Identification of Novel Resorcinol Amide Derivatives as Potent and Specific Pyruvate Dehydrogenase Kinase (PDHK) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b00565

Pyruvate dehydrogenase kinases (PDHKs) promote abnormal respiration in cancer cells. Studies with novel resorcinol amide derivatives based on VER-246608 (6) led to the identification of 19n and 19t containing five-membered heteroaromatic rings as unique structural features. These substances possess single-digit nanomolar activities against PDHKs. 19t exhibits higher potencies against PDHK1/2/4 than does 6 and inhibits only PDHKs among 366 kinases. Moreover, 19g, 19l, and 19s were found to be isotype-selective PDHK inhibitors. Molecular dynamics simulations provide a better understanding of how the heteroaromatic rings affect the activities of 19n and 19t on PDHK1/2/3/4. Moreover, 19n possesses a much higher antiproliferative activity against cancer cells than does 6. We demonstrated that the results of PDH assays better correlate with cellular activities than do those of PDHK kinase assays. Furthermore, 19n induces apoptosis of cancer cells via mitochondrial dysfunction, suppresses tumorigenesis, and displays a synergistic effect on satraplatin suppression of cancer cell proliferation.

Full text of DOI:10.1021/acs.jmedchem.9b00565

Subscriber access provided by Nottingham Trent University
Article
Identification of novel resorcinol amide derivatives as potent
and specific pyruvate dehydrogenase kinase (PDHK) inhibitors
Hanna Cho, Injae Shin, Kyung seon Cho, Hojong Yoon, Eun Kyung Yoo,
Mi-Jin Kim, Sungmi Park, In-Kyu Lee, Nam Doo Kim, and Taebo Sim
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00565 • Publication Date (Web): 30 Aug 2019
Downloaded from pubs.acs.org on August 31, 2019
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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
Identification of novel resorcinol amide derivatives as potent and specific pyruvate
dehydrogenase kinase (PDHK) inhibitors
8
9
Hanna Cho,†,1 Injae Shin,†,1 Kyungseon Cho,‡ Hojong Yoon,‡ Eun Kyung Yoo,§ Mi-Jin Kim,§ Sungmi
Park,§ In-Kyu Lee,§,∥Nam Doo Kim,⊥, # and Taebo Sim†, ‡,*
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
†KU-KIST Graduate School of Converging Science and Technology, Korea University, 145 Anam-ro,
Seongbuk-gu, Seoul 02841, Republic of Korea.
‡Chemical Kinomics Research Center, Korea Institute of Science and Technology (KIST), 5 Hwarangro
14-gil, Seongbuk-gu, Seoul 02792, Republic of Korea.
§Leading-edge Research Center for Drug Discovery and Development for Diabetes and Metabolic
Disease, Kyungpook National University Hospital, Daegu 41404, Republic of Korea.
∥Department of Internal Medicine, School of Medicine, Kyungpook National University, Kyungpook
National University Hospital, Daegu 41944, Republic of Korea.
⊥Daegu-Gyeongbuk Medical Innovation Foundation, 2387 dalgubeol-daero, Suseong-gu, Daegu 42019,
Republic of Korea.
#NDBio Therapeutics Inc., 32 Songdogwahak-ro, Yeonsu-gu, Incheon 21984, Republic of Korea.
1These authors are equally contributed to this work.
Abstract
Pyruvate dehydrogenase kinases (PDHKs) promote abnormal respiration in cancer cells. Studies with
novel resorcinol amide derivatives based on VER-246608 (6) led to the identification of 19n and 19t
containing 5-membered heteroaromatic ring as unique structural features. These substances possess
single-digit nanomolar activities against PDHKs. 19t exhibits higher potencies against PDHK1/2/4 than
does 6 and inhibits only PDHKs among 366 kinases. Moreover, 19g, 19l and 19s were found to be
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 61
1
2
3
4
5
6
7
8
9
isotype-selective PDHK inhibitors. Molecular dynamics simulations provide a better understanding
how the heteroaromatic rings affect the activities of 19n and 19t on PDHK1/2/3/4. Moreover, 19n
possesses a much higher antiproliferative activity against cancer cells than does 6. We demonstrated
that the results of PDH assays better correlate with cellular activities than do those of PDHK kinase
assays. Furthermore, 19n induces apoptosis of cancer cells via mitochondrial dysfunction, suppresses
tumorigenesis and displays a synergistic effect on satraplatin suppression of cancer cell proliferation.
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
Introduction
Cancer cells display a distinct metabolic phenomenon, known as the Warburg effect, in order to
rapidly produce energy to support growth. Specifically, cancer cells prefer to utilize aerobic glycolysis,
in which glucose is converted to lactic acid, rather than normal oxidative phosphorylation pathway to
produce ATP.1 For this reasons, enzymes involved in aerobic glycolysis including GLUT (Glucose
transporters), HK (hexokinase), PDHK (pyruvate dehydrogenase kinase), NADH-ubiquinone reductase
and ATP transporter have gained attention as a potential targets for cancer chemotherapy.2 Most cancer
cells express high levels of PDHK, a serine threonine kinase that negatively regulates activity of the
PDH complex by phosphorylation of the PDH E1a subunit. The PDH complex in the mitochondrial
membrane plays a key role in aerobic glycolysis by irreversibly converting pyruvate into acetyl-CoA
used for energy generation in normal cells.3 Indeed, aberrant expression of PDHKs occurs in multiple
human tumors. Not only is PDHK1 overexpressed in gastric cancer, RCC and myeloma,4 it also
promotes migration in NSCLC.5 In addition, PDHK2 and PDHK3 are highly activated in breast and
colon cancers, and NSCLC,6 and especially high levels of expression of these enzymes are associated
with poor prognosis in patients.7 PDHK4 is also intimately associated with colon8 and bladder9 cancers.
In contrast, low expression levels of the PDHKs occur in non-cancerous cells. This finding suggests
that side effects caused by PDHK inhibition should be minimal.10 Hence, PDHKs have become
attractive targets for cancer chemotherapy and many types of PDHK inhibitors have progressed to the
preclinical/clinical stage (Figure 1). However, no PDHK inhibitors has been approved for use thus far.
ACS Paragon Plus Environment

