Design and Synthesis of Simplified Largazole Analogues as Isoform-Selective Human Lysine Deacetylase Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.5b01632

Selective inhibition of KDAC isoforms while maintaining potency remains a challenge. Using the largazole macrocyclic depsipeptide structure as a starting point for developing new KDACIs with increased selectivity, a combination of four different simplified largazole analogue (SLA) scaffolds with diverse zinc-binding groups (for a total of 60 compounds) were designed, synthesized, and evaluated against class I KDACs 1, 3, and 8, and class II KDAC6. Experimental evidence as well as molecular docking poses converged to establish the cyclic tetrapeptides (CTPs) as the primary determinant of both potency and selectivity by influencing the correct alignment of the zinc-binding group in the KDAC active site, providing a further basis for developing new KDACIs of higher isoform selectivity and potency.

Full text of DOI:10.1021/acs.jmedchem.5b01632

Subscriber access provided by CMU Libraries - http://library.cmich.edu
Article
Design and Synthesis of Simplified Largazole Analogs as
Isoform-Selective Human Lysine Deacetylase Inhibitors.
Damodara Reddy Nandarapu, Flavio Ballante, Timothy Chuang,
Adele Pirolli, Biagina Marrocco, and Garland R Marshall
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01632 • Publication Date (Web): 17 Dec 2015
Downloaded from http://pubs.acs.org on December 21, 2015
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 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design and Synthesis of Simplified Largazole Analogs as Isoform-
Selective Human Lysine Deacetylase Inhibitors.
‡
‡
‡
¥
#
Damodara N. Reddy, Flavio Ballante, Timothy Chuang, Adele Pirolli, Biagina Marrocco and
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
‡
,
Garland R. Marshall *
‡
Department of Biochemistry and Molecular Biophysics, Washington University School of
Medicine, St. Louis, MO 63110
¥
Rome Center for Molecular Design, Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza
Università di Roma, P. le A. Moro 5, 00185 Roma, Italy.
#
Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza Università di Roma, P. le A. Moro 5,
00185 Roma, Italy.
Abstract
Selective inhibition of KDACꢀisoforms while maintaining potency remains a challenge. Using the
largazole macrocyclic depsipeptide structure as a starting point for developing new KDACIs with
increased selectivity, a combination of four different simplified largazole analog (SLA) scaffolds
with diverse zincꢀbinding groups (for a total of 60 compounds) were designed, synthesized and
evaluated against class I KDACs 1, 3 and 8, and class II KDAC6. Experimental evidence as well as
molecular docking poses converged to establish the cyclic tetrapeptides (CTPs) as the primary
determinant of both potency and selectivity, by influencing the correct alignment of the zincꢀbinding
group in the KDAC active site, providing a further basis for developing new KDACIs of higher
isoform selectivity and potency.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
Lysine Deacetylases and Epigenetics.
1ꢀ3
Lysine deacetylases (KDACs) are a class ofꢀenzymes found in bacteria, fungi, plants andꢀanimals
that remove the acetyl group from the εꢀꢀamino groups of lysine residues in order toꢀinfluence gene
expression and modulate molecular recognition (epigenetics).ꢀLysine acetyltransferase (KAT)
catalyzes theꢀacetylation of histone tails causing localizedꢀrelaxation of chromatin and
transcriptionalꢀactivation of nearby genes and many other proteins modulating signaling, while
KDACsꢀcatalyze the deacetylation of acetylated histones and other proteinsꢀleading to
transcriptional repression and modulation of functional responses. There areꢀ18 known human lysine
deacetylases. KDACsꢀ1ꢀ11 are zincꢀbased enzymes and hydrolyze an amide bond similar to a
variety of zinc proteases. KDAC inhibitorsꢀ(KDACIs) are an emerging class of antitumorꢀdrugs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
Three FDAꢀapprovedꢀKDAC inhibitors: Vorinostat (suberoylanilide hydroxamic acid,ꢀSAHA),
5
6
Romidepsin (FK228) and Belinostat (PXD101) are indicated for treatmentꢀof peripheral Tꢀcell
lymphomas, in particular, cutaneous Tꢀcell lymphoma, and other types ofꢀnonꢀHodgkin's lymphoma.
The role of KDACs in pathology is ubiquitous. Besides oncology, KDACIs have shown potential
therapeutic roles in addiction, asthma, cardiovascular disease, immunosuppression,
neurodegenerative diseases, sepsis, sickleꢀcell disease and termination of viral latency among others.
It has become obvious that isoformꢀselective KDACIs are essential to help dissect the complex
dynamics of epigenetic control of gene expression/signal transduction by lysine
acetylation/deacetylation.
2
ACS Paragon Plus Environment

