Quinazolin-4-one derivatives as selective histone deacetylase-6 inhibitors for the treatment of Alzheimer's disease

Article abstract of DOI:10.1021/jm400564j

Novel quinazolin-4-one derivatives containing a hydroxamic acid moiety were designed and synthesized. All compounds were subjected to histone deacetylase (HDAC) enzymatic assays to identify selective HDAC6 inhibitors with nanomolar IC₅₀ values. (E)-3-(2-Ethyl-7-fluoro-4-oxo-3-phenethyl-3,4- dihydroquinazolin-6-yl)-N-hydroxyacrylamide, 4b, is the most potent HDAC6 inhibitor (IC₅₀, 8 nM). In vitro, these compounds induced neurite outgrowth accompanied by growth-associated protein 43 expression, and they enhanced the synaptic activities of PC12 and SH-SY5Y neuronal cells without producing toxic or mitogenic effects. Several of the compounds dramatically increased nonhistone protein acetylation, specifically of α-tubulin. Some of the more potent HDAC6 inhibitors decreased zinc-mediated β-amyloid aggregation in vitro. N-Hydroxy-3-(2-methyl-4-oxo-3-phenethyl-3,4-dihydro- quinazolin-7-yl)-acrylamide, 3f, the most promising drug candidate, selectively inhibits HDAC6 (IC₅₀, 29 nM), practically does not affect human ether-a-go-go-related membrane channel activity (IC₅₀ >10 μM) or cytochrome P450 activity (IC₅₀ >6.5 μM) in vitro, and significantly improves learning-based performances of mice with β-amyloid-induced hippocampal lesions.

Full text of DOI:10.1021/jm400564j

Subscriber access provided by Universitätsbibliothek | Technische Universität München
Article
Quinazolin-4-one Derivatives as Selective Histone Deacetylase-6
Inhibitors for the Treatment of Alzheimer’s Disease
Chao-Wu Yu, Pei-Teh Chang, Ling-Wei Hsin, and Ji-Wang Chern
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400564j • Publication Date (Web): 01 Aug 2013
Downloaded from http://pubs.acs.org on August 5, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Quinazolin-4-one Derivatives as Selective
Histone Deacetylase-6 Inhibitors for the
Treatment of Alzheimer’s Disease
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†
,#,♦
†,#, ♦
†,#
,†,#,§
ChaoꢀWu Yu,
PeiꢀTeh Chang,
LingꢀWei Hsin, and JiꢀWang Chern*
†
§
School of Pharmacy, College of Medicine, Department of Life Science, College of Life
#
Science, and Center for Innovative Therapeutics Discovery, National Taiwan University,
3
3, LinSen South Road, Taipei 100, Taiwan, Republic of China
*
jwchern@ntu.edu.tw
ABSTRACT
Novel quinazolinꢀ4ꢀone derivatives each containing a hydroxamic acid moiety were
designed and synthesized. All compounds were subjected to histone deacetylase (HDAC)
enzymatic assays to identify selective HDAC6 inhibitors with nanomolar IC50 values.
(E)ꢀ3ꢀ(2ꢀEthylꢀ7ꢀfluoroꢀ4ꢀoxoꢀ3ꢀphenethylꢀ3,4ꢀdihydroquinazolinꢀ6ꢀyl)ꢀNꢀ
hydroxyacrylamide, 4b, is the most potent HDAC6 inhibitor (IC , 8 nM). In vitro, these
5
0
compounds induced neurite outgrowth accompanied by growthꢀassociated protein 43
expression, and enhanced the synaptic activities of PC12 and SHꢀSY5Y neuronal cells
without producing toxic or mitogenic effects. Several of the compounds dramatically
increased nonꢀhistone protein acetylation, specifically that of αꢀtubulin. Some of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 69
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
more potent HDAC6 inhibitors decreased zincꢀmediated βꢀamyloid aggregation in vitro.
NꢀHydroxyꢀ3ꢀ(2ꢀmethylꢀ4ꢀoxoꢀ3ꢀphenethylꢀ3,4ꢀdihydroꢀquinazolinꢀ7ꢀyl)ꢀacrylamide, 3f,
the most promising drug candidate, selectively inhibits HDAC6 (IC , 29 nM), practically
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
does not affect human etherꢀa goꢀgoꢀrelated membrane channel activity (IC >10 ꢀM) or
50
cytochrome P450 activity (IC >6.5 ꢀM) in vitro, and significantly improves learningꢀ
50
based performances of mice with βꢀamyloidꢀinduced hippocampal lesions.
INTRODUCTION
Histone deacetylase 6 (HDAC6) has a different activity profile than do other HDACs,
1
is expressed primarily in the cytoplasm, and is abundant in hippocampal neurons.
HDAC6 is a nonꢀhistone deacetylase, being active instead against acetylated cytoplasmic
proteins, e.g., αꢀacetylꢀtubulin (αꢀAcꢀtubulin), HSP90, and cortactin, thereby regulating
2
ꢀ6
microtubule stability, gene transcription, and cell motility, respectively.
HDAC6
expression increases when neurons are damaged, and the HDAC6 levels in the
7
hippocampi of patients with Alzheimer's disease (AD) seem to significantly increase,
which correlates with decreased neuronal survival. Upregulation of HDAC6 decreases αꢀ
8
, 9
Acꢀtubulin levels and consequently leads to microtubule dysfunction. Additionally, the
level of phosphorylated tau protein, which highly correlates with dysfunctional
7
microtubules, is attenuated by HDAC6 inhibitors. Moreover, selective inhibition of
HDAC6 by siRNA or a specific inhibitor promotes neurite extension or protects neurons
7
, 10
from oxidative stress, respectively.
However, the physiopathological implications of
ACS Paragon Plus Environment

Page 3 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
HDAC6 inhibition in conjunction with regulation of acetylation of nonꢀhistoneꢀtype
proteins as related to neuronal functions remains mostly to be explored.
During past investigations on HDAC inhibition of neurodegeneration, inhibitors
selective for HDAC6 were hardly examined, although nonꢀselective inhibitors have been
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
shown to have antiꢀneurodegeneration activity. However, nonꢀselective or partially
1
2
selective HDAC inhibition usually leads to undesirable biological responses, including
1
3
14
15
fatigue, nausea/vomiting, and cardiotoxicity. For these reasons, we assumed that
HDAC6 would be an appropriate target for the development of antiꢀAD agents. Selective
drugs directed against HDAC6 activity for clinical use remain to be developed, and the
contribution of HDAC6 inhibition in protecting against AD remains to be elucidated.
A perusal of the literature revealed that several HDAC6ꢀselective inhibitors have
been reported. Tubacin (Figure 1), discovered in 2003 during a highꢀthroughput
screening of 7392 compounds, decreases cell motility without destabilizing microtubules
and was, at that time, considered to be a potential antimetastatic and antiangiogenic
1
6, 17
agent.
In 2010, Tubastatin A (Figure 1), a hydroxamic acid derivative with a benzyl
linker and a tricyclic cap, was developed via structureꢀbased design and homology
modeling. At 15 nM, Tubastatin A displays HDAC6 inhibitory activity and a 50–2000ꢀ
1
8
fold selectivity versus other HDAC isozymes. In addition to hydroxamic acidꢀ
containing compounds, several thiolꢀbased compounds for HDAC6 inhibition are known,
1
9ꢀ21
but they are less stable than are the hydroxamic acid derivatives.
To develop new inhibitors of HDAC6 as drug candidates, we designed and
synthesized a series of compounds with structures based on that of quinazolinꢀ4ꢀone. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 69
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
quinazolinꢀ4ꢀone moiety is found in many bioactive compounds, which possess an
22ꢀ25
26
antimalarial, antidiabetic, anticonvulsant, anticancer,
or neuroprotective activity.
However, quinazolinꢀ4ꢀoneꢀbased compounds containing a hydroxamic acid moiety, as
do Tubacin and Tubastatin A, have not been assessed for HDAC activity.
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
Using wellꢀestablished synthetic protocols that employ palladiumꢀcatalyzed cross
coupling with microwave irradiation, we produced quinazolinꢀ4ꢀone derivatives each
containing a hydroxamic acid (Figure 2). We designed and synthesized these quinazolinꢀ
4
ꢀone derivatives so that they contain linkers of different lengths connected to the
hydroxamic acid moiety, which can bind zinc, to determine which of these compounds
has optimal HDAC6 selectivity. We evaluated the HDACꢀinhibitory activities of these
compounds against several HDAC isozymes and assessed the more potent of these
compounds for their abilities to stimulate in vitro neurite outgrowth, for their activity in
an in vitro antiꢀzincꢀmediated βꢀamyloid (Aβ) aggregation assay, and for their ability to
prevent Aβꢀmediated neurodegeneration in vivo.
RESULTS AND DISCUSSION
Chemistry
Commercially available chlorosubstituted anthranilic acids (1a-d) were the starting
materials for the synthesis of 3a-d, respectively (Scheme 1). Compounds 1a-d were each
cyclized with acetic anhydride or acetyl chloride to give the corresponding
benzoxazinone intermediate, which was then opened and recyclized with aniline to afford
2
7
2aꢀd. These cyclized quinazolinꢀ4ꢀones were each coupled to methyl or ethyl acrylate
ACS Paragon Plus Environment