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
H
N
O
OH
Cl
O
Cl
H
N
NH2
O
F3C
N
N
O
Cl
O
OH
N
O
F3C
HO
CN
S
Cl
O
O
O
8
9
DCA 1
( )
Nov3r 2
( )
AZD7545 3
( )
Pfz3 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
N
N
Cl
N
F
F
OH
Cl
O
O
S
O
O
O
N
O
NH
S
S
N
HO
HO
O
S
Radicicol 5
( )
OH
O
VER-246608 6
( )
3a 7
( )
Figure 1. Representative PDHK inhibitors.
Dichloroacetic acid (DCA, 1), a pyruvate analog that inhibits PDHK by binding to the N-terminal
regulatory region (R-domain) of the enzyme, has been found to suppress cancer growth.11 However,
clinical trials revealed the limitation of 1 associated with low potency and neurological toxicity.10 A
number of 1 analogues have been developed and progress has been made in enhancing anticancer
potency.12 The PDHK2 inhibitors Nov3r (2) and AZD7545 (3)13 also bind to the lipoyl group-binding
site in the R-domain and display limited cytotoxicity against cancer cells.13b, 14 Allosteric inhibitors
including Pfz3 (4) bind to a distant site located at the other side of the R domain and have been observed
to have moderate cellular activities.15 Moreover, radicicol (5) and 4,5-diarylisoxazoles are inhibitors
that competitively bind to the ATP binding pocket in PDHK. Because PDHK and HSP90 are members
of the GHKL ATPase/kinase super family that have conserved ATP binding sites, it is difficult to
develop inhibitors that display selectivity in binding to PDHK over other members of the GHKL
ATPase/kinase super family.16 It is worth noting that 6 displays a 1000-fold higher activity against
PDHK over HSP90 with cytostatic activity in a context dependent manner.17
Continuing investigations have identified novel PDHK inhibitors3a, 18 such as derivatives of the liver
specific inhibitor dihydroxyphenyl sulfonylisoindoline used for treatment of obesity and type 2
diabetes.19 The disulfide-based PDHK inhibitor 7, which covalently binds to a conserved cysteine in
the ATP binding region, displays a ca. 40-fold selectivity for PDHK1 over other subtypes and it
significantly suppresses the growth and tumor mass in the A549 xenograft model.20 A study aimed at
optimization of 6 derivatives as selective pan-PDHK inhibitors has recently been reported. In this effort,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 61
1
2
3
4
5
6
7
binding affinity was assessed by using a fluorescence polarization assay rather than by using an assay
based on inhibition of kinase activity.21
8
9
In an effort to discover potent and specific PDHK inhibitors, we designed and synthesized a series
of new resorcinol amide derivatives and performed an SAR study that is based on an evaluation of
inhibitory activities against PDHK1-4. This study led to identification of two outstanding lead
compounds, 19n and 19t, which exhibit single-digit nanomolar activities against the PDHKs, values
that are superior to that of 6. In a kinome-wide selectivity profiling investigation using 366 kinases, we
showed that 19t binds to PDHKs only. We also demonstrated that 19n suppresses proliferation H1299
and DU145 with a GI50 of 13.4 µM and 10.2 µM respectively, promotes apoptosis via mitochondrial
dysfunction and effectively blocks tumorigenesis of H1299 cells. Moreover, we found that 19n
synergistically increases the cytotoxicity of satraplatin against lung carcinoma H1299 cells.
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
Results and discussion
Chemistry
Two synthetic routes were employed to generate the resorcinol amide derivatives and analogs 14,
15a-g, 19a-t and 21a-b shown in Schemes 1 and 2. The synthesis of 15a-g commenced with reductive
amination of methyl 4-aminobenzoate (8) with 4-nitrobenzaldehyde to form the secondary amine 9. The
substituted benzoic acid 10a was converted to the corresponding acid chlorides using oxalyl chloride
and treated with 9 to generate tertiary amide 11. Reduction of the nitro group in 11 using catalytic
hydrogenation, followed by acetylation of the resulting amine produced 13, which upon removal of the
acetyl groups and hydrolysis of the methyl ester formed 14. Amide coupling reactions using HATU
carried out on 14 yielded the target amides 15a-g. Compounds 19a-t and 21a-b were prepared by the
synthetic strategy described in Scheme 2. The aniline 16 was submitted to reductive amination reactions
to afford the secondary amines 17a-t. 17a-t were coupled with the benzoic acid 10a to furnish the
tertiary amides 18a-t of which acetyl groups were hydrolyzed to yield the desired targets 19a-t. 21a-b
were synthesized by using 10b-c in the same synthetic route as for 19a-t synthesis.
ACS Paragon Plus Environment

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
Scheme 1. Synthesis of 14 and 15a-ga
8
9
O
O
OMe
O
O
OMe
O
OMe
O
OMe
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
b
c
AcO
NO2
AcO
NH2
NO2
AcO
N
N
HN
NH2
OH
OAc
OAc
R1
10a
OAc
O
8
9
11
12
O
OMe
O
OH
O
O
d
e
f
H
N
H
N
H
N
AcO
HO
HO
N
O
N
O
O
N
O
OAc
O
OH
OH
13
14
15a-g
aReagents and conditions: (a) 4-nitrobenzaldehyde, acetic acid, NaBH3CN, MeOH, rt, 5 h, 81%; (b) i)
10a, oxalyl chloride, cat. DMF, DCM, rt, 2 h ii) 9, DIPEA, DMF, 0 oC to rt, 2 h, 49% over 2 steps; (c)
Pd/C, H2, MeOH, rt, 12 h; (d) acetic anhydride, TEA, DCM, 0 oC to rt, 2 h, 71% over 2 steps (e) NaOH,
MeOH/H2O, 80 oC, 2 h, 79 %; (f) HATU, various amines, N-methylmorpholine, DMF, rt, 3 h, 21-58%.
Scheme 2. Synthesis of 19a-t and 21a-ba
O
O
N
O
O
N
O
N
O
N
a
b
c
AcO
HO
AcO
N
R2
N
R2
OH
HN
R2
NH2
10a
OAc
OH
OAc
O
16
17a-t
18a-t
19a-t
AcO
OH
b
O
N
O
O
N
10b-c
O
c
F
F
10b
10c
: 4-acetoxy
: 2-acetoxy
AcO
HO
N
N
21a
21b
: 4-hydroxy
: 2-hydorxy
O
20a-b
21a-b
aReagents and conditions: (a) various aldehydes, acetic acid, NaBH3CN, MeOH, rt, 5 h; (b) i) 10a-c,
oxalyl chloride, cat. DMF, DCM, rt, 2 h ii) 17a-t, DIPEA, DMF, 0 oC to rt, 2 h; (c) K2CO3, MeOH, rt,
1 h, 17-45%, over 4 steps.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 61
1
2
3
4
5
Structure and activity relationships
6
7
8
9
An analysis of the crystal structure of the complex between 6 and PDHK2 shows that the resorcinol
moiety of 6 forms direct and water-mediated hydrogen bonds with Asp282 and Thr346 in the ATP
binding site of PDHK2. This is the major reason why resorcinol-containing substances were selected
as potential PDHK inhibitors. In the complex, the difluorobenzylacetamide moiety of 6 is oriented
toward the solvent-exposed region of the enzyme and, as a result, it is not expected to be critical for
binding to the PDHKs. Thus, a phenylacetamide instead of the difluorobenzylacetamide of 6 was
incorporated in the inhibitors owing for synthetic accessibility reasons. The selected inhibitors 14 and
15a-g contain various amide substituents (R1) that replace the chloromethylpyrimidine moiety of 6.
Inspection of the X-ray structure of the complex between 6 and PDHK2 shows17 that the 2-nitrogen of
the chloromethylpyrimidine moiety of the inhibitor interacts with Arg250 of PDHK2. We assumed that
the amide carbonyl oxygen in R1 would mimic the 2-nitrogen in the chloromethylpyrimidine group by
engaging in H-bonding with Arg250 of PDHK2.
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
An in vitro kinase assay was used to evaluate the inhibitory activities of 14 and 15a-g against
PDHK2/4. As the data in Table 1 demonstrate, the R1 = carboxyl acid containing derivative 14 has only
moderate activities against both PDHK2 (IC50 = 7.220 µM) and PDHK4 (IC50 = 4.670 µM). Moreover,
changing the R1 group to a cyclopropyl amide (15a), methyl amide (15b) and ethanolamide (15c) does
not lead to an improvement of kinase-inhibitory potency. As expected, when a dimethylamide (15d) is
the R1 group, a remarkable enhancement in kinase-inhibitory activity against both PDHK2 (IC50 = 0.503
µM) and PDHK4 (IC50 = 0.311 µM) occurs. This observation is consistent with the previous report21
that incorporation of a dimethylamide moiety leads to submicromolar potencies against PDHK2. The
amides 15e-g, which contain the cyclic amines piperidine, morpholine and pyrrolidine, display lower
activities than does dimethylamide (15d). These results show that the dimethylamide in 15d is an
optimized R1 substituent.
Table 1. Inhibitory activities of 14, 15a-g against PDHK2/4.
ACS Paragon Plus Environment

