cancer,5ꢀ6 neurodegeneration,7ꢀ9 inflammation,10ꢀ11 and ischaemic injury.12
Currently our understanding of sirtuin function is far from complete. Conflicting reports abound even
for SIRT1, which is by far the most studied of the seven sirtuins. For example, in one study SIRT1
knockꢀout mice displayed increased weight gain and decreased glucose tolerance.13 While in another
study, SIRT1 deficient mice showed increased glucose tolerance, decreased weight gain and lower fat
accumulation in some tissues.14 What is clear is that the function of these enzymes is dependent on their
cellular localization and the type of tissue where the cells reside.1
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nevertheless, there is compelling evidence that sirtuin modulation could find significant clinical use.
SIRT1 and SIRT2 both fulfill the same function of deacetylating p53, which is a tumor suppressor
protein involved in regulating a pathway leading toward apoptosis. Though inactivating SIRT1 or
SIRT2 individually was insufficient for inhibiting the deacetylation of p53 in vivo, the simultaneous
inhibition of SIRT1 and SIRT2 was beneficial against cancer by completely blocking p53 deacetylation,
which lead to cell death.5 In a neurodegenerative setting, evidence suggests that the SIRT2 mediated
deacetylation of FOXO3a promotes neuronal damage leading to apoptotic pathways, and that SIRT2
knockout prevents neuronal death in a Parkinson disease mouse model.7
One of the common structural features of all seven sirtuins is a largely conserved catalytic core while
otherwise incorporating significantly variable Cꢀ and Nꢀterminal regions. This conserved core leads to
the assumption that it may be possible to identify pan inhibitors of these enzymes as a means of
therapeutic intervention. Alternatively, it also suggests that it may be difficult to design selective active
site inhibitors.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 45
To date, there have been several reports identifying sirtuin inhibitors, largely focused on SIRT1 and
SIRT2. Among the earliest SIRT1/2 inhibitors identified were sirtinol (1, Figure 1) and the closely
related salermide (2), both which display marginal selectivity for SIRT2.15ꢀ16 Suramin (3), a high
molecular weight urea derivative, also inhibited SIRT1 and SIRT2 with little selectivity (IC50 = 0.3 ꢀM
1
2
3
4
5
6
7
8
9
and 1.2 ꢀM respectively).16 The tetrahydrocarbazole inhibitor, EX527 (4) on the other hand, is more
selective for SIRT1 over SIRT2 and SIRT3.17 This compound is among the most studied of the
published inhibitors and has been used as a standard inhibitor in biological studies. Overall, a broad
spectrum of compound classes have demonstrated sirtuin inhibition18 ranging from peptide substrate
mimetics (5)19ꢀ21 to heterocyclic small molecules such as cambinol (6).22 In general, these inhibitors
exhibit micromolar or high nanomolar potencies and tend to have moderate selectivity for SIRT1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
More recently, selective SIRT2 inhibitors have been reported including 2ꢀanilinobenzamide analogues
23
(7)
and the chromanone derivative 8.24 Benzamide 7 exhibits >500:1 preference for SIRT2 over
SIRT1 and SIRT3. Similarly, chromanone 8 shows selectivity over both SIRT1 and SIRT3, showing
less than 10% inhibition at 200 ꢀM against SIRT1/3. There is also a recent report25 showing that 3ꢀ(1Hꢀ
1,2,3ꢀtriazolꢀ4ꢀyl)pyridine, a nicotinamide analogue, was modestly selective for SIRT3 (IC50 = 38 ꢀM)
over SIRT1 and SIRT2 (88% and 92% activity remaining at 1 mM for SIRT1 and SIRT2 respectively),
and it demonstrated modest antiproliferative effects against several cancer cell lines. Ultimately, the
known classes of inhibitors all possess significant liabilities even as tool compounds. With the possible
exception of 4, they are generally poorly potent or suffer from poor physicochemical properties. In order
to develop a complete understanding of sirtuin function, there is a need for better tool compounds
composed of both sirtuin activators and inhibitors.
Herein, we report a new class of potent sirtuin inhibitors based on a thieno[3,2ꢀd]pyrimidineꢀ6ꢀ
carboxamide scaffold, (9, Figure 2). These inhibitors were initially discovered through an in vitro
SIRT3 affinity selection using Encoded Library Technology (ELT).26 These compounds represent a
ACS Paragon Plus Environment

Page 5 of 45
Journal of Medicinal Chemistry
novel class of very potent, low nanomolar, pan SIRT1/2/3 inhibitors, which could serve as aids in
elucidating the diverse biology of sirtuins.
1
2
3
4
5
6
7
8
RESULTS AND DISCUSSION
9
ELT Screening
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In our efforts to discover novel small molecules that modulate sirtuin activity, we utilized Encoded
Library Technology (ELT)26ꢀ27, a novel and robust hit identification platform that employs large
collections of chemotypically diverse DNAꢀencoded small molecule libraries which are screened for
their affinity towards a desired protein target. The technology provides access to a broad set of
chemotypes with vast structural diversity in an evolving library collection. It is also an attractive
platform because it uses negligible amounts of target protein to carry out selection experiments, and it
identifies ligands regardless of their functional activity. In the past few years, ELT has been successfully
used to identify hits against a number of soluble targets.28ꢀ32
In pursuit of novel small molecule modulators of SIRT3, we carried out an ELT selection campaign
against a FlagꢀSBPꢀtagged SIRT3 construct. FlagꢀSIRT3ꢀSBP was immobilized on streptavidin matrix
tips, and selections were performed under three conditions: SIRT3 alone; SIRT3 plus βꢀNicotinamide
Adenine Dinucleotide (NAD+); and SIRT3 plus thioacetylꢀpeptide ACS2 substrate
(TRSGKsꢀAcVMRRLLR).33 These SIRT3 selection conditions were used to screen a 3ꢀcycle linear
library capped with heteroaryl moieties (10, Figure 3a). The library was established by coupling 16 bisꢀ
acid building blocks (cycle 1) to the ELT headpiece (HP) which allowed for further elaboration of the
second carboxylate with 134 diamines (cycle 2). The second amine from cycle 2 was then
functionalized with 570 heteroaryl building blocks (cycle 3) to afford an ELT library with 1.2 million
enumerated compounds. The details of the synthesis of this novel encoded library will be the subject of
a separate publication.
The library was screened against SIRT3 (see Experimental Methods) and the sequencing data was
transferred into a cubic scatter plot for visualization and analysis within Spotfire™, where each axis
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 45
represents a cycle of diversity in the library (Figure 3b). The background noise, single hits, and low
copy number molecules were removed to simplify the data analysis. This allowed for closer observation
of the more highly enriched families and features within the cube.
1
2
3
4
5
6
7
8
9
The main selected chemotype was represented by a horizontal and a vertical line intersecting at a
single point in the cube. These lines defined a plane in cycle 3 originating from the 4ꢀchlorothieno[3,2ꢀ
d]pyrimidineꢀ6ꢀcarboxamide building block connected to cycle 2 through an amine displacement of the
chloride. The horizontal and vertical lines, themselves, originated from a combination of the selected
cycle 3 building block (4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide) and a specific cycle 1
(thiopheneꢀ2,5ꢀdicarboxylic acid) or cycle 2 building block (2ꢀ(piperidinꢀ4ꢀyl)ethanamine).
Knowing that there was selection for a large variety of cycle 1 and cycle 2 building blocks, the
common pharmacophore intersection product (11c, large dot) was chosen for offꢀDNA activity
confirmation. For simplicity, the attachment point to DNA was substituted by an ethylamide. Given the
larger variability observed for the selection output for both cycle 1 and cycle 2 residues, we chose one
additional cycle 1 (isophthalic acid) and cycle 2 building block (2ꢀ(piperazinꢀ1ꢀyl)ethanamine)) and
synthesized a 2 × 2 library for offꢀDNA biochemical activity confirmation (Scheme 1). This produced a
sufficient number of offꢀDNA compounds to confirm activity of the chemotype and allowed for
potential offꢀDNA preliminary SAR studies.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemistry
Assembly of the 2 × 2 library for offꢀDNA confirmation of the selected chemotype was accomplished
by carboxylating commercially available 4ꢀchlorothieno[3,2ꢀd]pyrimidine (12, Scheme 1) to obtain
carboxylic acid 13. Treating 13 with oxalyl chloride, and quenching the intermediate acid chloride with
ammonia in dioxane afforded the versatile 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide intermediate
14. The chloride on 14 was subsequently displaced with 4ꢀ(2ꢀBocꢀaminoethyl)piperidine or 4ꢀ(2ꢀBocꢀ
aminoethyl)piperazine to afford the Bocꢀprotected precursors (15a-b). The Boc groups were removed
by treatment with trifluoroacetic acid in CH2Cl2, and the resulting amines (16aꢀb) were reacted under
ACS Paragon Plus Environment

Page 7 of 45
Journal of Medicinal Chemistry
amide coupling conditions (HATU, DIEA) with (3ꢀ(ethylcarbamoyl)benzoic acid or 5ꢀ
(ethylcarbamoyl)thiopheneꢀ2ꢀcarboxylic acid to afford 11a-d. Most other analogs in this study were
synthesized through similar transformations (see Supplemental Information).
