Structure-Based Design and Biological Characterization of Selective Histone Deacetylase 8 (HDAC8) Inhibitors with Anti-Neuroblastoma Activity

Article abstract of DOI:10.1021/acs.jmedchem.7b01447

Histone deacetylases (HDACs) are important modulators of epigenetic gene regulation and additionally control the activity of non-histone protein substrates. While for HDACs 1-3 and 6 many potent selective inhibitors have been obtained, for other subtypes much less is known on selective inhibitors and the consequences of their inhibition. The present report describes the development of substituted benzhydroxamic acids as potent and selective HDAC8 inhibitors. Docking studies using available crystal structures have been used for structure-based optimization of this series of compounds. Within this study, we have investigated the role of HDAC8 in the proliferation of cancer cells and optimized hits for potency and selectivity, both in vitro and in cell culture. The combination of structure-based design, synthesis, and in vitro screening to cellular testing resulted in potent and selective HDAC8 inhibitors that showed anti-neuroblastoma activity in cellular testing.

Full text of DOI:10.1021/acs.jmedchem.7b01447

Subscriber access provided by Gothenburg University Library
Article
Structure-based design and biological characterization of selective
histone deacetylase 8 (HDAC8) inhibitors with anti-neuroblastoma activity
Tino Heimburg, Fiona R. Kolbinger, Patrik Zeyen, Ehab Ghazy, Daniel Herp, Karin
Schmidtkunz, Jelena Melesina, Tajith Baba Shaik, Frank Erdmann, Matthias Schmidt,
Christophe Romier, Dina Robaa, Olaf Witt, Ina Oehme, Manfred Jung, and Wolfgang Sippl
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01447 • Publication Date (Web): 30 Nov 2017
Downloaded from http://pubs.acs.org on November 30, 2017
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 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structureꢀbased design and biological
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
characterization of selective histone deacetylase 8
(HDAC8) inhibitors with antiꢀneuroblastoma activity
1
2,3
1
1
4
Tino Heimburg , Fiona R. Kolbinger , Patrik Zeyen , Ehab Ghazy , Daniel Herp , Karin
4
1
5
1
1
Schmidtkunz , Jelena Melesina , Tajith Baba Shaik , Frank Erdmann , Matthias Schmidt ,
5
1
2,3,6
2,3
4
Christophe Romier , Dina Robaa , Olaf Witt , Ina Oehme , Manfred Jung , and Wolfgang
1
Sippl *
1
Institute of Pharmacy, MartinꢀLuther University of HalleꢀWittenberg, 06120 Halle/Saale,
Germany
2
Clinical Cooperation Unit Pediatric Oncology, German Cancer Research Center (DKFZ), INF
280, Dꢀ69120 Heidelberg, Germany and German Consortium for Translational Cancer Research
(DKTK)
3
Translational Program, Hopp Children’s Cancer Center at NCT Heidelberg (KiTZ)
Institute of Pharmaceutical Sciences, University of Freiburg, 79104 Freiburg, Germany
Département de Biologie Structurale Intégrative, Institut de Génétique et Biologie Moléculaire
4
5
et Cellulaire (IGBMC), Université de Strasbourg (UDS), CNRS, INSERM, 67404 Illkirch Cedex,
France
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 55
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
6
Department of Pediatric Oncology, Hematology and Immunology, University of Heidelberg
Medical Center
Keywords : HDAC8, structureꢀbased design, neuroblastoma, benzhydroxamic acids
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
ACS Paragon Plus Environment

Page 3 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
Histone deacetylases (HDACs) are important modulators of epigenetic gene regulation and
additionally control the activity of nonꢀhistone protein substrates. While for HDACs 1ꢀ3 and 6
many potent selective inhibitors have been obtained, for other subtypes much less is known on
selective inhibitors and the consequences of their inhibition. The present report describes the
development of substituted benzhydroxamic acids as potent and selective HDAC8 inhibitors.
Docking studies using available crystal structures have been used for structureꢀbased
optimization of this series of compounds. Within this study, we have investigated the role of
HDAC8 in the proliferation of cancer cells and optimized hits for potency and selectivity, both in
vitro and in cell culture. The combination of structureꢀbased design, synthesis, in vitro screening
to cellular testing resulted in potent and selective HDAC8 inhibitors that showed antiꢀ
neuroblastoma activity in cellular testing.
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
INTRODUCTION
Neuroblastoma, the most common extracranial solid tumor occurring in childhood, arises from
abnormal growth of the neuronal crest progenitor cells of the devolving sympathetic nervous
system. A distinguishing mark of neuroblastoma is its highly heterogeneous clinical behavior,
which ranges from spontaneous tumor regression to highly aggressive, treatmentꢀresistant and
fatal tumor progression. Despite numerous advances and the implementation of intensive
multimodal therapies, the prognosis for patients with high risk disease, which represent almost
1
half of all neuroblastoma patients, still remains poor with less than 50% longꢀterm survival rates.
Further, high risk patients suffer from severe treatmentꢀrelated immediate and longꢀterm adverse
2ꢀ3
effects, and a significant number will still relapse. This emphasizes the urgent need for novel,
more effective, and less toxic therapeutic approaches to improve the survival and the lifeꢀquality
of neuroblastoma patients. Diverse genomic abnormalities have been described in neuroblastoma,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 55
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
yet no common genomic aberration has been uniformly identified in all high risk patients. One of
the most important genetic marks of neuroblastoma is the amplification of the protoꢀoncogene
4
ꢀ6
MYCN which is usually associated with tumor aggressiveness, resistance to therapy, and poor
3
, 7
survival.
Besides MYCN, ALK has also been implicated as another causative gene in
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
neuroblastoma; activation mutations of ALK occur in all cases of familial neuroblastoma and in
8ꢀ9
ca. 14% of sporadic cases.
Other recurrent genetic alterations include deletions on
1
0
chromosomes 1p and 11q or gains on 17q, which are all used as prognostic markers.
HDAC inhibitors are an emerging class of drugs which have shown promise in the treatment of a
11ꢀ13
wide variety of malignancies.
We have previously demonstrated that specifically the isozyme
HDAC8 is involved in the pathogenesis of neuroblastoma and that its expression significantly
1
4
correlates with advanced tumor stage and markers for poor prognosis.
HDAC8 is a 42 kDa protein composed of 377 amino acids, which is ubiquitously expressed and
located in the nucleus and in the cytoplasm, where it supports smooth muscle cell differentiation
1
5ꢀ16
and contractility by interaction with αꢀactin and modulators of cytoskeleton dynamics.
Histones are not considered to be bona fide HDAC8 substrates and acetylation patterns observed
17
in vitro and in cells regularly differ. While SMC3, a cohesin complex protein exerting its role
during the separation of sister chromatids, is directly deacetylated by HDAC8, the enzyme can
also interact with other proteins as a scaffold.
1
8
HDAC8 plays a significant role in regulating spindle assembly during mouse oocyte meiosis. In
oral squamous cell carcinoma cells, HDAC8 knockdown induced apoptotic cell death through
caspases activation and proꢀsurvival autophagy, consequently, the additional treatment with
1
9
autophagy inhibitors enhanced cell death.
HDAC8 has also been shown to be one of the causative genes in Cornelia de Lange syndrome
(CdLS), a rare genetic disease causing congenital malformations, and has been identified as
ACS Paragon Plus Environment