Page 7 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
R1
6
7
8
9
H
HO
N
N
O
OH
O
14, 15a-g
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
IC50 (µM) a
IC50 (µM) a
Entry
R1
Entry
R1
PDHK2
7.220
PDHK4
4.670
PDHK2 PDHK4
O
N
O
14
15d
15e
15f
15g
0.503
1.620
2.660
1.660
0.311
1.980
5.793
1.810
OH
O
N
O
15a
15b
15c
1.960
8.590
9.910
14.750
7.840
N
H
O
O
N
N
H
O
O
O
OH
12.260
N
N
H
aRadiometric biochemical kinase assay.
Our attention next focused on optimization of the R2 substituent that corresponds to the
difluorobenzylacetamide moiety in 6 (Table 2). Utilizing phenyl in 19a as the R2 group causes only
slight enhancements of inhibitory activities against PDHK2 and PDHK4. This finding suggests that the
acetamide moiety in 6 does not greatly contribute to activity. We next explored derivatives 19b-f, which
contain electron-withdrawing groups (F, Cl, CF3, NO2 and CN, respectively) at the para-position of the
phenyl ring. Use of fluorine as the para-substituent (19b) results in a slightly higher potency than that
of 19a for inhibition of both PDHK1 (IC50 = 0.067 µM) and PDHK2 (IC50 = 0.113 µM). The activities
of the 19b analogs, 21a and 21b, that contain a phenol instead of a resorcinol moiety are significantly
lower, which is in accord with a previous report17 that the resorcinol moiety in 6 plays a pivotal role in
activity. Other electron-withdrawing groups (19c-f) have little influence on the inhibitory activities
against PDHK2/4. However, 19g, which contains a phenol moiety as R2, exhibits a two-fold higher
activity against PDHK4 than does 19a. It should be emphasized that 19g possesses a 5-fold selectivity
for PDHK4 over PDHK1/2/3 and, to the best our knowledge, it is the first reported isotype-selective
PDHK4 inhibitor. Use of the large hydrophobic naphthalene moiety as R2 (19h) brings about a moderate
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 61
1
2
3
4
5
6
7
8
9
decrease in activity against both PDHK2 and PDHK4 compared with that of 19a. Similarly,
incorporation of a bulky t-butyl group (19i) also diminishes activities against PDHK2/4. Using a
cyclohexyl (19j) in place of the phenyl group (19a) results in lower activity against both PDHK3 and
PDHK4 and comparable activity against PDHK1/2.
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
An exploration of various groups as R2 shows that the five-membered heteroaromatic ring
containing derivatives 19k-t have enhanced inhibitory activities. Specifically, the furan
containing analogue 19k displays a 3-fold enhancement of activity against PDHK3 (IC50 = 0.067
µM) relative to that of 19a. Significantly, the 2-bromofuran analogue 19n exhibits high potency
on both PDHK2 (IC50 = 10 nM) and PDHK4 (IC50 = 13 nM). Interestingly, the 2-iodofuran
derivative 19o has a 12-fold higher inhibitory activity on PDHK3 (IC50 = 5 nM) and a 10-fold
lower potency on PDHK2 (IC50 = 98 nM) compared with those of the 2-bromofuran analog 19n.
In addition, 19o has a 6 to 14-fold enhanced activity on PDHK3 (IC50 = 5 nM) and PDHK4 (IC50
= 11 nM) relative to that of the unsubstituted furan derivative 19k. Based on the results of this
study, it is possible to conclude that the 2-substituent on the furan ring of the R2 group
significantly influences inhibitory activity against PDHK2/3 but it does not greatly influence
the activity against PDHK1. Also, incorporation of 2-bromo and 2-iodo substituents on the furan
ring leads to substantially increased activity against PDHK2/4 and PDHK3/4, respectively.
We next investigated the activities of substances possessing a thiophene ring as the R2
substituent. The results show that the inhibitory activity of 19p, containing a 2-boromothiophene
group, are comparable to that of 19n. The 2-cyano substituted derivative 19q was found to have
an inferior activity against both PDHK2/4 as compared to that of the 2-bromo analog 19p.
Moreover, the 3-bromothiophene derivative 19t has a 5-fold enhanced activity against PDHK1
relative to the 2-bromo-substituted derivative 19p. It is of note that the potencies of 19t against
PDHK1/2/4 are 4 to 8-fold higher than those of 6. A study of analogs containing a thiazole as the
R2 group shows that 19r containing a 2-bromothiazole group has a 7 to 11-fold lower activity
for inhibition of PDHK2/4 as compared to that of the 2-boromothiophene derivative 19p and
that 19s, with a 5-bromothiazole group, is 4-fold less active against PDHK2 (IC50 = 64 nM) than
ACS Paragon Plus Environment

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
is 19p. It is worth noting that small changes in the nature and location of the substituent (19p vs 19t),
and the presence or absence of a nitrogen (19p vs 19r) in the 5-membered heteroaromatic ring markedly
influence inhibitory activity. The results of this SAR study show that substances in this series containing
5-membered heteroaromatic rings as R2 substituents have higher activities than those of derivatives
bearing phenyl rings.
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 2. Inhibitory activities of 15d, 19a-t and 21a-b against PDHK1/2/3/4.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 61
1
2
3
4
5
6
7
8
9
N
N
Cl
O
O
N
O
O
N
N
F
F
O
NH
HO
F
HO
HO
N
R2
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
21a
21b
: 4-hydroxy
: 2-hydorxy
OH
O
6
15d, 19a-t
PDHK IC50 (µM) a
PDHK IC50 (µM) a
Entry
R2
Entry
R2
1
2
3
4
1
2
3
4
H
N
15d
19a
19b
ND b
0.503
ND b
0.311
0.066
19j
19k
19l
0.140
0.184
0.858
0.775
O
O
O
0.109
0.206
0.186
0.048* 0.133* 0.067* 0.069*
F
0.067* 0.113* 0.142*
0.052*
0.390
> 10
0.020
0.024
0.252
0.193
0.255
0.181
0.052
0.082
Cl
Br
I
O
O
O
S
S
21a
21b
19c
19d
19e
19f
19g
19h
19i
-
-
ND b
ND b
ND b
ND b
ND b
0.151
2.040
> 10
ND b
ND b
ND b
ND b
ND b
0.461
19m
19n
19o
0.036* 0.010* 0.059* 0.013*
Cl
CF3
NO2
CN
0.126
0.140*
0.110
0.302
0.169
0.014
0.042
0.057
0.024
0.098
0.018
0.095
0.132
0.005
0.090
0.098
0.090
0.011
0.008
0.089
0.087
Br
CN
0.273* 19p
0.099
0.399
0.034*
0.234
5.330
19q
19r
19s
19t
6
Br
N
S
S
Br
OH
0.113* 0.150* 0.136*
0.020* 0.064* 0.064* 0.020*
0.009* 0.008* 0.080* 0.005*
N
Br
ND b
ND b
0.373
0.772
ND b
ND b
S
-
0.042* 0.065* 0.026* 0.021*
aRadiometric biochemical kinase assay. ND stands for Not Determined. Average value of two
b
*
independent assays.
ACS Paragon Plus Environment