1
2
3
4
5
6
7
8
9
SAR Studies
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The four offꢀDNA synthesized representative compounds (Table 1, 11a-d) were evaluated for their
ability to inhibit the sirtuin mediated deacetylation of a minimal natural peptide substrate (Acꢀ
RHKKAcWꢀNH2)34 in a biochemical assay with SIRT1, SIRT2 and SIRT3. The activity data was
evaluated, in addition to physicochemical and calculated properties (solubility, CHI LogD,35 tPSA,
CLogP) to determine drugꢀlike properties.
The representative common chemotype 11c displayed excellent SIRT3 potency with an IC50 of 4 nM,
confirming that the offꢀDNA chemotype was a functional inhibitor for SIRT3 and not merely a ligand
with high affinity. It was also observed to have analogous potency against SIRT1 and SIRT2.
Collectively, all the compounds in the 2 × 2 library (11a-d) were very potent pan inhibitors of SIRT1,
SIRT2 and SIRT3. Replacement of the piperidine of 11c with a piperazine (11d) only slightly reduced
the potency against SIRT2 (≤ 2ꢀfold) while reducing inhibition of SIRT1 and SIRT3 7ꢀ8 fold. The
corresponding phenyl based analogs (11a, 11b) showed a similar trend. Comparing the effect of
replacing the thiophene of 11c-d with a phenyl (11a-b) revealed a general decrease in potency for
SIRT1 and SIRT2 (2 fold). Analysis of the physiochemical properties of 11a-d revealed that the more
basic piperazines imparted a lowering of CLogP, LogD and a concurrent 10ꢀfold improvement in
aqueous solubility. The change from thiophene to benzene had little effect on physiochemical
properties.
The offꢀDNA compounds (11a-d) represented a novel and highly potent new class of SIRT1/2/3 pan
inhibitors. While the goal of identifying novel sirtuin inhibitor scaffolds was achieved, inhibitors 11a-d
tended to have suboptimal drugꢀlike properties (MW, PSA, solubility and aromatic ring count) which
limited their progression. To explore the possibility of improving the physiochemical properties of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 45
potent analog 11c, the thieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide core was kept stationary and 11c was
truncated from the DNA tag end. The systematically truncated analogs were prepared (Supplemental
Information) and evaluated in the SIRT1, SIRT2 and SIRT3 biochemical inhibition assays (Table 2).
Changing the ethylamide substituent on the thiophene to either a tertꢀbutyl ester 17, carboxylic acid
18 or hydrogen 19 resulted in only modest (ca. 2ꢀfold) reductions of SIRT1, SIRT2 or SIRT3 potency,
suggesting that the terminal ethylamide is not essential for activity. Encouraged by these results, we
next evaluated the impact of removing the thiophene ring. Replacing the thiophene in 19, with a tertꢀ
butyl carbamate (15a) resulted in a 3ꢀ4 fold reduction of SIRT1 and SIRT3 activity, while SIRT2
inhibition remained unchanged. When the thiophene in 19 was replaced with an acetamide (20), we
observed more pronounced reductions in SIRT1 (18ꢀfold), SIRT2 (3ꢀfold) and SIRT3 (7ꢀfold) activity.
This trend continued with the removal of the acetamide in the amine analog 16a, which was 5ꢀ to 15ꢀ
fold less potent than 20. Replacing the aminoethylpiperidine with a piperidine (21) dramatically reduced
the sirtuin inhibitory activity to micromolar levels, and SIRT1/2/3 inhibitory activity was abolished in
the severely truncated ethylamine 22 or chloride 14. These results indicated that the thieno[3,2ꢀ
d]pyrimidineꢀ6ꢀcarboxamide core alone was not sufficient to inhibit SIRT1/2/3.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Of the truncated analogs presented, compound 20 exhibited the best balance of SIRT1/2/3 inhibitory
potency (23 ꢀ 110 nM) and low molecular weight (MW = 347). To further optimize the potency, a larger
compound set (Table 3) was prepared where variation of three areas was explored. 1) The effect of
changing the piperidine to the moreꢀpolar piperazine (X = CH or N) ring system was evaluated. 2) The
linker length (n) was varied to determine the optimal distance between the functional group and the
thienopyrimidine core. 3) And the acetamide functionality on 20 was replaced with thioacetylꢀ,
pivaloylꢀ, sulfonamide or pyrrolidine groups.
ACS Paragon Plus Environment

Page 9 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In general, the ethyl piperidines (X = CH, n = 2) were more potent inhibitors than the ethyl
piperazines (X = N, n = 2), (compare 20 vs. 24, 25 vs. 26, 28 vs. 30, 31 vs. 32). However, for the
pyrrolidine variants (34 and 35) the piperazine (X = N) analog displayed similar or greater potency on
SIRT1, SIRT2 or SIRT3 versus the corresponding piperidine (X = CH) versions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Linker length changes were evaluated on the acetamides (20 and 23), pivaloylamides (27 - 29), and
the pyrrolidines (33 and 34). Taking the case of the pivaloylamides, the ethyl linker (28, n = 2) is
slightly more potent than the longer propyl linker (29, n = 3), and the shorter methylene linker (27, n =
1) led to a dramatic decrease in SIRT1, SIRT2 and SIRT3 potency. This dramatic decrease in potency
with a shorter (n = 1) linker length was broadly observed in other analogs (23 and 33).
Replacing the acetamide (20 and 24) with a thioamide (25 and 26) was well tolerated, resulting in
modest improvements in potency. Changing the acetyl groups to the more lipophilic pivaloylamide or
the polar sulfonamides was advantageous for sirtuin inhibition. Interestingly, the sulfonamide 31 which
contains the optimal structural elements for inhibition (X = CH, n = 2) is a single digit nanomolar
inhibitor of SIRT1, SIRT2 and SIRT3 and represents one of the most potent panꢀinhibitors in the study.
When evaluating the pyrrolidine analogs (33 - 35), the selectivity profile appears to slightly favor SIRT3
inhibition over SIRT1 and SIRT2. In general, there appears to be broad functional group tolerance in
this region of the inhibitors (e.g. lipophilic, polar, basic, Hꢀbond donor or acceptor groups are tolerated).
The SAR of the heteroaromatic thieno[3,2ꢀd]pyrimidine core was also evaluated. Utilizing the potent
pan inhibitor 28 as a comparator, a small series of heteroaromatic carboxamide cores were tested for
their ability to inhibit SIRT1/2/3 (Table 4). Replacing the core of 28 (X = S) with furo[3,2ꢀ
d]pyrimidineꢀ6ꢀcarboxamide (36, X = O) resulted in 15ꢀ to 40ꢀfold reductions in SIRT1/2/3 potency. To
evaluate the pyrimidine portion of the heteroaromatic core, two thienopyridine carboxamide scaffolds
were prepared. The first thienopyridine analog, where N3 (adjacent to the piperidine) was replaced with
a CH (37, Y = CH) had modest reductions in SIRT1/2/3 inhibition (3ꢀ to 4ꢀfold), whereas replacing N1
with a CH (38, Z = CH) resulted in more significant (30ꢀ to 63ꢀfold) reductions in potency. To assess
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 45
the sensitivity of the unsubstituted position of the thiophene ring, a methyl was added at the 7ꢀposition
(39) which reduced activity 4ꢀ12 fold across all three enzymes.
1
2
3
4
5
6
7
8
Lastly, the SAR of the carboxamide was also evaluated by making small adjustments at the 6ꢀposition
of the thieno[3,2ꢀd]pyrimidine core of 28 (Table 5). The Nꢀmethylcarboxamide (41) displayed a
dramatic loss in SIRT2 activity, and no measurable SIRT1 or SIRT3 activity. Similarly, changing the
carboxamide to a carboxylic acid (40) or complete removing of the carboxamide (42) resulted in no
abolishing the SIRT1, SIRT2 and SIRT3 activity. These results indicate that the carboxamide is
important for maintaining SIRT1/2/3 inhibition and it is likely involved in critical contacts with the
protein. This strong connection between structural modification of the carboxamide group and SIRT
inhibition is similar to the SAR that is observed with the carboxamide group of 4.17
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SIRT3 X-ray Structural Studies
There have been several reported crystal structures for the sirtuins33, 36ꢀ39 including SIRT1, SIRT2,
SIRT3 and SIRT5. The sirtuins have variable Nꢀ and Cꢀterminal regions, and a conserved catalytic core
containing two lobes; a large Rossmann lobe, and a smaller lobe with a structural zinc binding motif.
Acetylated substrates bind in the cleft formed at the interface of the two lobes with the acetylated lysine
projecting toward the nicotinamide riboside portion of NAD+. A flexible loop, on the smaller lobe,
closes down during the course of the deacetylation reaction to protect the imidate intermediate from
solvent exposure.