Page 5 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
0
SMC3 deacetylase. Chromatinꢀreleased acetylated SMC3 is deacetylated by the HDAC8
enzyme in early mitosis facilitating cohesin recycling for subsequent cell cycles. There have been
2
1
several lossꢀofꢀfunction mutations reported for HDAC8, that result in a strong loss of activity.
Therefore, specific HDAC8 activators, such as Nꢀ(phenylcarbamothioyl)benzamide, have been
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
2
2
reported as useful leads in the search for new therapeutic strategies in managing CdLS.
In INSS stage 4 neuroblastoma patient samples, HDAC8 is the only HDAC isoform showing
significant transcriptional upregulation as compared to stages 1, 2, 3 and 4S. High HDAC8 levels
were found to be correlated with predictors of poor prognosis in neuroblastoma such as
14
unfavorable Shimada histopathologic classification and loss of chromosomes 1p and 11q. Zhao
et al. revealed that targeted inhibition of HDAC8 by subtypeꢀselective HDAC8 inhibitors such as
2
3
PCIꢀ34051 (1, Figure 1) can suppress the growth of neuroblastoma cells and increase
24
Doxorubicin sensitivity via upregulation of miRꢀ137 and suppression of MDR1. The pivotal
role of this class I HDAC in differentiation and tumor progression has been confirmed in
2
5
malignant peripheral nerve sheath tumors and neuroblastoma, where selective HDAC8
inhibition induces a differentiated phenotype, enhances retinoic acidꢀmediated effects, and
reduces tumor growth in cell culture and in vivo with decreased adverse effects as compared to
26
unspecific HDAC inhibition. This endorses selective HDAC8 inhibition as a very promising
therapeutic option for neuroblastoma therapy.
The FDAꢀapproved broadꢀspectrum HDAC inhibitor vorinostat (SAHA), which is used to treat
refractory cutaneous Tꢀcell lymphoma, is a weaker inhibitor of HDAC8 (micromolar range) than
of HDACs 1–3 (nanomolar range). Selective HDAC8 inhibitors were identified by either
structureꢀbased design, for example, linkerless hydroxamic acids (IC50 0.3 ꢁM, >100ꢀfold
2
7
selectivity over HDAC1 and 6)
or from highꢀthroughput screens like the cyclic thiourea
2
8
29
SB379278A (IC 0.5 ꢁM) (Figure 1). Compound 1 (Figure 1) is probably the most widely
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 55
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
used HDAC8ꢀspecific inhibitor in research. Based on an indole linker, this hydroxamic acid
23
showed good HDAC8 selectivity and cellular activity. NCC149 (Figure 1) and further analogs
were identified using a rapid screen of a 151ꢀtriazole compound library generated using click
3
0ꢀ31
chemistry.
In vitro tests revealed an IC50 of 70 nM against HDAC8 with a very high
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
selectivity over major HDACs (HDAC1: IC 38 ꢁM, HDAC2 >100 ꢁM, HDAC4: IC : 44 ꢁM,
5
0
50
HDAC6: IC50 2.4 ꢁM). Huang et al. reported orthoꢀarylꢀNꢀhydroxycinnamides as further
3
2
selective HDAC8 inhibitors. The biphenyl substituted orthoꢀarylꢀNꢀhydroxycinnamide (Figure
) showed an IC value of 27 nM for HDAC8, good selectivity over other HDACs and moderate
1
5
0
growth inhibition of lung cancer cell lines. Tang et al. reported on the synthesis of a small library
3
3
of benzhydroxamic acids containing a hydrazide linker. The most promising compound A8B4
(Figure 1) showed an IC of 23 nM and good selectivity over HDAC2 and HDAC3. However,
5
0
no cellular data was reported. Very recently, Ingham et al. developed a highly potent HDAC8
inhibitor with good selectivity over HDAC1 and HDAC6. The triazole hydroxamic acid OJIꢀ1
34
(Figure 1) showed an IC of 0.8 nM. However, no cellular data has been provided so far.
5
0
3
5
Besides hydroxamic acids, inhibitors containing an amino acid group and an azetidinone
36
group were reported as selective HDAC8 inhibitors (Figure 1). Another potent, nonꢀ
3
7
hydroxamic acid HDAC8 inhibitor, that has been reported, is βꢀphenyltropolone. The betaꢀ
isopropyltropolone derivative showed a Ki of 8 nM and about 1400ꢀfold selectivity toward
HDAC8 than HDAC4. Synthetic as well as structureꢀbased approaches to develop
3
8ꢀ39
benzhydroxamic acids as HDAC8 inhibitors have been recently reported and reviewed.
4
0
Based on our previous study to develop inhibitors of Schistosoma mansoni HDAC8, we
identified the benzhydroxamic acids 2, 3, and 4 (Figure 2; IC values are listed in Table 1) as
5
0
promising starting points for the development of inhibitors of the homologous human HDAC8.
Our previous results showed that the lead benzamidoꢀhydroxamic acid 2 displays good inhibition
ACS Paragon Plus Environment