Page 5 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
by a Heck reaction in the presence of Pd (dba) , P(tꢀBu) , and Cy NMe in toluene to
2
3
3
2
28
afford the corresponding methyl or ethyl ester, which was subsequently converted to
aꢀd with hydroxylamine.
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 1. Synthesis of 3aꢀd
a, b, c (2a)
or
O
d (2b, 2d)
O
N
6
O
N
5
or
5
8
5
e, f
6
OH
6
a, b (2c)
N
R1
N
O
R1
NH2
Me
3
Me
8
HO NH
1a, R = 6-Cl
1
2a, R = 5-Cl
1
2b, R = 6-Cl
2c, R = 7-Cl
2d, R = 8-Cl
3a-d
1b, R = 5-Cl
1
1
1c, R = 4-Cl
1
1
1
d, R = 3-Cl
1
1
a
Reagents and conditions: (a) Ac O, reflux 4 h; (b) PhNH , AcOH, reflux 16 h; (c) i)
2
2
piperidine, ethyl acetate, rt, 36 h; ii) SiO , ethyl acetate, rt, 24 h; (d) i) P(OPh) , AcCl,
2
3
pyridine, MW 250 W, 15 min, ii) PhNH , MW 250 W, 10 min; (e) Pd (dba) ,P(tꢀBu) ,
2
2
3
3
Cy NMe, methyl acrylate or ethyl acrylate, toluene, 100 °C, 50 h; (f) NH OHꢁHCl, KOH,
2
2
MeOH, 0 °C to rt, 5–11 h. (MW, microwave irradiation)
Compounds 3e-n with various substitutions at positions 2 and 3 were prepared
according to Scheme 2. Compound 1c was cyclized in two steps or in a oneꢀpot reaction
with one of a variety of alkynoyl chlorides and alkyl amines in the presence of P(OPh)3,
2
9
with microwave irradiation to give the quinazolinꢀ4ꢀones 2e-n. Compounds 3e-n were
synthesized from 2e-n according to steps e and f of Scheme 1.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 69
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
Scheme 2. Synthesis of 3eꢀn
O
O
N
O
N
R2
R1
OH a,b or c
R2
R1
d, e
N
N
H
N
Cl
NH2
HO
Cl
O
1c
2
2
2
2
2
e, R = Me, R = CH Ph
3e, R = Me, R = CH Ph
1
2
2
1 2 2
f , R = Me, R = (CH ) Ph
3f , R = Me, R = (CH ) Ph
1 2 2 2
1
2
2 2
g, R = Me, R = ethyl-3-indolyl
3g, R = Me, R = ethyl-3-indolyl
1
2
1 2
h, R = H , R = (CH ) Ph
3h, R = H , R = (CH ) Ph
1 2 2 2
1
2
2 2
i , R = Et , R = (CH ) Ph
3i , R = Et , R = (CH ) Ph
1
2
2 2
1 2 2 2
2j , R = Et , R = (CH ) Ph
3j , R = Et , R = (CH ) Ph
1 2 2 3
1
2
2 3
2k, R = cPr, R = (CH ) Ph
3k, R = cPr, R = (CH ) Ph
1
2
2 2
1 2 2 2
2
2
2
l , R = iPr, R = (CH ) Ph
3l , R = iPr, R = (CH ) Ph
1 2 2 2
1
2
2 2
m, R = Et, R = (CH ) Ph-4OMe
3m, R = Et, R = (CH ) Ph-4OMe
1
2
2 2
1 2 2 2
n , R = Et, R = (CH ) Ph-4F
3n , R = Et, R = (CH ) Ph-4F
1 2 2 2
1
2
2 2
a
Reagents and conditions: (a) (R CO) O, reflux 4 h; (b) R NH , AcOH, reflux 16 h; (c) i)
1
2
2
2
P(OPh) , R COCl, pyridine, MW 250 W, 15 min, ii) R NH , MW 250 W, 10 min; (d)
3
1
2
2
Pd (dba) , P(tꢀBu) , Cy NMe, methyl acrylate or ethyl acrylate, toluene, 100 °C, 50 h; (e)
2
3
3
2
NH OHꢁHCl, KOH, MeOH, 0 °C to rt, 5–11 h. (MW, microwave irradiation)
2
ACS Paragon Plus Environment