Previously, we have reported the crystal structures of human SIRT3 (PDB code: 3GLS), substrate
bound AcꢀACS2/SIRT3 (PDB code: 3GLR), the imidate reaction intermediate mimetic SIRT3ꢀACS2ꢀ
KsꢀacꢀADPR (PDB code: 3GLT) and the ternary carbaNAD/ACS2/SIRT3 (PDB code: 4FVT). In this
study, the identified ELT hit (11c) and two of the most potent truncated pan SIRT1/2/3 inhibitors (28
and 31) were evaluated by Xꢀray crystallography in complex with human SIRT3 (Figure 4, S1ꢀ3).
Crystals of SIRT3 and the inhibitors were obtained by coꢀcrystallization with SIRT3(118ꢀ399). Crystals
of SIRT3/11c (PDB code: 4JSR) and SIRT3/31 (PDB code: 4JT9) diffracted to 1.70 and 2.25 Å,
ACS Paragon Plus Environment

Page 11 of 45
Journal of Medicinal Chemistry
respectively. However, for SIRT3/28 the diffraction quality of the crystals was poor. Therefore 28 was
soaked and exchanged into SIRT3/31 crystals to achieve a suitable diffraction of 2.26 Å for SIRT3/28
(PDB code: 4JT8).
1
2
3
4
5
6
7
8
9
DISCUSSION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Globally, the overall fold of the binary SIRT3/inhibitor structures is very similar to the previously
reported structures of human SIRT3 (PDB codes: 3GLR, 3GLS, 3GLT, 3GLU, 4FVT). The Rossmann
fold is highly superimposable, and there is domain closure upon binding to either substrates, cofactor
intermediates33, 37 or activeꢀsite targeting inhibitors (11c, 28, 31; this study). The largest divergence
among the SIRT3 structures occurs at the flexible loop Ile154ꢀTyr175, which closes down on the
nicotinamide Cꢀpocket (Figure 4a).
The three SIRT1/2/3 pan inhibitors (11c, 28 and 31) bind identically in the catalytic active site (RMS
= 0.29 Å, Figure S2), occupying the nicotinamide Cꢀpocket (Figures 4c, S3) and acetyl lysine substrate
channel (Figures 4d, S3). Closer evaluation of the binding interactions of 11c (Figures 4bꢀd) reveals that
the carboxamide makes four hydrogen bonds with the protein surface, similar to the analogous
nicotinamide portion of carbaꢀNAD+.37 The 6ꢀcarboxamide carbonyl of 11c accepts hydrogen bonds
from protein backbone NH’s of Ile230 and Asp231. One 6ꢀcarboxamide NH donates into the carboxylic
acid on Asp231, while the other carboxamide NH hydrogen bonds to structural bridging water (W31),
which is held in place by Ala146 and Ile154.
The nicotinamide of the SIRT3/ACS2/carbaꢀNAD+ complex makes very similar hydrogen bonding
contacts to Ile154, Ala146, Ile230 and the neighboring structural water (Figures 5aꢀb). This recognition
motif for the nicotinamide of NAD+ is mimicked very well by the sirtuin inhibitors (11c, 28 and 31) and
explains the observed SAR in Table 5. As a result, a substantial reduction of sirtuin inhibitory activity is
observed for compounds that lack the ability to make these critical hydrogen bonds in the nicotinamide
Cꢀpocket (40, R = CO2H; 41, R = CONHMe and 42, R = H).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 45
Moving toward the substrate channel, the thieno[3,2ꢀd]pyrimidine aromatic core lines the top portion
of the receptor pocket, along the hydrophobic zinc binding lobe. The thieno[3,2ꢀd]pyrimidine πꢀstacks
with the phenyl ring of Phe157 (Figures 4cꢀd) and the pyrimidine nitrogen (N1) is orientated to
hydrogen bond with the Phe157 backbone resulting in an improvement in inhibitory activity (compare
28 vs. 38). The other pyrimidine nitrogen (N3) is solvent exposed, allowing for hydrogen bonding with
water, resulting in an interaction that produces only a small improvement in inhibitory activity (compare
28 vs. 37). The slight loss of activity for 37 (compared to 28) could also be explained by the presence of
the CH twisting the neighboring piperidine slightly out of plane of the core.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The ethyl piperidine of 11c adopts an extended conformation sitting along the top of a hydrophobic
cleft under the small structural domain (defined by Tyr165, Phe180, Ile230, Ile291 and Phe294) while
the arylamide is directed toward the Nꢀacetyllysine substrate channel. The hydrophobic nature of this
shelf explains why the lipophilic piperidines (11a, 11c, 20, 28 and 31) are more potent sirtuin inhibitors
than the polar piperazine analogs (11b, 11d, 24, 30 and 32).
Further along the substrate channel, the aryl amide NH hydrogen bonds with Val292. In other
substrate bound structures (3GLR and 4FVT), Val292 forms a hydrogen bond with the Nꢀεꢀacetyl lysine
from the substrate (Figures 5aꢀb). For the SIRT1/2/3 inhibitors, a modest improvement in inhibition is
observed with increasing acidity of the NH donor that interacts with Val292. For instance, comparing
the SIRT1/2/3 inhibitory activity of the sulfonamide 25, with acetamide 20 reveals an 8ꢀ to 28ꢀfold
improvement in potency. However, the lack of an available NH donor to interact with Val292, as
exemplified by the pyrrolidine 34, resulted in only modest changes in sirtuin inhibitory activity.
The Xꢀray structures also provided an explanation for the SAR observed for changes in the linker
length (n). The ethyl linker (n = 2) for the piperidine (20 & 28) optimally align the amide NH to
hydrogen bond with Val292, whereas the methylene analogs (n = 1, 23 & 27) are too short to optimally
make this interaction, resulting in a reduction of potency. The longer propylpiperidine (n = 3, 29) is
envisioned in twisting to accomidate this interaction, but at a slight loss of activity (1.5 ꢀ 3 fold).
ACS Paragon Plus Environment

Page 13 of 45
Journal of Medicinal Chemistry
Lastly, the distal ethylamide substituted on the 2ꢀthiophene on 11c makes a hydrogen bond with
Glu296. This ethylamide, which was the point of attachment to the DNA linker in the ELT screen,
extends out of the protein surface. Removing this portion had little effect on sirtuin inhibitory activity
(see 19 vs. 11c) given the sum of the other interactions the small molecule inhibitors make with in
catalytic site.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The larger portion of the catalytic site, which is the space usually occupied by the ribofuranose to the
adenine site of NAD+, is still largely unoccupied in the SIRT3/11c, SIRT3/28 and SIRT3/31 structures,
except for bulk water or crystallization medium. This space may be more efficiently exploited in future
designs.
The residues that form the NAD+ binding pocket are highly conserved between SIRT1/2/3, which
likely explains why these compounds are pan inhibitors. Interestingly, several small molecule sirtuin
inhibitors that have been described in the literature possess carboxamides, e.g. nicotinamide, 4 and
benzamides (7) and structural modification of this group is strongly linked to changes in SIRT
inhibition.
Recently, the ternary complex of the catalytic core of SIRT1(241ꢀ516), NAD+ and a structural analog
of EX52717 (43, Figure 5c) was solved by Xꢀray crystallography40. Analysis of the ternary complex
revealed that the carboxamide of 43 occupies the nicotinamide Cꢀpocket with the indole projecting
upward into a hydrophobic pocket between the flexible loop and a helix on the small lobe. Meanwhile,
the nicotinamide of the cofactor swings out in an alternate inactive confirmation projecting toward the
substrate channel.
Structural comparisons of the crystal structures of SIRT1/43/NAD+ and SIRT3/11c (Figures 5cꢀd)
reveal several similarities and differences in how these two classes of inhibitors bind. Globally, the
Rossmann fold, NAD+ cofactor binding channel and nicotinamide binding pockets are largely
undisturbed. However, there is a shift of the SIRT1 small lobe onto the active site in the presence of 43.
Looking closer, the nicotinamide binding pockets, where the carboxamides of inhibitors 11c and 43
bind, have identically conserved hydrogen bonding networks, explaining the sensitivity of the SAR of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 45
the carboxamides in the EX527 analogs (4 and 43) and the inhibitors in this study. Interestingly, the
orientation of the indole scaffold in 43 is 74 degrees normal to the thienopyrimidine scaffold of 11c
projecting into a narrow hydrophobic pocket, which could be an area of future exploitation to find
additional classes of selective SIRT1 inhibitors.