Page 7 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of hHDAC8, high selectivity over HDAC1 and a moderate selectivity over HDAC6. Modifying
the capping group to quinoline and introducing a methyl substituent in the paraꢀposition of the
benzhydroxamic acid moiety (3) resulted in a significant improvement in HDAC8 inhibition but
was coupled with a loss of selectivity towards HDAC6. Meanwhile, switching from an amideꢀ to
an amineꢀlinker (4) resulted in a considerable enhancement of HDAC8 inhibition, but
simultaneously a decreased selectivity against HDAC1 was observed as compared to 2.
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
On the basis of our initial data, we wanted to get insights into the structureꢀactivity relationships
(SARs) of HDAC8 inhibitors with a metaꢀsubstituted benzhydroxamic acid structure. Here we
report that changes of the linker attached to the benzhydroxamic acid have a huge impact on
HDAC8 potency and selectivity (Figure 3). In addition, antiꢀneuroblastoma activity of the most
promising inhibitors was verified by viability and colony formation assays as well as detection of
neuriteꢀlike outgrowths and upregulation of differentiation markers, such as p21/CDKN1,
TrkA/NTRK1 and tyrosine hydroxylase (TH).
CHEMISTRY
In order to obtain the described 3ꢀaminobenzhydroxamic acids (8a-s), both nonꢀsubstituted and
4
ꢀsubstitutedꢀ3ꢀaminobenzoic acids (5a-s) were used as starting points. Alkyl and aryl residues
were introduced on the aromatic NH group via reductive amination, which was performed by the
2
reaction of the 3ꢀaminobenzoic acids 5a-s with the respective aldehyde using sodium
40
triacetoxyborohydride as a reducing agent. To avoid byproducts in the following synthetic
steps, the secondary amine 6 was protected by tertꢀbutyloxycarbonyl (Boc) group initially
(Scheme 1). However, we recognized in further trials that this step could be skipped because the
byproducts were only obtained in small amounts and could be neglected. The Nꢀphenethyl
substituted derivatives 6l and 6s were alternatively attained through Nꢀalkylation since the
products of the reductive amination using phenylacetaldehyde and sodium triacetoxyborohydride
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 55
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
showed poor purity. The desired hydroxamic acids 8a-s were finally obtained using PyBOP as an
activating agent and Oꢀ(tetrahydroꢀ2Hꢀpyranꢀ2ꢀyl)hydroxylamine followed by cleavage of the
protecting group (Scheme 1). Generally, using tetrahydropyran (THP)ꢀprotected hydroxylamine
increased the yields of the desired benzhydroxamic acids compared to other methods using
hydroxylamine hydrochloride and KOH or potassium methanolate.
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
The desired phenylcarbamoylꢀbenzhydroxamic acids 12a-d could be obtained starting from the
methoxycarbonylbenzoic acid derivatives 9a-d, which were first activated using thionyl chloride
or oxalyl chloride and subsequently coupled with the respective aniline to yield the
corresponding amides 10a-d. Alkaline hydrolysis of the methyl ester to the respective 3ꢀ
substituted carbamoylbenzoic acid 11a-d was followed by the conversion to the hydroxamic
acids 12a-d using the aforementioned protocol. While the 3ꢀphenylꢀ4ꢀcarbamoylbenzoic acid 11a
could be directly prepared from the commercially available 3ꢀmethoxycarbonylbenzoic acid 9a
using thionyl chloride and coupling with aniline, the starting materials 5ꢀmethoxycarbonylꢀ2ꢀ
substitutedꢀbenzoic acids 9b-d had to first be prepared as described in literature (see Scheme S1
4
1
Supporting Information). Activation of the three acids 9b-d was achieved using oxalyl chloride
as a mild and efficient activating agent. The activated acid chlorides were then coupled with 4ꢀ
chloroaniline using DIPEA as a base to form the respective amides 10b-d (Scheme 2).
The 3ꢀbenzyloxyꢀ4ꢀmethoxybenzhydroxamic acid 20a was directly synthesized from the
commercially available carboxylic acid 25 using the aforementioned procedure (Scheme 3). To
obtain 20b-g, the acid derivatives 21 were first converted to the methyl esters 22 to avoid
byproducts in the following steps. OꢀBenzylation to the ether derivatives 24b-g was performed
through Williamsonꢀetherꢀsynthesis using K CO and the respective substituted benzyl bromides
2
3
4
2
2
3. The corresponding 3ꢀbenzyloxyꢀbenzhydroxamic acids 20b-g were synthesized by cleavage
ACS Paragon Plus Environment