Page 7 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The syntheses of 4a-c and 5a-d are shown in Schemes 3 and 4, respectively.
Iodinated 1f and 1g were synthesized from the anthranilic acids 1c and 1e in the presence
of KI, NaIO , and NaCl. The quinazolinꢀ4ꢀones 2p and 2q were cyclized from 1f and 1g
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Scheme 3). 3ꢀChloroꢀ4ꢀfluoroꢀaniline was coupled to propionyl chloride and
subsequently brominated with Br to give the propionamide intermediate that was then
2
carbonylated with carbon monoxide catalyzed by Pd(OAc) and Xantphos to afford the
2
Nꢀsubstituted anthranilic acid that was then cyclized to give compound 2o (Scheme 3). A
Heck reaction with microwave irradiation was used to individually couple 2o-q with
ethyl acrylate to afford the corresponding ethyl esters that were then converted to the
hydroxamic acids 4a-c (Scheme 3). Negishi reactions with freshly prepared benzylzinc
bromide were used to convert 2r, 2i, and 2s to benzylic esters. Hydrolysis of their
benzylic esters with LiOH and subsequent amide coupling with Oꢀprotected
hydroxylamine furnished Oꢀbenzyl hydroxamic acids. Finally, reductive deprotection of
the benzyl moiety furnished the desired hydroxamic acids 5a-c (Scheme 4).
a
Scheme 3. Synthesis of 4aꢀc
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 69
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
O
O
O
I
5
b
c-e, b
F
a
OH
6
OH
NH2
N
R2
R1
R1
NH2
N
Cl
NH2
8
1c, R = Cl
1e, R = F
1f, R = Cl
1g, R = F
1
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
1
1
2
o, R = 6-F , 7-Cl
2
2p, R = 6-I , 7-Cl
2
2
q, R = 6-I , 7-F
2
f, g
H
4a, R = F, R =
N
3
4
O
N
OH
O
R3
R4
H
N
N
4b, R =
3
OH , R = F
4
O
O
H
N
4
c, R =
3
, R = Cl
4
OH
a
Reagents and conditions: (a) NaIO , KI, NaCl, AcOH/H O (9:1), rt, 8 h; (b) i) P(OPh) ,
4
2
3
EtCOCl, pyridine, MW 250 W, 15 min, ii) Ph(CH ) NH , MW 250 W, 10 min; (c)
2
2
2
EtCOCl, TEA, CH Cl , rt, 1 h; (d) Br , AcOH, rt, 7.5 h; (e) Pd(OAc) , Xantphos, CO,
2
2
2
2
TEA, toluene, 40 h; (f) Herrmann's palladacycle, [(tꢀBu) PH]BF , ethyl acrylate,
3
4
Cy NMe, DMF, MW 120 W, 25–50 min; (g) NH OHꢁHCl, KOH, MeOH, 0 °C to rt, 7 h.
2
2
(
MW, microwave irradiation)
a
Scheme 4. Synthesis of 5aꢀc
O
O
N
O
N
6
5
5
OH
NH2
a
6
b, c, d, e
N
N
X
X
R5
5a, 6-R
5
3
8
5b, 7-R
5
R =
5c, 8-R
5
1
h, R=5-I
5
2
2
r, R= 6-I
i, R= 7-Cl
2s, R= 8-Cl
H
1c, R=4-Cl
N
OH
1
d, R=3-Cl
O
a
Reagents and conditions:(a) i) P(OPh) , EtCOCl, pyridine, MW 250 W, 15 min, ii)
3
Ph(CH ) NH , MW 250 W, 10 min; (b) pꢀEtO CC H CH ZnBr/THF, Pd (dba) , [(tꢀ
2 2
2
2
6
4
2
2
3
Bu) PH]BF , NMP, MW 200 W, 15 min; (c) LiOH( 2.5 M), MeOH/THF (1:5), rt, 4 h; (d)
3
4
ACS Paragon Plus Environment

Page 9 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH OBnꢁHCl, EDC, HOBt, TEA, CH Cl or DMF, rt, 24 h; (e) 10% Pd/C, H , MeOH, rt,
2
2
2
2
5
h. (MW, microwave irradiation)
The synthesis of compound 7 is depicted in Scheme 5. Suzuki coupling of 4ꢀ
methoxycarbonyl phenylboronic acid and 2s with microwave irradiation furnished ester 6.
Hydrolysis of 6 followed by amide coupling with Oꢀbenzylhydroxylamine and reductive
deprotection resulted in the hydroxamic acid 7.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 5. Synthesis of 7
O
O
N
O
N
a
b, c, d
N
N
N
N
Cl
2s
CO Me
2
HN
OH
O
6
7
a
Reagents and conditions:(a) palladacycle, [(tꢀBu) PH]BF , pꢀMeO CC H B(OH) , DBU,
3
4
2
6
4
2
Cs CO , DMF, reflux, MW 120 W, 25 min; (b) 2.5 M LiOH, MeOH/THF (1:5), rt, 8.5 h;
2
3
(
c) NH OBnꢁHCl, EDC, HOBt, TEA, CH Cl , rt, 24 h; (d) 10% Pd/C, H , THF/MeOH
2
2
2
2
(1:3), rt, 5 h. (MW, microwave irradiation)
Profiling Inhibition of HDACs by 3a-d
Compounds 3a-d, containing a hydroxamic acid moiety connected via a vinyl linker to
different positions of a quinazolinꢀ4ꢀone skeleton were subjected to inhibitory assays
using different HDACs (Table 1). Compounds 3aꢀd preferentially inhibited HDAC6 and
HDAC8; whereas, Trichostatin A (TSA, a panꢀHDAC inhibitor), which served as the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 69
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
reference inhibitor, showed no selectivity as expected. Compound 3b with a hydroxamic
acid moiety at position 6 is the most potent of these HDAC6 inhibitors (IC , 32 nM),
5
0
and 3c with a hydroxamic acid moiety at position 7 is threefold less active than is 3b.
Compounds 3b and 3c are more active toward HDAC6 than are 3a and 3d, which have a
hydroxamic acid moiety at position 5 or 8, respectively. Possibly, the extended linear
structures of 3b and 3c can be better accommodated in the active site of HDAC6 (i.e.,
have a smaller steric effect) than can the angular structures of 3a and 3d. Conversely, the
relative positions of the hydroxamic acid moiety and the 2ꢀmethylꢀ3ꢀphenylꢀquinazolinꢀ
4ꢀone do not correlate with the HDAC8 inhibitory activities of 3a-d. Compound 3d with
the hydroxamic acid moiety at position 8 is more active against HDAC8 (IC , 0.42 ꢀM)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
than it is against HDAC6 (IC , 0.69 ꢀM), which suggests that the activeꢀsite pocket of
5
0
HDAC6 might be less flexible than that of HDAC8.
Compound 3c has a superior HDAC6/HDAC8 selectivity index than does 3b. Notably,
HDAC8 seems to possess a nonꢀrigid catalytic pocket, which would allow its
conformation to change substantially upon inhibitor binding and thereby better
30
accommodate the different geometries of the inhibitors. Given their relative activities
against the different HDACs, 3c, the inhibitor with the best HDAC6/HDAC8 selectivity
index among seriesꢀ3 compounds, containing an extended linker at position 7, was
chosen as the lead compound to further increase the selectivities of hydroxamic acidꢀ
containing quinazolinꢀ4ꢀone derivatives for HDAC6.
To investigate the effects of substitutions at positions 2 and 3 of quinazolinꢀ4ꢀone with
the entryway of the HDAC6 active site, the compounds 3eꢀn were synthesized. We
ACS Paragon Plus Environment

