Design, Synthesis, and Biological Evaluation of Quinazolin-4-one-Based Hydroxamic Acids as Dual PI3K/HDAC Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.0c00193

A series of quinazolin-4-one based hydroxamic acids was rationally designed and synthesized as novel dual PI3K/HDAC inhibitors by incorporating an HDAC pharmacophore into a PI3K inhibitor (Idelalisib) via an optimized linker. Several of these dual inhibitors were highly potent (IC₅₀ < 10 nM) and selective against PI3Kγ, δand HDAC6 enzymes and exhibited good antiproliferative activity against multiple cancer cell lines. The lead compound 48c, induced necrosis in several mutant and FLT3-resistant AML cell lines and primary blasts from AML patients, while showing no cytotoxicity against normal PBMCs, NIH3T3, and HEK293 cells. Target engagement of PI3Kδand HDAC6 by 48c was demonstrated in MV411 cells using the cellular thermal shift assay (CETSA). Compound 48c showed good pharmacokinetics properties in mice via intraperitoneal (ip) administration and provides a means to examine the biological effects of inhibiting these two important enzymes with a single molecule, either in vitro or in vivo.

Full text of DOI:10.1021/acs.jmedchem.0c00193

Subscriber access provided by University of Massachusetts Amherst Libraries
Article
Design, Synthesis and Biological Evaluation of Quinazolin-4-
one Based Hydroxamic Acids as Dual PI3K/HDAC Inhibitors
Ashish Thakur, Gregory James Tawa, Mark J. Henderson, Carina Danchik, Suiyang
Liu, Pranav Shah, Amy Q. Wang, Garrett Dunn, Md Kabir, Elias C Padilha, Xin
Xu, Anton Simeonov, Surender Kharbanda, Richard Stone, and Gurmit Grewal
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00193 • Publication Date (Web): 26 Mar 2020
Downloaded from pubs.acs.org on March 28, 2020
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 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 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.
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 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design, Synthesis and Biological Evaluation of
Quinazolin-4-one Based Hydroxamic Acids as Dual
PI3K/HDAC Inhibitors
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
Ashish Thakur*, † Gregory J. Tawa, Mark J. Henderson, Carina Danchik, Suiyang Liu, Pranav
†
†
†
‡
†
†
†
†
†
†
†
Shah, Amy Q. Wang, Garrett Dunn, Md Kabir, Elias C. Padilha, Xin Xu, Anton Simeonov,
‡
‡
*, †
Surender Kharbanda, Richard Stone and Gurmit Grewal
†National Center for Advancing Translational Sciences, National Institutes of Health, 9800
Medical Center Drive, Rockville, Maryland 20850, United States
‡Dana-Farber Cancer Institute, 450 Brookline Ave, Boston, MA 02215, USA
ABSTRACT: A series of quinazolin-4-one based hydroxamic acids was rationally designed and
synthesized as novel dual PI3K/HDAC inhibitors by incorporating an HDAC pharmacophore into
a PI3K inhibitor (Idelalisib) via an optimized linker. Several of these dual inhibitors were highly
potent (IC < 10 nM) and selective against PI3K,  and HDAC6 enzymes, and exhibited good
5
0
anti-proliferative activity against multiple cancer cell lines. The lead compound 48c, induced
necrosis in several mutant and FLT3-resistant AML cell lines and primary blasts from AML
patients, while showing no cytotoxicity against normal PBMCs, NIH3T3 and HEK293 cells.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 117
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
Target engagement of PI3K and HDAC6 by 48c was demonstrated in MV411 cells using the
Cellular Thermal Shift Assay (CETSA). Compound 48c showed good pharmacokinetics properties
in mice via intraperitoneal (ip) administration and provides a means to examine the biological
effects of inhibiting these two important enzymes with a single molecule, either in vitro or in vivo.
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
Most of the currently approved drugs based on the “single-target single drug” model are becoming
less effective in the treatment of complex, heterogeneous, multigenic diseases, like cancer.1, 2
While drug-combination therapies are employed as an alternative approach to achieve efficacy,
their benefits are often negated by adverse drug-drug interactions, unpredictable pharmacokinetic
(
PK) and safety profiles, and poor patient compliance.3, 4 In recent years, there has been an
increasing inclination towards discovering “multi-target drugs” to address the limited efficacy, and
resistance or toxicity, associated with many single-target or combination-based therapies.5-12
Class I phosphoinositide 3-kinases (PI3Ks) are lipid kinase enzymes that transduce signals
from cell surface receptors (RTK, GPCR, etc.) to downstream effectors (Akt, mTOR, etc.), leading
to a variety of cellular processes, including cell proliferation, survival, differentiation, metabolism
and angiogenesis.13, 14 Specifically, class I PI3Ks phosphorylate phosphatidylinositol (4,5)-
bisphosphate (PI(4,5)P2 or PIP2) in vivo to form the secondary messenger phosphatidylinositol
(3,4,5)-trisphosphate (PI(3,4,5)P3 or PIP3), which activates the pleckstrin homology(PH)-domain
of protein kinases (Akt, BTK, etc.), leading to downstream signaling.15, 16 The four isoforms of
Class I PI3Ks (PI3K PI3K PI3K and PI3K) exist as heterodimers, consisting of a catalytic
1
4
p110 (   ) subunit and a regulatory p85 (  ) or p101/p84 () subunit. The PI3K and
PI3K isoforms are ubiquitously expressed, whereas PI3K and PI3K expression is limited to
ACS Paragon Plus Environment
2

Page 3 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
16
leukocytes. Aberrant regulation of class I PI3Ks has been implicated in several cancer types and
their inhibition continues to be an active cancer therapeutic area with four recently approved PI3K
inhibitor drugs (Idelalisib, Copanlisib, Duvelisib and Alpelisib) (Figure 1) and several others in
ongoing clinical trials.17-20 Despite tremendous progress in the discovery of PI3K inhibitors
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
(
PI3Ki) over the last decade, clinical trials with PI3Ki as a monotherapy have shown poor
2
1
efficacy, prompting their evaluation in combination therapies and/or developing PI3K based
multi-target drugs.19
O
N
F
F
F
O
O
N
N
O
F
O
N
F
O
N
N
N
N
HN
S
N
N
HN
N
HN
N
HN
O
NH2
N
N
N
N
N
N
O
NH
NH2
NH
Idelalisib
PI3K Inhibitor
Copanlisib
PI3K Inhibitor
Duvelisib
PI3K Inhibitor
Approved (2018) for
Alpelisib
PI3K Inhibitor
Approved (2019) for (HR)-
positive, (HER2)-negative
Breast Cancer
Approved (2014) for relapsed CLL Approved (2017) for
relapsed FL
relapsed CLL/SL
Figure 1. FDA Approved PI3K inhibitor drugs for cancer. (CLL = chronic lymphocytic leukemia; FL = Follicular
lymphoma; SLL = small lymphocytic lymphoma).
Histone deacetylases (HDACs) are an important class of epigenetic enzymes that regulate
2
2
gene expression by removing acetyl groups from -amino lysine residues on histones. HDACs
play an important role in regulating expression of various proteins, including tumor suppressors
23
(
p53, p21, etc.) and transcription factors (e.g. TFIIE, TCF, SF1, etc.). Dysregulation of HDACs
2
4
is involved in cancer initiation and proliferation. HDAC inhibition has emerged as a therapeutic
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 117
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
approach for cancer and several inhibitors, including Vorinostat (SAHA), Romidepsin, Belinostat,
2
5
Panobinostat and Chidamide, have been approved in recent years (Figure 2A). Pan-HDAC
inhibitors, such as Panobinostat and Vorinostat (SAHA), modulate the acetylation status of a wide
range of protein targets leading to a therapeutic response; however, these molecules also have
undesirable effects, including hematological, gastrointestinal and cardiac toxicity.25-27 SAHA
monotherapy was approved by the U.S. Food and Drug Administration (FDA) in 2006 for the
treatment of cutaneous T-cell lymphoma (CTCL), however it has been demonstrated to have little
efficacy.28
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
OH
HN
HO
NH
F
O
N
OH
NH
O
O
HO
NH2
NH
O
NH
NH
N
O
N
S
HN
O
S
NH
O
NH
O
N
N
O
O
N
N
O
N
H
N
Panobinostat (LBH589)
pan-HDAC Inhibitor
Approved (2015) for
Multiple Myeloma
Chidamide (CS005)
pan-HDAC Inhibitor
Approved (2015) in China
for PTCL
Vorinostat (SAHA)
pan-HDAC Inhibitor
Approved (2006) for CTCL Approved (2014) for PTCL
Belinostat (PXD101)
pan-HDAC Inhibitor
Fimepinostat (CUDC-907)
pan-PI3K/HDAC Inhibitor
Phase II for DLBCL
Figure 2. A) Selected approved HDAC inhibitor drugs for cancer. B) Chemical structure of dual PI3K/HDAC
inhibitor, Fimepinostat. (CTCL = cutaneous T-cell lymphoma; PTCL = peripheral T-cell lymphoma; DLBCL =
diffuse large B-cell lymphoma)
Although both PI3K and HDAC inhibitor drugs are limited by insufficient efficacy and
resistance mutations, there is strong evidence that simultaneous inhibition of both PI3K and HDAC
can synergistically inhibit tumor growth and address these limitations by improving efficacy,
ACS Paragon Plus Environment
4