1
2
3
4
5
6
7
8
9
Several regions on the small domain show differences in the two complexes. At the substrate channel
entrance, the loop (Ile291ꢀLeu298) in SIRT3/11c is more open to accommodate the distal thiophene
carboxamide, whereas the identical loop (Ile411ꢀLeu418) in SIRT1/43/NAD+ closes downward to allow
Phe414 to piꢀstack with the backside of the nicotinamide of NAD+. The analogous residue (Phe294) in
SIRT3/11c is situated just above the Nꢀεꢀacetyl lysine binding site.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
At the flexible loop region (Asp156ꢀPro174) of the SIRT3/11c complex, Phe157 piꢀstacks with the
thienopyrimdine core of 11c, whereas in the SIRT1/43/NAD+ complex, the same sequence (Asp272ꢀ
Pro291) forms a helix and the analogous Phe273 side chain closes in to cap the hydrophobic cage for
43. Additional shifts are observed for other hydrophobic residues on the small lobe of SIRT1/NAD+/43
complex (Tyr280, Phe297 and Phe413) which collapse to form the hydrophobic cage, whereas in
SIRT3/11c the analogous residues (Tyr165, Phe180 and Phe293) form the upper hydrophobic shelf to
accommodate the binding of 11c.
The observed SAR is in strong agreement with the observed ligand interactions found in the
SIRT3/11c, SIRT3/28 and SIRT3/31 crystal structures. Compounds 11c and 28 were generally selective
(XC50 > 10 ꢁM) when profiled against a broad panel of kinases, nuclear receptors, ion channels,
transporters and GPCRs. They do not inhibit hERG (11c and 28; >50 ꢁM, dofetilide binding) and are
poor CYP inhibitors (28; 1A2, 2C19, 2D6 and 3A4 > 50 ꢁM, 2C9 = 7.2 ꢁM). Furthermore, compound
28 has a low LogD (2.73), high solubility (297 ꢁM) and was stable in microsomal incubations (Human
CLint = 15.8 ꢁL/min/mg, Mouse CLint = 12.7 ꢁL/min/mg).
ACS Paragon Plus Environment

Page 15 of 45
Journal of Medicinal Chemistry
CONCLUSION
1
2
3
4
5
6
7
8
9
A novel class of potent SIRT1/2/3 pan inhibitors was identified by utilizing encoded library
technology to enrich for molecules that interact with SIRT3. Based on the analysis of the ELT
sequencing data, the selected cycle 3 building block (thieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide) was a
preferred core scaffold which was critical for the chemotype inhibitory function.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The initially prepared offꢀDNA library confirmed that compounds 11a-d were very potent pan
SIRT1/2/3 inhibitors. Reduction of molecular weight, to improve physiochemical properties of 11c,
resulted in the identification of the acetamide 20 as a good compromise of potency and reduced
molecular weight. Further SAR showed that the thioacetyl 25, tertꢀbutylꢀamide 28 and sulfonamide 31
were particularly potent pan inhibitors.
To better understand binding interactions, three pan inhibitors (11c, 28 and 31) were crystallized with
SIRT3. The thieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide inhibitors bind in the active site cleft, in between
the large Rossmann fold and the small zinc binding domain occupying the nicotinamide Cꢀpocket in
addition to the substrate channel. Comparisons with previously described SIRT3 structures revealed
similar protein folding, except for the flexible loop region where Phe157 makes a πꢀstacking interaction
with the thienopyrimidine core. The carboxamide on the inhibitors, which is essential for activity,
makes key hydrogen bonding interactions with residues in the nicotinamide binding pocket, similar to
carbaꢀNAD+ and the recently described ternary complex of SIRT1/NAD+/43.
The pan inhibitor 11c and the truncated analogs 28 and 31, represent a significant advance over
currently available sirtuin inhibitors. Their binding mode, corroborated by Xꢀray crystallographic data,
is in good agreement with the observed SAR. The potency of this class of inhibitors makes them
valuable tools for understanding the biological effects of modulating the deacetylase activity of SIRT1,
SIRT2 and SIRT3.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 45
EXPERIMENTAL METHODS
Materials
1
2
3
4
5
6
7
8
Thioacetylated peptide (ACS2ꢀKsꢀAc: TRSGKthioacetylVMRRLLR) which has the same sequence as the
ACS2Kac peptide with the acetyl lysine replaced with thioacetyl lysine using FmocꢀLys(CSCH3)ꢀOH
was prepared as previously described.33
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ELT Affinity Selection
Selections were carried out by capturing 2 ꢂg of FlagꢀhSIRT3(118ꢀ399)ꢀSBP on streptavidin matrix
tips (Phynexus) in the presence of 1) no βꢀNAD/no peptide substrate, 2) 100 ꢂM βꢀNAD (Sigma), or 3)
20 ꢂM TRSGKthioacetylVMRRLLR for three rounds. A no target control selection with buffer was carried
out concurrently in the absence of SIRT3 protein.
Streptavidin tips were preꢀwashed in selection buffer: 50 mM Tris (pH 7.5), 150 mM NaCl, 0.1%
Tweenꢀ20, and 0.1 mg/mL sheared salmon sperm DNA (sssDNA, Ambion), 0.1 mg/mL BSA (Ambion)
and 5 mM βꢀmercaptoethanol (BME). In the first round of selection, 2 ꢂg of FlagꢀhSIRT3(118ꢀ399)ꢀ
SBP protein was immobilized on preꢀwashed tips in the presence of 1) no βꢀNAD/no peptide substrate,
2) 100 ꢂM βꢀNAD or 3) 20 ꢂM thioacetylated peptide. The tips were washed two times with buffer
containing the corresponding cofactor and substrate when necessary. Pooled ELT libraries (5 nmoles)
were passed over the immobilized SIRT3 in the presence of the corresponding cofactor and substrate for
1 hour at room temperature. The tips were washed 8 times with selection buffer containing the
corresponding cofactor and substrate and two times with BSA free selection buffer containing the
corresponding cofactor and substrate. Bound molecules were heat eluted by passing BSA free selection
buffer containing no cofactors and substrates over the tip at 72 °C for 10 minutes. The cooled heat
elution was postꢀcleared twice by passing the elution over streptavidin tips for 15 min to remove any
denatured SIRT3 and matrix binders. Fresh BSA and sssDNA were added to all samples and
corresponding cofactor and substrate was added to the elutions as needed. Round 2 was performed as
described for Round 1 using freshly immobilized SIRT3 on streptavidin tips in the presence of
ACS Paragon Plus Environment

Page 17 of 45
Journal of Medicinal Chemistry
corresponding cofactor and substrate and postꢀcleared Round 1 output. Round 3 was performed as
described for Round 1 using freshly immobilized SIRT3 on streptavidin tips in the presence of
corresponding cofactor and substrate and postꢀcleared Round 2 output with the exceptions that the last
two washes and elution were with BSAꢀfree and sssDNAꢀfree selection buffer and the round 3 output
was not postꢀcleared. Quantitative PCR was used to quantitate the outputs from each round of selection.
The round 3 output was sequenced using an Illumina sequencing platform.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Protein Cloning, Expression, and Purification
Human SIRT3ꢀ(118–399) was cloned into a proprietary modified pET21b vector (Viva Biotech). The
protein was expressed in E. coli BL21ꢀGold(DE3) cells (Stratagene) as an Nꢀterminal fusion to a
hexahistidine affinity tag with an integrated TEV protease site. A single colony was inoculated in LB
media containing 100 ꢀg/ml ampicillin at 37 °C, swirled at 250 rpm until the A600 reached 0.3. The
culture was then cooled to 18 °C, swirled at 250 rpm until the A600 reached 0.6–0.8. 1ꢀ(2ꢀisopropylthio)ꢀ
β-Dꢀgalactopyranoside (IPGT) was added to a final concentration of 0.3 mM, and expression was
continued at 18 °C, swirled at 160 rpm overnight. Cells were collected by centrifugation, and the pellet
was reꢀsuspended in lysis buffer (200 mM NaCl, 5% glycerol, 5 mM 2ꢀmercaptoethanol, and 25 mM
HEPESꢀNaOH, pH 7.5) and sonicated to break the cells. The supernatant was separated from the cell
debris by centrifugation at 10,000 × g for 40 min at 4 °C and loaded onto a NiꢀNTA column (Qiagen)
that was equilibrated with a buffer containing 200 mM NaCl, 5% glycerol, 5 mM 2ꢀmercaptoethanol, 20
mM imidazole, and 25 mM HEPESꢀNaOH, pH 7.5. The column was washed with 5 column volumes of
a buffer containing 200 mM NaCl, 5% glycerol, 5 mM 2ꢀmercaptoethanol, 50 mM imidazole, and 25
mM HEPESꢀNaOH, pH 7.5, then eluted with a buffer containing 200 mM NaCl, 5% glycerol, 5 mM 2ꢀ
mercaptoethanol, 250 mM imidazole, and 25 mM HEPESꢀNaOH, pH 7.5. The eluted protein was
dialyzed in lysis buffer and digested with TEV protease (Invitrogen) at 4 °C overnight to remove the Nꢀ
terminal His tag. The protein was loaded on a second NiꢀNTA column equilibrated with lysis buffer.