Page 11 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
assumed that modifications at these positions might impact the selectivity of the
inhibitors for the HDAC isozymes. Substituents longer than that of the 3c phenyl, i.e.,
benzyl (3e) and phenethyl (3f), were individually added at position 3 of quinazolinꢀ4ꢀone,
and the derivatives containing these substituents were better inhibitors of HDAC6 (IC ,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
2
4 nM and 29 nM, respectively, Table 2) than was 3c. Both compounds are less selective
for HDAC11, but they have increased selectivity for HDAC1/8 (Table 2). When position
was substituted with an ethylꢀ3ꢀindolyl moiety (3g), the inhibitory activity of the
3
quinazolinꢀ4ꢀone derivative against the tested HDACs improved in comparison with that
of 3e, suggesting that an additional interaction(s), perhaps a hydrogen bond involving the
indole ring and each HDAC, was responsible.
Next, the methyl group at position 2 of 3f was replaced with different substituents to
produce 3h-i, and 3k-l (Table 2). Interestingly, 3i, with an ethyl group at position 2, is
~2.6ꢀfold more active than is 3f against HDAC6 (IC50 values of 11 nM and 29 nM,
respectively). In addition, 3i has improved selectivity for HDAC6 in comparison with its
selectivity for HDAC1 (108ꢀfold), HDAC8 (107ꢀfold), and HDAC11 (268ꢀfold).
Conversely, replacing the methyl group at position 2 of 3f with a hydrogen (3h; IC , 35
5
0
nM) or a cyclopropyl moiety (3k; IC , 41 nM) did not substantially change the
50
selectivity in comparison with that of 3f for HDAC6 (Table 2). However, 3l, which
contains an isopropyl group at position 2, has an IC value of 13 nM. Further elongation
50
of the 3ꢀphenethyl group of 3i to a 3ꢀphenpropyl (insertion of an additional methylene
group) to form 3j causes the IC value to increase threeꢀfold from 11 nM to 33 nM.
5
0
However, 3m and 3n, which contain the electronꢀdonating group OMe or the hydrogen
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 69
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
bondꢀacceptor F, respectively, at the 3i phenethyl pꢀposition have reduced activity against
HDAC6 (IC values of 41 nM and 43 nM, respectively) in comparison with that of 3i.
5
0
Surprisingly, both 3m and 3n have increased activity against HDAC8 leading to a
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
decreased selectivity for HDAC6.
a
Table 1. Enzyme inhibition data for 3a–d against class I, II, and IV HDACs
O
5
6
N
O
7
N
CH3
8
HO NH
IC (nM)
Selectivity index
HDAC6/ HDAC6/
5
0
Linker
positio
n
Class I
Class II
Class
IV
Cpd
HDA HDAC
C1
HDA HDA HDAC1 HDAC
1
8
2
C8
C6
0
11
3
3
3
3
a
5
6
7
8
18600 >10000
1060
1920 >10000 >10000
10
65
56
70
6
<1
19
b
c
2090
4890
7480
1610
609
3880
420
32
88
5140
5220
1350
17500
44
d
48500 >10000
690
29800 >10000
<1
b
TSA
5
5
129
0.9
9
17.3
143
57
c
c
c
c
c
c
c
Tubustatin A
16400 >30000
854
15 >30000 >30000
1093
a
Compounds were tested in the 10ꢀdose IC50 mode in duplicate with threefold serial
dilutions starting at 10 ꢂM. The IC values are the means of at least two experiments
50
with intraꢀand interꢀassay variations of <10%.
b
TSA (a nonselective HDAC inhibitor) was used as the positive control.
c
The IC values of Tubustatin A (a HDAC6ꢀselective inhibitor, using as a positively
5
0
HDAC6ꢀspecific control) are cited from refꢀ18.
Using 3i as the new lead compound, compounds 4a and 4b were synthesized by
affixing a fluorine atom at position 6 or 7, respectively (Table 3). These modifications
ACS Paragon Plus Environment

Page 13 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were made assuming that a fluorine atom might enhance the binding affinity of a
31
molecule for its biological target. In comparison with the HDAC6 inhibitory activity of
3
i, 4a, containing the fluorine atom at position 6, did not substantially improve the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
activity (IC , 14 nM). However, it improved the selectivity of HDAC6 in comparison
5
0
with its selectivities for the other HDACs. Because a hydroxamic acid moiety at position
or 7 improved antiꢀHDAC6 activity and selectivity, we switched the positions of the
fluorine atom and the hydroxamic acid moiety in 4a to create 4b, which has an IC of 8
6
5
0
nM, the best value discussed so far. In addition, 4b is more selective for HDAC6 than it
is for HDAC1 and HDAC11 (243ꢀfold and 585ꢀfold more selective, respectively). Finally,
replacing the fluorine atom with a chlorine atom at position 7 to produce 4c resulted in
the loss of activity against all HDACs.
To enhance the flexibility and size of the hydroxamic acid linker, the vinyl of 3i was
replaced with a benzylic linker to form 5a-c (Table 4). Interestingly, 5a with the benzylic
linker at position 6 has an IC value of 11 nM for HDAC6, which is the same as that of
50
3i. Although the selectivity indexes of 5a for HDAC6 versus HDAC1 and for HDAC6
versus HDAC2 are of the same magnitude as those for 3i, those for HDAC6 versus
HDAC8 and HDAC11 decreased to 9ꢀ and 60ꢀfold, respectively. Furthermore, for 5b
with a benzylic linker at position 7, its HDAC6 IC value is 9 nM and its selectivity
5
0
index for HDAC6 versus HDAC1 is >500. However, the selectivity indexes for HDAC6
versus HDAC8 and HDAC11 by 5b were only 19 and 147, respectively. Although
moving the benzylic side chain to position 8 to form 5c resulted in an HDAC6 IC value
5
0
of 79 nM, which is a sevenꢀfold decrease in inhibitory activity in comparison with that of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 69
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
5a, 5c still had an improved selectivity for HDAC6 versus HDAC1 (633ꢀfold) and versus
HDAC11 (470ꢀfold), whereas the HDAC6/8 selectivity index was only four.
Because 5c shows better selectivity for HDAC6 than for the other HDACs except
HDAC8, it was of considerable interest to study the effect of a rigid linker on HDAC8
inhibition. Thus, compound 7 with a phenyl linker at position 8 was synthesized.
However, it inhibits only 50% of the HDAC6 and HDAC8 activities when present at 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
ꢀM and does not inhibit the other HDACs, which lends support to our conclusion that an
angular configuration might not be beneficial for HDAC6 inhibitory activity. Conversely,
the less flexible hydroxamic acid moiety at position 7 in the series 3 and 4 compounds
leads to an inhibition that is more specific for HDAC6 than for HDAC8 than does
changing the linker to a benzyl (series 5 compounds) as it enhanced HDAC6 selectivity
in comparison with its selectivity for the other HDACs, with the exception of HDAC8.
Effects of the HDAC6-Selective Inhibitors on Neurite Outgrowth and Synaptic
Activity
The effects of the HDAC6ꢀselective inhibitors on neurite outgrowth and synaptic
activity were assessed on undifferentiated PC12 and SHꢀSY5Y neuronal cells. Cells were
treated with each inhibitor at concentrations between 100 nM and 25 ꢀM. The
morphologies of the differentiated neuronal neurite outgrowth caused by some of the
seriesꢀ3, seriesꢀ5 compounds, and 4b at a concentration of 10 ꢀM for the PC12 and SHꢀ
SY5Y cells are shown in Figure 3 and Figure 4, respectively. TSA, 3e, 3h, 3k, 3l, and 3n
induced moderate neurite outgrowth for both types of cells; whereas, 3c, 3f, 3i, 3j, 4b, 5a,
ACS Paragon Plus Environment