Page 5 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
limiting resistance and providing a better therapeutic window than single inhibitors.28-30 Selective
inhibitors of specific isoforms of HDAC and PI3K potentially would have a better toxicity profile
3
1
and therefore bigger therapeutic window. In this context, several dual PI3K/HDAC inhibitors
have been recently reported in literature,32-34 with Curis Inc.’s Fimepinostat (CUDC-907) currently
being evaluated in Phase 2 clinical trials for treatment of diffuse large B-cell lymphoma (DLBCL)
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 2B].19, 35, 36 Dual inhibitors such as CUDC-907, may thus offer improved therapeutic
benefit through synergistic effects in inhibiting cancer cell proliferation and their compensatory
pathways for survival.19, 35 However, CUDC-907 exhibits pan-HDAC and pan-PI3K inhibition,
which might lead to toxicity and low tolerability.37, 38 Thus, there remains an unmet need for new
dual inhibitors having high potency and selectivity.
Herein we report the design, synthesis and biological evaluation of potent and selective
dual PI3K/HDAC inhibitors.
RESULTS AND DISCUSSION
Design of dual inhibitors. The rational design of our PI3K/HDAC dual inhibitors commenced by


analyzing the crystal structures of Idelalisib (1) bound to PI3K [PDB ID: 4XE0] (Figure 3A)
4
0
and Vorinostat (2) bound to HDAC2 [PDB ID: 4LXZ] (Figure 3B), respectively. As shown in
Figure 3A, the purine moiety in Idelalisib binds to the hinge (Glu826, Val828) and the
quinazolinone core extends up into a pocket formed by the G loop residues, Trp760 and Met752.
th
th
The carbonyl and fluoro group in the quinazolinone core at 4 and 5 positions, respectively,
extend into the solvent exposed area, and therefore, serve as desired locations to attach the HDAC
pharmacophore. The phenyl ring on the nitrogen in the quinazolinone core is largely solvent
exposed (on top) but makes hydrophobic contacts underneath to methyl groups on Thr833 and
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 117
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
Met900. In the Vorinostat-HDAC2 structure (Figure 3B), Zn2+ is penta-coordinated (Asp181,
Asp260, His183, and both oxygens on hydroxamic acid) and the hydroxamic acid group forms
hydrogen bonds to two charge relay histidines (His145 and His146) and a nearby Tyr308. The
Vorinostat aliphatic chain traverses a hydrophobic portion of the pocket (Leu144, Phe155). The
phenyl group forms a hydrogen bond to the rim of the pocket at Asp104 via the amino nitrogen.
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
Solvent Exposed Region
F
5
O
N
4
N
HN
N
N
N
NH Hinge
Idelalisib (1)
C
D
OH
NH
O
O
NH
Vorinostat [SAHA] (2)
Figure 3. A) Chemical structure of the PI3K inhibitor, Idelalisib (1). B) Crystal structures of Idelalisib (1) [top view]
bound to P (PDB ID: 4XE0) with protein surface (blue) and without protein surface, respectively. C) Chemical
structure of the HDAC inhibitor, Vorinostat (2). D) Crystal structures of Vorinostat (2) bound to HDAC2 (PDB ID:
4
LXZ) with protein surface (blue) and without protein surface, respectively.
ACS Paragon Plus Environment
6

Page 7 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In general, most of the HDAC inhibitors, including Vorinostat, share these three common
structural components; a zinc binding group (e.g. hydroxamic acid), a linker occupying the
hydrophobic tunnel in the active site and a cap group, that sits outside the hydrophobic tunnel and
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
41
interacts with the HDAC surface residues (Figure 4). More importantly, the HDAC inhibitors
can accommodate a wide variety of cap groups, which allows for replacement of the cap with a
second target pharmacophore to design HDAC inhibitor based multi-target drugs.32, 33, 42-50 Thus,
by assessing the binding features of Idelalisib and Vorinostat in PI3K and HDAC2 enzymes,
respectively, we designed quinazoline- and quinazolinone-based dual PI3K/HDAC inhibitors (3
and 4, Figure 4) by incorporating the PI3K inhibitor as the cap group of an HDAC pharmacophore
and optimizing for the appropriate linker-hydroxamic acid combination at quinazoline’s C-4
position (3) or quinazolinone’s C-5 position (4) to achieve potencies against both the targets.
Zinc Binding Group
HO
H
H
NH
O
N
O
N
Solvent exposed region
O
OH
OH
Linker
Linker
F
O
N
O
4
5
N
4
Linker
N3
N3
&
2
2
N
N
1
1
HN
N
O
HN
N
HN
N
HN
N
N
N
N
N
N
NH
NH
NH
Cap
Idelalisib (1)
3
4
Vorinostat (2)
PI3K/HDAC Dual Inhibitors
Figure 4. Design strategy for dual PI3K/HDAC inhibitor types, 3 and 4.
Chemistry. The C-4 substituted quinazoline compounds (10a, b and 14 a, b) were synthesized
5
1
from tert-butyl (S)-(1-(4-chloroquinazolin-2-yl)propyl)carbamate 5, in 6-7 steps as outlined in
Scheme 1. The compound 5 was subjected to Sonogashira cross-coupling reactions with alkynes
t
6a and 6b, using a copper-free Pd/P Bu -catalytic system to afford 7a, b. Reduction of alkyne
3
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 117
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
group to alkane in 7a, b using Pd/C hydrogenation, followed by boc-deprotection and nucleophilic
aromatic substitution of 6-chloro-9-(tetrahydro-2H-pyran-2-yl)-9H-purine 8 with the resulting free
amine, yielded compounds 9a, b. Hydrolysis of the alkyl ester in 9a, b to the respective carboxylic
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
acid, followed by coupling with NH OTHP (THP = tetrahydropyran) and THP-deprotection using
2
trifluoroacetic acid formed the target compounds 10a, b.
Scheme 1. Synthesis of C-4 alkyl/amino-alkyl substituted quinazolines [10a-b and 14a-b]a
O
O
n(H2C) OR1
O
n(H C) OR1
Cl
N
2
OH
O
N
N
n(H2C)
N
H
N
(H C) OR1
Cl
4
6
a n = 2, R = t-Bu
1
n
2
4
1
N
6
b n = 3, R = Me
a)
THP
8
N
N
(
(b)
N
(c)
N
N
N
N
HN
N
NHBoc
HN
N
N
N
NHBoc
N
N
1
N
5
7a, n = 2, R = t-Bu, 61%
NH
THP
O
1
7
b, n = 3, R = Me, 57%
1
10a, n = 2, 55%
10b, n = 3, 60%
9
9
a, n = 2, R = H, 56%
m(H C) OR2
2
1
b, n = 3, R = Me, 74%
NH .HCl
2
2
1
1a m = 3, R = Et
2
OR2
OH
NH
1
1b m = 5, R = Me
O
(d)
O
OR2
m
(H C)
2
O
NH
m(H2C)
N4H
m(H C)
2
N
NH
8
N
N
(c)
N
(e)
N
HN
N
N
N
HN
N
N
N
NHBoc
N
N
THP
NH
2
2
1
2a, m =3, R = Et, 94%
13a, m =3, R = Et, 77 %
14a, m =3, 38%
14b, m =5, 38%
2
2
1
2b, m =5, R = Me, 94%
13b, m =5, R = Me, 95 %
a
t
Reagents and Conditions: (a) 6a, 6b (1.2 equiv), [PdCl(allyl)] (0.05 equiv), P Bu HBF (0.20 equiv), 1,4-dioxane, 16 h, rt; (b) i) 10 wt% Pd/C, H
2
3
4
2
balloon, EtOAc, 20 h, rt; ii) CF CO H (20.0 equiv), DCM, 3 h, rt; iii) 8 (2.0 equiv), triethyamine (4.0 equiv), EtOH, 1h, MW, 100 °C; (c) i)
3
2
LiOHH O (2.0 equiv), MeOH/H O, 10 h, rt; ii) NH OTHP (3.1 equiv), N-methyl morpholine (3.0 equiv) 3-(((ethylimino)methylene)amino)-N,N-
2
2
2
dimethylpropan-1-amine hydrochloride [EDC.HCl] (1.4 equiv), 1H-[1,2,3]triazolo[4,5-b]pyridin-1-ol [HOAT] (1.2 equiv), DMF, 16h rt; iii)
CF CO H (20.0 equiv), DCM, 20 h, rt, (For 9a only c(ii) and c(iii) were used); (d) 11a,b (2.0 equiv), triethyamine (4.0 equiv), EtOH, 1h, 100 °C; (e)
3
2
8 (2.0 equiv), triethyamine (4.0 equiv), EtOH, 1h, 100 °C.
The synthesis of C-4 amino-alkyl substituted compounds (14a, b) began by subjecting
compound 5 to nucleophilic aromatic substitution with amines 11a, b to yield compounds 12a, b.
Boc-deprotection in 12a, b, followed by the installation of a purine kinase hinge binding moiety
using nucleophilic aromatic substitution (S Ar) on 8, afforded compounds 13a, b. Target
N
ACS Paragon Plus Environment
8