Page 11 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
Molecular dynamics simulation
8
9
19n possesses high potency against PDHK2/4 and selectivity over PDHK1/3. To gain insight into
the factors governing the activities of 19n in inhibiting the PDHKs, molecular dynamics simulations
were performed to understand its binding modes in the ATP-site of these kinases (Figure 2). The results
reveal that the resorcinol moiety of 19n participates in water-mediated hydrogen-bonding networks
with Gly322 in PDHK1, Leu244 in PDHK2, Ser252 in PDHK3 and Gly297 in PDHK4 as well as in
direct hydrogen bonding with Asp318 in PDHK1, Asp282 in PDHK2, Asp287 in PDHK3 and Asp293
in PDHK4. The carbonyl oxygen in the resorcylic acid amide forms a hydrogen bond with Thr383 in
PDHK1, Thr346 in PDHK2, Gly291 in PDHK3 and Thr358 in PDHK4. It is worth noting that the
carbonyl oxygen of resorcylic acid amide forms an additional water-mediated H-bond only with a
residue in the ATP binding site of PDHK2. This unique interaction could contribute to the high activity
(IC50 = 10 nM) of 19n against PDHK2. The noticeable differences between the inhibitory activities of
19n on the PDHK isotypes appears to be mainly a consequence of the degree of interactions between
its N,N-dimethylbenzamide moiety and Arg residues in the kinases. Indeed, both Arg250 in PDHK2
and Arg261 in PDHK4 strongly interact with the carbonyl oxygen of the N,N-dimethylbenzamide group
(respective interaction scores of 95% and 83%) while Arg286 in PDHK1 and Arg254 in PDHK3 only
weakly interact with the carbonyl oxygen of 19n (respective interaction scores of 31% and 35%). It
appears that cation-pi interaction (an interaction score of 84%) between the phenyl ring of the N,N-
dimethylbenzamide moiety and Arg261 in PDHK4 also contributes to the high potency (IC50 = 5 nM)
of 19n against PDHK4. The representative docking poses of the 19n with PDHKs are shown in Figure
3. Overall, carbonyl of N,N-dimethylbenzamide form hydrogen bond with Arg residue (Arg286, Arg250,
Arg254 and Arg261 respectively) and resorcinol moiety also makes interactions with Asp (Asp318
Asp282, Asp287 and Asp293 respectively) and Thr (Thr383, Thr346, Thr352 and Thr358 respectively)
in PDHKs. Unlike other isoforms, Thr352 of PDHK3 forms an indirect interaction with 19n.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 61
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Ligand interaction diagrams and interaction profile analysis of molecular dynamics
simulations of 19n within PDHK isoforms (a) PDHK1, (b) PDHK2, (c) PDHK3, (d) PDHK4. The
dotted arrows indicate H-bonds to side chains of the PDHKs and full arrows indicate H-bonds to
backbone of the PDHKs. Red connectors denote cation-pi stacking interactions. Green: H-bonds, pink:
ionic interactions, purple: hydrophobic interactions, blue: water bridges.
ACS Paragon Plus Environment

Page 13 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Docking structures of 19n with the ATP binding sites of the PDHKs. (a) PDHK1, (b) PDHK2,
(c) PDHK3 and (d) PDHK4
To gain an understanding of why 19t possesses excellent activities (IC50s = 5 to 9 nM) against
PDHK1/2/4 but a much lower potency (IC50 = 80 nM) against PDHK3, molecular dynamics simulations
were carried out to determine its binding modes to the PDHKs (Figure 4). In a manner that is similar to
the binding modes of 19n to PDHK1/2/3/4, the carbonyl oxygen of the resorcylic acid amide group in
19t forms a hydrogen bond with Thr383 in PDHK1 and Thr346 in PDHK2, which could contribute to
its potencies against PDHK1/2. In contrast, the amide carbonyl oxygen in 19t does not participate in
direct hydrogen bonding with residues in PDHK3/4. In addition, the carbonyl oxygen of the N,N-
dimethylbenzamide group in 19t engages in direct H-bonding interactions with Arg286 in PDHK1,
Arg250 in PDHK2, and Arg261 and Thr313 in PDHK4 but it is not involved in H-bonding with residues
in PDHK3. Neither of the two carbonyl oxygens in 19t participate in direct H-bonding interactions with
residues in PDHK3, which could be the source of its low inhibitory activity (IC50 = 80 nM) against
PDHK3. Furthermore, cation-pi interactions between the N,N-dimethylbenzamide moiety in 19t and
Arg286 in PDHK1 and Arg261 in PDHK4 could be the reason for its high activity against PDHK1/4.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 61
1
2
3
4
5
6
7
8
9
In summary, the results of molecular dynamics simulations provide a better understanding how the R2
substituent affects inhibitory activities of 19n and 19t against PDHK1/2/3/4. We also performed
molecular dynamics simulations of 6 on PDHK1/2/3/4 (Figure S1) for comparison. The results of
molecular dynamics simulation of 6 are quite similar with those of both 19n and 19t even though all
interaction % scores are not completely in accordance with the kinase-inhibitory activities of 19n, 19t
and 6.
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
Figure 4 Ligand interaction diagrams of molecular dynamic simulations of 19t with PDHK isoforms
(a) PDHK1, (b) PDHK2, (c) PDHK3, (d) PDHK4. The dotted arrows indicate H-bonds to side chains
of the PDHKs and full arrows indicate H-bonds to the backbone of PDHKs. Red connector denotes
cation-pi stacking interaction.
19b and 19t are specific PDHKs inhibitors
ACS Paragon Plus Environment

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Kinome-wide selectivity profiling of both 19t and 19b were carried out using a panel of 366 kinases.
To our surprise, the profiling data reveal that among 366 kinases only the PDHKs are inhibited by more
than 90% by 1 µM of 19t (Figure 5 and Table S2) and 19b (Figure S2 and Table S1). This observation
indicates that both 19t and 19b are specific PDHKs inhibitors. It is of note that 10 µM of 6 inhibits 8
out of 96 kinases by more than 65%.17 In addition to its exceptional kinome-wide selectivity, 19t has
ca. 10-fold greater kinase-inhibitory activity against PDHK1/2/4 than does 6 (Figure 5B). To assess the
selectivity on the GHKL (gyrase, Hsp90, histidine kinase, MutL), which have ATP binding sites that
are similar to those of the PDHKs,16b we measured the kinase-inhibitory activities of selected
compounds against HSP90. Even though 19k and 19n are reasonably strong inhibitors of HSP90 (IC50
values of ca. 1 µM), the other analogs listed in Figure 5C possess low kinase-inhibitory activity (IC50s
= ca. 8 to >10 µM) against HSP90 and excellent selectivity for the PDHKs over HSP90.
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
Figure 5. (A) Overall kinase selectivity profiling of 19t. Kinases showing >50 % inhibition are
clustered on the kinome tree. PDHK1/2/3/4 out of 366 kinases are inhibited more than 90% by 1 µM
19t. Illustration reproduced courtesy of Cell Signaling Technology, Inc. (www.cellsignal.com) (B) The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 61
1
2
3
4
5
6
7
8
9
percent inhibition of PDHKs activities in the presence of 19t and 19b (1 µM). (C) Inhibitory activities
of selected compounds against PDHK1/2/3/4 and HSP90. Compounds were tested in 10-doses with a
3-fold serial dilution starting at 10 μM.
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
In vitro potency of resorcinol amide derivatives against cancer cells and reconstituted PDC-
reaction system.
In order to assess the cellular activities of the resorcinol amide derivatives that have potent inhibitory
activity against the PDHKs (IC50 < 200 nM), we determined their antiproliferative activities in p53-
deficient human non-small cell lung carcinoma H1299 cells and DU145 prostate cancer cells (Table 3).
The results show that 19n, which has a one-digit nanomolar IC50 against PDHK2 and PDHK4,
suppresses growth of H1299 and DU145 with a GI50 of 13.4 µM and 10.2 µM respectively whereas 6
has an extremely low antiproliferative activity against H1299 (Table 3 and Figure S3). In order to
support the notion that 19n suppresses cancer cell growth through targeting PDHK, we investigated
whether knockdown of PDHK expression affects the antitumor activity of 19n (Figure S4). PDHK1/2
double knockdown in H1299 cells resulted in about 40% inhibition of cancer cell growth (Figure S4B).
In H1299 cells treated with siCTL (control siRNA), 10 µM of 19n reduced the proliferation of the
H1299 cells by around 50% compared to DMSO treated cells. On the other hands, in PDHK1/2 double
knockdown cells, 10 µM of 19n suppressed the proliferation of the H1299 cells by 32% compared to
DMSO treated cells suggesting that 19n suppresses the growth of H1299 cancer cells through targeting
PDHK1 and PDHK2. Interestingly, 19g and 19k have appreciable anti-proliferative activities against
both H1299 and DU145 cells with respective GI50 values ranging from 3.5 to 9.0 µM. Considering the
moderate PDHK inhibitory activities of 19g and 19k (respective IC50 of 30 nM to ca.150 nM), it is clear
that their inhibitory activities are not well-correlated with their cellular potencies. Finally, 19g does not
inhibit HSP90, indicating that its antiproliferative activity against H1299 and DU145 results from off-
target effects.
To understand the discrepancy between enzyme inhibition and cellular activities, we utilized a
pyruvate dehydrogenase (PDH) assay, which is a reconstituted PDC-reaction system used to measure
PDH activity in vitro (Figure 6A). PDH is a component of the PDC complex converting pyruvate to
ACS Paragon Plus Environment