Page 3 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HIV-1 Latency.
As another example of potential therapeutic applications of lysine deacetylase inhibitors, KDACIs
have been shown to reverse HIVꢀ1 latency as a potential cure for patients infected with HIV to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
, 8
prevent subsequent development of AIDS.
AIDS (acquired immunodeficiency syndrome) is a
chronic, lifeꢀthreatening condition due to infection by the human immunodeficiency virus (HIVꢀ1).
Highly active antiretroviral therapy (HAART) is an effective treatment to reduce HIV RNA in
plasma to undetectable levels (generally defined as 50 copies/ml) which improves patients' lives,
prevents subsequent development of AIDS, and reduces the risk of transmission to others. Even after
9
15 years of viral suppression by HAART, however, HIV can still be harvested from resting Tꢀcells,
10
11
and other cells with latent HIV infection , and viral rebound almost always occurs after HAART
1
2
interruption. Persistent proviral human immunodeficiency virus type 1 (HIVꢀ1) infection, primarily
9
within a small population of longꢀlived resting CD4+ T cells, is a major obstacle to the eradication
of HIVꢀ1 infection. For this purpose, "shock and kill" strategies have been proposed by forcing HIVꢀ
13
infected Tꢀcells to express their HIV genome, so that infected cells can be identified and eradicated
to eliminate HIV infection. Such efforts will be aided by further understanding of epigenetic
mechanisms that regulate transcription from the HIVꢀ1 long terminal repeat (LTR) promoter and the
14, 15
role of Tat/TAR interactions.
Evidence suggests that dynamic acetylation and methylation can
1
6
regulate HIVꢀ1 proviral expression. KDACs are recruited to the initiator and enhancer regions of
the HIVꢀ1 LTR by several transcription factors and coꢀrepressor complexes. More recently it has
4
been found that nonꢀselective histone deacetylase inhibitors (pan KDACIs), such as SAHA,
17
5
Trichostatin A (TSA) and FK228, have been shown to disrupt HIV latency and provides proof-of-
1
8, 19
20
principle,
but potential toxicity remains a significant concern with any epigenetic therapeutic.
Furthermore, no current KDACI is specific for any of the eleven zincꢀbased HDAC isoforms, and
there is no clinical experience to estimate the length of treatment necessary to eradicate the HIV
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
reservoir. Recently, it has shown that KDAC inhibitors specific for a limited number of class I
2
1, 22
KDACs may offer a targeted approach to the disruption of persistent HIVꢀ1 infection.
Even if a
vaccine were found to be effective, however, in preventing the spread of HIV, ~34 million of HIVꢀ
positive patients worldwide remain to be cured of their infections. To address this potential
therapeutic opportunity, however, isoformꢀselective KDACIs are essential to determine the
mechanism of overcoming HIV latency and to minimize adverse side effects in the clinic.
Furthermore, development of isoformꢀspecific inhibitors would enable pharmacological dissection of
the role of KDAC isoforms in both physiology as well as pathology.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KDAC Inhibitors.
23
24
Many KDAC inhibitors have been reported, both naturally occurring such as TSA, apicidin, and
25
5
26ꢀ28
29
trapoxin, FK228, azumamide analogs,
largazole , and a vast number of synthetic compounds,
30
such as SAHA. These KDAC inhibitors (Figure 1) can be divided into categories according to their
structural chemistry, such as hydroxamates, carboxylates, benzamides, and cyclic peptides, though
31
most are hydroxamic acid derivatives. All classes of KDACIs contain three key structural elements:
i) a zincꢀbinding group (warhead) which coordinates the zinc ion at the bottom of the 12 Å narrow,
(
activeꢀsite cavity, (ii) a capping or headgroup which interacts with the amino acids on the rim of the
isoform’s binding cavity as the substrateꢀprotein recognition surface and (iii) a linker domain whose
role is to ensure the correct positioning of the two former groups and to interact with the lipophilic
binding tunnel. The structures of naturally occurring, potent inhibitors of human KDACs often
contain a cyclic tetrapeptide “headgroup” linked to a pendant aliphatic spacer with a terminal
“
warhead” functional group that interacts with zinc in the active site. Despite an intensive search for
32
33
over a decade in both academia and industry, and the discovery of selective KDAC1, KDAC3,
3
4ꢀ36
KDAC6
inhibitors , the development of new KDACIs endowed with higher potency and
selectivity is still necessary to clarify the therapeutic effect of a single KDAC isoform inhibition.
4
ACS Paragon Plus Environment

Page 5 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
O
O
O
H
OH
N
OH
OH
N
OH
N
N
H
N
N
H
H
O
SAHA
TSA
PCI-34051
O
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
O
N
H
N
N
H N
2
N
NH2
N
O
N
H
H
H
N
N
N
S
O
O
N
MS-275
T247
Tubastatin A
O
O
O
O
N
O
N
O
O
H
N
N
N
HN
O
H
H
H
NH
O
N
O
HN
H
NH
NH
NH
O
H
N
H
H
O N
N
N
N
O
O
O
O
O
O
O
NH2
O
O
OH
O
Azumamide A
Apicidin
Azumamide E
Trapoxin B
S
N
S
N
S
N
S
S
O
O
N
O
N
O
HN
O
NH
N
H
HN
O
HN
HN
O
HN
HN
O
O
O
O
S
HN
O
S
O
O
O
O
H
N
O
O
S
SH
S
O
O
FK228
SD-L-256
Largazole
Largazole thiol
Figure 1: Examples of Potent KDAC inhibitors.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Schematic of KDAC inhibitors.
Design of Isoform-Selective KDAC Inhibitors.
Macrocyclic KDAC inhibitors (KDACIs) possess complex capping groups which interact with the
37
KDAC enzyme’s outer rim and show excellent KDACI potency with some isoform selectivity. The
structure of largazole in complex with human and Schistosoma mansoni KDAC8s have been
3
8, 39
solved.
Comparison of the crystal structures of KDAC isoforms (2, 3, 4, 7, and 8) show that the
length and conformation of the loops at the rim of the lysineꢀbinding channel where the macrocyclic
headgroups interact with the KDAC substrate are extremely variable, and a likely source of
enzyme/proteinꢀsubstrate specificity. Overemphasis on targeting the zinc in the essentially identical
active site of KDAC isoforms may explain the limited progress in obtaining isoform selectivity.
More recently, KDACIs have been developed without a zincꢀbinding group to examine this
40ꢀ42
strategy.
Variability in the headgroupꢀbinding region of KDAC isoforms provides one strategy
for selectively targeting KDAC isoforms. KDACs target specific sites on substrate proteins;
consequently, they must discriminate among the protein structures bearing the lysine residue to be
modified. In naturally occurring KDACIs, macrocyclic cyclic tetrapeptides (CTPs) are located as
capping groups at the recognition interface between the protein substrate and the KDAC. This
strategy focuses on proteinꢀsubstrate discrimination by different KDACs as seen in the example of
43
39
the complex of largazole with human KDAC8 and its homolog from Schistosoma mansoni.
6
ACS Paragon Plus Environment