Page 9 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds 14a, b were synthesized from 13a, b by converting alkyl ester to hydroxamic acid, in
a fashion similar to that described above for compounds 10a, b.
Scheme 2. Synthesis of C-5 alkyl/aryl substituted quinazolinones [19a-c and 24a-c]a
O
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
Z
OR1
O
Cl
N
O
OR1
O
OH
O
O
N
N
Z
1
N
Z
N
1
6a Z = (CH ) , R = t-Bu
2 2
H
OR1
1
Z
Br
O
N
16b Z = (CH ) , R = Me
2 3
6c Z = phenylene, R1 = Me
5
5
THP
8
N
1
N
O
N
N
N
(a)
(b)
N
(c) or (d)
N
HN
N
NHBoc
HN
N
N
N
N
NHBoc
1
N
N
1
NH
15a
THP
17a, Z = (CH ) , R = t-Bu, 75%
18a, Z = (CH ) , R = H, 54%
19a, Z = (CH ) , 43%
2
2
2 2
2 2
1
1
1
1
7b, Z = (CH ) , R = Me, 63%
18b, Z = (CH ) , R = Me, 92%
18c, Z = phenylene, R1 = Me, 74%
19a, Z = (CH2)3, 24%
19a, Z = phenylene, 25%
2 3
2 3
7c, Z = phenylene, R1 = Me, 93%
O
B
OEt
OEt
Ar
OH
HN
HO
Ar
O
O
N
OEt
O
Ar
O
X
5
O
N
5
OH
0a-b
N
O
Ar
O
N
8
N
2
N
N
(b)
(c) or (d)
N
(e)
HN
N
HN
N
NHBoc
N
NHBoc
N
N
N
N
THP
NH
1
5a, X = Br
22a, 96 %
22b, 90%
23a, 91%
23b, 80%
23c, 70%
24a, 71 %
24b, 36 %
24c, 32%
15a', X = Cl
O
2
2c, 70 %
Br
(f)
Ar OEt
0c
2
O
O
(g)
O
OEt
EtO
O
OEt
OEt
=
O
O
Ar
B
O
N
S
N
N
20a
20b
20c
NHBoc
21
a
t
Reagents and Conditions: (a) 16a, 16b, 16c (1.2 equiv), [PdCl(allyl)]2 (0.05 equiv), P Bu3HBF4 (0.20 equiv), 1,4-dioxane, 16 h, rt; (b) i) 10 wt% Pd/C, H2 balloon, EtOAc, 20 h, rt;
ii) CF3CO2H (20.0 equiv), DCM, 3 h, rt; iii) 8 (1.5-2.5 equiv), triethyamine (3.0-5.0 equiv), EtOH, 4h, MW, 100 °C; (c) i) LiOHH2O (2.0 equiv), MeOH/H2O, 10 h, rt; ii)
NH2OTHP (3.1 equiv), N-methyl morpholine (3.0 equiv) 3-(((ethylimino)methylene)amino)-N,N-dimethylpropan-1-amine hydrochloride [EDC.HCl] (1.4 equiv), 1H-
[
1,2,3]triazolo[4,5-b]pyridin-1-ol [HOAT] (1.2 equiv), DMF, 16h rt; iii) CF3CO2H (20.0 equiv), DCM/MeOH, 20 h, rt, (For 17a, used only c(ii) and c(iii)); (d) i) aq. 50 wt%
NH2OH (30.0 equiv), LiOHH2O (1.1 equiv), MeOH/H2O, 20h, 0 °C to rt; ii) CF3CO2H (20.0 equiv), DCM/MeOH, 20 h, rt; (e) Boronic acid/ester 20a-b (1.2 equiv), [XPhos
Pd(crotyl)Cl] (0.05 equiv), K3PO4 (3.0 equiv), 1,4-dioxane/water, 1-2h MW, 100 °C; (f) [Bpin]2 (1.2 equiv), [Pd(dppf)Cl2] (0.05 equiv), KOAc (3.0 equiv) 1,4-dioxane, 16h, 100
°
C; (g) 20c (0.85 equiv), [XPhos Pd(crotyl)Cl] (0.05 equiv), K3PO4 (3.0 equiv), 1,4-dioxane/water, 10h, 100 °C; (h) i) CF3CO2H (20.0 equiv), DCM, 3 h, rt; ii) 8 (1.5-2.5 equiv),
triethyamine (3.0-5.0 equiv), EtOH, 4h, 100 °C.
The procedure for syntheses of C-5 alkyl/aryl substituted quinazolinone-based compounds
5
2
is depicted in Scheme 2. The quinazolinone compound 15a was subjected to Sonogashira cross-
coupling reactions with alkynes 16a-c to yield compounds 17a-c. Hydrogenation of alkyne in 17a-
c followed by Boc-deprotection and nucleophilic aromatic substitution on purine 8 afforded
compounds 18a-c. Target compounds 19a-c were derived from 18a-c by either: a) hydrolyzing the
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 117
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
alkyl ester to carboxylic acid, coupling with NH OTHP followed by deprotection of THP-
2
protecting groups using TFA, or b) converting the alkyl ester directly to hydroxamic acid using
aq. NH OH/LiOH and subsequent removal of the THP-protecting group using TFA. The latter
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
method also formed carboxylic acid as the minor side product due to ester hydrolysis (See
Supporting Information).
For C-5 aryl substituted quinazolinone-based compounds, 5-bromo or 5-chloro-
quinazolinone (15a or 15a′) was either subjected to a Suzuki-Miyaura cross-coupling reaction with
boronic acids 20a, b or first converted to boronate ester 21 using Pd(dppf)Cl /Bpin followed by
2
2
Suzuki-Miyaura coupling with aryl bromide 20c, to form compounds 22a-c. These C-5 arylated
molecules 22a-c were converted to target compounds 24a-c in a sequence of steps similar to that
previously described for compounds 19a-c.
Scheme 3. Synthesis of C-5 amino-alkyl/aryl substituted quinazolinones [28a-i]a
O
OR2
R3
Z
HO
O
Cl
N
Z
OR2
N
N
O
Z
OR2
Z
N
ClH.H2N
NH
O
N
Br
O
N
N
O
NH
O
N
2
5a-h
THP
R2 = Me or Et
8
N
N
NH
O
N
N
R1
(a)
(
b)
(c) or (d) or
(e) (For 28i)
N
HN
N
NHBoc
R1
HN
N
N
NHBoc
N
N
N
N
NH
THP
1
1
2
2
3
1
5a, R = Et,
26a, R = Et, R = Et, 70 %
27a, R = Et, 53%
28a, R = H, 21%
1
1
2
2
3
1
5b, R = Me
26b, R = Et, R = Me, 63%
27b, R = Me, 60%
28b, R = H, 61%
1
2
2
3
2
6c, R = Et, R = Me, 91%
27c, R = Me, 90%
28c, R = H, 32%
1
2
2
3
2
6d, R = Et, R = Me, 89 %
27d, R = Me, 80%
28d, R = H, 50%
1
2
2
3
26e, R = Et, R = Me, 88%
27e, R = Me, 91%
28e, R = H, 55%
1
2
2
3
26f, R = Et, R = Me, 59%
27f, R = Me, 76%
28f, R = H, 32%
1
2
2
3
2
2
2
6g, R = Et, R = Me, 77%
27g, R = Me, 67%
28g, R = H, 36%
1
2
2
3
6h, R = Et, R = Me, 88%
27h, R = Me, 72%
28h, R = H, 78%
1
2
28i, R3 = Me, Z = phenylene, 30%
6i, R = R = Me, Z = phenylene, 77%
O
OMe
O
OMe
O
OMe
O
OMe
O
OMe
O
Z
O
O
O
=
N
OR2
(
)2 OEt
()3 OMe ()4 OMe
O
N
2
5a
25b
25c
25d
25e
25f
25g
25h
aReagents and Conditions: (a) 25a-h (1.3 equiv), [Xantphos Palladacycle Gen. 4] (0.03 equiv),Cs2CO3 (3.0 equiv), toluene or 1,4-dioxane, 16 h, rt; (b) i) CF3CO2H (20.0 equiv), DCM,
h, rt; ii) 8 (1.5-2.5 equiv), triethyamine (3.0-5.0 equiv), EtOH, 4h, MW, 100 °C; (c) i) LiOHH2O (2.0 equiv), MeOH/H2O, 10 h, rt; ii) NH2OTHP (3.1 equiv), N-methyl morpholine
3.0 equiv) 3-(((ethylimino)methylene)amino)-N,N-dimethylpropan-1-amine hydrochloride [EDC.HCl] (1.4 equiv), 1H-[1,2,3]triazolo[4,5-b]pyridin-1-ol [HOAT] (1.2 equiv), DMF, 16h
3
(
rt; iii) CF3CO2H (20.0 equiv), DCM/MeOH, 20 h, rt; (d) i) aq. 50 wt% NH2OH (30.0 equiv), LiOHH2O (1.1 equiv), MeOH/H2O, 20h, 0 °C to rt; ii) CF3CO2H (20.0 equiv),
DCM/MeOH, 20 h, rt; (e) i) LiOHH2O (1.1 equiv), MeOH/H2O, 20h, 0 °C to rt; ii) N-methylhydroxylamine hydrochloride (1.5 equiv), HATU (1.5), DIPEA (2.5 equiv), DMF, rt; iii)
CF3CO2H (20.0 equiv), DCM/MeOH, 20 h, rt.
ACS Paragon Plus Environment
10