The untagged protein was eluted with a buffer containing 200 mM NaCl, 5% glycerol, 5 mM 2ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 45
mercaptoethanol, 5 mM imidazole, and 25 mM HEPESꢀNaOH, pH 7.5. The purified protein was
dialyzed against a buffer containing 200 mM NaCl, 5 mM 2ꢀmercaptoethanol, and 20 mM TrisꢀHCl, pH
8.0, and concentrated. The protein was further purified by elution with dialyzing buffer over a S200
column (GE Healthcare) to 95% purity as assessed by SDSꢀPAGE analysis stained by Coomassie
Brilliant Blue Rꢀ250, and concentrated to 10–15 mg/ml in the dialyzing buffer.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SIRT1, SIRT2 and SIRT3 Biochemical assays
Deacetylation of a Trp 5ꢀmer peptide (AcꢀRHKKAcWꢀNH, Biopeptide, San Diego, CA) by Hisꢀ
SIRT1(1ꢀ747), HisꢀSIRT2(1ꢀ389) and HisꢀSIRT3(102ꢀ399) was measured by a discontinuous OAADPr
Mass Spec assay which measures OAADPr (2'ꢀOꢀacetylꢀADPꢀribose) production. All assays were
performed at room temperature in reaction buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM DTT,
0.05% BSA). Test compounds (1 ꢁL in DMSO) were preꢀincubated with either SIRT1 (5 nM), SIRT2
(10 nM) or SIRT3 (5 nM) in reaction buffer (50 ꢁL) for 20 minutes. For IC50 determination, Trp 5ꢀmer
peptide was added at KM conditions (2 ꢁM for SIRT1, 10 ꢁM for SIRT2 or 2.2 ꢁM for SIRT3) along
with NAD at KM (80 ꢁM for SIRT1, 50 ꢁM for SIRT2 and 130 ꢁM SIRT3) for a final volume of 100
ꢁL. The reaction was quenched after 30 minutes with 10 ꢁL of stop buffer (50 mM Nicotinamide in
10% formic acid) to give a final concentration of 0.9% formic acid and 4.5 mM nicotinamide. To
prepare the assays for analysis, 20 ꢁL of reaction volume was mixed in 80 ꢁL of 50:50
acetonitrile/methanol mixture. The plates were analyzed on an Agilent RapidFire 200 HighꢀThroughput
Mass Spectrometry System (Agilent, Wakefield) coupled to an AB Sciex API 4000 mass spectrometer
fitted with an electrospray ionization source in negative MRM mode monitoring the transition
600.1/345.9 for the parent/daughter ion under low resolution conditions. Peak data was integrated using
RapidFire Integrator software (Agilent, Santa Clara, CA).
Crystallization and Structural Determination of SIRT3 Complex
ACS Paragon Plus Environment

Page 19 of 45
Journal of Medicinal Chemistry
SIRT3/31 and SIRT3/11c crystals were obtained by hanging drop vapor diffusion method at 18 °C.
The drop was composed of 1 ꢁl of protein/compound mixture and 1 ꢁl crystallization buffer. For
SIRT3/31 the crystallization condition was 0.1 M HEPES pH 7.5 20% w/v PEG 8000. The
crystallization buffer for SIRT3/11c was 0.1M Tris pH 8.0, 20% PEG 4000 or 20% PEG 6000. The
SIRT3/31 crystals were soaked for 4 h in 10 mM 28 in crystallization buffer to obtain SIRT3/28
crystals.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The SIRT3/31 and SIRT3/11c crystals were cryoꢀprotected in mother liquor, which contained 20%
glycerol before being flashꢀfrozen in liquid nitrogen. The SIRT3/31 crystals soaked in 28 were cryoꢀ
protected in mother liquor, which contained 20% glycerol and 10 mM 28. The diffraction data was
collected at Shanghai Synchrotron Radiation Facility (SSRF) beamline workstations BL17U1 and APS
21ꢀIDꢀD and processed using the Xia241 or HKL200042 programs. The structures were solved by
molecular replacement with Phaser43 using the substrate bound ACS2/SIRT3 structure (PDB code:
3GLR) as a search model. Iterative structure refinement and model building were performed between
Refmac44 in the CCP4 suite45 and Coot46. Detailed information regarding the diffraction data,
refinement, and structure statistics is listed in the Supporting Information. All the parameters for each
diffraction data set were reprocessed using Mosflm47 and Scala48 and the refinement statistics were
obtained from Refmac44 in the CCP4 suite45.
Synthetic Methods
Unless otherwise noted, all reagents and solvents were obtained from commercial suppliers and used
without further purification. Purities of all assayed compounds were determined by analytical HPLC
using the area percentage method on the UV trace recorded at a wavelength of 254 nm, and were found
to have ≥ 95% purity unless otherwise specified. Analytical HPLC was performed on an Agilent 1100
Series HPLC equipped with a 3.5 ꢁm Eclipse XDBꢀC18 (4.6 mm × 100 mm) column with the following
conditions: CH3CN/H2O, modified with 0.1% formic acid mobile phase. Gradient elution: 5% CH3CN
hold (2 min), 5% to 95% CH3CN gradient (11 min), 95% to 5% CH3CN gradient (0.3 min), 5% CH3CN
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 45
hold (2.7 min), 15 min. total run time with a flow rate of 0.8 ml/min. Lowꢀresolution Mass spectroscopy
(LRMS) was performed on an Agilent 1100 Series LC/MSD single quadrupole in electrospray
ionization. High Resolution Mass Spec (HRMS) was completed on a Waters qTOF Premiere Mass
Spectrometer operating in W mode positive ionization with a resolving power of approximately 15,000.
Flow injection was completed using a Waters nanoꢀacquity LC. HRMS acceptable error is 3 mDa or 5
ppm, although most analyses are observed within 0.5 mDa with isotope fits in good agreement with the
proposed structures. Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker AVANCEꢀ
III 300 or 400 MHz instrument referenced with the appropriate internal standard (TMS: δ = 0 ppm).
Flash column chromatography was performed on silica gel, particle size 60 Å, mesh of 230−400
(Merck) under standard techniques. All microwave reactions were conducted on a Biotage Initiator
Microwave Synthesizer.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-Chlorothieno[3,2-d]pyrimidine-6-carboxylic acid (13). To a stirring solution of 2,2,6,6ꢀ
tetramethylpiperidine (1.484 mL, 8.79 mmol) in anhydrous THF (15 mL) at 0 °C under nitrogen was
added 2.5 M BuLi in hexanes (3.52 mL, 8.79 mmol) dropwise. The reaction mixture was stirred at 0 °C
for 30 minutes, then the mixture was added to a solution of 4ꢀchlorothieno[3,2ꢀd]pyrimidine (1.00 g,
5.86 mmol) in anhydrous THF (15 mL) at ꢀ78 °C dropwise over a period of 30 minutes. The reaction
was stirred at ꢀ78 °C for 1 hour, then to the reaction was added dry ice (2.58 g, 58.6 mmol). The
reaction was allowed to warm up to room temperature over a period of 2 hours. The reaction was diluted
with EtOAc (100 mL) and washed with 0.1 M HCl. The organic layer was dried over MgSO4 and
1
evaporated to dryness to obtain 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxylic acid (1.1 g, 83%). H
NMR (DMSOꢀd6) δ 14.5 (br s, 1H), 9.16 (s, 1H), 8.27 (s, 1H). LRMS (ESI) 215.0 [M + H]+.
4-Chlorothieno[3,2-d]pyrimidine-6-carboxamide (14). To a solution of oxalyl chloride (4.17 mL,
47 mmol) in anhydrous dichloromethane (50 mL) at 0 °C under nitrogen was added DMF (0.8 mL). The
solution was stirred at 0 °C for 30 minutes, and then a suspension of 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀ
carboxylic acid (5.04 grams, 23.5 mmol) in dichloromethane (50 mL) was added dropwise over 10
minutes at 0 °C. The reaction mixture was heated to reflux for 3.5 hours and concentrated to dryness.
ACS Paragon Plus Environment

Page 21 of 45
Journal of Medicinal Chemistry
The crude acid chloride was dissolved in dioxane (80 mL) and 182 mL (91 mmol) of a solution of 0.5 M
ammonia in dioxane was added dropwise over 10 min at 0 °C. The reaction mixture was stirred at 0 °C
for 30 min, warmed to room temperature, diluted with water (150 mL) and extracted with CH2Cl2 (3x).
The combined organic layers were washed with water (2x), dilute aq. NaHCO3, brine and concentrated
to dryness. The product was recrystallized from CH3CN to obtain 842 mg of 4ꢀchlorothieno[3,2ꢀ
d]pyrimidineꢀ6ꢀcarboxamide. The mother liquor was concentrated to obtain a second crop (1499 mg)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
and a third crop (259 mg), of 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide (2.6 g, 52%). H NMR
(DMSOꢀd6) δ 9.11 (s, 1H), 8.62 (br s, 1H), 8.35 (s, 1H), 8.15 (br s, 1H). 13C NMR (DMSOꢀd6) δ 161.53,
161.21, 154.75 (CH), 154.13, 150.77, 131.47, 124.80 (CH). LRMS (ESI) 214.0 [M + H]+.
tert-Butyl (2-(1-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperidin-4-yl)ethyl)carbamate (15a).