Page 15 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5b, and 5c induced substantial neurite outgrowth, and 3m triggered neurite outgrowth
only in PC12 cells. The EC values for these HDAC6-selective inhibitors in stimulating
50
neurite outgrowth and increasing synaptic activity are reported in Table 5. The potent
HDAC6ꢀselective inhibitors 3c, 3f, 3g, 3i, 3j, 4b, 5a, 5b, and 5c are superior inducers of
neurite outgrowth and increased synaptic activity. However, three other potent HDAC6ꢀ
selective inhibitors did not perform as well as might have been expected. Compound 3e
only increased synaptic activity, and 3m and 3n had no notable effects on neurite
outgrowth and synaptic activity. The less selective HDAC6 inhibitors 3h, 3k, and 3l had
no effect on neurite outgrowth. Compounds 3h and 3k had a weak effect on synaptic
activity, but 3l increased synaptic activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Given these results, the hydroxamic acidꢀbased quinazolinꢀ4ꢀone derivatives
containing a phenethyl moiety at position 3 and with a methyl (3f) or an ethyl group (3i)
at position 2 appear to be the most effective inducers of neurite outgrowth and synaptic
activity. Additionally, the presence of a vinyl (3i) or benzyl (5b) linker connecting the
hydroxamic acid moiety to position 7 of the quinazolineꢀ4ꢀone did not interfere with the
neurotrophic effect. Moreover, compounds with IC50 values <100 nM for HDAC6
inhibition and with selectivity indexes for HDAC6 versus HDAC1, 2, and 8 that are >40
more favorably induced neurite outgrowth and synaptic activity. We also characterized
the effects of the HDAC6ꢀselective inhibitors on the expression level of growthꢀ
associated protein 43 (GAPꢀ43), which is involved in neurite development and elongation,
3
2
synapse formation, and sprouting. An increase in GAPꢀ43 expression also leads to
3
3
enhanced learning in mice. Seven of the compounds were chosen according to their
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 69
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
efficacies as HDAC6ꢀselective inhibitors and neurite outgrowth inducers to examine their
effects on GAPꢀ43 expression (Figure 5). All seven compounds induced GAPꢀ43
expression in a doseꢀdependent manner and did not affect the expression of βꢀactin,
which served as the internal control. To our surprise, compounds 3c, 3f, 3j, 4b, and 5c
induced greater levels of GAPꢀ43 than did 3i, 5b, and TSA. It should be noted that 3c, 3f,
and 4b could induce higher GAPꢀ43 levels at low concentration (0.1 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
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 17 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Table 2. Enzyme inhibition data for 3eꢀn against HDAC1, 6, 8, and 11
O
R2
N
H
N
HO
N
R1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
IC (nM)
Selectivity index
HDAC HDAC
5
0
class I
class
IV
class II
6/1
6/8
Cp R1 R2
d
HDAC HDAC HDAC HDAC HDAC
1
2
8
6
24
29
15
11
3
3
e
f
Me PhCH
Me Ph(CH
Me 3ꢀindolylꢀ
ethylꢀ
2
1450
1880
758
4000
6450
2210
1270
1750
1590
2440
4080
1210
60
53
)
65
51
60
106
2
2
3g
3h H
Ph(CH
)
2
)
2
)
3
495
1190
3200
1220
4810
>1200
0
600
1180
1330
35
11
33
730
2950
5590
14
108
97
17
107
40
2
2
2
3
3
i
j
Et Ph(CH
Et Ph(CH
b
b
b
b
3
k cPr Ph(CH
3l iPr Ph(CH
3m Et 4ꢀCH
Ph(CH2)2
n Et 4ꢀFꢀPh(CH
Compounds were tested in the 10ꢀdose IC mode in duplicate with threefold serial
dilutions starting at 10 ꢂM. The IC values are the means of at least two experiments
with intraꢀand interꢀassay variations of <10%.
Not determined.
2
)
2
433
389
1810
ND
ND
ND
41
13
41
ND
ND
11
30
44
ND
ND
b
b
b
b
)
2
ND
2
Oꢀ
4780
594
3910
14
3
3
2
)
2
2110
6880
891
43
4250
49
21
a
5
0
50
b
a
Table 3. Enzyme inhibition data for 4a-c and 5a-c against HDAC1, 6, 8, and 11
O
R3
N
R4
N
Et
IC (nM)
Selectivity index
5
0
class I
HDAC HDAC HDAC
class II class IV
HDAC6/ HDAC6/
Cp
R3
F
R4
HDAC1
HDAC6
1
1
8
d
1
2
8
H
4a
N
>1000
0
HO
3
000
1400
766
14
8
3740
4680
214
243
100
96
O
H
4b
N
F
HO
1940
5680
O
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 69
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
H
4c
N
Cl
>1000 >1200
1880
HO
747 >30000
>13
3
O
0
0
a
Compounds were tested in the10ꢀdose IC mode in duplicate with threefold serial
dilutions starting at 10 ꢂM. The IC values are the means of at least two experiments
with intraꢀand interꢀassay variations of <10%.
5
0
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
9
0
1
2
3
4
5
6
7
8
9
0
a
Table 4. Enzyme inhibition data for 5a-c against HDAC1, 6, 8, and 11
O
5
6
N
7
N
Et
8
HN
HO
O
Side chain
position
IC (nM)
Selectivity index
HDAC6 HDAC6/8
/1
50
class I
class II class IV
HDAC HDAC2 HDAC8 HDAC6 HDAC1
1
1
5a
6
7
8
1050
4850
>50000 >10000
3610
ND
99
173
282
11
9
79
656
1320
37100
95
539
>633
9
19
4
b
5b
5c
a
Compounds were tested in the 10ꢀdose IC mode in duplicate with threefold serial
dilutions starting at 10 ꢂM. IC values are means of at least two experiments with intraꢀ
and interꢀassay variations of <10%.
Not determined.
5
0
5
0
b
α-Ac-Tubulin and Ac-Histone Levels in PC12 Neurons
Because we demonstrated that selective inhibition of HDAC6 by the quinazolinꢀ4ꢀone
derivatives positively affected neurite outgrowth and because nerve growth factor (NGF)
34
induces neurite outgrowth via αꢀtubulin acetylation, we assessed the effects of our
HDAC6 inhibitors on the levels of αꢀAcꢀtubulin and Acꢀhistone H3, for which
3
5
acetylation is controlled by constitutively expressed nuclear HDACs. βꢀActin
expression served as the control. Given their positive induction of neurite outgrowth, the
ACS Paragon Plus Environment

Page 19 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
effects of 3c, 3f, 4b, 5b and 5c on the levels of αꢀAcꢀtubulin and Acꢀhistone H3 from
PC12 cells were characterized. Compounds 3c, 4b and 5c induced substantial increases in
αꢀtubulin acetylation (3~6 fold increase) in a doseꢀdependent fashion without affecting
the acetylation of histone H3 and βꢀactin; whereas, 3f increased the amounts of αꢀAcꢀ
tubulin and Acꢀhistone H3 significantly at 10 ꢀM concentrations (Figure 6). Compound
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
b increased both of the acetylated proteins but its acetylating effect was more
preferentially on αꢀtubulin TSA, the nonꢀselective panꢀHDAC inhibitor, increased both
.
Acꢀhistone H3 and αꢀAcꢀtubulin levels. Given the effects of the quinazolinꢀ4ꢀone
derivatives on neuriteꢀoutgrowth morphology (Figure 3), inhibition of αꢀAcꢀtubulin
deacetylation appears to result from their selective inhibition of HDAC6, therefore
inhibition of αꢀAcꢀtubulin deacetylation might result in the accumulation of αꢀAcꢀtubulin
3
6
in neurites and stimulate neurite outgrowth.
Effects of the HDAC6-Selective Inhibitors on Aβ Aggregation In Vitro
37
Because zinc ions are involved in Aβ aggregation, we hypothesized that a
compound containing a zincꢀbinding hydroxamic acid moiety might prevent zincꢀ
mediated Aβ aggregation. Several seriesꢀ3 compounds were selected to assess their
effects on Aβ aggregation in the presence or absence of zinc (Table 6). Compounds 3f,
3
g, 3h, and 3i blocked Aβ aggregation more effectively in the presence of zinc than did
the compound 3c and 3e; whereas compounds 3a, 3b, and 3d did not alter the level of
zincꢀmediated Aβ aggregation. As expected, none of the compounds affected Aβ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 69
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
aggregation in the absence of zinc. The results from this study indicate that hydroxamic
acid containing HDAC6 inhibitors might provide an alternative mechanism by preventing
zincꢀmediated Aβ aggregation for the treatment of AD.
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
Effects on In Vivo Anti-Aβ-induced Degeneration
To further examine the antiꢀneurodegenerative aspects of the HDAC6ꢀselective
inhibitors, the effects of the inhibitors on learningꢀimpaired mice with hippocampal
lesions induced by zincꢀmediated Aβ aggregates were assessed using the accelerating
3
8
rotarod test as a motorꢀlearning task. Given their relative efficacies in such as GAPꢀ43
expression, neurite outgrowth, and synaptic activity assays and potential as drug leads, 3f
and 3c were initially selected for in vivo efficacy. After introducing the lesion, the
experimental mice underwent daily intraperitoneal injections of 3f or 3c at a
concentration of 10 mg/kg for 30 days, whereas the control mice were injected with only
the vehicle. The learningꢀbased performances of the mice injected with 3f or 3c improved
with time when compared with the performances of mice injected only with vehicle
(Figure 7a). Furthermore, we found the level of acetylated αꢀtubulin was extensively
increased by 3c and 3f within the hippocampus region of lesion mice (Figure 7b). These
results support that the learningꢀbased performance improvement was due to the
inhibition of HDAC6 in mice brain.
In terms of their biological profiles, i.e., antiꢀHDAC6 potency and selectivity,
inhibition of zincꢀmediated Aβ aggregation, and induction of neurite outgrowth, 3f is
ACS Paragon Plus Environment