Page 11 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The synthesis of C-5 amino-alkyl/aryl substituted quinazolinones involved the
Pd/XantPhos- catalyzed amination of 5-bromo-quinazolinone 15a, b with amines 25a-h yielding
compounds 26a-i (Scheme 3). Subsequent attachment of purine 8 and conversion of alkyl esters
to hydroxamic acids afforded the target compounds 28a-h. For compound 28i, the methyl ester in
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
7d was converted to N-methyl hydroxamic acid via hydrolysis to carboxylic acid followed by its
coupling with N-methyl hydroxylamine using HATU/DIPEA.
The procedure for synthesizing C-5 alkyl-amino substituted quinazolinones commenced
from the Pd-catalyzed cyanation of C-5 bromide in 15a, b and subsequent hydrogenation of the
nitrile group to form compounds 30a, b (Scheme 4). The free amine in 30a was reacted with ethyl
2-bromothiazole-4-carboxylate, 31 to yield 32a, which was converted to final compound 33 using
a similar sequence of steps as described in previous schemes. For target compounds 37, 40a, b and
4
3, compound 29a was subjected to Boc-deprotection and nucleophilic aromatic substitution on
purine 8 to afford 34 (Scheme 4). Hydrogenation of the nitrile group in 34 to free amine was
utilized in synthesizing a variety of linkers using: a) amide-coupling with 35, b) reductive
amination with 38a, b, and c) substitution with 41, to form compounds 36, 39a, b and 42.
Subsequent conversion of alkyl esters to hydroxamic acids and removal of the THP protecting
groups afforded the target compounds, 37, 40a, b and 43.
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 117
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
C-5 alkyl-substituted quinazolinones with pyrimidine kinase hinge binding groups (46a-d)
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
12

Page 13 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were readily synthesized from 17b, c in four steps as shown in Scheme 5. The alkyne group in
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
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 117
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
7b, c was hydrogenated using Raney-Ni followed by Boc-deprotection and S Ar on
N
chloropyrimidines 44a-c to form compounds 45a-d. Subsequent conversion of methyl esters in
Scheme 4. Synthesis of C-5 alkyl-amino substituted quinazolinones [33, 37, 40 and 43]a
45a-
O
Cl
N
OH
NH
N
N
N
O
O
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
9
0
1
2
3
4
5
6
7
8
9
0
S
N
d
to hydroxamic
OEt
acids
OEt
H2N
S
THP
8
HN
Br
O
CN
O
N
O
N
N
O
N
S
(d), (e)
Br 31
N
N
N
N
HN
R1
R1
R1
O
N
(a)
(b)
(c)
N
N
NHBoc
NHBoc
NHBoc
HN
N
N
NHBoc
N
1
NH
1
1
30a, R = Et
1
1
5a, R = Et,
29a, R = Et, 98%
32a, 38%
33, 53%
1
1
1
30b, R = Me
5b, R = Me
29b, R = Me, 96%
O
N
O
N
CO2Me
OMe
OH
N
CN
O
N
O
H
O
N
HN
CN
O
N
HN
CO2H
35
O
N
O
N
8
N
HN
N
(
d)
(b), (f)
(e)
N
NHBoc
N
HN
N
HN
N
N
N
THP
O
N
N
N
N
NH
THP
29a
34, 81%
OMe
36, 87%
3
7, 31%
O
OR1
OH
NH
OEt
N
O
O
Z
or
Z
S
HN
O
N
HN
O
N
O
O
N
3
8a
38b
3
4
N
(e)
(b), (g)
HN
HN
N
N
N
N
N
N
N
THP
N
H
9a, R1 = Et, Z = thiazolyl, 65%
3
3
OEt
4
0a, 39%
40b, 41%
1
9b, R = Me, Z = phenylene, 94%
O
N
H
N
O
OEt
O
N
OH
N
N
N
N
HN
Cl
O
N
HN
O
N
4
1
3
4
N
(
b), (h)
(e)
N
HN
HN
N
N
N
N
N
N
THP
N
N
H
42, 82%
43, 57%
aReagents and Conditions: (a) Pd(PPh3)4 (0.05 equiv), Zn(CN)2 (1.2 equiv), DMF, 100 °C, 16h; (b) Raney Ni, H , MeOH, 20h; (c) 31 (2.0 equiv) DIPEA
2
(
4.0 equiv), DMF, 180C, MW, 30 min; (d) i) CF CO H (20.0 equiv), DCM, 3 h, rt; ii) 8 (1.5-2.5 equiv), triethyamine (3.0-5.0 equiv), EtOH, 4h, MW, 100
3 2
°
C; (e) i) LiOHH O (2.0 equiv), MeOH/H O, 10 h, rt; ii) NH OTHP (3.1 equiv), N-methyl morpholine (3.0 equiv) 3-(((ethylimino)methylene)amino)-N,N-
2 2 2
dimethylpropan-1-amine hydrochloride [EDC.HCl] (1.4 equiv), 1H-[1,2,3]triazolo[4,5-b]pyridin-1-ol [HOAT] (1.2 equiv), DMF, 16h rt; iii) CF CO H (20.0
3
2
equiv), DCM/MeOH, 20 h, rt; (f) 35 (1.5 equiv), HATU (1.2 equiv), DIPEA (3.0 equiv), DMF, 16h, rt; (g) 38 (1 equiv), cat. AcOH, DCE, 2h, then
NaBH(OAC) (3.0 equiv), 2h; (h) 41 (2.0 equiv), triethylamine (4.0 equiv), EtOH, 100 °C, MW, 3h.
3
using hydroxylamine afforded the target compounds 46a-d.
ACS Paragon Plus Environment
14