Page 7 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Strategy for Isoform Selectivity in KDACIs.
In this work, we have chosen to explore four different simplified largazole analog (SLA) scaffolds
with a wide variety of potential zincꢀbinding groups and directly compared their activities for a series
of KDAC isoforms. Largazole thiol (Figure 1) is the most potent KDACI (picomolar against
KDACs 1, 2 & 3) and selective to class I KDACs. Bowers et al. made one analog ~3ꢀ4ꢀtimes as
potent as largazole against KDAC1, 2 & 3 by substituing a pyridine group for one of the thiazole
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
44
rings. In our search of isoformꢀselective KDACIs, largazole has been modified to simplify the
synthesis of analogs with aminoꢀacid synthons as shown in Figure 4. Replacing the thiazole ring
with heterochiral proline dipeptides (ProꢀDꢀpro & Dꢀ
Figure 3. Original docking of largazole
4
5ꢀ47
proꢀPro) that stabilize reverse turns
generated two
to KDAC8 (left) compared with crystal
structure of largazole in complex; (right)
docked structure of simplified largazole
analog overlapped with crystal structure
of largazole when complexed to KDAC8.
CTP scaffolds. Preliminary conformational analysis
of the simplified largazole analogs (SLAs) gave
promising overlaps with the conformation of
4
3
largazole bound to KDAC8 (Figure 3). Further
simplification replaced the unusual βꢀhydroxy acid
bearing the thiol warhead with aspartic acid to which
a variety of warheads could be linked by an amide
bond when the trans olefin was replaced with an
equivalent amide bond (Figure 1). Replacement of
the depsipeptide ester link in the largazole ring with an amide bond by Bowers et al. showed little
4
8
differences in isoform selectivity, and only a fourꢀfold reduction in affinity to 4 nM. The proline
rings of the SLA provide a semiꢀrigid structure for potential replacement with substituted prolines to
further enhance specificity for KDAC isoforms. Varying the stereogenic center (Asp vs. Dꢀasp)
where the linker and warhead were attached might further optimize the orientation of the zincꢀ
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
binding group. This proposed scaffold allowed combinatorial optimization of the CTP headgroup
and the zincꢀbinding group by preassembly of the CTP headgroup followed by attachment of the
linkerꢀzincꢀbinding group through an amide bond. Optimization of both the linker and the zincꢀ
binding group would hopefully generate maximal KDACꢀisoform specificity. Unfortunately, the
simplified largazole analogs did not bind to KDACs in the same orientation as largazole itself, based
on the loss of inhibitory activity observed. A detectable KDAC6 inhibitory selectivity (among
KDAC1, 3, 6 and 8) was measured for compound 25a, to help improve ligandꢀKDAC isoform
recognition. Due to the loses of biological responses, an comprehensive docking assessment protocol
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
49
has been developed and applied to the human lysine deacetylases (KDACs), determining the best
docking protocol to be used in rationalizing SLAsꢀKDACs interactions.
8
ACS Paragon Plus Environment

Page 9 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
S
S
N
O
N
HN
HN
O
O
O
D-pro
Pro
Pro
D-pro
S
N
N
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
N
O
O
N
O
O
Largazole
O
O
O
Asp
HN
Val
Asp
HN
Val
O
O
HN
HN
NH
NH
O
O
R
R
5
6
a) n = 1, R = SH
a) n = 2, R = SH
7a) n = 2, R = SMe
12a) n = 2, R = COOMe
13a) n = 2, R = COOH
14a) n = 2, R = CONHOH
a) n = 2, R = S(O)Me 15a) n = 2, R = S(CH2)2NH2
a) n = 2, R = S(O)2Me 16a) n = 2, R = NH2
17a) n = 1, R = SH
18a) n = 2, R = SH
19a) n = 2, R = SMe
20a) n = 2, R = S(O)Me 25a) n = 2, R = CONHOH
21a) n = 2, R = S(O)2Me 26a) n = 2, R = S(CH2)2NH2
22a) n = 2, R = OH
23a) n = 2, R = COOMe
24a) n = 2, R = COOH
8
9
10a) n = 2, R = OH
1a) n = 2, R = S(CH2)2NHP(O)(OPh)2
1
D-pro
Pro
Pro
D-pro
N
N
N
O
N
O
O
O
O
O
D-asp
HN
Val
D-asp
HN
Val
O
O
HN
HN
NH
NH
O
O
R
R
5
b) n = 1, R = SH
6b) n = 2, R = SH
b) n = 2, R = SMe
12b) n = 2, R = COOMe
13b) n = 2, R = COOH
14b) n = 2, R = CONHOH
17b) n = 1, R = SH
18b) n = 2, R = SH
19b) n = 2, R = SMe
20b) n = 2, R = S(O)Me 25b) n = 2, R = CONHOH
21b) n = 2, R = S(O) Me 26b) n = 2, R = S(CH ) NH
2
22b) n = 2, R = OH
23b) n = 2, R = COOMe
24b) n = 2, R = COOH
7
8
b) n = 2, R = S(O)Me 15b) n = 2, R = S(CH2)2NH2
9b) n = 2, R = S(O)2Me 16b) n = 2, R = NH2
0b) n = 2, R = OH
2
2 2
1
1
1b) n = 2, R = S(CH2)2NHP(O)(OPh)2
Figure 4: Designing of selective KDAC inhibitors.
The Yoshida group has shown significant isoformꢀselectivity enhancement against human KDACs
with chlamydocin analogs by varying the chemical nature of the warhead interacting with the activeꢀ
50
site zinc. The choice of chemical functionality for binding to zinc at the active site is quite
51
diverse. In addition to optimizing the headgroup for isoform selectivity, we varied the chemical
nature of the warhead interacting with the zinc in the active site from the sulfhydryl, present in
FK228 and largazole, to thioether, sulfone, carboxyl ester, carboxylic acid, amine, hydroxyl and
hydroxymate as well as methylsulphoxide functionality presents in sulforaphane, a naturally
52, 53
occurring HDAC inhibitor commonly found in cruciferous vegetables
. Optimization of the
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
warhead interacting with the zinc in the active site by transitionꢀstate analogs (TSAs) of amide
51
hydrolysis may further improves the selectivity and potency. One can imagine extension of the
54
zincꢀbinding headgroup to access the cavity associated with acetate release. Thus, we plan to utilize
experimental data from other classes of KDACIs to further optimize the isoformꢀselectivity
advantages for KDAC as seen for each of the three components; CTP, linker and warhead.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results and Discussions
First, we synthesized CTPs 1a, 1b, 2a and 2b containing a pendant OBnꢀprotected carboxylic acid
on the alpha carbon of Asp, which was easily elongated with aminopropyl zincꢀbinding warheads.
t
Synthesis of CTP shown in Scheme 1. Using a solutionꢀphase strategy, we started with Oꢀ Butyl
valine ester hydrochloride and coupled it with FmocꢀPro/DꢀproꢀOH using standard EDC coupling to
give protected dipeptide 27 and 28 respectively. The Fmoc group at Nꢀterminal of dipeptide was
deprotected using 20% piperidine in dichlomethane (DCM) and coupled with FmocꢀDꢀpro/ProꢀOH
to yield the linear tripeptides 29 and 30, these tripeptides aggregated to form an insoluble material
once solidified that decomposed on silica gel. To overcome this difficulty, we proceeded further
without purification and characterization of the tripeptides. The Fmoc groups at the Nꢀterminals of
the tripeptides 28 and 29 were deprotected and coupled with BocꢀAsp(OH)ꢀOBn or with BocꢀDꢀ
asp(OH)ꢀOBn to yield linear tetrapeptides 31a, 31b, 32a and 32b respectively. Nꢀterminal Boc and
t
Cꢀterminal O Bu ester protection groups were then removed quantitatively using 30% trifluoroacetic
acid in DCM to obtain the desired products. Then, macrolactamization was optimized at 2 mM in
DMF using diphenylphosphoryl azide (DPPA) and diisopropylethylamine (DIPA); the CTPs were
obtained in good yields as shown in Scheme 1. No epimerized or dimerized compound was observed
during cyclization at 2 mM concentration in DMF. By synthesizing four CTP scaffolds with a
pendant carboxyl group, a variety of warheads could be attached through an amide bond to determine
10
ACS Paragon Plus Environment