Page 9 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the methyl ester followed by the conversion of the carboxylic acid as previously described
(Scheme 3).
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 1. Chemical structures of HDAC8 inhibitors.
Figure 2. Previously identified benzhydroxamic acidꢀbased HDAC8 inhibitors.
Figure 3. Series of the herein described benzhydroxamic acids as HDAC8 inhibitors.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 55
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 1.
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) i. aldehyde, toluene; ii. Na(AcO) BH, AcOH; (b) Boc O, MeOH, tꢀBuOH; (c) i. PyBOP,
3
2
DIPEA, H NOTHP, THF; ii. cat. HCl, THF; (7a,b) iii. TFA, CHCl3
2
Scheme 2.
(
a) SOCl , aniline; (b) (COCl) , 4ꢀchloroaniline, cat. DMF, DIPEA, CH Cl (c) aq. NaOH,
2 2 2 2
MeOH; (d) i. PyBOP, DIPEA, H NOTHP, THF; ii. cat. HCl, THF
2
Scheme 3.
ACS Paragon Plus Environment

Page 11 of 55
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
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) SOCl , MeOH; (b) K CO , DMF; (c) aq. NaOH, MeOH; (d) PyBOP, DIPEA, H NOTHP,
2
2
3
2
THF; cat. HCl, THF
STRUCTUREꢀACTIVITY RELATIONSHIPS
Initial optimization studies focused on the linker between the benzhydroxamic acid and the
terminal aromatic ring used as capping group. Our previously reported lead benzamidoꢀ
40
hydroxamic acid 2 showed a moderate HDAC8 inhibitory activity as compared to the selective
HDAC8 inhibitor 1 (Table 1) and exhibited a good selectivity profile against HDAC1 and to a
lesser extent HDAC6. As previously described, switching from an amideꢀ to an amineꢀlinker
(compound 4, Table 1) resulted in a slight improvement in the HDAC8 inhibitory potency, a
considerable loss in the selectivity over HDAC1 and a slight increase in the selectivity over
HDAC6.
In order to guide further optimization efforts, docking studies were performed using the lead
43
structures in available crystal structures of HDAC8 (PDB ID: 2V5X ), HDAC1 (PDB ID:
4
4
45
46
4
BKX ) and HDAC6 (PDB ID: 5EDU and 5G0I ), in an attempt to detect the structural
aspects determinant for achieving high HDAC8 inhibitory potency as well as selectivity over
HDAC1 and ꢀ6. As could be expected, the benzhydroxamic acid group of all lead compounds
showed a similar interaction pattern in the catalytic pocket of HDAC8: The hydroxamic acid
group is able to chelate the zinc ion in a bidentate fashion and undergo three Hꢀbond interactions
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 55
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
with the side chains of His142, His143 and Tyr306. Meanwhile the benzene ring is sandwiched
between Phe152 and His180, where it exhibits πꢀπ stacking interactions. The capping group of
the different leads showed, however, more diverse interactions, which might give a primary
explanation for the differences in the HDAC8 inhibitory activity. The benzyl moiety of 1 exhibits
πꢀπ stacking interactions with Tyr306 and Phe152, which are lost in the case of the less potent
benzamide derivative 2. The corresponding highly active quinoline derivative 3 seems to benefit
from additional hydrophobic interactions with Tyr306 and Pro273, while the 3ꢀ
aminobenzhydroxamic acid 4 shows, like 1, πꢀπ stacking interactions with Tyr306 and Phe152.
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 4)
The selectivity of the compounds over HDAC1 could be easily explained by the docking results
(see Supporting Information Figure S1) which clearly show that the compounds are not able to
properly chelate the catalytic zinc ion. HDAC1 is characterized by its narrow pocket as compared
to HDAC8 and ꢀ6, and, hence, metaꢀsubstituted benzhydroxamic acids generally show a
3
1, 40,
HDAC6/8 over HDAC1 selectivity as has been frequently demonstrated in previous studies.
4
7ꢀ49
Meanwhile, the docking results cannot completely rationalize the selectivity of the compounds
for HDAC8 over HDAC6. Crystal structures of HDAC6 and HDAC8 clearly demonstrate that the
substrate binding grooves in both isoforms show a significantly different shape, with one of the
main differences in HDAC6 being the presence of an additional loop (L1ꢀloop: Asp460ꢀPro464,
DrHDAC6 numbering). While in HDAC8 the benzyl capping group of 1 is placed perpendicular
to Tyr306 and Phe152 allowing for a perfect πꢀπ stacking interaction, the additional loop in
HDAC6 hinders the ligand from binding to this hydrophobic subpocket. In the case of 1, only a
hydrophobic interaction with Leu712 of HDAC6 can be observed. The benzamidoꢀhydroxamic
acids 2 and 3 seem to benefit from additional waterꢀmediated Hꢀbond interactions with the
ACS Paragon Plus Environment