Page 15 of 117
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
C-5 amino-alkyl substituted quinazolinones (48a-o) were synthesized in a similar fashion
from compounds 26b, d, g, i by attaching the pyrimidine kinase binding moieties using S Ar
N
reactions on 44a, b, d-l and converting methyl esters in 47a-o to hydroxamic acids.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 117
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 5. Synthesis of C-5 substituted quinazolinones with pyrimidine hinge binding groups [46a-d and 48a-o]a
O
Z
OMe
O
O
OH
O
R2 Cl
Z
OMe
O
Z
N
O
N
N
H
4
4a-c
(a)
(b)
N
N
NHBoc
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
9
0
1
2
3
4
5
6
7
8
9
0
N
NHR2
NHR2
1
1
7b Z = (CH2)3
7c Z = phenylene
4
5a-d
46a-d
2
Z = (CH ) , R
=
2
3
NC
NC
NC
N
N
N
2
N
Z =
, R =
H N
N
NH2
N
NH2
N
NH2
H2N
N
NH2
2
4
6a
46b
46c
46d
HO
OMe
Z
OMe
Z
NH
O
R2 Cl
O
O
Z
NH
O
N
44a,b,d-l
NH O
NH
O
(
a)ii), (a)iii)
(b)
N
N
N
R1
R1
R1
N
N
NHR2
NHBoc
1
NHR2
2
2
2
6b, Z = (CH ) , R = Et
2 3
4
7a-o
48a-o
1
6d, Z = phenylene, R = Et
1
6g, Z = 2-pyridyl, R = Et
1
26i, Z = phenylene, R = Me
1
2
Z = (CH ) , R = Et, R =
2
3
1
2
Z =
, R = Et, R =
NC
N
N
NH2
NC
NC
NC
Cl
Cl
48a
N
N
N
N
N
N
N
H N
N
NH2
N
NH2
N
48d
NH2 H N
N
H N
N
NH2
2
2
2
4
8b
48c
48e
48f
48h
1
2
Z =
, R = Et, R =
N
NC
Cl
NC
NC
NC
N
N
N
N
N
H N
N
H N
N
NH2 F3C
N
NH2
N
NH2 HF C
2
N
NH2
Cl
H N
2
2
N
N
4
8i
48j
48k
48l
48m
2
48g
NC
NC
1
2
N
Z =
, R = Me, R =
N
H N
N
NH2
2
N
NH2
4
8n
48o
aReagents and Conditions: (a) i) 10 wt% Pd/C, H2 balloon, EtOAc, 20 h, rt; ii) CF CO H (20.0 equiv), DCM, 3 h, rt; iii) 44 (1.5-2.0 equiv), DIPEA (3.0-4.0
3
2
equiv), n-Butanol, 2-3h, MW, 130 °C, 38-88% (3 steps);(b) aq. 50 wt% NH OH (30.0 equiv), LiOHH O (1.1 equiv), MeOH/H O, 20h, 0 °C to rt; 36-82%.
2
2
2
For synthesis of quinazolinones 51 and 54, the nitrile group in compound 29b was
hydrogenated using Raney Ni/H to generate free amine, which was either coupled to 4-
2
(
methoxycarbonyl)benzoic acid 35 using HATU/DIPEA or subjected to S Ar reaction with ethyl
N
ACS Paragon Plus Environment
16

Page 17 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
-chloropyrimidine-5-carboxylate 41, to form compounds 49 and 52, respectively (Scheme 6).
Subsequent attachment of the pyrimidine kinase hinge binding group and transformation of alkyl
esters to hydroxamic acids yielded the target compounds 51 and 54.
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
Scheme 6. Synthesis of C-5 alkyl-amino substituted quinazolinones [51 and 54]a
O
N
O
N
CO Me
2
OH
Cl
OMe
O
N
N
N
H
O
O
N
OMe
CN
O
N
O
HN
HN
CO H
H N
N
44a
(c)
NH2
O
N
2
2
O
N
35
N
HN
O
N
(a), (b)
(d)
NHBoc
HN
N
NH2
HN
N
NH2
N
N
NHBoc
N
N
NH2
NH2
2
9b
49, 96%
50, 77%
51, 23%
H
N
O
N
OEt
O
OMe
O
N
OH
O
N
OEt
Cl
N
N
N
N
N
N
CN
O
N
Cl
HN
N
H2N
N
44a
(c)
NH2
HN
O
N
O
N
41
HN
O
N
N
(d)
N
(a), (e)
NHBoc
N
N
HN
N
NH2
HN
N
NH2
NHBoc
N
N
N
N
NH2
NH2
2
9b
52, 83%
53, 79%
54, 50%
a
Reagents and Conditions: (a) Raney Ni, H , MeOH, 20h; (b) 35 (1.5 equiv), HATU (1.2 equiv), DIPEA (3.0 equiv), DMF, 16h, ; (c) i) CF CO H (20.0 equiv),
2
3
2
DCM, 3 h, rt; ii) 44a (1.5 equiv), DIPEA (3.0 equiv), n-Butanol, 2-5h, MW, 130-150 °C; (d) aq. 50 wt% NH OH (30.0 equiv), LiOHH O (1.1 equiv), MeOH/H O,
2
2
2
2
0h, 0 °C to rt; (e) 41 (1.5 equiv), DIPEA (3.0 equiv), n-Butanol, 2h, MW, 130 °C.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 117
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
PI3K and HDAC: Enzyme Inhibition Activity and SAR. The dual inhibitors were screened
against PI3K and HDAC enzymes at Reaction Biology Corporation, Malvern, PA, USA. Initially
we compared the enzyme inhibitory activities of 4-substituted quinazolines vs 5-substituted
quinazolinones by screening against PI3K and HDAC 1-11 enzymes (Table 1). In general, 5-
substituted quinazolinones were consistently found to be more potent than the 4-substituted
quinazolines for PI3K and HDAC enzyme inhibition. As shown for a few examples in Table 1, 5-
substituted quinazolinones having 4- and 5-carbon alkylene linkers (19a, 19b) and amino-
propylene linker (28a), showed more potent PI3K/HDAC activities than their 4-substituted
quinazoline counterparts (10a, 10b, 14a). Additionally, the 5-substituted compounds showed high
potency and selectivity for HDAC6 inhibition, leading us to focus only on 5-substituted
quinazolinones.
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
Table 1. Enzyme Inhibitory Activities of 4-substituted quinazolines (10a, b, 14a) vs 5-substituted quinazolinones
a
(19a, b, 28a)
HO
HO
O
NH
NH
O
Z
O
N
Z
N
N
N
HN
HN
N
N
N
N
N
N
NH
N
H
1
0a Z = (CH2)4
19a Z = (CH2)4
19b Z = (CH2)5
10b Z = (CH2)5
1
4a Z = NH(CH2)3
28a Z = NH(CH2)3
IC50 (nM)
Compound
10a
HDACs
PI3K
49
1
2
3
4
5
6
7
8
9
10
11
5846
429
18190
1231
14100
7075
427
2169
36
1937
385
353
253
6194
33730
2073
16900
1502
23030
2089
9819
2155
74140
>105
1
9a
<5
70910
>105
23610
43700
8106
90810
24400
1
1
0b
9b
68
5390
388
10600
39
0.2
11290
>105
5
34850
14a
520
>105
>105
>105
431
ACS Paragon Plus Environment
18

