Discovery of novel N-hydroxy-2-arylisoindoline-4-carboxamides as potent and selective inhibitors of HDAC11

Article abstract of DOI:10.1016/j.bmcl.2018.05.021

N-Hydroxy-2-arylisoindoline-4-carboxamides are potent and selective inhibitors of HDAC11. The discovery, synthesis, and structure activity relationships of this novel series of inhibitors are reported. An advanced analog (FT895) displays promising cellular activity and pharmacokinetic properties that make it a useful tool to study the biology of HDAC11 and its potential use as a therapeutic target for oncology and inflammation indications.

Full text of DOI:10.1016/j.bmcl.2018.05.021

Accepted Manuscript
Discovery of novel N-hydroxy-2-arylisoindoline-4-carboxamides as potent and
selective inhibitors of HDAC11
Matthew W. Martin, Jennifer Lee, David R. Lancia Jr, Pui Yee Ng, Bingsong
Han, Jennifer R. Thomason, Maureen S. Lynes, C. Gary Marshall, Chiara Conti,
Alan Collis, Monica Alvarez Morales, Kshama Doshi, Aleksandra Rudnitskaya,
Lili Yao, Xiaozhang Zheng
PII:
S0960-894X(18)30421-9
https://doi.org/10.1016/j.bmcl.2018.05.021
BMCL 25838
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
15 March 2018
2 May 2018
9 May 2018
Please cite this article as: Martin, M.W., Lee, J., Lancia, D.R. Jr, Yee Ng, P., Han, B., Thomason, J.R., Lynes, M.S.,
Marshall, C.G., Conti, C., Collis, A., Morales, M.A., Doshi, K., Rudnitskaya, A., Yao, L., Zheng, X., Discovery of
novel N-hydroxy-2-arylisoindoline-4-carboxamides as potent and selective inhibitors of HDAC11, Bioorganic &
Medicinal Chemistry Letters (2018), doi: https://doi.org/10.1016/j.bmcl.2018.05.021
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
Discovery of novel N-hydroxy-2-arylisoindoline-4-carboxamides as potent and
selective inhibitors of HDAC11
Matthew W. Martin,a, Jennifer Lee,a David R. Lancia, Jr.,a Pui Yee Ng,a Bingsong Han,a Jennifer R.
Thomason,a Maureen S. Lynes,a C. Gary Marshall,a Chiara Conti,a Alan Collis,a Monica Alvarez Morales,a
Kshama Doshi,a Aleksandra Rudnitskaya,a Lili Yao,b and Xiaozhang Zhenga
aFORMA Therapeutics, 500 Arsenal Street, Suite 100, Watertown, MA 02472, USA
bFORMA Therapeutics, 35 Northeast Industrial Road, Branford, CT 06405, USA
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
N-Hydroxy-2-arylisoindoline-4-carboxamides are potent and selective inhibitors of
HDAC11. The discovery, synthesis, and structure activity relationships of this novel
series of inhibitors are reported. An advanced analog (FT895) displays promising
cellular activity and pharmacokinetic properties that make it a useful tool to study the
biology of HDAC11 and its potential use as a therapeutic target for oncology and
inflammation indications.
Keywords:
HDAC11, HDACs, oncology, inflammation,
isoindoline, hydroxamic acid
2009 Elsevier Ltd. All rights reserved.
———
 Corresponding author. Tel.: +1-857-209-2357; fax: +1-617-679-1977; e-mail: mmartin@formatherapeutics.com