Page 21 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
more efficient than is 3c in recovering the learningꢀbased performances of the
hippocampusꢀlesioned mice. However, inhibition of HDAC6ꢀrelated mechanisms other
than the aforementioned ones might also be involved in the performance recovery of the
mice by 3c and 3f. Overexpression of HDAC6 appears to destructively regulate
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
39
intracellular microtubuleꢀdependent transportation by reducing αꢀAcꢀtubulin levels.
4
0, 41
This phenomenon is considered a sign of neurodegeneration
and is supported by the
observation that inhibition of HDAC6 reverses the transport deficit in peripheral nerve
4
2
43
disorders in vivo and in vitro degeneration caused by of the presence of Aβ or by
Q109
44
knockꢀin of the hhdh
CAG gene. Accordingly, the improved learningꢀbased
behaviors of these HDAC6ꢀselective inhibitors might be partly a result of upregulated
motor activity mediated by reconstructing the microtubuleꢀdependent intracellular
transportation system that may be another consequence of HDAC6 inhibition in neurons.
Additionally, the quinazolinone moiety had been reported to inhibit poly ADP ribose
2
6
polymerase 1 (PARPꢀ1) leading to its neuroprotective effect. Therefore, the effects on
PARP activity of quinazolinone moiety bearing compounds 3c, 3f, 3i, 3j, 4b, 5b, and 5c
were estimated. We found that our quinazolinone moiety bearing compounds didn’t
inhibit or affect PARP activity at concentration up to 25 ꢀM.
In Vitro Drug Safety Assessments
The most active compounds according to their biological effects, from series 3 and 5,
3
f and 5c, were evaluated for their abilities to inhibit cytochrome P450 (CYP) and the
human etherꢀa goꢀgoꢀrelated (hERG) channel to assess their effects on liver and cardiac
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 69
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
function, respectively, in vitro. The inhibitory effects of 3f and 5c on CYP subtypes
(CYP3A4, 2C9, 2C19, 2D6, and 1A2) and hERG at concentrations up to 10 ꢀM are
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
shown in Figure 8. Compound 3f had a very weak effect on CYP3A4/2C9 and none on
CYP2C19/2D6/1A2; whereas, compound 5c inhibited CYP3A4/2C9/2C19 in the
submicro to micromolar range and had no effect on CYP 2D6/1A2. Compound 3f did not
inhibit hERG activity (>10 ꢀM), but 5c weakly inhibited hERG activity (IC , 4.7 ꢀM).
5
0
Thus, 3f is considered to be nonꢀcardiotoxic. Additionally, 3f, 5c, and all other
compounds are nonꢀtoxic to PC12 and SYꢀSY5Y neuronal cells (>25 ꢀM) and Vero cells
(>25 ꢀM) (data not shown). This study showed that 3f is somewhat safer than 5c and that
our HDAC6ꢀselective inhibitors are not inherently toxic, which provides further evidence
that they may prove useful as lead compounds for the development of drugs against AD.
CONCLUSIONS
Novel quinazolinꢀ4ꢀone hydroxamic acid derivatives with optimized substitutions at
the 2/3 positions were developed and were found to be selective inhibitors of HDAC6
activity. These compounds are not toxic to neurons and Vero cells, inhibit zincꢀmediated
Aβ aggregation, induce neurite outgrowth, and synaptic activity in vitro. Combining their
multiple actions as antiꢀneurodegeneration agents, Nꢀhydroxyꢀ3ꢀ(2ꢀmethylꢀ4ꢀoxoꢀ3ꢀ
phenylꢀ3,4ꢀdihydroꢀquinazolinꢀ7ꢀyl)ꢀacrylamide (3c) and Nꢀhydroxyꢀ3ꢀ(2ꢀmethylꢀ4ꢀoxoꢀ
3
ꢀphenethylꢀ3,4ꢀdihydroꢀquinazolinꢀ7ꢀyl)ꢀacrylamide
(3f)
induced
learning
enhancements in hippocampusꢀlesioned mice caused by βꢀamyloid aggregates. The
results of the CYP450 and hERG activity assays that were used as measures of liver and
ACS Paragon Plus Environment

Page 23 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cardiac function, respectively, indicated that 3f should not impair these functions. Further
chemical optimization of these novel HDAC6ꢀselective inhibitors may lead to a drug that
can reverse or decrease the progression of neurodegenerative disorders.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Effects of seriesꢀ3 , -4, and ꢀ5 compounds on inducing neurite outgrowth and
synaptic activity in PC12 and SHꢀSY5Y cells
aꢀc
EC (ꢂM)
EC (ꢂM)
50
50
Neurite outgrowth
Synaptic activity
Compound
PC12
SHꢀ
SY5Y
11.9
>25
9.2
PC12
SHꢀ
SY5Y
12.7
8.2
3c
9.2
20.9
7.3
15.9
8.2
3e
3f
6.5
6.8
3
3
3
h
i
j
>25
5.3
6.6
10.5
11.5
7.2
18.4
7.2
15.0
11.0
11.7
>25
15.0
>25
>25
8.6
14.5
7.2
4.9
0.9
7.0
10.0
>25
6.7
20.6
22.2
4.3
d
3k
ND
3l
5.3
20.6
13.9
3.3
3
m
3
4
n
b
5
5b
5c
a
9.4
9.3
6.2
18.4
9.8
8.1
2.9
3.6
6.7
4.1
5.5
8.3
a
Compounds were tested in triplicate at eight dose levels 0, 0.1, 0.5, 1, 5, 10, 15, and 25
ꢀ
M. The EC values are the means of at least three experiments for both assays. In all
cases, the standard deviation was <12%.
5
0
b
NGF was used as the positive control for both assays.
c
At a concentration of 25 ꢂM, the compounds did not cause neuronal proliferation or
neuronal death.
d
ND, not determined.
Table 6. Effect on βꢀamyloid (Aβ) aggregation of hydroxamic acid containing
2+ a
compounds in the presence or absence of Zn .
IC (ꢂM) for Aβ aggregation
5
0
Compound
detected by the thioflavin S
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 69
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
a
assay
2
+
2+
+
Zn
–Zn
3
a
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
3b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
c
18.4
>50
18.1
9.5
3d
3
e
3
f
3
g
7.4
3
h
5.5
3i
6.8
b
EDTA
Nicotinic
7.7
28.4
13.1
c
43.7
acid
a
Compounds were tested in the eightꢀdose IC mode in triplicate with twofold serial
dilutions starting at 50 ꢂM. The IC values are the means of at least three experiments
for both assays. In all cases, the standard deviation was <10%.
Chelator control. Nonꢀchelator control.
5
0
50
b
c
EXPERIMENTAL SECTION
General Information Concerning Synthetic Methods. Hꢀ and CꢀNMR spectra of the
compounds in CDCl or DMSOꢀd were recorded using a Bruker spectrometer (DPX200,
1
13
3
6
AV400). Infrared (IR) spectra were recorded on a Nicolet iS5 FTꢀIR Spectrometer. Mass
spectrometry was performed using a Bruker spectrometer (Esquire 2000) with
electrospray ionization. Column chromatography used Merck silica gel (40–60 ꢂm).
Melting points were measured using a BOXꢀ6420 instrument (Laboratory Devices Inc)
that was not calibrated. Elemental analyses for C, H, and N were performed using a
ACS Paragon Plus Environment