Page 19 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
8a
46
3
1575
NT
3247
NT
1031
NT
9087
NT
37
1758
NT
NT
NT
3362
NT
PI-103
NT
NT
NT
NT
TSA
NT
NT
9
25
14
NT
NT
3
NT
371
NT
NT
49
61
3470
NT
TMP 269
NT
NT
NT
281
333
NT
108
NT
aCompounds were tested in singlet 10-dose IC50 mode with 3-fold serial dilution starting at 1M for PI3K and 10 M for HDAC1-
11 enzymes. Empty cells indicate no inhibition or compound activity that could not be fit to an IC50 curve. NT = Compound not
tested against enzyme. PI-103 was used as a control compound for PI3K, whereas Trichostatin A (TSA) and TMP 269 were used
as control compounds for HDAC enzymes.
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
Next, we explored the linker region on 5-substituted quinazolinones by incorporating a
variety of aryl, heteroaryl and amino-aryl groups on the PI3K core. Initially we screened against
PI3K isoforms (   ) and HDAC isoforms (1, 2, 3, 6, 8, and 10) at a single dose (1 M for
PI3Ks and 10 M for HDAC enzymes). As shown in Table 2, dual inhibitors with a variety of
linkers were synthesized and tested to establish SAR. Dual inhibitors with aryl (24a) and heteroaryl
linkers such as benzothiazole, (24c) showed high potency for PI3K and HDAC6, 8, whereas
benzyl linker (24b) only showed high PI3K activity. The use of amino-alkyl heterocycle linkers
such as 3-(methylamino) pyridine (28f) and 2-(methylamino) thiazole (33) showed good inhibition
for PI3K and HDAC6, but the corresponding alkyl-amino furan (28h) lost its potency against both
enzymes. Linkers containing basic amines (40a, b) exhibited modest to high HDAC6 activity (80%
for 40a, 99% for 40b), but were inactive against PI3Ks. Select dual inhibitors that showed high
single-point potency (% inhibition) in both PI3K and HDAC6 were subsequently tested in 10-
dose IC titration against PI3Ks (   ) and HDAC 1-11 isozymes.
5
0
Table 2. PI3K/HDAC Enzyme Inhibition (%) Activities of 5-substituted quinazolinones with purine as the hinge
binding group
a
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 117
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
Z
O
N
OH
NH
N
O
;
where Z =
HN
N
Linker
N
N
NH
OH
OH
NH
HO
NH
OH
NH
OH
NH
O
N
OH
O
O
NH
O
O
OH
NH
O
OH
NH
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
HN
O
O
N
O
S
S
N
Linker:
HN
S
N
HN
HN
33
HN
HN
28f
2
4a
24b
24c
28h
40a
40b
Inhibition (%) at 1 M
Inhibition (%) at 10 M
Compound
24a
PI3K
HDAC




1
2
3
6
8
10
84.5 ± 0.3
64.1 ± 1.0
84.7 ± 0.2
45.8 ± 2.7
18.1 ± 0.4
45.6 ± 1.0
1.8 ± 0.7
44.0 ± 0.6
37.5 ± 2.4
38.6 ± 1.4
4.4 ± 4.3
3.8 ± 1.3
6.8 ± 0.7
5.3 ± 0.3
-6.0 ± 0.3
81.6 ± 0.2
79.1 ± 0.02
88.0 ± 0.6
39.9 ± 5.0
48.0 ± 1.3
59.9 ± 0.4
-12.2 ± 0.7
-9.1 ± 3.4
94.7 ± 0.02
95.8 ± 1.3
92.6 ± 0.6
88.3 ± 0.5
61.2 ± 1.0
95.2 ± 0.1
56.0 ± 1.0
20.5 ± 0.7
63.3 ± 3.2
-0.9 ± 2.8
50.0 ± 2.3
20.5 ± 0.3
18.3 ± 6.7
5.2 ± 3.9
52.3 ± 0.4
11.5 ± 0.1
40.5 ± 1.7
16.3 ± 0.9
28.0 ± 3.6
4.1 ± 1.1
48.2 ± 0.9
-13.4 ± 0.3
20.9 ± 7.2
14.3 ± 1.6
-6.3 ± 3.2
2.0 ± 3.5
93.3 ± 0.3
70.7 ± 2.9
95.6 ± 0.4
90.5 ± 0.7
72.3 ± 0.03
89.1 ± 0.1
79.9 ± 0.7
99.2 ± 0.7
99.1 ± 0.1
70.1 ± 0.6
98.2 ± 0.3
87.2 ± 0.1
68.4 ± 3.3
58.4 ± 0.8
54.5 ± 0.3
93.5 ± 0.3
54.0 ± 0.3
20.0 ± 0.6
33.1 ± 0.1
-7.0 ± 1.6
17.0 ± 2.8
-4.6 ± 4.7
26.4 ± 0.7
57.2 ± 1.2
2
4b
4c
8f
8h
33
0a
0b
PI-103
2
2
2
4
24.3 ± 3.3
65.6 ± 2.0
30.4 ± 3.8
57.2 ± 5.7
-4.3 ± 1.1
19.1 ± 1.4
4
-6.8 ± 0.2
5
13
63
7
NT
18
NT
48
NT
40
NT
4
NT
NT
63
(IC50
)
TSA (IC50
)
NT
NT
NT
NT
519
aCompounds were tested in single dose duplicate at 1 M for PI3Ks and 10 M for HDAC enzymes. NT = Compound not tested
against enzyme. PI-103 was used as a control compound for PI3K, whereas Trichostatin A (TSA) was used as control compounds
for HDAC enzymes.
Table 3 shows the PI3K and HDAC enzyme potencies of the 5-substituted quinazolinone
with purine hinge binding group. The dual inhibitor containing the pentyl group as linker (19b)
exhibited high activity and selectivity for both PI3K (IC = 0.2 nM, >210-fold selectivity over
5
0
PI3K) and HDAC6 (IC = 13 nM, >65 fold-selectivity over HDAC1-5,7-11). Substitution
5
0
with a homo-benzyl (19c) or phenyl linker (24a) showed high PI3K inhibition but poor activity
against HDAC6. Swapping the homo-benzylic linker with 1-(methylamino) benzyl group (28d)
restored HDAC6 activity and selectivity along with good PI3K activity. Interestingly, substituting
the 1-(methylamino) benzyl linker in 28d with a 1-(methylamino)-3-methylbenzyl group in 28e
ACS Paragon Plus Environment
20