A solution of 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide (1.30 g, 6.08 mmol), tertꢀbutyl (2ꢀ
(piperidinꢀ4ꢀyl)ethyl)carbamate (1.39 g, 6.08 mmol) and DIEA (1.05 mL, 6.08 mmol) in CH3CN (80
mL) was heated at reflux for 1 hour. The reaction mixture was cooled to room temperature and
concentrated to dryness. The residue was suspended in MeOH (10 mL) then water (90 mL) was added.
The mixture was sonicated and the precipitate was collected by filtration, washed with water and dried
under high vacuum to obtain tertꢀButyl (2ꢀ(1ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperidinꢀ4ꢀ
1
yl)ethyl)carbamate (2.23 g, 90%). H NMR (DMSOꢀd6) δ 8.48 (s, 1H), 8.40 (br s, 1H), 8.03 (s, 1H),
7.87 (br s, 1H), 6.81 (t, 1H), 4.69 (d, 2H), 3.16 (t, 2H), 2.976 (q, 2H), 1.83 (d, 2H), 1.66 (br, 1H), 1.37
(s, 9H), 1.34 (m, 2H), 1.16 (m, 2H). LRMS (ESI) 406.1 [M + H]+. HRMS (m/z): [M + H]+ calc for
C19H28N5O3S, 406.1913; found, 406.1915.
tert-Butyl (2-(4-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperazin-1-yl)ethyl)carbamate (15b).
A solution of 4ꢀchlorothieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide (250 mg, 1.17 mmol), DIEA (245 ꢂL,
1.40 mmol) and tertꢀbutyl (2ꢀ(piperazinꢀ1ꢀyl)ethyl)carbamate (332 mg, 1.40 mmol) in CH3CN (15 mL)
was heated at 60 °C for 18 hours. The reaction mixture was cooled to room temperature and filtered to
collect the product as a solid. The solid was washed with CH3CN (2 x 10 mL) and dried to obtain tertꢀ
butyl (2ꢀ(4ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperazinꢀ1ꢀyl)ethyl)carbamate as a white solid
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 45
(440 mg, 93% yield). 1H NMR (DMSOꢀd6) δ 8.51 (s, 1H), 8.41 (br s, 1H), 8.05 (s, 1H), 7.88 (br s, 1H),
6.71 (t, 1H), 3.92 (m, 4H), 3.08 (q, 2H), 2.55 (m, 4H), 2.38 (t, 2H), 1.38 (s, 9H). LRMS (ESI): 407.2 [M
+ H]+. HRMS (m/z): [M + H]+ calc for C18H27N6O3S, 407.1865; found, 407.1866.
1
2
3
4
5
6
7
8
9
4-(4-(2-Aminoethyl)piperidin-1-yl)thieno[3,2-d]pyrimidine-6-carboxamide (16a). To a solution of
tertꢀButyl (2ꢀ(1ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperidinꢀ4ꢀyl)ethyl)carbamate (15a) (840
mg, 2.07 mmol) in CH2Cl2 (40 mL) was added Trifluoroacetic acid (10 mL). The reaction mixture was
stirred for 72 hours, concentrated to dryness and chased with CH2Cl2 (2x). The residue was diluted with
MeOH, then a solution of 1.25 M HCl in MeOH (2 mL) was added. To the resulting oil was added
diethyl ether (15 mL) and pentane (5 mL). The solution was sonicated to produce solid which was
isolated by decantation, and dried under vacuum to obtain 4ꢀ(4ꢀ(2ꢀaminoethyl)piperidinꢀ1ꢀyl)thieno[3,2ꢀ
d]pyrimidineꢀ6ꢀcarboxamide as the hydrochloride salt (1.09 g) 1H NMR (DMSOꢀd6) δ 8.65 (s, 1H), 8.57
(br s, 1H), 8.08 (s, 1H), 7.987 (br s, 1H), 7.75 (br, 3H), 4.76 (d, 2H), 3.29 (t, 2H), 2.85 (m, 2H), 1.89 (d,
2H), 1.79 (s, 1H), 1.50 (m, 2H), 1.23 (m, 2H). LRMS (ESI): 306.1 [M + H]+. HRMS (m/z): [M + H]+
calc for C14H20N5OS, 306.1389; found, 306.1389.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-(4-(2-Aminoethyl)piperazin-1-yl)thieno[3,2-d]pyrimidine-6-carboxamide (16b). A solution of
tertꢀbutyl (2ꢀ(4ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperazinꢀ1ꢀyl)ethyl)carbamate (15b) (440
mg, 2.46 mmol) was stirred with 25% TFA in CH2Cl2 (8 mL) for 6 hours. The solution was
concentrated to dryness and triturated with a mixture of diethyl ether and pentane mixture to obtain solid
4ꢀ(4ꢀ(2ꢀaminoethyl)piperazinꢀ1ꢀyl)thieno[3,2ꢀd]pyrimidineꢀ6ꢀcarboxamide as a tan solid as the bisꢀTFA
1
salt (844 mg, 64%). H NMR (DMSOꢀd6) δ 8.68 (s, 1H), 8.53 (br s, 1H), 8.13 (s, 1H), 7.97 (br s, 4H),
4.14 (br, 4H), 3.1ꢀ3.4 (m, 8H). LRMS (ESI): 307.2 [M + H]+. HRMS (m/z): [M + H]+ calc for
C13H19N6OS, 307.1341; found, 307.1341.
Methyl
5-(ethylcarbamoyl)thiophene-2-carboxylate.
To
a
solution
of
5ꢀ
(methoxycarbonyl)thiopheneꢀ2ꢀcarboxylic acid (500 mg, 2.68 mmol) and HATU (1.23g, 3.23 mmol) in
DMF (10 mL) was added DIEA (1.16 mL, 6.70 mmol). The reaction mixture was stirred at room
temperature for 10 min, then ethylamine hydrochloride (219 mg 2.68 mmol) was added as a solid. The
ACS Paragon Plus Environment

Page 23 of 45
Journal of Medicinal Chemistry
reaction mixture was stirred for 7 h, and another batch of ethylamine hydrochloride (219 mg 2.68 mmol)
and DIEA (1.16 mL, 6.70 mmol) was added and stirring was continued for a total of 24 h. To the
reaction mixture was added sat. NaHCO3 (30 mL) and water (50mL). The precipitated product was
collect by filtration, and washed with water to obtain methyl 5ꢀ(ethylcarbamoyl)thiopheneꢀ2ꢀ
1
2
3
4
5
6
7
8
9
1
carboxylate as a tan solid (555 mg, 97%). H NMR (DMSOꢀd6) δ 8.76 (t, 1H), 7.79 (d, 1H), 7.75 (d,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H), 3.84 (s, 3H), 3.25 (m,2H), 1.12 (t, 3H). LRMS (ESI): 214.1 [M + H]+.
5-(Ethylcarbamoyl)thiophene-2-carboxylic
acid.
To
a
solution
of
methyl
5ꢀ
(ethylcarbamoyl)thiopheneꢀ2ꢀcarboxylate (555 mg, 2.60 mmol) in THF (5 mL) was added a solution of
LiOH (124 mg, 5.18 mmol) in water (5 mL). The reaction mixture was stirred at room temperature for
72 h, concentrated to dryness, dissolved in water, and acidified with conc. HCl to pH =1. The precipitate
was collected by filtration, washed with water and dried under vacuum to obtain 5ꢀ
1
(ethylcarbamoyl)thiopheneꢀ2ꢀcarboxylic acid as an off white solid. (221 mg, 43%). H NMR (DMSOꢀ
d6) δ 13.40 (s, 1H), 8.71 (t, 1H), 7.71 (d, 1H), 7.68 (d, 1H), 3.26 (m, 2H), 1.12 (t, 3H). LRMS (ESI):
200.1 [M + H]+.
Methyl 3-(ethylcarbamoyl)benzoate. A solution of 3ꢀ(methoxycarbonyl)benzoic acid (1.0 g, 5.55
mmol), HATU (2.53g, 6.65 mmol), DIEA (1.44 mL, 8.31 mmol) in DMF (15 mL) was stirred at room
temperature for 10 min, then ethylamine (70% in water, 8.84 mL) was added and the reaction mixture
was stirred at room temperature for 18 hours. The mixture was diluted with ethyl acetate and the organic
layer was washed with sat. NaHCO3, water (3 x), dried (Na2SO4), concentrated and chased with
methanol to obtain methyl 3ꢀ(ethylcarbamoyl)benzoate as an orange oil (1.20 g) which was used
without further purification in the next step.