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acetylCoA via NAD+ reduction. PDHKs inhibit PDH activity and cause decreased levels of NADH
production. Thus, inhibition of PDHK results in higher PDH activity and an increase in the amount of
NADH formed from NAD+.22 The effects of 19g, 19k and 19n, which have high antiproliferative
activities (GI50 <15 µM) against H1299 and DU145, on PDH activity were determined by measuring
the rate of NAD+ reduction. As the results in Figure 6B show, 19n has the lowest IC50 value against
both PDHK2 (IC50 = 0.58 µM) and PDHK4 (IC50 = 0.31 µM). The activities of 19g and 19k are
approximately three times lower than that of 19n in PDC assay while 19g and 19k are over 15 times
less potent than 19n against PDHK2 in biochemical kinase assay. Considering the antiproliferative
activities of 19g, 19k and 19n on H1299 and DU145 cells, it appears that the PDH assay results better
correlate with cellular activities than do the PDHK kinase-inhibitory activities. The IC50 values
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
derived from biochemical kinase assay depend on ATP concentration.23 We assessed the IC50
values of 19n against PDHKs using three different ATP concentrations (10/50/200 µM) (Table
S3). 200 μM of ATP is close to a physiological ATP concentration (ca. 1 mM). Compared with
10 µM of ATP condition, potency of 19n against PDHK1 is decreased by 11-fold at 200 µM
of ATP concentration. PDHK-inhibitory activity of 19n would be much lower at physiological
ATP concentration, which might be a reason of the discrepancy between its enzymatic activity
and cellular potency. Also, the discrepancy might be in part attributed to low cellular
permeability and/or high efflux ratio.
Table 3. Anti-proliferative activities against H1299 and DU145 cells.
H1299a
DU145a
H1299a
DU145a
Entry
Entry
GI50 (µM)
GI50 (µM)
GI50 (µM)
GI50 (µM)
19b
57.1 ± 4.3
69.7 ± 29.0
19o
39.19 ± 5.3
15.9 ± 2.0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 61
1
2
3
4
5
6
19c
31.5 ± 2.6
51.0 ± 5.54
19p
49.8 ± 8.9
42.25 ± 1.5
7
8
9
19e
19g
19k
19m
19n
97.1 ± 2.9
4.9 ± 0.9
44.5 ± 20.0
3.5 ± 0.2
19q
19r
19s
19t
6
107.2 ± 24.3
32.16 ± 2.15
60.2 ± 29.4
51.6 ± 3.2
> 50
59.8 ± 6.8
18.9 ± 8.9
40.8 ± 2.2
41.9 ± 4.9
> 50
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
7.8 ± 1.7
9.0 ± 0.8
30.0 ± 4.58
13.4 ± 2.6
29.9 ± 3.9
10.2 ± 22.4
a GI50 is the concentration at which half-maximal growth inhibition occurs. The cells were treated with
compounds for 72 h in a dose escalation manner. A CellTiter-Glo assay was performed to assess cell
viability. Average GI50 values with SD (n = 2) are shown.
Figure 6. PDH assay results. (A) Schematic diagram depicting the principle of reconstituted PDC-
reaction system. NAD+ reduction simultaneously occurs when PDH complex (E1, E2 and E3) converts
pyruvate to Acetyl-CoA. Phosphorylation of E1 by PDHKs suppresses the activity of PDC complex,
thus NAD+ reduction is decreased. The activity of PDH complex (E1, E2 and E3) was evaluated by
measuring the rate of NAD+ reduction for 30 min in the presence of 19g, 19k and 19n respectively. (B)
The result of PDH assays of 19g, 19k and 19n.
CYP inhibition profiles of 19g and 19n
Because the resorcinol amide derivatives 19g and 19n display high activities in the PDH assay, we
determined their inhibitory properties against five major cytochrome P450s including CYP1A2,
ACS Paragon Plus Environment

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CYP2C9, CYP2C19, CYP2D6 and CYP3A4. As the data in Table 4 show, these substances have little
to no inhibitory effects on CYP1A2, CYP2C9 and CYP2D6. 19n is a moderate inhibitor (IC50 >1 µM)
of CYP2C19 and CYP3A4 while 19g has favourable CYPs inhibition profiles on all of the five major
cytochrome P450s. We also assessed cytochromes P450-inhibitory activities of 19t possessing excellent
kinase-inhibitory activity on PDHKs and found that 19t has little or no inhibition effects on CYP1A2,
CYP2C9, CYP2C19 and CYP2D6 while it inhibits CYP3A4 with IC50 value less than 1 µM.
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 4. CYP inhibition profiles of 19g, 19n and 19t.
CYP IC50 (µM)
Entry
CYP1A2
> 20
CYP2C9
13.4
CYP2C19
> 20
CYP2D6
> 20
CYP3A4
6.47
19g
19n
19t
> 20
> 20
> 20
18.0
4.3
7.8
> 20
> 20
1.4
< 1
19n blocks phosphorylation of PDHE1a in H1299 and DU145 cells.
In order to determine if the PDHK inhibitory activities of the resorcinol amide derivatives have
cellular consequences, western blot analysis of H1299 cells was carried out to assess the effects of 10
µM of 19d, 19g, 19k, and 19n on levels of phosphorylation of Ser293 residue of PDHE1a promoted by
all four PDHK isozymes. The results of western blot analysis (Figure 7A) are consistent with those
arising from the PDH assays (Figure 6). Specifically, among the four derivatives, 19n most strongly
decreases the level of phosphorylation of PDHE1a. Moreover, 19n inhibits phosphorylation of PDHE1a
in a dose-dependent fashion (Figure 7B) and at 20 µM it completely abolishes phosphorylation of
PDHE1a (S293), an effect that is comparable to that of 6. We also found that 19n dose dependently
inhibited phosphorylation of Ser293 of PDHE1a in DU145 cells (Figure 7C).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 61
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. 19n blocks PDHE1a phosphorylation in H1299 cells and DU145 cells. (A) Western blot
analysis of H1299 cells treated with the selected compounds at 10 µM for 2 h (110 mM) and a graph of
the results (right panel). 19n attenuated pPDHE1a phosphorylation 70-80% while the other substances
show only moderate inhibitory activities. (B) Phosphorylation of PDHE1a (S293) is attenuated by 19n
and 6 in a dose-dependent manner. (C) Western blot analysis of DU145 cells treated with indicated
concentration of 19n for 2 h.
19n induces OCR, ROS production and mitochondrial depolarization.
To assess the energy metabolism in cancer cells, we measured OCR (oxygen consumption rate) in
H1299 cells treated with 19n and 1. We found that 19n treatment increased both the spare respiratory
capacity and maximal OCR, which were comparable to the effects of 1. (Figure 8A). This result implies
that inhibition of PDHK by 19n can switch pyruvate consumption from lactate production to oxidative
phosphorylation. We also determined whether PDHK inhibition by 19n promotes mitochondrial
dysfunction and ROS production. By using a flow cytometry based assay, we found that 19n causes a
dose-dependent increase in ROS production (Figure 8B). It has been reported that inhibition of PDHK
activity by 1 or its analogues decreases the mitochondrial membrane potential of cancer cells., 12c, 24 In
order to assess whether 19n also alters the mitochondrial membrane potential, we measured changes
in the ratio of red to green fluorescence intensities of the JC-1 probe in DU145 cells. JC-1 forms red
fluorescent aggregates in the normal mitochondrial membrane and the color of the fluorescence
reversibly shifts from red (aggregate) to green (monomer) as the membrane potential decreases in the
ACS Paragon Plus Environment