Page 21 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
maintained the PI3K activity but completely abolished the HDAC activity. A similar trend in
PI3K/HDAC6 potency was observed with (3-methylamino)pyridyl linker in 28f, where HDAC6
HN
HN
activity was
presumably
hydrogen-
lost,
due
His
His
N
N
H
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
Zn
to
O
H
N
Chelation decreases
HDAC inhibition
O
NH
O
O
H
bonding
of
N
interaction
pyridyl N-
N
the
2
8f
28g
atom with
Figure 5. Hydrogen bonding rationale for HDAC activity comparison in 28f and 28g
O-H group on hydroxamic acid, thereby impacting the chelation of hydroxamic acid with Zn in
the HDAC6 enzyme (Figure 5). To further support this hydrogen bonding rationale, we designed
and synthesized compound 28g with a (2-methylamino)pyridyl linker, where the possibility of H-
bonding interaction with the pyridyl N-atom and -OH group of hydroxamic acid was eliminated.
Gratifyingly, the HDAC6 potency was restored in 28g (IC = 9 nM) when compared with 28f
5
0
(
IC = 169 nM) (Figure 5).
50
Substituting the hydroxamic acid in 28d to N-methyl hydroxamic acid as zinc-binding
group in 28i, led to a 2-fold drop in PI3K potency and complete loss in activity against all HDACs
tested. The dual inhibitor containing N-methylbenzamide (37) as the linker resulted in high activity
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 117
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 selectivity in both PI3K and HDAC6. Compound 43 with a 2-(aminomethyl)-pyrimidine
linker also showed good potency and selectivity against PI3K and HDAC6 inhibition.
Table 3. PI3K/HDAC Enzyme Inhibitory Activities of 5-substituted quinazolinones with purine hinge binding
groupa
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
Z
O
N
OH
NH
O
N
;
where Z =
HN
N
Linker
N
N
NH
OH
NH
OH
NH
HO
O
OH
NH
OH
NH
OH
NH
OH
NH
OH
N
OH
NH
OH
NH
O
NH
O
O
O
O
O
O
O
N
O
N
Linker:
N
N
O
NH
NH
HN
HN
HN
HN
HN
19b
19c
24a
28d
28e
28f
28g
28i
37
43
IC50 (nM)
Compound
PI3K
HDAC




1
2
3
4
5
6
7
8
9
10
11
1
9b
9c
24a
8d
8e
8f
8g
28i
122
NT
NT
254
517
NT
392
387
266
422
4
286
NT
NT
43
0.2
5
1390
3030
2400
8060
>105
NT
13
889
362
5
856
3840
71
1240
1840
9000
NT
1
NT
NT
103
203
NT
135
162
234
995
81
0.7
47
41
43
26
81
0.6
49
8
NT
NT
NT
NT
1751
NT
NT
NT
3335
NT
NT
2
3925
7173
6699
2144
NT
229
250
6922
NT
14700
324
NT
2
2699
169
9
2
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
2
6067
10990
381
439
104
204
922
9996
2255
3
4
7
3
8434
67870
NT
10
6
23120
60
138
320
NT
445
NT
19890
1778
25480
NT
1428
NT
3827
NT
5349
NT
NT
19
PI-103
TSA
7
NT
19
NT
2
NT
NT
62
NT
45
NT
4826
NT
NT
NT
NT
NT
NT
NT
NT
NT
10
34
NT
NT
TMP 269
NT
NT
NT
244
278
NT
NT
aCompounds were tested in singlet 10-dose IC50 mode with 3-fold serial dilution starting at 1M for PI3K    and 10 M for
HDAC1-11 enzymes.Empty cells indicate no inhibition or compound activity that could not be fit to an IC50 curve. NT = Compound
not tested against enzyme. PI-103 was used as a control compound for PI3Ks, whereas Trichostatin A (TSA) and TMP 269 were
used as control compounds for HDAC enzymes.
ACS Paragon Plus Environment
22

Page 23 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Despite high activity and selectivity observed with several 5-substituted quinazolinones
against PI3K/HDAC6, most didn’t show expected anti-proliferative/cell kill activities when
tested against the NCI-60 panel (See section titled, “NCI-60 Human Tumor Cell Lines Screen”).
We hypothesized that their poor cellular potency might be due to low permeability (predicted by
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
3
PAMPA permeability) generally observed with these dual inhibitors. Table 4 shows the PAMPA
permeability values of selected 5-substituted quinazolinones with a purine hinge binding group.
Table 4. PAMPA Permeability data of selected 5-substituted quinazolinones with purine hinge binding groupa
PAMPA permeabilitya,b
Compound
9b
4a
4c
1
<1
2
1.2 ± 0.2
2.4 ± 0.1
6.4 ± 0.8
2.6 ± 0.1
1.3 ± 0.4
<1
2
28d
28e
3
3
3
7
4
0b
<1
<1
4
3
a
-6
Average of two runs ± standard deviation (S. D.). PAMPA permeability values were determined at pH 7.4 (10 cm/s).
bNo S.D. values provided if all PAMPA permeability values are less than 1.
To improve upon the cellular potencies of 5-substituted quinazolinones, we investigated a
new series of dual inhibitors by substituting the purine kinase hinge binding moiety with amino-
pyrimidine groups.52, 54-56 Table 5 shows the PI3K and HDAC enzyme inhibition data for selected
5
-substituted quinazolinones with amino-pyrimidine hinge binding groups. Compounds 46a,b
with 2,4-diamino-6-pyrimidine-5-carbonitrile and 2-amino-6-methylpyrimidine-5-carbonitrile
hinge binding groups, respectively, showed excellent PI3K activity (IC = 0.2-0.7 nM), but a
5
0
weaker HDAC6 potency (IC = 28-69 nM), when compared with compound, 19b from the purine
5
0
series (PI3K IC = 0.2 nM, HDAC6 IC = 13 nM). Compound 46d with a homo-benzyl linker
50
50
and 2,4-diamino-6-pyrimidine-5-carbonitrile hinge binding group was highly active against
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 117
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
PI3K IC = 0.2 nM) but showed a modest potency for HDAC6 (IC = 77 nM). Dual inhibitors
5
0
50
with amino-benzyl/amino-methylpyridyl linkers were explored with a variety of amino-pyrimidine
kinase hinge binders to establish SAR. Replacement of the pentyl linker in 46b with aminobutyl
linker in 48a with the same hinge binding group, showed high potencies for PI3K, but no
activity against HDACs. Compound 48b with 2,4-diamino-6-pyrimidine-5-carbonitrile hinge
binder showed moderate potency for PI3K but high potency and selectivity for HDCA6. The
PI3K potency was restored in 48c using 2-amino-6-methylpyrimidine-5-carbonitrile hinge binding
group, while maintaining good potency for HDAC6. Removing the nitrile group in 48c gave
compound 48d, which was inactive against PI3K. Dual inhibitors with amino-pyrimidines lacking
the 2-amino group in the hinge region (48e,f, g) showed high HDAC6 activity (IC = 3-4 nM),
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
but poor PI3K potency (PI3K IC = 60 nM for 48e, 133 nM for 48f, 189 nM for 48g). Similar
5
0
detrimental effects on PI3K potencies were observed when the C-5 nitrile group on the
aminopyrimidine was changed to Cl (48h, 48j) and the C-2 amino group was changed to Me (48i).
Replacement of the C-6 methyl/amino group in the kinase hinge to CF , cyclopropyl and CF H,
3
2
respectively, were also not good for PI3K activity (48k, 48l, 48m). Overall, only 2-amino-6-
methylpyrimidine-5-carbonitrile and 2,4-diamino-6-pyrimidine-5-carbonitrile hinge binding
groups were revealed as optimal hinge binders for achieving high potencies for both PI3K and
HDAC enzymes. While HDAC6 activity and selectivity obtained by using these amino-pyrimidine
hinge binders were similar to their purine hinge binder counterparts, less PI3K selectivity was
observed with aminopyrimidine series, with almost equipotent activities against PI3K and PI3K
1
in many examples. Swapping the ethyl group at R with a methyl group was tolerated well to obtain
compounds (48n, 48o) with single-digit nM activities against PI3K and HDAC6. Compounds
having 2,4-diamino-6-pyrimidine-5-carbonitrile as hinge binder with N-methylbenzamide and 2-
ACS Paragon Plus Environment
24