Acetylation of lysine residues is an important post-
impact was observed when the hydroxamic acid was substituted
at the 5-position. While compound 2 exhibited only modest
HDAC11 activity (IC50 =1.2 M), no significant HDAC6
potency was observed (IC50 > 10 M). Having identified a
preferred regiochemistry for the hydroxamic acid, optimization
of the tetrahydroisoquinoline ring and the N-aryl substituent was
initiated to identify the optimal shape for HDAC11 inhibition.
translational modification that occurs on cellular proteins
including, but not limited to, histones. Protein acetylation levels
are controlled by histone deacetylases (HDACs) that catalyze the
removal of acetyl groups and histone acetyltransferases (HATs)
that are responsible for the addition of acetyl groups. The 11
HDAC family members require zinc as a cofactor for deacetylase
activity and are grouped into four classes: class I (HDAC1,
HDAC2, HDAC3, HDAC8), class IIa (HDAC4, HDAC5,
HDAC7, HDAC9), class IIb (HDAC6, HDAC10), and class IV
(HDAC11). HDACs regulate a range of cellular processes
including gene expression, transcription factor activity, cell
signaling pathways, and protein degradation.1
Table 1. SAR of the Hydroxamic Acid Regiochemistry
As a result, HDACs are associated with cancer and other
diseases and have been an active area of drug development.
Currently, four HDAC inhibitors have been approved for use in
hematological cancers (vorinostat, romidepsin, belinostat, and
panobinostat) and many more are currently in clinical trials.2
Most of the reported HDAC inhibitors are pan-inhibitors,
meaning that they inhibit most if not all HDAC isoforms, and
their full utility has been limited by toxicity thought to be
attributable to broad HDAC inhibition.3 Therefore, isoform-
selective inhibitors have the potential to maximize therapeutic
benefit while minimizing side effects.
HDAC11 is the most recently identified HDAC4 and has been
reported to regulate various immune cells, such as antigen-
presenting cells, neutrophils, myeloid-derived suppressor cells, T
cells, and regulatory T cells,5–9 and function in RNA splicing.10
HDAC11-selective inhibitors could have utility for cancer or
inflammatory and immunological diseases. In addition to
deacetylase activity, HDAC11 has also been reported to have
fatty acid deacylase activity.11 Herein, we describe work to
identify selective HDAC11 deacetylase inhibitors with cellular
activity and pharmacokinetic properties in mice that can be used
to further interrogate HDAC11 biology.
Position of
Hydroxamic
Acid
HDAC 11
HDAC 6
IC50 (M)a
IC50 (M)a
7
5
6
8
2.1
1.2
>10
4.6
0.004
>10
1
2
3
4
0.002
0.082
aActivity was measured using electrophoretic mobility shift assays
with full length human recombinant HDAC proteins and
fluorescently labeled peptide substrates. Reported as the mean of at
least two separate assay runs.
Table 2. Optimization of Hydroxamic Acid core
HDAC11a HDAC6a m-CLint
IC50 (M) IC50 (M) (L/min/mg)
b
R1
LipEc
4.5
As part of our broad efforts to design and synthesize potent
and selective inhibitors of the various HDAC family members,
N-hydroxy-tetrahydroisoquinoline-7-carboxamide 1 was
identified as a potent inhibitor of HDAC6 with modest activity
against HDAC11 (Figure 1). The discovery of compound 1
provided an opportunity to identify selective HDAC11 inhibitors,
and efforts to explore this hypothesis were initiated.
1.2
>10
>10
<7
17
2
5
0.17
5.7
5.2
0.03
2.8
161
6
Figure 1. Structure and activity of Compound 1
aActivity was measured using electrophoretic mobility shift
assays with full length human recombinant HDAC proteins and
fluorescently labeled peptide substrates. Reported as the mean of
Hydroxamic acids are well known as one of the common
chemotypes found in HDAC inhibitors.2 As hypothesized for
compounds such as 1, this group coordinates to the zinc atom
located within the active site of all members of the HDAC family
and is essential for HDAC activity.3 Initial studies explored the
impact of the hydroxamic acid position on both the HDAC6 and
HDAC11 activities. As shown in Table 1, the regiochemistry of
the hydroxamic acid substituent affected the potency of both
HDAC11 and HDAC6. Substitution at the 6-position (compound
3) afforded similar HDAC6 activity but resulted in loss of
HDAC11 potency relative to the initial lead (1). Installing the
hydroxamic acid group at the 8-position (4), resulted in a 20-fold
loss in potency vs. HDAC6 but only a 2-fold loss in activity vs.
HDAC11, meaning it could potentially be tolerated. The largest
b
at least two separate assay runs. In vitro intrinsic clearance after
incubation with mouse liver microsomes. cLipophilic efficiency =
pIC50 HDAC11 – clogD7.4.
Table 2 highlights efforts to optimize the hydroxamic acid
core. Working from the tetrahydroisoquinoline, changes to the
ring size of the saturated ring were examined. Thus, isoindoline
and 2,3,4,5-tetrahydrobenzodiazepine cores were explored. Both
isoindolines (5) and benzodiazepine (6) cores were tolerated and
exhibited >10-fold increases in potency against HDAC11 (IC50
170 nM and 3 nM, respectively) while maintaining selectivity
over HDAC6 (IC50 >10 M and 2.8 M, respectively). While
=
tetrahydrobenzodiazepine 6 showed better potency vs. HDAC11,