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
response to stimuli.25 Thus, a decrease in the ratio of red to green fluorescence indicates the occurrence
of depolarization of the mitochondrial membrane and/or dysfunction of the mitochondria. To detect the
red/green fluorescence ratio, DU145 cells were incubated with indicated doses of 19n and then stained
with JC-1. Fluorescence microscopic images of the cells were quantified by using image J (Figure 8C).
As expected, 19n dose dependently promotes a decrease in the red/green fluorescence ratio. This result
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
clearly demonstrates that mitochondrial function is impaired by 19n
.
Figure 8. The effects of 19n on mitochondrial respiration, ROS production and mitochondrial
membrane potential. (A) 19n enhances both mitochondrial respiration in H1299 cells. OCRs were
measured in H1299 cells in the presence or absence of DCA (10 mM) and 19n (10 µM) for 1 h and
oligomycin A (1 μM), FCCP (2 μM) and rotenone/antimycin A (1 μM) were sequentially delivered to
the XFe96 assay medium through injection ports in the sensor cartridge. (n=6, +/- SD, t-test, * < 0.05
vs. control). (B) 19n promotes strong generation of ROS in a dose-dependent manner (left panel: H1299,
right panel: DU145). Fluorescence of DCF was detected by using FACS (n = 3, +/- SD, One way
ANOVA; **** p <0.0001, *** p <0.0009, * <0.05). (C) DU145 cells were treated with increasing
concentrations of 19n for 5 h and stained with JC-1 for 30 min. 19n dose dependently reduces the
red/green fluorescence ratio. Scale bar = 25 µm, intensity of fluorescence was automatically measured
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 61
1
2
3
4
5
by image J (n=3, +/- SD, One way ANOVA; *** p <0.0007, ** <0.01).
6
7
19n promotes apoptosis and suppresses tumorigenesis of H1299 and DU145 cells.
8
9
Because alteration of the mitochondrial membrane potential is a cellular event in the apoptotic
process,26 we explored the apoptosis-inducing effect of 19n on H1299 and DU145 cells. After treatment
with 19n for 48 h, cells were co-stained with FITC conjugated Annexin-V and propidium iodide, and
subjected to flow cytometric analysis (Figure 9A). The results show that 19n causes a gradual dose-
dependent increase in the population of apoptotic cells. The apoptosis ratio promoted by treatment with
30 µM of 19n was found to be 24.8% in H1299 (Figure 9A left panel). In DU145, 19n dramatically
induced apoptosis over 60 % at 30 µM. 1 does not promote apoptosis of H1299 and DU145 cells
indicating that the apoptosis-inducing effect of 1 may be dependent on cell type. Because caspase 3
cleavage is activated by cytochrome c released from mitochondria, it is a marker for mitochondria-
mediated apoptosis.27 To determine whether 19n induces apoptosis via mitochondrial signaling, the
degree of caspase 3 cleavage was measured by using western blotting (Figure 9B left panel). The results
show that 19n effectively induces cleavage of caspase 3 in a dose dependent manner. Cleavage of PARP,
a substrate of caspase 3, by 19n was also assessed in DU145 (Figure 9B right panel). 19n increased the
level of cleaved PARP in a dose dependent fashion while cleaved PARP was not enhanced by 1, which
is consistent with the results of FACS analysis.
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
Anchorage independent growth in vitro reflects the tumorigenesis capacities of cells in vivo. Hence,
we evaluated the ability of 19n to suppress colony formation of H1299 cells in the soft agar assay. Agar
embedded H1299 cells were treated with increasing concentrations of 19n for 14 d and then subjected
to colony formation analysis (Figure 9C). It was found that both the colony number and size of 19n
treated cells were greatly lower than those of the control. In addition, 19n causes complete inhibition
of colony formation at 20 µM concentration. It is noteworthy that 19n suppresses the proliferation of
H1299 cells in a 2D culture system by 70% at 20 µM, which indicates that antiproliferative activity of
19n is higher in the 3D soft agar colony formation assay than in the 2D CellTiter-Glo assay. These
results suggest that 19n effectively prevents tumorigenesis of H1299 cells.
ACS Paragon Plus Environment

Page 23 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. 19n induces apoptosis of H1299 and DU145 and 19n inhibits anchorage independent growth
in H1299. (A) Numbers of apoptotic cells after treatment of 19n for 48 h were determined by using
Annexin V/PI staining. 19n causes a dose dependent increase in apoptotic cells (left panel: H1299, right
panel: DU145). (B) Caspase 3 cleavage or PARP cleavage are increased by dose dependent treatment
with 19n (left panel: H1299, right panel: DU145). (C) Anchorage independent growth inhibition of
H1299 cells promoted by 19n. Average numbers of colonies per well were automatically counted using
Image J software and are shown in the bar-graph. (n = 3, +/- SD, One way ANOVA; **** p <0.0001,
**p <0.01, * p <0.05).
19n increases in vitro cytotoxicity of satraplatin.
Co-treatment with 1 increases the anti-proliferative activities in cancer cells of platinum compounds
such as carboplatin and satraplatin.28 Stimulated by this finding, we evaluated whether 19n has
synergistic effects on the activities of platinum compounds against H1299 cells. The antiproliferative
effects of satraplatin, 19n and a combination of both compounds were determined (Figure 10). The
results show that satraplatin and 19n moderately suppress proliferation of H1299 cells with respective
GI50 value of 12.0 and 13.1 µM. Interestingly, satraplatin in combination with 10 µM of 19n suppresses
(GI50 = 5.5 µM) proliferation of H1299 cells more effectively than does satraplatin alone (GI50 = 12.0
µM), while a combination of satraplatin and 10 mM 1 causes only a slight increase (GI50 = 8.5 µM) in
proliferation. We next investigated the synergistic effect of 19n and satraplatin promoted apoptosis
(Figure 10B). Treatment with 10 µM satraplatin alone results in production of 7% apoptotic cells while
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 61
1
2
3
4
5
6
7
8
9
treatment with a combination of 10 µM satraplatin and 10 µM 19n results in a 3 to 4-fold increase in
the number of apoptotic cells (30-34 %). Western blot analysis also revealed that caspase 3 cleavage
increases upon co-treatment with satraplatin and 19n (Figure 10C). Together, the observations show
that PDHK inhibition by 19n synergistically increases the antiproliferative activity of satraplatin on
H1299 cells via apoptosis. It is worthwhile noting that 6 does not synergistically effect the activity of
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
cisplatin but it causes a 2-fold increase in the activity of doxorubicin against PC-3 prostate cancer cells.
17
Figure 10. 19n increases the cytotoxicity of satraplatin. (A) Anti-proliferative activity of 19n was
determined by using the CellTiter-Glo assay. A combination of satraplatin and 19n (10 µM) promotes
an increase in the anti-proliferation effect of satraplatin alone. (B-C) A combination of satraplatin and
19n have a synergistic effect in promoting apoptosis in H1299 cells. (n = 3, One way ANOVA; **** p
<0.0001, *** p <0.0005, ** p <0.005)
Conclusions
Alteration of cellular metabolism is a cancer hallmark. Consequently, substances that affect cancer
metabolism are attractive targets for cancer chemotherapy. Among the differences between normal and
cancer cells is aberrant expressions of PDHK observed in various cancers including RCC, NSCLS, and
breast and colon cancer. Thus, intensive efforts have been made to discover efficacious PDHK
inhibitors. Recently, it was observed that 6 is a ATP competitive PDHK inhibitor and that it displays
cytostatic activity in context dependent manner.16a Also, binding affinity-based optimization led to the
development of active derivatives of 6.19 In order to uncover more potent and selective PDHK inhibitors,
ACS Paragon Plus Environment