Page 13 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conserved water molecule at the rim of the catalytic tunnel. Interestingly, the quinoline group of 3
is able to undergo additional hydrophobic interactions which extend to Pro464 of the L1ꢀloop,
which might explain the observed increase in HDAC6 inhibitory potency. Meanwhile, due to the
flexibility of the aminomethyl linker of compound 4, the capping benzyl group can undergo πꢀπ
stacking interactions with Phe583 (see Supporting Information Figure S2).
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
In order to obtain more potent and selective HDAC8 inhibitors, we started with the derivatization
of the 3ꢀaminoꢀbenzhydroxamic acid 4 by introducing different substituents at the 4ꢀposition of
the benzhydroxamic acid core and modifying the capping groups to maximize the interactions in
the previously described hydrophobic side pocket of HDAC8.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 55
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 4. Predicted binding modes in HDAC8: a) the reference inhibitor 1, b) the benzamidoꢀ
hydroxamic acid 2, c) the corresponding quinoline derivative 3, and d) the 3ꢀ
aminobenzhydroxamic acid derivative 4. Side chains of active site residues are shown as white
sticks and the ligands as yellow sticks. The zinc ion is depicted as cyan ball, while cocrystallized
water molecules are shown as red spheres. Hydrogen bonds are shown as yellow dashed lines and
coordination of the zinc ion as paleꢀcyan dashed lines.
ACS Paragon Plus Environment

Page 15 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Regarding the 3ꢀaminoꢀbenzhydroxamic acid series (8a-s), the introduction of a hydrophobic
substituent in the paraꢀposition of the benzhydroxamic acid moiety generally resulted in a
decreased inhibitory activity against HDAC1 as compared to the unsubstituted derivative 4,
however, had little impact on the HDAC6 activities (Table 2). Only compounds with a methoxy
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
(8b) and a fluoro substituent (8o) showed increased HDAC8 inhibitory activity and consequently
higher selectivity over HDAC1 and ꢀ6, as compared to the respective 4ꢀmethyl and the
unsubstituted derivative (8a and 4). Docking of 8b to HDAC8 shows that it adopts a highly
similar binding conformation as that predicted for the unsubstituted 3ꢀaminobenzhydroxamic acid
4
, where the hydroxamic acid group chelates the zinc ion and forms three Hꢀbond interactions
with the conserved His and Tyr residues, the core benzene moiety is sandwiched between Phe152
and His180 and the capping benzyl group exhibits πꢀπ stacking interactions with Tyr306 and
Phe152. Additionally, the 4ꢀmethoxy group can undergo hydrophobic interactions with Phe208
(Figure 5a). Meanwhile, the 4ꢀfluoro derivative 8o and the 4ꢀchloro derivative 8m also showed a
similar binding mode to 4 (not shown). Whereas the fluoro substitution (8o) resulted in a similar
inhibitory activity, the chloro substitution (8m) was found to be less favourable. However, the
obtained docking results offer no explanation for the differences in HDAC8 inhibition. Similarly,
increasing the length of the linker to ethylene and propylene moieties (8k, 8l, 8r and 8s) did not
result in an increase in the inhibitory activity against HDAC8 or an enhancement of the
compounds’ selectivity. With respect to the capping groups, most of the probed moieties were
well tolerated by HDAC8, even bulkier groups like biphenyl (8p) and naphthyl (8q). A
significant enhancement in the inhibitory activity against HDAC8, together with an improvement
in the selectivity over HDAC1 and HDAC6, was only observed for the 2,4ꢀdichloro derivative 8f.
This might be explained by the dichlorophenyl capping group undergoing additional vdW
interaction with the hydrophobic side pocket of HDAC8 (Figure 5b). Among this series, the 4ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 55
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
methoxyꢀbenzyl derivative 8b and the 4ꢀmethylꢀ2,4ꢀdichloro derivative 8f displayed the highest
potency against HDAC8 (IC 69 nM and 35 nM; respectively) and simultaneously showed the
5
0
best selectivity over HDAC1 (210 and 534 fold HDAC8/HDAC1; respectively) and HDAC6 (74
and 411 HDAC8/HDAC6; respectively).
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 5. Predicted binding mode of the most potent HDAC8 inhibitors: a) the 4ꢀmethoxyꢀ3ꢀ
benzylaminoꢀbenzhydroxamic acid 8b, b) the 2,4ꢀdichlorobenzylaminoꢀbenzhydroxamic acid
derivative 8f, c) the inverted amide derivative 12b, and d) the 3ꢀbenzyloxyꢀbenzhydroxamic acid
ACS Paragon Plus Environment