optimization efforts were focused on the isoindoline 5 due to its
combination of reasonable potency and significant microsomal
stability relative to 6 (CLint = 17 L/min/mg vs. 161 L/min/mg
after incubation with mouse liver microsomes).
internal co-crystal structures of HDAC8 was generated.
Modeling of compound 9 into the homology model supported a
lipophilic binding hypothesis with the trifluoromethyl group
predicted to efficiently fill a small pocket adjacent to the zinc
binding site formed by hydrophobic residues and backbone
carbonyls.
Table 3. Optimization of N-Aryl substituent
While altering the core and N-aryl substituent led to
improvements in both potency and LipE, these changes also
resulted in compounds with poor microsomal stability. It was
hypothesized that the two benzylic methylene groups present in
the isoindoline core were the most likely sites of metabolism and
hypothesized that blocking these sites would afford compounds
with improved microsomal stability.
b
HDAC11a m-CLint
Ar
LipEc
5.7
IC50 (M) (L/min/mg)
Initial efforts focused on substitution of the benzyl group at
the 1-position (Table 4). The less-hindered benzylic group was
expected to be more prone to oxidation relative to the 3-position
(which is presumably shielded by the neighboring hydroxamic
acid). Thus, introduction of the gem-dimethyl group afforded
compound 14 which retained the potency and LipE of the parent
compound (9) with a significant increase in stability (CLint = 12
L/min/mg vs. 140 L/min/mg after incubation with mouse liver
microsomes). This trend was consistent across all analogs,
including benzoxazole 15 and pyridines 17 and 18. The
solubility and permeability of the isoindolines was also examined
in hopes of identifying a compound suitable for in vivo studies.
Pyridines, such as 17 and 18 showed a nice balance of properties,
as did the related pyrimidine (19 and 20) and pyrazine (FT895)
analogs.
0.17
17
5
0.007
0.035
313
323
6.6
5.5
7
8
0.002
140
6.3
9
0.007
0.026
81
92
6.2
5.5
10
11
Having identified a series of potent compounds with suitable
in vitro ADME properties, the cellular activity of these HDAC11
inhibitors was examined. Cellular activity was measured via a
bioluminescence resonance energy transfer (BRET) target
engagement assay using HDAC11 fused to Nanoluc luciferase
and a proprietary compound labeled with a fluorescent tracer.12 In
general, compounds exhibited a five- to twenty-fold shift
between the biochemical and cellular assays, with many having
cellular IC50 values of less than 100 nM. Specifically,
2.3
88
6.5
12
0.30
317
5.8
13
aActivity was measured using electrophoretic mobility shift
assays with full length human recombinant HDAC proteins and
fluorescently labeled peptide substrates. Reported as the mean of
b
at least two separate assay runs. In vitro intrinsic clearance after
incubation with mouse liver microsomes. cLipophilic efficiency =
(trifluoromethyl)pyridine (18) and (trifluoromethyl)pyrazine
(FT895) analogs, displayed IC50 values of less than 20 nM.
pIC50 HDAC11 – clogD7.4.
FT895 was envisioned as a potentially useful tool compound
based on its overall potency and in vitro ADME profile. To
enable a more thorough exploration of HDAC11 biology, a
structurally matched companion inactive control analog was
sought to use in tandem with FT895. Thus, compound 21 was
synthesized and profiled for this purpose, and, as expected,
replacing the hydroxamic acid necessary for zinc binding in the
active site with a primary amide resulted in loss of all HDAC11
activity (Table 4). Furthermore, FT895 was determined to be a
highly-selective HDAC11 inhibitor showing greater than 1000-
fold selectivity against the other 10 members of the HDAC
family (Table 5), while 21 was found to be inactive against all
HDACs.
The optimization of the heterocyclic ring at the 2-position of
the isoindoline ring is summarized in Table 3. A variety of
replacements for the benzimidazole ring were explored, with
benzoxazole (7), benzothiazole (8), pyridine (10), and quinoline
(11) analogs all showing improved potency vs. HDAC11, albeit
with a significant loss of microsomal stability. Saturating one of
the aromatic rings (13) or eliminating one of the rings (to afford
imidazole 12) resulted in loss of potency and microsomal
stability. Of note, introduction of the lipophilic trifluoromethyl
group (compound 9) resulted in a significant increase in both
potency (2 nM vs. 170 nM) and LipE (6.3 vs. 5.7) but again with
loss of microsomal stability relative to 5. In the absence of a co-
crystal structure of HDAC11, a homology model based on
Table 4. Optimization of N-Arylisoindolines
c
HDAC11a
IC50 (M)
HDAC11 BRET b m-CLint
IC50 (M)
Solubilityd
PAMPAe
f
clogD7.4 LipEg
Ar
R
(10-6 cm/s)
(L/min/mg) (M)