3-(Ethylcarbamoyl)benzoic acid. To a solution of methyl 3ꢀ(ethylcarbamoyl)benzoate (1.20 g) in
methanol (50 mL) was added LiOH (666mg, 27.8 mmol) in water (10 mL). The reaction mixture was
stirred at room temperature for 72 h, concentrated to dryness, dissolved in water, and acidified with
conc. HCl to pH =1ꢀ2. The precipitate was collected by filtration, washed with water and dried to obtain
3ꢀ(ethylcarbamoyl)benzoic acid (793 mg, 74% after 2ꢀsteps). 1H NMR (DMSOꢀd6) δ 13.17 (s, 1H), 8.67
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 45
(t, 1H), 8.42 (m, 1H), 8.07 (m, 1H), 7.59 (t, 1H), 3.30 (m, 2H), 1.13 (t, 3H). LRMS (ESI): 194.1 [M +
H]+.
1
2
3
4
5
6
7
8
General procedure for the synthesis of 11a-11d. Step 1) To a solution of appropriate Boc protected
amine (0.166 mmol) in dichloromethane (6 mL) was added TFA (0.309 mL, 4.0 mmol). The reaction
mixture was stirred for 1 h. The excess TFA and solvents were evaporated, then the residue was
azeotroped with DCM (2 x). The resulting amine was carried to the next step as is. Step 2) To a mixture
of desired carboxylic acid (0.200 mmol) and HATU (76 mg, 0.200 mmol) in DMF (3 mL) was added
DIEA (0.175 mL, 1.0 mmol). The reaction mixture was stirred for 10 min, then the appropriate amine
from step 1 was added. The reaction mixtures were allowed to stir at room temperature overnight, and
the products were purified by prep HPLC.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N1-(2-(1-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperidin-4-yl)ethyl)-N3-ethylisophthalamide
(11a). The title compound was synthesized according to the general procedure using tertꢀbutyl (2ꢀ(1ꢀ(6ꢀ
carbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperidinꢀ4ꢀyl)ethyl)carbamate (67.2 mg, 0.166 mmol) and 3ꢀ
(ethylcarbamoyl)benzoic acid (40 mg, 0.207 mmol) to afford 11a (90 mg, 77%). 1H NMR (DMSOꢀd6) δ
8.57 (m, 2H), 8.48 (s, 1H), 8.40 (br, 1H), 8.29 (m, 1H), 8.04 (s, 1H), 7.95 (m, 2H), 7.86 (br, 1H), 7.54 (t,
1H), 4.71 (m, 2H), 3.25ꢀ3.40 (m, 5H) 3.1ꢀ3.25 (m, 2H), 1.91 (m, 2H), 1.75 (m, 1H), 1.51 (q, 2H), 1.2ꢀ
1.3 (m, 2H), 1.13 (t, 3H). LRMS (ESI): 481.35 [M + H]+. HRMS (m/z): [M + H]+ calc for C24H29N6O3S,
481.2022; found, 481.2021.
N1-(2-(4-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperazin-1-yl)ethyl)-N3-ethylisophthalamide
(11b). The title compound was synthesized according to the general procedure using tertꢀbutyl (2ꢀ(4ꢀ(6ꢀ
carbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperazinꢀ1ꢀyl)ethyl)carbamate (67.3 mg, 0.161 mmol) and 3ꢀ
(ethylcarbamoyl)benzoic acid (40 mg, 0.207 mmol) to afford 11b (95 mg, 86%). 1H NMR (DMSOꢀd6) δ
8.57 (m, 2H), 8.52 (s, 1H), 8.41 (br, 1H), 8.29 (m, 1H), 8.05 (s, 1H), 7.95 (dd, 2H), 7.87 (br, 1H), 7.55
(t, 1H), 3.95 (m, 4H), 3.45 (q, 2H), 3.30 (m, 2H), 2.63 (m, 4H), 2.54 (m, 2H), 1.13 (t, 3H). LRMS (ESI):
482.30 [M + H]+. HRMS (m/z): [M + H]+ calc for C23H28N7O3S, 482.1974; found, 482.1974.
ACS Paragon Plus Environment

Page 25 of 45
Journal of Medicinal Chemistry
N2-(2-(1-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperidin-4-yl)ethyl)-N5-ethylthiophene-2,5-
dicarboxamide (11c). The title compound was synthesized according to the general procedure using
tertꢀbutyl (2ꢀ(1ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperidinꢀ4ꢀyl)ethyl)carbamate (65.1 mg,
0.161 mmol) and 5ꢀ[(ethylamino)carbonyl]ꢀ2ꢀthiophenecarboxylic acid (40 mg, 0.201 mmol) to afford
1
2
3
4
5
6
7
8
9
1
11c (70 mg, 68%). H NMR (DMSOꢀd6) δ 8.67 (s, 1H), 8.56ꢀ8.62 (m, 3H), 8.06 (s, 1H), 8.00 (br, 1H),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7.68 (m, 2H), 4.76 (m, 2H), 3.21ꢀ3.37 (m, 6H), 1.95 (m, 2H), 1.76 (m, 1H), 1.49 (m, 2H), 1.25 (m, 2H),
1.11 (t, 3H). LRMS (ESI): 487.26 [M + H]+. HRMS (m/z): [M + H]+ calc for C22H27N6O3S2, 487.1586;
found, 482.1586.
N2-(2-(4-(6-carbamoylthieno[3,2-d]pyrimidin-4-yl)piperazin-1-yl)ethyl)-N5-ethylthiophene-2,5-
dicarboxamide (11d). The title compound was synthesized according to the general procedure using
tertꢀbutyl (2ꢀ(4ꢀ(6ꢀcarbamoylthieno[3,2ꢀd]pyrimidinꢀ4ꢀyl)piperazinꢀ1ꢀyl)ethyl)carbamate (65.3 mg,
0.161 mmol) and 5ꢀ[(ethylamino)carbonyl]ꢀ2ꢀthiophenecarboxylic acid (40 mg, 0.201 mmol) to afford
11d (90 mg, 94%). 1H NMR (DMSOꢀd6) δ 8.60 (m, 2H), 8.51 (s, 1H), 8.41 (br, 1H), 8.05 (s, 1H), 7.87
(br, 1H), 7.68 (s, 2H), 3.94 (m, 4H), 3.40 (q, 2H), 3.25 (m, 2H), 2.61 (m, 4H), 2.53 (m, 2H), 1.11 (t,
3H). LRMS (ESI): 488.15 [M + H]+. HRMS (m/z): [M + H]+ calc for C21H26N7O3S2, 488.1539; found,
488.1540.
ANCILLARY INFORMATION
Supporting Information
Xꢀray diffraction and refinement parameters, view of the superposition of SIRT3/11c, SIRT3/28 and
SIRT3/31, binding site views of SIRT3/28 and SIRT3/31, a view of the orientation of Phe157 in SIRT3
structures, and the synthetic details for compounds 17 – 42. This material is available free of charge via
the Internet at http://pubs.acs.org.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 45
Accession Codes
1
2
3
4
5
6
7
8
9
The atomic coordinates and experimental data for SIRT3/11c (PDB code: 4JSR), SIRT3/28 (PDB
code: 4JT8), and SIRT3/31 (PDB code: 4JT9) have been deposited into the RCSB Protein Data Bank
(www.rcsb.org).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Corresponding Authors
For J.S.D.: phone, (508)ꢀ934ꢀ6975; eꢀmail: jdisch92@gmail.com. For G.E.: phone, (781)ꢀ795ꢀ4423;
fax, (781)ꢀ795ꢀ4496; eꢀmail, ghotas.x.evindar@gsk.com.
Acknowledgements
The authors thank Matthew Lochansky for HRMS analysis and Andrew Haser for Microsomal
Stability results. We are grateful for protein production and Xꢀray diffraction data collection by
Dongming Qian, Fan Jiang, Yu Xia, Wentao Wei, and Jianhua Cai at Viva Biotech Ltd. Lastly, the
authors thank Meidy Lontoh for compound plating and logistics, in addition to Andrew Maynard and
Aaron Miller for modeling support.
Abbreviations Used
SIRT, sirtuin; ELT, Encoded Library Technology; NAD+, nicotinamide adenine dinucleotide; SAR,
structure activity relationship; ACS2, acetylꢀCoA synthetase 2; ADPR, Adenosine Diphosphate Ribose;
HATU, (Oꢀ(7ꢀazabenzotriazolꢀ1ꢀyl)ꢀN,N,N',N'ꢀtetramethyluronium hexafluorophosphate); DIEA,
Diisopropylethylamine.
ACS Paragon Plus Environment

Page 27 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
REFERENCES
6
7
8
9
1.
Bauer, J. A.; Ungvari, Z.; Minor, R. K.; Le Couteur, D. G.; de Cabo, R., Are Sirtuins Viable
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Targets for Improving Healthspan and Lifespan? Nat. Rev. Drug Disc. 2012, 11 (6), 443ꢀ461.
2. Finkel, T.; Deng, C. X.; Mostoslavsky, R., Recent progress in the biology and physiology of
sirtuins. Nature 2009, 460 (7255), 587ꢀ591.