Page 11 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the relative affinities of different warheads in the same structural context against different KDAC
isoforms.
Scheme 1: Synthesis of CTP reverseꢀturn scaffolds.
(
ii), (iii)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(i)
t
t
Fmoc-Yaa-Val-O Bu
Fmoc-Yaa-OH
H-Val-O Bu.HCl
+
2
2
7) Yaa = Pro, 87%
8) Yaa = D-pro, 92%
(
ii), (iv)
t
t
Boc-Zaa(Xaa-Yaa-Val-O Bu)-OBn
Fmoc-Xaa-Yaa-Val-O Bu
29) Xaa = D-pro, Yaa = Pro, 83%
30) Xaa = Pro, Yaa = D-pro, 76%
31a) Xaa = D-pro, Yaa = Pro, Zaa = Asp(OH)-OBn, 77%
31b) Xaa = D-pro, Yaa = Pro, Zaa = D-asp(OH)-OBn, 75%
3
3
2a) Xaa = Pro, Yaa = D-pro, Zaa = Asp(OH)-OBn, 73%
2b) Xaa = Pro, Yaa = D-pro, Zaa = D-asp(OH)-OBn, 70%
(
v)
(
vi)
H-Zaa(Xaa-Yaa-Val-OH)-OBn
Cyclo-Zaa(Xaa-Yaa-Val)-OBn
3
3a) Xaa = D-pro, Yaa = Pro, Zaa = Asp
1a) Xaa = D-pro, Yaa = Pro, Zaa = Asp, 79%
1b) Xaa = D-pro, Yaa = Pro, Zaa =D-asp, 78%
D
3
3
3b) Xaa = D-pro, Yaa = Pro, Zaa = Asp
2
2
a) Xaa = Pro, Yaa = D-pro, Zaa = Asp, 67%
b) Xaa = Pro, Yaa = D-pro, Zaa = D-asp, 73%
L
4a) Xaa = Pro, Yaa = D-pro, Zaa = Asp
D
34b) Xaa = Pro, Yaa = D-pro, Zaa = Asp
(
v)
Cyclo-Zaa(Xaa-Yaa-Val)-OH
3
3
4
4
a) Xaa = D-pro, Yaa = Pro, Zaa = Asp, 96%
b) Xaa = D-pro, Yaa = Pro, Zaa =D-asp, 94%
a) Xaa = Pro, Yaa = D-pro, Zaa = Asp, 94%
b) Xaa = Pro, Yaa = D-pro, Zaa = D-asp, 91%
Reagents and conditions: (i) EDCl, HOBt, DIEA, DCM, 9 h; (ii) 20% Piperidine/DCM, rt, 0.5 h, 100%;
iii) Fmoc-D-pro-OH, EDCl, HOBt, DIEA, DCM, 12 h; (iv) Boc-Zaa-OBn, EDCl, HOBt, DIEA, DCM, 24
(
o
o
h; (v) 30% TFA/DCM, 0 C-rt, 3 h, 100%; (vi) DPPA, DIEA,DMF, 0 C-rt, 72 h; (v) H , Pd/C, EtOH, rt,
2
5-6 h.
Diastereomerically pure CTPs 3a, 3b, 4a and 4b containing free pendant carboxyl groups were
obtained quantitatively by hydrogenation of the OBn ester in the presence of Pd/C (see Supporting
Information sections S11ꢀS14). Epimerization at the Asp αꢀcarbon stereocenter was observed,
however, upon saponification of the OBn ester by lithium hydroxide (LiOH). The epimerized
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
compound was separated by HPLC and confirmed as 3b by comparing LCꢀMS data, H and
1
3
CNMR spectra of 3b (see Supporting Information section S10), which was authentically
synthesized by Pd/C hydrogenation. CTP scaffolds with a pendant carboxyl group was coupled to
,2′ꢀdiaminodiethyl disulfide and 3,3'ꢀdiaminodipropyl disulfide in presence of EDCL yielded the
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
desired CTPs having disulfide bridges of two and three carbons, respectively; phosphamide warhead
CTP analogs 11a and 11b where obtained in the presence of DPPA. Upon treatment of disulphide
containing CTPs (35a-38a and 35b-38b) with (2S,3S)ꢀ1,4ꢀbis(sulfanyl)butaneꢀ2,3ꢀdiol (DTT),
sulfhydrylꢀwarhead SLA analogs (5a, 6a, 5b, 6b, 17a, 17b, 18a and 18b) were obtained as shown in
the Scheme 2.
Scheme 2: Introduction of sulfhydryl zincꢀbinding functional group on CTP reverseꢀturn scaffolds.
Methylthioether, hydroxyl and amine zincꢀbinding, threeꢀcarbon warheads were introduced using
standard EDCL coupling to give a wide variety of SLA analogs. Sulfoxides and sulfones were
12
ACS Paragon Plus Environment