0.002
0.001
NT
NT
12
10
51
4
0.1
7
2.7
3.1
6.0
6.1
14
15
OH
OH
0.002
0.006
0.003
0.005
0.020
0.13
7
78
8
0.7
12
8
1.6
2.6
2.5
2.2
7.2
5.7
6.1
6.1
16
17
18
19
OH
OH
OH
OH
28
12
21
0.012
0.085
51
32
18
0.027
0.003
>10
0.27
0.015
>25
7
82
31
0.3
NT
13
1.1
2.3
2.7
6.5
6.3
2.3
20
OH
OH
H
22
7
FT895
21
17
aActivity was measured using electrophoretic mobility shift assays with full length human recombinant HDAC proteins and
fluorescently labeled peptide substrates. Reported as the mean of at least two separate assay runs. NT = not tested. bBRET Cell assay.
cIn vitro intrinsic clearance after incubation with mouse liver microsomes. dKinetic solubility at pH 7.4. ePermeability measured via
parallel artificial membrane permeability assay. fCalculated logD value at pH 7.4. gLipophilic efficiency = pIC50 HDAC11 – clogD7.4.
Based on its overall in vitro profile, FT895 was advanced to mouse PK studies to measure its suitability as an in vivo tool
compound. The pharmacokinetic properties of FT895 were assessed in male Balb/c nude mice following both intravenous (i.v.) and
intraperitoneal (i.p.) dosing (Table 6).13 The compound displayed a moderate clearance (42 mL/min/kg) and high volume of
distribution, resulting in a half-life of 9.4 h after i.v. dosing. When dosed i.p., FT895 had a similar t1/2 (10.2 h) and improved
exposure, resulting in a bioavailability of 81%. FT895 also maintained free drug levels over the cellular IC50 for up to 4 h after a
single 5 mg/kg i.p. dose, thus providing a potentially useful tool for further understanding the biology of HDAC11 in vitro and in
vivo.
a
Table 5. HDAC Activity Profiles of FT895 and 21
HDAC IC50 (M)
1
2
3
4
5
6
7
8
9
10 11
>10 >10 >10 >10 >10 >10 >10 5.6 >10 >10 0.003
FT895
21
>10 >10 >10 >10 >10 >10 >10 >10 >10 >10 >10
aActivity was measured using electrophoretic mobility shift assays with full length human recombinant HDAC proteins and fluorescently
labeled peptide substrates. Reported as the mean of at least two separate assay runs.
a
Table 6. Pharmacokinetic Profile of FT895
C0
(M)
3.5
AUClast
(M*h)
1.0
IV Dose
(mg/kg)
t1/2
(h)
Vss
(L/kg) (mL/min/kg)
18 42
CL
1
9.4
Cmax
(M)
4.2
AUClast
(M*h)
3.8
IP Dose
(mg/kg)
t1/2
(h)
Bioavailability (%F)
81
5
10.2
aDosed in male Balb/c nude mice (n=3). Dosing formulation: 5% DMA/1% Tween 80/94% sterile water