Page 25 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
HERAEUS VarioELꢀIII or a Thermo Flash 2000 instrument, and the experimental values
were within ±0.4% of their expected theoretical values. Microwave irradiation used a
Discover instrument (CEM Corporation).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-Amino-4-chloro-5-iodobenzoic acid (1f)
Procedure A (Iodination) with the Synthesis of 1f as the Example
Potassium iodide (9.48 g, 57.11 mmol) in H O (20 mL) was added dropwise into AcOH
2
(180 mL) containing 2ꢀaminoꢀ5ꢀchlorobenzoic acid (10 g, 57.11 mmol), NaIO (12.22 g,
4
5
7.13 mmol), and NaCl (6.66 g, 113.96 mmol) over a 2 h period. The mixture was stirred
at rt for 8 h, after which it was poured into ice water (300 mL), and the reaction was
quenched with excess Na S O . Then, the solution was extracted with CH Cl (6 × 50
2
2
3
2
2
mL). The acidity of the extracted solution was adjusted to pH 5 with 1 N NaOH, and the
solution was decolorized with NaHSO . Then, the CH Cl layer was evaporated to yield a
3
2
2
brown solid, which was washed with excess water to afford 1f as brown solid (14.08 g,
1
83% yield). R = 0.50 (EtOAc/hexane = 1: 1); H NMR (400 MHz, DMSOꢀd ) δ, 7.00 (s,
f
6
1
3
1
H), 8.07 (s, 1H); C NMR (100 MHz, DMSOꢀd ) δ, 77.7, 111.0, 116.3, 141.7, 141.9,
6
1
51.8, 167.5;
2-Amino-4-fluoro-5-iodobenzoic acid (1g) Procedure A was followed to give 1g as
1
brown solid. Yield, 66%; R = 0.37 (EtOAc/hexane = 1:2); H NMR (400 MHz, DMSOꢀ
f
13
d ) δ, 6.60 (d, J = 11.2 Hz, 1H), 8.01 (d, J = 7.6 Hz, 1H); C NMR (100 MHz, DMSOꢀ
6
d6) δ, 62.2 (d, J = 28 Hz), 102.0 (d, J = 26 Hz), 109.2, 141.8 (d, J = 6 Hz), 153.4 (d, J =
1
3 Hz, 1H), 163.9 (d, J = 244 Hz), 167.6;
5
-Chloro-2-methyl-3-phenyl-3H-quinazolin-4-one (2a)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 69
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
6ꢀChloroanthranilic acid (6 g, 34.27 mmol) mixed with Ac O (25 mL) was refluxed for 4
2
h. The resulting mixture was diluted with excess hexane and filtered to provide a light
yellow solid. The solid was dried and then mixed with AcOH (25 mL). Aniline (3.3 mL,
36.14 mmol) was added into the mixture, and it was refluxed for 16 h. Next, the mixture
was evaporated to obtain a brown residue. The residue was mixed with piperidine (20
mL) and EtOAc (50 mL) and was stirred at rt for 36 h. Then the solution was evaporated
to yield a red residue. Silica gel (10 g) was added into EtOAc (30 mL) that contained the
residue, and the mixture was then stirred at rt for 24 h. The resulting mixture was filtered
and washed with EtOAc to afford a red filtrate. The filtrate was concentrated and several
drops of hexane were then added to afford 2a as a beige crystalline solid (5.83 g, 64%).
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
R = 0.5 (EtOAc /hexane = 1:2); mp, 157–159 °C (recrystallized from EtOAc/hexane,
f
1
beige plate crystal); H NMR (400 MHz, DMSOꢀd ), δ, 2.07 (s, 3H, CH ), 7.42–7.44 (m,
6
3
13
2
H, ArH), 7.49–7.59 (m, 5H, ArH), 7.27 (t, J = 8 Hz, 1H, ArH); C NMR (100 MHz,
DMSOꢀd ), δ, 23.9, 117.4, 126.2, 128.4, 128.7, 128.9, 129.5, 132.6, 134.3, 137.7, 149.8,
6
+
155.3, 159.2; ESIMS(+), m/z = 271.0 [M+H] . Anal. Calcd for C H ClN O: C, 66.55;
15
11
2
H, 4.10; N, 10.35. Found: C, 66.59; H, 3.99; N, 10.10.
-Chloro-2-methyl-3-phenylquinazolin-4(3H)-one (2b)
Procedure B (Cyclization) with the Synthesis of 2b as the Example
ꢀChloroanthranilic acid (6g, 34.27 mmol) mixed with triphenyl phosphite (11.1 mL,
6
5
41.08 mmol) in pyridine (25 mL) was added into acetyl chloride (3.7 mL, 51.33 mmol)
and irradiated by microwave (250 W) with refluxing for 15 min. The resulting mixture
was added into aniline (4.7 mL, 51.47 mmol) and heated by microwave irradiation (250
ACS Paragon Plus Environment