Page 17 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
derivative 20a in human HDAC8. Ligands are colored in yellow, side chains of active site
residues are shown as white sticks, the zinc ion is depicted as cyan ball, while the cocrystallized
water molecule is shown as red balls. Hydrogen bonds are shown as yellow dashed lines and
coordination of the zinc ion as paleꢀcyan dashed lines.
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
Another series of inhibitors that we prepared was obtained by inverting the amide linker of 2. The
49
inverted amide 12a (which has already been reported) showed about four times increased
HDAC8 inhibitory activity compared to 2 (Table 3). Adding a methoxy group at the paraꢀ
position of the benzhydroxamic acid and modifying the terminal phenyl ring with a paraꢀchloro
substituent resulted in a highly potent HDAC8 inhibitor (12b, IC 38 nM), but was coupled with
5
0
an increase in the inhibition of HDAC1 and ꢀ6. Docking studies revealed that the inverted amide
is able to benefit from an additional waterꢀmediated Hꢀbond interaction which is missing in the
analogous benzamidoꢀhydroxamic acid 2 (Figure 5c). Further modifications did not result in an
improvement in the potency against HDAC8. Due to the strong HDAC6 inhibition, we did not
further study the most active compound in this series (12b; HDAC6: 59% inhibition at 1ꢀM), in
the cellular testing.
The third linker that we tested was a methyleneoxy group (Table 4). As described for the 3ꢀ
aminobenzhydroxamic acids, the paraꢀposition of the benzhydroxamic acid scaffold was
modified and hydrophobic substituents at the terminal benzyl group were used in an attempt to
increase the potency of the compounds. The 4ꢀmethoxy substituted inhibitor 20a showed the
strongest HDAC8 inhibition among all tested compounds with good selectivity indices (448 fold
against HDAC1, 107 fold against HDAC6). Our docking studies revealed that the ether linker in
2
0a is able to undergo waterꢀmediated Hꢀbond interaction with the water molecule at the rim of
the catalytic tunnel of HDAC8, while the capping benzyl group shows πꢀπ stacking with Phe152
and Tyr306 (Figure 5d). Derivatives bearing a 4ꢀchloro substituent (20e-g) showed a decreased
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 55
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
HDAC8 inhibition as compared to their 4ꢀmethoxy counterparts. Meanwhile, introducing chloroꢀ
substituents at the capping benzyl group generally resulted in a significant loss in the HDAC8
inhibitory activity. Interestingly, a 2,4ꢀdichloro substitution of 20d and 20g resulted in a
pronounced decrease in HDAC8 inhibitory potency, which is in contrast to our previous
observation where a similar substitution of the 3ꢀbenzylaminoꢀbenzhydroxamic acid scaffold (8f)
led to an increase in HDAC8 inhibitory activity and improved selectivity.
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
Among all herein described compounds, the 3ꢀbenzylaminoꢀbenzhydroxamic acids 8b, 8f, and
the 3ꢀbenzyloxyꢀbenzhydroxamic acid 20a were the most potent and selective compounds in
comparison with 1 as reference and were, hence, selected for further biological/cellular testing in
comparison with reference inhibitors.
ACS Paragon Plus Environment

Page 19 of 55
Journal of Medicinal Chemistry
1
2
3
4
0
Table 1: In vitro IC values of reference compounds (values of 2, 3, and 4 were taken from Heimburg et al. ). (± Standard error, SI =
5
0
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
selectivity index).
Cpd. no.
Structure
hHDAC8
IC nM
hHDAC1
hHDAC6
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
5
0
IC µM
50
IC µM
50
H
N
N
OH
1
92 ± 15
28.3 ± 2.0 (SI 307)
48.2 ± 6.2 (SI 524)
O
O
O
H
N
2
3
4
N
H
OH
582 ± 48
33.6 ± 1.8 (SI 58)
2.7 ± 0.8 (SI 90)
2.3 ± 1.2 (SI 16)
3.0 ± 0.3 (SI 5)
O
O
H
N
N
OH 30 ± 7
0.089 ± 0.014 (SI 3)
2.5 ± 1.1 (SI 18)
N
H
O
H
N
143 ± 7
N
H
OH
O
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 55
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
Table 2. In vitro IC values of 3ꢀaminobenzhydroxamic acid derivatives. (± Standard error, SI = selectivity index).
5
0
X
H
N
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
Y
N
H
OH
8
O
X
Y
hHDAC8
IC nM
hHDAC1
hHDAC1
hHDAC6
hHDAC6
Cpd. no.
% inhibition
IC µM
50
% inhibition
IC µM
50
5
0
3
9% @ 25 µM
69% @ 25 µM
35% @ 5 µM
8
a
methyl
benzyl
benzyl
260 ± 36
69 ± 17
21.8 ± 2.1 (SI 84)
14.5 ± 1.4 (SI 210)
5.1 ± 0.3 (SI 20)
5.1 ± 0.8 (SI 74)
1
6% @ 5 µM
7% @ 100 µM
9
70% @ 10 µM
12% @ 1 µM.
8
8
b
c
methoxy
4
9% @ 10 µM
methyl
methyl
2ꢀchlorobenzyl
4ꢀchlorobenzyl
286 ± 41
161 ± 21
n.d
n.d
n.d.
n.d.
n.d
n.d
n.d.
n.d.
8d
8e
3
2% @ 10 µM
61% @ 10 µM
11 % @ 1 µM
methyl
2,6ꢀdichlorobenzyl
335 ± 55
n.d.
n.d.
3% @ 1 µM
6
6% @ 50 µM
75% @ 50 µM
1.6% @ 1 µM
8
8
f
methyl
2,4ꢀdichlorobenzyl
3ꢀmethoxybenzyl
35 ± 4
18.7 ± 2.5 (SI 534)
n.d.
14.4 ± 2.4 (SI 411)
n.d.
1
6% @ 1 µM
g
methoxy
167 ± 59
31% @ 10 µM
38% @ 10 µM
ACS Paragon Plus Environment