Scheme 1. Synthesis of 2-arylisoindoline-4-carboxamides. Reagents and conditions: (a) NBS, benzoyl peroxide, CCl4, 80 °C, 16 h, 100%;
(b) p-methoxybenzylamine, Et3N, MeOH, 40 °C, 4 h, 37%; (c) NaH, THF, rt, 3 h then MeI, rt, 16 h, 61 %; (d) BH3-THF, 80 °C, 16 h,
83%; (e) CO (60 atm), PdCl2(dppf), Et3N, MeOH, 130 °C, 24 h, 47%; (f) H2, Pd/C, conc. HCl, MeOH, rt, 16 h, 79%; (g) ArCl, RuPhos,
RuPhos 2G precatalyst, Cs2CO3, toluene, 100 °C, 16 h, 44-88%; (h) NH2OH, NaOH, THF, MeOH, rt, 4 h, 12-48%; (i) LiOH, THF, H2O, rt,
16 h, 78-94%; (j) NH4Cl, HATU, (i-Pr2)NEt, DMF, rt, 16 h, 20-34%. PMB = p-methoxybenzyl.
The N-hydroxy-2-arylisoindoline-4-carboxamides described in Tables 3 and 4 were synthesized according to the route illustrated
in Scheme 1. Bromination of ethyl 2-bromo-6-methylbenzoate (22) with NBS afforded benzyl bromide 23 in quantitative yield.
Amination and intramolecular cyclization provided isoindolinone 24 which was subsequently deprotonated with sodium hydride
and alkylated with methyl iodide to afford 25. Chemoselective reduction of the carbonyl group with borane provided isoindoline
26, which underwent a palladium-catalyzed carbonylation in the presence of carbon monoxide and methanol to yield 27. Removal
of the p-methoxybenzyl protecting group afforded isoindoline 28. Palladium-catalyzed Buchwald coupling14 of 28 with the
appropriate aryl chloride and the second generation RuPhos precatalyst yielded isoindoline carboxylate 29. Treatment of 29 with
hydroxyl amine generated the hydroxamic acid and provided the desired N-hydroxy-2-arylisoindoline-4-carboxamides (30a) in
good yield. For the preparation of amides, such as 21, ester 29 was saponified with lithium hydroxide, and the resulting acid was
coupled with ammonium chloride in the presence of HATU to give amide 30b.
In summary, a novel series of N-hydroxy-2-arylisoindoline-4-carboxamides have been identified as potent and highly selective
inhibitors of HDAC11. These first-in-class inhibitors show promising cellular activity and pharmacokinetic properties that make
them useful tools to better understand the biology of HDAC11 and its potential use as a therapeutic target for oncology and
inflammation indications. Future work will include further optimization of FT895 and its utility in in vivo models.
Acknowledgments
We thank Kyle Bassoli, Tanner Burk, Christopher Delude, Hien Diep, Christopher Lombardi, Fu-ni Luan, and Megan Walton
for analytical, in vitro ADME, and compound management support.
Supplementary Material
Supplementary material associated with this article (including details of the in vitro HDAC 11 assay and experimental
procedures for the synthesis of compounds FT895 and 21) can be found in the online version.
References and notes
1. West AC, Johnstone RW. New and emerging HDAC inhibitors for cancer treatment. J Clin Invest. 2014;124(1):30-39. doi:10.1172/JCI69738
2. Eckschlager T, Plch J, Stiborova M, Hrabeta J. Histone Deacetylase Inhibitors as Anticancer Drugs. Int J Mol Sci. 2017;18(7).
doi:10.3390/ijms18071414
3. Thaler F, Mercurio C. Towards selective inhibition of histone deacetylase isoforms: what has been achieved, where we are and what will be next.
ChemMedChem. 2014;9(3):523-526. doi:10.1002/cmdc.201300413
4. Gao L, Cueto MA, Asselbergs F, Atadja P. Cloning and functional characterization of HDAC11, a novel member of the human histone deacetylase
family. J Biol Chem. 2002;277(28):25748-25755. doi:10.1074/jbc.M111871200
5. Woods DM, Woan KV, Cheng F, et al. T-cells lacking HDAC11 have increased effector functions and mediate enhanced alloreactivity in a murine
model. Blood. May 2017. doi:10.1182/blood-2016-08-731505