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
we performed SAR studies on a series of resorcinol amide derivatives utilizing kinase-inhibitory and
PDC activity assays, which are different from the previous binding affinity-based SAR study19 on the
derivatives 6. We observed that introduction of substituted five-membered heteroaromatic groups on
the resorcinol amide core led to the potent PDHK inhibitors 19m-t which have structural features that
are distinct from those of 6 and its derivatives.19
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
In this effort, we showed that 19n and 19t are extremely potent and highly selective PDHK inhibitors.
We observed that 19t has excellent kinase-inhibitory activity against PDHK1, PDHK2 and PDHK4
with IC50 values in the range of 5-9 nM that are superior to those of 6. Also, 19t selectively inhibits the
four PDHK isoforms out of 366 kinases and it has a 70-fold lower activity against HSP90. It is
noteworthy that this investigation showed that the resorcinol amide derivatives 19g and 19l are isotype-
selective PDHK inhibitors. Since the expression level and distribution of each isotypes are tissue
specific, each of PDHK isoforms is associated with specific disease.29 For example, PDHK4 is abundant
in skeletal muscle in type 1 diabetes animal model30 and highly associated with diabetes. Thus, PDHK4
specific inhibitor could be potentially applied to diabetes disease. In addition, an isoform specific
inhibitor should be a useful tool in studying a function of specific isoform.
Molecular dynamics simulation studies with PDHK isoforms revealed that the differences that exist
between the activities of 19n on different PDHK isotypes appear to be associated with the degree of
hydrogen bonding interactions between N,N-dimethylbenzamide moiety and the corresponding Arg
residues (Arg250 in PDHK2 and Arg261 in PDHK4). The carbonyl oxygen of the resorcylic acid amide
moiety in 19t participated in H-bonding with Thr383 in PDHK1 and Thr346 in PDHK2, and cation-pi
interaction between N,N-dimethylbenzamide phenyl ring of 19t and Arg286 in PDHK1 and Arg261 in
PDHK4 are responsible for the excellent activity (IC50 = 5 to 9 nM) of 19t against PDHK1/2/4. Finally,
oxygens of neither of the two carbonyl groups in 19t participate in direct H-bonding interactions with
residues in PDHK3, which is consistent with its only mild activity (IC50 = 80 nM) against PDHK3.
An assessment was made of the anti-proliferative activities p53-deficient human non-small cell lung
carcinoma H1299 and prostate cancer DU145 cells of substances that have potent enzymatic inhibitory
activities. We observed that 19n, which possesses excellent potency against both PDHK2 and PDHK4,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 61
1
2
3
4
5
6
7
8
9
suppresses growth of H1299 and DU145 cells with a GI50 of 13.4 µM and 10.2 µM respectively. We
also found that the enzyme inhibitory activity of the resorcinol amide derivatives is not well-correlated
with their cellular activity. It is worth noting that 6 has an extremely low antiproliferative activity
against H1299 cells. A better understanding of the discrepancy between the enzymatic and cellular
activities of 19n was gained using a PDH assay. The results show that 19n inhibits PDC activity in the
reconstituted PDC reaction with a sub-micromolar IC50. The results of the PDH assay better correlate
with cellular activity than do the results of PDHK kinase assay, which is also true for 19g and 19k.
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
In this effort, we also demonstrated that 19n dose dependently decreases PDHE1a phosphorylation
in H1299 and DU145 cells and that it induces oxygen consumption, apoptosis via mitochondrial
dysfunction and, thus, it effectively suppresses tumorigenesis. It is worthwhile noting that the
antiproliferative activity of 19n is higher in the 3D soft agar colony formation assay than it is in a 2D
culture system. In addition, 19n treatment leads to an increase in the apoptosis-mediated
antiproliferative activity of satraplatin. Taken together, the results of this study provide insight into the
structural design of potent and selective PDHK inhibitors and the biological significance of PDHK
inhibition in cancer cells.
Experimental
Chemistry
General Information.
Unless otherwise noted, all reagents and solvents were purchased from commercial suppliers and
used without further purification. All reactions were performed under a N2 atmosphere in flame-dried
glassware. Reactions were monitored by using TLC with 0.25 mm E. Merck precoated silica gel plates
(60 F254). The progress of reactions was monitored by using TLC analysis and a UV lamp, ninhydrin
or p-anisaldehyde stain for detection. All solvents were purified by using standard techniques.
Purification of reaction products was carried out using silica gel column chromatography with Kieselgel
60 Art. 9385 (230−400 mesh). The purity of all compounds was over 95%, and and purity of all
ACS Paragon Plus Environment