Page 27 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
W) with refluxing for 10 min. The reaction mixture was then diluted with EtOAc (50 mL)
and washed with H O (3 × 50 mL). The organic phase was held at rt to precipitate 2b,
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
which was isolated by filtration as a white solid (5.32g, 57%). R = 0.56 (EtOAc /hexane
f
1
=
1:1); mp, 181–183 °C; H NMR (400 MHz, DMSOꢀd ) δ, 2.10 (s, 3H), 7.43–7.45 (m,
6
2
H), 7.51–7.58 (m, 3H), 7.66 (d, J = 8.8 Hz, 1H), 7.83 (dd, J = 8.8, 0.4 Hz, 1H), 7.98 (d,
13
J = 2.4 Hz, 1H); C NMR (100 MHz, DMSOꢀd ) δ, 24.0, 121.7, 125.1, 128.2, 128.8,
6
1
29.0, 129.5, 130.5, 134.5, 137.5, 145.9, 155.0, 160.3.
7
-Chloro-2-methyl-3-phenyl-3H-quinazolin-4-one (2c)
Procedure C (Cyclization) with the Synthesis of 2c as the Example
ꢀChloroanthranilic acid (6 g, 34.27 mmol) mixed with Ac O (25 mL) was refluxed for 4
4
2
h. The resulting mixture was diluted with excess hexane and filtered to afford a light
yellow solid. The solid was dried, mixed with AcOH (25 mL), added into aniline (3.3
mL, 36.14 mmol), and refluxed for 16 h. The product mixture was added into H O (50
2
mL) to afford 2c as white sand crystals. Yield, 86%; R = 0.38 (EtOAc/hexane = 1:2);
f
4
5
1
mp, 173–175 °C (lit 173–175 °C); H NMR (400 MHz, DMSOꢀd ) δ, 2.11 (s, 3H),
6
7
.25–7.27 (m, 2H), 7.41 (dd, J = 8.4, 2 Hz, 1H), 7.51–7.59 (m, 3H), 7.68 (d, J = 2 Hz,
1
3
1H), 8.18 (d, J = 8.4 Hz, 1H); C NMR (100 MHz, DMSOꢀd ) δ, 24.1, 119.3, 125.8,
6
1
26.7, 128.3, 129.1, 129.6, 137.6, 139.1, 148.4, 156.2, 160.7.
8
-Chloro-2-methyl-3-phenyl-3H-quinazolin-4-one (2d)
Procedure C was followed to give 2d as beige needle crystal. Yield, 89%; R = 0.63
f
1
(
EtOAc/hexane = 1:1); mp 253–255 °C (recrystallized from EtOAc); H NMR (400
MHz, DMSOꢀd ) δ, 2.15 (s, 3H), 7.48–7.59 (m, 6H), 7.97 (d, J = 7.6 Hz, 1H), 8.04 (d, J
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 69
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
3
= 8 Hz, 1H); C NMR (100 MHz, DMSOꢀd ) δ, 24.3, 122.2, 125.4, 126.7, 128.2, 129.0,
6
+
1
29.6, 130.1, 134.5, 137.6, 143.7, 155.6, 160.8; ESIMS(+), m/z = 271.0 [M+H] . Anal.
Calcd for C H ClN O: C,66.55; H, 4.10; N, 10.35. Found: C, 66.77; H, 4.12; N, 10.47.
1
5
11
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
3-Benzyl-7-chloro-2-methyl-3H-quinazolin-4-one (2e)
Procedure C was followed to give 2e as beige sand crystal. Yield, 65%; R = 0.38 (EtOAc
f
1
/
hexane = 1:2); mp, 115–116 °C (recrystallized from EtOAc); H NMR (400 MHz,
DMSOꢀd ) δ, 2.49 (s, 3H), 5.37 (s, 2H), 7.20–7.35 (m, 5H), 7.49 (br d, J = 8.4 Hz, 1H),
6
13
7
.62 (br s, 1H), 8.11 (d, J = 8.4 Hz, 1H); C NMR (50 MHz, DMSOꢀd ) δ, 22.9, 46.4,
6
118.5, 125.6, 126.2, 126.5, 127.2, 128.3, 128.6, 136.1, 139.0, 148.0, 156.6, 160.8;
+
ESIMS(+), m/z = 285.0 [M+H] . Anal. Calcd for C H ClN O: C, 67.49; H, 4.60; N,
1
6
13
2
9
.84. Found: C, 67.22; H, 4.50; N, 9.87.
7-Chloro-2-methyl-3-phenethyl-3H-quinazolin-4-one (2f) Procedure C was followed
to give 2f as colorless cube crystal. Yield, 51%; R = 0.3 ( EtOAc/hexane = 1:2); mp,
f
1
1
33–135 °C (recrystallized from MeOH); H NMR (400 MHz, DMSOꢀd ) δ, 2.46 (s, 3H),
6
2.95 (dd, J = 8.0, 7.6 Hz, 2H), 4.20 (dd, J = 8.0, 7.6 Hz, 2H), 7.21–7.32 (m, 5H), 7.49
13
(
dd, 1H, J = 8.4, 2.0 Hz), 7.60 (d, J = 2.0 Hz, 1H), 8.09 (d, J = 8.4 Hz, 1H); C NMR (50
MHz, DMSOꢀd ) δ, 22.7, 33.4, 45.7, 118.7, 125.6, 126.4, 126.5, 128.2, 128.5, 128.7,
6
+
1
38.1, 138.8, 148.0, 156.5, 160.4; ESIMS(+), m/z = 299.0 [M+H] . Anal. Calcd for
C H ClN O: C, 68.34; H, 5.06; N, 9.38. Found: C, 68.56; H, 5.07; N, 9.57.
17
15
2
7-Chloro-3-[2-(1H-indol-3-yl)-ethyl]-2-methyl-3H-quinazolin-4-one (2g) Procedure C
was followed to give 2g as beige sand crystal. Yield, 55%; mp, 218–220 °C
1
(
recrystallized from CH Cl ); H NMR (400 MHz, DMSOꢀd ) δ, 2.46 (s, 3H), 3.08 (dd, J
2
2
6
ACS Paragon Plus Environment

Page 29 of 69
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
= 7.6, 7.6 Hz, 2H), 4.24 (dd, J = 8.0, 7.2 Hz, 2H), 6.97 (dd, J = 7.6, 7.2 Hz, 1H), 7.07 (dd,
J = 7.6, 7.2 Hz, 1H), 7.18 (br s, 1H), 7.35 (d, J = 8 Hz, 1H), 7.50 (d, J = 8.4 Hz, 1H),
1
3
7
.60–7.62 (m, 2H), 8.13 (d, J = 8.8 Hz, 1H), 10.88 (br s, 1H); C NMR (100 MHz,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DMSOꢀd ) δ, 22.7, 23.5, 45.2, 110.4, 111.5, 118.1, 118.4, 118.8, 121.1, 123.2, 125.5,
6
+
1
26.4, 127.0, 128.2, 136.2, 138.8, 148.1, 156.6, 160.5; ESIMS(+), m/z = 338.1 [M+H] ;
Anal. Calcd for C H ClN O: C, 67.56; H, 4.77; N, 12.44. Found: C, 67.43; H, 4.73; N,
19 16
3
1
2.28.
7
-Chloro-3-phenethylquinazolin-4(3H)-one (2h) Procedure C was followed to give 2h
1
as white sand crystal. Yield, 35%; mp, 151–153 °C; H NMR (400 MHz, DMSOꢀd ) δ,
6
3
.00 (dd, J = 7.6, 7.2 Hz, 2H), 4.19 (dd, J = 7.6, 7.2 Hz, 2H), 7.18–7.28 (m, 5H), 7.54
(dd, J = 8.4, 2.0 Hz, 1H), 7.67 (d, J = 2.0 Hz, 1H), 8.13 (d, J = 8.8 Hz, 1H), 8.18 (br s,
1
3
1H); C NMR (100 MHz, DMSOꢀd ) δ, 34.1, 47.4, 120.2, 126.2, 126.5, 127.2, 128.1,
6
1
28.4, 128.7, 137.7, 138.8, 148.9, 149.2, 159.5; Anal. Calcd for C H ClN O: C, 67.49;
16 13 2
H, 4.60; N, 9.84. Found: C, 67.28; H, 4.54; N, 9.76.
7
-Chloro-2-ethyl-3-phenethyl-3H-quinazolin-4-one (2i)
Procedure B was followed to give 2i as white needle crystal. Yield, 34%; mp, 118–120
1
°
C; H NMR (400 MHz, CDCl ) δ, 1.24 (t, J = 7.2 Hz, 3H), 2.60 (q, J = 7.2 Hz, 2H), 2.95
3
(dd, J = 8.0, 7.6 Hz, 2H), 4.20 (dd, J = 8.4, 7.2 Hz, 2H), 7.14–7.26 (m, 5H), 7.30–7.33
13
(
m, 1H), 7.57 (d, J = 1.6 Hz, 1H), 8.12 (dd, J = 8.4, 1.2 Hz, 1H); C NMR (100 MHz,
CDCl ) δ, 10.2, 27.1, 33.7, 44.4, 117.8, 125.5, 125.96, 125.99, 127.1, 127.7, 127.8, 136.7,
3
+
1
39.2, 147.1, 157.9, 160.5; ESIMS(+), m/z = 335.1 [M+Na] ; Anal. Calcd for
C H ClN O: C, 69.12; H, 5.48; N, 8.96. Found: C, 69.00; H, 5.43; N, 8.85.
18
17
2
ACS Paragon Plus Environment