Page 13 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
introduced by boraxꢀcatalyzed, pHꢀcontrolled selective oxidation of organic sulfides by hydrogen
55
2 2
peroxide (H O ) according to a reported procedure. Methylthioether was treated with hydrogen
peroxide in presence of borax in methanol (MeOH) and water mixture furnished the desired
sulfoxide analogs (8a, 8b, 20a and 20b) at acidic pH and sulfones (9a, 9b, 21a and 21b) at basic pH
(Scheme 3).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3: Introduction of thiol, thioether, sulfoxide and sulfone zincꢀbinding groups.
Pendant carboxylic acids on CTPs were coupled with methyl 4ꢀaminobutanoate to give methylester
warheads (12a, 12b, 23a and 23b), which upon treatment with hydroxylamine in presence of
potassium hydroxide (KOH) furnished hydroxamic acid zincꢀbinding warheads (14a, 14b, 25a and
25b). SLA analogs 13a, 13b, 24a and 24b having a carboxylic acid group were obtained
quantitatively upon
saponification
of methyl
esters
by
LiOH
(Scheme
4).
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 4: Introduction of carboxylic acid, ester, and hydroxamate zincꢀbinding groups
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Bioactivity Against Isolated KDAC Isoforms.
A selection of synthetized SLA compounds (a total of 60, Supporting Information Section S18) was
assayed for their inhibitory activity against four human KDACs (herein KDACs 1, 3, 6 and 8), to
clarify how the different CTP headgroups determine isoform selectivity (see Tables 1 and 2). In vitro
assays were performed on recombinant human KDACs 1, 3 and 8 (Class I KDACs) and KDAC 6
(Class II KDACs) using the Caliper EZ Reader II system (PerkinꢀElmer Caliper Life Sciences,
USA).
Isoform selectivity.
Except for CTPs with hydroxymate groups (compounds 14a, 14b, 25a and 25b), the SAR results
clearly demonstrate the limited ability of the simplified largazole scaffolds to determine strong or
selective interactions with the 4 used KDAC isoforms, showing generally weak activities (Table 1
14
ACS Paragon Plus Environment

Page 15 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and Supporting Information Tables 1-2). Conversely, SLAs with hydroxamate groups (compounds
4a, 14b, 25a and 25b) show a general higher inhibitor potency compared to other SLAs (see
1
Supporting Information Tables 1-2), confirming a prominent role of the hydroxamate group in
driving the ligandꢀprotein recognition by interacting with the zinc ion. Compounds 14a and 25a
show IC s in a micromolar range with a detectable selectivity for KDAC6 (Table 1 and Supporting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
Information Tables 1-2). These two compounds are DꢀproꢀPro and ProꢀDꢀpro SLAs, and both are
Asp derivatives. Thus, the presence of an Asp group instead of Dꢀasp consistently justifies the
general increased selectivity vs KDAC6 (compare inhibition percentages of 14a with 14b and 25a
with 25b in Supporting Information Tables 1-2) which is worth of consideration in designing new
KDAC6 selective SLAs. This is consistent with the hypothesis that the selected SLA scaffolds
oriented the linker, the warhead and the capping group differently from largazole by binding in a
different orientation, as confirmed by docking results (see molecular docking results of SLAs section
and Figure 10). Nevertheless, some KDACꢀisoform selectivity between KDACs 1, 3, 6 & 8 was
observed for nonꢀhydroxamate derivatives (Figures 6-7 and Supporting Information Tables 1-2) that
might be useful for future optimization. DꢀproꢀPro CTP analogs with thiomethyl, sulfoxide and
sulfone warheads (compounds 8a-b and 9a-b, Supporting Information Table 1) showed some
selectivity for KDAC8 compared with KDACs 1, 3 & 6. The data further suggest:
•
The strong hydroxamic acid zincꢀbinding group generates, as reported above, the highest
KDAC inhibitory activity (except in the case of the DꢀaspꢀProꢀDꢀpro scaffolds), In these
cases, the strong zincꢀbinding group drives isoform recognition.
•
•
Thiol zincꢀbinding groups (along with the disulfides which decompose to thiols) showed the
highest KDAC6 inhibitory activity next to hydroxamates.
Both the Asp αꢀcarbon stereocenter and CTP type can modulate and change the activity
profile and drive ligandꢀprotein selectivity if a weaker zincꢀbinding groups are present: in the
case of sulfoxides, sulfones, hydroxyls, thiomethyl and carboxylic acid zinc-binding groups,
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the interaction seems to be dominated by the CTP moiety that favors interactions with the
KDAC8 isoform.
•
AspꢀDꢀproꢀPro derivatives with sulfoxides, sulfones, hydroxyls zinc-binding groups and Aspꢀ
ProꢀDꢀpro derivatives with thiomethyl, sulfoxides, sulfones and carboxylic acids zinc-binding
group show the highest KDAC8 selectivity. About the DꢀproꢀPro derivatives, with the above
zincꢀbinding groups, the replacement of LꢀAsp with Dꢀasp doesn’t appear to affect selectivity
on KDAC8 (except for the thiomethyl case, in fact 7b is completely selective on KDAC8, but
not 7a). For ProꢀDꢀpro scaffolds with the above zincꢀbinding groups, the replacement of Lꢀ
Asp with Dꢀasp eliminates selectivity on KDAC8 (except for the thiomethyl case, in fact 19a
and 19b have the same activity profile).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cases with no high selectivity, but with significant activity changes.
•
As shown in Figures 6 and 7, hydroxamic acid warheads (generally recognized as most
potent, but with weak selectivity between different KDACs) provided, as in the case of the
AspꢀDꢀproꢀPro compound 14a (Figure 6A and Supporting Information Table 1), the highest
potency among the three different isoforms (KDACs 3, 6 and 8) with a detectable preference
for KDACs 3 and 6 (2.16 and 2.65 times more active than with KDAC8, respectively) while
the KDAC8 inhibitor activity was only average. The activity trend of compound 14a was
confirmed also for the DꢀaspꢀDꢀproꢀPro analog 14b (Figure 6B), but interestingly without
the large potency difference between KDAC3 and KDAC8 (1.29 times more active with
respect to KDAC8, while the KDAC6 inhibitor potency was 2.52 times with respect to
KDAC8). This clearly suggests a different KDAC3 conformational “availability and
response” based on the Asp αꢀcarbon stereocenter. Conversely, the ProꢀDꢀpro hydroxamic
acid derivatives showed a different activity potency and selectivity trend (compare compound
16
ACS Paragon Plus Environment