Page 25 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
aminomethyl)-pyrimidine linkers (51 and 54) also showed high PI3K/HDAC6 activity. In
general, dual inhibitors with aminopyrimidine hinge binders showed comparatively better
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
PAMPA permeability (Table 6) than their purine hinge binding counterparts. Additionally, 5-
substituted quinazolinones with amino-pyrimidine hinge binders showed potent cell activities
when tested in the NCI-60 panel (See section NCI-60 assays and Supplementary Information).
Table 5. PI3K/HDAC Enzyme Inhibitory Activities of 5-substituted quinazolinones with amino-pyrimidine
hinge binders
a
O
HO
N
L
O
N
H
N
R1
NHR2
4
6a-d, 48a-o, 51, 54
IC50 (nM)
PI3K
HDAC6 isoform
Compound
46a
L
R1
R2
b
selectivity
HDAC6




NC
N
Et
52
33
3
2
0.7
69
>13
>36
>22
H2N
N
NH2
NH2
NH2
NC
NC
N
N
4
6b
Et
Et
9
12
0.2
96
28
45
N
4
6c
5210 1769
N
NC
N
N
N
4
6d
Et
Et
Et
13
186
50
26
2
8
0.2
5
77
1238
4
>7
>2
H2N
N
NH2
NH2
NH2
NC
4
8a
279
842
N
NH
NC
48b
45
32
>73
H2N
N
HN
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 117
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
NC
N
48c
Et
Et
Et
47
9
7.0
12
4
>39
>58
N
N
NH2
HN
HN
HN
4
8d
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
N
NH2
NC
N
4
8e
444
752
60
4
>113
H2N
N
Cl
N
N
4
8f
Et
Et
Et
133
189
663
4
3
>5
H2N
N
N
HN
HN
HN
Cl
N
4
8g
>81
>15
H2N
Cl
N
48h
22
N
NH2
NC
N
4
8i
Et
Et
Et
Et
Et
Me
491
269
14
9
>20
>24
>13
>9
H2N
N
HN
HN
HN
HN
HN
HN
Cl
N
48j
H2N
N
NH2
NC
N
4
8k
52
47
27
6
F3C
N
NH2
NH2
NC
N
4
8l
N
NC
N
4
8m
222
4
390
4
>6
HF2C
N
NH2
NC
N
4
8n
42
159
>31
H2N
N
NH2
ACS Paragon Plus Environment
26

Page 27 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NC
N
48o
Me
Me
75
18
188
379
2
4
8
>15
>6
N
NH2
HN
NC
N
5
5
1
4
10
0.7
16
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
H2N
N
NH2
O
NH
NC
N
N
NH
N
Me
-
42
4
293
7
15
55
3
5
3
>25
-
H2N
N
NH2
PI-103
-
-
NT
TSA
TMP269
-
-
-
-
-
-
NT
NT
NT
NT
NT
NT
NT
NT
6
NT
-
-
aCompounds were tested in singlet 10-dose IC50 mode with 3-fold serial dilution starting at 1M for PI3K   
and 10 M for HDAC1-11 enzymes. Empty cells indicate no inhibition or compound activity that could not be fit to
an IC50 curve. NT = Compound not tested against enzyme. PI-103 was used as a control compound for PI3Ks, whereas
b
Trichostatin A (TSA) and TMP 269 were used as control compounds for HDAC enzymes. Selectivity of HDAC6
inhibition over other HDAC isoforms (1, 2, 3, 4, 5, 7, 8, 9, 10 and 11). Please see Table S1 in Supplementary
Information for full IC50 data against HDAC1-11.
Table 6. PAMPA Permeability data of selected 5-substituted quinazolinones with amino-pyrimidine hinge
binding groups
PAMPA
permeability
Compound
a, b
4
6b
6c
8b
8c
8d
48f
8g
8h
25.3 ± 0.8
22.7 ± 7.6
4.9 ± 0.1
4
4
4
12.0 ± 0.7
15.7 ± 5.6
50.5 ± 12.6
77.5 ± 11.7
30.2 ± 3.2
4
4
4
4
8i
12.4 ± 1.2
4
8j
149.1 ± 2.1
a
-6
b
Average of two runs ± standard deviation (S. D.). PAMPA permeability values were determined at pH 7.4 (10 cm/s). No S.D.
values provided if all PAMPA permbeability values are less than 1.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 117
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
Molecular modeling: HDAC6 Binding and Selectivity. The dual inhibitor 19b was docked into
5
7
the solved crystal structure of hHDAC6 (PDB ID: 5EDU.pdb) [Figure 6A]. The geometry of the
protein around the 19b hydroxamic acid head group mirrors that of Vorinostat (see Figure 3). The
aliphatic linker chain, in similar fashion, traverses a hydrophobic region of the pocket. The
capping group, which is composed of a quinazolinone ring system, on the other hand, forms a
hydrogen bond to Ser568 via the quinazolinone carbonyl. This hydrogen bonding interaction on
the rim of the pocket is distinctly different than that made by Vorinostat.
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
HDAC6 selectivity. Using the procedure described in methods, we identified Ser568 as the only
residue on HDAC6 that could be responsible for the HDAC selectivity profile of 19b. As shown
in Figure 6A, Ser568 in HDAC6 forms a hydrogen bond with the quinazolinone carbonyl on 19b.
In the case of HDACs-1, 2, 3, and 8, Ser568 becomes Asp568 (Figure 6B). The negatively charged
carboxyl group on Asp568 is a hydrogen bond acceptor and does not form a hydrogen bond with
the quinazolinone carbonyl on 19b [Figure S1-S3 (Supporting Information)]. Rather the Asp568
carboxyl repulses the quinazolinone carbonyl and swings away from the ligand. The nearby
presence of Glu567 [HDAC1,2, Figure S1 (Supporting Information)] and Asp567 [HDAC3,
Figure S2 (Supporting Information)] contributes to this process as these residues have negatively
charged carboxyl side chains.
A
B
Residue Pair
HDAC-6 (Asp567, Ser568)
HDAC-1, 2 (Glu567, Asp568)
HDAC-3 (Asp567, Asp568)
HDAC-8 (Tyr567, Asp568)
HDAC-4, 5, 7, 9 (Asp567, Thr568)
HDAC-10 (Asp567, Ala568)
HDAC-11 (Pro567, Asn568)
ACS Paragon Plus Environment
28

Page 29 of 117
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 6. A) Compound 19b bound to HDAC6 (PDB ID: 5EDU.pdb). B) Variation in amino acid residues at positions
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
67 and 568 across different HDACs.
In the case of HDAC8 the residue adjacent to Asp568 is Tyrosine which contains a neutral
phenyl side chain [Figure S3 (Supporting Information)]. As a result, the conformational shift in
the carboxyl group on position 568 is less significant than for HDACs-1,2,3. Nevertheless, the H
bond to the quinazolinone carbonyl is still broken. In the case of HDACs-4,5,7, and 9, Threonine
occupies position 568. Thr568 contains a hydrogen bond donor (hydroxyl), but the hydroxyl group
does not interact with the quinazolinone carbonyl on the ligand. That is because the extra methyl
group on the Thr568 side chain clashes with nearby Gly619 and prevents the OH group from
obtaining a favorable hydrogen bonding position with respect to the quinazolinone carbonyl. In
fact, it is easier for the Thr568 OH group to hydrogen bond with the carboxyl side chain of nearby
Asp567 [Figure S4 (Supporting Information)]. In the case of HDAC10, Ser568 becomes Ala568
resulting in an obvious loss of hydrogen bonding to the quinazolinone carbonyl on the ligand
[Figure S5 (Supporting Information)]. In HDAC11, Ser568 becomes Asn568. Although
Asparagine has a hydrogen bond donating nitrogen group, its side chain is one carbon too long to
adopt a reasonably orientation for hydrogen bonding to the quinazolinone carbonyl on the ligand
[Figure S6 (Supporting Information)]. The 19b-HDAC interaction energies relative to 19b-
HDAC6 were computed for each of the scenarios described above. These interaction energies are
completely consistent with the idea that there is one less hydrogen bond between 19b and HDACs-
ACS Paragon Plus Environment
29