Page 21 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
3
.5% @ 1 µM
2% @ 10 µM
0 % @ 1 µM
5% @ 10 µM
% @ 1 µM
1% @ 10 µM
4 % @ 1 µM
n.i. % @ 1 µM
3
63% @ 10 µM
11% @ 1 µM
8
8
8
8
8
8
h
methoxy
methoxy
methoxy
methoxy
methyl
chloro
3ꢀphenoxybenzyl
2ꢀquinolinylmethyl
4ꢀchlorobenzyl
202 ± 22
481 ± 93
147 ± 18
240 ± 22
130 ± 20
254 ± 41
136 ± 17
72 ± 12
n.d.
n.d.
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
1
4
5
73% @ 10 µM
19% @ 1 µM
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
i
n.d.
n.d.
8
59% @ 10 µM
11% @ 1 µM
j
n.d.
n.d.
1
1
ꢀ(4ꢀchlorophenyl)ꢀ
ethyl
31% @ 10 µM
5 % @ 1 µM
46% @ 10 µM
5% @ 1 µM
k
l
n.d.
n.d.
3
4% @ 10 µM
63% @ 10 µM
20% @ 1 µM
2ꢀphenylethyl
benzyl
n.d.
10.4 ± 1.2 (SI 41)
n.d
n.d.
4.0 ± 0.2 (SI 16)
n.d
9
% @ 1 µM
4
8% @ 10µM
74% @ 10µM
22% @ 1 µM
m
3
% @ 1 µM
33% @ 10 µM
% @ 1 µM
62% @ 10 µM
n.i. @ 10 µM
8n
8o
8p
ethyl
benzyl
8
40% @ 10 µM
n.i. @ 1 µM
67% @ 10 µM
20%. @ 1 µM
fluoro
benzyl
25.0 ± 6.5 (SI 347)
5.0 ± 0.7 (SI 69)
43% @ 10 µM
5% @ 1 µM
58% @ 10µM
11% @ 1 µM
methyl
4ꢀphenylbenzyl
227 ± 54
n.d.
n.d.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 55
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
3
7% @ 10 µM
% @ 1 µM
52% @ 10 µM
11%. @ 1 µM
8
q
methyl
2ꢀnaphtylmethyl
119 ± 34
n.d.
n.d.
8
4
2
1% @ 10 µM
% @ 1 µM
9% @ 10 µM
% @ 1 µM
63% @ 10 µM
20% @ 1 µM
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
8
8
r
s
methoxy
methyl
2ꢀphenylethyl
151 ± 33
248 ± 52
19.1 ± 3.0 (SI 126)
25.6 ± 5.5 (SI 103)
7.5 ± 2.8 (SI 50)
8.9 ± 1.7 (SI 36)
0
55% @ 10 µM
42% @ 1 µM
3ꢀphenylpropyl
3
ACS Paragon Plus Environment

Page 23 of 55
Journal of Medicinal Chemistry
1
2
3
Table 3: In vitro IC values of 3ꢀ(phenylcarbamoyl)benzhydroxamic acid derivatives. (± Standard error, SI = selectivity index)..
5
0
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
hHDAC8
IC nM
hHDAC1
hHDAC1
hHDAC6
hHDAC6
Cpd. no.
Structure
% inhibition
IC µM
50
% inhibition
IC µM
50
5
0
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
120 ± 37
10% @ 5 µM
3% @ 0.5 µM
80% @ 5 µM
H
N
H
N
12a*
39.1 ± 8.9 (SI 326)
3.1 ± 0.1 (SI 25)
OH
4
9
O
O
(110 ± 47)
27% @ 0.5 µM
O
72% @ 10 µM
24 % @ 1 µM
93% @ 10 µM
59 % @ 1 µM
H
N
H
N
12b
38 ± 2
n.d.
n.d.
OH
O
O
Cl
Cl
H
N
H
N
26% @ 10 µM
0.1 % @ 1 µM
78% @ 10 µM
25 % @ 1 µM
1
1
2c
OH
OH
175 ± 45
114 ± 14
n.d.
n.d.
n.d.
n.d.
O
O
O
Cl
Cl
H
N
H
N
15% @ 10 µM
1.8 % @ 1 µM
69% @ 10 µM
18 % @ 1 µM
2d
O
4
9
*
IC value of compound 12a was previously reported by Olson et. al. Difference in the herein determined IC value to the reported
50 50
one might be due to the use of different substrates.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 55
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
Table 4. In vitro IC values of 4ꢀbenzyloxyꢀ3ꢀsubstitutedꢀbenzhydroxamic acid derivatives. (± Standard error, SI = selectivity index).
5
0
hHDAC8
IC nM
hHDAC1
hHDAC1
hHDAC6
hHDAC6
Cpd. no.
Structure
% inhibition
IC µM
50
% inhibition
IC µM
50
5
0
O
O
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
62% @ 25 µM
34% @ 5 µM
73% @ 25 µM
53% @ 5 µM
H
N
20a
27 ± 3
12.1 ± 5.7 (SI 448)
2.9 ± 0.3 (SI 107)
OH
O
O
45% @ 10 µM
8.7% @ 1 µM
78% @ 10 µM
32% @ 1 µM
H
N
20b
O
OH
181 ± 26
120 ± 20
840 ± 120
1020 ± 100
2100 ± 445
14.5 ± 3.5 (SI 80)
4.6 ± 0.4 (SI 25)
O
Cl
2
2
2
2
0c
0d
0e
0f
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
ACS Paragon Plus Environment