Page 17 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
25a with 25b, Figures 7A-B, Supporting Information Table 2), with a loss of activity for
KDAC3 and KDAC6 in the case of the Dꢀasp analog (compound 25b). The effect on activity
from the Asp αꢀcarbon stereocenter dominates considering the ProꢀDꢀpro derivative over the
three KDAC isoforms even when a strong zincꢀbinding group (hydroxamic acid) was present.
Another interesting case was represented by CTPs characterized with thiol warheads: among
the 3ꢀcarbon linkerꢀdomain thiols, the AspꢀDꢀproꢀPro and the DꢀaspꢀProꢀDꢀpro compounds
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
•
6a and 18b, respectively, showed the higher inhibition potency on KDAC6 (compounds 6b
and 18b) suggesting, also in this case, a modulating role of the different CTPs. In fact,
considering the DꢀproꢀPro derivatives, the Asp derivative (compound 6a) was generally more
active compared with the Dꢀasp analog, whereas for ProꢀDꢀpro derivatives, the Asp analog
(
compound 18a) was less active (but not for KDAC8) compared with the Dꢀasp derivative
compound 18b).
(
•
A third interesting case was the disulfide compounds with a 3ꢀcarbon linker (compounds 37a,
37b, 38a and 38b): among the DꢀproꢀPro analogs, it is appreciable how the Dꢀasp derivative
(
compound 37b) maintained detectable KDAC3 inhibitory activity that was totally lost with
the Asp analog (compound 37a).
Cases with high selectivity.
•
•
•
Sulfoxides (compounds 8a, 8b and 20a) derivatives showed a clear selectivity for KDAC8
regardless of the CTP type. Only in the case of the DꢀaspꢀProꢀDꢀpro analog (compound 20b),
no selectivity and potency was detected.
Sulfones (compounds 9a, 9b and 21a) derivatives showed clear selectivity for KDAC8
regardless the CTP type. Only in the case of the DꢀaspꢀProꢀDꢀpro analog (compound 21b), no
selectivity or potency was detected.
Hydroxyls (compound 10a, AspꢀDꢀproꢀPro) showed selectivity for KDAC8, other analogs
(10b, 22a and 22b) did not share that selectivity.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
•
•
Carboxylic Acids (only ProꢀDꢀpro derivatives, compounds 24a and 24b) showed selectivity
on KDAC8.
Thiomethyl derivatives (7b, 19a and 19b) showed selectivity on KDAC8, the presence of
AspꢀDꢀproꢀPro (as in the case of 7a) neutralized the selectivity and decreased the potency.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Our hypothesis that the warhead groups of the SLA scaffolds were not oriented correctly when
bound to KDAC isoforms so that they did not interact optimally with the activeꢀsite zinc An
intensive effort to find docking protocols that demonstrates such a difference by developing a
49
protocol capable to detect a reliable structureꢀbased procedure and applied to the lysine deacetylase
4
9
(
KDAC) inhibitors, providing a mechanism for optimizing novel CTP scaffolds. Molecular
docking results confirmed this hypothesis (see Molecular Docking Results of SLAs paragraph and
Figure 10).
18
ACS Paragon Plus Environment

Page 19 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1: Most active SLAs and standard compounds (positive controls) IC50 (ꢁM) values.
IC50 ꢁM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KDAC isoform
ID
#
1
3
6
8
a
1
(12)
14a
25a
25.1
13.7
5.07
>30
>30
>30
>30
0.49
0.50
>30
>30
4.55
2.34
a
2 (43)
>30
>30
7.66
>30
5
6
3
4
5
6
7
8
9
MS-275
Largazole
1.48
2.33
14.41
0.0070
3.48
1.11
0.015
2.87
0.79
1.36
>30
b44
9.29
4.57
0.0014
1.61
>30
32
PCI-34051
4
SAHA
0.0014
0.47
3.94
0.020
0.77
5
7
SD-L-256
3
3
T247
17
TSA
Tubastatin A
0.038
0.014
3
4
10
a
Number in the complete list of compounds (see Supporting Information section S18)
b
44
The thioester form of largazole (as depicted in Figure 1) was assayed.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A
B
Figure 6. Simplified Largazole (DꢀproꢀPro) analogs inhibition profiles plots. A: Asp derivatives; B:
Dꢀasp derivatives.
20
ACS Paragon Plus Environment

Page 21 of 74
Journal of Medicinal Chemistry
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
A
B
Figure 7. Simplified Largazole (ProꢀDꢀpro) analogs inhibition profiles plots. A: Asp derivatives; B:
Dꢀasp derivatives.
Molecular Modeling
Molecular docking was used to rationalize the SLA binding poses and help clarify any correlation
with the obtained KDAC1, 3, 6 and 8 isoforms’ biological activities. A comprehensive molecular
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
docking investigation on lysine deacetylases was performed by means of the recently released
49
Clusterizer and DockAccessor software, with the purpose of detecting a best docking protocol to be
58
59
49
applied, and used also to develop predictive COMBINEr or 3ꢀD QSAR models. As reported,
a
6
0
comprehensive docking assessment proposed PLANTS (using the PLP scoring function and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
49
considering the best docked poses, BD) eligible for predicting reliable SLAs binding poses. Due to
the loss in activity of SLAs, a reliable binding pose analysis was not possible among the different
KDAC isoforms, except for comparing their predicted binding poses with respect to the largazole
thiol experimental one characterizing the KDAC8 coꢀcrystal structure (PDB code 3RQD).
KDAC8 complex structure preparation.
4
3
49
KDAC8 coꢀcrystal structures (PDB code 3RQD) was prepared as previously reported,
Subsequently, compounds 14a and 25a were docked as described.
Docking Assessment.
49
As previously reported, an extensive docking assessment was applied to detect the best docking
61
62
63
strategy using 9 different docking programs: AutoDock4, AutoDock4(Zn), AutoDock Vina,
64
65
60
66
60
DOCK6, MpSDockZn, PLANTS, and SurflexꢀDock. ). PLANTSPLP has been detected (using
the Hungarian symmetryꢀcorrected heavyꢀatom RMSD, HA_RMSDh and the obtained best docked
49
poses, BD) as the preferable docking program (showing the best compromise between performance
and speed) to be used to dock largazole and its analogs, confirming its capability in predicting the
binding pose of largazole thiol in the native protein (3RQD, Figure 8), thus validating it for docking
SLAs.
22
ACS Paragon Plus Environment