6. Sahakian E, Powers JJ, Chen J, et al. Histone deacetylase 11: A novel epigenetic regulator of myeloid derived suppressor cell expansion and
function. Mol Immunol. 2015;63(2):579-585. doi:10.1016/j.molimm.2014.08.002
7. Sahakian E, Chen J, Powers JJ, et al. Essential role for histone deacetylase 11 (HDAC11) in neutrophil biology. J Leukoc Biol. May 2017.
doi:10.1189/jlb.1A0415-176RRR
8. Huang J, Wang L, Dahiya S, et al. Histone/protein deacetylase 11 targeting promotes Foxp3+ Treg function. Sci Rep. 2017;7(1):8626.
doi:10.1038/s41598-017-09211-3
9. Villagra A, Cheng F, Wang H-W, et al. The histone deacetylase HDAC11 regulates the expression of interleukin 10 and immune tolerance. Nat
Immunol. 2009;10(1):92-100. doi:10.1038/ni.1673
10. Joshi P, Greco TM, Guise AJ, et al. The functional interactome landscape of the human histone deacetylase family. Mol Syst Biol. 2013;9:672.
doi:10.1038/msb.2013.26
11. Kutil Z, Novakova Z, Meleshin M, Mikesova J, Schutkowski M, Barinka C. Histone Deacetylase 11 Is a Fatty-Acid Deacylase. ACS Chem Biol.
2018;13(3):685-693. doi:10.1021/acschembio.7b00942
12. Robers MB, Dart ML, Woodroofe CC, et al. Target engagement and drug residence time can be observed in living cells with BRET. Nat Commun.
2015;6:10091. doi:10.1038/ncomms10091
13. FT895 was also dosed orally at 5 mg/kg, but no significant exposure was observed.
14. For a review of palladium-catalyzed C-N cross-coupling reactions, see: Ruiz-Castillo P, Buchwald SL. Applications of Palladium-Catalyzed C-N
Cross-Coupling Reactions. Chem Rev. 2016;116(19):12564-12649. doi:10.1021/acs.chemrev.6b00512

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Discovery of novel N-hydroxy-2-
Leave this area blank for abstract info.
arylisoindoline-4-carboxamides as potent
and selective inhibitors of HDAC11
Matthew W. Martin, Jennifer Lee, David R. Lancia, Jr., Pui Yee Ng, Bingsong Han, Jennifer R. Thomason,
Maureen S. Lynes, C. Gary Marshall, Chiara Conti, Alan Collis, Monica Alvarez Morales, Kshama Doshi,
Aleksandra Rudnitskaya, Lili Yao, and Xiaozhang Zheng

Highlights
 Identified novel, potent, and selective first-
in-class inhibitors of HDAC11
 Synthesis and structure activity
relationships of compounds are reported
 FT895 displays promising cellular activity
and pharmacokinetic properties
 FT895 is a valuable tool for further study of
the biology of HDAC11