Page 25 of 55
Journal of Medicinal Chemistry
1
2
3
4
2
0g
4150 ± 700
n.d.
n.d.
n.d.
n.d.
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
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 55
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
Table 5. Cytotoxicity studies in HEK293 cells.
Cpd.
no.
Viability*
(%)
Cpd.
no.
Viability*
Cpd.
no.
Viability*
Cpd.
no.
Viability*
(%)
(%)
(%)
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
2
3
4
72.0 ± 2.9
66.2 ± 0.7
n.d.
8h
8i
55.0 ± 2.4
84.8 ± 4.9
82.2 ± 1.3
78.3 ± 2.6
n.d.
8r
92.0 ± 4.1
80.9 ± 2.7
86.0 ± 3.2
73.2 ± 4.1
83.8 ± 5.7
79.7 ± 7.4
69.1 ± 1.1
62.8 ± 8.4
77.6 ± 14.4
70.1 ± 6.7
20e
20f
20g
63.6 ± 14.9
74.7 ± 1.4
65.4 ± 2.9
8s
8j
12a
12b
12c
12d
20a
20b
20c
20d
8
a
90.4 ± 1.7
68.1 ± 1.2
95.3 ± 2.7
82.9 ± 3.4
65.7 ± 3.6
78.9 ± 3.89
77.3 ± 0.3
8k
8l
8b
8
c
8m
8n
8o
8p
8q
87.4 ± 3.4
n.d.
8d
8
e
86.9 ± 6.3
77.0 ± 3.9
89.9 ± 3.0
8f
8g
* Viability of cells in presence of 50 µM compound in comparison to an untreated sample. As positive control
daunorubicin was used and an IC value of 12.55 ± 0.07 µM was determined.
5
0
n.d. not determined
ACS Paragon Plus Environment

Page 27 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CYTOTOXICITY ASSAY
The cytotoxicity of the compounds was tested against a human embryonic kidney cell line
HEK293). The cells were incubated with the compounds at a concentration of 50 µM for 45 h,
(
and the cell viability was determined using the Alamar Blue assay. All tested compounds showed
only a relatively weak cytotoxicity at the used concentration (Table 5).
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
CELLULAR ACTIVITY
To show cellular target engagement and selectivity, Western blotting experiments were
performed with a selection of compounds in neuroblastoma BE(2)ꢀC cells. The tumor cells were
incubated for 6 h with 10 µM of compound 3, 8a, 8b, 8f or 20a. Cell lysates were blotted against
AcꢀSMC3 (HDAC8 substrate), Acꢀαꢀtubulin (HDAC6 substrate) and Acꢀhistone H3 (HDAC1
substrate). As shown in Figure 6A, all compounds induced increase of both SMC3 and tubulin
acetylation compared to solvent (DMSO) control, thus they inhibit both HDAC6 and HDAC8 in
BE(2)ꢀC cells. In contrast, no significant effect on the acetylation level of histone H3 was
observed. This indicates that the observed in vitro selectivity is also detected in a cellular setting.
A summary and quantification of all Western Blot repetitions is displayed in the corresponding
bar diagram (Figure 6B).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 55
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 6. Cellular target engagement and selectivity. (A) Acetylation of SMC3, αꢀtubulin and
histone H3 is shown in Western blots of wholeꢀcell lysates 6 h after treatment of BE(2)ꢀC cells
with 10 µM of compound 3, 8a, 8b, 8f or 20a. HSC70, tubulin, histone H3, βꢀActin and Ponceau
staining were used as a loading control. (B) The normalized ratios from three independent
experiments are shown in the bar graphs on the right.
Anti-neuroblastoma phenotype
ACS Paragon Plus Environment

Page 29 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In order to measure the functional consequence of HDAC8 inhibition, the BE(2)ꢀC
neuroblastoma cell model was applied, which responds with growth arrest and signs of neuronal
26
differentiation upon knockdown or selective inhibition of HDAC8. Growth arrest was
determined by colony formation and viability assays. The induction of a more differentiated
phenotype was detected by analysis of marker gene expression (CDKN1A, TH and NTRK1).
Tyrosine hydroxylase (TH) and neurotrophin receptor kinase A (NTRK1/TrkA) are neuronal
marker genes. Tyrosine hydroxylase is encoded by the TH gene and is present in the central
nervous system (CNS), peripheral sympathetic neurons and the adrenal medulla. The enzyme
catalyzes the rate limiting step in this synthesis of catecholamines. Differentiation of
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
5
0
neuroblastoma cells is accompanied by an increase in tyrosine hydroxylase (TH) mRNA.
Activation of the neurotrophin receptor TrkA signaling pathway leads to neuronal differentiation
and high expression of NTRK1/TrkA is present in neuroblastomas with favorable biological
5
1
features and is highly correlated with patient survival. Differentiation is typically associated
WAF1/CIP1
with a cell cycle arrest and CDK inhibitor p21
(CDKN1A) is known to mediate cell cycle
arrest in response to pharmacological HDAC inhibition.
For colony formation assays, BE(2)ꢀC cells were treated with 10 µM of each of the selected
compound (3, 8a, 8b, 8f, 20a) for 96 h, followed by culturing for another 6 days without
treatment. This assesses whether the treatment impairs the clonogenic growth capacity of tumor
cells, indicating effectiveness of compounds on the survival and proliferation of tumor cells. All
compounds significantly decreased the ability to form colonies (Figure 7A). The WSTꢀ8
metabolic activity assay was applied to determine metabolic IC50 values in BE(2)ꢀC cells,
employing six different inhibitor concentrations per compound and six replicates per inhibitor
concentration.
ACS Paragon Plus Environment