Page 23 of 74
Journal of Medicinal Chemistry
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
A
B
4
3
Figure 8. Docking of Largazole in 3RQD chain ꢀ A. Best Docked poses (BD); A: Experimental
Conformation ReꢀDocking (ECRD); B: Random Conformation ReꢀDocking (RCRD). Experimental
binding pose is green colored, BD predicted binding pose is magenta colored.
Molecular Docking Results of compounds 14a and 25a.
60
PLANTSPLP was used to dock compounds 14a and 25a (the most active ones) in KDAC8 using a
49
same strategy as previously reported. A first comparison between the predicted binding poses of
1
4a and 25a and the experimental crystal structure of largazole thiol in KDAC8 (PDB code:3RQD)
clearly show how the heteroproline dipeptide reverseꢀturn mimetics bind in a different manner
compared to the largazole thiol (Figure 9), suggesting that the loss of planarity between the two
heteroproline determines negative interactions and, as a consequence, low inhibition potency against
KDAC8 (Table 1). The presence of a dihedral angle between the two heteroprolines leading to nonꢀ
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
planarity has a negative effect also on the inhibitor potency against other KDACs preventing the
necessary interaction with the zinc ion.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A
B
Figure 9. Comparison between largazole thiol (green) experimental binding pose and SLAs 14a and
5a BD poses in KDAC8 (PDB code: 3RQD). A: Predicted BD poses of compound 14a (AspꢀDꢀproꢀ
2
Pro, cyan); B: Predicted BD poses of compound 25a (AspꢀProꢀDꢀpro, grey).
Conclusions
In conclusion, we have explored four different simplified largazole analog (SLA) scaffolds by
replacing the thiazole ring in largazole with heterochiral proline dipeptides (ProꢀDꢀpro & DꢀproꢀPro).
We have synthesized totally 52 SLA analogs by introducing a wide variety of potential zincꢀbinding
groups on pendant carboxylic acids in CTPs. The effort to reduce the synthetic complexity of
largazole by replacement of the bisꢀthiazole and unusual amino acid with readily accessible amino
24
ACS Paragon Plus Environment

Page 25 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
acid synthons did not give the anticipated results. Preliminary docking studies with KDAC8 of the
simplified largazole analog (SLA) indicated a plausible overlay that encouraged the synthetic effort
described herein. The experimental results obtained, however, clearly:
•
indicate that binding of the SLA analogs reorients the linker and warhead in different
positions that precluded their anticipated interaction with the zinc in the active site;
suggest the detrimental effect of a dihedral angle between the two heteroproline;
confirm the hydroxamate group capable of establishing stronger interactions with the zinc
ion, partly reversing the depressing effect of the used CTP headgroups.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
•
•
• Reveal the effect of the Asp group (where the linker and warhead are attached) which directs
to a higher KDAC6 selectivity, suggesting its use when design new compounds with
increase potency and selectivity against KDAC6.
A recent docking protocol has been developed and assessed based on these results and will be used,
together with the above reported experimental evidencies, in the design of new SLA headgroups
capable to properly interact with the zinc and endowed with higher activity and selectivity.
Experimental Section
Chemistry General. All the reactions were performed in ovenꢀdried apparatus and were stirred
using magnetic stirbars. Starting materials, reagents, and solvents were purchased from commercial
vendors unless otherwise noted. Chromatography grade ethylacetate, dichloromethane, acetonitrile,
and hexanes were obtained from SigmaꢀAldrich. Column chromatography was performed on silica
gel (100ꢀ200 mess) purchased from Sorbent Technologies. TLC was carried out on Analtech 200
microns silicaꢀgel coated plasticꢀfiber sheets. All reactions were monitored by thin layer
chromatography (TLC) carried out on Merck silicaꢀgel plates (0.25 mm thick, 60F254), visualized by
using UV (254 nm) or dyes such as ninhydrin, KMnO4, pꢀanisaldehyde or CAM (ceric ammonium
molybdate). Highꢀperformance liquid chromatography (HPLC) was carried out on GILSON GXꢀ281
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
using Waters C18 5ꢁM, 4.6*50mm and Waters Prep C18 5ꢁM, 19*150mm reverse phase columns.
2 3
The mobile phases used were A: H O with 0.05% TFA, B: CH CN with 0.05% TFA using a solvent
gradient of AꢀB over 30 min with a flow rate of 14.8 mL/min, with detection at 220 and 254 nm UV
detectors. Purity assessment and mass spectra (MS) data were obtained using a HewlettꢀPackard
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
13
HPLC/MSD using electrospray ionization (ESI) for detection. H and C NMR spectra were
measured on a Varian 400 MHz NMR instrument. Chemical shifts are expressed in parts per million
(ppm) from the residual of nondeuterated solvents as internal standard. (1H NMR: TMS δ = 0.00
13
ppm, CDCl
3
δ
= 7.26 ppm, DMSOꢀd
6
δ
= 2.50 ppm, D
2
O: δ= 4.79 ppm; C NMR (APT): TMS δ =
0.00 ppm, CDCl
3
δ
= 77.16 ppm, DMSOꢀd δ
6
= 39.52 ppm). Coupling constants (J) are given in
hertz (Hz). The following abbreviations were used to express the multiplicities: s = singlet; d =
doublet; t = triplet; q = quartet; p = pentet; quin = quintet; sep = septet; hept = heptet; m = multiplet;
dd = doublet of doublets; dt = doublet of triplet; td = triplet of doublet; m = multiplet; bs = broad
singlet. All compounds used for biological assays are >95% pure based on NMR and LCꢀMS by UV
absorbance at 210 nm and 254 nm wavelengths.
General procedure for synthesis of CTPs by using macrolactamization (1a, 1b, 2a and 2b)
protocol:
To a cold (0° C) stirred solution of N- and Cꢀterminal deprotected tetrapeptide (1 mmol) in dry DMF
(500 mL) was added N,Nꢀdiisopropylethylamine (6 mmol) dropwise over 1 min followed by
Diphenyl phosphorazidate (DPPA) (4 mmol) then slowly bring it to room temperature and stirred
until the complete consumption of starting material, then DMF was removed under reduced
pressure. The resulting viscous solution was diluted with water (5 mL) and thoroughly extracted with
ethyl acetate (15 mL). The combined organic extracts were washed with 1 N HCl (5 mL), saturated
aqueous sodium bicarbonate (NaHCO ) (5 mL) and dried over anhydrous sodium sulphate (Na SO )
3
2
4
26
ACS Paragon Plus Environment