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds was analyzed using LC/mass spec analysis with a Waters LCMS system (Waters 2998
Photodiode Array Detector, Waters 3100 Mass Detector, Waters SFO System Fluidics Organizer, Water
2545 Binary Gradient Module, Waters Reagent Manager, Waters 2767 Sample Manager) using
SunFireTM C18 column (4.6 × 50 mm, 5 μm particle size): solvent gradient = 60% (or 95%) A at 0
min, 1% A at 5 min. Solvent A = 0.035% TFA in H2O; Solvent B = 0.035% TFA in MeOH; flow rate
3.0 (or 2.5) mL/min. HPLC analysis was performed on an Agilent 1100 using YMC-Pack Pro C18
column (250 × 4.6 mm, 5 μm particle size): solvent gradient = 5% B at 0 min, 5% B at 3 min, 100% B
at 20 min, 100% B at 25 min. Solvent A = 0.03% TFA in H2O; Solvent B = 0.03% TFA in acetonitrile;
flow rate 0.5 mL/min. 1H and 13C NMR spectra were obtained using a Bruker 400 MHz FT-NMR (400
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
1
13
MHz for H, and 100 MHz for C) spectrometer. Standard abbreviations are used for denoting the
signal multiplicities.
Methyl 4-((4-nitrobenzyl)amino)benzoate (9). To a solution of 8 (6.0 g, 39.74 mmol) in dry MeOH
(265 mL) was added 4-nitrobenzaldehyde (4.67 mL, 47.69 mmol) and acetic acid (0.3 mL) at room
temperature. The mixture was stirred for 10 min and then treated slowly with NaBH3CN (3.75 g, 59.61
mmol). The reaction mixture was stirred for 5 h at room temperature, diluted with H2O and NaHCO3.
The aqueous layer was extracted with DCM. The combined organic layers were dried over MgSO4,
filtered and concentrated under reduced pressure. The residue was subjected to flash column
chromatography on silica gel (30% EtOAc/hexane) to afford 9 (9.21 g, 81%) as yellow solid. 1H NMR
(400 MHz, CDCl3) δ 8.19 (d, J = 8.4 Hz, 2H), 7.85 (d, J = 8.4 Hz, 2H), 7.50 (d, J = 8.4 Hz, 2H), 6.55
(d, J = 8.4 Hz, 2H), 4.53 (s, 2H), 3.84 (s, 3H); 13C NMR (100 MHz, CDCl3)  167.1, 151.0, 147.3,
146.2, 131.6, 127.6, 124.0, 119.5, 111.8, 51.6, 47.0. LRMS (ESI) m/z 287 [M + H]+.
2,4-Diacetoxybenzoic acid (10a). To a solution of 2,4-dihydroxybenzoic acid (10.0 g, 64.94
mmol) in acetic anhydride (61.4 mL, 649.35 mmol) was added H2SO4 (5 mL) dropwise at 0 oC. The
mixture was then stirred for 2 h at 60 oC, cooled to room temperature and poured into ice-water. The
mixture was stirred at room temperature for 1 h and the resulting solid was separated by filtration,
washed with water and dried using an air flow to give 10a (8.66 g, 56%) as a white solid. 1H NMR
(400 MHz, CD3OD-d4) 8.05 (d, J = 8.8 Hz, 1H), 7.10 (dd, J = 8.8, 2.0 Hz, 1H), 6.97 (d, J = 2.0 Hz,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 61
1
2
3
4
5
6
7
1H), 2.27 (s, 3H), 2.25 (s, 3H); 13C NMR (100 MHz, CD3OD-d4)  171.2, 170.2, 167.0, 156.1, 153.1,
133.9, 122.5, 120.5, 118.6, 21.1, 21.0. LRMS (ESI) m/z 239 [M + H]+.
8
9
4-Acetoxybenzoic acid (10b). To a solution of 4-hydroxybenzoic acid (10.0 g, 72.46 mmol) in acetic
anhydride (34.2 mL, 362.32 mmol) was added H2SO4 (5 mL) dropwise at room temperature. The
mixture was stirred for 2 h at 60 oC, cooled to room temperature and poured into ice-water. The mixture
was stirred at room temperature for 1 h and the resulting solid was separated by filtration, washed with
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
1
water, and dried using an air flow to give 10b (7.04 g, 54%) as a white solid. H NMR (400 MHz,
CD3OD-d4) δ 8.06-8.02 (m, 2H), 7.21-7.17 (m, 2H), 2.27 (s, 3H); 13C NMR (100 MHz, CD3OD-d4) δ
170.50, 168.93, 155.89, 132.19, 129.41, 122.82, 20.86. LRMS (ESI) m/z 181 [M + H]+.
2-Acetoxybenzoic acid (10c). To a solution of 2-hydroxybenzoic acid (10.0 g, 74.07 mmol) in acetic
anhydride (34.2 mL, 362.32 mmol) was added H2SO4 (5 mL) dropwise at room temperature. The
mixture was then stirred for 2 h at 60 oC, cooled to room temperature and poured into ice-water. The
mixture was stirred at room temperature for 1 h and the resulting solid was separated by filtration,
1
washed with water, and dried using an air flow to give 10c (8.53 g, 64%) as a white solid. H NMR
(400 MHz, CD3OD-d4) δ 8.01 (dd, J = 8.0, 1.6 Hz, 1H), 7.57 (td, J = 7.6, 1.6 Hz, 1H), 7.32 (td, J = 7.6,
1.2 Hz, 1H), 7.12 (dd, J = 8.0, 0.8 Hz, 1H), 2.27 (s, 3H); 13C NMR (100 MHz, CD3OD-d4) δ 171.36,
167.51, 152.02, 134.79, 132.72, 126.94, 124.91, 124.68, 20.91. LRMS (ESI) m/z 181 [M + H]+.
4-((4-(Methoxycarbonyl)phenyl)(4-nitrobenzyl)carbamoyl)-1,3-phenylene diacetate (11). To a
solution of 10a (3.0 g, 12.61 mmol) in DCM(42 mL) was added oxalyl chloride (5.34 mL, 63.05 mmol)
and DMF (0.1 mL). The mixture was stirred for 2 h and concentrated under reduced pressure. To a
solution of the resulting residue in dry DMF (42 mL) was added 9 (1.80 g, 6.31 mmol). After cooling
the solution to 0 oC, DIPEA (3.21 mL, 18.93 mmol) was treated dropwise and the mixture was stirred
for 2 h at room temperature. The mixture was diluted with H2O, extracted with EtOAc. The combined
organic layers were dried over MgSO4, filtered and concentrated under reduced pressure. The residue
was subjected to flash column chromatography on silica gel (30% EtOAc/hexane to 70% EtOAc/hexane)
to afford 11 (2. 94 g, 49% over 2 steps) as a white solid. 1H NMR (400 MHz, DMSO-d6)  8.19 (d, J =
8.0 Hz, 2H), 7.75 (d, J = 8.0 Hz, 2H), 7.58 (d, J = 8.0 Hz, 2H), 7.39 (d, J = 8.0 Hz, 1H), 7.31 (d, J = 8.0
ACS Paragon Plus Environment

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hz, 2H), 7.00 (s, 1H), 6.96 (d, J = 8.0 Hz, 2H), 5.29 (s, 2H), 3.76 (s, 3H), 2.28 (s, 3H), 2.21 (s, 3H); 13C
NMR (100 MHz, DMSO-d6)  168.51, 168.31, 166.11, 165.33, 151.40, 147.66, 146.75, 145.68, 144.81,
129.82, 128.80, 127.80, 127.21, 126.04, 123.60, 119.06, 116.74, 52.12, 51.46, 20.71. LRMS (ESI) m/z
507 [M + H]+.
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
4-((4-Acetamidobenzyl)(4-(methoxycarbonyl)phenyl)carbamoyl)-1,3-phenylene diacetate (13).
To a solution of 11 (1.71 g, 3.37 mmol) in MeOH (34 mL) was added Pd/C (170 mg) at room
temperature. The reaction mixture was stirred under a hydrogen atmosphere for 12 h at room
temperature, filtered through a celite pad and concentrated. To a solution of the resulting residue in dry
DCM (34 mL) was added acetic anhydride (0.45 mL, 4.72 mmol). The mixture was cooled to 0 oC, and
treated dropwise with TEA (1.41 mL, 10.11 mmol). The resulting solution was stirred at room
temperature for 2 h and diluted with H2O. The aqueous layer was extracted with DCM. The combined
organic layers were dried over MgSO4, filtered and concentrated under reduced pressure. The residue
was subjected to flash column chromatography on silica gel (70% EtOAc/hexane to 85% EtOAc/hexane)
to afford 13 (1.24 g, 71% over 2 steps) as a white solid. 1H NMR (400 MHz, CD3OD-d4) δ 7.76 (d, J =
8.8 Hz, 2H), 7.47 (d, J = 8.4 Hz, 2H), 7.22-7.18 (m, 5H), 6.92 (s, 1H), 6.81 (dd, J = 8.4, 2.0 Hz, 1H),
5.08 (s, 2H), 2.23 (s, 3H), 2.16 (s, 3H), 2.07 (s, 3H), 1.39 (s, 6H); 13C NMR (100 MHz, CD3OD-d4) δ
171.47, 170.00, 168.56, 167.41, 153.32, 149.67, 147.33, 139.38, 139.04, 133.58, 131.18, 130.72, 128.78,
127.50, 126.07, 121.01, 119.84, 117.79, 52.63, 35.30, 30.85, 23.79, 20.93, 20.78. LRMS (ESI) m/z 519
[M + H]+.
4-(N-(4-Acetamidobenzyl)-2,4-dihydroxybenzamido)benzoic acid (14). To a solution of 13 (0.41g,
0.75 mmol) in MeOH/H2O (20:1, 15 mL) was added NaOH (0.15 g, 3.77 mmol) at room temperature.
The reaction mixture was stirred at 80 oC for 2 h, cooled to room temperature, diluted sequentially with
EtOAc and 1 N HCl to adjust the pH to 2. The mixture was then extracted with EtOAc, and the extracts
were washed with H2O, dried over MgSO4, filtered and concentrated under reduced pressure. Diethyl
ether was added and the mixture was stirred at room temperature for 1 h. The resulting solid was filtered,
washed with diethyl ether, and dried using an air flow to give 14 (248.9 mg, 79%) as a white solid. 1H
NMR (400 MHz, CD3OD-d4)  7.82 (d, J = 8.4 Hz, 2H), 7.45 (d, J = 8.4 Hz, 2H), 7.26 (d, J = 8.4 Hz,
ACS Paragon Plus Environment