3. Blum, C. A.; Ellis, J. L.; Loh, C.; Ng, P. Y.; Perni, R. B.; Stein, R., SIRT1 Modulation as a
Novel Approach to the Treatment of Diseases of Aging. J. Med. Chem. 2011, 54 (2), 417ꢀ432.
4. Banks, A. S.; Kon, N.; Knight, C.; Matsumoto, M.; GutierrezꢀJuarez, R.; Rossetti, L.; Gu, W.;
Accili, D., SirT1 gain of function increases energy efficiency and prevents diabetes in mice. Cell Metab.
2008, 8 (4), 333ꢀ341.
5.
Peck, B.; Chen, C.ꢀY.; Ho, K.ꢀK.; P., D. F.; Myatt, S. S.; Coombes, R. C.; Fuchter, M. J.; Hsiao,
C.ꢀD.; Lam, E. W.ꢀF., SIRT Inhibitors Induce Cell Death and p53 Acetylation through Targetting Both
SIRT1 and SIRT2. Mol. Cancer Ther. 2010, 9 (4), 844ꢀ855.
6.
Wang, R. H.; Sengupta, K.; Li, C.; Kim, H. S.; Cao, L.; Xiao, C.; Kim, S.; Xu, X.; Zheng, Y.;
Chilton, B.; Jia, R.; Zheng, Z. M.; Appella, E.; Wang, X. W.; Ried, T.; Deng, C. X., Impaired DNA
damage response, genome instability, and tumorigenesis in SIRT1 mutant mice. Cancer Cell 2008, 14
(4), 312ꢀ323.
7.
Liu, L.; Arun, A.; Ellis, L.; Peritore, C.; Donmez, G., Sirtuin 2 (SIRT2) enhances 1ꢀmethylꢀ4ꢀ
phenylꢀ1,2,3,6ꢀtetrahydropyridine (MPTP)ꢀinduced nigrostriatal damage via deacetylating forkhead box
O3a (Foxo3a) and activating Bim protein. J. Biol. Chem. 2012, 287 (39), 32307ꢀ32311.
8.
Tang, B. L., Sirt1's Complex Roles in Neuroprotection. Cell. Mol. Neurobiol. 2009, 29 (8),
1093ꢀ1103.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 45
9.
Outeiro, T. F.; Kontopoulos, E.; Altmann, S. M.; Kufareva, I.; Strathearn, K. E.; Amore, A. M.;
1
2
3
4
5
6
7
8
Volk, C. B.; Maxwell, M. M.; Rochet, J.ꢀC.; McLean, P. J.; Young, A. B.; Abagyan, R.; Feany, M. B.;
Hyman, B. T.; Kazantsev, A. G., Sirtuin 2 Inhibitors Rescue αꢀSynucleinꢀMediated Toxicity in Models
of Parkinson's Disease. Science 2007, 317 (5837), 516ꢀ519.
9
10.
Yoshizaki, T.; Milne, J. C.; Imamura, T.; Schenk, S.; Sonoda, N.; Babendure, J. L.; Lu, J. C.;
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Smith, J. J.; Jirousek, M. R.; Olefsky, J. M., SIRT1 exerts antiꢀinflammatory effects and improves
insulin sensitivity in adipocytes. Mol. Cell. Biol. 2009, 29 (5), 1363ꢀ1374.
11.
Yoshizaki, T.; Schenk, S.; Imamura, T.; Babendure, J. L.; Sonoda, N.; Bae, E. J.; Oh, D.; Lu,
M.; Milne, J. C.; Westphal, C.; Bandyopadhyay, G.; Olefsky, J. M., SIRT1 Inhibits Inflammatory
Pathways in Macrophages and Modulates Insulin Sensitivity. Am. J. Physiol.: Endocrinol. Metab. 2010,
298 (3), E419ꢀE428.
12.
Narayan, N.; Lee, I. H.; Borenstein, R.; Sun, J.; Wong, R.; Tong, G.; Fergusson, M. M.; Liu, J.;
Rovira, I. I.; Cheng, H.ꢀL.; Wang, G.; Gucek, M.; Lombard, D.; Alt, F. W.; Sack, M. N.; Murphy, E.;
Cao, L.; Finkel, T., The NADꢀdependent deacetylase SIRT2 is required for programmed necrosis.
Nature 2012, 492, 199ꢀ204.
13.
Purushotham, A.; Schug, T. T.; Xu, Q.; Surapureddi, S.; Guo, X.; Li, X., Hepatocyteꢀspecific
deletion of SIRT1 alters fatty acid metabolism and results in hepatic steatosis and inflammation. Cell
Metab. 2009, 9 (4), 327ꢀ338.
14.
regulation of SIRT1 by calorie restriction. Genes Dev. 2008, 22 (13), 1753ꢀ1757.
15. Lara, E.; Mai, A.; Calvanese, V.; Altucci, L.; LopezꢀNieva, P.; MartinezꢀChantar, M. L.; Varelaꢀ
Chen, D.; Bruno, J.; Easlon, E.; Lin, S. J.; Cheng, H. L.; Alt, F. W.; Guarente, L., Tissueꢀspecific
Rey, M.; Rotili, D.; Nebbioso, A.; Ropero, S.; Montoya, G.; Oyarzabal, J.; Velasco, S.; Serrano, M.;
Witt, M.; VillarꢀGarea, A.; Imhof, A.; Mato, J. M.; Esteller, M.; Fraga, M. F., Salermide, a Sirtuin
inhibitor with a strong cancerꢀspecific proapoptotic effect. Oncogene 2009, 28 (6), 781ꢀ791.
ACS Paragon Plus Environment

Page 29 of 45
Journal of Medicinal Chemistry
16.
Trapp, J.; Meier, R.; Hongwiset, D.; Kassack, M. U.; Sippl, W.; Jung, M., Structureꢀactivity
1
2
3
studies on suramin analogues as inhibitors of NAD+ꢀdependent histone deacetylases (sirtuins).
4
5
6
ChemMedChem 2007, 2 (10), 1419ꢀ1431.
7
8
17.
Napper, A. D.; Hixon, J.; McDonagh, T.; Keavey, K.; Pons, J. F.; Barker, J.; Yau, W. T.;
9
Amouzegh, P.; Flegg, A.; Hamelin, E.; Thomas, R. J.; Kates, M.; Jones, S.; Navia, M. A.; Saunders, J.
O.; DiStefano, P. S.; Curtis, R., Discovery of indoles as potent and selective inhibitors of the
deacetylase SIRT1. J. Med. Chem. 2005, 48 (25), 8045ꢀ8054.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18.
Sanders, B. D.; Jackson, B.; Brent, M.; Taylor, A. M.; Dang, W.; Berger, S. L.; Schreiber, S. L.;
Howitz, K.; Marmorstein, R., Identification and characterization of novel sirtuin inhibitor scaffolds.
Bioorg. Med. Chem. 2009, 17 (19), 7031ꢀ7041.
19.
Huhtiniemi, T.; Salo, H. S.; Suuronen, T.; Poso, A.; Salminen, A.; Leppänen, J.; Jarho, E.;
LahtelaꢀKakkonen, M., StructureꢀBased Design of Pseudopeptidic Inhibitors for SIRT1 and SIRT2. J.
Med. Chem. 2011, 54 (19), 6456ꢀ6468.
20.
Kiviranta, P. H.; Suuronen, T.; Wallén, E. A. A.; Leppa nen, J.; Tervonen, J.; Kyrylenko, S.;
Salminen, A.; Poso, A.; Jarho, E. M., NꢀεꢀThioacetylꢀLysineꢀContaining Triꢀ, Tetraꢀ, and Pentapeptides
as SIRT1 and SIRT2 Inhibitors. J. Med. Chem. 2009, 52 (7), 2153ꢀ2156.
21.
Tervo, A. J.; Suuronen, T.; Kyrylenko, S.; Kuusisto, E.; Kiviranta, P. H.; Salminen, A.;
Leppänen, J.; Poso, A., Discovering Inhibitors of Human Sirtuin Type 2: Novel Structural Scaffolds. J.
Med. Chem. 2006, 49, 7239ꢀ7241.
22.
Heltweg, B.; Gatbonton, T.; Schuler, A. D.; Posankony, J.; Li, H.; Goehlem, S.; Kollipara, R.;
DePinho, R. A.; Gu, Y.; Simon, J. A.; Bedalov, A., Antitumor Activity of a SmallꢀMolecule Inhibitor of
Silent Information Regulator 2 Enzymes. Cancer Res. 2006, 66 (8), 4368ꢀ4377.
23.
Suzuki, T.; Khan, M. N. A.; Sawada, H.; Imai, E.; Itoh, Y.; Yamatsuta, K.; Tokuda, N.;
Takeuchi, J.; Seko, T.; Nakagawa, H.; Miyata, N., Design, Synthesis, and Biological Activity of a Novel
Series of Human Sirtuinꢀ2ꢀSelective Inhibitors. J. Med. Chem. 2012, 55 (12), 5760ꢀ5773.
ACS Paragon Plus Environment