Page 27 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and concentrated to give a residue, which was purified by silica gel (100ꢀ200 mesh) flash column
chromatograph.
General procedure for catalytic hydrogenation of benzyl ester (OBn) protecting group (3a, 3b,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4a and 4b):
A stirred solution of the benzyl ester (1 mmol) and 10% Pd/C (0.1 mmol) in EtOH (10 mL) at 25 °C
was placed under an atmosphere of hydrogen. After 4 h, the mixture was filtered through Celite
using EtOH as eluent, and concentrated under reduced pressure and proceeded further without
purification.
General procedure for introducing warhead on pendent carboxylic acid by using EDCl peptide
coupling (7a, 7b, 10a, 10b, 12a, 12b, 16a, 16b, 19a, 19b, 22a, 22b, 23a and 23b) protocol:
To a solution (0° C) of 1ꢀ(3ꢀdimethylaminopropyl)ꢀ3ꢀethylcarbodiimide hydrochloride (1.3 mmol),
1ꢀhydroxybenzotriazole (1.3 mmol), cycloꢀZaa(XaaꢀYaaꢀVal)ꢀOH (1 mmol) in N,Nꢀ
dimethylformamide (10 mL) added N,Nꢀdiisopropylethylamine (3 mmol) was stirred for 15 min.
Then added solution of amino propyl linker (1 mmol) in N,Nꢀdimethylformamide (2 mL) flask and
the reaction mixture was stirred at room temperature till the starting material was consumed
completely. N,Nꢀdimethylformamide and dichloromethane were removed under reduced pressure
and the resulting viscous solution was diluted with water (5 mL) and thoroughly extracted with ethyl
acetate (15 mL). The combined organic extracts were washed with 1 N HCl (5 mL), saturated
aqueous sodium bicarbonate (NaHCO ) (5 mL) and dried over anhydrous sodium sulphate (Na SO )
3
2
4
and concentrated to give a residue, which was purified by purified by HPLC.
General procedure for LiOH base-mediated hydrolysis of methyl ester ( 13a, 13b, 24a and
24b):
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To a stirred solution of lithium hydroxide (1.5 mmol) in of distill water (1 mL) was added to the
methyl ester (1 mmol) in MeOH (3 mL) and then stirred at room temperature for 3.5 h, then MeOH
o
was removed under reduced pressure, acidified with 1N HCl until pH was between 2ꢀ3 at 0 C. Then
extracted with EtOAc, dried over anhydrous sodium sulphate (Na SO ) and concentrated to give the
2
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
desired acid, which was used further without purification.
General procedure for synthesis of hydroxamic acids from corresponding methyl esters (14a,
4b, 25a and 25b):
1
To a stirred solution of methyl ester (1 mmol) in MeOH (3 mL), hydroxylamine hydrochloride (10
mmol) was added followed by 0.1 N potassium hydroxide (12 mmol) in distilled water (1 mL) and
then stirred at room temperature for 30 minutes, then MeOH was removed under reduced pressure,
o
acidified with 1N HCl until pH was between 2ꢀ3 at 0 C. The solution was extracted with EtOAc,
2 4
dried over anhydrous sodium sulphate (Na SO ) and concentrated to give the desired hydroxamic
acid in high yield, which was further purified by HPLC.
General Procedure for Sulfoxide synthesis from Organic Sulfides by H
ans 20b):
2 2
O /Borax (8a, 8b, 20a
In a typical experiment to a 25 mL flask equipped with a magnetic stirrer and 30% H
2 2
O
(6.0 mmol) in water was added borax (0.2 mmol) and MeOH (5 mL) followed by methyl thioether
containing peptide (1 mmol ). The reaction was monitored by TLC. After complete disappearance of
the reactant added Na S O to destroy the excess amount of H O and then removed MeOH, the
2
2
5
2
2
product was extracted with EtOAc, dried over anhydrous sodium sulphate (Na
concentrated to give a desired sulfoxide, which was purified by reverse phase HPLC.
2 4
SO ) and
2 2
General Procedure for Sulfones synthesis from Organic Sulfides by H O /Borax (9a, 9b, 21a
ans 21b):
28
ACS Paragon Plus Environment

Page 29 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In a typical experiment to a 25ꢀmL flask equipped with a magnetic stirrer and 30% H
2 2
O
(6.0 mmol) was added borax (0.2 mmol) and MeOH (5 mL) followed by methyl thioether containing
peptide (1 mmol ). To the resulting solution was added 0.1 N NaOH to maintain the pH of the
solution at 10, and the mixture was stirred at room temperature for 12 ꢀ15 h and monitored by LCꢀ
MS. After complete disappearance of the reactant added Na S O to destroy the excess amount of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
5
2 2
H O and then removed MeOH, the product was extracted with EtOAc, dried over anhydrous sodium
2 4
sulphate (Na SO ) and concentrated to give a desired sulfone, which was purified by reverse phase
HPLC.
General procedure for synthesis of sulfhydryl from corresponding homodimer (5a, 5b, 6a, 6b,
17a, 17a, 18a and 18b):
To a stirred solution of disulfide containg homodimer of SLA analog (1 mmol) in DMSOꢀwater
(1:1) (4 mL), Dithiothreitol (DTT) (20 mmol) was added followed by 0.1 N sodium hydroxide (0.01
mL) in distilled water and then stirred at room temperature for 2 hours. Then which was further
purified by HPLC using 10ꢀ40 acetonitrile gradient over 30 minutes.
D
Cyclo-Asp( Pro-Pro-Val)-OBn (1a):
Cyclic tetrapeptide 1a was synthesized by following the above general procedure for DPPA
macrolactamisation and purified by silical gel column chromatography (EtOAc) as a white solid
1
(
yield: 79%, purity by LCꢀMS: 99%). H NMR (400 MHz, CDCl
3
) δ ppm: 7.50 ꢀ 7.41 (m, 2 H), 7.41
ꢀ
7.28 (m, 3 H), 7.03 (d, J = 9.4 Hz, 1 H), 6.48 (d, J = 9.8 Hz, 1 H), 5.28 (d, J = 11.7 Hz, 1 H), 5.19
(d, J = 12.1 Hz, 1 H), 5.12 (t, J = 9.6 Hz, 1 H), 4.58 (dd, J = 3.6, 9.0 Hz, 1 H), 4.56 (t, J = 7.0 Hz, 1
H), 4.45 (dd, J = 3.9, 9.0 Hz, 1 H), 4.18 ꢀ 4.09 (m, 1 H), 3.63 ꢀ 3.54 (m, 1 H), 3.31 ꢀ 3.22 (m, 1 H),
3.12 (d, J = 12.9 Hz, 1 H), 3.01 (dt, J = 5.7, 9.7 Hz, 1 H), 2.79 (dd, J = 9.8, 12.5 Hz, 1 H), 2.53 (dt, J
=
3.3, 6.7 Hz, 1 H), 2.41 ꢀ 2.29 (m, 1 H), 2.17 ꢀ 2.02 (m, 4 H), 1.93 ꢀ 1.79 (m, 2 H), 1.76 ꢀ 1.63 (m, 1
29
ACS Paragon Plus Environment